the effects of minimum wages on snap enrollments and expenditures

1
The Eects of Minimum Wages on SNAP Enrollments and Expenditures By Rachel West and Michael Reich March 2014  WWW.AMERICANPROGRESS.ORG A  S  S  O  C I   A T E  D P R E  S  S  /  M A T T Y  O R K 

Upload: center-for-american-progress

Post on 03-Jun-2018

215 views

Category:

Documents


0 download

TRANSCRIPT

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 143

The Effects of Minimum Wages on

SNAP Enrollments and ExpenditureBy Rachel West and Michael Reich March 2014

WWWAMERICANPROGRESSO

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 243

The Effects of Minimum Wageson SNAP Enrollments andExpendituresBy Rachel West and Michael Reich March 2014

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 343

1 Introduction and summary

5 Background

9 Methods and data

15 Results

19 Conclusion

20 About the authors and acknowledgments

21 References

23 Appendix A

25 Appendix B

33 Appendix C

38 Endnotes

Contents

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 443

Introduction and summary | wwwamericanprogress

Introduction and summary

How do minimum wage policy increases affec enrollmens and expendiures on

means-esed public assisance programs In his repor we address his quesion

or he case o he Supplemenal Nuriion Assisance Program or SNAP

ormerly known as he ood samp program

By definiion governmen spending on a means-esed program should decline

as average earnings increase insoar as benefi levels all wih increased earnings

and insoar as he earnings increase makes some individuals ineligible orany benefis Boh o hese condiions are saisfied in he case o he effec o

minimum wages on SNAP benefis SNAP benefis decline 30 cens or every $1

increase in amily earnings and phase ou enirely a abou he ederal povery

level1 Low-wage workers are disproporionaely enrolled in SNAP A minimum

wage increase ha lifs many amilies ou o povery should hereore reduce

public expendiure on his program

Bu he relaionship may be more complex I a minimum wage increase reduces

employmen hereby adding o he number o unemployed he number o SNAP

recipiens could increase SNAP recipiens who are unemployed disabled or

reired will no be affeced by a minimum wage increase Conversely i many

SNAP recipiens have earnings ha already bring hem close o becoming

ineligible or he program a minimum wage increase may have a very small effec

on SNAP expendiures Te quaniaive effec o minimum wages on SNAP

spending is no sel-eviden I requires a causal analysis

In an era o hisorically low real ederal minimum wage raes rising income

inequaliy job-marke sagnaion and conenious debae abou governmen defici

spending he possibiliy ha a higher minimum wage may lead o increased orreduced public spending has grea relevance o he public and o policymakers

Tis repor presens an iniial empirical analysis o he effecs o minimum wage

policy on SNAP paricipaion and expendiures We do so by exploiing more han

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 543

2 Center for American Progress | The Effects of M inim um Wages on SNAP Enro llments and Expenditures

wo decades o variaion in binding sae and ederal minimum wage changes in an

economeric ramework Our uure research will examine he effecs on SNAP

urher and apply an analogous ramework o wo oher public assisance spending

programs he Earned Income ax Credi and Medicaid

According o he finding in his repor a 10 percen increase in he minimum wagereduces SNAP enrollmen by beween 24 percen and 32 percen and reduces

program expendiures by an esimaed 19 percen aking ino accoun each

saersquos 2014 minimum wage level we apply hese resuls o he legislaive proposal

pu orward by Sen om Harkin (D-IA) and Rep George Miller (D-CA) o raise

he ederal minimum wage o $1010 per hour2 Our resuls imply ha he effecs

o he Harkin-Miller proposal on wage increases would reduce SNAP enrollmens

by beween 75 percen and 87 percen (31 million o 36 million persons) Te

oal anicipaed annual decrease in program expendiures is nearly $46 billion or

abou 6 percen o curren SNAP program expendiures

Harkin-Miller proposes o index minimum wage levels in subsequen years o he

consumer price index or CPI Te minimum wage would hen increase a he

same rae as SNAP benefi and eligibiliy levels which are also indexed o he CPI

Consequenly he savings over 10 years in 2014 dollars would be 10 imes he

one-year savings or a oal o approximaely $46 billion

Some o he reducion in SNAP program enrollmen and expendiures would

occur among workers making less han $1010 per hour991252hose whose pay would

be direcly increased by he minimum wage law Anoher par o he reducion

would occur among workers currenly earning beween $1010 and $1150 who

would also receive pay increases3

Alhough a large number o sudies have examined he impac o minimum

wage increases on earnings and employmen he impac o such minimum

wage policies on public assisance enrollmens and expendiures remains an

under-explored subjec in he economic lieraure Only a ew sudies discuss

he relaion beween he minimum wage and governmen ranser spending

much less atemp o ideniy he causal effec o one upon he oher Proessors

Marianne Paige Joanne Spez and Jane Millar find posiive effecs o minimum wage increases on welare caseloads as hey sae however heir resuls vary

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 643

Introduction and summary | wwwamericanprogress

considerably wih differen sample periods and assumpions abou sae rends4

Proessors Marianne Biler and Hilary Hoynes discuss he imporance o SNAP

as a saey ne program bu hey do no examine is relaion o minimum wage

policy5 Research economis Sylvia Allegreto and her Universiy o Caliornia a

Berkeley colleagues show ha low-wage workers in general and as-ood workers

in paricular are much more likely o be SNAP recipiens han all workers6

Several sudies have examined he relaionship beween he minimum wage and

he Earned Income ax Credi or EIC Proessor David Neumark and William

Wascher a researcher a he Federal Reserve Board o Governors find ha a

higher minimum wage increases EIC benefis or amilies in deep povery

while reducing EIC benefis or some sub-groups7 Proessors David Lee and

Emmanuel Saez argue ha he minimum wage and EIC are complemenary

policies no subsiues8 Te Congressional Budge Office or CBO argues ha

a minimum wage increase will no have a subsanial effec on EIC spending9

while Proessor Jesse Rohsein examines wheher he posiive effec o he EICon emale labor supply has lowered wages10 While hese sudies are o ineres

he EIC is quie differen rom SNAP in having a subsanial phase-in period in

which EIC benefis increase as well as a long phase-ou period wih complee

phase-ou a an annual income o abou $48000 or a amily o our quie a bi

above he reach o he minimum wage11

Research by Proessor Arindraji Dube on he causal effec o he minimum wage

on amily povery represens he sudy mos relaed o he one a hand12 Dube finds

ha Harkin-Miller would raise abou 46 million non-elderly Americans above

he ederal povery level or FPL In conras when CBO uses a simple simulaion

mehod o address he same quesion hey find ha Harkin-Miller would raise

900000 people above FPL13 Te difference beween hese wo esimaes highlighs

he imporance o underaking a causal analysis Te mehods used in his paper are

in many respecs similar o Dubersquos Moreover since eligibiliy and benefi levels or

programs such as SNAP and Medicaid are ied o he ederal povery level Dubersquos

findings have direc implicaions or his sudy Noneheless his repor appears o

be he firs sudy o examine he effecs o he minimum wage on SNAP In uure

work we plan o underake similar analyses or he EIC and Medicaid

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 743

4 Center for American Progress | The Effects of M inim um Wages on SNAP Enro llments and Expenditures

Te repor proceeds as ollows

bull Secion 1 provides background inormaion on he ederal minimum wage sae

minimum wages and he SNAP program

bullSecion 2 describes our mehods and daa

bull Secion 3 provides our main resuls including a simulaion o he effecs o a

Harkin-Miller minimum wage increase and a sae-by-sae analysis

bull Secion 4 presens our conclusions

Furher deails are provided in a series o appendices

Correction April 28 2014 Tis report incorrectly stated the potential reduction in

SNAP enrollment fom the Harkin-Miller proposal Te correct amount is 75 percentand 87 percent (31 million to 36 million persons) as stated in the reportrsquos tables

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 843

Background | wwwamericanprogress

Background

The federal minimum wage

Te ederal minimum wage was las increased in July o 2009

nearly five years ago During he pas wo decades many saes

have passed legislaion fixing he minimum wage a a higher

level han he ederal minimum Te maps in Figure 1 show ha

while saes in every region o he Unied Saes have adoped

higher minimum wages hey are no disribued randomly bygeography As shown in he 2013 sudy ldquoCredible Research

Designs or Minimum Wage Sudiesrdquo by economiss Sylvia

Allegreto Arindraji Dube Michael Reich and Ben Zipperer

hese saes vary sysemaically rom he oher saes by a number

o characerisics ha affec low-wage employmen rends bu

which are no hemselves relaed o minimum wage policy14

Te nonrandom patern o minimum wage adopion has

imporan implicaions or obaining unbiased esimaes o

minimum wages on employmen In paricular naional panel

sudies ha use sae and ime fixed effec models991252such as

a 1992 sudy by David Neumark and William Wascher991252

spuriously esimae negaive employmen effecs Te reason

or his resul is uncovered using ess or pre-rends Tese

ess find ha low-wage employmen was already declining wo

years beore minimum wages were implemened By making a

saisically large number o local comparisons ha conrol or

heerogeneiy among saes and by ime eliminaes his pre-

rend For his reason we conduc similar ess or our SNAPoucomes and use model specificaions ha include local

comparisons as in he sudy cied above

FIGURE 1

High versus low minimum wage stat

from 1990 to 2012

Means and variances

More than $533

Average minimum wage over 1990ndash2012

Less than or equal to $533

More than $121

Minimum wage variance over 1990ndash2012

Less than or equal to $121

Notes State means and variances were calculated using annual stat

minimum wage data from 1990 to 2012 The shading on the maps

partitions the states into above- and below-median values

Source Sylvia Allegretto and others ldquoCredible Research Designs forMinimum Wage Studiesrdquo Working Paper 148-13 (Institute for Resear

on Labor and Employment 2013) available at httpwwwirleberkeleyeduworkingpapers148-13pdf

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 943

6 Center for American Progress | The Effects of M inim um Wages on SNAP Enro llments and Expenditures

Te CBO recenly projeced ha in 2016 17 million workers will earn less han

he $1010 hourly wage proposed in he Harkin-Miller bill Furhermore he

CBO esimaes ha an addiional 8 million workers earned beween $1010 and

$1150 per hour and were also likely o experience a wage increase15

Supplemental Nutrition Assistance Program

We ocus our inquiry on SNAP Benefis under he program are enirely

ederally unded he program is adminisered by he US Deparmen o

Agriculure ogeher wih sae agencies which share in adminisraion

coss Spending on SNAP has grown in he pas decade reaching $78 bil lion

in 2011 wih SNAP enrollmen increasing o 45 million people abou one-

sevenh o he US populaion16 According o he CBO changes since 1990 in

SNAP spending and enrollmens are primarily he resul o cyclical economic

condiions noably changes in he unemploymen rae and changes in per capiaincome17 Te 2009 American Recovery and Reinvesmen Ac emporarily

increased SNAP benefi amouns by 136 percen as repored by he US

Deparmen o Agriculurersquos Food and Nuriion Service hese higher benefi

levels expired on November 1 201318 Te CBO esimaes ha abou wo-

hirds o he changes in SNAP expendiure are associaed wih changes in he

number o recipiens and one-hird wih changes in he benefis received when

recipiensrsquo incomes change19

In fiscal year 2014 SNAPrsquos maximum monhly benefis are $189 or a single

individual $497 or a amily o hree and $750 or a amily o five Benefis are

reduced by 30 cens per dollar received and phase ou enirely a gross monhly

household incomes o 130 percen o he ederal povery level $1245 or a

single individual $2116 or a amily o hree and $2987 or a amily o five

o deermine benefis SNAP also defines a ne monhly income concep and

ses benefis a 100 percen o he ederal povery level using his concep

Calculaion o ne monhly income can include cerain deducions rom

monhly gross income such as medical expenses and child care coss Alhough

saes are permited some laiude on wha deducions are allowed in pracice

hese vary by very small amouns Our saisical model akes accoun o sae-specific differences in benefis20

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1043

Background | wwwamericanprogress

SNAP imposes an employmen or work-raining requiremen or able-bodied

prime-age aduls991252hose beween he ages o 18 and 50 and wihou disabiliies

or dependen children Such households can receive only hree monhs o benefis

in a hree-year period In recen years abou 85 percen o households receiving

benefis have incomes below he ederal povery level 49 percen have dependen

children 16 percen are age 60 or older 20 percen are disabled and 30 percenrepor some earned income21

A 2012 CBO repor also noes ha ake-up raes among eligible SNAP recipiens

average abou 70 percen wih much lower ake-up among elderly households

Te ake-up rae increases in harder economic imes I also increased when sigma

issues were reduced as SNAP debi cards replaced acual ood samps ake-up

is especially high among hose mos needy Adminisraive spending equaled 91

percen o he poenial spending ha would have occurred i all eligible recipiens

were enrolled Alhough some SNAP policy changes have occurred since 1990

mos were relaively minor and all were naional in scope Te 1996 welare reorm bill eliminaed SNAP eligibiliy or some legal immigrans limied he ime lengh

o eligibiliy or able-bodied childless aduls and reduced maximum benefis

Some o hese resricions were relaxed in 2002 and again in he American

Recovery and Reinvesmen Ac in 200922

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1143

8 Center for American Progress | The Effects of M inim um Wages on SNAP Enro llments and Expenditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1243

Methods and data | wwwamericanprogress

Methods and data

As previously menioned we exploi variaion in minimum wages by sae and ime

o examine heir causal effecs on SNAP enrollmens and expendiures o do so we

merge daa rom 1990 hrough 2012 drawn rom he Annual Social and Economic

Supplemen o he Curren Populaion Survey991252an annual Census Bureau survey

commonly known as he March CPS ha includes23 inormaion on SNAP

enrollmens a he amily level991252wih sae-level daa on minimum wages SNAP

expendiures populaion unemploymen raes and sae median income levels

o conrol or ime-varying heerogeneiy among saes our specificaions includeconrols or sae linear rends and effecs by Census division and ime We esimae

effecs a wo levels allowing or amily variaion and allowing only or sae-level

variaion We also employ a se o sandard demographic conrols such as amily size

and composiion and race and ehnic composiion

Distinguishing causation from correlation

How can we ensure ha our analysis does no pick up a spurious correlaion or

example he endency o more economically vibran saes o implemen higher

minimum wages Disinguishing correlaion and policy endogeneiy rom rue

causal effecs is he primary moivaion or economeric analysis In he ideal

experimen researchers would begin wih wo saes991252ha are alike in every

respec prior o he policy991252 and ldquoreardquo only one o hese saes wih a higher

minimum wage Tey would atemp o shield hese saes rom any influence ha

could obscure heir undersanding o he minimum wagersquos direc effec on SNAP

enrollmen Researchers o course canno conduc such experimens

We can however use saisical mehods o conrol simulaneously heindependen effecs on SNAP o sae unemploymen raes sae income levels

and common rajecories among saes wihin he same Census division By

ensuring similariy along hese dimensions we maximize he likelihood ha

SNAP aciviy in wo saes would have comparable oucomes in he absence

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1343

10 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

o a minimum wage policy change Tus i a new minimum wage policy were

implemened in one sae only he researchers could atribue all o he difference

hey observe in SNAP aciviy o he new minimum wage policy

In oher words we approximae he ideal experimen by using non-experimenal

saisical mehods Te desirable ldquopre-exising similariiesrdquo beween saes ha wehave defined above inorm our choice o conrol variables in a saisical seting

More precisely in our muliple regression models we use median amily income

he unemploymen rae he employmen-o-populaion raio and regional and

ime idenifiers o consruc an appropriae group o peers or each sae on he

eve o a policy change

Data description

wo daa ses include inormaion abou boh income and paricipaion inpublic programs Te Survey o Income and Program Paricipaion or SIPP

which is conduced in inermiten years has he advanage o ollowing he

same individuals over a period o ime In oher words i is a longiudinal daa

se I also has he advanage o conaining monhly daa However he sample

size o he SIPP is no sufficien or analyzing variaions in sae-level minimum

wages Te March CPS has he advanage o a much larger sample size and i is

conduced annually wihou any breaks in ime I has he disadvanage o being

a cross-secional daa se so we canno ollow he same individuals over ime991252

sricly speaking over more han one year On ne he March CPS is much more

suiable or our sudy We examine he empirical relaionship beween minimum

wage policy and ood samp aciviy a wo levels o aggregaion he amily level

and he sae level Family-level daa are drawn rom he March CPS

Te March CPS comprises responses rom he residens o 50000 o 60000

dwelling places surveyed per year and conains deailed inormaion on he

residensrsquo employmen and income including income rom ranser paymens

Te sample or our analysis comprises more han 128 million amily unis during

he period rom 1990 o 2012 (inclusive) Survey weighs allow us o analyze

SNAP paricipaion in a manner ha is represenaive o he US populaion alarge Over all years he share o amilies reporing ood samp receip in he

weighed March CPS sample is 91 percen Te enrollmen rae was a a low o 6

percen in he year 2000 In 2012 he mos recen year in our panel 133 percen

o amilies repored paricipaing in SNAP a some poin during he survey year

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1443

Methods and data | wwwamericanprogresso

Te March CPS also collecs inormaion on he number o SNAP recipiens in

he household he number o monhs he household receives SNAP benefis

and he cash-equivalen value o he SNAP benefis received However he

value o SNAP benefis is severely underrepored among recipiens perhaps

because recipiens are unaware o he exac monhly cash-equivalen value o

benefis hey receive

Our firs empirical sraegy ocuses solely on SNAP enrollmen By using he

amily as he uni o analysis we are able o inser saisical conrols o accoun

or non-wage-relaed acors ha influence any paricular amilyrsquos likelihood o

program paricipaion wih he inenion o isolaing any differences in program

paricipaion ha are due purely o changes in wage policy Tis approach

idenifies he effecs o low-wage labor policy on he exernal margin991252ha is

he effec o he minimum wage on he likelihood ha a amily paricipaes in he

SNAP program a all991252as opposed o he inernal margin or how much SNAP

unding he amily would receive

Our second empirical ramework uses sae-level adminisraive daa Ta is

we aggregae he daa o obain a single daa poin or each saeyear back

o 1990 represening he mean o he oucome or he sae Te sae-level

esimaion serves as a robusness check on he amily-level resuls or SNAP

paricipaion Also using aggregaed daa allows us o esimae direcly he

causal effec o minimum wage changes on SNAP spending Tis is no possible

a he amily level as discussed above daa on cash-equivalen value o ood

samps or SNAP recipiens is very requenly no repored in he March CPS

and when i is repored he inormaion may be unreliable By conras he

Bureau o Economic Analysis publishes aggregae SNAP spending a he sae

level in is Naional Income and Produc Accoun or NIPA ables Tus

while we are unable o observe he heerogeneiy in he cash value o SNAP

or amilies in each sae we are able o calculae average SNAP spending

per residen in each sae per year Supporing covariaes include he annual

unemploymen and employmen daa rom he Bureau o Labor Saisics or

BLS and sae-level populaion series rom he iner-decennial census releases

Minimum wage daa are available rom he BLSrsquos wages and hours division For

sae minimum wage changes enaced a oher imes han he firs o he yearan average value or he year is used

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1543

12 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

Family level model specifications

We firs examine he effec o he minimum wage on paricipaion in public

assisance programs For amily residing in sae and during year we esimae

an equaion o he ollowing orm

(1)

is a binary variable ha is se equal o 1 i a leas one member o amily

received ood samps during he survey year is a se o sae-level

characerisics including annual averages o he unemploymen rae he

employmen-o-populaion raio and he naural log o median amily income

is a vecor o amily atribues including indicaors or he race and marial saus

o he amily head size o he amily he presence o children and he presence oan adul male Sae fixed effecs are capured by o conrol or ime-varying

heerogeneiy our preerred model specificaion also includes year fixed effecs

ha vary by Census division ( ) and sae-level linear ime rends In

Appendix B we jusiy he inclusion o hese las wo erms We also compare he

resuls rom our preerred specificaion wih less sauraed specificaions

Te effec o ineres which is capured by is he expeced change in he

probabiliy o receiving SNAP benefis wih respec o a change in he (log o he)

binding minimum wage in sae during year We repor robus sandard errors

clusered a he sae level We esimae he parameers using linear regression

producing a linear probabiliy model Deails o he model selecion process are

covered in Appendix B below

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1643

Methods and data | wwwamericanprogresso

State-level model specifications

Te sae-level models are similarly specified For sae in year we assume ha

(2)

In his model is now eiher he SNAP enrollmen rae in sae during year

or he naural logarihm o per capia SNAP expendiures in sae during

year is once again a se o sae-level characerisics including he same

sae-level covariaes as in he amily regressions (annual average unemploymen

rae employmen-o-populaion raio naural log o median amily income)

wih he addiion o amily level characerisics averaged across he sae (average

amily size and he shares o populaion consiued by each o five racialehnicgroups) Sae fixed effecs are represened by As above our preerred model

specificaion includes year fixed effecs ha vary by Census division ( ) and

sae-level linear ime rends as elaboraed in Appendix B Te effec o

ineres is capured by

We esimae boh sae-level models (enrollmen and expendiures) using

ordinary leas squares regression Tus he inerpreaion o he coefficien is no

longer ha o a change in probabiliy as in he binary oucome models described

above Raher or he sae-level SNAP enrollmen model represens he

expeced change (in percenage poins) in he saersquos SNAP enrollmen rae ha

is due o a 1 percen change in he minimum wage For he SNAP expendiures

model is simply he elasiciy o SNAP spending wih respec o he minimum

wage991252ha is he percenage change in sae expendiures expeced o resul rom

a 1 percen change in ha saersquos minimum wage For urher deails on model

specificaion reer o Appendix B below

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1743

14 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1843

Results | wwwamericanprogresso

Results

Estimated minimum wage effects on SNAP enrollment and

expenditures

able 1 shows he esimaed parameer o ineres991252he coefficien o he

minimum wage or he preerred model o each ype Coefficiens on he

minimum wage variable are no direcly comparable across models because all

our models have a differen uncional orm o undersand and compare hese

esimaes we compue he change in SNAP aciviy prediced or a paricular wage scenario Te final column in able 1 answers he quesion Wha would be

he expeced change in SNAP aciviy in response o a 10 percen increase in he

minimum wage Te answer o his quesion varies wih he value o he inpu

parameers in he able we calculae he percenage decrease in enrollmen or

expendiures prediced or he average sae wih a minimum wage o $725 in

2014 Te sae-level SNAP expendiure model which is a consan-elasiciy

model conveys elasiciy inormaion direcly or he change in expendiures per

capia in he sae

TABLE 1

Comparison of national SNAP predictions for a 10 percent increase in the federal minimum wage

Model LevelRegression

type

Predicted outcome Coefficient of log

(minimum wage)

(Standard error)

Effect of a 10 percent in

in the minimum wage

VariableForm of

variable

Total

enrollment

Tot

expend

1 Family Linear

probability Enrollment Binary (enrolled=1)

-0042

(0008) -317 N

2

StateLinear regression

(ordinary least

squares)

Enrollment State enrollment rate (percent) -0031

(0012) -235 N

3 Expenditures Log (state expenditures per capita) -0190

(0103) NA -19

plt01 plt005 plt001Note Predicted changes are calculated for the average state with a minimum wage of $725 in 2014

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1943

16 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

According o his model a 10 percen increase in he minimum wage would resul

in a 19 percen reducion in SNAP expendiures Tis resul is reassuringly similar

o Dubersquos resul or povery reducion Te wo enrollmen models are much more

precisely esimaed han he expendiure model Te sae-level enrollmen model

finds ha a 10 percen minimum wage increase in a low-wage sae is associaed wih

a 235 percen decrease in SNAP enrollmens Te amily-level linear probabiliymodel predics a somewha greaer elasiciy or low minimum wage saes an

increase o 10 percen in he ederal minimum would resul in a 317 percen decline

in SNAP enrollmen24 Te differences in hese esimaes sem rom a number o

acors including difference in model uncional orm and daa used We rea his

range o elasiciy esimaes as an upper and lower bound on enrollmen impacs

Harkin-Miller bill National and state-level predicted impacts

Wha would be he prediced change or he SNAP program i he ederalminimum were raised o $1010 as proposed in he Harkin-Miller bill In order o

make his inerence we accoun or he ac ha no all saes are currenly subjec

o he ederal minimum wage a he beginning o 2014 21 saes mainained

higher minimum wages han $725 In hose saes an increase in he ederal

minimum wage may or may no be binding or employers in he sae depending

upon wheher he new ederal minimum exceeds he sae-level minimum Bu

regardless o wheher a minimum wage change is binding he impac on SNAP

aciviy will be lower in high minimum wage saes In order o accoun or his

properly we calculae sae by sae he percenage wage change ha would resul

rom he Harkin-Miller proposal and apply he parameers rom each o he hree

models above o compue he expeced decrease in SNAP aciviy or each sae

In his exercise we use saesrsquo curren (2014) minimum wage levels and assume

as a baseline he 2012 levels o SNAP enrollmen and expendiure as 2012 is he

mos recen year or which SNAP daa are available

able 10 and able 11 in Appendix C repor he esimaed effecs on SNAP

enrollmen and expendiures respecively or each sae under he Harkin-Miller

bill25 An increase o $1010 i enaced oday would represen beween a 393

percen wage increase in a $725 minimum wage sae and an 84 percen increasein Washingon sae which has he highes minimum wage in he naion a $932

as o January 201426 Slighly more han 56 percen o he decrease in expendiures

and abou 59 percen o he decrease in enrollmen would occur in saes wih

presen-day minimum wages o $725

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2043

Results | wwwamericanprogresso

In 2012 hese saes were home o 46 percen o he American populaion

and accouned or a roughly equivalen percenage o oal naional SNAP

expendiures No surprisingly he larges enrollmen decreases come rom

populous saes wih very high SNAP enrollmen raes andor very low minimum

wages Te larges prediced enrollmen reducion991252beween 319000 individuals

and 362000 individuals991252would occur in exas which has a $725 minimum wage In Caliornia which has a 206 percen SNAP paricipaion rae and an $8

per hour minimum wage we anicipae SNAP enrollmen declines o 310000

persons o 371000 persons And in Florida which had a SNAP paricipaion rae

o 166 percen and a $793 minimum wage enrollmen could decline by beween

164000 individuals and 196000 individuals For he our saes ha ogeher

accouned or he greaes amoun o SNAP spending in 2012991252exas Caliornia

Florida and New York respecively991252he combined expendiure reducion rom

he Harkin-Miller bill is prediced o be $14 billion

able 2 summarizes he prediced declines in SNAP aciviy or he naion as a whole ha would resul rom he direc and indirec effecs o he Harkin-Miller

bill Enrollmen would all beween 31 million persons and 36 million persons

represening 75 percen o 87 percen o curren enrollmen Te anicipaed

reducion in program expendiures would be nearly $46 billion or 61 percen o

program expendiures

TABLE 2

Comparison of national SNAP predictions under the Harkin-Miller billrsquos $1010 minimum wage

Model

Enrollment

(persons)

Expenditures

(millions of dollars)

Current (2012) Predicted Change Current (2012) Predicted Chan

Family enrollment (linear probability)

41866195

45489339 -3623144

$74861

NA NA

State enrollment (ordinary least squares) 38745435 -3120759 NA NA

State expenditures (ordinary least squares) NA NA $70305 -$45

Note Calculations use 2014 state minimum wages and the most recent SNAP data from 2012 They assume that per-enrollee expenditures remain constant

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2143

18 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

Tere are o course oher possibiliies or minimum wage legislaion able 3

shows he expeced SNAP changes or he Unied Saes under a variey o wage

scenarios calculaed using he sae-level models I saes were no able o se

heir minimum wages independenly such ha all saes were consrained by

he ederal minimum o $725 SNAP would be received by abou 514000 more

people across he Unied Saes a an addiional program cos o nearly hree-quarers o a billion dollars In conras he effecs o a higher minimum wage

proposal991252a ederal wage floor o $11 per hour991252would decrease enrollmen in

SNAP by more han 10 percen and decrease program coss by 83 percen

TABLE 3

Summary of par ticipation and expenditures under wage scenarios

If all states had

minimum wages of

Enrollment(persons)

Expenditures

(millions of dollars)

Predicted Change Predicted Change

Recent levels (2014) 41866195 $74861

$725 42380520 514326 $75604 $743

$800 41423919 -442276 $74209 -$652

$900 40148451 -1717744 $72350 -$2511

$1000 38872982 -2993212 $70490 -$4371

$1010 38745435 -3120759 $70305 -$4556

$1100 37597514 -4268681 $68631 -$6230

Note Calculations use state-level enrollment model coefficient

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2243

Conclusion | wwwamericanprogresso

Conclusion

An exensive body o lieraure examines employmen effecs o he minimum

wage A much smaller se o sudies examines how he minimum wage affecs

povery and only a handul o papers examine he effecs o he minimum wage

on he EIC Our analysis is he firs o examine he effecs o he minimum

wage on SNAP

Our findings indicae ha increased earnings rom minimum wage changes do

reduce SNAP enrollmens and expendiures We esimae ha he Harkin-Miller bill would save axpayers nearly $46 billion per year equivalen o 61 percen

o SNAP expendiures in 2012 he las year or which daa are available Over a

10-year period he esimaed savings amoun o nearly $46 billion

Our repor is subjec o limiaions ha we expec o overcome in our uure

research Firs he findings do no ake ino accoun possible ineracions among

SNAP he EIC and Medicaid Te eligibiliy cuoffs among hese programs

are quie differen suggesing ha such ineracions may be minor Noneheless

he join effecs can only be deermined by urher research using a causal

model Second i would be useul o know he disribuion o SNAP reducions

along he wage disribuion Using he Congressional Budge Officersquos calculaions

o how much he oal dollar value o wage would increase under he Harkin-

Miller proposal our findings imply ha he decline in overall SNAP spending

equals abou 15 percen o he oal resuling increase in wages Te amoun and

disribuion o his offse are o considerable ineres Minimum wage beneficiaries

who come rom working amilies already well above he povery line would no

see any offse while hose who are currenly considerably below he povery line

will see larger offses Tese issues will also be a subjec or our uure research

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2343

20 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

About the authors

Rachel West is a maser o public policy candidae a he Goldman School

o Public Policy Universiy o Caliornia Berkeley Her research ocuses on

economic policy in he areas o low-wage labor and povery

Michael Reich is proessor o economics and direcor o he Insiue or

Research on Labor and Employmen a he Universiy o Caliornia a Berkeley

His research publicaions cover numerous areas o labor economics including

racial inequaliy labor marke segmenaion high-perormance workplaces

union-managemen cooperaion Japanese labor-managemen sysems living

wages and minimum wages He received his docorae in economics rom

Harvard Universiy

Acknowledgments

We are graeul o Sylvia Allegreto Arindraji Dube Bill Leser Jesse Rohsein

Daniel Tompson and Ben Zipperer or heir valuable suggesions

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2443

References | wwwamericanprogresso

References

Allegreto Sylvia and ohers 2013 ldquoFas Food Povery Wages Te Public Cos o Low-Wage Jobsin he Fas-Food Indusryrdquo Universiy o Caliornia Berkeley Cener or Labor Research andEducaion Available a htplaborcenerberkeleyedupubliccossas_ood_povery_wage

Allegreto Sylvia and ohers 2013 ldquoCredible Research Designs or Minimum Wage Sudiesrdquo

Working Paper 148-13 Universiy o Caliornia Berkeley Insiue or Research on Labor andEmploymen Available a htpwwwirleberkeleyeduworkingpapers148-13pd

Biler Marianne and Hilary Hoynes 2013 ldquo Te More Tings Change he More Tey Say heSame Te Saey Ne Living Arrangemens and Povery in he Grea Recessionrdquo Working Paper19449 Naional Bureau o Economic Research

Congressional Budge Office 2012 ldquoTe Supplemenal Nuriion Assisance Programrdquo Washingon Available a htpwwwcbogovsiesdeaulfilescbofilesatachmens04-19-SNAPpd

991252 991252 991252 2014 ldquoTe Effec o a Minimum-Wage Increase on Employmen and Family Incomerdquo Washingon Available a htpwwwcbogovsiesdeaulfilescbofilesatachmens44995-MinimumWagepd

Dube Arindraji 2013 rdquoMinimum Wages and he Disribuion o Family Incomerdquo Unpublished working paper Available a htpsdldropboxuserconencomu15038936Dube_ MinimumWagesFamilyIncomespd

991252 991252 991252 2014 ldquoTe Povery o Minimum Wage Facsrdquo Arindraji Dube Blog January 22 Availablea htparindubecom20140122he-povery-o-minimum-wage-acs

Lee David and Emmanuel Saez 2012 ldquoOpimal Minimum Wage Policy in Compeiive LaborMarkesrdquo Journal o Public Economics 96 (9) 739ndash749

Neumark David and William Wascher 1992 ldquoEmploymen Effecs o Minimum and Subminimum Wages Panel Daa on Sae Minimum Wage Lawsrdquo Industrial and Labor Relations Review 46 (1)

55ndash81

Neumark David and William Wascher 2011 ldquoDoes a Higher Minimum Wage Enhance heEffeciveness o he Earned Income ax Credirdquo Industrial and Labor Relations Review 64 (4)712ndash746

Page Marianne Joanne Spez and Jane Millar 2005 ldquoDoes he Minimum Wage Affec WelareCaseloadsrdquo Journal o Policy Analysis and Management 24 (2) 273ndash295

Rohsein Jesse 2010 ldquoIs he EIC as Good as an NI Condiional Cash ransers and ax

Incidencerdquo American Economic Journal Economic Policy 2 (1) 177ndash208

Wage and Hours Division 2014 ldquoMinimum Wage Laws in he Saes ndash January 1 2014rdquo USDeparmen o Labor (htpwwwdolgovwhdminwageamericahm [February 2014])

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2543

22 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2643

Appendix A | wwwamericanprogresso

Appendix A

Pre-trend falsification check

Recen minimum wage research27 highlighs a common flaw in previous sudies

ailure o veriy ha he oucome variable is ree o negaive pre-exising rends I

or example SNAP aciviy was already rending down in saes ha raised heir

minimum wages beore hese changes came ino effec our regression analysis could

(misakenly) atribue ha reducion o he minimum wage We check or such

pre-rends by inroducing variables ha represen he prior yearrsquos value or leads o

he minimum wage I he model esimaes he minimum wage o have an effec on

he oucome variable beore he wage change wen ino effec hen an unobservedacor no he minimum wage change caused he change in SNAP aciviy

We es he specificaions above or pre-rends by including a one-year lead in

all hree specificaions We find ha he lead erms are small posiive and no

saisically significan indicaing ha he concurren minimum wage991252no

he wage level in prior periods991252is driving he observed changes in SNAP

oucomes28 In paricular he coefficien (sandard error) on he lead erm in

our preerred amily-level enrollmen regression is 011 and no significan

while he coefficien and sandard error o he conemporaneous minimum

wage is unchanged In he sae-level preerred enrollmen regression he

coefficien o he lead erm is again small (07) and i is no significan Te

corresponding coefficien on he lead erm in he sae-level expendiure

regression is 16 and is no significan Te posiive poin esimaes on hese lead

erms resuls no only rule ou disoring negaive pre-rends Tey also sugges

ha our main resuls may underesimae he rue effecs

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2743

24 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2843

Appendix B | wwwamericanprogresso

Appendix B

Model selection process

For boh he amily-level and sae-level models we es hree mehods o conrol

or unobserved geographic- and ime-varying characerisics as suggesed by he

minimum wage lieraure Firs we include only independen sae-specific fixed

effecs and year-specific fixed effecs Tis specificaion (specificaion 1) implicily

assumes ha amilies in any sae consiue an equally good saisical ldquoconrolrdquo

group or hose in any randomly chosen sae afer accouning or various

characerisics (median income and unemploymen rae among ohers) Similarly

simple ime fixed effecs assume ha amilies surveyed in any year can crediblyserve as a conrol group or amilies surveyed in every oher year o he sample

(1990 hrough 2012)

In oher words specificaion 1 assumes ha a saersquos immediae neighbor provides

no beter a couneracual or he effec o a minimum wage change han does a

sae across he counry We relax his resricive specificaion sequenially in wo

seps In specificaion 2 we replace simple year fixed effecs wih fixed effecs or

each Census divisionyear (capured as an addiional variable in he vecor By

using division-year effecs we remove he resricion ha amilies in each sae

are equally good saisical conrols or all oher amilies Raher we allow or he

possibiliy ha amilies in similar geographic regions (or example he Souh or

he Norheas) may be more similar o one anoher han amilies arher away

Finally in specificaion 3 we add sae-specific linear ime rends o he previous

specificaion Tus specificaion 3 is he mos rigorous model specificaion in ha

i allows or heerogeneiy along hree dimensions Ta is specificaion 3 allows

each sae o have is own ime-varying rends raher han imposing he resricion

ha saes evolve idenically over he 22 years in he sample

We begin building he heoreical specificaion above rom a se o simpleuncondiional models regression o SNAP aciviy (enrollmen or expendiures)

on he log o he minimum wage and a se o geographic- and ime-specific

effecs (specificaions 1 2 and 3 described above) As shown in ables 1ndash3 (or

specificaion 3) we hen add covariaes sequenially o hese models including

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2943

26 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

firs he vecor o amily-level conrols ollowed by each o several sae-level

covariaes in urn he unemploymen rae log o median-amily income and he

employmen-o-populaion raio Comparable resuls or specificaions 2 and 3

will be available in our orhcoming working paper

As expeced he simple uncondiional models indicae ha he relaionship beween he minimum wage and SNAP enrollmen i one exiss is a more complex

one influenced by oher acors In he uncondiional model he coefficien on he

variable o ineres991252he log o he minimum wage991252is small in magniude and no

saisically differen rom zero Once we accoun or he influence o labor marke

condiions and variaion in income levels on program paricipaion (by including

unemploymen rae and median-amily income conrol variables respecively)

he effec o he minimum wage on SNAP enrollmen is precisely esimaed Te

coefficien o he log minimum wage is slighly higher (-0042) in he amily-level

analysis han he coefficien (-031) in he sae-level analysis Te level o precision

is also higher in he amily-level analysis Tis is o be expeced when using 124million observaions compared o 1127

Te second dimension o model choice concerns he effec specificaion ables

7ndash9 compare he primary coefficiens o ineres or he SNAP enrollmen and

expendiure models For boh he enrollmen models he effec sizes are smalles

or specificaion 1 larges or specificaion 2 and inermediae beween hese wo

in specificaion 3 Recall ha Specificaion 3 conains sae-specific linear ime

rends in addiion o he census divisionyear conrols included in specificaion

2 In he amily-level enrollmen model he sandard error o he minimum wage

coefficien is smaller han in he oher wo specificaions Sandard errors on he

oher variables are much smaller in specificaions 2 and 3 han in specificaion

1 On he basis o coefficien significance (join and individual) specificaions 2

and 3 are sricly preerred in boh enrollmen models o specificaion 1 which

conains only sae and year fixed effecs

A concern wih specificaions 2 and 3 is ha rend conrols such as sae linear

rends may incorrecly absorb some o he delayed impac o a minimum wage

When we es his issue by including lagged minimum wages we do no find ha

delayed effecs are significan Anoher concern is ha more sauraed modelsuse less o he saisical variaion which could reduce he saisical power o

he resuls However he sandard errors or our more sauraed models are no

higher and are lower in some cases han or he less sauraed models Overall

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3043

Appendix B | wwwamericanprogresso

his evidence suppors our use o he sauraed model as he preerred model

specificaion Moreover Dubersquos 2013 sudy shows ha more sauraed models

perorm beter han models wih jus sae and ime fixed effecs

Te esimaed enrollmen regressions a boh he amily and sae levels show large

and saisically significan coefficiens Te esimaed minimum wage effec in heexpendiures regressions991252or which we have only sae-level daa991252is also large

and saisically significan

We do no use weighed regression or he sae-level models preerring o keep

analysis o he ldquoreamenrdquo (ha is o say a minimum wage change) appropriae

o he average sae raher han he average amily or individual I insead our

primary ineres were he impac o a minimum wage change on he average amily

or he average individual we migh choose o designae he number o amilies

in each sae or he sae populaion respecively as analyic weighs in order o

obain a coefficien beter suied or such inerence

TABLE 4

SNAP enrollment

Family-level linear probability

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0048 -0047 -0040 -0043 -0042

(0013) (0013) (001) (0008) (0009)

Unemployment rate (100) 0505 0420 0280

(0083) (0086) (0082)

Log median income -0057 -0039

(0011) (0011)

Employment-to-population ratio -0239

(0038)

N 1242022 1242022 1242022 1242022 1242022

plt01 plt005 plt001

Note Robust standard errors are in parentheses Observations are clustered at the state level The outcome variable is binary and equal to one if a family is enrolledin SNAP All models include state fixed effects Census division x-year fixed effects and state-specific linear time trends All specifications except 3a include additional

controls for family size race and marital status of the family head presence of children and presence of an adult maleSource Annual data from the March Current Population Survey for 1990 to 2012 estimation includes Current Population Survey probability weights

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3143

28 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

TABLE 5

SNAP enrollment

State-level linear regression

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0028 -0024 -0026 -0031 -0031

(0014) (0013) (0013) (0012) (0012)

Unemployment rate (100) 0625 0466 0320 0339

(0087) (0088) (0085) (0083)

Log median income -0090 -0065 -0061

(0013) (0013) (0013)

Employment-to-population ratio -0282 -0248

(0037) (0038)

N 1127 1127 1127 1127 1127

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the SNAP enrollment rate All models include state fixed effects Census division x-year fixedeffects and state-specific linear time trends

Source Annual data from the March Current Population Survey for 1990 to 2012

TABLE 6

SNAP expenditures

State-level linear regression

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0181 -0149 -0156 -0153 -0190

(011) (0103) (0102) (0103) (0103)

Unemployment rate (100) 4587 4152 4232 4313

(0622) (0621) (0633) (0628)

Log median income -0246 -0261 -0294

(0075) (0078) (0078)

Employment-to-population ratio 0155 0244

(0237) (024)

N 1127 1127 1127 1127 1127

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the natural log of state SNAP expenditures per capita for 1990 to 2012 All models include state

fixed effects Census division x-year fixed effects and state-specific linear time trends

Source Annual data from Bureau of Economic Analysis national income and product accounts tables for 1990 to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3243

Appendix B | wwwamericanprogresso

TABLE 7

Comparison of specifications SNAP enrollment

Family-level linear probability

(1) (2) (3)

Log minimum wage -0028 -0049 -0042

(0014) (0017) (0009)

Unemployment rate (100) 0275 0297 0280

(0161) (0076) (0082)

Log median income -0077 -0055 -0039

(0014) (0012) (0011)

Employment-to-population ratio -0238 -0250 -0239

(0054) (004) (0038)

N 1242022 1242022 1242022

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses Observations are clustered at the state level The outcomevariable is binary or equal to one if a family is enrolled in SNAP All specifications include additional controls forfamily size race and marital status of the family head presence of children and presence of an adult male

Source Annual data from the March Current Population Survey for 1990 to 2012 estimation includes CurrentPopulation Survey probability weights

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3343

30 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

TABLE 8

Comparison of specifications SNAP enrollment

State-level linear regression

(1) (2) (3)

Log minimum wage -0019 -0035 -0031

(0009) (0012) (0012)

Unemployment rate (100) 0401 0370 0339

(0063) (0077) (0083)

Log median income -0081 -0073 -0061

(0011) (0013) (0013)

Employment-to-population ratio -0183 -0222 -0248

(0039) (0039) (0038)

N 1127 1127 1127

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the SNAP enrollment rate Allregressions include the share of households with children and the racial shares of the population

Source Annual data from the March Current Population Survey for 1990 to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3443

Appendix B | wwwamericanprogresso

TABLE 9

Comparison of specifications SNAP expenditures

State-level linear regression

(1) (2) (3)

Log minimum wage -0121 -0203 -0190

(0075) (0103) (0103)

Unemployment rate (100) 5292 5152 4313

(0464) (0576) (0628)

Log median income -0437 -0417 -0294

(008) (0086) (0078)

Employment-to-population ratio -0040 -0220 0244

(0261) (0260) (0240)

N 1127 1127 1127

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the natural log of state SNAPexpenditures per capita All regressions include the share of households with children and the racial shares ofthe state population

Source Annual data from Bureau of Economic Analysis national income and product accounts tables for 1990to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3543

32 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3643

Appendix C | wwwamericanprogresso

Appendix C Harkin-Miller

policy simulation results

StateMinimum wage

(2014)

SNAP enrollment (2012) Predicted change in SNAP enrollment

RateTotal

(persons)

Model 1 family

enrollment (linearprobability)

Model 2 state

enrollment (linearregression)

Model 3 s

expenditu(linear regre

Alabama $725 164 788682 -66922 -58897 -58906

Alaska $775 120 87436 -8104 -6887 -3288

Arizona $790 201 1319323 -67435 -56738 -64356

Arkansas $725 78 230489 -40977 -36063 -35248

California $800 206 7813680 -371131 -310222 -18223

Colorado $800 164 853155 -50684 -42365 -23926

Connecticut $870 91 326621 -22456 -17975 -13711

Delaware $725 186 170262 -12739 -11211 -10647

District of Columbia $825 133 84009 -5370 -4417 -3632

Florida $793 166 3208026 -195813 -164426 -13046

Georgia $725 160 1586336 -137741 -121224 -11004

Hawaii $725 96 133662 -19310 -16995 -14933

Idaho $725 92 147501 -22165 -19507 -15809

Illinois $825 95 1225084 -109088 -89742 -70955

Indiana $725 125 816233 -90818 -79928 -83985

Iowa $725 155 478011 -42716 -37594 -28556

Kansas $725 135 388269 -40082 -35275 -27461

Kentucky $725 130 568821 -60840 -53544 -52259

Louisiana $725 149 683832 -63929 -56263 -66083

Maine $750 77 101976 -16567 -14323 -15234

Maryland $725 144 846415 -81748 -71946 -38370

Massachusetts $800 130 864721 -64902 -54251 -42913

Michigan $740 146 1439141 -128801 -112140 -11022

Minnesota $725 133 713646 -74730 -65769 -37878

TABLE 10

SNAP enrollments Predicted changes by state under the Harkin-Miller bill

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3743

34 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

StateMinimum wage

(2014)

SNAP enrollment (2012) Predicted change in SNAP enrollment

RateTotal

(persons)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 state

expenditures

(linear regression)

Mississippi $725 129 386501 -41486 -36511 -46467

Missouri $750 172 1036182 -75131 -64952 -56244

Montana $790 132 132452 -10350 -8708 -5846

Nebraska $725 124 230683 -25773 -22683 -12189

Nevada $825 162 446035 -23349 -19209 -11894

New Hampshire $725 127 168404 -18359 -16157 -5735

New Jersey $825 160 1416666 -75175 -61843 -28236

New Mexico $750 149 310896 -25983 -22463 -22512

New York $800 192 3763553 -191193 -159815 -142182

North Carolina $725 174 1697193 -135417 -119179 -113503

North Dakota $725 87 61225 -9743 -8574 -4021

Ohio $795 143 1647345 -115869 -97169 -88580

Oklahoma $725 129 494053 -53006 -46650 -46854

Oregon $910 124 485326 -17036 -13328 -16398

Pennsylvania $725 161 2053643 -177315 -156052 -125586

Rhode Island $800 156 163730 -10258 -8574 -8698

South Carolina $725 94 445277 -65614 -57746 -50304

South Dakota $725 208 173749 -11586 -10197 -7458

Tennessee $725 142 914903 -89667 -78915 -99134

Texas $725 110 2863779 -362018 -318607 -253285

Utah $725 88 251107 -39658 -34902 -19390

Vermont $873 156 97792 -3823 -3055 -2475

Virginia $725 101 829771 -113723 -100086 -58212

Washington $932 72 496934 -23221 -17947 -17756

West Virginia $725 58 107875 -25792 -22699 -21665

Wisconsin $725 75 427822 -79521 -69986 -53210

Wyoming $725 164 94590 -8010 -7050 -3104

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3843

Appendix C | wwwamericanprogresso

StateMinimum wage

(2014)

SNAP expenditures (2012)Predicted change in SNAP expenditures

(millions of dollars)

Per resident

(millions of

dollars)

Total

(millions of

dollars)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 s

expenditu

(linear regre

Alabama $725 $290 $1397 -$1185 -$1043 -$1043

Alaska $775 $253 $185 -$262 -$223 -$106

Arizona $790 $257 $1686 -$935 -$787 -$892

Arkansas $725 $248 $733 -$636 -$560 -$547

California $800 $189 $7164 -$7276 -$6082 -$3573

Colorado $800 $157 $816 -$862 -$721 -$407

Connecticut $870 $191 $686 -$343 -$275 -$210

Delaware $725 $250 $229 -$205 -$180 -$171

District of Columbia $825 $366 $232 -$146 -$120 -$99

Florida $793 $294 $5676 -$4429 -$3719 -$2951

Georgia $725 $317 $3140 -$2936 -$2584 -$2346

Hawaii $725 $335 $465 -$449 -$395 -$347

Idaho $725 $225 $359 -$376 -$331 -$268

Illinois $825 $249 $3200 -$2096 -$1725 -$1364

Indiana $725 $220 $1439 -$1162 -$1023 -$1075

Iowa $725 $192 $589 -$658 -$579 -$440

Kansas $725 $159 $460 -$502 -$441 -$344

Kentucky $725 $298 $1303 -$1133 -$997 -$973

Louisiana $725 $315 $1450 -$1047 -$922 -$1083

Maine $750 $281 $373 -$267 -$231 -$246

Maryland $725 $188 $1109 -$1765 -$1553 -$828

Massachusetts $800 $206 $1366 -$1030 -$861 -$681

Michigan $740 $300 $2963 -$2400 -$2090 -$2054

Minnesota $725 $140 $755 -$1113 -$980 -$564

Mississippi $725 $326 $973 -$649 -$571 -$726

Missouri $750 $241 $1452 -$1278 -$1104 -$956

Montana $790 $190 $191 -$179 -$151 -$101

Nebraska $725 $140 $259 -$409 -$360 -$194

Nevada $825 $191 $527 -$441 -$363 -$225

New Hampshire $725 $126 $167 -$399 -$351 -$125

TABLE 11

SNAP expenditures Predicted changes by state under the Harkin-Miller bill

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3943

36 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

StateMinimum wage

(2014)

SNAP expenditures (2012)Predicted change in SNAP expenditures

(millions of dollars)

Per resident

(millions of

dollars)

Total

(millions of

dollars)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 state

expenditures

(linear regression)

New Jersey $825 $160 $1420 -$1611 -$1325 -$605

New Mexico $750 $324 $675 -$513 -$444 -$445New York $800 $287 $5616 -$3766 -$3148 -$2801

North Carolina $725 $252 $2454 -$2187 -$1925 -$1833

North Dakota $725 $128 $90 -$162 -$143 -$67

Ohio $795 $259 $2995 -$2013 -$1688 -$1539

Oklahoma $725 $248 $945 -$799 -$703 -$706

Oregon $910 $322 $1255 -$272 -$213 -$262

Pennsylvania $725 $218 $2779 -$2930 -$2579 -$2075

Rhode Island $800 $280 $294 -$173 -$144 -$147

South Carolina $725 $291 $1373 -$1337 -$1177 -$1025South Dakota $725 $198 $165 -$192 -$169 -$123

Tennessee $725 $324 $2091 -$1413 -$1243 -$1562

Texas $725 $230 $5997 -$6402 -$5634 -$4479

Utah $725 $141 $402 -$614 -$541 -$300

Vermont $873 $230 $144 -$66 -$53 -$43

Virginia $725 $173 $1413 -$2062 -$1815 -$1056

Washington $932 $244 $1682 -$350 -$270 -$267

West Virginia $725 $273 $508 -$451 -$397 -$379

Wisconsin $725 $204 $1166 -$1302 -$1146 -$871Wyoming $725 $95 $55 -$105 -$93 -$41

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4043

Appendix C | wwwamericanprogresso

FIGURE 2

Raising the minimum wage to $1010would cut taxpayer costs in every state

Predicted decreases in cost and enrollment

in SNAP in 50 states

$200+$51ndash$100

$101ndash$200

0ndash$25

$26ndash$50

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4143

38 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

Endnotes

1 US Department of Agriculture ldquoSupplementalNutrition Assistance Program (SNAP)rdquo available athttpwwwfnsusdagovsnapeligibility (last accessedFebruary 2014)

2 For this initial analysis we do not consider Harkin-Millerrsquos increase in subminimum wages for tippedworkers To do s o would increase the estimated SNAP

savings by an unknown amount

3 The Congressional Budget Office estimates thatworkers currently earning between $1010 and $1150per hour would see their wages rise under the Harkin-Miller proposal Congressional Budget O ffice ldquoTheEffects of a Minimum Wage Increase on Employmentand Family Incomerdquo (2014)

4 Marianne Page Joanne Spetz and Jane Millar ldquoDoesthe Minimum Wage Affect Welfare Caseloadsrdquo Journalof Policy Analysis and Management 24 (2) (2005)273ndash295

5 Marianne Bitler and Hilary Hoynes ldquoThe More ThingsChange the More They Stay the Same The SafetyNet Living Arrangements and Poverty in the GreatRecessionrdquo NBER Working Paper 194 49 2013

6 Sylvia Allegretto and others ldquoFast Food Poverty Wages The Public Cost of Low-Wage Jobs in the Fast-FoodIndustryrdquo (Berkeley California Center for LaborResearch and Education 2013) available at httplaborcenterberkeleyedupubliccostsfast_food_poverty_wage

7 David Neumark and William Wascher ldquoDoes a HigherMinimum Wage Enhance the Effectiveness of theEarned Income Tax Creditrdquo Industrial and LaborRelations Review 64 (4) (2011) 712ndash746

8 David Lee and Emmanuel Saez ldquoOptimal MinimumWage Policy in Competitive Labor Marketsrdquo Journal ofPublic Economics 96 (9) (2012) 739ndash749

9 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family IncomerdquoFebruary 2014

10 Jesse Rothstein ldquoIs the EITC as Good as an NITConditional Cash Transfers and Tax Incidencerdquo AmericanEconomic Journal Economic Policy 2 (1) (2010) 177ndash208

11 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family Incomerdquo

12 Dube Arindrajit 2013 rdquoMinimum Wagesand the Distribution of Family IncomerdquoUnpublished working paper Available at httpsdldropboxusercontentcomu15038936Dube_MinimumWagesFamilyIncomespdf

13 As Dube explains in ldquoThe poverty of Minimum WageFactsrdquo the simulation approach underestimate stemsfrom a number of unwarranted assumptions includingthe range of actual wage increases and the accuracy ofwage data in the Current Population Survey The causal

approach does not make these assumptions

14 Allegretto Sylvia and others 2013 ldquoCredible ResearchDesigns for Minimum Wage Studiesrdquo Working Paper148-13 University of California Berkeley Institutefor Research on Labor and Employment Available athttpwwwirleberkeleyeduworkingpapers148-13pdf

15 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family In comerdquo

16 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo (2012)

17 Ibid

18 Ibid

19 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo

20 US Department of Agriculture ldquoSupplementalNutrition Assistance Program (SNAP)rdquo

21 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo

22 Ibid

23 We deviate from the Census Bureaursquos definition ofa family unit which is ldquotwo people or more (on e ofwhom is the householder) related by bir th marriageor adoption and residing togetherrdquo US Bureau ofthe Census ldquoCurrent Population Survey DefinitionsFamilyrdquo available at httpwwwcensusgovcpsabout

cpsdefhtml (last accessed February 2014) We countas a family unit any individual residing on his or herown two or more persons residing together whodo not belong to a family in the March CPS sampleare constructed as one family in our analysis For thepurposes of food stamp allocations the consumptionresulting from this transfer is probably distributed tofamily members (rather than household members ora single individual within the household) Howeversingle individuals canmdashand domdashreceive SNAPbenefits Excluding them would fail to make theanalysis reflective of the population at large

24 Strictly the family level linear probability modelpredicts the percentage-point decrease in theprobability that an individual family will receive SNAPpayments When applied to a large number of familieshowever we are able to interpret the coefficient asa decrease in the mean of enrollmentmdashthat is a

decrease in the enrollment ratemdashby applying the lawof iterated expectations

25 We generate expenditure predictions from theenrollment modelsmdashand conversely generateenrollment predictions from the expenditure modelmdashby assuming that expenditures per enrolled familyremains the same before and after the minimum wagechange In practice this is likely to be a conservativeestimatemdashthat is to underestimate the decrease inSNAP activity Average SNAP benefits per family willalso decrease as many families that remain eligible forSNAP experience income gains

26 Wage and Hour Division ldquoMinimum Wage Laws inthe States ndash Januar y 1 2014rdquo available at httpwwwdolgovwhdminwageamericahtm (last accessedFebruary 2014)

27 See for example Sylvia Allegretto and others ldquoCredibleResearch Designs for Minimum Wage Studiesrdquo WorkingPaper 148-113 (Berkeley California Institute forResearch on Labor and Employment 2013) available athttpwwwirleberkeleyeduworkingpapers148-13pdf

28 We will report these results in a forthcoming workingpaper

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4243

Endnotes | wwwamericanprogresso

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4343

The Center for American Progress is a nonpartisan research and educational institute

dedicated to promoting a strong just and free America that ensures opportunity

for all We believe that Americans are bound together by a common commitment to

these values and we aspire to ensure that our national policies reflect these values

We work to find progressive and pragmatic solutions to significant domestic and

international problems and develop policy proposals that foster a government that

is ldquoof the people by the people and for the peoplerdquo

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 243

The Effects of Minimum Wageson SNAP Enrollments andExpendituresBy Rachel West and Michael Reich March 2014

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 343

1 Introduction and summary

5 Background

9 Methods and data

15 Results

19 Conclusion

20 About the authors and acknowledgments

21 References

23 Appendix A

25 Appendix B

33 Appendix C

38 Endnotes

Contents

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 443

Introduction and summary | wwwamericanprogress

Introduction and summary

How do minimum wage policy increases affec enrollmens and expendiures on

means-esed public assisance programs In his repor we address his quesion

or he case o he Supplemenal Nuriion Assisance Program or SNAP

ormerly known as he ood samp program

By definiion governmen spending on a means-esed program should decline

as average earnings increase insoar as benefi levels all wih increased earnings

and insoar as he earnings increase makes some individuals ineligible orany benefis Boh o hese condiions are saisfied in he case o he effec o

minimum wages on SNAP benefis SNAP benefis decline 30 cens or every $1

increase in amily earnings and phase ou enirely a abou he ederal povery

level1 Low-wage workers are disproporionaely enrolled in SNAP A minimum

wage increase ha lifs many amilies ou o povery should hereore reduce

public expendiure on his program

Bu he relaionship may be more complex I a minimum wage increase reduces

employmen hereby adding o he number o unemployed he number o SNAP

recipiens could increase SNAP recipiens who are unemployed disabled or

reired will no be affeced by a minimum wage increase Conversely i many

SNAP recipiens have earnings ha already bring hem close o becoming

ineligible or he program a minimum wage increase may have a very small effec

on SNAP expendiures Te quaniaive effec o minimum wages on SNAP

spending is no sel-eviden I requires a causal analysis

In an era o hisorically low real ederal minimum wage raes rising income

inequaliy job-marke sagnaion and conenious debae abou governmen defici

spending he possibiliy ha a higher minimum wage may lead o increased orreduced public spending has grea relevance o he public and o policymakers

Tis repor presens an iniial empirical analysis o he effecs o minimum wage

policy on SNAP paricipaion and expendiures We do so by exploiing more han

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 543

2 Center for American Progress | The Effects of M inim um Wages on SNAP Enro llments and Expenditures

wo decades o variaion in binding sae and ederal minimum wage changes in an

economeric ramework Our uure research will examine he effecs on SNAP

urher and apply an analogous ramework o wo oher public assisance spending

programs he Earned Income ax Credi and Medicaid

According o he finding in his repor a 10 percen increase in he minimum wagereduces SNAP enrollmen by beween 24 percen and 32 percen and reduces

program expendiures by an esimaed 19 percen aking ino accoun each

saersquos 2014 minimum wage level we apply hese resuls o he legislaive proposal

pu orward by Sen om Harkin (D-IA) and Rep George Miller (D-CA) o raise

he ederal minimum wage o $1010 per hour2 Our resuls imply ha he effecs

o he Harkin-Miller proposal on wage increases would reduce SNAP enrollmens

by beween 75 percen and 87 percen (31 million o 36 million persons) Te

oal anicipaed annual decrease in program expendiures is nearly $46 billion or

abou 6 percen o curren SNAP program expendiures

Harkin-Miller proposes o index minimum wage levels in subsequen years o he

consumer price index or CPI Te minimum wage would hen increase a he

same rae as SNAP benefi and eligibiliy levels which are also indexed o he CPI

Consequenly he savings over 10 years in 2014 dollars would be 10 imes he

one-year savings or a oal o approximaely $46 billion

Some o he reducion in SNAP program enrollmen and expendiures would

occur among workers making less han $1010 per hour991252hose whose pay would

be direcly increased by he minimum wage law Anoher par o he reducion

would occur among workers currenly earning beween $1010 and $1150 who

would also receive pay increases3

Alhough a large number o sudies have examined he impac o minimum

wage increases on earnings and employmen he impac o such minimum

wage policies on public assisance enrollmens and expendiures remains an

under-explored subjec in he economic lieraure Only a ew sudies discuss

he relaion beween he minimum wage and governmen ranser spending

much less atemp o ideniy he causal effec o one upon he oher Proessors

Marianne Paige Joanne Spez and Jane Millar find posiive effecs o minimum wage increases on welare caseloads as hey sae however heir resuls vary

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 643

Introduction and summary | wwwamericanprogress

considerably wih differen sample periods and assumpions abou sae rends4

Proessors Marianne Biler and Hilary Hoynes discuss he imporance o SNAP

as a saey ne program bu hey do no examine is relaion o minimum wage

policy5 Research economis Sylvia Allegreto and her Universiy o Caliornia a

Berkeley colleagues show ha low-wage workers in general and as-ood workers

in paricular are much more likely o be SNAP recipiens han all workers6

Several sudies have examined he relaionship beween he minimum wage and

he Earned Income ax Credi or EIC Proessor David Neumark and William

Wascher a researcher a he Federal Reserve Board o Governors find ha a

higher minimum wage increases EIC benefis or amilies in deep povery

while reducing EIC benefis or some sub-groups7 Proessors David Lee and

Emmanuel Saez argue ha he minimum wage and EIC are complemenary

policies no subsiues8 Te Congressional Budge Office or CBO argues ha

a minimum wage increase will no have a subsanial effec on EIC spending9

while Proessor Jesse Rohsein examines wheher he posiive effec o he EICon emale labor supply has lowered wages10 While hese sudies are o ineres

he EIC is quie differen rom SNAP in having a subsanial phase-in period in

which EIC benefis increase as well as a long phase-ou period wih complee

phase-ou a an annual income o abou $48000 or a amily o our quie a bi

above he reach o he minimum wage11

Research by Proessor Arindraji Dube on he causal effec o he minimum wage

on amily povery represens he sudy mos relaed o he one a hand12 Dube finds

ha Harkin-Miller would raise abou 46 million non-elderly Americans above

he ederal povery level or FPL In conras when CBO uses a simple simulaion

mehod o address he same quesion hey find ha Harkin-Miller would raise

900000 people above FPL13 Te difference beween hese wo esimaes highlighs

he imporance o underaking a causal analysis Te mehods used in his paper are

in many respecs similar o Dubersquos Moreover since eligibiliy and benefi levels or

programs such as SNAP and Medicaid are ied o he ederal povery level Dubersquos

findings have direc implicaions or his sudy Noneheless his repor appears o

be he firs sudy o examine he effecs o he minimum wage on SNAP In uure

work we plan o underake similar analyses or he EIC and Medicaid

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 743

4 Center for American Progress | The Effects of M inim um Wages on SNAP Enro llments and Expenditures

Te repor proceeds as ollows

bull Secion 1 provides background inormaion on he ederal minimum wage sae

minimum wages and he SNAP program

bullSecion 2 describes our mehods and daa

bull Secion 3 provides our main resuls including a simulaion o he effecs o a

Harkin-Miller minimum wage increase and a sae-by-sae analysis

bull Secion 4 presens our conclusions

Furher deails are provided in a series o appendices

Correction April 28 2014 Tis report incorrectly stated the potential reduction in

SNAP enrollment fom the Harkin-Miller proposal Te correct amount is 75 percentand 87 percent (31 million to 36 million persons) as stated in the reportrsquos tables

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 843

Background | wwwamericanprogress

Background

The federal minimum wage

Te ederal minimum wage was las increased in July o 2009

nearly five years ago During he pas wo decades many saes

have passed legislaion fixing he minimum wage a a higher

level han he ederal minimum Te maps in Figure 1 show ha

while saes in every region o he Unied Saes have adoped

higher minimum wages hey are no disribued randomly bygeography As shown in he 2013 sudy ldquoCredible Research

Designs or Minimum Wage Sudiesrdquo by economiss Sylvia

Allegreto Arindraji Dube Michael Reich and Ben Zipperer

hese saes vary sysemaically rom he oher saes by a number

o characerisics ha affec low-wage employmen rends bu

which are no hemselves relaed o minimum wage policy14

Te nonrandom patern o minimum wage adopion has

imporan implicaions or obaining unbiased esimaes o

minimum wages on employmen In paricular naional panel

sudies ha use sae and ime fixed effec models991252such as

a 1992 sudy by David Neumark and William Wascher991252

spuriously esimae negaive employmen effecs Te reason

or his resul is uncovered using ess or pre-rends Tese

ess find ha low-wage employmen was already declining wo

years beore minimum wages were implemened By making a

saisically large number o local comparisons ha conrol or

heerogeneiy among saes and by ime eliminaes his pre-

rend For his reason we conduc similar ess or our SNAPoucomes and use model specificaions ha include local

comparisons as in he sudy cied above

FIGURE 1

High versus low minimum wage stat

from 1990 to 2012

Means and variances

More than $533

Average minimum wage over 1990ndash2012

Less than or equal to $533

More than $121

Minimum wage variance over 1990ndash2012

Less than or equal to $121

Notes State means and variances were calculated using annual stat

minimum wage data from 1990 to 2012 The shading on the maps

partitions the states into above- and below-median values

Source Sylvia Allegretto and others ldquoCredible Research Designs forMinimum Wage Studiesrdquo Working Paper 148-13 (Institute for Resear

on Labor and Employment 2013) available at httpwwwirleberkeleyeduworkingpapers148-13pdf

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 943

6 Center for American Progress | The Effects of M inim um Wages on SNAP Enro llments and Expenditures

Te CBO recenly projeced ha in 2016 17 million workers will earn less han

he $1010 hourly wage proposed in he Harkin-Miller bill Furhermore he

CBO esimaes ha an addiional 8 million workers earned beween $1010 and

$1150 per hour and were also likely o experience a wage increase15

Supplemental Nutrition Assistance Program

We ocus our inquiry on SNAP Benefis under he program are enirely

ederally unded he program is adminisered by he US Deparmen o

Agriculure ogeher wih sae agencies which share in adminisraion

coss Spending on SNAP has grown in he pas decade reaching $78 bil lion

in 2011 wih SNAP enrollmen increasing o 45 million people abou one-

sevenh o he US populaion16 According o he CBO changes since 1990 in

SNAP spending and enrollmens are primarily he resul o cyclical economic

condiions noably changes in he unemploymen rae and changes in per capiaincome17 Te 2009 American Recovery and Reinvesmen Ac emporarily

increased SNAP benefi amouns by 136 percen as repored by he US

Deparmen o Agriculurersquos Food and Nuriion Service hese higher benefi

levels expired on November 1 201318 Te CBO esimaes ha abou wo-

hirds o he changes in SNAP expendiure are associaed wih changes in he

number o recipiens and one-hird wih changes in he benefis received when

recipiensrsquo incomes change19

In fiscal year 2014 SNAPrsquos maximum monhly benefis are $189 or a single

individual $497 or a amily o hree and $750 or a amily o five Benefis are

reduced by 30 cens per dollar received and phase ou enirely a gross monhly

household incomes o 130 percen o he ederal povery level $1245 or a

single individual $2116 or a amily o hree and $2987 or a amily o five

o deermine benefis SNAP also defines a ne monhly income concep and

ses benefis a 100 percen o he ederal povery level using his concep

Calculaion o ne monhly income can include cerain deducions rom

monhly gross income such as medical expenses and child care coss Alhough

saes are permited some laiude on wha deducions are allowed in pracice

hese vary by very small amouns Our saisical model akes accoun o sae-specific differences in benefis20

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1043

Background | wwwamericanprogress

SNAP imposes an employmen or work-raining requiremen or able-bodied

prime-age aduls991252hose beween he ages o 18 and 50 and wihou disabiliies

or dependen children Such households can receive only hree monhs o benefis

in a hree-year period In recen years abou 85 percen o households receiving

benefis have incomes below he ederal povery level 49 percen have dependen

children 16 percen are age 60 or older 20 percen are disabled and 30 percenrepor some earned income21

A 2012 CBO repor also noes ha ake-up raes among eligible SNAP recipiens

average abou 70 percen wih much lower ake-up among elderly households

Te ake-up rae increases in harder economic imes I also increased when sigma

issues were reduced as SNAP debi cards replaced acual ood samps ake-up

is especially high among hose mos needy Adminisraive spending equaled 91

percen o he poenial spending ha would have occurred i all eligible recipiens

were enrolled Alhough some SNAP policy changes have occurred since 1990

mos were relaively minor and all were naional in scope Te 1996 welare reorm bill eliminaed SNAP eligibiliy or some legal immigrans limied he ime lengh

o eligibiliy or able-bodied childless aduls and reduced maximum benefis

Some o hese resricions were relaxed in 2002 and again in he American

Recovery and Reinvesmen Ac in 200922

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1143

8 Center for American Progress | The Effects of M inim um Wages on SNAP Enro llments and Expenditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1243

Methods and data | wwwamericanprogress

Methods and data

As previously menioned we exploi variaion in minimum wages by sae and ime

o examine heir causal effecs on SNAP enrollmens and expendiures o do so we

merge daa rom 1990 hrough 2012 drawn rom he Annual Social and Economic

Supplemen o he Curren Populaion Survey991252an annual Census Bureau survey

commonly known as he March CPS ha includes23 inormaion on SNAP

enrollmens a he amily level991252wih sae-level daa on minimum wages SNAP

expendiures populaion unemploymen raes and sae median income levels

o conrol or ime-varying heerogeneiy among saes our specificaions includeconrols or sae linear rends and effecs by Census division and ime We esimae

effecs a wo levels allowing or amily variaion and allowing only or sae-level

variaion We also employ a se o sandard demographic conrols such as amily size

and composiion and race and ehnic composiion

Distinguishing causation from correlation

How can we ensure ha our analysis does no pick up a spurious correlaion or

example he endency o more economically vibran saes o implemen higher

minimum wages Disinguishing correlaion and policy endogeneiy rom rue

causal effecs is he primary moivaion or economeric analysis In he ideal

experimen researchers would begin wih wo saes991252ha are alike in every

respec prior o he policy991252 and ldquoreardquo only one o hese saes wih a higher

minimum wage Tey would atemp o shield hese saes rom any influence ha

could obscure heir undersanding o he minimum wagersquos direc effec on SNAP

enrollmen Researchers o course canno conduc such experimens

We can however use saisical mehods o conrol simulaneously heindependen effecs on SNAP o sae unemploymen raes sae income levels

and common rajecories among saes wihin he same Census division By

ensuring similariy along hese dimensions we maximize he likelihood ha

SNAP aciviy in wo saes would have comparable oucomes in he absence

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1343

10 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

o a minimum wage policy change Tus i a new minimum wage policy were

implemened in one sae only he researchers could atribue all o he difference

hey observe in SNAP aciviy o he new minimum wage policy

In oher words we approximae he ideal experimen by using non-experimenal

saisical mehods Te desirable ldquopre-exising similariiesrdquo beween saes ha wehave defined above inorm our choice o conrol variables in a saisical seting

More precisely in our muliple regression models we use median amily income

he unemploymen rae he employmen-o-populaion raio and regional and

ime idenifiers o consruc an appropriae group o peers or each sae on he

eve o a policy change

Data description

wo daa ses include inormaion abou boh income and paricipaion inpublic programs Te Survey o Income and Program Paricipaion or SIPP

which is conduced in inermiten years has he advanage o ollowing he

same individuals over a period o ime In oher words i is a longiudinal daa

se I also has he advanage o conaining monhly daa However he sample

size o he SIPP is no sufficien or analyzing variaions in sae-level minimum

wages Te March CPS has he advanage o a much larger sample size and i is

conduced annually wihou any breaks in ime I has he disadvanage o being

a cross-secional daa se so we canno ollow he same individuals over ime991252

sricly speaking over more han one year On ne he March CPS is much more

suiable or our sudy We examine he empirical relaionship beween minimum

wage policy and ood samp aciviy a wo levels o aggregaion he amily level

and he sae level Family-level daa are drawn rom he March CPS

Te March CPS comprises responses rom he residens o 50000 o 60000

dwelling places surveyed per year and conains deailed inormaion on he

residensrsquo employmen and income including income rom ranser paymens

Te sample or our analysis comprises more han 128 million amily unis during

he period rom 1990 o 2012 (inclusive) Survey weighs allow us o analyze

SNAP paricipaion in a manner ha is represenaive o he US populaion alarge Over all years he share o amilies reporing ood samp receip in he

weighed March CPS sample is 91 percen Te enrollmen rae was a a low o 6

percen in he year 2000 In 2012 he mos recen year in our panel 133 percen

o amilies repored paricipaing in SNAP a some poin during he survey year

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1443

Methods and data | wwwamericanprogresso

Te March CPS also collecs inormaion on he number o SNAP recipiens in

he household he number o monhs he household receives SNAP benefis

and he cash-equivalen value o he SNAP benefis received However he

value o SNAP benefis is severely underrepored among recipiens perhaps

because recipiens are unaware o he exac monhly cash-equivalen value o

benefis hey receive

Our firs empirical sraegy ocuses solely on SNAP enrollmen By using he

amily as he uni o analysis we are able o inser saisical conrols o accoun

or non-wage-relaed acors ha influence any paricular amilyrsquos likelihood o

program paricipaion wih he inenion o isolaing any differences in program

paricipaion ha are due purely o changes in wage policy Tis approach

idenifies he effecs o low-wage labor policy on he exernal margin991252ha is

he effec o he minimum wage on he likelihood ha a amily paricipaes in he

SNAP program a all991252as opposed o he inernal margin or how much SNAP

unding he amily would receive

Our second empirical ramework uses sae-level adminisraive daa Ta is

we aggregae he daa o obain a single daa poin or each saeyear back

o 1990 represening he mean o he oucome or he sae Te sae-level

esimaion serves as a robusness check on he amily-level resuls or SNAP

paricipaion Also using aggregaed daa allows us o esimae direcly he

causal effec o minimum wage changes on SNAP spending Tis is no possible

a he amily level as discussed above daa on cash-equivalen value o ood

samps or SNAP recipiens is very requenly no repored in he March CPS

and when i is repored he inormaion may be unreliable By conras he

Bureau o Economic Analysis publishes aggregae SNAP spending a he sae

level in is Naional Income and Produc Accoun or NIPA ables Tus

while we are unable o observe he heerogeneiy in he cash value o SNAP

or amilies in each sae we are able o calculae average SNAP spending

per residen in each sae per year Supporing covariaes include he annual

unemploymen and employmen daa rom he Bureau o Labor Saisics or

BLS and sae-level populaion series rom he iner-decennial census releases

Minimum wage daa are available rom he BLSrsquos wages and hours division For

sae minimum wage changes enaced a oher imes han he firs o he yearan average value or he year is used

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1543

12 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

Family level model specifications

We firs examine he effec o he minimum wage on paricipaion in public

assisance programs For amily residing in sae and during year we esimae

an equaion o he ollowing orm

(1)

is a binary variable ha is se equal o 1 i a leas one member o amily

received ood samps during he survey year is a se o sae-level

characerisics including annual averages o he unemploymen rae he

employmen-o-populaion raio and he naural log o median amily income

is a vecor o amily atribues including indicaors or he race and marial saus

o he amily head size o he amily he presence o children and he presence oan adul male Sae fixed effecs are capured by o conrol or ime-varying

heerogeneiy our preerred model specificaion also includes year fixed effecs

ha vary by Census division ( ) and sae-level linear ime rends In

Appendix B we jusiy he inclusion o hese las wo erms We also compare he

resuls rom our preerred specificaion wih less sauraed specificaions

Te effec o ineres which is capured by is he expeced change in he

probabiliy o receiving SNAP benefis wih respec o a change in he (log o he)

binding minimum wage in sae during year We repor robus sandard errors

clusered a he sae level We esimae he parameers using linear regression

producing a linear probabiliy model Deails o he model selecion process are

covered in Appendix B below

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1643

Methods and data | wwwamericanprogresso

State-level model specifications

Te sae-level models are similarly specified For sae in year we assume ha

(2)

In his model is now eiher he SNAP enrollmen rae in sae during year

or he naural logarihm o per capia SNAP expendiures in sae during

year is once again a se o sae-level characerisics including he same

sae-level covariaes as in he amily regressions (annual average unemploymen

rae employmen-o-populaion raio naural log o median amily income)

wih he addiion o amily level characerisics averaged across he sae (average

amily size and he shares o populaion consiued by each o five racialehnicgroups) Sae fixed effecs are represened by As above our preerred model

specificaion includes year fixed effecs ha vary by Census division ( ) and

sae-level linear ime rends as elaboraed in Appendix B Te effec o

ineres is capured by

We esimae boh sae-level models (enrollmen and expendiures) using

ordinary leas squares regression Tus he inerpreaion o he coefficien is no

longer ha o a change in probabiliy as in he binary oucome models described

above Raher or he sae-level SNAP enrollmen model represens he

expeced change (in percenage poins) in he saersquos SNAP enrollmen rae ha

is due o a 1 percen change in he minimum wage For he SNAP expendiures

model is simply he elasiciy o SNAP spending wih respec o he minimum

wage991252ha is he percenage change in sae expendiures expeced o resul rom

a 1 percen change in ha saersquos minimum wage For urher deails on model

specificaion reer o Appendix B below

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1743

14 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1843

Results | wwwamericanprogresso

Results

Estimated minimum wage effects on SNAP enrollment and

expenditures

able 1 shows he esimaed parameer o ineres991252he coefficien o he

minimum wage or he preerred model o each ype Coefficiens on he

minimum wage variable are no direcly comparable across models because all

our models have a differen uncional orm o undersand and compare hese

esimaes we compue he change in SNAP aciviy prediced or a paricular wage scenario Te final column in able 1 answers he quesion Wha would be

he expeced change in SNAP aciviy in response o a 10 percen increase in he

minimum wage Te answer o his quesion varies wih he value o he inpu

parameers in he able we calculae he percenage decrease in enrollmen or

expendiures prediced or he average sae wih a minimum wage o $725 in

2014 Te sae-level SNAP expendiure model which is a consan-elasiciy

model conveys elasiciy inormaion direcly or he change in expendiures per

capia in he sae

TABLE 1

Comparison of national SNAP predictions for a 10 percent increase in the federal minimum wage

Model LevelRegression

type

Predicted outcome Coefficient of log

(minimum wage)

(Standard error)

Effect of a 10 percent in

in the minimum wage

VariableForm of

variable

Total

enrollment

Tot

expend

1 Family Linear

probability Enrollment Binary (enrolled=1)

-0042

(0008) -317 N

2

StateLinear regression

(ordinary least

squares)

Enrollment State enrollment rate (percent) -0031

(0012) -235 N

3 Expenditures Log (state expenditures per capita) -0190

(0103) NA -19

plt01 plt005 plt001Note Predicted changes are calculated for the average state with a minimum wage of $725 in 2014

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1943

16 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

According o his model a 10 percen increase in he minimum wage would resul

in a 19 percen reducion in SNAP expendiures Tis resul is reassuringly similar

o Dubersquos resul or povery reducion Te wo enrollmen models are much more

precisely esimaed han he expendiure model Te sae-level enrollmen model

finds ha a 10 percen minimum wage increase in a low-wage sae is associaed wih

a 235 percen decrease in SNAP enrollmens Te amily-level linear probabiliymodel predics a somewha greaer elasiciy or low minimum wage saes an

increase o 10 percen in he ederal minimum would resul in a 317 percen decline

in SNAP enrollmen24 Te differences in hese esimaes sem rom a number o

acors including difference in model uncional orm and daa used We rea his

range o elasiciy esimaes as an upper and lower bound on enrollmen impacs

Harkin-Miller bill National and state-level predicted impacts

Wha would be he prediced change or he SNAP program i he ederalminimum were raised o $1010 as proposed in he Harkin-Miller bill In order o

make his inerence we accoun or he ac ha no all saes are currenly subjec

o he ederal minimum wage a he beginning o 2014 21 saes mainained

higher minimum wages han $725 In hose saes an increase in he ederal

minimum wage may or may no be binding or employers in he sae depending

upon wheher he new ederal minimum exceeds he sae-level minimum Bu

regardless o wheher a minimum wage change is binding he impac on SNAP

aciviy will be lower in high minimum wage saes In order o accoun or his

properly we calculae sae by sae he percenage wage change ha would resul

rom he Harkin-Miller proposal and apply he parameers rom each o he hree

models above o compue he expeced decrease in SNAP aciviy or each sae

In his exercise we use saesrsquo curren (2014) minimum wage levels and assume

as a baseline he 2012 levels o SNAP enrollmen and expendiure as 2012 is he

mos recen year or which SNAP daa are available

able 10 and able 11 in Appendix C repor he esimaed effecs on SNAP

enrollmen and expendiures respecively or each sae under he Harkin-Miller

bill25 An increase o $1010 i enaced oday would represen beween a 393

percen wage increase in a $725 minimum wage sae and an 84 percen increasein Washingon sae which has he highes minimum wage in he naion a $932

as o January 201426 Slighly more han 56 percen o he decrease in expendiures

and abou 59 percen o he decrease in enrollmen would occur in saes wih

presen-day minimum wages o $725

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2043

Results | wwwamericanprogresso

In 2012 hese saes were home o 46 percen o he American populaion

and accouned or a roughly equivalen percenage o oal naional SNAP

expendiures No surprisingly he larges enrollmen decreases come rom

populous saes wih very high SNAP enrollmen raes andor very low minimum

wages Te larges prediced enrollmen reducion991252beween 319000 individuals

and 362000 individuals991252would occur in exas which has a $725 minimum wage In Caliornia which has a 206 percen SNAP paricipaion rae and an $8

per hour minimum wage we anicipae SNAP enrollmen declines o 310000

persons o 371000 persons And in Florida which had a SNAP paricipaion rae

o 166 percen and a $793 minimum wage enrollmen could decline by beween

164000 individuals and 196000 individuals For he our saes ha ogeher

accouned or he greaes amoun o SNAP spending in 2012991252exas Caliornia

Florida and New York respecively991252he combined expendiure reducion rom

he Harkin-Miller bill is prediced o be $14 billion

able 2 summarizes he prediced declines in SNAP aciviy or he naion as a whole ha would resul rom he direc and indirec effecs o he Harkin-Miller

bill Enrollmen would all beween 31 million persons and 36 million persons

represening 75 percen o 87 percen o curren enrollmen Te anicipaed

reducion in program expendiures would be nearly $46 billion or 61 percen o

program expendiures

TABLE 2

Comparison of national SNAP predictions under the Harkin-Miller billrsquos $1010 minimum wage

Model

Enrollment

(persons)

Expenditures

(millions of dollars)

Current (2012) Predicted Change Current (2012) Predicted Chan

Family enrollment (linear probability)

41866195

45489339 -3623144

$74861

NA NA

State enrollment (ordinary least squares) 38745435 -3120759 NA NA

State expenditures (ordinary least squares) NA NA $70305 -$45

Note Calculations use 2014 state minimum wages and the most recent SNAP data from 2012 They assume that per-enrollee expenditures remain constant

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2143

18 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

Tere are o course oher possibiliies or minimum wage legislaion able 3

shows he expeced SNAP changes or he Unied Saes under a variey o wage

scenarios calculaed using he sae-level models I saes were no able o se

heir minimum wages independenly such ha all saes were consrained by

he ederal minimum o $725 SNAP would be received by abou 514000 more

people across he Unied Saes a an addiional program cos o nearly hree-quarers o a billion dollars In conras he effecs o a higher minimum wage

proposal991252a ederal wage floor o $11 per hour991252would decrease enrollmen in

SNAP by more han 10 percen and decrease program coss by 83 percen

TABLE 3

Summary of par ticipation and expenditures under wage scenarios

If all states had

minimum wages of

Enrollment(persons)

Expenditures

(millions of dollars)

Predicted Change Predicted Change

Recent levels (2014) 41866195 $74861

$725 42380520 514326 $75604 $743

$800 41423919 -442276 $74209 -$652

$900 40148451 -1717744 $72350 -$2511

$1000 38872982 -2993212 $70490 -$4371

$1010 38745435 -3120759 $70305 -$4556

$1100 37597514 -4268681 $68631 -$6230

Note Calculations use state-level enrollment model coefficient

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2243

Conclusion | wwwamericanprogresso

Conclusion

An exensive body o lieraure examines employmen effecs o he minimum

wage A much smaller se o sudies examines how he minimum wage affecs

povery and only a handul o papers examine he effecs o he minimum wage

on he EIC Our analysis is he firs o examine he effecs o he minimum

wage on SNAP

Our findings indicae ha increased earnings rom minimum wage changes do

reduce SNAP enrollmens and expendiures We esimae ha he Harkin-Miller bill would save axpayers nearly $46 billion per year equivalen o 61 percen

o SNAP expendiures in 2012 he las year or which daa are available Over a

10-year period he esimaed savings amoun o nearly $46 billion

Our repor is subjec o limiaions ha we expec o overcome in our uure

research Firs he findings do no ake ino accoun possible ineracions among

SNAP he EIC and Medicaid Te eligibiliy cuoffs among hese programs

are quie differen suggesing ha such ineracions may be minor Noneheless

he join effecs can only be deermined by urher research using a causal

model Second i would be useul o know he disribuion o SNAP reducions

along he wage disribuion Using he Congressional Budge Officersquos calculaions

o how much he oal dollar value o wage would increase under he Harkin-

Miller proposal our findings imply ha he decline in overall SNAP spending

equals abou 15 percen o he oal resuling increase in wages Te amoun and

disribuion o his offse are o considerable ineres Minimum wage beneficiaries

who come rom working amilies already well above he povery line would no

see any offse while hose who are currenly considerably below he povery line

will see larger offses Tese issues will also be a subjec or our uure research

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2343

20 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

About the authors

Rachel West is a maser o public policy candidae a he Goldman School

o Public Policy Universiy o Caliornia Berkeley Her research ocuses on

economic policy in he areas o low-wage labor and povery

Michael Reich is proessor o economics and direcor o he Insiue or

Research on Labor and Employmen a he Universiy o Caliornia a Berkeley

His research publicaions cover numerous areas o labor economics including

racial inequaliy labor marke segmenaion high-perormance workplaces

union-managemen cooperaion Japanese labor-managemen sysems living

wages and minimum wages He received his docorae in economics rom

Harvard Universiy

Acknowledgments

We are graeul o Sylvia Allegreto Arindraji Dube Bill Leser Jesse Rohsein

Daniel Tompson and Ben Zipperer or heir valuable suggesions

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2443

References | wwwamericanprogresso

References

Allegreto Sylvia and ohers 2013 ldquoFas Food Povery Wages Te Public Cos o Low-Wage Jobsin he Fas-Food Indusryrdquo Universiy o Caliornia Berkeley Cener or Labor Research andEducaion Available a htplaborcenerberkeleyedupubliccossas_ood_povery_wage

Allegreto Sylvia and ohers 2013 ldquoCredible Research Designs or Minimum Wage Sudiesrdquo

Working Paper 148-13 Universiy o Caliornia Berkeley Insiue or Research on Labor andEmploymen Available a htpwwwirleberkeleyeduworkingpapers148-13pd

Biler Marianne and Hilary Hoynes 2013 ldquo Te More Tings Change he More Tey Say heSame Te Saey Ne Living Arrangemens and Povery in he Grea Recessionrdquo Working Paper19449 Naional Bureau o Economic Research

Congressional Budge Office 2012 ldquoTe Supplemenal Nuriion Assisance Programrdquo Washingon Available a htpwwwcbogovsiesdeaulfilescbofilesatachmens04-19-SNAPpd

991252 991252 991252 2014 ldquoTe Effec o a Minimum-Wage Increase on Employmen and Family Incomerdquo Washingon Available a htpwwwcbogovsiesdeaulfilescbofilesatachmens44995-MinimumWagepd

Dube Arindraji 2013 rdquoMinimum Wages and he Disribuion o Family Incomerdquo Unpublished working paper Available a htpsdldropboxuserconencomu15038936Dube_ MinimumWagesFamilyIncomespd

991252 991252 991252 2014 ldquoTe Povery o Minimum Wage Facsrdquo Arindraji Dube Blog January 22 Availablea htparindubecom20140122he-povery-o-minimum-wage-acs

Lee David and Emmanuel Saez 2012 ldquoOpimal Minimum Wage Policy in Compeiive LaborMarkesrdquo Journal o Public Economics 96 (9) 739ndash749

Neumark David and William Wascher 1992 ldquoEmploymen Effecs o Minimum and Subminimum Wages Panel Daa on Sae Minimum Wage Lawsrdquo Industrial and Labor Relations Review 46 (1)

55ndash81

Neumark David and William Wascher 2011 ldquoDoes a Higher Minimum Wage Enhance heEffeciveness o he Earned Income ax Credirdquo Industrial and Labor Relations Review 64 (4)712ndash746

Page Marianne Joanne Spez and Jane Millar 2005 ldquoDoes he Minimum Wage Affec WelareCaseloadsrdquo Journal o Policy Analysis and Management 24 (2) 273ndash295

Rohsein Jesse 2010 ldquoIs he EIC as Good as an NI Condiional Cash ransers and ax

Incidencerdquo American Economic Journal Economic Policy 2 (1) 177ndash208

Wage and Hours Division 2014 ldquoMinimum Wage Laws in he Saes ndash January 1 2014rdquo USDeparmen o Labor (htpwwwdolgovwhdminwageamericahm [February 2014])

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2543

22 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2643

Appendix A | wwwamericanprogresso

Appendix A

Pre-trend falsification check

Recen minimum wage research27 highlighs a common flaw in previous sudies

ailure o veriy ha he oucome variable is ree o negaive pre-exising rends I

or example SNAP aciviy was already rending down in saes ha raised heir

minimum wages beore hese changes came ino effec our regression analysis could

(misakenly) atribue ha reducion o he minimum wage We check or such

pre-rends by inroducing variables ha represen he prior yearrsquos value or leads o

he minimum wage I he model esimaes he minimum wage o have an effec on

he oucome variable beore he wage change wen ino effec hen an unobservedacor no he minimum wage change caused he change in SNAP aciviy

We es he specificaions above or pre-rends by including a one-year lead in

all hree specificaions We find ha he lead erms are small posiive and no

saisically significan indicaing ha he concurren minimum wage991252no

he wage level in prior periods991252is driving he observed changes in SNAP

oucomes28 In paricular he coefficien (sandard error) on he lead erm in

our preerred amily-level enrollmen regression is 011 and no significan

while he coefficien and sandard error o he conemporaneous minimum

wage is unchanged In he sae-level preerred enrollmen regression he

coefficien o he lead erm is again small (07) and i is no significan Te

corresponding coefficien on he lead erm in he sae-level expendiure

regression is 16 and is no significan Te posiive poin esimaes on hese lead

erms resuls no only rule ou disoring negaive pre-rends Tey also sugges

ha our main resuls may underesimae he rue effecs

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2743

24 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2843

Appendix B | wwwamericanprogresso

Appendix B

Model selection process

For boh he amily-level and sae-level models we es hree mehods o conrol

or unobserved geographic- and ime-varying characerisics as suggesed by he

minimum wage lieraure Firs we include only independen sae-specific fixed

effecs and year-specific fixed effecs Tis specificaion (specificaion 1) implicily

assumes ha amilies in any sae consiue an equally good saisical ldquoconrolrdquo

group or hose in any randomly chosen sae afer accouning or various

characerisics (median income and unemploymen rae among ohers) Similarly

simple ime fixed effecs assume ha amilies surveyed in any year can crediblyserve as a conrol group or amilies surveyed in every oher year o he sample

(1990 hrough 2012)

In oher words specificaion 1 assumes ha a saersquos immediae neighbor provides

no beter a couneracual or he effec o a minimum wage change han does a

sae across he counry We relax his resricive specificaion sequenially in wo

seps In specificaion 2 we replace simple year fixed effecs wih fixed effecs or

each Census divisionyear (capured as an addiional variable in he vecor By

using division-year effecs we remove he resricion ha amilies in each sae

are equally good saisical conrols or all oher amilies Raher we allow or he

possibiliy ha amilies in similar geographic regions (or example he Souh or

he Norheas) may be more similar o one anoher han amilies arher away

Finally in specificaion 3 we add sae-specific linear ime rends o he previous

specificaion Tus specificaion 3 is he mos rigorous model specificaion in ha

i allows or heerogeneiy along hree dimensions Ta is specificaion 3 allows

each sae o have is own ime-varying rends raher han imposing he resricion

ha saes evolve idenically over he 22 years in he sample

We begin building he heoreical specificaion above rom a se o simpleuncondiional models regression o SNAP aciviy (enrollmen or expendiures)

on he log o he minimum wage and a se o geographic- and ime-specific

effecs (specificaions 1 2 and 3 described above) As shown in ables 1ndash3 (or

specificaion 3) we hen add covariaes sequenially o hese models including

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2943

26 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

firs he vecor o amily-level conrols ollowed by each o several sae-level

covariaes in urn he unemploymen rae log o median-amily income and he

employmen-o-populaion raio Comparable resuls or specificaions 2 and 3

will be available in our orhcoming working paper

As expeced he simple uncondiional models indicae ha he relaionship beween he minimum wage and SNAP enrollmen i one exiss is a more complex

one influenced by oher acors In he uncondiional model he coefficien on he

variable o ineres991252he log o he minimum wage991252is small in magniude and no

saisically differen rom zero Once we accoun or he influence o labor marke

condiions and variaion in income levels on program paricipaion (by including

unemploymen rae and median-amily income conrol variables respecively)

he effec o he minimum wage on SNAP enrollmen is precisely esimaed Te

coefficien o he log minimum wage is slighly higher (-0042) in he amily-level

analysis han he coefficien (-031) in he sae-level analysis Te level o precision

is also higher in he amily-level analysis Tis is o be expeced when using 124million observaions compared o 1127

Te second dimension o model choice concerns he effec specificaion ables

7ndash9 compare he primary coefficiens o ineres or he SNAP enrollmen and

expendiure models For boh he enrollmen models he effec sizes are smalles

or specificaion 1 larges or specificaion 2 and inermediae beween hese wo

in specificaion 3 Recall ha Specificaion 3 conains sae-specific linear ime

rends in addiion o he census divisionyear conrols included in specificaion

2 In he amily-level enrollmen model he sandard error o he minimum wage

coefficien is smaller han in he oher wo specificaions Sandard errors on he

oher variables are much smaller in specificaions 2 and 3 han in specificaion

1 On he basis o coefficien significance (join and individual) specificaions 2

and 3 are sricly preerred in boh enrollmen models o specificaion 1 which

conains only sae and year fixed effecs

A concern wih specificaions 2 and 3 is ha rend conrols such as sae linear

rends may incorrecly absorb some o he delayed impac o a minimum wage

When we es his issue by including lagged minimum wages we do no find ha

delayed effecs are significan Anoher concern is ha more sauraed modelsuse less o he saisical variaion which could reduce he saisical power o

he resuls However he sandard errors or our more sauraed models are no

higher and are lower in some cases han or he less sauraed models Overall

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3043

Appendix B | wwwamericanprogresso

his evidence suppors our use o he sauraed model as he preerred model

specificaion Moreover Dubersquos 2013 sudy shows ha more sauraed models

perorm beter han models wih jus sae and ime fixed effecs

Te esimaed enrollmen regressions a boh he amily and sae levels show large

and saisically significan coefficiens Te esimaed minimum wage effec in heexpendiures regressions991252or which we have only sae-level daa991252is also large

and saisically significan

We do no use weighed regression or he sae-level models preerring o keep

analysis o he ldquoreamenrdquo (ha is o say a minimum wage change) appropriae

o he average sae raher han he average amily or individual I insead our

primary ineres were he impac o a minimum wage change on he average amily

or he average individual we migh choose o designae he number o amilies

in each sae or he sae populaion respecively as analyic weighs in order o

obain a coefficien beter suied or such inerence

TABLE 4

SNAP enrollment

Family-level linear probability

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0048 -0047 -0040 -0043 -0042

(0013) (0013) (001) (0008) (0009)

Unemployment rate (100) 0505 0420 0280

(0083) (0086) (0082)

Log median income -0057 -0039

(0011) (0011)

Employment-to-population ratio -0239

(0038)

N 1242022 1242022 1242022 1242022 1242022

plt01 plt005 plt001

Note Robust standard errors are in parentheses Observations are clustered at the state level The outcome variable is binary and equal to one if a family is enrolledin SNAP All models include state fixed effects Census division x-year fixed effects and state-specific linear time trends All specifications except 3a include additional

controls for family size race and marital status of the family head presence of children and presence of an adult maleSource Annual data from the March Current Population Survey for 1990 to 2012 estimation includes Current Population Survey probability weights

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3143

28 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

TABLE 5

SNAP enrollment

State-level linear regression

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0028 -0024 -0026 -0031 -0031

(0014) (0013) (0013) (0012) (0012)

Unemployment rate (100) 0625 0466 0320 0339

(0087) (0088) (0085) (0083)

Log median income -0090 -0065 -0061

(0013) (0013) (0013)

Employment-to-population ratio -0282 -0248

(0037) (0038)

N 1127 1127 1127 1127 1127

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the SNAP enrollment rate All models include state fixed effects Census division x-year fixedeffects and state-specific linear time trends

Source Annual data from the March Current Population Survey for 1990 to 2012

TABLE 6

SNAP expenditures

State-level linear regression

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0181 -0149 -0156 -0153 -0190

(011) (0103) (0102) (0103) (0103)

Unemployment rate (100) 4587 4152 4232 4313

(0622) (0621) (0633) (0628)

Log median income -0246 -0261 -0294

(0075) (0078) (0078)

Employment-to-population ratio 0155 0244

(0237) (024)

N 1127 1127 1127 1127 1127

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the natural log of state SNAP expenditures per capita for 1990 to 2012 All models include state

fixed effects Census division x-year fixed effects and state-specific linear time trends

Source Annual data from Bureau of Economic Analysis national income and product accounts tables for 1990 to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3243

Appendix B | wwwamericanprogresso

TABLE 7

Comparison of specifications SNAP enrollment

Family-level linear probability

(1) (2) (3)

Log minimum wage -0028 -0049 -0042

(0014) (0017) (0009)

Unemployment rate (100) 0275 0297 0280

(0161) (0076) (0082)

Log median income -0077 -0055 -0039

(0014) (0012) (0011)

Employment-to-population ratio -0238 -0250 -0239

(0054) (004) (0038)

N 1242022 1242022 1242022

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses Observations are clustered at the state level The outcomevariable is binary or equal to one if a family is enrolled in SNAP All specifications include additional controls forfamily size race and marital status of the family head presence of children and presence of an adult male

Source Annual data from the March Current Population Survey for 1990 to 2012 estimation includes CurrentPopulation Survey probability weights

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3343

30 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

TABLE 8

Comparison of specifications SNAP enrollment

State-level linear regression

(1) (2) (3)

Log minimum wage -0019 -0035 -0031

(0009) (0012) (0012)

Unemployment rate (100) 0401 0370 0339

(0063) (0077) (0083)

Log median income -0081 -0073 -0061

(0011) (0013) (0013)

Employment-to-population ratio -0183 -0222 -0248

(0039) (0039) (0038)

N 1127 1127 1127

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the SNAP enrollment rate Allregressions include the share of households with children and the racial shares of the population

Source Annual data from the March Current Population Survey for 1990 to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3443

Appendix B | wwwamericanprogresso

TABLE 9

Comparison of specifications SNAP expenditures

State-level linear regression

(1) (2) (3)

Log minimum wage -0121 -0203 -0190

(0075) (0103) (0103)

Unemployment rate (100) 5292 5152 4313

(0464) (0576) (0628)

Log median income -0437 -0417 -0294

(008) (0086) (0078)

Employment-to-population ratio -0040 -0220 0244

(0261) (0260) (0240)

N 1127 1127 1127

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the natural log of state SNAPexpenditures per capita All regressions include the share of households with children and the racial shares ofthe state population

Source Annual data from Bureau of Economic Analysis national income and product accounts tables for 1990to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3543

32 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3643

Appendix C | wwwamericanprogresso

Appendix C Harkin-Miller

policy simulation results

StateMinimum wage

(2014)

SNAP enrollment (2012) Predicted change in SNAP enrollment

RateTotal

(persons)

Model 1 family

enrollment (linearprobability)

Model 2 state

enrollment (linearregression)

Model 3 s

expenditu(linear regre

Alabama $725 164 788682 -66922 -58897 -58906

Alaska $775 120 87436 -8104 -6887 -3288

Arizona $790 201 1319323 -67435 -56738 -64356

Arkansas $725 78 230489 -40977 -36063 -35248

California $800 206 7813680 -371131 -310222 -18223

Colorado $800 164 853155 -50684 -42365 -23926

Connecticut $870 91 326621 -22456 -17975 -13711

Delaware $725 186 170262 -12739 -11211 -10647

District of Columbia $825 133 84009 -5370 -4417 -3632

Florida $793 166 3208026 -195813 -164426 -13046

Georgia $725 160 1586336 -137741 -121224 -11004

Hawaii $725 96 133662 -19310 -16995 -14933

Idaho $725 92 147501 -22165 -19507 -15809

Illinois $825 95 1225084 -109088 -89742 -70955

Indiana $725 125 816233 -90818 -79928 -83985

Iowa $725 155 478011 -42716 -37594 -28556

Kansas $725 135 388269 -40082 -35275 -27461

Kentucky $725 130 568821 -60840 -53544 -52259

Louisiana $725 149 683832 -63929 -56263 -66083

Maine $750 77 101976 -16567 -14323 -15234

Maryland $725 144 846415 -81748 -71946 -38370

Massachusetts $800 130 864721 -64902 -54251 -42913

Michigan $740 146 1439141 -128801 -112140 -11022

Minnesota $725 133 713646 -74730 -65769 -37878

TABLE 10

SNAP enrollments Predicted changes by state under the Harkin-Miller bill

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3743

34 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

StateMinimum wage

(2014)

SNAP enrollment (2012) Predicted change in SNAP enrollment

RateTotal

(persons)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 state

expenditures

(linear regression)

Mississippi $725 129 386501 -41486 -36511 -46467

Missouri $750 172 1036182 -75131 -64952 -56244

Montana $790 132 132452 -10350 -8708 -5846

Nebraska $725 124 230683 -25773 -22683 -12189

Nevada $825 162 446035 -23349 -19209 -11894

New Hampshire $725 127 168404 -18359 -16157 -5735

New Jersey $825 160 1416666 -75175 -61843 -28236

New Mexico $750 149 310896 -25983 -22463 -22512

New York $800 192 3763553 -191193 -159815 -142182

North Carolina $725 174 1697193 -135417 -119179 -113503

North Dakota $725 87 61225 -9743 -8574 -4021

Ohio $795 143 1647345 -115869 -97169 -88580

Oklahoma $725 129 494053 -53006 -46650 -46854

Oregon $910 124 485326 -17036 -13328 -16398

Pennsylvania $725 161 2053643 -177315 -156052 -125586

Rhode Island $800 156 163730 -10258 -8574 -8698

South Carolina $725 94 445277 -65614 -57746 -50304

South Dakota $725 208 173749 -11586 -10197 -7458

Tennessee $725 142 914903 -89667 -78915 -99134

Texas $725 110 2863779 -362018 -318607 -253285

Utah $725 88 251107 -39658 -34902 -19390

Vermont $873 156 97792 -3823 -3055 -2475

Virginia $725 101 829771 -113723 -100086 -58212

Washington $932 72 496934 -23221 -17947 -17756

West Virginia $725 58 107875 -25792 -22699 -21665

Wisconsin $725 75 427822 -79521 -69986 -53210

Wyoming $725 164 94590 -8010 -7050 -3104

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3843

Appendix C | wwwamericanprogresso

StateMinimum wage

(2014)

SNAP expenditures (2012)Predicted change in SNAP expenditures

(millions of dollars)

Per resident

(millions of

dollars)

Total

(millions of

dollars)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 s

expenditu

(linear regre

Alabama $725 $290 $1397 -$1185 -$1043 -$1043

Alaska $775 $253 $185 -$262 -$223 -$106

Arizona $790 $257 $1686 -$935 -$787 -$892

Arkansas $725 $248 $733 -$636 -$560 -$547

California $800 $189 $7164 -$7276 -$6082 -$3573

Colorado $800 $157 $816 -$862 -$721 -$407

Connecticut $870 $191 $686 -$343 -$275 -$210

Delaware $725 $250 $229 -$205 -$180 -$171

District of Columbia $825 $366 $232 -$146 -$120 -$99

Florida $793 $294 $5676 -$4429 -$3719 -$2951

Georgia $725 $317 $3140 -$2936 -$2584 -$2346

Hawaii $725 $335 $465 -$449 -$395 -$347

Idaho $725 $225 $359 -$376 -$331 -$268

Illinois $825 $249 $3200 -$2096 -$1725 -$1364

Indiana $725 $220 $1439 -$1162 -$1023 -$1075

Iowa $725 $192 $589 -$658 -$579 -$440

Kansas $725 $159 $460 -$502 -$441 -$344

Kentucky $725 $298 $1303 -$1133 -$997 -$973

Louisiana $725 $315 $1450 -$1047 -$922 -$1083

Maine $750 $281 $373 -$267 -$231 -$246

Maryland $725 $188 $1109 -$1765 -$1553 -$828

Massachusetts $800 $206 $1366 -$1030 -$861 -$681

Michigan $740 $300 $2963 -$2400 -$2090 -$2054

Minnesota $725 $140 $755 -$1113 -$980 -$564

Mississippi $725 $326 $973 -$649 -$571 -$726

Missouri $750 $241 $1452 -$1278 -$1104 -$956

Montana $790 $190 $191 -$179 -$151 -$101

Nebraska $725 $140 $259 -$409 -$360 -$194

Nevada $825 $191 $527 -$441 -$363 -$225

New Hampshire $725 $126 $167 -$399 -$351 -$125

TABLE 11

SNAP expenditures Predicted changes by state under the Harkin-Miller bill

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3943

36 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

StateMinimum wage

(2014)

SNAP expenditures (2012)Predicted change in SNAP expenditures

(millions of dollars)

Per resident

(millions of

dollars)

Total

(millions of

dollars)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 state

expenditures

(linear regression)

New Jersey $825 $160 $1420 -$1611 -$1325 -$605

New Mexico $750 $324 $675 -$513 -$444 -$445New York $800 $287 $5616 -$3766 -$3148 -$2801

North Carolina $725 $252 $2454 -$2187 -$1925 -$1833

North Dakota $725 $128 $90 -$162 -$143 -$67

Ohio $795 $259 $2995 -$2013 -$1688 -$1539

Oklahoma $725 $248 $945 -$799 -$703 -$706

Oregon $910 $322 $1255 -$272 -$213 -$262

Pennsylvania $725 $218 $2779 -$2930 -$2579 -$2075

Rhode Island $800 $280 $294 -$173 -$144 -$147

South Carolina $725 $291 $1373 -$1337 -$1177 -$1025South Dakota $725 $198 $165 -$192 -$169 -$123

Tennessee $725 $324 $2091 -$1413 -$1243 -$1562

Texas $725 $230 $5997 -$6402 -$5634 -$4479

Utah $725 $141 $402 -$614 -$541 -$300

Vermont $873 $230 $144 -$66 -$53 -$43

Virginia $725 $173 $1413 -$2062 -$1815 -$1056

Washington $932 $244 $1682 -$350 -$270 -$267

West Virginia $725 $273 $508 -$451 -$397 -$379

Wisconsin $725 $204 $1166 -$1302 -$1146 -$871Wyoming $725 $95 $55 -$105 -$93 -$41

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4043

Appendix C | wwwamericanprogresso

FIGURE 2

Raising the minimum wage to $1010would cut taxpayer costs in every state

Predicted decreases in cost and enrollment

in SNAP in 50 states

$200+$51ndash$100

$101ndash$200

0ndash$25

$26ndash$50

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4143

38 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

Endnotes

1 US Department of Agriculture ldquoSupplementalNutrition Assistance Program (SNAP)rdquo available athttpwwwfnsusdagovsnapeligibility (last accessedFebruary 2014)

2 For this initial analysis we do not consider Harkin-Millerrsquos increase in subminimum wages for tippedworkers To do s o would increase the estimated SNAP

savings by an unknown amount

3 The Congressional Budget Office estimates thatworkers currently earning between $1010 and $1150per hour would see their wages rise under the Harkin-Miller proposal Congressional Budget O ffice ldquoTheEffects of a Minimum Wage Increase on Employmentand Family Incomerdquo (2014)

4 Marianne Page Joanne Spetz and Jane Millar ldquoDoesthe Minimum Wage Affect Welfare Caseloadsrdquo Journalof Policy Analysis and Management 24 (2) (2005)273ndash295

5 Marianne Bitler and Hilary Hoynes ldquoThe More ThingsChange the More They Stay the Same The SafetyNet Living Arrangements and Poverty in the GreatRecessionrdquo NBER Working Paper 194 49 2013

6 Sylvia Allegretto and others ldquoFast Food Poverty Wages The Public Cost of Low-Wage Jobs in the Fast-FoodIndustryrdquo (Berkeley California Center for LaborResearch and Education 2013) available at httplaborcenterberkeleyedupubliccostsfast_food_poverty_wage

7 David Neumark and William Wascher ldquoDoes a HigherMinimum Wage Enhance the Effectiveness of theEarned Income Tax Creditrdquo Industrial and LaborRelations Review 64 (4) (2011) 712ndash746

8 David Lee and Emmanuel Saez ldquoOptimal MinimumWage Policy in Competitive Labor Marketsrdquo Journal ofPublic Economics 96 (9) (2012) 739ndash749

9 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family IncomerdquoFebruary 2014

10 Jesse Rothstein ldquoIs the EITC as Good as an NITConditional Cash Transfers and Tax Incidencerdquo AmericanEconomic Journal Economic Policy 2 (1) (2010) 177ndash208

11 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family Incomerdquo

12 Dube Arindrajit 2013 rdquoMinimum Wagesand the Distribution of Family IncomerdquoUnpublished working paper Available at httpsdldropboxusercontentcomu15038936Dube_MinimumWagesFamilyIncomespdf

13 As Dube explains in ldquoThe poverty of Minimum WageFactsrdquo the simulation approach underestimate stemsfrom a number of unwarranted assumptions includingthe range of actual wage increases and the accuracy ofwage data in the Current Population Survey The causal

approach does not make these assumptions

14 Allegretto Sylvia and others 2013 ldquoCredible ResearchDesigns for Minimum Wage Studiesrdquo Working Paper148-13 University of California Berkeley Institutefor Research on Labor and Employment Available athttpwwwirleberkeleyeduworkingpapers148-13pdf

15 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family In comerdquo

16 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo (2012)

17 Ibid

18 Ibid

19 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo

20 US Department of Agriculture ldquoSupplementalNutrition Assistance Program (SNAP)rdquo

21 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo

22 Ibid

23 We deviate from the Census Bureaursquos definition ofa family unit which is ldquotwo people or more (on e ofwhom is the householder) related by bir th marriageor adoption and residing togetherrdquo US Bureau ofthe Census ldquoCurrent Population Survey DefinitionsFamilyrdquo available at httpwwwcensusgovcpsabout

cpsdefhtml (last accessed February 2014) We countas a family unit any individual residing on his or herown two or more persons residing together whodo not belong to a family in the March CPS sampleare constructed as one family in our analysis For thepurposes of food stamp allocations the consumptionresulting from this transfer is probably distributed tofamily members (rather than household members ora single individual within the household) Howeversingle individuals canmdashand domdashreceive SNAPbenefits Excluding them would fail to make theanalysis reflective of the population at large

24 Strictly the family level linear probability modelpredicts the percentage-point decrease in theprobability that an individual family will receive SNAPpayments When applied to a large number of familieshowever we are able to interpret the coefficient asa decrease in the mean of enrollmentmdashthat is a

decrease in the enrollment ratemdashby applying the lawof iterated expectations

25 We generate expenditure predictions from theenrollment modelsmdashand conversely generateenrollment predictions from the expenditure modelmdashby assuming that expenditures per enrolled familyremains the same before and after the minimum wagechange In practice this is likely to be a conservativeestimatemdashthat is to underestimate the decrease inSNAP activity Average SNAP benefits per family willalso decrease as many families that remain eligible forSNAP experience income gains

26 Wage and Hour Division ldquoMinimum Wage Laws inthe States ndash Januar y 1 2014rdquo available at httpwwwdolgovwhdminwageamericahtm (last accessedFebruary 2014)

27 See for example Sylvia Allegretto and others ldquoCredibleResearch Designs for Minimum Wage Studiesrdquo WorkingPaper 148-113 (Berkeley California Institute forResearch on Labor and Employment 2013) available athttpwwwirleberkeleyeduworkingpapers148-13pdf

28 We will report these results in a forthcoming workingpaper

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4243

Endnotes | wwwamericanprogresso

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4343

The Center for American Progress is a nonpartisan research and educational institute

dedicated to promoting a strong just and free America that ensures opportunity

for all We believe that Americans are bound together by a common commitment to

these values and we aspire to ensure that our national policies reflect these values

We work to find progressive and pragmatic solutions to significant domestic and

international problems and develop policy proposals that foster a government that

is ldquoof the people by the people and for the peoplerdquo

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 343

1 Introduction and summary

5 Background

9 Methods and data

15 Results

19 Conclusion

20 About the authors and acknowledgments

21 References

23 Appendix A

25 Appendix B

33 Appendix C

38 Endnotes

Contents

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 443

Introduction and summary | wwwamericanprogress

Introduction and summary

How do minimum wage policy increases affec enrollmens and expendiures on

means-esed public assisance programs In his repor we address his quesion

or he case o he Supplemenal Nuriion Assisance Program or SNAP

ormerly known as he ood samp program

By definiion governmen spending on a means-esed program should decline

as average earnings increase insoar as benefi levels all wih increased earnings

and insoar as he earnings increase makes some individuals ineligible orany benefis Boh o hese condiions are saisfied in he case o he effec o

minimum wages on SNAP benefis SNAP benefis decline 30 cens or every $1

increase in amily earnings and phase ou enirely a abou he ederal povery

level1 Low-wage workers are disproporionaely enrolled in SNAP A minimum

wage increase ha lifs many amilies ou o povery should hereore reduce

public expendiure on his program

Bu he relaionship may be more complex I a minimum wage increase reduces

employmen hereby adding o he number o unemployed he number o SNAP

recipiens could increase SNAP recipiens who are unemployed disabled or

reired will no be affeced by a minimum wage increase Conversely i many

SNAP recipiens have earnings ha already bring hem close o becoming

ineligible or he program a minimum wage increase may have a very small effec

on SNAP expendiures Te quaniaive effec o minimum wages on SNAP

spending is no sel-eviden I requires a causal analysis

In an era o hisorically low real ederal minimum wage raes rising income

inequaliy job-marke sagnaion and conenious debae abou governmen defici

spending he possibiliy ha a higher minimum wage may lead o increased orreduced public spending has grea relevance o he public and o policymakers

Tis repor presens an iniial empirical analysis o he effecs o minimum wage

policy on SNAP paricipaion and expendiures We do so by exploiing more han

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 543

2 Center for American Progress | The Effects of M inim um Wages on SNAP Enro llments and Expenditures

wo decades o variaion in binding sae and ederal minimum wage changes in an

economeric ramework Our uure research will examine he effecs on SNAP

urher and apply an analogous ramework o wo oher public assisance spending

programs he Earned Income ax Credi and Medicaid

According o he finding in his repor a 10 percen increase in he minimum wagereduces SNAP enrollmen by beween 24 percen and 32 percen and reduces

program expendiures by an esimaed 19 percen aking ino accoun each

saersquos 2014 minimum wage level we apply hese resuls o he legislaive proposal

pu orward by Sen om Harkin (D-IA) and Rep George Miller (D-CA) o raise

he ederal minimum wage o $1010 per hour2 Our resuls imply ha he effecs

o he Harkin-Miller proposal on wage increases would reduce SNAP enrollmens

by beween 75 percen and 87 percen (31 million o 36 million persons) Te

oal anicipaed annual decrease in program expendiures is nearly $46 billion or

abou 6 percen o curren SNAP program expendiures

Harkin-Miller proposes o index minimum wage levels in subsequen years o he

consumer price index or CPI Te minimum wage would hen increase a he

same rae as SNAP benefi and eligibiliy levels which are also indexed o he CPI

Consequenly he savings over 10 years in 2014 dollars would be 10 imes he

one-year savings or a oal o approximaely $46 billion

Some o he reducion in SNAP program enrollmen and expendiures would

occur among workers making less han $1010 per hour991252hose whose pay would

be direcly increased by he minimum wage law Anoher par o he reducion

would occur among workers currenly earning beween $1010 and $1150 who

would also receive pay increases3

Alhough a large number o sudies have examined he impac o minimum

wage increases on earnings and employmen he impac o such minimum

wage policies on public assisance enrollmens and expendiures remains an

under-explored subjec in he economic lieraure Only a ew sudies discuss

he relaion beween he minimum wage and governmen ranser spending

much less atemp o ideniy he causal effec o one upon he oher Proessors

Marianne Paige Joanne Spez and Jane Millar find posiive effecs o minimum wage increases on welare caseloads as hey sae however heir resuls vary

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 643

Introduction and summary | wwwamericanprogress

considerably wih differen sample periods and assumpions abou sae rends4

Proessors Marianne Biler and Hilary Hoynes discuss he imporance o SNAP

as a saey ne program bu hey do no examine is relaion o minimum wage

policy5 Research economis Sylvia Allegreto and her Universiy o Caliornia a

Berkeley colleagues show ha low-wage workers in general and as-ood workers

in paricular are much more likely o be SNAP recipiens han all workers6

Several sudies have examined he relaionship beween he minimum wage and

he Earned Income ax Credi or EIC Proessor David Neumark and William

Wascher a researcher a he Federal Reserve Board o Governors find ha a

higher minimum wage increases EIC benefis or amilies in deep povery

while reducing EIC benefis or some sub-groups7 Proessors David Lee and

Emmanuel Saez argue ha he minimum wage and EIC are complemenary

policies no subsiues8 Te Congressional Budge Office or CBO argues ha

a minimum wage increase will no have a subsanial effec on EIC spending9

while Proessor Jesse Rohsein examines wheher he posiive effec o he EICon emale labor supply has lowered wages10 While hese sudies are o ineres

he EIC is quie differen rom SNAP in having a subsanial phase-in period in

which EIC benefis increase as well as a long phase-ou period wih complee

phase-ou a an annual income o abou $48000 or a amily o our quie a bi

above he reach o he minimum wage11

Research by Proessor Arindraji Dube on he causal effec o he minimum wage

on amily povery represens he sudy mos relaed o he one a hand12 Dube finds

ha Harkin-Miller would raise abou 46 million non-elderly Americans above

he ederal povery level or FPL In conras when CBO uses a simple simulaion

mehod o address he same quesion hey find ha Harkin-Miller would raise

900000 people above FPL13 Te difference beween hese wo esimaes highlighs

he imporance o underaking a causal analysis Te mehods used in his paper are

in many respecs similar o Dubersquos Moreover since eligibiliy and benefi levels or

programs such as SNAP and Medicaid are ied o he ederal povery level Dubersquos

findings have direc implicaions or his sudy Noneheless his repor appears o

be he firs sudy o examine he effecs o he minimum wage on SNAP In uure

work we plan o underake similar analyses or he EIC and Medicaid

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 743

4 Center for American Progress | The Effects of M inim um Wages on SNAP Enro llments and Expenditures

Te repor proceeds as ollows

bull Secion 1 provides background inormaion on he ederal minimum wage sae

minimum wages and he SNAP program

bullSecion 2 describes our mehods and daa

bull Secion 3 provides our main resuls including a simulaion o he effecs o a

Harkin-Miller minimum wage increase and a sae-by-sae analysis

bull Secion 4 presens our conclusions

Furher deails are provided in a series o appendices

Correction April 28 2014 Tis report incorrectly stated the potential reduction in

SNAP enrollment fom the Harkin-Miller proposal Te correct amount is 75 percentand 87 percent (31 million to 36 million persons) as stated in the reportrsquos tables

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 843

Background | wwwamericanprogress

Background

The federal minimum wage

Te ederal minimum wage was las increased in July o 2009

nearly five years ago During he pas wo decades many saes

have passed legislaion fixing he minimum wage a a higher

level han he ederal minimum Te maps in Figure 1 show ha

while saes in every region o he Unied Saes have adoped

higher minimum wages hey are no disribued randomly bygeography As shown in he 2013 sudy ldquoCredible Research

Designs or Minimum Wage Sudiesrdquo by economiss Sylvia

Allegreto Arindraji Dube Michael Reich and Ben Zipperer

hese saes vary sysemaically rom he oher saes by a number

o characerisics ha affec low-wage employmen rends bu

which are no hemselves relaed o minimum wage policy14

Te nonrandom patern o minimum wage adopion has

imporan implicaions or obaining unbiased esimaes o

minimum wages on employmen In paricular naional panel

sudies ha use sae and ime fixed effec models991252such as

a 1992 sudy by David Neumark and William Wascher991252

spuriously esimae negaive employmen effecs Te reason

or his resul is uncovered using ess or pre-rends Tese

ess find ha low-wage employmen was already declining wo

years beore minimum wages were implemened By making a

saisically large number o local comparisons ha conrol or

heerogeneiy among saes and by ime eliminaes his pre-

rend For his reason we conduc similar ess or our SNAPoucomes and use model specificaions ha include local

comparisons as in he sudy cied above

FIGURE 1

High versus low minimum wage stat

from 1990 to 2012

Means and variances

More than $533

Average minimum wage over 1990ndash2012

Less than or equal to $533

More than $121

Minimum wage variance over 1990ndash2012

Less than or equal to $121

Notes State means and variances were calculated using annual stat

minimum wage data from 1990 to 2012 The shading on the maps

partitions the states into above- and below-median values

Source Sylvia Allegretto and others ldquoCredible Research Designs forMinimum Wage Studiesrdquo Working Paper 148-13 (Institute for Resear

on Labor and Employment 2013) available at httpwwwirleberkeleyeduworkingpapers148-13pdf

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 943

6 Center for American Progress | The Effects of M inim um Wages on SNAP Enro llments and Expenditures

Te CBO recenly projeced ha in 2016 17 million workers will earn less han

he $1010 hourly wage proposed in he Harkin-Miller bill Furhermore he

CBO esimaes ha an addiional 8 million workers earned beween $1010 and

$1150 per hour and were also likely o experience a wage increase15

Supplemental Nutrition Assistance Program

We ocus our inquiry on SNAP Benefis under he program are enirely

ederally unded he program is adminisered by he US Deparmen o

Agriculure ogeher wih sae agencies which share in adminisraion

coss Spending on SNAP has grown in he pas decade reaching $78 bil lion

in 2011 wih SNAP enrollmen increasing o 45 million people abou one-

sevenh o he US populaion16 According o he CBO changes since 1990 in

SNAP spending and enrollmens are primarily he resul o cyclical economic

condiions noably changes in he unemploymen rae and changes in per capiaincome17 Te 2009 American Recovery and Reinvesmen Ac emporarily

increased SNAP benefi amouns by 136 percen as repored by he US

Deparmen o Agriculurersquos Food and Nuriion Service hese higher benefi

levels expired on November 1 201318 Te CBO esimaes ha abou wo-

hirds o he changes in SNAP expendiure are associaed wih changes in he

number o recipiens and one-hird wih changes in he benefis received when

recipiensrsquo incomes change19

In fiscal year 2014 SNAPrsquos maximum monhly benefis are $189 or a single

individual $497 or a amily o hree and $750 or a amily o five Benefis are

reduced by 30 cens per dollar received and phase ou enirely a gross monhly

household incomes o 130 percen o he ederal povery level $1245 or a

single individual $2116 or a amily o hree and $2987 or a amily o five

o deermine benefis SNAP also defines a ne monhly income concep and

ses benefis a 100 percen o he ederal povery level using his concep

Calculaion o ne monhly income can include cerain deducions rom

monhly gross income such as medical expenses and child care coss Alhough

saes are permited some laiude on wha deducions are allowed in pracice

hese vary by very small amouns Our saisical model akes accoun o sae-specific differences in benefis20

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1043

Background | wwwamericanprogress

SNAP imposes an employmen or work-raining requiremen or able-bodied

prime-age aduls991252hose beween he ages o 18 and 50 and wihou disabiliies

or dependen children Such households can receive only hree monhs o benefis

in a hree-year period In recen years abou 85 percen o households receiving

benefis have incomes below he ederal povery level 49 percen have dependen

children 16 percen are age 60 or older 20 percen are disabled and 30 percenrepor some earned income21

A 2012 CBO repor also noes ha ake-up raes among eligible SNAP recipiens

average abou 70 percen wih much lower ake-up among elderly households

Te ake-up rae increases in harder economic imes I also increased when sigma

issues were reduced as SNAP debi cards replaced acual ood samps ake-up

is especially high among hose mos needy Adminisraive spending equaled 91

percen o he poenial spending ha would have occurred i all eligible recipiens

were enrolled Alhough some SNAP policy changes have occurred since 1990

mos were relaively minor and all were naional in scope Te 1996 welare reorm bill eliminaed SNAP eligibiliy or some legal immigrans limied he ime lengh

o eligibiliy or able-bodied childless aduls and reduced maximum benefis

Some o hese resricions were relaxed in 2002 and again in he American

Recovery and Reinvesmen Ac in 200922

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1143

8 Center for American Progress | The Effects of M inim um Wages on SNAP Enro llments and Expenditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1243

Methods and data | wwwamericanprogress

Methods and data

As previously menioned we exploi variaion in minimum wages by sae and ime

o examine heir causal effecs on SNAP enrollmens and expendiures o do so we

merge daa rom 1990 hrough 2012 drawn rom he Annual Social and Economic

Supplemen o he Curren Populaion Survey991252an annual Census Bureau survey

commonly known as he March CPS ha includes23 inormaion on SNAP

enrollmens a he amily level991252wih sae-level daa on minimum wages SNAP

expendiures populaion unemploymen raes and sae median income levels

o conrol or ime-varying heerogeneiy among saes our specificaions includeconrols or sae linear rends and effecs by Census division and ime We esimae

effecs a wo levels allowing or amily variaion and allowing only or sae-level

variaion We also employ a se o sandard demographic conrols such as amily size

and composiion and race and ehnic composiion

Distinguishing causation from correlation

How can we ensure ha our analysis does no pick up a spurious correlaion or

example he endency o more economically vibran saes o implemen higher

minimum wages Disinguishing correlaion and policy endogeneiy rom rue

causal effecs is he primary moivaion or economeric analysis In he ideal

experimen researchers would begin wih wo saes991252ha are alike in every

respec prior o he policy991252 and ldquoreardquo only one o hese saes wih a higher

minimum wage Tey would atemp o shield hese saes rom any influence ha

could obscure heir undersanding o he minimum wagersquos direc effec on SNAP

enrollmen Researchers o course canno conduc such experimens

We can however use saisical mehods o conrol simulaneously heindependen effecs on SNAP o sae unemploymen raes sae income levels

and common rajecories among saes wihin he same Census division By

ensuring similariy along hese dimensions we maximize he likelihood ha

SNAP aciviy in wo saes would have comparable oucomes in he absence

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1343

10 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

o a minimum wage policy change Tus i a new minimum wage policy were

implemened in one sae only he researchers could atribue all o he difference

hey observe in SNAP aciviy o he new minimum wage policy

In oher words we approximae he ideal experimen by using non-experimenal

saisical mehods Te desirable ldquopre-exising similariiesrdquo beween saes ha wehave defined above inorm our choice o conrol variables in a saisical seting

More precisely in our muliple regression models we use median amily income

he unemploymen rae he employmen-o-populaion raio and regional and

ime idenifiers o consruc an appropriae group o peers or each sae on he

eve o a policy change

Data description

wo daa ses include inormaion abou boh income and paricipaion inpublic programs Te Survey o Income and Program Paricipaion or SIPP

which is conduced in inermiten years has he advanage o ollowing he

same individuals over a period o ime In oher words i is a longiudinal daa

se I also has he advanage o conaining monhly daa However he sample

size o he SIPP is no sufficien or analyzing variaions in sae-level minimum

wages Te March CPS has he advanage o a much larger sample size and i is

conduced annually wihou any breaks in ime I has he disadvanage o being

a cross-secional daa se so we canno ollow he same individuals over ime991252

sricly speaking over more han one year On ne he March CPS is much more

suiable or our sudy We examine he empirical relaionship beween minimum

wage policy and ood samp aciviy a wo levels o aggregaion he amily level

and he sae level Family-level daa are drawn rom he March CPS

Te March CPS comprises responses rom he residens o 50000 o 60000

dwelling places surveyed per year and conains deailed inormaion on he

residensrsquo employmen and income including income rom ranser paymens

Te sample or our analysis comprises more han 128 million amily unis during

he period rom 1990 o 2012 (inclusive) Survey weighs allow us o analyze

SNAP paricipaion in a manner ha is represenaive o he US populaion alarge Over all years he share o amilies reporing ood samp receip in he

weighed March CPS sample is 91 percen Te enrollmen rae was a a low o 6

percen in he year 2000 In 2012 he mos recen year in our panel 133 percen

o amilies repored paricipaing in SNAP a some poin during he survey year

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1443

Methods and data | wwwamericanprogresso

Te March CPS also collecs inormaion on he number o SNAP recipiens in

he household he number o monhs he household receives SNAP benefis

and he cash-equivalen value o he SNAP benefis received However he

value o SNAP benefis is severely underrepored among recipiens perhaps

because recipiens are unaware o he exac monhly cash-equivalen value o

benefis hey receive

Our firs empirical sraegy ocuses solely on SNAP enrollmen By using he

amily as he uni o analysis we are able o inser saisical conrols o accoun

or non-wage-relaed acors ha influence any paricular amilyrsquos likelihood o

program paricipaion wih he inenion o isolaing any differences in program

paricipaion ha are due purely o changes in wage policy Tis approach

idenifies he effecs o low-wage labor policy on he exernal margin991252ha is

he effec o he minimum wage on he likelihood ha a amily paricipaes in he

SNAP program a all991252as opposed o he inernal margin or how much SNAP

unding he amily would receive

Our second empirical ramework uses sae-level adminisraive daa Ta is

we aggregae he daa o obain a single daa poin or each saeyear back

o 1990 represening he mean o he oucome or he sae Te sae-level

esimaion serves as a robusness check on he amily-level resuls or SNAP

paricipaion Also using aggregaed daa allows us o esimae direcly he

causal effec o minimum wage changes on SNAP spending Tis is no possible

a he amily level as discussed above daa on cash-equivalen value o ood

samps or SNAP recipiens is very requenly no repored in he March CPS

and when i is repored he inormaion may be unreliable By conras he

Bureau o Economic Analysis publishes aggregae SNAP spending a he sae

level in is Naional Income and Produc Accoun or NIPA ables Tus

while we are unable o observe he heerogeneiy in he cash value o SNAP

or amilies in each sae we are able o calculae average SNAP spending

per residen in each sae per year Supporing covariaes include he annual

unemploymen and employmen daa rom he Bureau o Labor Saisics or

BLS and sae-level populaion series rom he iner-decennial census releases

Minimum wage daa are available rom he BLSrsquos wages and hours division For

sae minimum wage changes enaced a oher imes han he firs o he yearan average value or he year is used

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1543

12 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

Family level model specifications

We firs examine he effec o he minimum wage on paricipaion in public

assisance programs For amily residing in sae and during year we esimae

an equaion o he ollowing orm

(1)

is a binary variable ha is se equal o 1 i a leas one member o amily

received ood samps during he survey year is a se o sae-level

characerisics including annual averages o he unemploymen rae he

employmen-o-populaion raio and he naural log o median amily income

is a vecor o amily atribues including indicaors or he race and marial saus

o he amily head size o he amily he presence o children and he presence oan adul male Sae fixed effecs are capured by o conrol or ime-varying

heerogeneiy our preerred model specificaion also includes year fixed effecs

ha vary by Census division ( ) and sae-level linear ime rends In

Appendix B we jusiy he inclusion o hese las wo erms We also compare he

resuls rom our preerred specificaion wih less sauraed specificaions

Te effec o ineres which is capured by is he expeced change in he

probabiliy o receiving SNAP benefis wih respec o a change in he (log o he)

binding minimum wage in sae during year We repor robus sandard errors

clusered a he sae level We esimae he parameers using linear regression

producing a linear probabiliy model Deails o he model selecion process are

covered in Appendix B below

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1643

Methods and data | wwwamericanprogresso

State-level model specifications

Te sae-level models are similarly specified For sae in year we assume ha

(2)

In his model is now eiher he SNAP enrollmen rae in sae during year

or he naural logarihm o per capia SNAP expendiures in sae during

year is once again a se o sae-level characerisics including he same

sae-level covariaes as in he amily regressions (annual average unemploymen

rae employmen-o-populaion raio naural log o median amily income)

wih he addiion o amily level characerisics averaged across he sae (average

amily size and he shares o populaion consiued by each o five racialehnicgroups) Sae fixed effecs are represened by As above our preerred model

specificaion includes year fixed effecs ha vary by Census division ( ) and

sae-level linear ime rends as elaboraed in Appendix B Te effec o

ineres is capured by

We esimae boh sae-level models (enrollmen and expendiures) using

ordinary leas squares regression Tus he inerpreaion o he coefficien is no

longer ha o a change in probabiliy as in he binary oucome models described

above Raher or he sae-level SNAP enrollmen model represens he

expeced change (in percenage poins) in he saersquos SNAP enrollmen rae ha

is due o a 1 percen change in he minimum wage For he SNAP expendiures

model is simply he elasiciy o SNAP spending wih respec o he minimum

wage991252ha is he percenage change in sae expendiures expeced o resul rom

a 1 percen change in ha saersquos minimum wage For urher deails on model

specificaion reer o Appendix B below

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1743

14 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1843

Results | wwwamericanprogresso

Results

Estimated minimum wage effects on SNAP enrollment and

expenditures

able 1 shows he esimaed parameer o ineres991252he coefficien o he

minimum wage or he preerred model o each ype Coefficiens on he

minimum wage variable are no direcly comparable across models because all

our models have a differen uncional orm o undersand and compare hese

esimaes we compue he change in SNAP aciviy prediced or a paricular wage scenario Te final column in able 1 answers he quesion Wha would be

he expeced change in SNAP aciviy in response o a 10 percen increase in he

minimum wage Te answer o his quesion varies wih he value o he inpu

parameers in he able we calculae he percenage decrease in enrollmen or

expendiures prediced or he average sae wih a minimum wage o $725 in

2014 Te sae-level SNAP expendiure model which is a consan-elasiciy

model conveys elasiciy inormaion direcly or he change in expendiures per

capia in he sae

TABLE 1

Comparison of national SNAP predictions for a 10 percent increase in the federal minimum wage

Model LevelRegression

type

Predicted outcome Coefficient of log

(minimum wage)

(Standard error)

Effect of a 10 percent in

in the minimum wage

VariableForm of

variable

Total

enrollment

Tot

expend

1 Family Linear

probability Enrollment Binary (enrolled=1)

-0042

(0008) -317 N

2

StateLinear regression

(ordinary least

squares)

Enrollment State enrollment rate (percent) -0031

(0012) -235 N

3 Expenditures Log (state expenditures per capita) -0190

(0103) NA -19

plt01 plt005 plt001Note Predicted changes are calculated for the average state with a minimum wage of $725 in 2014

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1943

16 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

According o his model a 10 percen increase in he minimum wage would resul

in a 19 percen reducion in SNAP expendiures Tis resul is reassuringly similar

o Dubersquos resul or povery reducion Te wo enrollmen models are much more

precisely esimaed han he expendiure model Te sae-level enrollmen model

finds ha a 10 percen minimum wage increase in a low-wage sae is associaed wih

a 235 percen decrease in SNAP enrollmens Te amily-level linear probabiliymodel predics a somewha greaer elasiciy or low minimum wage saes an

increase o 10 percen in he ederal minimum would resul in a 317 percen decline

in SNAP enrollmen24 Te differences in hese esimaes sem rom a number o

acors including difference in model uncional orm and daa used We rea his

range o elasiciy esimaes as an upper and lower bound on enrollmen impacs

Harkin-Miller bill National and state-level predicted impacts

Wha would be he prediced change or he SNAP program i he ederalminimum were raised o $1010 as proposed in he Harkin-Miller bill In order o

make his inerence we accoun or he ac ha no all saes are currenly subjec

o he ederal minimum wage a he beginning o 2014 21 saes mainained

higher minimum wages han $725 In hose saes an increase in he ederal

minimum wage may or may no be binding or employers in he sae depending

upon wheher he new ederal minimum exceeds he sae-level minimum Bu

regardless o wheher a minimum wage change is binding he impac on SNAP

aciviy will be lower in high minimum wage saes In order o accoun or his

properly we calculae sae by sae he percenage wage change ha would resul

rom he Harkin-Miller proposal and apply he parameers rom each o he hree

models above o compue he expeced decrease in SNAP aciviy or each sae

In his exercise we use saesrsquo curren (2014) minimum wage levels and assume

as a baseline he 2012 levels o SNAP enrollmen and expendiure as 2012 is he

mos recen year or which SNAP daa are available

able 10 and able 11 in Appendix C repor he esimaed effecs on SNAP

enrollmen and expendiures respecively or each sae under he Harkin-Miller

bill25 An increase o $1010 i enaced oday would represen beween a 393

percen wage increase in a $725 minimum wage sae and an 84 percen increasein Washingon sae which has he highes minimum wage in he naion a $932

as o January 201426 Slighly more han 56 percen o he decrease in expendiures

and abou 59 percen o he decrease in enrollmen would occur in saes wih

presen-day minimum wages o $725

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2043

Results | wwwamericanprogresso

In 2012 hese saes were home o 46 percen o he American populaion

and accouned or a roughly equivalen percenage o oal naional SNAP

expendiures No surprisingly he larges enrollmen decreases come rom

populous saes wih very high SNAP enrollmen raes andor very low minimum

wages Te larges prediced enrollmen reducion991252beween 319000 individuals

and 362000 individuals991252would occur in exas which has a $725 minimum wage In Caliornia which has a 206 percen SNAP paricipaion rae and an $8

per hour minimum wage we anicipae SNAP enrollmen declines o 310000

persons o 371000 persons And in Florida which had a SNAP paricipaion rae

o 166 percen and a $793 minimum wage enrollmen could decline by beween

164000 individuals and 196000 individuals For he our saes ha ogeher

accouned or he greaes amoun o SNAP spending in 2012991252exas Caliornia

Florida and New York respecively991252he combined expendiure reducion rom

he Harkin-Miller bill is prediced o be $14 billion

able 2 summarizes he prediced declines in SNAP aciviy or he naion as a whole ha would resul rom he direc and indirec effecs o he Harkin-Miller

bill Enrollmen would all beween 31 million persons and 36 million persons

represening 75 percen o 87 percen o curren enrollmen Te anicipaed

reducion in program expendiures would be nearly $46 billion or 61 percen o

program expendiures

TABLE 2

Comparison of national SNAP predictions under the Harkin-Miller billrsquos $1010 minimum wage

Model

Enrollment

(persons)

Expenditures

(millions of dollars)

Current (2012) Predicted Change Current (2012) Predicted Chan

Family enrollment (linear probability)

41866195

45489339 -3623144

$74861

NA NA

State enrollment (ordinary least squares) 38745435 -3120759 NA NA

State expenditures (ordinary least squares) NA NA $70305 -$45

Note Calculations use 2014 state minimum wages and the most recent SNAP data from 2012 They assume that per-enrollee expenditures remain constant

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2143

18 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

Tere are o course oher possibiliies or minimum wage legislaion able 3

shows he expeced SNAP changes or he Unied Saes under a variey o wage

scenarios calculaed using he sae-level models I saes were no able o se

heir minimum wages independenly such ha all saes were consrained by

he ederal minimum o $725 SNAP would be received by abou 514000 more

people across he Unied Saes a an addiional program cos o nearly hree-quarers o a billion dollars In conras he effecs o a higher minimum wage

proposal991252a ederal wage floor o $11 per hour991252would decrease enrollmen in

SNAP by more han 10 percen and decrease program coss by 83 percen

TABLE 3

Summary of par ticipation and expenditures under wage scenarios

If all states had

minimum wages of

Enrollment(persons)

Expenditures

(millions of dollars)

Predicted Change Predicted Change

Recent levels (2014) 41866195 $74861

$725 42380520 514326 $75604 $743

$800 41423919 -442276 $74209 -$652

$900 40148451 -1717744 $72350 -$2511

$1000 38872982 -2993212 $70490 -$4371

$1010 38745435 -3120759 $70305 -$4556

$1100 37597514 -4268681 $68631 -$6230

Note Calculations use state-level enrollment model coefficient

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2243

Conclusion | wwwamericanprogresso

Conclusion

An exensive body o lieraure examines employmen effecs o he minimum

wage A much smaller se o sudies examines how he minimum wage affecs

povery and only a handul o papers examine he effecs o he minimum wage

on he EIC Our analysis is he firs o examine he effecs o he minimum

wage on SNAP

Our findings indicae ha increased earnings rom minimum wage changes do

reduce SNAP enrollmens and expendiures We esimae ha he Harkin-Miller bill would save axpayers nearly $46 billion per year equivalen o 61 percen

o SNAP expendiures in 2012 he las year or which daa are available Over a

10-year period he esimaed savings amoun o nearly $46 billion

Our repor is subjec o limiaions ha we expec o overcome in our uure

research Firs he findings do no ake ino accoun possible ineracions among

SNAP he EIC and Medicaid Te eligibiliy cuoffs among hese programs

are quie differen suggesing ha such ineracions may be minor Noneheless

he join effecs can only be deermined by urher research using a causal

model Second i would be useul o know he disribuion o SNAP reducions

along he wage disribuion Using he Congressional Budge Officersquos calculaions

o how much he oal dollar value o wage would increase under he Harkin-

Miller proposal our findings imply ha he decline in overall SNAP spending

equals abou 15 percen o he oal resuling increase in wages Te amoun and

disribuion o his offse are o considerable ineres Minimum wage beneficiaries

who come rom working amilies already well above he povery line would no

see any offse while hose who are currenly considerably below he povery line

will see larger offses Tese issues will also be a subjec or our uure research

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2343

20 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

About the authors

Rachel West is a maser o public policy candidae a he Goldman School

o Public Policy Universiy o Caliornia Berkeley Her research ocuses on

economic policy in he areas o low-wage labor and povery

Michael Reich is proessor o economics and direcor o he Insiue or

Research on Labor and Employmen a he Universiy o Caliornia a Berkeley

His research publicaions cover numerous areas o labor economics including

racial inequaliy labor marke segmenaion high-perormance workplaces

union-managemen cooperaion Japanese labor-managemen sysems living

wages and minimum wages He received his docorae in economics rom

Harvard Universiy

Acknowledgments

We are graeul o Sylvia Allegreto Arindraji Dube Bill Leser Jesse Rohsein

Daniel Tompson and Ben Zipperer or heir valuable suggesions

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2443

References | wwwamericanprogresso

References

Allegreto Sylvia and ohers 2013 ldquoFas Food Povery Wages Te Public Cos o Low-Wage Jobsin he Fas-Food Indusryrdquo Universiy o Caliornia Berkeley Cener or Labor Research andEducaion Available a htplaborcenerberkeleyedupubliccossas_ood_povery_wage

Allegreto Sylvia and ohers 2013 ldquoCredible Research Designs or Minimum Wage Sudiesrdquo

Working Paper 148-13 Universiy o Caliornia Berkeley Insiue or Research on Labor andEmploymen Available a htpwwwirleberkeleyeduworkingpapers148-13pd

Biler Marianne and Hilary Hoynes 2013 ldquo Te More Tings Change he More Tey Say heSame Te Saey Ne Living Arrangemens and Povery in he Grea Recessionrdquo Working Paper19449 Naional Bureau o Economic Research

Congressional Budge Office 2012 ldquoTe Supplemenal Nuriion Assisance Programrdquo Washingon Available a htpwwwcbogovsiesdeaulfilescbofilesatachmens04-19-SNAPpd

991252 991252 991252 2014 ldquoTe Effec o a Minimum-Wage Increase on Employmen and Family Incomerdquo Washingon Available a htpwwwcbogovsiesdeaulfilescbofilesatachmens44995-MinimumWagepd

Dube Arindraji 2013 rdquoMinimum Wages and he Disribuion o Family Incomerdquo Unpublished working paper Available a htpsdldropboxuserconencomu15038936Dube_ MinimumWagesFamilyIncomespd

991252 991252 991252 2014 ldquoTe Povery o Minimum Wage Facsrdquo Arindraji Dube Blog January 22 Availablea htparindubecom20140122he-povery-o-minimum-wage-acs

Lee David and Emmanuel Saez 2012 ldquoOpimal Minimum Wage Policy in Compeiive LaborMarkesrdquo Journal o Public Economics 96 (9) 739ndash749

Neumark David and William Wascher 1992 ldquoEmploymen Effecs o Minimum and Subminimum Wages Panel Daa on Sae Minimum Wage Lawsrdquo Industrial and Labor Relations Review 46 (1)

55ndash81

Neumark David and William Wascher 2011 ldquoDoes a Higher Minimum Wage Enhance heEffeciveness o he Earned Income ax Credirdquo Industrial and Labor Relations Review 64 (4)712ndash746

Page Marianne Joanne Spez and Jane Millar 2005 ldquoDoes he Minimum Wage Affec WelareCaseloadsrdquo Journal o Policy Analysis and Management 24 (2) 273ndash295

Rohsein Jesse 2010 ldquoIs he EIC as Good as an NI Condiional Cash ransers and ax

Incidencerdquo American Economic Journal Economic Policy 2 (1) 177ndash208

Wage and Hours Division 2014 ldquoMinimum Wage Laws in he Saes ndash January 1 2014rdquo USDeparmen o Labor (htpwwwdolgovwhdminwageamericahm [February 2014])

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2543

22 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2643

Appendix A | wwwamericanprogresso

Appendix A

Pre-trend falsification check

Recen minimum wage research27 highlighs a common flaw in previous sudies

ailure o veriy ha he oucome variable is ree o negaive pre-exising rends I

or example SNAP aciviy was already rending down in saes ha raised heir

minimum wages beore hese changes came ino effec our regression analysis could

(misakenly) atribue ha reducion o he minimum wage We check or such

pre-rends by inroducing variables ha represen he prior yearrsquos value or leads o

he minimum wage I he model esimaes he minimum wage o have an effec on

he oucome variable beore he wage change wen ino effec hen an unobservedacor no he minimum wage change caused he change in SNAP aciviy

We es he specificaions above or pre-rends by including a one-year lead in

all hree specificaions We find ha he lead erms are small posiive and no

saisically significan indicaing ha he concurren minimum wage991252no

he wage level in prior periods991252is driving he observed changes in SNAP

oucomes28 In paricular he coefficien (sandard error) on he lead erm in

our preerred amily-level enrollmen regression is 011 and no significan

while he coefficien and sandard error o he conemporaneous minimum

wage is unchanged In he sae-level preerred enrollmen regression he

coefficien o he lead erm is again small (07) and i is no significan Te

corresponding coefficien on he lead erm in he sae-level expendiure

regression is 16 and is no significan Te posiive poin esimaes on hese lead

erms resuls no only rule ou disoring negaive pre-rends Tey also sugges

ha our main resuls may underesimae he rue effecs

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2743

24 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2843

Appendix B | wwwamericanprogresso

Appendix B

Model selection process

For boh he amily-level and sae-level models we es hree mehods o conrol

or unobserved geographic- and ime-varying characerisics as suggesed by he

minimum wage lieraure Firs we include only independen sae-specific fixed

effecs and year-specific fixed effecs Tis specificaion (specificaion 1) implicily

assumes ha amilies in any sae consiue an equally good saisical ldquoconrolrdquo

group or hose in any randomly chosen sae afer accouning or various

characerisics (median income and unemploymen rae among ohers) Similarly

simple ime fixed effecs assume ha amilies surveyed in any year can crediblyserve as a conrol group or amilies surveyed in every oher year o he sample

(1990 hrough 2012)

In oher words specificaion 1 assumes ha a saersquos immediae neighbor provides

no beter a couneracual or he effec o a minimum wage change han does a

sae across he counry We relax his resricive specificaion sequenially in wo

seps In specificaion 2 we replace simple year fixed effecs wih fixed effecs or

each Census divisionyear (capured as an addiional variable in he vecor By

using division-year effecs we remove he resricion ha amilies in each sae

are equally good saisical conrols or all oher amilies Raher we allow or he

possibiliy ha amilies in similar geographic regions (or example he Souh or

he Norheas) may be more similar o one anoher han amilies arher away

Finally in specificaion 3 we add sae-specific linear ime rends o he previous

specificaion Tus specificaion 3 is he mos rigorous model specificaion in ha

i allows or heerogeneiy along hree dimensions Ta is specificaion 3 allows

each sae o have is own ime-varying rends raher han imposing he resricion

ha saes evolve idenically over he 22 years in he sample

We begin building he heoreical specificaion above rom a se o simpleuncondiional models regression o SNAP aciviy (enrollmen or expendiures)

on he log o he minimum wage and a se o geographic- and ime-specific

effecs (specificaions 1 2 and 3 described above) As shown in ables 1ndash3 (or

specificaion 3) we hen add covariaes sequenially o hese models including

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2943

26 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

firs he vecor o amily-level conrols ollowed by each o several sae-level

covariaes in urn he unemploymen rae log o median-amily income and he

employmen-o-populaion raio Comparable resuls or specificaions 2 and 3

will be available in our orhcoming working paper

As expeced he simple uncondiional models indicae ha he relaionship beween he minimum wage and SNAP enrollmen i one exiss is a more complex

one influenced by oher acors In he uncondiional model he coefficien on he

variable o ineres991252he log o he minimum wage991252is small in magniude and no

saisically differen rom zero Once we accoun or he influence o labor marke

condiions and variaion in income levels on program paricipaion (by including

unemploymen rae and median-amily income conrol variables respecively)

he effec o he minimum wage on SNAP enrollmen is precisely esimaed Te

coefficien o he log minimum wage is slighly higher (-0042) in he amily-level

analysis han he coefficien (-031) in he sae-level analysis Te level o precision

is also higher in he amily-level analysis Tis is o be expeced when using 124million observaions compared o 1127

Te second dimension o model choice concerns he effec specificaion ables

7ndash9 compare he primary coefficiens o ineres or he SNAP enrollmen and

expendiure models For boh he enrollmen models he effec sizes are smalles

or specificaion 1 larges or specificaion 2 and inermediae beween hese wo

in specificaion 3 Recall ha Specificaion 3 conains sae-specific linear ime

rends in addiion o he census divisionyear conrols included in specificaion

2 In he amily-level enrollmen model he sandard error o he minimum wage

coefficien is smaller han in he oher wo specificaions Sandard errors on he

oher variables are much smaller in specificaions 2 and 3 han in specificaion

1 On he basis o coefficien significance (join and individual) specificaions 2

and 3 are sricly preerred in boh enrollmen models o specificaion 1 which

conains only sae and year fixed effecs

A concern wih specificaions 2 and 3 is ha rend conrols such as sae linear

rends may incorrecly absorb some o he delayed impac o a minimum wage

When we es his issue by including lagged minimum wages we do no find ha

delayed effecs are significan Anoher concern is ha more sauraed modelsuse less o he saisical variaion which could reduce he saisical power o

he resuls However he sandard errors or our more sauraed models are no

higher and are lower in some cases han or he less sauraed models Overall

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3043

Appendix B | wwwamericanprogresso

his evidence suppors our use o he sauraed model as he preerred model

specificaion Moreover Dubersquos 2013 sudy shows ha more sauraed models

perorm beter han models wih jus sae and ime fixed effecs

Te esimaed enrollmen regressions a boh he amily and sae levels show large

and saisically significan coefficiens Te esimaed minimum wage effec in heexpendiures regressions991252or which we have only sae-level daa991252is also large

and saisically significan

We do no use weighed regression or he sae-level models preerring o keep

analysis o he ldquoreamenrdquo (ha is o say a minimum wage change) appropriae

o he average sae raher han he average amily or individual I insead our

primary ineres were he impac o a minimum wage change on he average amily

or he average individual we migh choose o designae he number o amilies

in each sae or he sae populaion respecively as analyic weighs in order o

obain a coefficien beter suied or such inerence

TABLE 4

SNAP enrollment

Family-level linear probability

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0048 -0047 -0040 -0043 -0042

(0013) (0013) (001) (0008) (0009)

Unemployment rate (100) 0505 0420 0280

(0083) (0086) (0082)

Log median income -0057 -0039

(0011) (0011)

Employment-to-population ratio -0239

(0038)

N 1242022 1242022 1242022 1242022 1242022

plt01 plt005 plt001

Note Robust standard errors are in parentheses Observations are clustered at the state level The outcome variable is binary and equal to one if a family is enrolledin SNAP All models include state fixed effects Census division x-year fixed effects and state-specific linear time trends All specifications except 3a include additional

controls for family size race and marital status of the family head presence of children and presence of an adult maleSource Annual data from the March Current Population Survey for 1990 to 2012 estimation includes Current Population Survey probability weights

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3143

28 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

TABLE 5

SNAP enrollment

State-level linear regression

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0028 -0024 -0026 -0031 -0031

(0014) (0013) (0013) (0012) (0012)

Unemployment rate (100) 0625 0466 0320 0339

(0087) (0088) (0085) (0083)

Log median income -0090 -0065 -0061

(0013) (0013) (0013)

Employment-to-population ratio -0282 -0248

(0037) (0038)

N 1127 1127 1127 1127 1127

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the SNAP enrollment rate All models include state fixed effects Census division x-year fixedeffects and state-specific linear time trends

Source Annual data from the March Current Population Survey for 1990 to 2012

TABLE 6

SNAP expenditures

State-level linear regression

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0181 -0149 -0156 -0153 -0190

(011) (0103) (0102) (0103) (0103)

Unemployment rate (100) 4587 4152 4232 4313

(0622) (0621) (0633) (0628)

Log median income -0246 -0261 -0294

(0075) (0078) (0078)

Employment-to-population ratio 0155 0244

(0237) (024)

N 1127 1127 1127 1127 1127

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the natural log of state SNAP expenditures per capita for 1990 to 2012 All models include state

fixed effects Census division x-year fixed effects and state-specific linear time trends

Source Annual data from Bureau of Economic Analysis national income and product accounts tables for 1990 to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3243

Appendix B | wwwamericanprogresso

TABLE 7

Comparison of specifications SNAP enrollment

Family-level linear probability

(1) (2) (3)

Log minimum wage -0028 -0049 -0042

(0014) (0017) (0009)

Unemployment rate (100) 0275 0297 0280

(0161) (0076) (0082)

Log median income -0077 -0055 -0039

(0014) (0012) (0011)

Employment-to-population ratio -0238 -0250 -0239

(0054) (004) (0038)

N 1242022 1242022 1242022

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses Observations are clustered at the state level The outcomevariable is binary or equal to one if a family is enrolled in SNAP All specifications include additional controls forfamily size race and marital status of the family head presence of children and presence of an adult male

Source Annual data from the March Current Population Survey for 1990 to 2012 estimation includes CurrentPopulation Survey probability weights

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3343

30 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

TABLE 8

Comparison of specifications SNAP enrollment

State-level linear regression

(1) (2) (3)

Log minimum wage -0019 -0035 -0031

(0009) (0012) (0012)

Unemployment rate (100) 0401 0370 0339

(0063) (0077) (0083)

Log median income -0081 -0073 -0061

(0011) (0013) (0013)

Employment-to-population ratio -0183 -0222 -0248

(0039) (0039) (0038)

N 1127 1127 1127

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the SNAP enrollment rate Allregressions include the share of households with children and the racial shares of the population

Source Annual data from the March Current Population Survey for 1990 to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3443

Appendix B | wwwamericanprogresso

TABLE 9

Comparison of specifications SNAP expenditures

State-level linear regression

(1) (2) (3)

Log minimum wage -0121 -0203 -0190

(0075) (0103) (0103)

Unemployment rate (100) 5292 5152 4313

(0464) (0576) (0628)

Log median income -0437 -0417 -0294

(008) (0086) (0078)

Employment-to-population ratio -0040 -0220 0244

(0261) (0260) (0240)

N 1127 1127 1127

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the natural log of state SNAPexpenditures per capita All regressions include the share of households with children and the racial shares ofthe state population

Source Annual data from Bureau of Economic Analysis national income and product accounts tables for 1990to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3543

32 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3643

Appendix C | wwwamericanprogresso

Appendix C Harkin-Miller

policy simulation results

StateMinimum wage

(2014)

SNAP enrollment (2012) Predicted change in SNAP enrollment

RateTotal

(persons)

Model 1 family

enrollment (linearprobability)

Model 2 state

enrollment (linearregression)

Model 3 s

expenditu(linear regre

Alabama $725 164 788682 -66922 -58897 -58906

Alaska $775 120 87436 -8104 -6887 -3288

Arizona $790 201 1319323 -67435 -56738 -64356

Arkansas $725 78 230489 -40977 -36063 -35248

California $800 206 7813680 -371131 -310222 -18223

Colorado $800 164 853155 -50684 -42365 -23926

Connecticut $870 91 326621 -22456 -17975 -13711

Delaware $725 186 170262 -12739 -11211 -10647

District of Columbia $825 133 84009 -5370 -4417 -3632

Florida $793 166 3208026 -195813 -164426 -13046

Georgia $725 160 1586336 -137741 -121224 -11004

Hawaii $725 96 133662 -19310 -16995 -14933

Idaho $725 92 147501 -22165 -19507 -15809

Illinois $825 95 1225084 -109088 -89742 -70955

Indiana $725 125 816233 -90818 -79928 -83985

Iowa $725 155 478011 -42716 -37594 -28556

Kansas $725 135 388269 -40082 -35275 -27461

Kentucky $725 130 568821 -60840 -53544 -52259

Louisiana $725 149 683832 -63929 -56263 -66083

Maine $750 77 101976 -16567 -14323 -15234

Maryland $725 144 846415 -81748 -71946 -38370

Massachusetts $800 130 864721 -64902 -54251 -42913

Michigan $740 146 1439141 -128801 -112140 -11022

Minnesota $725 133 713646 -74730 -65769 -37878

TABLE 10

SNAP enrollments Predicted changes by state under the Harkin-Miller bill

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3743

34 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

StateMinimum wage

(2014)

SNAP enrollment (2012) Predicted change in SNAP enrollment

RateTotal

(persons)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 state

expenditures

(linear regression)

Mississippi $725 129 386501 -41486 -36511 -46467

Missouri $750 172 1036182 -75131 -64952 -56244

Montana $790 132 132452 -10350 -8708 -5846

Nebraska $725 124 230683 -25773 -22683 -12189

Nevada $825 162 446035 -23349 -19209 -11894

New Hampshire $725 127 168404 -18359 -16157 -5735

New Jersey $825 160 1416666 -75175 -61843 -28236

New Mexico $750 149 310896 -25983 -22463 -22512

New York $800 192 3763553 -191193 -159815 -142182

North Carolina $725 174 1697193 -135417 -119179 -113503

North Dakota $725 87 61225 -9743 -8574 -4021

Ohio $795 143 1647345 -115869 -97169 -88580

Oklahoma $725 129 494053 -53006 -46650 -46854

Oregon $910 124 485326 -17036 -13328 -16398

Pennsylvania $725 161 2053643 -177315 -156052 -125586

Rhode Island $800 156 163730 -10258 -8574 -8698

South Carolina $725 94 445277 -65614 -57746 -50304

South Dakota $725 208 173749 -11586 -10197 -7458

Tennessee $725 142 914903 -89667 -78915 -99134

Texas $725 110 2863779 -362018 -318607 -253285

Utah $725 88 251107 -39658 -34902 -19390

Vermont $873 156 97792 -3823 -3055 -2475

Virginia $725 101 829771 -113723 -100086 -58212

Washington $932 72 496934 -23221 -17947 -17756

West Virginia $725 58 107875 -25792 -22699 -21665

Wisconsin $725 75 427822 -79521 -69986 -53210

Wyoming $725 164 94590 -8010 -7050 -3104

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3843

Appendix C | wwwamericanprogresso

StateMinimum wage

(2014)

SNAP expenditures (2012)Predicted change in SNAP expenditures

(millions of dollars)

Per resident

(millions of

dollars)

Total

(millions of

dollars)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 s

expenditu

(linear regre

Alabama $725 $290 $1397 -$1185 -$1043 -$1043

Alaska $775 $253 $185 -$262 -$223 -$106

Arizona $790 $257 $1686 -$935 -$787 -$892

Arkansas $725 $248 $733 -$636 -$560 -$547

California $800 $189 $7164 -$7276 -$6082 -$3573

Colorado $800 $157 $816 -$862 -$721 -$407

Connecticut $870 $191 $686 -$343 -$275 -$210

Delaware $725 $250 $229 -$205 -$180 -$171

District of Columbia $825 $366 $232 -$146 -$120 -$99

Florida $793 $294 $5676 -$4429 -$3719 -$2951

Georgia $725 $317 $3140 -$2936 -$2584 -$2346

Hawaii $725 $335 $465 -$449 -$395 -$347

Idaho $725 $225 $359 -$376 -$331 -$268

Illinois $825 $249 $3200 -$2096 -$1725 -$1364

Indiana $725 $220 $1439 -$1162 -$1023 -$1075

Iowa $725 $192 $589 -$658 -$579 -$440

Kansas $725 $159 $460 -$502 -$441 -$344

Kentucky $725 $298 $1303 -$1133 -$997 -$973

Louisiana $725 $315 $1450 -$1047 -$922 -$1083

Maine $750 $281 $373 -$267 -$231 -$246

Maryland $725 $188 $1109 -$1765 -$1553 -$828

Massachusetts $800 $206 $1366 -$1030 -$861 -$681

Michigan $740 $300 $2963 -$2400 -$2090 -$2054

Minnesota $725 $140 $755 -$1113 -$980 -$564

Mississippi $725 $326 $973 -$649 -$571 -$726

Missouri $750 $241 $1452 -$1278 -$1104 -$956

Montana $790 $190 $191 -$179 -$151 -$101

Nebraska $725 $140 $259 -$409 -$360 -$194

Nevada $825 $191 $527 -$441 -$363 -$225

New Hampshire $725 $126 $167 -$399 -$351 -$125

TABLE 11

SNAP expenditures Predicted changes by state under the Harkin-Miller bill

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3943

36 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

StateMinimum wage

(2014)

SNAP expenditures (2012)Predicted change in SNAP expenditures

(millions of dollars)

Per resident

(millions of

dollars)

Total

(millions of

dollars)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 state

expenditures

(linear regression)

New Jersey $825 $160 $1420 -$1611 -$1325 -$605

New Mexico $750 $324 $675 -$513 -$444 -$445New York $800 $287 $5616 -$3766 -$3148 -$2801

North Carolina $725 $252 $2454 -$2187 -$1925 -$1833

North Dakota $725 $128 $90 -$162 -$143 -$67

Ohio $795 $259 $2995 -$2013 -$1688 -$1539

Oklahoma $725 $248 $945 -$799 -$703 -$706

Oregon $910 $322 $1255 -$272 -$213 -$262

Pennsylvania $725 $218 $2779 -$2930 -$2579 -$2075

Rhode Island $800 $280 $294 -$173 -$144 -$147

South Carolina $725 $291 $1373 -$1337 -$1177 -$1025South Dakota $725 $198 $165 -$192 -$169 -$123

Tennessee $725 $324 $2091 -$1413 -$1243 -$1562

Texas $725 $230 $5997 -$6402 -$5634 -$4479

Utah $725 $141 $402 -$614 -$541 -$300

Vermont $873 $230 $144 -$66 -$53 -$43

Virginia $725 $173 $1413 -$2062 -$1815 -$1056

Washington $932 $244 $1682 -$350 -$270 -$267

West Virginia $725 $273 $508 -$451 -$397 -$379

Wisconsin $725 $204 $1166 -$1302 -$1146 -$871Wyoming $725 $95 $55 -$105 -$93 -$41

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4043

Appendix C | wwwamericanprogresso

FIGURE 2

Raising the minimum wage to $1010would cut taxpayer costs in every state

Predicted decreases in cost and enrollment

in SNAP in 50 states

$200+$51ndash$100

$101ndash$200

0ndash$25

$26ndash$50

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4143

38 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

Endnotes

1 US Department of Agriculture ldquoSupplementalNutrition Assistance Program (SNAP)rdquo available athttpwwwfnsusdagovsnapeligibility (last accessedFebruary 2014)

2 For this initial analysis we do not consider Harkin-Millerrsquos increase in subminimum wages for tippedworkers To do s o would increase the estimated SNAP

savings by an unknown amount

3 The Congressional Budget Office estimates thatworkers currently earning between $1010 and $1150per hour would see their wages rise under the Harkin-Miller proposal Congressional Budget O ffice ldquoTheEffects of a Minimum Wage Increase on Employmentand Family Incomerdquo (2014)

4 Marianne Page Joanne Spetz and Jane Millar ldquoDoesthe Minimum Wage Affect Welfare Caseloadsrdquo Journalof Policy Analysis and Management 24 (2) (2005)273ndash295

5 Marianne Bitler and Hilary Hoynes ldquoThe More ThingsChange the More They Stay the Same The SafetyNet Living Arrangements and Poverty in the GreatRecessionrdquo NBER Working Paper 194 49 2013

6 Sylvia Allegretto and others ldquoFast Food Poverty Wages The Public Cost of Low-Wage Jobs in the Fast-FoodIndustryrdquo (Berkeley California Center for LaborResearch and Education 2013) available at httplaborcenterberkeleyedupubliccostsfast_food_poverty_wage

7 David Neumark and William Wascher ldquoDoes a HigherMinimum Wage Enhance the Effectiveness of theEarned Income Tax Creditrdquo Industrial and LaborRelations Review 64 (4) (2011) 712ndash746

8 David Lee and Emmanuel Saez ldquoOptimal MinimumWage Policy in Competitive Labor Marketsrdquo Journal ofPublic Economics 96 (9) (2012) 739ndash749

9 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family IncomerdquoFebruary 2014

10 Jesse Rothstein ldquoIs the EITC as Good as an NITConditional Cash Transfers and Tax Incidencerdquo AmericanEconomic Journal Economic Policy 2 (1) (2010) 177ndash208

11 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family Incomerdquo

12 Dube Arindrajit 2013 rdquoMinimum Wagesand the Distribution of Family IncomerdquoUnpublished working paper Available at httpsdldropboxusercontentcomu15038936Dube_MinimumWagesFamilyIncomespdf

13 As Dube explains in ldquoThe poverty of Minimum WageFactsrdquo the simulation approach underestimate stemsfrom a number of unwarranted assumptions includingthe range of actual wage increases and the accuracy ofwage data in the Current Population Survey The causal

approach does not make these assumptions

14 Allegretto Sylvia and others 2013 ldquoCredible ResearchDesigns for Minimum Wage Studiesrdquo Working Paper148-13 University of California Berkeley Institutefor Research on Labor and Employment Available athttpwwwirleberkeleyeduworkingpapers148-13pdf

15 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family In comerdquo

16 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo (2012)

17 Ibid

18 Ibid

19 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo

20 US Department of Agriculture ldquoSupplementalNutrition Assistance Program (SNAP)rdquo

21 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo

22 Ibid

23 We deviate from the Census Bureaursquos definition ofa family unit which is ldquotwo people or more (on e ofwhom is the householder) related by bir th marriageor adoption and residing togetherrdquo US Bureau ofthe Census ldquoCurrent Population Survey DefinitionsFamilyrdquo available at httpwwwcensusgovcpsabout

cpsdefhtml (last accessed February 2014) We countas a family unit any individual residing on his or herown two or more persons residing together whodo not belong to a family in the March CPS sampleare constructed as one family in our analysis For thepurposes of food stamp allocations the consumptionresulting from this transfer is probably distributed tofamily members (rather than household members ora single individual within the household) Howeversingle individuals canmdashand domdashreceive SNAPbenefits Excluding them would fail to make theanalysis reflective of the population at large

24 Strictly the family level linear probability modelpredicts the percentage-point decrease in theprobability that an individual family will receive SNAPpayments When applied to a large number of familieshowever we are able to interpret the coefficient asa decrease in the mean of enrollmentmdashthat is a

decrease in the enrollment ratemdashby applying the lawof iterated expectations

25 We generate expenditure predictions from theenrollment modelsmdashand conversely generateenrollment predictions from the expenditure modelmdashby assuming that expenditures per enrolled familyremains the same before and after the minimum wagechange In practice this is likely to be a conservativeestimatemdashthat is to underestimate the decrease inSNAP activity Average SNAP benefits per family willalso decrease as many families that remain eligible forSNAP experience income gains

26 Wage and Hour Division ldquoMinimum Wage Laws inthe States ndash Januar y 1 2014rdquo available at httpwwwdolgovwhdminwageamericahtm (last accessedFebruary 2014)

27 See for example Sylvia Allegretto and others ldquoCredibleResearch Designs for Minimum Wage Studiesrdquo WorkingPaper 148-113 (Berkeley California Institute forResearch on Labor and Employment 2013) available athttpwwwirleberkeleyeduworkingpapers148-13pdf

28 We will report these results in a forthcoming workingpaper

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4243

Endnotes | wwwamericanprogresso

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4343

The Center for American Progress is a nonpartisan research and educational institute

dedicated to promoting a strong just and free America that ensures opportunity

for all We believe that Americans are bound together by a common commitment to

these values and we aspire to ensure that our national policies reflect these values

We work to find progressive and pragmatic solutions to significant domestic and

international problems and develop policy proposals that foster a government that

is ldquoof the people by the people and for the peoplerdquo

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 443

Introduction and summary | wwwamericanprogress

Introduction and summary

How do minimum wage policy increases affec enrollmens and expendiures on

means-esed public assisance programs In his repor we address his quesion

or he case o he Supplemenal Nuriion Assisance Program or SNAP

ormerly known as he ood samp program

By definiion governmen spending on a means-esed program should decline

as average earnings increase insoar as benefi levels all wih increased earnings

and insoar as he earnings increase makes some individuals ineligible orany benefis Boh o hese condiions are saisfied in he case o he effec o

minimum wages on SNAP benefis SNAP benefis decline 30 cens or every $1

increase in amily earnings and phase ou enirely a abou he ederal povery

level1 Low-wage workers are disproporionaely enrolled in SNAP A minimum

wage increase ha lifs many amilies ou o povery should hereore reduce

public expendiure on his program

Bu he relaionship may be more complex I a minimum wage increase reduces

employmen hereby adding o he number o unemployed he number o SNAP

recipiens could increase SNAP recipiens who are unemployed disabled or

reired will no be affeced by a minimum wage increase Conversely i many

SNAP recipiens have earnings ha already bring hem close o becoming

ineligible or he program a minimum wage increase may have a very small effec

on SNAP expendiures Te quaniaive effec o minimum wages on SNAP

spending is no sel-eviden I requires a causal analysis

In an era o hisorically low real ederal minimum wage raes rising income

inequaliy job-marke sagnaion and conenious debae abou governmen defici

spending he possibiliy ha a higher minimum wage may lead o increased orreduced public spending has grea relevance o he public and o policymakers

Tis repor presens an iniial empirical analysis o he effecs o minimum wage

policy on SNAP paricipaion and expendiures We do so by exploiing more han

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 543

2 Center for American Progress | The Effects of M inim um Wages on SNAP Enro llments and Expenditures

wo decades o variaion in binding sae and ederal minimum wage changes in an

economeric ramework Our uure research will examine he effecs on SNAP

urher and apply an analogous ramework o wo oher public assisance spending

programs he Earned Income ax Credi and Medicaid

According o he finding in his repor a 10 percen increase in he minimum wagereduces SNAP enrollmen by beween 24 percen and 32 percen and reduces

program expendiures by an esimaed 19 percen aking ino accoun each

saersquos 2014 minimum wage level we apply hese resuls o he legislaive proposal

pu orward by Sen om Harkin (D-IA) and Rep George Miller (D-CA) o raise

he ederal minimum wage o $1010 per hour2 Our resuls imply ha he effecs

o he Harkin-Miller proposal on wage increases would reduce SNAP enrollmens

by beween 75 percen and 87 percen (31 million o 36 million persons) Te

oal anicipaed annual decrease in program expendiures is nearly $46 billion or

abou 6 percen o curren SNAP program expendiures

Harkin-Miller proposes o index minimum wage levels in subsequen years o he

consumer price index or CPI Te minimum wage would hen increase a he

same rae as SNAP benefi and eligibiliy levels which are also indexed o he CPI

Consequenly he savings over 10 years in 2014 dollars would be 10 imes he

one-year savings or a oal o approximaely $46 billion

Some o he reducion in SNAP program enrollmen and expendiures would

occur among workers making less han $1010 per hour991252hose whose pay would

be direcly increased by he minimum wage law Anoher par o he reducion

would occur among workers currenly earning beween $1010 and $1150 who

would also receive pay increases3

Alhough a large number o sudies have examined he impac o minimum

wage increases on earnings and employmen he impac o such minimum

wage policies on public assisance enrollmens and expendiures remains an

under-explored subjec in he economic lieraure Only a ew sudies discuss

he relaion beween he minimum wage and governmen ranser spending

much less atemp o ideniy he causal effec o one upon he oher Proessors

Marianne Paige Joanne Spez and Jane Millar find posiive effecs o minimum wage increases on welare caseloads as hey sae however heir resuls vary

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 643

Introduction and summary | wwwamericanprogress

considerably wih differen sample periods and assumpions abou sae rends4

Proessors Marianne Biler and Hilary Hoynes discuss he imporance o SNAP

as a saey ne program bu hey do no examine is relaion o minimum wage

policy5 Research economis Sylvia Allegreto and her Universiy o Caliornia a

Berkeley colleagues show ha low-wage workers in general and as-ood workers

in paricular are much more likely o be SNAP recipiens han all workers6

Several sudies have examined he relaionship beween he minimum wage and

he Earned Income ax Credi or EIC Proessor David Neumark and William

Wascher a researcher a he Federal Reserve Board o Governors find ha a

higher minimum wage increases EIC benefis or amilies in deep povery

while reducing EIC benefis or some sub-groups7 Proessors David Lee and

Emmanuel Saez argue ha he minimum wage and EIC are complemenary

policies no subsiues8 Te Congressional Budge Office or CBO argues ha

a minimum wage increase will no have a subsanial effec on EIC spending9

while Proessor Jesse Rohsein examines wheher he posiive effec o he EICon emale labor supply has lowered wages10 While hese sudies are o ineres

he EIC is quie differen rom SNAP in having a subsanial phase-in period in

which EIC benefis increase as well as a long phase-ou period wih complee

phase-ou a an annual income o abou $48000 or a amily o our quie a bi

above he reach o he minimum wage11

Research by Proessor Arindraji Dube on he causal effec o he minimum wage

on amily povery represens he sudy mos relaed o he one a hand12 Dube finds

ha Harkin-Miller would raise abou 46 million non-elderly Americans above

he ederal povery level or FPL In conras when CBO uses a simple simulaion

mehod o address he same quesion hey find ha Harkin-Miller would raise

900000 people above FPL13 Te difference beween hese wo esimaes highlighs

he imporance o underaking a causal analysis Te mehods used in his paper are

in many respecs similar o Dubersquos Moreover since eligibiliy and benefi levels or

programs such as SNAP and Medicaid are ied o he ederal povery level Dubersquos

findings have direc implicaions or his sudy Noneheless his repor appears o

be he firs sudy o examine he effecs o he minimum wage on SNAP In uure

work we plan o underake similar analyses or he EIC and Medicaid

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 743

4 Center for American Progress | The Effects of M inim um Wages on SNAP Enro llments and Expenditures

Te repor proceeds as ollows

bull Secion 1 provides background inormaion on he ederal minimum wage sae

minimum wages and he SNAP program

bullSecion 2 describes our mehods and daa

bull Secion 3 provides our main resuls including a simulaion o he effecs o a

Harkin-Miller minimum wage increase and a sae-by-sae analysis

bull Secion 4 presens our conclusions

Furher deails are provided in a series o appendices

Correction April 28 2014 Tis report incorrectly stated the potential reduction in

SNAP enrollment fom the Harkin-Miller proposal Te correct amount is 75 percentand 87 percent (31 million to 36 million persons) as stated in the reportrsquos tables

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 843

Background | wwwamericanprogress

Background

The federal minimum wage

Te ederal minimum wage was las increased in July o 2009

nearly five years ago During he pas wo decades many saes

have passed legislaion fixing he minimum wage a a higher

level han he ederal minimum Te maps in Figure 1 show ha

while saes in every region o he Unied Saes have adoped

higher minimum wages hey are no disribued randomly bygeography As shown in he 2013 sudy ldquoCredible Research

Designs or Minimum Wage Sudiesrdquo by economiss Sylvia

Allegreto Arindraji Dube Michael Reich and Ben Zipperer

hese saes vary sysemaically rom he oher saes by a number

o characerisics ha affec low-wage employmen rends bu

which are no hemselves relaed o minimum wage policy14

Te nonrandom patern o minimum wage adopion has

imporan implicaions or obaining unbiased esimaes o

minimum wages on employmen In paricular naional panel

sudies ha use sae and ime fixed effec models991252such as

a 1992 sudy by David Neumark and William Wascher991252

spuriously esimae negaive employmen effecs Te reason

or his resul is uncovered using ess or pre-rends Tese

ess find ha low-wage employmen was already declining wo

years beore minimum wages were implemened By making a

saisically large number o local comparisons ha conrol or

heerogeneiy among saes and by ime eliminaes his pre-

rend For his reason we conduc similar ess or our SNAPoucomes and use model specificaions ha include local

comparisons as in he sudy cied above

FIGURE 1

High versus low minimum wage stat

from 1990 to 2012

Means and variances

More than $533

Average minimum wage over 1990ndash2012

Less than or equal to $533

More than $121

Minimum wage variance over 1990ndash2012

Less than or equal to $121

Notes State means and variances were calculated using annual stat

minimum wage data from 1990 to 2012 The shading on the maps

partitions the states into above- and below-median values

Source Sylvia Allegretto and others ldquoCredible Research Designs forMinimum Wage Studiesrdquo Working Paper 148-13 (Institute for Resear

on Labor and Employment 2013) available at httpwwwirleberkeleyeduworkingpapers148-13pdf

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 943

6 Center for American Progress | The Effects of M inim um Wages on SNAP Enro llments and Expenditures

Te CBO recenly projeced ha in 2016 17 million workers will earn less han

he $1010 hourly wage proposed in he Harkin-Miller bill Furhermore he

CBO esimaes ha an addiional 8 million workers earned beween $1010 and

$1150 per hour and were also likely o experience a wage increase15

Supplemental Nutrition Assistance Program

We ocus our inquiry on SNAP Benefis under he program are enirely

ederally unded he program is adminisered by he US Deparmen o

Agriculure ogeher wih sae agencies which share in adminisraion

coss Spending on SNAP has grown in he pas decade reaching $78 bil lion

in 2011 wih SNAP enrollmen increasing o 45 million people abou one-

sevenh o he US populaion16 According o he CBO changes since 1990 in

SNAP spending and enrollmens are primarily he resul o cyclical economic

condiions noably changes in he unemploymen rae and changes in per capiaincome17 Te 2009 American Recovery and Reinvesmen Ac emporarily

increased SNAP benefi amouns by 136 percen as repored by he US

Deparmen o Agriculurersquos Food and Nuriion Service hese higher benefi

levels expired on November 1 201318 Te CBO esimaes ha abou wo-

hirds o he changes in SNAP expendiure are associaed wih changes in he

number o recipiens and one-hird wih changes in he benefis received when

recipiensrsquo incomes change19

In fiscal year 2014 SNAPrsquos maximum monhly benefis are $189 or a single

individual $497 or a amily o hree and $750 or a amily o five Benefis are

reduced by 30 cens per dollar received and phase ou enirely a gross monhly

household incomes o 130 percen o he ederal povery level $1245 or a

single individual $2116 or a amily o hree and $2987 or a amily o five

o deermine benefis SNAP also defines a ne monhly income concep and

ses benefis a 100 percen o he ederal povery level using his concep

Calculaion o ne monhly income can include cerain deducions rom

monhly gross income such as medical expenses and child care coss Alhough

saes are permited some laiude on wha deducions are allowed in pracice

hese vary by very small amouns Our saisical model akes accoun o sae-specific differences in benefis20

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1043

Background | wwwamericanprogress

SNAP imposes an employmen or work-raining requiremen or able-bodied

prime-age aduls991252hose beween he ages o 18 and 50 and wihou disabiliies

or dependen children Such households can receive only hree monhs o benefis

in a hree-year period In recen years abou 85 percen o households receiving

benefis have incomes below he ederal povery level 49 percen have dependen

children 16 percen are age 60 or older 20 percen are disabled and 30 percenrepor some earned income21

A 2012 CBO repor also noes ha ake-up raes among eligible SNAP recipiens

average abou 70 percen wih much lower ake-up among elderly households

Te ake-up rae increases in harder economic imes I also increased when sigma

issues were reduced as SNAP debi cards replaced acual ood samps ake-up

is especially high among hose mos needy Adminisraive spending equaled 91

percen o he poenial spending ha would have occurred i all eligible recipiens

were enrolled Alhough some SNAP policy changes have occurred since 1990

mos were relaively minor and all were naional in scope Te 1996 welare reorm bill eliminaed SNAP eligibiliy or some legal immigrans limied he ime lengh

o eligibiliy or able-bodied childless aduls and reduced maximum benefis

Some o hese resricions were relaxed in 2002 and again in he American

Recovery and Reinvesmen Ac in 200922

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1143

8 Center for American Progress | The Effects of M inim um Wages on SNAP Enro llments and Expenditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1243

Methods and data | wwwamericanprogress

Methods and data

As previously menioned we exploi variaion in minimum wages by sae and ime

o examine heir causal effecs on SNAP enrollmens and expendiures o do so we

merge daa rom 1990 hrough 2012 drawn rom he Annual Social and Economic

Supplemen o he Curren Populaion Survey991252an annual Census Bureau survey

commonly known as he March CPS ha includes23 inormaion on SNAP

enrollmens a he amily level991252wih sae-level daa on minimum wages SNAP

expendiures populaion unemploymen raes and sae median income levels

o conrol or ime-varying heerogeneiy among saes our specificaions includeconrols or sae linear rends and effecs by Census division and ime We esimae

effecs a wo levels allowing or amily variaion and allowing only or sae-level

variaion We also employ a se o sandard demographic conrols such as amily size

and composiion and race and ehnic composiion

Distinguishing causation from correlation

How can we ensure ha our analysis does no pick up a spurious correlaion or

example he endency o more economically vibran saes o implemen higher

minimum wages Disinguishing correlaion and policy endogeneiy rom rue

causal effecs is he primary moivaion or economeric analysis In he ideal

experimen researchers would begin wih wo saes991252ha are alike in every

respec prior o he policy991252 and ldquoreardquo only one o hese saes wih a higher

minimum wage Tey would atemp o shield hese saes rom any influence ha

could obscure heir undersanding o he minimum wagersquos direc effec on SNAP

enrollmen Researchers o course canno conduc such experimens

We can however use saisical mehods o conrol simulaneously heindependen effecs on SNAP o sae unemploymen raes sae income levels

and common rajecories among saes wihin he same Census division By

ensuring similariy along hese dimensions we maximize he likelihood ha

SNAP aciviy in wo saes would have comparable oucomes in he absence

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1343

10 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

o a minimum wage policy change Tus i a new minimum wage policy were

implemened in one sae only he researchers could atribue all o he difference

hey observe in SNAP aciviy o he new minimum wage policy

In oher words we approximae he ideal experimen by using non-experimenal

saisical mehods Te desirable ldquopre-exising similariiesrdquo beween saes ha wehave defined above inorm our choice o conrol variables in a saisical seting

More precisely in our muliple regression models we use median amily income

he unemploymen rae he employmen-o-populaion raio and regional and

ime idenifiers o consruc an appropriae group o peers or each sae on he

eve o a policy change

Data description

wo daa ses include inormaion abou boh income and paricipaion inpublic programs Te Survey o Income and Program Paricipaion or SIPP

which is conduced in inermiten years has he advanage o ollowing he

same individuals over a period o ime In oher words i is a longiudinal daa

se I also has he advanage o conaining monhly daa However he sample

size o he SIPP is no sufficien or analyzing variaions in sae-level minimum

wages Te March CPS has he advanage o a much larger sample size and i is

conduced annually wihou any breaks in ime I has he disadvanage o being

a cross-secional daa se so we canno ollow he same individuals over ime991252

sricly speaking over more han one year On ne he March CPS is much more

suiable or our sudy We examine he empirical relaionship beween minimum

wage policy and ood samp aciviy a wo levels o aggregaion he amily level

and he sae level Family-level daa are drawn rom he March CPS

Te March CPS comprises responses rom he residens o 50000 o 60000

dwelling places surveyed per year and conains deailed inormaion on he

residensrsquo employmen and income including income rom ranser paymens

Te sample or our analysis comprises more han 128 million amily unis during

he period rom 1990 o 2012 (inclusive) Survey weighs allow us o analyze

SNAP paricipaion in a manner ha is represenaive o he US populaion alarge Over all years he share o amilies reporing ood samp receip in he

weighed March CPS sample is 91 percen Te enrollmen rae was a a low o 6

percen in he year 2000 In 2012 he mos recen year in our panel 133 percen

o amilies repored paricipaing in SNAP a some poin during he survey year

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1443

Methods and data | wwwamericanprogresso

Te March CPS also collecs inormaion on he number o SNAP recipiens in

he household he number o monhs he household receives SNAP benefis

and he cash-equivalen value o he SNAP benefis received However he

value o SNAP benefis is severely underrepored among recipiens perhaps

because recipiens are unaware o he exac monhly cash-equivalen value o

benefis hey receive

Our firs empirical sraegy ocuses solely on SNAP enrollmen By using he

amily as he uni o analysis we are able o inser saisical conrols o accoun

or non-wage-relaed acors ha influence any paricular amilyrsquos likelihood o

program paricipaion wih he inenion o isolaing any differences in program

paricipaion ha are due purely o changes in wage policy Tis approach

idenifies he effecs o low-wage labor policy on he exernal margin991252ha is

he effec o he minimum wage on he likelihood ha a amily paricipaes in he

SNAP program a all991252as opposed o he inernal margin or how much SNAP

unding he amily would receive

Our second empirical ramework uses sae-level adminisraive daa Ta is

we aggregae he daa o obain a single daa poin or each saeyear back

o 1990 represening he mean o he oucome or he sae Te sae-level

esimaion serves as a robusness check on he amily-level resuls or SNAP

paricipaion Also using aggregaed daa allows us o esimae direcly he

causal effec o minimum wage changes on SNAP spending Tis is no possible

a he amily level as discussed above daa on cash-equivalen value o ood

samps or SNAP recipiens is very requenly no repored in he March CPS

and when i is repored he inormaion may be unreliable By conras he

Bureau o Economic Analysis publishes aggregae SNAP spending a he sae

level in is Naional Income and Produc Accoun or NIPA ables Tus

while we are unable o observe he heerogeneiy in he cash value o SNAP

or amilies in each sae we are able o calculae average SNAP spending

per residen in each sae per year Supporing covariaes include he annual

unemploymen and employmen daa rom he Bureau o Labor Saisics or

BLS and sae-level populaion series rom he iner-decennial census releases

Minimum wage daa are available rom he BLSrsquos wages and hours division For

sae minimum wage changes enaced a oher imes han he firs o he yearan average value or he year is used

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1543

12 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

Family level model specifications

We firs examine he effec o he minimum wage on paricipaion in public

assisance programs For amily residing in sae and during year we esimae

an equaion o he ollowing orm

(1)

is a binary variable ha is se equal o 1 i a leas one member o amily

received ood samps during he survey year is a se o sae-level

characerisics including annual averages o he unemploymen rae he

employmen-o-populaion raio and he naural log o median amily income

is a vecor o amily atribues including indicaors or he race and marial saus

o he amily head size o he amily he presence o children and he presence oan adul male Sae fixed effecs are capured by o conrol or ime-varying

heerogeneiy our preerred model specificaion also includes year fixed effecs

ha vary by Census division ( ) and sae-level linear ime rends In

Appendix B we jusiy he inclusion o hese las wo erms We also compare he

resuls rom our preerred specificaion wih less sauraed specificaions

Te effec o ineres which is capured by is he expeced change in he

probabiliy o receiving SNAP benefis wih respec o a change in he (log o he)

binding minimum wage in sae during year We repor robus sandard errors

clusered a he sae level We esimae he parameers using linear regression

producing a linear probabiliy model Deails o he model selecion process are

covered in Appendix B below

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1643

Methods and data | wwwamericanprogresso

State-level model specifications

Te sae-level models are similarly specified For sae in year we assume ha

(2)

In his model is now eiher he SNAP enrollmen rae in sae during year

or he naural logarihm o per capia SNAP expendiures in sae during

year is once again a se o sae-level characerisics including he same

sae-level covariaes as in he amily regressions (annual average unemploymen

rae employmen-o-populaion raio naural log o median amily income)

wih he addiion o amily level characerisics averaged across he sae (average

amily size and he shares o populaion consiued by each o five racialehnicgroups) Sae fixed effecs are represened by As above our preerred model

specificaion includes year fixed effecs ha vary by Census division ( ) and

sae-level linear ime rends as elaboraed in Appendix B Te effec o

ineres is capured by

We esimae boh sae-level models (enrollmen and expendiures) using

ordinary leas squares regression Tus he inerpreaion o he coefficien is no

longer ha o a change in probabiliy as in he binary oucome models described

above Raher or he sae-level SNAP enrollmen model represens he

expeced change (in percenage poins) in he saersquos SNAP enrollmen rae ha

is due o a 1 percen change in he minimum wage For he SNAP expendiures

model is simply he elasiciy o SNAP spending wih respec o he minimum

wage991252ha is he percenage change in sae expendiures expeced o resul rom

a 1 percen change in ha saersquos minimum wage For urher deails on model

specificaion reer o Appendix B below

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1743

14 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1843

Results | wwwamericanprogresso

Results

Estimated minimum wage effects on SNAP enrollment and

expenditures

able 1 shows he esimaed parameer o ineres991252he coefficien o he

minimum wage or he preerred model o each ype Coefficiens on he

minimum wage variable are no direcly comparable across models because all

our models have a differen uncional orm o undersand and compare hese

esimaes we compue he change in SNAP aciviy prediced or a paricular wage scenario Te final column in able 1 answers he quesion Wha would be

he expeced change in SNAP aciviy in response o a 10 percen increase in he

minimum wage Te answer o his quesion varies wih he value o he inpu

parameers in he able we calculae he percenage decrease in enrollmen or

expendiures prediced or he average sae wih a minimum wage o $725 in

2014 Te sae-level SNAP expendiure model which is a consan-elasiciy

model conveys elasiciy inormaion direcly or he change in expendiures per

capia in he sae

TABLE 1

Comparison of national SNAP predictions for a 10 percent increase in the federal minimum wage

Model LevelRegression

type

Predicted outcome Coefficient of log

(minimum wage)

(Standard error)

Effect of a 10 percent in

in the minimum wage

VariableForm of

variable

Total

enrollment

Tot

expend

1 Family Linear

probability Enrollment Binary (enrolled=1)

-0042

(0008) -317 N

2

StateLinear regression

(ordinary least

squares)

Enrollment State enrollment rate (percent) -0031

(0012) -235 N

3 Expenditures Log (state expenditures per capita) -0190

(0103) NA -19

plt01 plt005 plt001Note Predicted changes are calculated for the average state with a minimum wage of $725 in 2014

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1943

16 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

According o his model a 10 percen increase in he minimum wage would resul

in a 19 percen reducion in SNAP expendiures Tis resul is reassuringly similar

o Dubersquos resul or povery reducion Te wo enrollmen models are much more

precisely esimaed han he expendiure model Te sae-level enrollmen model

finds ha a 10 percen minimum wage increase in a low-wage sae is associaed wih

a 235 percen decrease in SNAP enrollmens Te amily-level linear probabiliymodel predics a somewha greaer elasiciy or low minimum wage saes an

increase o 10 percen in he ederal minimum would resul in a 317 percen decline

in SNAP enrollmen24 Te differences in hese esimaes sem rom a number o

acors including difference in model uncional orm and daa used We rea his

range o elasiciy esimaes as an upper and lower bound on enrollmen impacs

Harkin-Miller bill National and state-level predicted impacts

Wha would be he prediced change or he SNAP program i he ederalminimum were raised o $1010 as proposed in he Harkin-Miller bill In order o

make his inerence we accoun or he ac ha no all saes are currenly subjec

o he ederal minimum wage a he beginning o 2014 21 saes mainained

higher minimum wages han $725 In hose saes an increase in he ederal

minimum wage may or may no be binding or employers in he sae depending

upon wheher he new ederal minimum exceeds he sae-level minimum Bu

regardless o wheher a minimum wage change is binding he impac on SNAP

aciviy will be lower in high minimum wage saes In order o accoun or his

properly we calculae sae by sae he percenage wage change ha would resul

rom he Harkin-Miller proposal and apply he parameers rom each o he hree

models above o compue he expeced decrease in SNAP aciviy or each sae

In his exercise we use saesrsquo curren (2014) minimum wage levels and assume

as a baseline he 2012 levels o SNAP enrollmen and expendiure as 2012 is he

mos recen year or which SNAP daa are available

able 10 and able 11 in Appendix C repor he esimaed effecs on SNAP

enrollmen and expendiures respecively or each sae under he Harkin-Miller

bill25 An increase o $1010 i enaced oday would represen beween a 393

percen wage increase in a $725 minimum wage sae and an 84 percen increasein Washingon sae which has he highes minimum wage in he naion a $932

as o January 201426 Slighly more han 56 percen o he decrease in expendiures

and abou 59 percen o he decrease in enrollmen would occur in saes wih

presen-day minimum wages o $725

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2043

Results | wwwamericanprogresso

In 2012 hese saes were home o 46 percen o he American populaion

and accouned or a roughly equivalen percenage o oal naional SNAP

expendiures No surprisingly he larges enrollmen decreases come rom

populous saes wih very high SNAP enrollmen raes andor very low minimum

wages Te larges prediced enrollmen reducion991252beween 319000 individuals

and 362000 individuals991252would occur in exas which has a $725 minimum wage In Caliornia which has a 206 percen SNAP paricipaion rae and an $8

per hour minimum wage we anicipae SNAP enrollmen declines o 310000

persons o 371000 persons And in Florida which had a SNAP paricipaion rae

o 166 percen and a $793 minimum wage enrollmen could decline by beween

164000 individuals and 196000 individuals For he our saes ha ogeher

accouned or he greaes amoun o SNAP spending in 2012991252exas Caliornia

Florida and New York respecively991252he combined expendiure reducion rom

he Harkin-Miller bill is prediced o be $14 billion

able 2 summarizes he prediced declines in SNAP aciviy or he naion as a whole ha would resul rom he direc and indirec effecs o he Harkin-Miller

bill Enrollmen would all beween 31 million persons and 36 million persons

represening 75 percen o 87 percen o curren enrollmen Te anicipaed

reducion in program expendiures would be nearly $46 billion or 61 percen o

program expendiures

TABLE 2

Comparison of national SNAP predictions under the Harkin-Miller billrsquos $1010 minimum wage

Model

Enrollment

(persons)

Expenditures

(millions of dollars)

Current (2012) Predicted Change Current (2012) Predicted Chan

Family enrollment (linear probability)

41866195

45489339 -3623144

$74861

NA NA

State enrollment (ordinary least squares) 38745435 -3120759 NA NA

State expenditures (ordinary least squares) NA NA $70305 -$45

Note Calculations use 2014 state minimum wages and the most recent SNAP data from 2012 They assume that per-enrollee expenditures remain constant

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2143

18 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

Tere are o course oher possibiliies or minimum wage legislaion able 3

shows he expeced SNAP changes or he Unied Saes under a variey o wage

scenarios calculaed using he sae-level models I saes were no able o se

heir minimum wages independenly such ha all saes were consrained by

he ederal minimum o $725 SNAP would be received by abou 514000 more

people across he Unied Saes a an addiional program cos o nearly hree-quarers o a billion dollars In conras he effecs o a higher minimum wage

proposal991252a ederal wage floor o $11 per hour991252would decrease enrollmen in

SNAP by more han 10 percen and decrease program coss by 83 percen

TABLE 3

Summary of par ticipation and expenditures under wage scenarios

If all states had

minimum wages of

Enrollment(persons)

Expenditures

(millions of dollars)

Predicted Change Predicted Change

Recent levels (2014) 41866195 $74861

$725 42380520 514326 $75604 $743

$800 41423919 -442276 $74209 -$652

$900 40148451 -1717744 $72350 -$2511

$1000 38872982 -2993212 $70490 -$4371

$1010 38745435 -3120759 $70305 -$4556

$1100 37597514 -4268681 $68631 -$6230

Note Calculations use state-level enrollment model coefficient

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2243

Conclusion | wwwamericanprogresso

Conclusion

An exensive body o lieraure examines employmen effecs o he minimum

wage A much smaller se o sudies examines how he minimum wage affecs

povery and only a handul o papers examine he effecs o he minimum wage

on he EIC Our analysis is he firs o examine he effecs o he minimum

wage on SNAP

Our findings indicae ha increased earnings rom minimum wage changes do

reduce SNAP enrollmens and expendiures We esimae ha he Harkin-Miller bill would save axpayers nearly $46 billion per year equivalen o 61 percen

o SNAP expendiures in 2012 he las year or which daa are available Over a

10-year period he esimaed savings amoun o nearly $46 billion

Our repor is subjec o limiaions ha we expec o overcome in our uure

research Firs he findings do no ake ino accoun possible ineracions among

SNAP he EIC and Medicaid Te eligibiliy cuoffs among hese programs

are quie differen suggesing ha such ineracions may be minor Noneheless

he join effecs can only be deermined by urher research using a causal

model Second i would be useul o know he disribuion o SNAP reducions

along he wage disribuion Using he Congressional Budge Officersquos calculaions

o how much he oal dollar value o wage would increase under he Harkin-

Miller proposal our findings imply ha he decline in overall SNAP spending

equals abou 15 percen o he oal resuling increase in wages Te amoun and

disribuion o his offse are o considerable ineres Minimum wage beneficiaries

who come rom working amilies already well above he povery line would no

see any offse while hose who are currenly considerably below he povery line

will see larger offses Tese issues will also be a subjec or our uure research

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2343

20 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

About the authors

Rachel West is a maser o public policy candidae a he Goldman School

o Public Policy Universiy o Caliornia Berkeley Her research ocuses on

economic policy in he areas o low-wage labor and povery

Michael Reich is proessor o economics and direcor o he Insiue or

Research on Labor and Employmen a he Universiy o Caliornia a Berkeley

His research publicaions cover numerous areas o labor economics including

racial inequaliy labor marke segmenaion high-perormance workplaces

union-managemen cooperaion Japanese labor-managemen sysems living

wages and minimum wages He received his docorae in economics rom

Harvard Universiy

Acknowledgments

We are graeul o Sylvia Allegreto Arindraji Dube Bill Leser Jesse Rohsein

Daniel Tompson and Ben Zipperer or heir valuable suggesions

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2443

References | wwwamericanprogresso

References

Allegreto Sylvia and ohers 2013 ldquoFas Food Povery Wages Te Public Cos o Low-Wage Jobsin he Fas-Food Indusryrdquo Universiy o Caliornia Berkeley Cener or Labor Research andEducaion Available a htplaborcenerberkeleyedupubliccossas_ood_povery_wage

Allegreto Sylvia and ohers 2013 ldquoCredible Research Designs or Minimum Wage Sudiesrdquo

Working Paper 148-13 Universiy o Caliornia Berkeley Insiue or Research on Labor andEmploymen Available a htpwwwirleberkeleyeduworkingpapers148-13pd

Biler Marianne and Hilary Hoynes 2013 ldquo Te More Tings Change he More Tey Say heSame Te Saey Ne Living Arrangemens and Povery in he Grea Recessionrdquo Working Paper19449 Naional Bureau o Economic Research

Congressional Budge Office 2012 ldquoTe Supplemenal Nuriion Assisance Programrdquo Washingon Available a htpwwwcbogovsiesdeaulfilescbofilesatachmens04-19-SNAPpd

991252 991252 991252 2014 ldquoTe Effec o a Minimum-Wage Increase on Employmen and Family Incomerdquo Washingon Available a htpwwwcbogovsiesdeaulfilescbofilesatachmens44995-MinimumWagepd

Dube Arindraji 2013 rdquoMinimum Wages and he Disribuion o Family Incomerdquo Unpublished working paper Available a htpsdldropboxuserconencomu15038936Dube_ MinimumWagesFamilyIncomespd

991252 991252 991252 2014 ldquoTe Povery o Minimum Wage Facsrdquo Arindraji Dube Blog January 22 Availablea htparindubecom20140122he-povery-o-minimum-wage-acs

Lee David and Emmanuel Saez 2012 ldquoOpimal Minimum Wage Policy in Compeiive LaborMarkesrdquo Journal o Public Economics 96 (9) 739ndash749

Neumark David and William Wascher 1992 ldquoEmploymen Effecs o Minimum and Subminimum Wages Panel Daa on Sae Minimum Wage Lawsrdquo Industrial and Labor Relations Review 46 (1)

55ndash81

Neumark David and William Wascher 2011 ldquoDoes a Higher Minimum Wage Enhance heEffeciveness o he Earned Income ax Credirdquo Industrial and Labor Relations Review 64 (4)712ndash746

Page Marianne Joanne Spez and Jane Millar 2005 ldquoDoes he Minimum Wage Affec WelareCaseloadsrdquo Journal o Policy Analysis and Management 24 (2) 273ndash295

Rohsein Jesse 2010 ldquoIs he EIC as Good as an NI Condiional Cash ransers and ax

Incidencerdquo American Economic Journal Economic Policy 2 (1) 177ndash208

Wage and Hours Division 2014 ldquoMinimum Wage Laws in he Saes ndash January 1 2014rdquo USDeparmen o Labor (htpwwwdolgovwhdminwageamericahm [February 2014])

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2543

22 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2643

Appendix A | wwwamericanprogresso

Appendix A

Pre-trend falsification check

Recen minimum wage research27 highlighs a common flaw in previous sudies

ailure o veriy ha he oucome variable is ree o negaive pre-exising rends I

or example SNAP aciviy was already rending down in saes ha raised heir

minimum wages beore hese changes came ino effec our regression analysis could

(misakenly) atribue ha reducion o he minimum wage We check or such

pre-rends by inroducing variables ha represen he prior yearrsquos value or leads o

he minimum wage I he model esimaes he minimum wage o have an effec on

he oucome variable beore he wage change wen ino effec hen an unobservedacor no he minimum wage change caused he change in SNAP aciviy

We es he specificaions above or pre-rends by including a one-year lead in

all hree specificaions We find ha he lead erms are small posiive and no

saisically significan indicaing ha he concurren minimum wage991252no

he wage level in prior periods991252is driving he observed changes in SNAP

oucomes28 In paricular he coefficien (sandard error) on he lead erm in

our preerred amily-level enrollmen regression is 011 and no significan

while he coefficien and sandard error o he conemporaneous minimum

wage is unchanged In he sae-level preerred enrollmen regression he

coefficien o he lead erm is again small (07) and i is no significan Te

corresponding coefficien on he lead erm in he sae-level expendiure

regression is 16 and is no significan Te posiive poin esimaes on hese lead

erms resuls no only rule ou disoring negaive pre-rends Tey also sugges

ha our main resuls may underesimae he rue effecs

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2743

24 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2843

Appendix B | wwwamericanprogresso

Appendix B

Model selection process

For boh he amily-level and sae-level models we es hree mehods o conrol

or unobserved geographic- and ime-varying characerisics as suggesed by he

minimum wage lieraure Firs we include only independen sae-specific fixed

effecs and year-specific fixed effecs Tis specificaion (specificaion 1) implicily

assumes ha amilies in any sae consiue an equally good saisical ldquoconrolrdquo

group or hose in any randomly chosen sae afer accouning or various

characerisics (median income and unemploymen rae among ohers) Similarly

simple ime fixed effecs assume ha amilies surveyed in any year can crediblyserve as a conrol group or amilies surveyed in every oher year o he sample

(1990 hrough 2012)

In oher words specificaion 1 assumes ha a saersquos immediae neighbor provides

no beter a couneracual or he effec o a minimum wage change han does a

sae across he counry We relax his resricive specificaion sequenially in wo

seps In specificaion 2 we replace simple year fixed effecs wih fixed effecs or

each Census divisionyear (capured as an addiional variable in he vecor By

using division-year effecs we remove he resricion ha amilies in each sae

are equally good saisical conrols or all oher amilies Raher we allow or he

possibiliy ha amilies in similar geographic regions (or example he Souh or

he Norheas) may be more similar o one anoher han amilies arher away

Finally in specificaion 3 we add sae-specific linear ime rends o he previous

specificaion Tus specificaion 3 is he mos rigorous model specificaion in ha

i allows or heerogeneiy along hree dimensions Ta is specificaion 3 allows

each sae o have is own ime-varying rends raher han imposing he resricion

ha saes evolve idenically over he 22 years in he sample

We begin building he heoreical specificaion above rom a se o simpleuncondiional models regression o SNAP aciviy (enrollmen or expendiures)

on he log o he minimum wage and a se o geographic- and ime-specific

effecs (specificaions 1 2 and 3 described above) As shown in ables 1ndash3 (or

specificaion 3) we hen add covariaes sequenially o hese models including

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2943

26 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

firs he vecor o amily-level conrols ollowed by each o several sae-level

covariaes in urn he unemploymen rae log o median-amily income and he

employmen-o-populaion raio Comparable resuls or specificaions 2 and 3

will be available in our orhcoming working paper

As expeced he simple uncondiional models indicae ha he relaionship beween he minimum wage and SNAP enrollmen i one exiss is a more complex

one influenced by oher acors In he uncondiional model he coefficien on he

variable o ineres991252he log o he minimum wage991252is small in magniude and no

saisically differen rom zero Once we accoun or he influence o labor marke

condiions and variaion in income levels on program paricipaion (by including

unemploymen rae and median-amily income conrol variables respecively)

he effec o he minimum wage on SNAP enrollmen is precisely esimaed Te

coefficien o he log minimum wage is slighly higher (-0042) in he amily-level

analysis han he coefficien (-031) in he sae-level analysis Te level o precision

is also higher in he amily-level analysis Tis is o be expeced when using 124million observaions compared o 1127

Te second dimension o model choice concerns he effec specificaion ables

7ndash9 compare he primary coefficiens o ineres or he SNAP enrollmen and

expendiure models For boh he enrollmen models he effec sizes are smalles

or specificaion 1 larges or specificaion 2 and inermediae beween hese wo

in specificaion 3 Recall ha Specificaion 3 conains sae-specific linear ime

rends in addiion o he census divisionyear conrols included in specificaion

2 In he amily-level enrollmen model he sandard error o he minimum wage

coefficien is smaller han in he oher wo specificaions Sandard errors on he

oher variables are much smaller in specificaions 2 and 3 han in specificaion

1 On he basis o coefficien significance (join and individual) specificaions 2

and 3 are sricly preerred in boh enrollmen models o specificaion 1 which

conains only sae and year fixed effecs

A concern wih specificaions 2 and 3 is ha rend conrols such as sae linear

rends may incorrecly absorb some o he delayed impac o a minimum wage

When we es his issue by including lagged minimum wages we do no find ha

delayed effecs are significan Anoher concern is ha more sauraed modelsuse less o he saisical variaion which could reduce he saisical power o

he resuls However he sandard errors or our more sauraed models are no

higher and are lower in some cases han or he less sauraed models Overall

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3043

Appendix B | wwwamericanprogresso

his evidence suppors our use o he sauraed model as he preerred model

specificaion Moreover Dubersquos 2013 sudy shows ha more sauraed models

perorm beter han models wih jus sae and ime fixed effecs

Te esimaed enrollmen regressions a boh he amily and sae levels show large

and saisically significan coefficiens Te esimaed minimum wage effec in heexpendiures regressions991252or which we have only sae-level daa991252is also large

and saisically significan

We do no use weighed regression or he sae-level models preerring o keep

analysis o he ldquoreamenrdquo (ha is o say a minimum wage change) appropriae

o he average sae raher han he average amily or individual I insead our

primary ineres were he impac o a minimum wage change on he average amily

or he average individual we migh choose o designae he number o amilies

in each sae or he sae populaion respecively as analyic weighs in order o

obain a coefficien beter suied or such inerence

TABLE 4

SNAP enrollment

Family-level linear probability

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0048 -0047 -0040 -0043 -0042

(0013) (0013) (001) (0008) (0009)

Unemployment rate (100) 0505 0420 0280

(0083) (0086) (0082)

Log median income -0057 -0039

(0011) (0011)

Employment-to-population ratio -0239

(0038)

N 1242022 1242022 1242022 1242022 1242022

plt01 plt005 plt001

Note Robust standard errors are in parentheses Observations are clustered at the state level The outcome variable is binary and equal to one if a family is enrolledin SNAP All models include state fixed effects Census division x-year fixed effects and state-specific linear time trends All specifications except 3a include additional

controls for family size race and marital status of the family head presence of children and presence of an adult maleSource Annual data from the March Current Population Survey for 1990 to 2012 estimation includes Current Population Survey probability weights

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3143

28 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

TABLE 5

SNAP enrollment

State-level linear regression

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0028 -0024 -0026 -0031 -0031

(0014) (0013) (0013) (0012) (0012)

Unemployment rate (100) 0625 0466 0320 0339

(0087) (0088) (0085) (0083)

Log median income -0090 -0065 -0061

(0013) (0013) (0013)

Employment-to-population ratio -0282 -0248

(0037) (0038)

N 1127 1127 1127 1127 1127

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the SNAP enrollment rate All models include state fixed effects Census division x-year fixedeffects and state-specific linear time trends

Source Annual data from the March Current Population Survey for 1990 to 2012

TABLE 6

SNAP expenditures

State-level linear regression

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0181 -0149 -0156 -0153 -0190

(011) (0103) (0102) (0103) (0103)

Unemployment rate (100) 4587 4152 4232 4313

(0622) (0621) (0633) (0628)

Log median income -0246 -0261 -0294

(0075) (0078) (0078)

Employment-to-population ratio 0155 0244

(0237) (024)

N 1127 1127 1127 1127 1127

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the natural log of state SNAP expenditures per capita for 1990 to 2012 All models include state

fixed effects Census division x-year fixed effects and state-specific linear time trends

Source Annual data from Bureau of Economic Analysis national income and product accounts tables for 1990 to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3243

Appendix B | wwwamericanprogresso

TABLE 7

Comparison of specifications SNAP enrollment

Family-level linear probability

(1) (2) (3)

Log minimum wage -0028 -0049 -0042

(0014) (0017) (0009)

Unemployment rate (100) 0275 0297 0280

(0161) (0076) (0082)

Log median income -0077 -0055 -0039

(0014) (0012) (0011)

Employment-to-population ratio -0238 -0250 -0239

(0054) (004) (0038)

N 1242022 1242022 1242022

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses Observations are clustered at the state level The outcomevariable is binary or equal to one if a family is enrolled in SNAP All specifications include additional controls forfamily size race and marital status of the family head presence of children and presence of an adult male

Source Annual data from the March Current Population Survey for 1990 to 2012 estimation includes CurrentPopulation Survey probability weights

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3343

30 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

TABLE 8

Comparison of specifications SNAP enrollment

State-level linear regression

(1) (2) (3)

Log minimum wage -0019 -0035 -0031

(0009) (0012) (0012)

Unemployment rate (100) 0401 0370 0339

(0063) (0077) (0083)

Log median income -0081 -0073 -0061

(0011) (0013) (0013)

Employment-to-population ratio -0183 -0222 -0248

(0039) (0039) (0038)

N 1127 1127 1127

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the SNAP enrollment rate Allregressions include the share of households with children and the racial shares of the population

Source Annual data from the March Current Population Survey for 1990 to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3443

Appendix B | wwwamericanprogresso

TABLE 9

Comparison of specifications SNAP expenditures

State-level linear regression

(1) (2) (3)

Log minimum wage -0121 -0203 -0190

(0075) (0103) (0103)

Unemployment rate (100) 5292 5152 4313

(0464) (0576) (0628)

Log median income -0437 -0417 -0294

(008) (0086) (0078)

Employment-to-population ratio -0040 -0220 0244

(0261) (0260) (0240)

N 1127 1127 1127

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the natural log of state SNAPexpenditures per capita All regressions include the share of households with children and the racial shares ofthe state population

Source Annual data from Bureau of Economic Analysis national income and product accounts tables for 1990to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3543

32 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3643

Appendix C | wwwamericanprogresso

Appendix C Harkin-Miller

policy simulation results

StateMinimum wage

(2014)

SNAP enrollment (2012) Predicted change in SNAP enrollment

RateTotal

(persons)

Model 1 family

enrollment (linearprobability)

Model 2 state

enrollment (linearregression)

Model 3 s

expenditu(linear regre

Alabama $725 164 788682 -66922 -58897 -58906

Alaska $775 120 87436 -8104 -6887 -3288

Arizona $790 201 1319323 -67435 -56738 -64356

Arkansas $725 78 230489 -40977 -36063 -35248

California $800 206 7813680 -371131 -310222 -18223

Colorado $800 164 853155 -50684 -42365 -23926

Connecticut $870 91 326621 -22456 -17975 -13711

Delaware $725 186 170262 -12739 -11211 -10647

District of Columbia $825 133 84009 -5370 -4417 -3632

Florida $793 166 3208026 -195813 -164426 -13046

Georgia $725 160 1586336 -137741 -121224 -11004

Hawaii $725 96 133662 -19310 -16995 -14933

Idaho $725 92 147501 -22165 -19507 -15809

Illinois $825 95 1225084 -109088 -89742 -70955

Indiana $725 125 816233 -90818 -79928 -83985

Iowa $725 155 478011 -42716 -37594 -28556

Kansas $725 135 388269 -40082 -35275 -27461

Kentucky $725 130 568821 -60840 -53544 -52259

Louisiana $725 149 683832 -63929 -56263 -66083

Maine $750 77 101976 -16567 -14323 -15234

Maryland $725 144 846415 -81748 -71946 -38370

Massachusetts $800 130 864721 -64902 -54251 -42913

Michigan $740 146 1439141 -128801 -112140 -11022

Minnesota $725 133 713646 -74730 -65769 -37878

TABLE 10

SNAP enrollments Predicted changes by state under the Harkin-Miller bill

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3743

34 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

StateMinimum wage

(2014)

SNAP enrollment (2012) Predicted change in SNAP enrollment

RateTotal

(persons)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 state

expenditures

(linear regression)

Mississippi $725 129 386501 -41486 -36511 -46467

Missouri $750 172 1036182 -75131 -64952 -56244

Montana $790 132 132452 -10350 -8708 -5846

Nebraska $725 124 230683 -25773 -22683 -12189

Nevada $825 162 446035 -23349 -19209 -11894

New Hampshire $725 127 168404 -18359 -16157 -5735

New Jersey $825 160 1416666 -75175 -61843 -28236

New Mexico $750 149 310896 -25983 -22463 -22512

New York $800 192 3763553 -191193 -159815 -142182

North Carolina $725 174 1697193 -135417 -119179 -113503

North Dakota $725 87 61225 -9743 -8574 -4021

Ohio $795 143 1647345 -115869 -97169 -88580

Oklahoma $725 129 494053 -53006 -46650 -46854

Oregon $910 124 485326 -17036 -13328 -16398

Pennsylvania $725 161 2053643 -177315 -156052 -125586

Rhode Island $800 156 163730 -10258 -8574 -8698

South Carolina $725 94 445277 -65614 -57746 -50304

South Dakota $725 208 173749 -11586 -10197 -7458

Tennessee $725 142 914903 -89667 -78915 -99134

Texas $725 110 2863779 -362018 -318607 -253285

Utah $725 88 251107 -39658 -34902 -19390

Vermont $873 156 97792 -3823 -3055 -2475

Virginia $725 101 829771 -113723 -100086 -58212

Washington $932 72 496934 -23221 -17947 -17756

West Virginia $725 58 107875 -25792 -22699 -21665

Wisconsin $725 75 427822 -79521 -69986 -53210

Wyoming $725 164 94590 -8010 -7050 -3104

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3843

Appendix C | wwwamericanprogresso

StateMinimum wage

(2014)

SNAP expenditures (2012)Predicted change in SNAP expenditures

(millions of dollars)

Per resident

(millions of

dollars)

Total

(millions of

dollars)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 s

expenditu

(linear regre

Alabama $725 $290 $1397 -$1185 -$1043 -$1043

Alaska $775 $253 $185 -$262 -$223 -$106

Arizona $790 $257 $1686 -$935 -$787 -$892

Arkansas $725 $248 $733 -$636 -$560 -$547

California $800 $189 $7164 -$7276 -$6082 -$3573

Colorado $800 $157 $816 -$862 -$721 -$407

Connecticut $870 $191 $686 -$343 -$275 -$210

Delaware $725 $250 $229 -$205 -$180 -$171

District of Columbia $825 $366 $232 -$146 -$120 -$99

Florida $793 $294 $5676 -$4429 -$3719 -$2951

Georgia $725 $317 $3140 -$2936 -$2584 -$2346

Hawaii $725 $335 $465 -$449 -$395 -$347

Idaho $725 $225 $359 -$376 -$331 -$268

Illinois $825 $249 $3200 -$2096 -$1725 -$1364

Indiana $725 $220 $1439 -$1162 -$1023 -$1075

Iowa $725 $192 $589 -$658 -$579 -$440

Kansas $725 $159 $460 -$502 -$441 -$344

Kentucky $725 $298 $1303 -$1133 -$997 -$973

Louisiana $725 $315 $1450 -$1047 -$922 -$1083

Maine $750 $281 $373 -$267 -$231 -$246

Maryland $725 $188 $1109 -$1765 -$1553 -$828

Massachusetts $800 $206 $1366 -$1030 -$861 -$681

Michigan $740 $300 $2963 -$2400 -$2090 -$2054

Minnesota $725 $140 $755 -$1113 -$980 -$564

Mississippi $725 $326 $973 -$649 -$571 -$726

Missouri $750 $241 $1452 -$1278 -$1104 -$956

Montana $790 $190 $191 -$179 -$151 -$101

Nebraska $725 $140 $259 -$409 -$360 -$194

Nevada $825 $191 $527 -$441 -$363 -$225

New Hampshire $725 $126 $167 -$399 -$351 -$125

TABLE 11

SNAP expenditures Predicted changes by state under the Harkin-Miller bill

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3943

36 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

StateMinimum wage

(2014)

SNAP expenditures (2012)Predicted change in SNAP expenditures

(millions of dollars)

Per resident

(millions of

dollars)

Total

(millions of

dollars)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 state

expenditures

(linear regression)

New Jersey $825 $160 $1420 -$1611 -$1325 -$605

New Mexico $750 $324 $675 -$513 -$444 -$445New York $800 $287 $5616 -$3766 -$3148 -$2801

North Carolina $725 $252 $2454 -$2187 -$1925 -$1833

North Dakota $725 $128 $90 -$162 -$143 -$67

Ohio $795 $259 $2995 -$2013 -$1688 -$1539

Oklahoma $725 $248 $945 -$799 -$703 -$706

Oregon $910 $322 $1255 -$272 -$213 -$262

Pennsylvania $725 $218 $2779 -$2930 -$2579 -$2075

Rhode Island $800 $280 $294 -$173 -$144 -$147

South Carolina $725 $291 $1373 -$1337 -$1177 -$1025South Dakota $725 $198 $165 -$192 -$169 -$123

Tennessee $725 $324 $2091 -$1413 -$1243 -$1562

Texas $725 $230 $5997 -$6402 -$5634 -$4479

Utah $725 $141 $402 -$614 -$541 -$300

Vermont $873 $230 $144 -$66 -$53 -$43

Virginia $725 $173 $1413 -$2062 -$1815 -$1056

Washington $932 $244 $1682 -$350 -$270 -$267

West Virginia $725 $273 $508 -$451 -$397 -$379

Wisconsin $725 $204 $1166 -$1302 -$1146 -$871Wyoming $725 $95 $55 -$105 -$93 -$41

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4043

Appendix C | wwwamericanprogresso

FIGURE 2

Raising the minimum wage to $1010would cut taxpayer costs in every state

Predicted decreases in cost and enrollment

in SNAP in 50 states

$200+$51ndash$100

$101ndash$200

0ndash$25

$26ndash$50

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4143

38 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

Endnotes

1 US Department of Agriculture ldquoSupplementalNutrition Assistance Program (SNAP)rdquo available athttpwwwfnsusdagovsnapeligibility (last accessedFebruary 2014)

2 For this initial analysis we do not consider Harkin-Millerrsquos increase in subminimum wages for tippedworkers To do s o would increase the estimated SNAP

savings by an unknown amount

3 The Congressional Budget Office estimates thatworkers currently earning between $1010 and $1150per hour would see their wages rise under the Harkin-Miller proposal Congressional Budget O ffice ldquoTheEffects of a Minimum Wage Increase on Employmentand Family Incomerdquo (2014)

4 Marianne Page Joanne Spetz and Jane Millar ldquoDoesthe Minimum Wage Affect Welfare Caseloadsrdquo Journalof Policy Analysis and Management 24 (2) (2005)273ndash295

5 Marianne Bitler and Hilary Hoynes ldquoThe More ThingsChange the More They Stay the Same The SafetyNet Living Arrangements and Poverty in the GreatRecessionrdquo NBER Working Paper 194 49 2013

6 Sylvia Allegretto and others ldquoFast Food Poverty Wages The Public Cost of Low-Wage Jobs in the Fast-FoodIndustryrdquo (Berkeley California Center for LaborResearch and Education 2013) available at httplaborcenterberkeleyedupubliccostsfast_food_poverty_wage

7 David Neumark and William Wascher ldquoDoes a HigherMinimum Wage Enhance the Effectiveness of theEarned Income Tax Creditrdquo Industrial and LaborRelations Review 64 (4) (2011) 712ndash746

8 David Lee and Emmanuel Saez ldquoOptimal MinimumWage Policy in Competitive Labor Marketsrdquo Journal ofPublic Economics 96 (9) (2012) 739ndash749

9 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family IncomerdquoFebruary 2014

10 Jesse Rothstein ldquoIs the EITC as Good as an NITConditional Cash Transfers and Tax Incidencerdquo AmericanEconomic Journal Economic Policy 2 (1) (2010) 177ndash208

11 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family Incomerdquo

12 Dube Arindrajit 2013 rdquoMinimum Wagesand the Distribution of Family IncomerdquoUnpublished working paper Available at httpsdldropboxusercontentcomu15038936Dube_MinimumWagesFamilyIncomespdf

13 As Dube explains in ldquoThe poverty of Minimum WageFactsrdquo the simulation approach underestimate stemsfrom a number of unwarranted assumptions includingthe range of actual wage increases and the accuracy ofwage data in the Current Population Survey The causal

approach does not make these assumptions

14 Allegretto Sylvia and others 2013 ldquoCredible ResearchDesigns for Minimum Wage Studiesrdquo Working Paper148-13 University of California Berkeley Institutefor Research on Labor and Employment Available athttpwwwirleberkeleyeduworkingpapers148-13pdf

15 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family In comerdquo

16 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo (2012)

17 Ibid

18 Ibid

19 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo

20 US Department of Agriculture ldquoSupplementalNutrition Assistance Program (SNAP)rdquo

21 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo

22 Ibid

23 We deviate from the Census Bureaursquos definition ofa family unit which is ldquotwo people or more (on e ofwhom is the householder) related by bir th marriageor adoption and residing togetherrdquo US Bureau ofthe Census ldquoCurrent Population Survey DefinitionsFamilyrdquo available at httpwwwcensusgovcpsabout

cpsdefhtml (last accessed February 2014) We countas a family unit any individual residing on his or herown two or more persons residing together whodo not belong to a family in the March CPS sampleare constructed as one family in our analysis For thepurposes of food stamp allocations the consumptionresulting from this transfer is probably distributed tofamily members (rather than household members ora single individual within the household) Howeversingle individuals canmdashand domdashreceive SNAPbenefits Excluding them would fail to make theanalysis reflective of the population at large

24 Strictly the family level linear probability modelpredicts the percentage-point decrease in theprobability that an individual family will receive SNAPpayments When applied to a large number of familieshowever we are able to interpret the coefficient asa decrease in the mean of enrollmentmdashthat is a

decrease in the enrollment ratemdashby applying the lawof iterated expectations

25 We generate expenditure predictions from theenrollment modelsmdashand conversely generateenrollment predictions from the expenditure modelmdashby assuming that expenditures per enrolled familyremains the same before and after the minimum wagechange In practice this is likely to be a conservativeestimatemdashthat is to underestimate the decrease inSNAP activity Average SNAP benefits per family willalso decrease as many families that remain eligible forSNAP experience income gains

26 Wage and Hour Division ldquoMinimum Wage Laws inthe States ndash Januar y 1 2014rdquo available at httpwwwdolgovwhdminwageamericahtm (last accessedFebruary 2014)

27 See for example Sylvia Allegretto and others ldquoCredibleResearch Designs for Minimum Wage Studiesrdquo WorkingPaper 148-113 (Berkeley California Institute forResearch on Labor and Employment 2013) available athttpwwwirleberkeleyeduworkingpapers148-13pdf

28 We will report these results in a forthcoming workingpaper

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4243

Endnotes | wwwamericanprogresso

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4343

The Center for American Progress is a nonpartisan research and educational institute

dedicated to promoting a strong just and free America that ensures opportunity

for all We believe that Americans are bound together by a common commitment to

these values and we aspire to ensure that our national policies reflect these values

We work to find progressive and pragmatic solutions to significant domestic and

international problems and develop policy proposals that foster a government that

is ldquoof the people by the people and for the peoplerdquo

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 543

2 Center for American Progress | The Effects of M inim um Wages on SNAP Enro llments and Expenditures

wo decades o variaion in binding sae and ederal minimum wage changes in an

economeric ramework Our uure research will examine he effecs on SNAP

urher and apply an analogous ramework o wo oher public assisance spending

programs he Earned Income ax Credi and Medicaid

According o he finding in his repor a 10 percen increase in he minimum wagereduces SNAP enrollmen by beween 24 percen and 32 percen and reduces

program expendiures by an esimaed 19 percen aking ino accoun each

saersquos 2014 minimum wage level we apply hese resuls o he legislaive proposal

pu orward by Sen om Harkin (D-IA) and Rep George Miller (D-CA) o raise

he ederal minimum wage o $1010 per hour2 Our resuls imply ha he effecs

o he Harkin-Miller proposal on wage increases would reduce SNAP enrollmens

by beween 75 percen and 87 percen (31 million o 36 million persons) Te

oal anicipaed annual decrease in program expendiures is nearly $46 billion or

abou 6 percen o curren SNAP program expendiures

Harkin-Miller proposes o index minimum wage levels in subsequen years o he

consumer price index or CPI Te minimum wage would hen increase a he

same rae as SNAP benefi and eligibiliy levels which are also indexed o he CPI

Consequenly he savings over 10 years in 2014 dollars would be 10 imes he

one-year savings or a oal o approximaely $46 billion

Some o he reducion in SNAP program enrollmen and expendiures would

occur among workers making less han $1010 per hour991252hose whose pay would

be direcly increased by he minimum wage law Anoher par o he reducion

would occur among workers currenly earning beween $1010 and $1150 who

would also receive pay increases3

Alhough a large number o sudies have examined he impac o minimum

wage increases on earnings and employmen he impac o such minimum

wage policies on public assisance enrollmens and expendiures remains an

under-explored subjec in he economic lieraure Only a ew sudies discuss

he relaion beween he minimum wage and governmen ranser spending

much less atemp o ideniy he causal effec o one upon he oher Proessors

Marianne Paige Joanne Spez and Jane Millar find posiive effecs o minimum wage increases on welare caseloads as hey sae however heir resuls vary

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 643

Introduction and summary | wwwamericanprogress

considerably wih differen sample periods and assumpions abou sae rends4

Proessors Marianne Biler and Hilary Hoynes discuss he imporance o SNAP

as a saey ne program bu hey do no examine is relaion o minimum wage

policy5 Research economis Sylvia Allegreto and her Universiy o Caliornia a

Berkeley colleagues show ha low-wage workers in general and as-ood workers

in paricular are much more likely o be SNAP recipiens han all workers6

Several sudies have examined he relaionship beween he minimum wage and

he Earned Income ax Credi or EIC Proessor David Neumark and William

Wascher a researcher a he Federal Reserve Board o Governors find ha a

higher minimum wage increases EIC benefis or amilies in deep povery

while reducing EIC benefis or some sub-groups7 Proessors David Lee and

Emmanuel Saez argue ha he minimum wage and EIC are complemenary

policies no subsiues8 Te Congressional Budge Office or CBO argues ha

a minimum wage increase will no have a subsanial effec on EIC spending9

while Proessor Jesse Rohsein examines wheher he posiive effec o he EICon emale labor supply has lowered wages10 While hese sudies are o ineres

he EIC is quie differen rom SNAP in having a subsanial phase-in period in

which EIC benefis increase as well as a long phase-ou period wih complee

phase-ou a an annual income o abou $48000 or a amily o our quie a bi

above he reach o he minimum wage11

Research by Proessor Arindraji Dube on he causal effec o he minimum wage

on amily povery represens he sudy mos relaed o he one a hand12 Dube finds

ha Harkin-Miller would raise abou 46 million non-elderly Americans above

he ederal povery level or FPL In conras when CBO uses a simple simulaion

mehod o address he same quesion hey find ha Harkin-Miller would raise

900000 people above FPL13 Te difference beween hese wo esimaes highlighs

he imporance o underaking a causal analysis Te mehods used in his paper are

in many respecs similar o Dubersquos Moreover since eligibiliy and benefi levels or

programs such as SNAP and Medicaid are ied o he ederal povery level Dubersquos

findings have direc implicaions or his sudy Noneheless his repor appears o

be he firs sudy o examine he effecs o he minimum wage on SNAP In uure

work we plan o underake similar analyses or he EIC and Medicaid

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 743

4 Center for American Progress | The Effects of M inim um Wages on SNAP Enro llments and Expenditures

Te repor proceeds as ollows

bull Secion 1 provides background inormaion on he ederal minimum wage sae

minimum wages and he SNAP program

bullSecion 2 describes our mehods and daa

bull Secion 3 provides our main resuls including a simulaion o he effecs o a

Harkin-Miller minimum wage increase and a sae-by-sae analysis

bull Secion 4 presens our conclusions

Furher deails are provided in a series o appendices

Correction April 28 2014 Tis report incorrectly stated the potential reduction in

SNAP enrollment fom the Harkin-Miller proposal Te correct amount is 75 percentand 87 percent (31 million to 36 million persons) as stated in the reportrsquos tables

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 843

Background | wwwamericanprogress

Background

The federal minimum wage

Te ederal minimum wage was las increased in July o 2009

nearly five years ago During he pas wo decades many saes

have passed legislaion fixing he minimum wage a a higher

level han he ederal minimum Te maps in Figure 1 show ha

while saes in every region o he Unied Saes have adoped

higher minimum wages hey are no disribued randomly bygeography As shown in he 2013 sudy ldquoCredible Research

Designs or Minimum Wage Sudiesrdquo by economiss Sylvia

Allegreto Arindraji Dube Michael Reich and Ben Zipperer

hese saes vary sysemaically rom he oher saes by a number

o characerisics ha affec low-wage employmen rends bu

which are no hemselves relaed o minimum wage policy14

Te nonrandom patern o minimum wage adopion has

imporan implicaions or obaining unbiased esimaes o

minimum wages on employmen In paricular naional panel

sudies ha use sae and ime fixed effec models991252such as

a 1992 sudy by David Neumark and William Wascher991252

spuriously esimae negaive employmen effecs Te reason

or his resul is uncovered using ess or pre-rends Tese

ess find ha low-wage employmen was already declining wo

years beore minimum wages were implemened By making a

saisically large number o local comparisons ha conrol or

heerogeneiy among saes and by ime eliminaes his pre-

rend For his reason we conduc similar ess or our SNAPoucomes and use model specificaions ha include local

comparisons as in he sudy cied above

FIGURE 1

High versus low minimum wage stat

from 1990 to 2012

Means and variances

More than $533

Average minimum wage over 1990ndash2012

Less than or equal to $533

More than $121

Minimum wage variance over 1990ndash2012

Less than or equal to $121

Notes State means and variances were calculated using annual stat

minimum wage data from 1990 to 2012 The shading on the maps

partitions the states into above- and below-median values

Source Sylvia Allegretto and others ldquoCredible Research Designs forMinimum Wage Studiesrdquo Working Paper 148-13 (Institute for Resear

on Labor and Employment 2013) available at httpwwwirleberkeleyeduworkingpapers148-13pdf

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 943

6 Center for American Progress | The Effects of M inim um Wages on SNAP Enro llments and Expenditures

Te CBO recenly projeced ha in 2016 17 million workers will earn less han

he $1010 hourly wage proposed in he Harkin-Miller bill Furhermore he

CBO esimaes ha an addiional 8 million workers earned beween $1010 and

$1150 per hour and were also likely o experience a wage increase15

Supplemental Nutrition Assistance Program

We ocus our inquiry on SNAP Benefis under he program are enirely

ederally unded he program is adminisered by he US Deparmen o

Agriculure ogeher wih sae agencies which share in adminisraion

coss Spending on SNAP has grown in he pas decade reaching $78 bil lion

in 2011 wih SNAP enrollmen increasing o 45 million people abou one-

sevenh o he US populaion16 According o he CBO changes since 1990 in

SNAP spending and enrollmens are primarily he resul o cyclical economic

condiions noably changes in he unemploymen rae and changes in per capiaincome17 Te 2009 American Recovery and Reinvesmen Ac emporarily

increased SNAP benefi amouns by 136 percen as repored by he US

Deparmen o Agriculurersquos Food and Nuriion Service hese higher benefi

levels expired on November 1 201318 Te CBO esimaes ha abou wo-

hirds o he changes in SNAP expendiure are associaed wih changes in he

number o recipiens and one-hird wih changes in he benefis received when

recipiensrsquo incomes change19

In fiscal year 2014 SNAPrsquos maximum monhly benefis are $189 or a single

individual $497 or a amily o hree and $750 or a amily o five Benefis are

reduced by 30 cens per dollar received and phase ou enirely a gross monhly

household incomes o 130 percen o he ederal povery level $1245 or a

single individual $2116 or a amily o hree and $2987 or a amily o five

o deermine benefis SNAP also defines a ne monhly income concep and

ses benefis a 100 percen o he ederal povery level using his concep

Calculaion o ne monhly income can include cerain deducions rom

monhly gross income such as medical expenses and child care coss Alhough

saes are permited some laiude on wha deducions are allowed in pracice

hese vary by very small amouns Our saisical model akes accoun o sae-specific differences in benefis20

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1043

Background | wwwamericanprogress

SNAP imposes an employmen or work-raining requiremen or able-bodied

prime-age aduls991252hose beween he ages o 18 and 50 and wihou disabiliies

or dependen children Such households can receive only hree monhs o benefis

in a hree-year period In recen years abou 85 percen o households receiving

benefis have incomes below he ederal povery level 49 percen have dependen

children 16 percen are age 60 or older 20 percen are disabled and 30 percenrepor some earned income21

A 2012 CBO repor also noes ha ake-up raes among eligible SNAP recipiens

average abou 70 percen wih much lower ake-up among elderly households

Te ake-up rae increases in harder economic imes I also increased when sigma

issues were reduced as SNAP debi cards replaced acual ood samps ake-up

is especially high among hose mos needy Adminisraive spending equaled 91

percen o he poenial spending ha would have occurred i all eligible recipiens

were enrolled Alhough some SNAP policy changes have occurred since 1990

mos were relaively minor and all were naional in scope Te 1996 welare reorm bill eliminaed SNAP eligibiliy or some legal immigrans limied he ime lengh

o eligibiliy or able-bodied childless aduls and reduced maximum benefis

Some o hese resricions were relaxed in 2002 and again in he American

Recovery and Reinvesmen Ac in 200922

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1143

8 Center for American Progress | The Effects of M inim um Wages on SNAP Enro llments and Expenditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1243

Methods and data | wwwamericanprogress

Methods and data

As previously menioned we exploi variaion in minimum wages by sae and ime

o examine heir causal effecs on SNAP enrollmens and expendiures o do so we

merge daa rom 1990 hrough 2012 drawn rom he Annual Social and Economic

Supplemen o he Curren Populaion Survey991252an annual Census Bureau survey

commonly known as he March CPS ha includes23 inormaion on SNAP

enrollmens a he amily level991252wih sae-level daa on minimum wages SNAP

expendiures populaion unemploymen raes and sae median income levels

o conrol or ime-varying heerogeneiy among saes our specificaions includeconrols or sae linear rends and effecs by Census division and ime We esimae

effecs a wo levels allowing or amily variaion and allowing only or sae-level

variaion We also employ a se o sandard demographic conrols such as amily size

and composiion and race and ehnic composiion

Distinguishing causation from correlation

How can we ensure ha our analysis does no pick up a spurious correlaion or

example he endency o more economically vibran saes o implemen higher

minimum wages Disinguishing correlaion and policy endogeneiy rom rue

causal effecs is he primary moivaion or economeric analysis In he ideal

experimen researchers would begin wih wo saes991252ha are alike in every

respec prior o he policy991252 and ldquoreardquo only one o hese saes wih a higher

minimum wage Tey would atemp o shield hese saes rom any influence ha

could obscure heir undersanding o he minimum wagersquos direc effec on SNAP

enrollmen Researchers o course canno conduc such experimens

We can however use saisical mehods o conrol simulaneously heindependen effecs on SNAP o sae unemploymen raes sae income levels

and common rajecories among saes wihin he same Census division By

ensuring similariy along hese dimensions we maximize he likelihood ha

SNAP aciviy in wo saes would have comparable oucomes in he absence

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1343

10 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

o a minimum wage policy change Tus i a new minimum wage policy were

implemened in one sae only he researchers could atribue all o he difference

hey observe in SNAP aciviy o he new minimum wage policy

In oher words we approximae he ideal experimen by using non-experimenal

saisical mehods Te desirable ldquopre-exising similariiesrdquo beween saes ha wehave defined above inorm our choice o conrol variables in a saisical seting

More precisely in our muliple regression models we use median amily income

he unemploymen rae he employmen-o-populaion raio and regional and

ime idenifiers o consruc an appropriae group o peers or each sae on he

eve o a policy change

Data description

wo daa ses include inormaion abou boh income and paricipaion inpublic programs Te Survey o Income and Program Paricipaion or SIPP

which is conduced in inermiten years has he advanage o ollowing he

same individuals over a period o ime In oher words i is a longiudinal daa

se I also has he advanage o conaining monhly daa However he sample

size o he SIPP is no sufficien or analyzing variaions in sae-level minimum

wages Te March CPS has he advanage o a much larger sample size and i is

conduced annually wihou any breaks in ime I has he disadvanage o being

a cross-secional daa se so we canno ollow he same individuals over ime991252

sricly speaking over more han one year On ne he March CPS is much more

suiable or our sudy We examine he empirical relaionship beween minimum

wage policy and ood samp aciviy a wo levels o aggregaion he amily level

and he sae level Family-level daa are drawn rom he March CPS

Te March CPS comprises responses rom he residens o 50000 o 60000

dwelling places surveyed per year and conains deailed inormaion on he

residensrsquo employmen and income including income rom ranser paymens

Te sample or our analysis comprises more han 128 million amily unis during

he period rom 1990 o 2012 (inclusive) Survey weighs allow us o analyze

SNAP paricipaion in a manner ha is represenaive o he US populaion alarge Over all years he share o amilies reporing ood samp receip in he

weighed March CPS sample is 91 percen Te enrollmen rae was a a low o 6

percen in he year 2000 In 2012 he mos recen year in our panel 133 percen

o amilies repored paricipaing in SNAP a some poin during he survey year

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1443

Methods and data | wwwamericanprogresso

Te March CPS also collecs inormaion on he number o SNAP recipiens in

he household he number o monhs he household receives SNAP benefis

and he cash-equivalen value o he SNAP benefis received However he

value o SNAP benefis is severely underrepored among recipiens perhaps

because recipiens are unaware o he exac monhly cash-equivalen value o

benefis hey receive

Our firs empirical sraegy ocuses solely on SNAP enrollmen By using he

amily as he uni o analysis we are able o inser saisical conrols o accoun

or non-wage-relaed acors ha influence any paricular amilyrsquos likelihood o

program paricipaion wih he inenion o isolaing any differences in program

paricipaion ha are due purely o changes in wage policy Tis approach

idenifies he effecs o low-wage labor policy on he exernal margin991252ha is

he effec o he minimum wage on he likelihood ha a amily paricipaes in he

SNAP program a all991252as opposed o he inernal margin or how much SNAP

unding he amily would receive

Our second empirical ramework uses sae-level adminisraive daa Ta is

we aggregae he daa o obain a single daa poin or each saeyear back

o 1990 represening he mean o he oucome or he sae Te sae-level

esimaion serves as a robusness check on he amily-level resuls or SNAP

paricipaion Also using aggregaed daa allows us o esimae direcly he

causal effec o minimum wage changes on SNAP spending Tis is no possible

a he amily level as discussed above daa on cash-equivalen value o ood

samps or SNAP recipiens is very requenly no repored in he March CPS

and when i is repored he inormaion may be unreliable By conras he

Bureau o Economic Analysis publishes aggregae SNAP spending a he sae

level in is Naional Income and Produc Accoun or NIPA ables Tus

while we are unable o observe he heerogeneiy in he cash value o SNAP

or amilies in each sae we are able o calculae average SNAP spending

per residen in each sae per year Supporing covariaes include he annual

unemploymen and employmen daa rom he Bureau o Labor Saisics or

BLS and sae-level populaion series rom he iner-decennial census releases

Minimum wage daa are available rom he BLSrsquos wages and hours division For

sae minimum wage changes enaced a oher imes han he firs o he yearan average value or he year is used

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1543

12 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

Family level model specifications

We firs examine he effec o he minimum wage on paricipaion in public

assisance programs For amily residing in sae and during year we esimae

an equaion o he ollowing orm

(1)

is a binary variable ha is se equal o 1 i a leas one member o amily

received ood samps during he survey year is a se o sae-level

characerisics including annual averages o he unemploymen rae he

employmen-o-populaion raio and he naural log o median amily income

is a vecor o amily atribues including indicaors or he race and marial saus

o he amily head size o he amily he presence o children and he presence oan adul male Sae fixed effecs are capured by o conrol or ime-varying

heerogeneiy our preerred model specificaion also includes year fixed effecs

ha vary by Census division ( ) and sae-level linear ime rends In

Appendix B we jusiy he inclusion o hese las wo erms We also compare he

resuls rom our preerred specificaion wih less sauraed specificaions

Te effec o ineres which is capured by is he expeced change in he

probabiliy o receiving SNAP benefis wih respec o a change in he (log o he)

binding minimum wage in sae during year We repor robus sandard errors

clusered a he sae level We esimae he parameers using linear regression

producing a linear probabiliy model Deails o he model selecion process are

covered in Appendix B below

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1643

Methods and data | wwwamericanprogresso

State-level model specifications

Te sae-level models are similarly specified For sae in year we assume ha

(2)

In his model is now eiher he SNAP enrollmen rae in sae during year

or he naural logarihm o per capia SNAP expendiures in sae during

year is once again a se o sae-level characerisics including he same

sae-level covariaes as in he amily regressions (annual average unemploymen

rae employmen-o-populaion raio naural log o median amily income)

wih he addiion o amily level characerisics averaged across he sae (average

amily size and he shares o populaion consiued by each o five racialehnicgroups) Sae fixed effecs are represened by As above our preerred model

specificaion includes year fixed effecs ha vary by Census division ( ) and

sae-level linear ime rends as elaboraed in Appendix B Te effec o

ineres is capured by

We esimae boh sae-level models (enrollmen and expendiures) using

ordinary leas squares regression Tus he inerpreaion o he coefficien is no

longer ha o a change in probabiliy as in he binary oucome models described

above Raher or he sae-level SNAP enrollmen model represens he

expeced change (in percenage poins) in he saersquos SNAP enrollmen rae ha

is due o a 1 percen change in he minimum wage For he SNAP expendiures

model is simply he elasiciy o SNAP spending wih respec o he minimum

wage991252ha is he percenage change in sae expendiures expeced o resul rom

a 1 percen change in ha saersquos minimum wage For urher deails on model

specificaion reer o Appendix B below

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1743

14 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1843

Results | wwwamericanprogresso

Results

Estimated minimum wage effects on SNAP enrollment and

expenditures

able 1 shows he esimaed parameer o ineres991252he coefficien o he

minimum wage or he preerred model o each ype Coefficiens on he

minimum wage variable are no direcly comparable across models because all

our models have a differen uncional orm o undersand and compare hese

esimaes we compue he change in SNAP aciviy prediced or a paricular wage scenario Te final column in able 1 answers he quesion Wha would be

he expeced change in SNAP aciviy in response o a 10 percen increase in he

minimum wage Te answer o his quesion varies wih he value o he inpu

parameers in he able we calculae he percenage decrease in enrollmen or

expendiures prediced or he average sae wih a minimum wage o $725 in

2014 Te sae-level SNAP expendiure model which is a consan-elasiciy

model conveys elasiciy inormaion direcly or he change in expendiures per

capia in he sae

TABLE 1

Comparison of national SNAP predictions for a 10 percent increase in the federal minimum wage

Model LevelRegression

type

Predicted outcome Coefficient of log

(minimum wage)

(Standard error)

Effect of a 10 percent in

in the minimum wage

VariableForm of

variable

Total

enrollment

Tot

expend

1 Family Linear

probability Enrollment Binary (enrolled=1)

-0042

(0008) -317 N

2

StateLinear regression

(ordinary least

squares)

Enrollment State enrollment rate (percent) -0031

(0012) -235 N

3 Expenditures Log (state expenditures per capita) -0190

(0103) NA -19

plt01 plt005 plt001Note Predicted changes are calculated for the average state with a minimum wage of $725 in 2014

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1943

16 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

According o his model a 10 percen increase in he minimum wage would resul

in a 19 percen reducion in SNAP expendiures Tis resul is reassuringly similar

o Dubersquos resul or povery reducion Te wo enrollmen models are much more

precisely esimaed han he expendiure model Te sae-level enrollmen model

finds ha a 10 percen minimum wage increase in a low-wage sae is associaed wih

a 235 percen decrease in SNAP enrollmens Te amily-level linear probabiliymodel predics a somewha greaer elasiciy or low minimum wage saes an

increase o 10 percen in he ederal minimum would resul in a 317 percen decline

in SNAP enrollmen24 Te differences in hese esimaes sem rom a number o

acors including difference in model uncional orm and daa used We rea his

range o elasiciy esimaes as an upper and lower bound on enrollmen impacs

Harkin-Miller bill National and state-level predicted impacts

Wha would be he prediced change or he SNAP program i he ederalminimum were raised o $1010 as proposed in he Harkin-Miller bill In order o

make his inerence we accoun or he ac ha no all saes are currenly subjec

o he ederal minimum wage a he beginning o 2014 21 saes mainained

higher minimum wages han $725 In hose saes an increase in he ederal

minimum wage may or may no be binding or employers in he sae depending

upon wheher he new ederal minimum exceeds he sae-level minimum Bu

regardless o wheher a minimum wage change is binding he impac on SNAP

aciviy will be lower in high minimum wage saes In order o accoun or his

properly we calculae sae by sae he percenage wage change ha would resul

rom he Harkin-Miller proposal and apply he parameers rom each o he hree

models above o compue he expeced decrease in SNAP aciviy or each sae

In his exercise we use saesrsquo curren (2014) minimum wage levels and assume

as a baseline he 2012 levels o SNAP enrollmen and expendiure as 2012 is he

mos recen year or which SNAP daa are available

able 10 and able 11 in Appendix C repor he esimaed effecs on SNAP

enrollmen and expendiures respecively or each sae under he Harkin-Miller

bill25 An increase o $1010 i enaced oday would represen beween a 393

percen wage increase in a $725 minimum wage sae and an 84 percen increasein Washingon sae which has he highes minimum wage in he naion a $932

as o January 201426 Slighly more han 56 percen o he decrease in expendiures

and abou 59 percen o he decrease in enrollmen would occur in saes wih

presen-day minimum wages o $725

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2043

Results | wwwamericanprogresso

In 2012 hese saes were home o 46 percen o he American populaion

and accouned or a roughly equivalen percenage o oal naional SNAP

expendiures No surprisingly he larges enrollmen decreases come rom

populous saes wih very high SNAP enrollmen raes andor very low minimum

wages Te larges prediced enrollmen reducion991252beween 319000 individuals

and 362000 individuals991252would occur in exas which has a $725 minimum wage In Caliornia which has a 206 percen SNAP paricipaion rae and an $8

per hour minimum wage we anicipae SNAP enrollmen declines o 310000

persons o 371000 persons And in Florida which had a SNAP paricipaion rae

o 166 percen and a $793 minimum wage enrollmen could decline by beween

164000 individuals and 196000 individuals For he our saes ha ogeher

accouned or he greaes amoun o SNAP spending in 2012991252exas Caliornia

Florida and New York respecively991252he combined expendiure reducion rom

he Harkin-Miller bill is prediced o be $14 billion

able 2 summarizes he prediced declines in SNAP aciviy or he naion as a whole ha would resul rom he direc and indirec effecs o he Harkin-Miller

bill Enrollmen would all beween 31 million persons and 36 million persons

represening 75 percen o 87 percen o curren enrollmen Te anicipaed

reducion in program expendiures would be nearly $46 billion or 61 percen o

program expendiures

TABLE 2

Comparison of national SNAP predictions under the Harkin-Miller billrsquos $1010 minimum wage

Model

Enrollment

(persons)

Expenditures

(millions of dollars)

Current (2012) Predicted Change Current (2012) Predicted Chan

Family enrollment (linear probability)

41866195

45489339 -3623144

$74861

NA NA

State enrollment (ordinary least squares) 38745435 -3120759 NA NA

State expenditures (ordinary least squares) NA NA $70305 -$45

Note Calculations use 2014 state minimum wages and the most recent SNAP data from 2012 They assume that per-enrollee expenditures remain constant

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2143

18 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

Tere are o course oher possibiliies or minimum wage legislaion able 3

shows he expeced SNAP changes or he Unied Saes under a variey o wage

scenarios calculaed using he sae-level models I saes were no able o se

heir minimum wages independenly such ha all saes were consrained by

he ederal minimum o $725 SNAP would be received by abou 514000 more

people across he Unied Saes a an addiional program cos o nearly hree-quarers o a billion dollars In conras he effecs o a higher minimum wage

proposal991252a ederal wage floor o $11 per hour991252would decrease enrollmen in

SNAP by more han 10 percen and decrease program coss by 83 percen

TABLE 3

Summary of par ticipation and expenditures under wage scenarios

If all states had

minimum wages of

Enrollment(persons)

Expenditures

(millions of dollars)

Predicted Change Predicted Change

Recent levels (2014) 41866195 $74861

$725 42380520 514326 $75604 $743

$800 41423919 -442276 $74209 -$652

$900 40148451 -1717744 $72350 -$2511

$1000 38872982 -2993212 $70490 -$4371

$1010 38745435 -3120759 $70305 -$4556

$1100 37597514 -4268681 $68631 -$6230

Note Calculations use state-level enrollment model coefficient

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2243

Conclusion | wwwamericanprogresso

Conclusion

An exensive body o lieraure examines employmen effecs o he minimum

wage A much smaller se o sudies examines how he minimum wage affecs

povery and only a handul o papers examine he effecs o he minimum wage

on he EIC Our analysis is he firs o examine he effecs o he minimum

wage on SNAP

Our findings indicae ha increased earnings rom minimum wage changes do

reduce SNAP enrollmens and expendiures We esimae ha he Harkin-Miller bill would save axpayers nearly $46 billion per year equivalen o 61 percen

o SNAP expendiures in 2012 he las year or which daa are available Over a

10-year period he esimaed savings amoun o nearly $46 billion

Our repor is subjec o limiaions ha we expec o overcome in our uure

research Firs he findings do no ake ino accoun possible ineracions among

SNAP he EIC and Medicaid Te eligibiliy cuoffs among hese programs

are quie differen suggesing ha such ineracions may be minor Noneheless

he join effecs can only be deermined by urher research using a causal

model Second i would be useul o know he disribuion o SNAP reducions

along he wage disribuion Using he Congressional Budge Officersquos calculaions

o how much he oal dollar value o wage would increase under he Harkin-

Miller proposal our findings imply ha he decline in overall SNAP spending

equals abou 15 percen o he oal resuling increase in wages Te amoun and

disribuion o his offse are o considerable ineres Minimum wage beneficiaries

who come rom working amilies already well above he povery line would no

see any offse while hose who are currenly considerably below he povery line

will see larger offses Tese issues will also be a subjec or our uure research

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2343

20 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

About the authors

Rachel West is a maser o public policy candidae a he Goldman School

o Public Policy Universiy o Caliornia Berkeley Her research ocuses on

economic policy in he areas o low-wage labor and povery

Michael Reich is proessor o economics and direcor o he Insiue or

Research on Labor and Employmen a he Universiy o Caliornia a Berkeley

His research publicaions cover numerous areas o labor economics including

racial inequaliy labor marke segmenaion high-perormance workplaces

union-managemen cooperaion Japanese labor-managemen sysems living

wages and minimum wages He received his docorae in economics rom

Harvard Universiy

Acknowledgments

We are graeul o Sylvia Allegreto Arindraji Dube Bill Leser Jesse Rohsein

Daniel Tompson and Ben Zipperer or heir valuable suggesions

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2443

References | wwwamericanprogresso

References

Allegreto Sylvia and ohers 2013 ldquoFas Food Povery Wages Te Public Cos o Low-Wage Jobsin he Fas-Food Indusryrdquo Universiy o Caliornia Berkeley Cener or Labor Research andEducaion Available a htplaborcenerberkeleyedupubliccossas_ood_povery_wage

Allegreto Sylvia and ohers 2013 ldquoCredible Research Designs or Minimum Wage Sudiesrdquo

Working Paper 148-13 Universiy o Caliornia Berkeley Insiue or Research on Labor andEmploymen Available a htpwwwirleberkeleyeduworkingpapers148-13pd

Biler Marianne and Hilary Hoynes 2013 ldquo Te More Tings Change he More Tey Say heSame Te Saey Ne Living Arrangemens and Povery in he Grea Recessionrdquo Working Paper19449 Naional Bureau o Economic Research

Congressional Budge Office 2012 ldquoTe Supplemenal Nuriion Assisance Programrdquo Washingon Available a htpwwwcbogovsiesdeaulfilescbofilesatachmens04-19-SNAPpd

991252 991252 991252 2014 ldquoTe Effec o a Minimum-Wage Increase on Employmen and Family Incomerdquo Washingon Available a htpwwwcbogovsiesdeaulfilescbofilesatachmens44995-MinimumWagepd

Dube Arindraji 2013 rdquoMinimum Wages and he Disribuion o Family Incomerdquo Unpublished working paper Available a htpsdldropboxuserconencomu15038936Dube_ MinimumWagesFamilyIncomespd

991252 991252 991252 2014 ldquoTe Povery o Minimum Wage Facsrdquo Arindraji Dube Blog January 22 Availablea htparindubecom20140122he-povery-o-minimum-wage-acs

Lee David and Emmanuel Saez 2012 ldquoOpimal Minimum Wage Policy in Compeiive LaborMarkesrdquo Journal o Public Economics 96 (9) 739ndash749

Neumark David and William Wascher 1992 ldquoEmploymen Effecs o Minimum and Subminimum Wages Panel Daa on Sae Minimum Wage Lawsrdquo Industrial and Labor Relations Review 46 (1)

55ndash81

Neumark David and William Wascher 2011 ldquoDoes a Higher Minimum Wage Enhance heEffeciveness o he Earned Income ax Credirdquo Industrial and Labor Relations Review 64 (4)712ndash746

Page Marianne Joanne Spez and Jane Millar 2005 ldquoDoes he Minimum Wage Affec WelareCaseloadsrdquo Journal o Policy Analysis and Management 24 (2) 273ndash295

Rohsein Jesse 2010 ldquoIs he EIC as Good as an NI Condiional Cash ransers and ax

Incidencerdquo American Economic Journal Economic Policy 2 (1) 177ndash208

Wage and Hours Division 2014 ldquoMinimum Wage Laws in he Saes ndash January 1 2014rdquo USDeparmen o Labor (htpwwwdolgovwhdminwageamericahm [February 2014])

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2543

22 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2643

Appendix A | wwwamericanprogresso

Appendix A

Pre-trend falsification check

Recen minimum wage research27 highlighs a common flaw in previous sudies

ailure o veriy ha he oucome variable is ree o negaive pre-exising rends I

or example SNAP aciviy was already rending down in saes ha raised heir

minimum wages beore hese changes came ino effec our regression analysis could

(misakenly) atribue ha reducion o he minimum wage We check or such

pre-rends by inroducing variables ha represen he prior yearrsquos value or leads o

he minimum wage I he model esimaes he minimum wage o have an effec on

he oucome variable beore he wage change wen ino effec hen an unobservedacor no he minimum wage change caused he change in SNAP aciviy

We es he specificaions above or pre-rends by including a one-year lead in

all hree specificaions We find ha he lead erms are small posiive and no

saisically significan indicaing ha he concurren minimum wage991252no

he wage level in prior periods991252is driving he observed changes in SNAP

oucomes28 In paricular he coefficien (sandard error) on he lead erm in

our preerred amily-level enrollmen regression is 011 and no significan

while he coefficien and sandard error o he conemporaneous minimum

wage is unchanged In he sae-level preerred enrollmen regression he

coefficien o he lead erm is again small (07) and i is no significan Te

corresponding coefficien on he lead erm in he sae-level expendiure

regression is 16 and is no significan Te posiive poin esimaes on hese lead

erms resuls no only rule ou disoring negaive pre-rends Tey also sugges

ha our main resuls may underesimae he rue effecs

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2743

24 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2843

Appendix B | wwwamericanprogresso

Appendix B

Model selection process

For boh he amily-level and sae-level models we es hree mehods o conrol

or unobserved geographic- and ime-varying characerisics as suggesed by he

minimum wage lieraure Firs we include only independen sae-specific fixed

effecs and year-specific fixed effecs Tis specificaion (specificaion 1) implicily

assumes ha amilies in any sae consiue an equally good saisical ldquoconrolrdquo

group or hose in any randomly chosen sae afer accouning or various

characerisics (median income and unemploymen rae among ohers) Similarly

simple ime fixed effecs assume ha amilies surveyed in any year can crediblyserve as a conrol group or amilies surveyed in every oher year o he sample

(1990 hrough 2012)

In oher words specificaion 1 assumes ha a saersquos immediae neighbor provides

no beter a couneracual or he effec o a minimum wage change han does a

sae across he counry We relax his resricive specificaion sequenially in wo

seps In specificaion 2 we replace simple year fixed effecs wih fixed effecs or

each Census divisionyear (capured as an addiional variable in he vecor By

using division-year effecs we remove he resricion ha amilies in each sae

are equally good saisical conrols or all oher amilies Raher we allow or he

possibiliy ha amilies in similar geographic regions (or example he Souh or

he Norheas) may be more similar o one anoher han amilies arher away

Finally in specificaion 3 we add sae-specific linear ime rends o he previous

specificaion Tus specificaion 3 is he mos rigorous model specificaion in ha

i allows or heerogeneiy along hree dimensions Ta is specificaion 3 allows

each sae o have is own ime-varying rends raher han imposing he resricion

ha saes evolve idenically over he 22 years in he sample

We begin building he heoreical specificaion above rom a se o simpleuncondiional models regression o SNAP aciviy (enrollmen or expendiures)

on he log o he minimum wage and a se o geographic- and ime-specific

effecs (specificaions 1 2 and 3 described above) As shown in ables 1ndash3 (or

specificaion 3) we hen add covariaes sequenially o hese models including

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2943

26 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

firs he vecor o amily-level conrols ollowed by each o several sae-level

covariaes in urn he unemploymen rae log o median-amily income and he

employmen-o-populaion raio Comparable resuls or specificaions 2 and 3

will be available in our orhcoming working paper

As expeced he simple uncondiional models indicae ha he relaionship beween he minimum wage and SNAP enrollmen i one exiss is a more complex

one influenced by oher acors In he uncondiional model he coefficien on he

variable o ineres991252he log o he minimum wage991252is small in magniude and no

saisically differen rom zero Once we accoun or he influence o labor marke

condiions and variaion in income levels on program paricipaion (by including

unemploymen rae and median-amily income conrol variables respecively)

he effec o he minimum wage on SNAP enrollmen is precisely esimaed Te

coefficien o he log minimum wage is slighly higher (-0042) in he amily-level

analysis han he coefficien (-031) in he sae-level analysis Te level o precision

is also higher in he amily-level analysis Tis is o be expeced when using 124million observaions compared o 1127

Te second dimension o model choice concerns he effec specificaion ables

7ndash9 compare he primary coefficiens o ineres or he SNAP enrollmen and

expendiure models For boh he enrollmen models he effec sizes are smalles

or specificaion 1 larges or specificaion 2 and inermediae beween hese wo

in specificaion 3 Recall ha Specificaion 3 conains sae-specific linear ime

rends in addiion o he census divisionyear conrols included in specificaion

2 In he amily-level enrollmen model he sandard error o he minimum wage

coefficien is smaller han in he oher wo specificaions Sandard errors on he

oher variables are much smaller in specificaions 2 and 3 han in specificaion

1 On he basis o coefficien significance (join and individual) specificaions 2

and 3 are sricly preerred in boh enrollmen models o specificaion 1 which

conains only sae and year fixed effecs

A concern wih specificaions 2 and 3 is ha rend conrols such as sae linear

rends may incorrecly absorb some o he delayed impac o a minimum wage

When we es his issue by including lagged minimum wages we do no find ha

delayed effecs are significan Anoher concern is ha more sauraed modelsuse less o he saisical variaion which could reduce he saisical power o

he resuls However he sandard errors or our more sauraed models are no

higher and are lower in some cases han or he less sauraed models Overall

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3043

Appendix B | wwwamericanprogresso

his evidence suppors our use o he sauraed model as he preerred model

specificaion Moreover Dubersquos 2013 sudy shows ha more sauraed models

perorm beter han models wih jus sae and ime fixed effecs

Te esimaed enrollmen regressions a boh he amily and sae levels show large

and saisically significan coefficiens Te esimaed minimum wage effec in heexpendiures regressions991252or which we have only sae-level daa991252is also large

and saisically significan

We do no use weighed regression or he sae-level models preerring o keep

analysis o he ldquoreamenrdquo (ha is o say a minimum wage change) appropriae

o he average sae raher han he average amily or individual I insead our

primary ineres were he impac o a minimum wage change on he average amily

or he average individual we migh choose o designae he number o amilies

in each sae or he sae populaion respecively as analyic weighs in order o

obain a coefficien beter suied or such inerence

TABLE 4

SNAP enrollment

Family-level linear probability

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0048 -0047 -0040 -0043 -0042

(0013) (0013) (001) (0008) (0009)

Unemployment rate (100) 0505 0420 0280

(0083) (0086) (0082)

Log median income -0057 -0039

(0011) (0011)

Employment-to-population ratio -0239

(0038)

N 1242022 1242022 1242022 1242022 1242022

plt01 plt005 plt001

Note Robust standard errors are in parentheses Observations are clustered at the state level The outcome variable is binary and equal to one if a family is enrolledin SNAP All models include state fixed effects Census division x-year fixed effects and state-specific linear time trends All specifications except 3a include additional

controls for family size race and marital status of the family head presence of children and presence of an adult maleSource Annual data from the March Current Population Survey for 1990 to 2012 estimation includes Current Population Survey probability weights

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3143

28 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

TABLE 5

SNAP enrollment

State-level linear regression

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0028 -0024 -0026 -0031 -0031

(0014) (0013) (0013) (0012) (0012)

Unemployment rate (100) 0625 0466 0320 0339

(0087) (0088) (0085) (0083)

Log median income -0090 -0065 -0061

(0013) (0013) (0013)

Employment-to-population ratio -0282 -0248

(0037) (0038)

N 1127 1127 1127 1127 1127

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the SNAP enrollment rate All models include state fixed effects Census division x-year fixedeffects and state-specific linear time trends

Source Annual data from the March Current Population Survey for 1990 to 2012

TABLE 6

SNAP expenditures

State-level linear regression

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0181 -0149 -0156 -0153 -0190

(011) (0103) (0102) (0103) (0103)

Unemployment rate (100) 4587 4152 4232 4313

(0622) (0621) (0633) (0628)

Log median income -0246 -0261 -0294

(0075) (0078) (0078)

Employment-to-population ratio 0155 0244

(0237) (024)

N 1127 1127 1127 1127 1127

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the natural log of state SNAP expenditures per capita for 1990 to 2012 All models include state

fixed effects Census division x-year fixed effects and state-specific linear time trends

Source Annual data from Bureau of Economic Analysis national income and product accounts tables for 1990 to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3243

Appendix B | wwwamericanprogresso

TABLE 7

Comparison of specifications SNAP enrollment

Family-level linear probability

(1) (2) (3)

Log minimum wage -0028 -0049 -0042

(0014) (0017) (0009)

Unemployment rate (100) 0275 0297 0280

(0161) (0076) (0082)

Log median income -0077 -0055 -0039

(0014) (0012) (0011)

Employment-to-population ratio -0238 -0250 -0239

(0054) (004) (0038)

N 1242022 1242022 1242022

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses Observations are clustered at the state level The outcomevariable is binary or equal to one if a family is enrolled in SNAP All specifications include additional controls forfamily size race and marital status of the family head presence of children and presence of an adult male

Source Annual data from the March Current Population Survey for 1990 to 2012 estimation includes CurrentPopulation Survey probability weights

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3343

30 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

TABLE 8

Comparison of specifications SNAP enrollment

State-level linear regression

(1) (2) (3)

Log minimum wage -0019 -0035 -0031

(0009) (0012) (0012)

Unemployment rate (100) 0401 0370 0339

(0063) (0077) (0083)

Log median income -0081 -0073 -0061

(0011) (0013) (0013)

Employment-to-population ratio -0183 -0222 -0248

(0039) (0039) (0038)

N 1127 1127 1127

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the SNAP enrollment rate Allregressions include the share of households with children and the racial shares of the population

Source Annual data from the March Current Population Survey for 1990 to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3443

Appendix B | wwwamericanprogresso

TABLE 9

Comparison of specifications SNAP expenditures

State-level linear regression

(1) (2) (3)

Log minimum wage -0121 -0203 -0190

(0075) (0103) (0103)

Unemployment rate (100) 5292 5152 4313

(0464) (0576) (0628)

Log median income -0437 -0417 -0294

(008) (0086) (0078)

Employment-to-population ratio -0040 -0220 0244

(0261) (0260) (0240)

N 1127 1127 1127

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the natural log of state SNAPexpenditures per capita All regressions include the share of households with children and the racial shares ofthe state population

Source Annual data from Bureau of Economic Analysis national income and product accounts tables for 1990to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3543

32 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3643

Appendix C | wwwamericanprogresso

Appendix C Harkin-Miller

policy simulation results

StateMinimum wage

(2014)

SNAP enrollment (2012) Predicted change in SNAP enrollment

RateTotal

(persons)

Model 1 family

enrollment (linearprobability)

Model 2 state

enrollment (linearregression)

Model 3 s

expenditu(linear regre

Alabama $725 164 788682 -66922 -58897 -58906

Alaska $775 120 87436 -8104 -6887 -3288

Arizona $790 201 1319323 -67435 -56738 -64356

Arkansas $725 78 230489 -40977 -36063 -35248

California $800 206 7813680 -371131 -310222 -18223

Colorado $800 164 853155 -50684 -42365 -23926

Connecticut $870 91 326621 -22456 -17975 -13711

Delaware $725 186 170262 -12739 -11211 -10647

District of Columbia $825 133 84009 -5370 -4417 -3632

Florida $793 166 3208026 -195813 -164426 -13046

Georgia $725 160 1586336 -137741 -121224 -11004

Hawaii $725 96 133662 -19310 -16995 -14933

Idaho $725 92 147501 -22165 -19507 -15809

Illinois $825 95 1225084 -109088 -89742 -70955

Indiana $725 125 816233 -90818 -79928 -83985

Iowa $725 155 478011 -42716 -37594 -28556

Kansas $725 135 388269 -40082 -35275 -27461

Kentucky $725 130 568821 -60840 -53544 -52259

Louisiana $725 149 683832 -63929 -56263 -66083

Maine $750 77 101976 -16567 -14323 -15234

Maryland $725 144 846415 -81748 -71946 -38370

Massachusetts $800 130 864721 -64902 -54251 -42913

Michigan $740 146 1439141 -128801 -112140 -11022

Minnesota $725 133 713646 -74730 -65769 -37878

TABLE 10

SNAP enrollments Predicted changes by state under the Harkin-Miller bill

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3743

34 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

StateMinimum wage

(2014)

SNAP enrollment (2012) Predicted change in SNAP enrollment

RateTotal

(persons)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 state

expenditures

(linear regression)

Mississippi $725 129 386501 -41486 -36511 -46467

Missouri $750 172 1036182 -75131 -64952 -56244

Montana $790 132 132452 -10350 -8708 -5846

Nebraska $725 124 230683 -25773 -22683 -12189

Nevada $825 162 446035 -23349 -19209 -11894

New Hampshire $725 127 168404 -18359 -16157 -5735

New Jersey $825 160 1416666 -75175 -61843 -28236

New Mexico $750 149 310896 -25983 -22463 -22512

New York $800 192 3763553 -191193 -159815 -142182

North Carolina $725 174 1697193 -135417 -119179 -113503

North Dakota $725 87 61225 -9743 -8574 -4021

Ohio $795 143 1647345 -115869 -97169 -88580

Oklahoma $725 129 494053 -53006 -46650 -46854

Oregon $910 124 485326 -17036 -13328 -16398

Pennsylvania $725 161 2053643 -177315 -156052 -125586

Rhode Island $800 156 163730 -10258 -8574 -8698

South Carolina $725 94 445277 -65614 -57746 -50304

South Dakota $725 208 173749 -11586 -10197 -7458

Tennessee $725 142 914903 -89667 -78915 -99134

Texas $725 110 2863779 -362018 -318607 -253285

Utah $725 88 251107 -39658 -34902 -19390

Vermont $873 156 97792 -3823 -3055 -2475

Virginia $725 101 829771 -113723 -100086 -58212

Washington $932 72 496934 -23221 -17947 -17756

West Virginia $725 58 107875 -25792 -22699 -21665

Wisconsin $725 75 427822 -79521 -69986 -53210

Wyoming $725 164 94590 -8010 -7050 -3104

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3843

Appendix C | wwwamericanprogresso

StateMinimum wage

(2014)

SNAP expenditures (2012)Predicted change in SNAP expenditures

(millions of dollars)

Per resident

(millions of

dollars)

Total

(millions of

dollars)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 s

expenditu

(linear regre

Alabama $725 $290 $1397 -$1185 -$1043 -$1043

Alaska $775 $253 $185 -$262 -$223 -$106

Arizona $790 $257 $1686 -$935 -$787 -$892

Arkansas $725 $248 $733 -$636 -$560 -$547

California $800 $189 $7164 -$7276 -$6082 -$3573

Colorado $800 $157 $816 -$862 -$721 -$407

Connecticut $870 $191 $686 -$343 -$275 -$210

Delaware $725 $250 $229 -$205 -$180 -$171

District of Columbia $825 $366 $232 -$146 -$120 -$99

Florida $793 $294 $5676 -$4429 -$3719 -$2951

Georgia $725 $317 $3140 -$2936 -$2584 -$2346

Hawaii $725 $335 $465 -$449 -$395 -$347

Idaho $725 $225 $359 -$376 -$331 -$268

Illinois $825 $249 $3200 -$2096 -$1725 -$1364

Indiana $725 $220 $1439 -$1162 -$1023 -$1075

Iowa $725 $192 $589 -$658 -$579 -$440

Kansas $725 $159 $460 -$502 -$441 -$344

Kentucky $725 $298 $1303 -$1133 -$997 -$973

Louisiana $725 $315 $1450 -$1047 -$922 -$1083

Maine $750 $281 $373 -$267 -$231 -$246

Maryland $725 $188 $1109 -$1765 -$1553 -$828

Massachusetts $800 $206 $1366 -$1030 -$861 -$681

Michigan $740 $300 $2963 -$2400 -$2090 -$2054

Minnesota $725 $140 $755 -$1113 -$980 -$564

Mississippi $725 $326 $973 -$649 -$571 -$726

Missouri $750 $241 $1452 -$1278 -$1104 -$956

Montana $790 $190 $191 -$179 -$151 -$101

Nebraska $725 $140 $259 -$409 -$360 -$194

Nevada $825 $191 $527 -$441 -$363 -$225

New Hampshire $725 $126 $167 -$399 -$351 -$125

TABLE 11

SNAP expenditures Predicted changes by state under the Harkin-Miller bill

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3943

36 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

StateMinimum wage

(2014)

SNAP expenditures (2012)Predicted change in SNAP expenditures

(millions of dollars)

Per resident

(millions of

dollars)

Total

(millions of

dollars)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 state

expenditures

(linear regression)

New Jersey $825 $160 $1420 -$1611 -$1325 -$605

New Mexico $750 $324 $675 -$513 -$444 -$445New York $800 $287 $5616 -$3766 -$3148 -$2801

North Carolina $725 $252 $2454 -$2187 -$1925 -$1833

North Dakota $725 $128 $90 -$162 -$143 -$67

Ohio $795 $259 $2995 -$2013 -$1688 -$1539

Oklahoma $725 $248 $945 -$799 -$703 -$706

Oregon $910 $322 $1255 -$272 -$213 -$262

Pennsylvania $725 $218 $2779 -$2930 -$2579 -$2075

Rhode Island $800 $280 $294 -$173 -$144 -$147

South Carolina $725 $291 $1373 -$1337 -$1177 -$1025South Dakota $725 $198 $165 -$192 -$169 -$123

Tennessee $725 $324 $2091 -$1413 -$1243 -$1562

Texas $725 $230 $5997 -$6402 -$5634 -$4479

Utah $725 $141 $402 -$614 -$541 -$300

Vermont $873 $230 $144 -$66 -$53 -$43

Virginia $725 $173 $1413 -$2062 -$1815 -$1056

Washington $932 $244 $1682 -$350 -$270 -$267

West Virginia $725 $273 $508 -$451 -$397 -$379

Wisconsin $725 $204 $1166 -$1302 -$1146 -$871Wyoming $725 $95 $55 -$105 -$93 -$41

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4043

Appendix C | wwwamericanprogresso

FIGURE 2

Raising the minimum wage to $1010would cut taxpayer costs in every state

Predicted decreases in cost and enrollment

in SNAP in 50 states

$200+$51ndash$100

$101ndash$200

0ndash$25

$26ndash$50

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4143

38 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

Endnotes

1 US Department of Agriculture ldquoSupplementalNutrition Assistance Program (SNAP)rdquo available athttpwwwfnsusdagovsnapeligibility (last accessedFebruary 2014)

2 For this initial analysis we do not consider Harkin-Millerrsquos increase in subminimum wages for tippedworkers To do s o would increase the estimated SNAP

savings by an unknown amount

3 The Congressional Budget Office estimates thatworkers currently earning between $1010 and $1150per hour would see their wages rise under the Harkin-Miller proposal Congressional Budget O ffice ldquoTheEffects of a Minimum Wage Increase on Employmentand Family Incomerdquo (2014)

4 Marianne Page Joanne Spetz and Jane Millar ldquoDoesthe Minimum Wage Affect Welfare Caseloadsrdquo Journalof Policy Analysis and Management 24 (2) (2005)273ndash295

5 Marianne Bitler and Hilary Hoynes ldquoThe More ThingsChange the More They Stay the Same The SafetyNet Living Arrangements and Poverty in the GreatRecessionrdquo NBER Working Paper 194 49 2013

6 Sylvia Allegretto and others ldquoFast Food Poverty Wages The Public Cost of Low-Wage Jobs in the Fast-FoodIndustryrdquo (Berkeley California Center for LaborResearch and Education 2013) available at httplaborcenterberkeleyedupubliccostsfast_food_poverty_wage

7 David Neumark and William Wascher ldquoDoes a HigherMinimum Wage Enhance the Effectiveness of theEarned Income Tax Creditrdquo Industrial and LaborRelations Review 64 (4) (2011) 712ndash746

8 David Lee and Emmanuel Saez ldquoOptimal MinimumWage Policy in Competitive Labor Marketsrdquo Journal ofPublic Economics 96 (9) (2012) 739ndash749

9 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family IncomerdquoFebruary 2014

10 Jesse Rothstein ldquoIs the EITC as Good as an NITConditional Cash Transfers and Tax Incidencerdquo AmericanEconomic Journal Economic Policy 2 (1) (2010) 177ndash208

11 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family Incomerdquo

12 Dube Arindrajit 2013 rdquoMinimum Wagesand the Distribution of Family IncomerdquoUnpublished working paper Available at httpsdldropboxusercontentcomu15038936Dube_MinimumWagesFamilyIncomespdf

13 As Dube explains in ldquoThe poverty of Minimum WageFactsrdquo the simulation approach underestimate stemsfrom a number of unwarranted assumptions includingthe range of actual wage increases and the accuracy ofwage data in the Current Population Survey The causal

approach does not make these assumptions

14 Allegretto Sylvia and others 2013 ldquoCredible ResearchDesigns for Minimum Wage Studiesrdquo Working Paper148-13 University of California Berkeley Institutefor Research on Labor and Employment Available athttpwwwirleberkeleyeduworkingpapers148-13pdf

15 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family In comerdquo

16 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo (2012)

17 Ibid

18 Ibid

19 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo

20 US Department of Agriculture ldquoSupplementalNutrition Assistance Program (SNAP)rdquo

21 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo

22 Ibid

23 We deviate from the Census Bureaursquos definition ofa family unit which is ldquotwo people or more (on e ofwhom is the householder) related by bir th marriageor adoption and residing togetherrdquo US Bureau ofthe Census ldquoCurrent Population Survey DefinitionsFamilyrdquo available at httpwwwcensusgovcpsabout

cpsdefhtml (last accessed February 2014) We countas a family unit any individual residing on his or herown two or more persons residing together whodo not belong to a family in the March CPS sampleare constructed as one family in our analysis For thepurposes of food stamp allocations the consumptionresulting from this transfer is probably distributed tofamily members (rather than household members ora single individual within the household) Howeversingle individuals canmdashand domdashreceive SNAPbenefits Excluding them would fail to make theanalysis reflective of the population at large

24 Strictly the family level linear probability modelpredicts the percentage-point decrease in theprobability that an individual family will receive SNAPpayments When applied to a large number of familieshowever we are able to interpret the coefficient asa decrease in the mean of enrollmentmdashthat is a

decrease in the enrollment ratemdashby applying the lawof iterated expectations

25 We generate expenditure predictions from theenrollment modelsmdashand conversely generateenrollment predictions from the expenditure modelmdashby assuming that expenditures per enrolled familyremains the same before and after the minimum wagechange In practice this is likely to be a conservativeestimatemdashthat is to underestimate the decrease inSNAP activity Average SNAP benefits per family willalso decrease as many families that remain eligible forSNAP experience income gains

26 Wage and Hour Division ldquoMinimum Wage Laws inthe States ndash Januar y 1 2014rdquo available at httpwwwdolgovwhdminwageamericahtm (last accessedFebruary 2014)

27 See for example Sylvia Allegretto and others ldquoCredibleResearch Designs for Minimum Wage Studiesrdquo WorkingPaper 148-113 (Berkeley California Institute forResearch on Labor and Employment 2013) available athttpwwwirleberkeleyeduworkingpapers148-13pdf

28 We will report these results in a forthcoming workingpaper

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4243

Endnotes | wwwamericanprogresso

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4343

The Center for American Progress is a nonpartisan research and educational institute

dedicated to promoting a strong just and free America that ensures opportunity

for all We believe that Americans are bound together by a common commitment to

these values and we aspire to ensure that our national policies reflect these values

We work to find progressive and pragmatic solutions to significant domestic and

international problems and develop policy proposals that foster a government that

is ldquoof the people by the people and for the peoplerdquo

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 643

Introduction and summary | wwwamericanprogress

considerably wih differen sample periods and assumpions abou sae rends4

Proessors Marianne Biler and Hilary Hoynes discuss he imporance o SNAP

as a saey ne program bu hey do no examine is relaion o minimum wage

policy5 Research economis Sylvia Allegreto and her Universiy o Caliornia a

Berkeley colleagues show ha low-wage workers in general and as-ood workers

in paricular are much more likely o be SNAP recipiens han all workers6

Several sudies have examined he relaionship beween he minimum wage and

he Earned Income ax Credi or EIC Proessor David Neumark and William

Wascher a researcher a he Federal Reserve Board o Governors find ha a

higher minimum wage increases EIC benefis or amilies in deep povery

while reducing EIC benefis or some sub-groups7 Proessors David Lee and

Emmanuel Saez argue ha he minimum wage and EIC are complemenary

policies no subsiues8 Te Congressional Budge Office or CBO argues ha

a minimum wage increase will no have a subsanial effec on EIC spending9

while Proessor Jesse Rohsein examines wheher he posiive effec o he EICon emale labor supply has lowered wages10 While hese sudies are o ineres

he EIC is quie differen rom SNAP in having a subsanial phase-in period in

which EIC benefis increase as well as a long phase-ou period wih complee

phase-ou a an annual income o abou $48000 or a amily o our quie a bi

above he reach o he minimum wage11

Research by Proessor Arindraji Dube on he causal effec o he minimum wage

on amily povery represens he sudy mos relaed o he one a hand12 Dube finds

ha Harkin-Miller would raise abou 46 million non-elderly Americans above

he ederal povery level or FPL In conras when CBO uses a simple simulaion

mehod o address he same quesion hey find ha Harkin-Miller would raise

900000 people above FPL13 Te difference beween hese wo esimaes highlighs

he imporance o underaking a causal analysis Te mehods used in his paper are

in many respecs similar o Dubersquos Moreover since eligibiliy and benefi levels or

programs such as SNAP and Medicaid are ied o he ederal povery level Dubersquos

findings have direc implicaions or his sudy Noneheless his repor appears o

be he firs sudy o examine he effecs o he minimum wage on SNAP In uure

work we plan o underake similar analyses or he EIC and Medicaid

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 743

4 Center for American Progress | The Effects of M inim um Wages on SNAP Enro llments and Expenditures

Te repor proceeds as ollows

bull Secion 1 provides background inormaion on he ederal minimum wage sae

minimum wages and he SNAP program

bullSecion 2 describes our mehods and daa

bull Secion 3 provides our main resuls including a simulaion o he effecs o a

Harkin-Miller minimum wage increase and a sae-by-sae analysis

bull Secion 4 presens our conclusions

Furher deails are provided in a series o appendices

Correction April 28 2014 Tis report incorrectly stated the potential reduction in

SNAP enrollment fom the Harkin-Miller proposal Te correct amount is 75 percentand 87 percent (31 million to 36 million persons) as stated in the reportrsquos tables

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 843

Background | wwwamericanprogress

Background

The federal minimum wage

Te ederal minimum wage was las increased in July o 2009

nearly five years ago During he pas wo decades many saes

have passed legislaion fixing he minimum wage a a higher

level han he ederal minimum Te maps in Figure 1 show ha

while saes in every region o he Unied Saes have adoped

higher minimum wages hey are no disribued randomly bygeography As shown in he 2013 sudy ldquoCredible Research

Designs or Minimum Wage Sudiesrdquo by economiss Sylvia

Allegreto Arindraji Dube Michael Reich and Ben Zipperer

hese saes vary sysemaically rom he oher saes by a number

o characerisics ha affec low-wage employmen rends bu

which are no hemselves relaed o minimum wage policy14

Te nonrandom patern o minimum wage adopion has

imporan implicaions or obaining unbiased esimaes o

minimum wages on employmen In paricular naional panel

sudies ha use sae and ime fixed effec models991252such as

a 1992 sudy by David Neumark and William Wascher991252

spuriously esimae negaive employmen effecs Te reason

or his resul is uncovered using ess or pre-rends Tese

ess find ha low-wage employmen was already declining wo

years beore minimum wages were implemened By making a

saisically large number o local comparisons ha conrol or

heerogeneiy among saes and by ime eliminaes his pre-

rend For his reason we conduc similar ess or our SNAPoucomes and use model specificaions ha include local

comparisons as in he sudy cied above

FIGURE 1

High versus low minimum wage stat

from 1990 to 2012

Means and variances

More than $533

Average minimum wage over 1990ndash2012

Less than or equal to $533

More than $121

Minimum wage variance over 1990ndash2012

Less than or equal to $121

Notes State means and variances were calculated using annual stat

minimum wage data from 1990 to 2012 The shading on the maps

partitions the states into above- and below-median values

Source Sylvia Allegretto and others ldquoCredible Research Designs forMinimum Wage Studiesrdquo Working Paper 148-13 (Institute for Resear

on Labor and Employment 2013) available at httpwwwirleberkeleyeduworkingpapers148-13pdf

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 943

6 Center for American Progress | The Effects of M inim um Wages on SNAP Enro llments and Expenditures

Te CBO recenly projeced ha in 2016 17 million workers will earn less han

he $1010 hourly wage proposed in he Harkin-Miller bill Furhermore he

CBO esimaes ha an addiional 8 million workers earned beween $1010 and

$1150 per hour and were also likely o experience a wage increase15

Supplemental Nutrition Assistance Program

We ocus our inquiry on SNAP Benefis under he program are enirely

ederally unded he program is adminisered by he US Deparmen o

Agriculure ogeher wih sae agencies which share in adminisraion

coss Spending on SNAP has grown in he pas decade reaching $78 bil lion

in 2011 wih SNAP enrollmen increasing o 45 million people abou one-

sevenh o he US populaion16 According o he CBO changes since 1990 in

SNAP spending and enrollmens are primarily he resul o cyclical economic

condiions noably changes in he unemploymen rae and changes in per capiaincome17 Te 2009 American Recovery and Reinvesmen Ac emporarily

increased SNAP benefi amouns by 136 percen as repored by he US

Deparmen o Agriculurersquos Food and Nuriion Service hese higher benefi

levels expired on November 1 201318 Te CBO esimaes ha abou wo-

hirds o he changes in SNAP expendiure are associaed wih changes in he

number o recipiens and one-hird wih changes in he benefis received when

recipiensrsquo incomes change19

In fiscal year 2014 SNAPrsquos maximum monhly benefis are $189 or a single

individual $497 or a amily o hree and $750 or a amily o five Benefis are

reduced by 30 cens per dollar received and phase ou enirely a gross monhly

household incomes o 130 percen o he ederal povery level $1245 or a

single individual $2116 or a amily o hree and $2987 or a amily o five

o deermine benefis SNAP also defines a ne monhly income concep and

ses benefis a 100 percen o he ederal povery level using his concep

Calculaion o ne monhly income can include cerain deducions rom

monhly gross income such as medical expenses and child care coss Alhough

saes are permited some laiude on wha deducions are allowed in pracice

hese vary by very small amouns Our saisical model akes accoun o sae-specific differences in benefis20

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1043

Background | wwwamericanprogress

SNAP imposes an employmen or work-raining requiremen or able-bodied

prime-age aduls991252hose beween he ages o 18 and 50 and wihou disabiliies

or dependen children Such households can receive only hree monhs o benefis

in a hree-year period In recen years abou 85 percen o households receiving

benefis have incomes below he ederal povery level 49 percen have dependen

children 16 percen are age 60 or older 20 percen are disabled and 30 percenrepor some earned income21

A 2012 CBO repor also noes ha ake-up raes among eligible SNAP recipiens

average abou 70 percen wih much lower ake-up among elderly households

Te ake-up rae increases in harder economic imes I also increased when sigma

issues were reduced as SNAP debi cards replaced acual ood samps ake-up

is especially high among hose mos needy Adminisraive spending equaled 91

percen o he poenial spending ha would have occurred i all eligible recipiens

were enrolled Alhough some SNAP policy changes have occurred since 1990

mos were relaively minor and all were naional in scope Te 1996 welare reorm bill eliminaed SNAP eligibiliy or some legal immigrans limied he ime lengh

o eligibiliy or able-bodied childless aduls and reduced maximum benefis

Some o hese resricions were relaxed in 2002 and again in he American

Recovery and Reinvesmen Ac in 200922

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1143

8 Center for American Progress | The Effects of M inim um Wages on SNAP Enro llments and Expenditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1243

Methods and data | wwwamericanprogress

Methods and data

As previously menioned we exploi variaion in minimum wages by sae and ime

o examine heir causal effecs on SNAP enrollmens and expendiures o do so we

merge daa rom 1990 hrough 2012 drawn rom he Annual Social and Economic

Supplemen o he Curren Populaion Survey991252an annual Census Bureau survey

commonly known as he March CPS ha includes23 inormaion on SNAP

enrollmens a he amily level991252wih sae-level daa on minimum wages SNAP

expendiures populaion unemploymen raes and sae median income levels

o conrol or ime-varying heerogeneiy among saes our specificaions includeconrols or sae linear rends and effecs by Census division and ime We esimae

effecs a wo levels allowing or amily variaion and allowing only or sae-level

variaion We also employ a se o sandard demographic conrols such as amily size

and composiion and race and ehnic composiion

Distinguishing causation from correlation

How can we ensure ha our analysis does no pick up a spurious correlaion or

example he endency o more economically vibran saes o implemen higher

minimum wages Disinguishing correlaion and policy endogeneiy rom rue

causal effecs is he primary moivaion or economeric analysis In he ideal

experimen researchers would begin wih wo saes991252ha are alike in every

respec prior o he policy991252 and ldquoreardquo only one o hese saes wih a higher

minimum wage Tey would atemp o shield hese saes rom any influence ha

could obscure heir undersanding o he minimum wagersquos direc effec on SNAP

enrollmen Researchers o course canno conduc such experimens

We can however use saisical mehods o conrol simulaneously heindependen effecs on SNAP o sae unemploymen raes sae income levels

and common rajecories among saes wihin he same Census division By

ensuring similariy along hese dimensions we maximize he likelihood ha

SNAP aciviy in wo saes would have comparable oucomes in he absence

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1343

10 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

o a minimum wage policy change Tus i a new minimum wage policy were

implemened in one sae only he researchers could atribue all o he difference

hey observe in SNAP aciviy o he new minimum wage policy

In oher words we approximae he ideal experimen by using non-experimenal

saisical mehods Te desirable ldquopre-exising similariiesrdquo beween saes ha wehave defined above inorm our choice o conrol variables in a saisical seting

More precisely in our muliple regression models we use median amily income

he unemploymen rae he employmen-o-populaion raio and regional and

ime idenifiers o consruc an appropriae group o peers or each sae on he

eve o a policy change

Data description

wo daa ses include inormaion abou boh income and paricipaion inpublic programs Te Survey o Income and Program Paricipaion or SIPP

which is conduced in inermiten years has he advanage o ollowing he

same individuals over a period o ime In oher words i is a longiudinal daa

se I also has he advanage o conaining monhly daa However he sample

size o he SIPP is no sufficien or analyzing variaions in sae-level minimum

wages Te March CPS has he advanage o a much larger sample size and i is

conduced annually wihou any breaks in ime I has he disadvanage o being

a cross-secional daa se so we canno ollow he same individuals over ime991252

sricly speaking over more han one year On ne he March CPS is much more

suiable or our sudy We examine he empirical relaionship beween minimum

wage policy and ood samp aciviy a wo levels o aggregaion he amily level

and he sae level Family-level daa are drawn rom he March CPS

Te March CPS comprises responses rom he residens o 50000 o 60000

dwelling places surveyed per year and conains deailed inormaion on he

residensrsquo employmen and income including income rom ranser paymens

Te sample or our analysis comprises more han 128 million amily unis during

he period rom 1990 o 2012 (inclusive) Survey weighs allow us o analyze

SNAP paricipaion in a manner ha is represenaive o he US populaion alarge Over all years he share o amilies reporing ood samp receip in he

weighed March CPS sample is 91 percen Te enrollmen rae was a a low o 6

percen in he year 2000 In 2012 he mos recen year in our panel 133 percen

o amilies repored paricipaing in SNAP a some poin during he survey year

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1443

Methods and data | wwwamericanprogresso

Te March CPS also collecs inormaion on he number o SNAP recipiens in

he household he number o monhs he household receives SNAP benefis

and he cash-equivalen value o he SNAP benefis received However he

value o SNAP benefis is severely underrepored among recipiens perhaps

because recipiens are unaware o he exac monhly cash-equivalen value o

benefis hey receive

Our firs empirical sraegy ocuses solely on SNAP enrollmen By using he

amily as he uni o analysis we are able o inser saisical conrols o accoun

or non-wage-relaed acors ha influence any paricular amilyrsquos likelihood o

program paricipaion wih he inenion o isolaing any differences in program

paricipaion ha are due purely o changes in wage policy Tis approach

idenifies he effecs o low-wage labor policy on he exernal margin991252ha is

he effec o he minimum wage on he likelihood ha a amily paricipaes in he

SNAP program a all991252as opposed o he inernal margin or how much SNAP

unding he amily would receive

Our second empirical ramework uses sae-level adminisraive daa Ta is

we aggregae he daa o obain a single daa poin or each saeyear back

o 1990 represening he mean o he oucome or he sae Te sae-level

esimaion serves as a robusness check on he amily-level resuls or SNAP

paricipaion Also using aggregaed daa allows us o esimae direcly he

causal effec o minimum wage changes on SNAP spending Tis is no possible

a he amily level as discussed above daa on cash-equivalen value o ood

samps or SNAP recipiens is very requenly no repored in he March CPS

and when i is repored he inormaion may be unreliable By conras he

Bureau o Economic Analysis publishes aggregae SNAP spending a he sae

level in is Naional Income and Produc Accoun or NIPA ables Tus

while we are unable o observe he heerogeneiy in he cash value o SNAP

or amilies in each sae we are able o calculae average SNAP spending

per residen in each sae per year Supporing covariaes include he annual

unemploymen and employmen daa rom he Bureau o Labor Saisics or

BLS and sae-level populaion series rom he iner-decennial census releases

Minimum wage daa are available rom he BLSrsquos wages and hours division For

sae minimum wage changes enaced a oher imes han he firs o he yearan average value or he year is used

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1543

12 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

Family level model specifications

We firs examine he effec o he minimum wage on paricipaion in public

assisance programs For amily residing in sae and during year we esimae

an equaion o he ollowing orm

(1)

is a binary variable ha is se equal o 1 i a leas one member o amily

received ood samps during he survey year is a se o sae-level

characerisics including annual averages o he unemploymen rae he

employmen-o-populaion raio and he naural log o median amily income

is a vecor o amily atribues including indicaors or he race and marial saus

o he amily head size o he amily he presence o children and he presence oan adul male Sae fixed effecs are capured by o conrol or ime-varying

heerogeneiy our preerred model specificaion also includes year fixed effecs

ha vary by Census division ( ) and sae-level linear ime rends In

Appendix B we jusiy he inclusion o hese las wo erms We also compare he

resuls rom our preerred specificaion wih less sauraed specificaions

Te effec o ineres which is capured by is he expeced change in he

probabiliy o receiving SNAP benefis wih respec o a change in he (log o he)

binding minimum wage in sae during year We repor robus sandard errors

clusered a he sae level We esimae he parameers using linear regression

producing a linear probabiliy model Deails o he model selecion process are

covered in Appendix B below

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1643

Methods and data | wwwamericanprogresso

State-level model specifications

Te sae-level models are similarly specified For sae in year we assume ha

(2)

In his model is now eiher he SNAP enrollmen rae in sae during year

or he naural logarihm o per capia SNAP expendiures in sae during

year is once again a se o sae-level characerisics including he same

sae-level covariaes as in he amily regressions (annual average unemploymen

rae employmen-o-populaion raio naural log o median amily income)

wih he addiion o amily level characerisics averaged across he sae (average

amily size and he shares o populaion consiued by each o five racialehnicgroups) Sae fixed effecs are represened by As above our preerred model

specificaion includes year fixed effecs ha vary by Census division ( ) and

sae-level linear ime rends as elaboraed in Appendix B Te effec o

ineres is capured by

We esimae boh sae-level models (enrollmen and expendiures) using

ordinary leas squares regression Tus he inerpreaion o he coefficien is no

longer ha o a change in probabiliy as in he binary oucome models described

above Raher or he sae-level SNAP enrollmen model represens he

expeced change (in percenage poins) in he saersquos SNAP enrollmen rae ha

is due o a 1 percen change in he minimum wage For he SNAP expendiures

model is simply he elasiciy o SNAP spending wih respec o he minimum

wage991252ha is he percenage change in sae expendiures expeced o resul rom

a 1 percen change in ha saersquos minimum wage For urher deails on model

specificaion reer o Appendix B below

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1743

14 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1843

Results | wwwamericanprogresso

Results

Estimated minimum wage effects on SNAP enrollment and

expenditures

able 1 shows he esimaed parameer o ineres991252he coefficien o he

minimum wage or he preerred model o each ype Coefficiens on he

minimum wage variable are no direcly comparable across models because all

our models have a differen uncional orm o undersand and compare hese

esimaes we compue he change in SNAP aciviy prediced or a paricular wage scenario Te final column in able 1 answers he quesion Wha would be

he expeced change in SNAP aciviy in response o a 10 percen increase in he

minimum wage Te answer o his quesion varies wih he value o he inpu

parameers in he able we calculae he percenage decrease in enrollmen or

expendiures prediced or he average sae wih a minimum wage o $725 in

2014 Te sae-level SNAP expendiure model which is a consan-elasiciy

model conveys elasiciy inormaion direcly or he change in expendiures per

capia in he sae

TABLE 1

Comparison of national SNAP predictions for a 10 percent increase in the federal minimum wage

Model LevelRegression

type

Predicted outcome Coefficient of log

(minimum wage)

(Standard error)

Effect of a 10 percent in

in the minimum wage

VariableForm of

variable

Total

enrollment

Tot

expend

1 Family Linear

probability Enrollment Binary (enrolled=1)

-0042

(0008) -317 N

2

StateLinear regression

(ordinary least

squares)

Enrollment State enrollment rate (percent) -0031

(0012) -235 N

3 Expenditures Log (state expenditures per capita) -0190

(0103) NA -19

plt01 plt005 plt001Note Predicted changes are calculated for the average state with a minimum wage of $725 in 2014

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1943

16 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

According o his model a 10 percen increase in he minimum wage would resul

in a 19 percen reducion in SNAP expendiures Tis resul is reassuringly similar

o Dubersquos resul or povery reducion Te wo enrollmen models are much more

precisely esimaed han he expendiure model Te sae-level enrollmen model

finds ha a 10 percen minimum wage increase in a low-wage sae is associaed wih

a 235 percen decrease in SNAP enrollmens Te amily-level linear probabiliymodel predics a somewha greaer elasiciy or low minimum wage saes an

increase o 10 percen in he ederal minimum would resul in a 317 percen decline

in SNAP enrollmen24 Te differences in hese esimaes sem rom a number o

acors including difference in model uncional orm and daa used We rea his

range o elasiciy esimaes as an upper and lower bound on enrollmen impacs

Harkin-Miller bill National and state-level predicted impacts

Wha would be he prediced change or he SNAP program i he ederalminimum were raised o $1010 as proposed in he Harkin-Miller bill In order o

make his inerence we accoun or he ac ha no all saes are currenly subjec

o he ederal minimum wage a he beginning o 2014 21 saes mainained

higher minimum wages han $725 In hose saes an increase in he ederal

minimum wage may or may no be binding or employers in he sae depending

upon wheher he new ederal minimum exceeds he sae-level minimum Bu

regardless o wheher a minimum wage change is binding he impac on SNAP

aciviy will be lower in high minimum wage saes In order o accoun or his

properly we calculae sae by sae he percenage wage change ha would resul

rom he Harkin-Miller proposal and apply he parameers rom each o he hree

models above o compue he expeced decrease in SNAP aciviy or each sae

In his exercise we use saesrsquo curren (2014) minimum wage levels and assume

as a baseline he 2012 levels o SNAP enrollmen and expendiure as 2012 is he

mos recen year or which SNAP daa are available

able 10 and able 11 in Appendix C repor he esimaed effecs on SNAP

enrollmen and expendiures respecively or each sae under he Harkin-Miller

bill25 An increase o $1010 i enaced oday would represen beween a 393

percen wage increase in a $725 minimum wage sae and an 84 percen increasein Washingon sae which has he highes minimum wage in he naion a $932

as o January 201426 Slighly more han 56 percen o he decrease in expendiures

and abou 59 percen o he decrease in enrollmen would occur in saes wih

presen-day minimum wages o $725

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2043

Results | wwwamericanprogresso

In 2012 hese saes were home o 46 percen o he American populaion

and accouned or a roughly equivalen percenage o oal naional SNAP

expendiures No surprisingly he larges enrollmen decreases come rom

populous saes wih very high SNAP enrollmen raes andor very low minimum

wages Te larges prediced enrollmen reducion991252beween 319000 individuals

and 362000 individuals991252would occur in exas which has a $725 minimum wage In Caliornia which has a 206 percen SNAP paricipaion rae and an $8

per hour minimum wage we anicipae SNAP enrollmen declines o 310000

persons o 371000 persons And in Florida which had a SNAP paricipaion rae

o 166 percen and a $793 minimum wage enrollmen could decline by beween

164000 individuals and 196000 individuals For he our saes ha ogeher

accouned or he greaes amoun o SNAP spending in 2012991252exas Caliornia

Florida and New York respecively991252he combined expendiure reducion rom

he Harkin-Miller bill is prediced o be $14 billion

able 2 summarizes he prediced declines in SNAP aciviy or he naion as a whole ha would resul rom he direc and indirec effecs o he Harkin-Miller

bill Enrollmen would all beween 31 million persons and 36 million persons

represening 75 percen o 87 percen o curren enrollmen Te anicipaed

reducion in program expendiures would be nearly $46 billion or 61 percen o

program expendiures

TABLE 2

Comparison of national SNAP predictions under the Harkin-Miller billrsquos $1010 minimum wage

Model

Enrollment

(persons)

Expenditures

(millions of dollars)

Current (2012) Predicted Change Current (2012) Predicted Chan

Family enrollment (linear probability)

41866195

45489339 -3623144

$74861

NA NA

State enrollment (ordinary least squares) 38745435 -3120759 NA NA

State expenditures (ordinary least squares) NA NA $70305 -$45

Note Calculations use 2014 state minimum wages and the most recent SNAP data from 2012 They assume that per-enrollee expenditures remain constant

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2143

18 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

Tere are o course oher possibiliies or minimum wage legislaion able 3

shows he expeced SNAP changes or he Unied Saes under a variey o wage

scenarios calculaed using he sae-level models I saes were no able o se

heir minimum wages independenly such ha all saes were consrained by

he ederal minimum o $725 SNAP would be received by abou 514000 more

people across he Unied Saes a an addiional program cos o nearly hree-quarers o a billion dollars In conras he effecs o a higher minimum wage

proposal991252a ederal wage floor o $11 per hour991252would decrease enrollmen in

SNAP by more han 10 percen and decrease program coss by 83 percen

TABLE 3

Summary of par ticipation and expenditures under wage scenarios

If all states had

minimum wages of

Enrollment(persons)

Expenditures

(millions of dollars)

Predicted Change Predicted Change

Recent levels (2014) 41866195 $74861

$725 42380520 514326 $75604 $743

$800 41423919 -442276 $74209 -$652

$900 40148451 -1717744 $72350 -$2511

$1000 38872982 -2993212 $70490 -$4371

$1010 38745435 -3120759 $70305 -$4556

$1100 37597514 -4268681 $68631 -$6230

Note Calculations use state-level enrollment model coefficient

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2243

Conclusion | wwwamericanprogresso

Conclusion

An exensive body o lieraure examines employmen effecs o he minimum

wage A much smaller se o sudies examines how he minimum wage affecs

povery and only a handul o papers examine he effecs o he minimum wage

on he EIC Our analysis is he firs o examine he effecs o he minimum

wage on SNAP

Our findings indicae ha increased earnings rom minimum wage changes do

reduce SNAP enrollmens and expendiures We esimae ha he Harkin-Miller bill would save axpayers nearly $46 billion per year equivalen o 61 percen

o SNAP expendiures in 2012 he las year or which daa are available Over a

10-year period he esimaed savings amoun o nearly $46 billion

Our repor is subjec o limiaions ha we expec o overcome in our uure

research Firs he findings do no ake ino accoun possible ineracions among

SNAP he EIC and Medicaid Te eligibiliy cuoffs among hese programs

are quie differen suggesing ha such ineracions may be minor Noneheless

he join effecs can only be deermined by urher research using a causal

model Second i would be useul o know he disribuion o SNAP reducions

along he wage disribuion Using he Congressional Budge Officersquos calculaions

o how much he oal dollar value o wage would increase under he Harkin-

Miller proposal our findings imply ha he decline in overall SNAP spending

equals abou 15 percen o he oal resuling increase in wages Te amoun and

disribuion o his offse are o considerable ineres Minimum wage beneficiaries

who come rom working amilies already well above he povery line would no

see any offse while hose who are currenly considerably below he povery line

will see larger offses Tese issues will also be a subjec or our uure research

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2343

20 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

About the authors

Rachel West is a maser o public policy candidae a he Goldman School

o Public Policy Universiy o Caliornia Berkeley Her research ocuses on

economic policy in he areas o low-wage labor and povery

Michael Reich is proessor o economics and direcor o he Insiue or

Research on Labor and Employmen a he Universiy o Caliornia a Berkeley

His research publicaions cover numerous areas o labor economics including

racial inequaliy labor marke segmenaion high-perormance workplaces

union-managemen cooperaion Japanese labor-managemen sysems living

wages and minimum wages He received his docorae in economics rom

Harvard Universiy

Acknowledgments

We are graeul o Sylvia Allegreto Arindraji Dube Bill Leser Jesse Rohsein

Daniel Tompson and Ben Zipperer or heir valuable suggesions

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2443

References | wwwamericanprogresso

References

Allegreto Sylvia and ohers 2013 ldquoFas Food Povery Wages Te Public Cos o Low-Wage Jobsin he Fas-Food Indusryrdquo Universiy o Caliornia Berkeley Cener or Labor Research andEducaion Available a htplaborcenerberkeleyedupubliccossas_ood_povery_wage

Allegreto Sylvia and ohers 2013 ldquoCredible Research Designs or Minimum Wage Sudiesrdquo

Working Paper 148-13 Universiy o Caliornia Berkeley Insiue or Research on Labor andEmploymen Available a htpwwwirleberkeleyeduworkingpapers148-13pd

Biler Marianne and Hilary Hoynes 2013 ldquo Te More Tings Change he More Tey Say heSame Te Saey Ne Living Arrangemens and Povery in he Grea Recessionrdquo Working Paper19449 Naional Bureau o Economic Research

Congressional Budge Office 2012 ldquoTe Supplemenal Nuriion Assisance Programrdquo Washingon Available a htpwwwcbogovsiesdeaulfilescbofilesatachmens04-19-SNAPpd

991252 991252 991252 2014 ldquoTe Effec o a Minimum-Wage Increase on Employmen and Family Incomerdquo Washingon Available a htpwwwcbogovsiesdeaulfilescbofilesatachmens44995-MinimumWagepd

Dube Arindraji 2013 rdquoMinimum Wages and he Disribuion o Family Incomerdquo Unpublished working paper Available a htpsdldropboxuserconencomu15038936Dube_ MinimumWagesFamilyIncomespd

991252 991252 991252 2014 ldquoTe Povery o Minimum Wage Facsrdquo Arindraji Dube Blog January 22 Availablea htparindubecom20140122he-povery-o-minimum-wage-acs

Lee David and Emmanuel Saez 2012 ldquoOpimal Minimum Wage Policy in Compeiive LaborMarkesrdquo Journal o Public Economics 96 (9) 739ndash749

Neumark David and William Wascher 1992 ldquoEmploymen Effecs o Minimum and Subminimum Wages Panel Daa on Sae Minimum Wage Lawsrdquo Industrial and Labor Relations Review 46 (1)

55ndash81

Neumark David and William Wascher 2011 ldquoDoes a Higher Minimum Wage Enhance heEffeciveness o he Earned Income ax Credirdquo Industrial and Labor Relations Review 64 (4)712ndash746

Page Marianne Joanne Spez and Jane Millar 2005 ldquoDoes he Minimum Wage Affec WelareCaseloadsrdquo Journal o Policy Analysis and Management 24 (2) 273ndash295

Rohsein Jesse 2010 ldquoIs he EIC as Good as an NI Condiional Cash ransers and ax

Incidencerdquo American Economic Journal Economic Policy 2 (1) 177ndash208

Wage and Hours Division 2014 ldquoMinimum Wage Laws in he Saes ndash January 1 2014rdquo USDeparmen o Labor (htpwwwdolgovwhdminwageamericahm [February 2014])

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2543

22 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2643

Appendix A | wwwamericanprogresso

Appendix A

Pre-trend falsification check

Recen minimum wage research27 highlighs a common flaw in previous sudies

ailure o veriy ha he oucome variable is ree o negaive pre-exising rends I

or example SNAP aciviy was already rending down in saes ha raised heir

minimum wages beore hese changes came ino effec our regression analysis could

(misakenly) atribue ha reducion o he minimum wage We check or such

pre-rends by inroducing variables ha represen he prior yearrsquos value or leads o

he minimum wage I he model esimaes he minimum wage o have an effec on

he oucome variable beore he wage change wen ino effec hen an unobservedacor no he minimum wage change caused he change in SNAP aciviy

We es he specificaions above or pre-rends by including a one-year lead in

all hree specificaions We find ha he lead erms are small posiive and no

saisically significan indicaing ha he concurren minimum wage991252no

he wage level in prior periods991252is driving he observed changes in SNAP

oucomes28 In paricular he coefficien (sandard error) on he lead erm in

our preerred amily-level enrollmen regression is 011 and no significan

while he coefficien and sandard error o he conemporaneous minimum

wage is unchanged In he sae-level preerred enrollmen regression he

coefficien o he lead erm is again small (07) and i is no significan Te

corresponding coefficien on he lead erm in he sae-level expendiure

regression is 16 and is no significan Te posiive poin esimaes on hese lead

erms resuls no only rule ou disoring negaive pre-rends Tey also sugges

ha our main resuls may underesimae he rue effecs

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2743

24 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2843

Appendix B | wwwamericanprogresso

Appendix B

Model selection process

For boh he amily-level and sae-level models we es hree mehods o conrol

or unobserved geographic- and ime-varying characerisics as suggesed by he

minimum wage lieraure Firs we include only independen sae-specific fixed

effecs and year-specific fixed effecs Tis specificaion (specificaion 1) implicily

assumes ha amilies in any sae consiue an equally good saisical ldquoconrolrdquo

group or hose in any randomly chosen sae afer accouning or various

characerisics (median income and unemploymen rae among ohers) Similarly

simple ime fixed effecs assume ha amilies surveyed in any year can crediblyserve as a conrol group or amilies surveyed in every oher year o he sample

(1990 hrough 2012)

In oher words specificaion 1 assumes ha a saersquos immediae neighbor provides

no beter a couneracual or he effec o a minimum wage change han does a

sae across he counry We relax his resricive specificaion sequenially in wo

seps In specificaion 2 we replace simple year fixed effecs wih fixed effecs or

each Census divisionyear (capured as an addiional variable in he vecor By

using division-year effecs we remove he resricion ha amilies in each sae

are equally good saisical conrols or all oher amilies Raher we allow or he

possibiliy ha amilies in similar geographic regions (or example he Souh or

he Norheas) may be more similar o one anoher han amilies arher away

Finally in specificaion 3 we add sae-specific linear ime rends o he previous

specificaion Tus specificaion 3 is he mos rigorous model specificaion in ha

i allows or heerogeneiy along hree dimensions Ta is specificaion 3 allows

each sae o have is own ime-varying rends raher han imposing he resricion

ha saes evolve idenically over he 22 years in he sample

We begin building he heoreical specificaion above rom a se o simpleuncondiional models regression o SNAP aciviy (enrollmen or expendiures)

on he log o he minimum wage and a se o geographic- and ime-specific

effecs (specificaions 1 2 and 3 described above) As shown in ables 1ndash3 (or

specificaion 3) we hen add covariaes sequenially o hese models including

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2943

26 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

firs he vecor o amily-level conrols ollowed by each o several sae-level

covariaes in urn he unemploymen rae log o median-amily income and he

employmen-o-populaion raio Comparable resuls or specificaions 2 and 3

will be available in our orhcoming working paper

As expeced he simple uncondiional models indicae ha he relaionship beween he minimum wage and SNAP enrollmen i one exiss is a more complex

one influenced by oher acors In he uncondiional model he coefficien on he

variable o ineres991252he log o he minimum wage991252is small in magniude and no

saisically differen rom zero Once we accoun or he influence o labor marke

condiions and variaion in income levels on program paricipaion (by including

unemploymen rae and median-amily income conrol variables respecively)

he effec o he minimum wage on SNAP enrollmen is precisely esimaed Te

coefficien o he log minimum wage is slighly higher (-0042) in he amily-level

analysis han he coefficien (-031) in he sae-level analysis Te level o precision

is also higher in he amily-level analysis Tis is o be expeced when using 124million observaions compared o 1127

Te second dimension o model choice concerns he effec specificaion ables

7ndash9 compare he primary coefficiens o ineres or he SNAP enrollmen and

expendiure models For boh he enrollmen models he effec sizes are smalles

or specificaion 1 larges or specificaion 2 and inermediae beween hese wo

in specificaion 3 Recall ha Specificaion 3 conains sae-specific linear ime

rends in addiion o he census divisionyear conrols included in specificaion

2 In he amily-level enrollmen model he sandard error o he minimum wage

coefficien is smaller han in he oher wo specificaions Sandard errors on he

oher variables are much smaller in specificaions 2 and 3 han in specificaion

1 On he basis o coefficien significance (join and individual) specificaions 2

and 3 are sricly preerred in boh enrollmen models o specificaion 1 which

conains only sae and year fixed effecs

A concern wih specificaions 2 and 3 is ha rend conrols such as sae linear

rends may incorrecly absorb some o he delayed impac o a minimum wage

When we es his issue by including lagged minimum wages we do no find ha

delayed effecs are significan Anoher concern is ha more sauraed modelsuse less o he saisical variaion which could reduce he saisical power o

he resuls However he sandard errors or our more sauraed models are no

higher and are lower in some cases han or he less sauraed models Overall

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3043

Appendix B | wwwamericanprogresso

his evidence suppors our use o he sauraed model as he preerred model

specificaion Moreover Dubersquos 2013 sudy shows ha more sauraed models

perorm beter han models wih jus sae and ime fixed effecs

Te esimaed enrollmen regressions a boh he amily and sae levels show large

and saisically significan coefficiens Te esimaed minimum wage effec in heexpendiures regressions991252or which we have only sae-level daa991252is also large

and saisically significan

We do no use weighed regression or he sae-level models preerring o keep

analysis o he ldquoreamenrdquo (ha is o say a minimum wage change) appropriae

o he average sae raher han he average amily or individual I insead our

primary ineres were he impac o a minimum wage change on he average amily

or he average individual we migh choose o designae he number o amilies

in each sae or he sae populaion respecively as analyic weighs in order o

obain a coefficien beter suied or such inerence

TABLE 4

SNAP enrollment

Family-level linear probability

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0048 -0047 -0040 -0043 -0042

(0013) (0013) (001) (0008) (0009)

Unemployment rate (100) 0505 0420 0280

(0083) (0086) (0082)

Log median income -0057 -0039

(0011) (0011)

Employment-to-population ratio -0239

(0038)

N 1242022 1242022 1242022 1242022 1242022

plt01 plt005 plt001

Note Robust standard errors are in parentheses Observations are clustered at the state level The outcome variable is binary and equal to one if a family is enrolledin SNAP All models include state fixed effects Census division x-year fixed effects and state-specific linear time trends All specifications except 3a include additional

controls for family size race and marital status of the family head presence of children and presence of an adult maleSource Annual data from the March Current Population Survey for 1990 to 2012 estimation includes Current Population Survey probability weights

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3143

28 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

TABLE 5

SNAP enrollment

State-level linear regression

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0028 -0024 -0026 -0031 -0031

(0014) (0013) (0013) (0012) (0012)

Unemployment rate (100) 0625 0466 0320 0339

(0087) (0088) (0085) (0083)

Log median income -0090 -0065 -0061

(0013) (0013) (0013)

Employment-to-population ratio -0282 -0248

(0037) (0038)

N 1127 1127 1127 1127 1127

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the SNAP enrollment rate All models include state fixed effects Census division x-year fixedeffects and state-specific linear time trends

Source Annual data from the March Current Population Survey for 1990 to 2012

TABLE 6

SNAP expenditures

State-level linear regression

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0181 -0149 -0156 -0153 -0190

(011) (0103) (0102) (0103) (0103)

Unemployment rate (100) 4587 4152 4232 4313

(0622) (0621) (0633) (0628)

Log median income -0246 -0261 -0294

(0075) (0078) (0078)

Employment-to-population ratio 0155 0244

(0237) (024)

N 1127 1127 1127 1127 1127

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the natural log of state SNAP expenditures per capita for 1990 to 2012 All models include state

fixed effects Census division x-year fixed effects and state-specific linear time trends

Source Annual data from Bureau of Economic Analysis national income and product accounts tables for 1990 to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3243

Appendix B | wwwamericanprogresso

TABLE 7

Comparison of specifications SNAP enrollment

Family-level linear probability

(1) (2) (3)

Log minimum wage -0028 -0049 -0042

(0014) (0017) (0009)

Unemployment rate (100) 0275 0297 0280

(0161) (0076) (0082)

Log median income -0077 -0055 -0039

(0014) (0012) (0011)

Employment-to-population ratio -0238 -0250 -0239

(0054) (004) (0038)

N 1242022 1242022 1242022

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses Observations are clustered at the state level The outcomevariable is binary or equal to one if a family is enrolled in SNAP All specifications include additional controls forfamily size race and marital status of the family head presence of children and presence of an adult male

Source Annual data from the March Current Population Survey for 1990 to 2012 estimation includes CurrentPopulation Survey probability weights

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3343

30 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

TABLE 8

Comparison of specifications SNAP enrollment

State-level linear regression

(1) (2) (3)

Log minimum wage -0019 -0035 -0031

(0009) (0012) (0012)

Unemployment rate (100) 0401 0370 0339

(0063) (0077) (0083)

Log median income -0081 -0073 -0061

(0011) (0013) (0013)

Employment-to-population ratio -0183 -0222 -0248

(0039) (0039) (0038)

N 1127 1127 1127

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the SNAP enrollment rate Allregressions include the share of households with children and the racial shares of the population

Source Annual data from the March Current Population Survey for 1990 to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3443

Appendix B | wwwamericanprogresso

TABLE 9

Comparison of specifications SNAP expenditures

State-level linear regression

(1) (2) (3)

Log minimum wage -0121 -0203 -0190

(0075) (0103) (0103)

Unemployment rate (100) 5292 5152 4313

(0464) (0576) (0628)

Log median income -0437 -0417 -0294

(008) (0086) (0078)

Employment-to-population ratio -0040 -0220 0244

(0261) (0260) (0240)

N 1127 1127 1127

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the natural log of state SNAPexpenditures per capita All regressions include the share of households with children and the racial shares ofthe state population

Source Annual data from Bureau of Economic Analysis national income and product accounts tables for 1990to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3543

32 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3643

Appendix C | wwwamericanprogresso

Appendix C Harkin-Miller

policy simulation results

StateMinimum wage

(2014)

SNAP enrollment (2012) Predicted change in SNAP enrollment

RateTotal

(persons)

Model 1 family

enrollment (linearprobability)

Model 2 state

enrollment (linearregression)

Model 3 s

expenditu(linear regre

Alabama $725 164 788682 -66922 -58897 -58906

Alaska $775 120 87436 -8104 -6887 -3288

Arizona $790 201 1319323 -67435 -56738 -64356

Arkansas $725 78 230489 -40977 -36063 -35248

California $800 206 7813680 -371131 -310222 -18223

Colorado $800 164 853155 -50684 -42365 -23926

Connecticut $870 91 326621 -22456 -17975 -13711

Delaware $725 186 170262 -12739 -11211 -10647

District of Columbia $825 133 84009 -5370 -4417 -3632

Florida $793 166 3208026 -195813 -164426 -13046

Georgia $725 160 1586336 -137741 -121224 -11004

Hawaii $725 96 133662 -19310 -16995 -14933

Idaho $725 92 147501 -22165 -19507 -15809

Illinois $825 95 1225084 -109088 -89742 -70955

Indiana $725 125 816233 -90818 -79928 -83985

Iowa $725 155 478011 -42716 -37594 -28556

Kansas $725 135 388269 -40082 -35275 -27461

Kentucky $725 130 568821 -60840 -53544 -52259

Louisiana $725 149 683832 -63929 -56263 -66083

Maine $750 77 101976 -16567 -14323 -15234

Maryland $725 144 846415 -81748 -71946 -38370

Massachusetts $800 130 864721 -64902 -54251 -42913

Michigan $740 146 1439141 -128801 -112140 -11022

Minnesota $725 133 713646 -74730 -65769 -37878

TABLE 10

SNAP enrollments Predicted changes by state under the Harkin-Miller bill

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3743

34 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

StateMinimum wage

(2014)

SNAP enrollment (2012) Predicted change in SNAP enrollment

RateTotal

(persons)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 state

expenditures

(linear regression)

Mississippi $725 129 386501 -41486 -36511 -46467

Missouri $750 172 1036182 -75131 -64952 -56244

Montana $790 132 132452 -10350 -8708 -5846

Nebraska $725 124 230683 -25773 -22683 -12189

Nevada $825 162 446035 -23349 -19209 -11894

New Hampshire $725 127 168404 -18359 -16157 -5735

New Jersey $825 160 1416666 -75175 -61843 -28236

New Mexico $750 149 310896 -25983 -22463 -22512

New York $800 192 3763553 -191193 -159815 -142182

North Carolina $725 174 1697193 -135417 -119179 -113503

North Dakota $725 87 61225 -9743 -8574 -4021

Ohio $795 143 1647345 -115869 -97169 -88580

Oklahoma $725 129 494053 -53006 -46650 -46854

Oregon $910 124 485326 -17036 -13328 -16398

Pennsylvania $725 161 2053643 -177315 -156052 -125586

Rhode Island $800 156 163730 -10258 -8574 -8698

South Carolina $725 94 445277 -65614 -57746 -50304

South Dakota $725 208 173749 -11586 -10197 -7458

Tennessee $725 142 914903 -89667 -78915 -99134

Texas $725 110 2863779 -362018 -318607 -253285

Utah $725 88 251107 -39658 -34902 -19390

Vermont $873 156 97792 -3823 -3055 -2475

Virginia $725 101 829771 -113723 -100086 -58212

Washington $932 72 496934 -23221 -17947 -17756

West Virginia $725 58 107875 -25792 -22699 -21665

Wisconsin $725 75 427822 -79521 -69986 -53210

Wyoming $725 164 94590 -8010 -7050 -3104

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3843

Appendix C | wwwamericanprogresso

StateMinimum wage

(2014)

SNAP expenditures (2012)Predicted change in SNAP expenditures

(millions of dollars)

Per resident

(millions of

dollars)

Total

(millions of

dollars)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 s

expenditu

(linear regre

Alabama $725 $290 $1397 -$1185 -$1043 -$1043

Alaska $775 $253 $185 -$262 -$223 -$106

Arizona $790 $257 $1686 -$935 -$787 -$892

Arkansas $725 $248 $733 -$636 -$560 -$547

California $800 $189 $7164 -$7276 -$6082 -$3573

Colorado $800 $157 $816 -$862 -$721 -$407

Connecticut $870 $191 $686 -$343 -$275 -$210

Delaware $725 $250 $229 -$205 -$180 -$171

District of Columbia $825 $366 $232 -$146 -$120 -$99

Florida $793 $294 $5676 -$4429 -$3719 -$2951

Georgia $725 $317 $3140 -$2936 -$2584 -$2346

Hawaii $725 $335 $465 -$449 -$395 -$347

Idaho $725 $225 $359 -$376 -$331 -$268

Illinois $825 $249 $3200 -$2096 -$1725 -$1364

Indiana $725 $220 $1439 -$1162 -$1023 -$1075

Iowa $725 $192 $589 -$658 -$579 -$440

Kansas $725 $159 $460 -$502 -$441 -$344

Kentucky $725 $298 $1303 -$1133 -$997 -$973

Louisiana $725 $315 $1450 -$1047 -$922 -$1083

Maine $750 $281 $373 -$267 -$231 -$246

Maryland $725 $188 $1109 -$1765 -$1553 -$828

Massachusetts $800 $206 $1366 -$1030 -$861 -$681

Michigan $740 $300 $2963 -$2400 -$2090 -$2054

Minnesota $725 $140 $755 -$1113 -$980 -$564

Mississippi $725 $326 $973 -$649 -$571 -$726

Missouri $750 $241 $1452 -$1278 -$1104 -$956

Montana $790 $190 $191 -$179 -$151 -$101

Nebraska $725 $140 $259 -$409 -$360 -$194

Nevada $825 $191 $527 -$441 -$363 -$225

New Hampshire $725 $126 $167 -$399 -$351 -$125

TABLE 11

SNAP expenditures Predicted changes by state under the Harkin-Miller bill

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3943

36 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

StateMinimum wage

(2014)

SNAP expenditures (2012)Predicted change in SNAP expenditures

(millions of dollars)

Per resident

(millions of

dollars)

Total

(millions of

dollars)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 state

expenditures

(linear regression)

New Jersey $825 $160 $1420 -$1611 -$1325 -$605

New Mexico $750 $324 $675 -$513 -$444 -$445New York $800 $287 $5616 -$3766 -$3148 -$2801

North Carolina $725 $252 $2454 -$2187 -$1925 -$1833

North Dakota $725 $128 $90 -$162 -$143 -$67

Ohio $795 $259 $2995 -$2013 -$1688 -$1539

Oklahoma $725 $248 $945 -$799 -$703 -$706

Oregon $910 $322 $1255 -$272 -$213 -$262

Pennsylvania $725 $218 $2779 -$2930 -$2579 -$2075

Rhode Island $800 $280 $294 -$173 -$144 -$147

South Carolina $725 $291 $1373 -$1337 -$1177 -$1025South Dakota $725 $198 $165 -$192 -$169 -$123

Tennessee $725 $324 $2091 -$1413 -$1243 -$1562

Texas $725 $230 $5997 -$6402 -$5634 -$4479

Utah $725 $141 $402 -$614 -$541 -$300

Vermont $873 $230 $144 -$66 -$53 -$43

Virginia $725 $173 $1413 -$2062 -$1815 -$1056

Washington $932 $244 $1682 -$350 -$270 -$267

West Virginia $725 $273 $508 -$451 -$397 -$379

Wisconsin $725 $204 $1166 -$1302 -$1146 -$871Wyoming $725 $95 $55 -$105 -$93 -$41

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4043

Appendix C | wwwamericanprogresso

FIGURE 2

Raising the minimum wage to $1010would cut taxpayer costs in every state

Predicted decreases in cost and enrollment

in SNAP in 50 states

$200+$51ndash$100

$101ndash$200

0ndash$25

$26ndash$50

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4143

38 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

Endnotes

1 US Department of Agriculture ldquoSupplementalNutrition Assistance Program (SNAP)rdquo available athttpwwwfnsusdagovsnapeligibility (last accessedFebruary 2014)

2 For this initial analysis we do not consider Harkin-Millerrsquos increase in subminimum wages for tippedworkers To do s o would increase the estimated SNAP

savings by an unknown amount

3 The Congressional Budget Office estimates thatworkers currently earning between $1010 and $1150per hour would see their wages rise under the Harkin-Miller proposal Congressional Budget O ffice ldquoTheEffects of a Minimum Wage Increase on Employmentand Family Incomerdquo (2014)

4 Marianne Page Joanne Spetz and Jane Millar ldquoDoesthe Minimum Wage Affect Welfare Caseloadsrdquo Journalof Policy Analysis and Management 24 (2) (2005)273ndash295

5 Marianne Bitler and Hilary Hoynes ldquoThe More ThingsChange the More They Stay the Same The SafetyNet Living Arrangements and Poverty in the GreatRecessionrdquo NBER Working Paper 194 49 2013

6 Sylvia Allegretto and others ldquoFast Food Poverty Wages The Public Cost of Low-Wage Jobs in the Fast-FoodIndustryrdquo (Berkeley California Center for LaborResearch and Education 2013) available at httplaborcenterberkeleyedupubliccostsfast_food_poverty_wage

7 David Neumark and William Wascher ldquoDoes a HigherMinimum Wage Enhance the Effectiveness of theEarned Income Tax Creditrdquo Industrial and LaborRelations Review 64 (4) (2011) 712ndash746

8 David Lee and Emmanuel Saez ldquoOptimal MinimumWage Policy in Competitive Labor Marketsrdquo Journal ofPublic Economics 96 (9) (2012) 739ndash749

9 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family IncomerdquoFebruary 2014

10 Jesse Rothstein ldquoIs the EITC as Good as an NITConditional Cash Transfers and Tax Incidencerdquo AmericanEconomic Journal Economic Policy 2 (1) (2010) 177ndash208

11 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family Incomerdquo

12 Dube Arindrajit 2013 rdquoMinimum Wagesand the Distribution of Family IncomerdquoUnpublished working paper Available at httpsdldropboxusercontentcomu15038936Dube_MinimumWagesFamilyIncomespdf

13 As Dube explains in ldquoThe poverty of Minimum WageFactsrdquo the simulation approach underestimate stemsfrom a number of unwarranted assumptions includingthe range of actual wage increases and the accuracy ofwage data in the Current Population Survey The causal

approach does not make these assumptions

14 Allegretto Sylvia and others 2013 ldquoCredible ResearchDesigns for Minimum Wage Studiesrdquo Working Paper148-13 University of California Berkeley Institutefor Research on Labor and Employment Available athttpwwwirleberkeleyeduworkingpapers148-13pdf

15 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family In comerdquo

16 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo (2012)

17 Ibid

18 Ibid

19 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo

20 US Department of Agriculture ldquoSupplementalNutrition Assistance Program (SNAP)rdquo

21 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo

22 Ibid

23 We deviate from the Census Bureaursquos definition ofa family unit which is ldquotwo people or more (on e ofwhom is the householder) related by bir th marriageor adoption and residing togetherrdquo US Bureau ofthe Census ldquoCurrent Population Survey DefinitionsFamilyrdquo available at httpwwwcensusgovcpsabout

cpsdefhtml (last accessed February 2014) We countas a family unit any individual residing on his or herown two or more persons residing together whodo not belong to a family in the March CPS sampleare constructed as one family in our analysis For thepurposes of food stamp allocations the consumptionresulting from this transfer is probably distributed tofamily members (rather than household members ora single individual within the household) Howeversingle individuals canmdashand domdashreceive SNAPbenefits Excluding them would fail to make theanalysis reflective of the population at large

24 Strictly the family level linear probability modelpredicts the percentage-point decrease in theprobability that an individual family will receive SNAPpayments When applied to a large number of familieshowever we are able to interpret the coefficient asa decrease in the mean of enrollmentmdashthat is a

decrease in the enrollment ratemdashby applying the lawof iterated expectations

25 We generate expenditure predictions from theenrollment modelsmdashand conversely generateenrollment predictions from the expenditure modelmdashby assuming that expenditures per enrolled familyremains the same before and after the minimum wagechange In practice this is likely to be a conservativeestimatemdashthat is to underestimate the decrease inSNAP activity Average SNAP benefits per family willalso decrease as many families that remain eligible forSNAP experience income gains

26 Wage and Hour Division ldquoMinimum Wage Laws inthe States ndash Januar y 1 2014rdquo available at httpwwwdolgovwhdminwageamericahtm (last accessedFebruary 2014)

27 See for example Sylvia Allegretto and others ldquoCredibleResearch Designs for Minimum Wage Studiesrdquo WorkingPaper 148-113 (Berkeley California Institute forResearch on Labor and Employment 2013) available athttpwwwirleberkeleyeduworkingpapers148-13pdf

28 We will report these results in a forthcoming workingpaper

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4243

Endnotes | wwwamericanprogresso

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4343

The Center for American Progress is a nonpartisan research and educational institute

dedicated to promoting a strong just and free America that ensures opportunity

for all We believe that Americans are bound together by a common commitment to

these values and we aspire to ensure that our national policies reflect these values

We work to find progressive and pragmatic solutions to significant domestic and

international problems and develop policy proposals that foster a government that

is ldquoof the people by the people and for the peoplerdquo

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 743

4 Center for American Progress | The Effects of M inim um Wages on SNAP Enro llments and Expenditures

Te repor proceeds as ollows

bull Secion 1 provides background inormaion on he ederal minimum wage sae

minimum wages and he SNAP program

bullSecion 2 describes our mehods and daa

bull Secion 3 provides our main resuls including a simulaion o he effecs o a

Harkin-Miller minimum wage increase and a sae-by-sae analysis

bull Secion 4 presens our conclusions

Furher deails are provided in a series o appendices

Correction April 28 2014 Tis report incorrectly stated the potential reduction in

SNAP enrollment fom the Harkin-Miller proposal Te correct amount is 75 percentand 87 percent (31 million to 36 million persons) as stated in the reportrsquos tables

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 843

Background | wwwamericanprogress

Background

The federal minimum wage

Te ederal minimum wage was las increased in July o 2009

nearly five years ago During he pas wo decades many saes

have passed legislaion fixing he minimum wage a a higher

level han he ederal minimum Te maps in Figure 1 show ha

while saes in every region o he Unied Saes have adoped

higher minimum wages hey are no disribued randomly bygeography As shown in he 2013 sudy ldquoCredible Research

Designs or Minimum Wage Sudiesrdquo by economiss Sylvia

Allegreto Arindraji Dube Michael Reich and Ben Zipperer

hese saes vary sysemaically rom he oher saes by a number

o characerisics ha affec low-wage employmen rends bu

which are no hemselves relaed o minimum wage policy14

Te nonrandom patern o minimum wage adopion has

imporan implicaions or obaining unbiased esimaes o

minimum wages on employmen In paricular naional panel

sudies ha use sae and ime fixed effec models991252such as

a 1992 sudy by David Neumark and William Wascher991252

spuriously esimae negaive employmen effecs Te reason

or his resul is uncovered using ess or pre-rends Tese

ess find ha low-wage employmen was already declining wo

years beore minimum wages were implemened By making a

saisically large number o local comparisons ha conrol or

heerogeneiy among saes and by ime eliminaes his pre-

rend For his reason we conduc similar ess or our SNAPoucomes and use model specificaions ha include local

comparisons as in he sudy cied above

FIGURE 1

High versus low minimum wage stat

from 1990 to 2012

Means and variances

More than $533

Average minimum wage over 1990ndash2012

Less than or equal to $533

More than $121

Minimum wage variance over 1990ndash2012

Less than or equal to $121

Notes State means and variances were calculated using annual stat

minimum wage data from 1990 to 2012 The shading on the maps

partitions the states into above- and below-median values

Source Sylvia Allegretto and others ldquoCredible Research Designs forMinimum Wage Studiesrdquo Working Paper 148-13 (Institute for Resear

on Labor and Employment 2013) available at httpwwwirleberkeleyeduworkingpapers148-13pdf

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 943

6 Center for American Progress | The Effects of M inim um Wages on SNAP Enro llments and Expenditures

Te CBO recenly projeced ha in 2016 17 million workers will earn less han

he $1010 hourly wage proposed in he Harkin-Miller bill Furhermore he

CBO esimaes ha an addiional 8 million workers earned beween $1010 and

$1150 per hour and were also likely o experience a wage increase15

Supplemental Nutrition Assistance Program

We ocus our inquiry on SNAP Benefis under he program are enirely

ederally unded he program is adminisered by he US Deparmen o

Agriculure ogeher wih sae agencies which share in adminisraion

coss Spending on SNAP has grown in he pas decade reaching $78 bil lion

in 2011 wih SNAP enrollmen increasing o 45 million people abou one-

sevenh o he US populaion16 According o he CBO changes since 1990 in

SNAP spending and enrollmens are primarily he resul o cyclical economic

condiions noably changes in he unemploymen rae and changes in per capiaincome17 Te 2009 American Recovery and Reinvesmen Ac emporarily

increased SNAP benefi amouns by 136 percen as repored by he US

Deparmen o Agriculurersquos Food and Nuriion Service hese higher benefi

levels expired on November 1 201318 Te CBO esimaes ha abou wo-

hirds o he changes in SNAP expendiure are associaed wih changes in he

number o recipiens and one-hird wih changes in he benefis received when

recipiensrsquo incomes change19

In fiscal year 2014 SNAPrsquos maximum monhly benefis are $189 or a single

individual $497 or a amily o hree and $750 or a amily o five Benefis are

reduced by 30 cens per dollar received and phase ou enirely a gross monhly

household incomes o 130 percen o he ederal povery level $1245 or a

single individual $2116 or a amily o hree and $2987 or a amily o five

o deermine benefis SNAP also defines a ne monhly income concep and

ses benefis a 100 percen o he ederal povery level using his concep

Calculaion o ne monhly income can include cerain deducions rom

monhly gross income such as medical expenses and child care coss Alhough

saes are permited some laiude on wha deducions are allowed in pracice

hese vary by very small amouns Our saisical model akes accoun o sae-specific differences in benefis20

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1043

Background | wwwamericanprogress

SNAP imposes an employmen or work-raining requiremen or able-bodied

prime-age aduls991252hose beween he ages o 18 and 50 and wihou disabiliies

or dependen children Such households can receive only hree monhs o benefis

in a hree-year period In recen years abou 85 percen o households receiving

benefis have incomes below he ederal povery level 49 percen have dependen

children 16 percen are age 60 or older 20 percen are disabled and 30 percenrepor some earned income21

A 2012 CBO repor also noes ha ake-up raes among eligible SNAP recipiens

average abou 70 percen wih much lower ake-up among elderly households

Te ake-up rae increases in harder economic imes I also increased when sigma

issues were reduced as SNAP debi cards replaced acual ood samps ake-up

is especially high among hose mos needy Adminisraive spending equaled 91

percen o he poenial spending ha would have occurred i all eligible recipiens

were enrolled Alhough some SNAP policy changes have occurred since 1990

mos were relaively minor and all were naional in scope Te 1996 welare reorm bill eliminaed SNAP eligibiliy or some legal immigrans limied he ime lengh

o eligibiliy or able-bodied childless aduls and reduced maximum benefis

Some o hese resricions were relaxed in 2002 and again in he American

Recovery and Reinvesmen Ac in 200922

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1143

8 Center for American Progress | The Effects of M inim um Wages on SNAP Enro llments and Expenditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1243

Methods and data | wwwamericanprogress

Methods and data

As previously menioned we exploi variaion in minimum wages by sae and ime

o examine heir causal effecs on SNAP enrollmens and expendiures o do so we

merge daa rom 1990 hrough 2012 drawn rom he Annual Social and Economic

Supplemen o he Curren Populaion Survey991252an annual Census Bureau survey

commonly known as he March CPS ha includes23 inormaion on SNAP

enrollmens a he amily level991252wih sae-level daa on minimum wages SNAP

expendiures populaion unemploymen raes and sae median income levels

o conrol or ime-varying heerogeneiy among saes our specificaions includeconrols or sae linear rends and effecs by Census division and ime We esimae

effecs a wo levels allowing or amily variaion and allowing only or sae-level

variaion We also employ a se o sandard demographic conrols such as amily size

and composiion and race and ehnic composiion

Distinguishing causation from correlation

How can we ensure ha our analysis does no pick up a spurious correlaion or

example he endency o more economically vibran saes o implemen higher

minimum wages Disinguishing correlaion and policy endogeneiy rom rue

causal effecs is he primary moivaion or economeric analysis In he ideal

experimen researchers would begin wih wo saes991252ha are alike in every

respec prior o he policy991252 and ldquoreardquo only one o hese saes wih a higher

minimum wage Tey would atemp o shield hese saes rom any influence ha

could obscure heir undersanding o he minimum wagersquos direc effec on SNAP

enrollmen Researchers o course canno conduc such experimens

We can however use saisical mehods o conrol simulaneously heindependen effecs on SNAP o sae unemploymen raes sae income levels

and common rajecories among saes wihin he same Census division By

ensuring similariy along hese dimensions we maximize he likelihood ha

SNAP aciviy in wo saes would have comparable oucomes in he absence

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1343

10 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

o a minimum wage policy change Tus i a new minimum wage policy were

implemened in one sae only he researchers could atribue all o he difference

hey observe in SNAP aciviy o he new minimum wage policy

In oher words we approximae he ideal experimen by using non-experimenal

saisical mehods Te desirable ldquopre-exising similariiesrdquo beween saes ha wehave defined above inorm our choice o conrol variables in a saisical seting

More precisely in our muliple regression models we use median amily income

he unemploymen rae he employmen-o-populaion raio and regional and

ime idenifiers o consruc an appropriae group o peers or each sae on he

eve o a policy change

Data description

wo daa ses include inormaion abou boh income and paricipaion inpublic programs Te Survey o Income and Program Paricipaion or SIPP

which is conduced in inermiten years has he advanage o ollowing he

same individuals over a period o ime In oher words i is a longiudinal daa

se I also has he advanage o conaining monhly daa However he sample

size o he SIPP is no sufficien or analyzing variaions in sae-level minimum

wages Te March CPS has he advanage o a much larger sample size and i is

conduced annually wihou any breaks in ime I has he disadvanage o being

a cross-secional daa se so we canno ollow he same individuals over ime991252

sricly speaking over more han one year On ne he March CPS is much more

suiable or our sudy We examine he empirical relaionship beween minimum

wage policy and ood samp aciviy a wo levels o aggregaion he amily level

and he sae level Family-level daa are drawn rom he March CPS

Te March CPS comprises responses rom he residens o 50000 o 60000

dwelling places surveyed per year and conains deailed inormaion on he

residensrsquo employmen and income including income rom ranser paymens

Te sample or our analysis comprises more han 128 million amily unis during

he period rom 1990 o 2012 (inclusive) Survey weighs allow us o analyze

SNAP paricipaion in a manner ha is represenaive o he US populaion alarge Over all years he share o amilies reporing ood samp receip in he

weighed March CPS sample is 91 percen Te enrollmen rae was a a low o 6

percen in he year 2000 In 2012 he mos recen year in our panel 133 percen

o amilies repored paricipaing in SNAP a some poin during he survey year

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1443

Methods and data | wwwamericanprogresso

Te March CPS also collecs inormaion on he number o SNAP recipiens in

he household he number o monhs he household receives SNAP benefis

and he cash-equivalen value o he SNAP benefis received However he

value o SNAP benefis is severely underrepored among recipiens perhaps

because recipiens are unaware o he exac monhly cash-equivalen value o

benefis hey receive

Our firs empirical sraegy ocuses solely on SNAP enrollmen By using he

amily as he uni o analysis we are able o inser saisical conrols o accoun

or non-wage-relaed acors ha influence any paricular amilyrsquos likelihood o

program paricipaion wih he inenion o isolaing any differences in program

paricipaion ha are due purely o changes in wage policy Tis approach

idenifies he effecs o low-wage labor policy on he exernal margin991252ha is

he effec o he minimum wage on he likelihood ha a amily paricipaes in he

SNAP program a all991252as opposed o he inernal margin or how much SNAP

unding he amily would receive

Our second empirical ramework uses sae-level adminisraive daa Ta is

we aggregae he daa o obain a single daa poin or each saeyear back

o 1990 represening he mean o he oucome or he sae Te sae-level

esimaion serves as a robusness check on he amily-level resuls or SNAP

paricipaion Also using aggregaed daa allows us o esimae direcly he

causal effec o minimum wage changes on SNAP spending Tis is no possible

a he amily level as discussed above daa on cash-equivalen value o ood

samps or SNAP recipiens is very requenly no repored in he March CPS

and when i is repored he inormaion may be unreliable By conras he

Bureau o Economic Analysis publishes aggregae SNAP spending a he sae

level in is Naional Income and Produc Accoun or NIPA ables Tus

while we are unable o observe he heerogeneiy in he cash value o SNAP

or amilies in each sae we are able o calculae average SNAP spending

per residen in each sae per year Supporing covariaes include he annual

unemploymen and employmen daa rom he Bureau o Labor Saisics or

BLS and sae-level populaion series rom he iner-decennial census releases

Minimum wage daa are available rom he BLSrsquos wages and hours division For

sae minimum wage changes enaced a oher imes han he firs o he yearan average value or he year is used

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1543

12 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

Family level model specifications

We firs examine he effec o he minimum wage on paricipaion in public

assisance programs For amily residing in sae and during year we esimae

an equaion o he ollowing orm

(1)

is a binary variable ha is se equal o 1 i a leas one member o amily

received ood samps during he survey year is a se o sae-level

characerisics including annual averages o he unemploymen rae he

employmen-o-populaion raio and he naural log o median amily income

is a vecor o amily atribues including indicaors or he race and marial saus

o he amily head size o he amily he presence o children and he presence oan adul male Sae fixed effecs are capured by o conrol or ime-varying

heerogeneiy our preerred model specificaion also includes year fixed effecs

ha vary by Census division ( ) and sae-level linear ime rends In

Appendix B we jusiy he inclusion o hese las wo erms We also compare he

resuls rom our preerred specificaion wih less sauraed specificaions

Te effec o ineres which is capured by is he expeced change in he

probabiliy o receiving SNAP benefis wih respec o a change in he (log o he)

binding minimum wage in sae during year We repor robus sandard errors

clusered a he sae level We esimae he parameers using linear regression

producing a linear probabiliy model Deails o he model selecion process are

covered in Appendix B below

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1643

Methods and data | wwwamericanprogresso

State-level model specifications

Te sae-level models are similarly specified For sae in year we assume ha

(2)

In his model is now eiher he SNAP enrollmen rae in sae during year

or he naural logarihm o per capia SNAP expendiures in sae during

year is once again a se o sae-level characerisics including he same

sae-level covariaes as in he amily regressions (annual average unemploymen

rae employmen-o-populaion raio naural log o median amily income)

wih he addiion o amily level characerisics averaged across he sae (average

amily size and he shares o populaion consiued by each o five racialehnicgroups) Sae fixed effecs are represened by As above our preerred model

specificaion includes year fixed effecs ha vary by Census division ( ) and

sae-level linear ime rends as elaboraed in Appendix B Te effec o

ineres is capured by

We esimae boh sae-level models (enrollmen and expendiures) using

ordinary leas squares regression Tus he inerpreaion o he coefficien is no

longer ha o a change in probabiliy as in he binary oucome models described

above Raher or he sae-level SNAP enrollmen model represens he

expeced change (in percenage poins) in he saersquos SNAP enrollmen rae ha

is due o a 1 percen change in he minimum wage For he SNAP expendiures

model is simply he elasiciy o SNAP spending wih respec o he minimum

wage991252ha is he percenage change in sae expendiures expeced o resul rom

a 1 percen change in ha saersquos minimum wage For urher deails on model

specificaion reer o Appendix B below

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1743

14 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1843

Results | wwwamericanprogresso

Results

Estimated minimum wage effects on SNAP enrollment and

expenditures

able 1 shows he esimaed parameer o ineres991252he coefficien o he

minimum wage or he preerred model o each ype Coefficiens on he

minimum wage variable are no direcly comparable across models because all

our models have a differen uncional orm o undersand and compare hese

esimaes we compue he change in SNAP aciviy prediced or a paricular wage scenario Te final column in able 1 answers he quesion Wha would be

he expeced change in SNAP aciviy in response o a 10 percen increase in he

minimum wage Te answer o his quesion varies wih he value o he inpu

parameers in he able we calculae he percenage decrease in enrollmen or

expendiures prediced or he average sae wih a minimum wage o $725 in

2014 Te sae-level SNAP expendiure model which is a consan-elasiciy

model conveys elasiciy inormaion direcly or he change in expendiures per

capia in he sae

TABLE 1

Comparison of national SNAP predictions for a 10 percent increase in the federal minimum wage

Model LevelRegression

type

Predicted outcome Coefficient of log

(minimum wage)

(Standard error)

Effect of a 10 percent in

in the minimum wage

VariableForm of

variable

Total

enrollment

Tot

expend

1 Family Linear

probability Enrollment Binary (enrolled=1)

-0042

(0008) -317 N

2

StateLinear regression

(ordinary least

squares)

Enrollment State enrollment rate (percent) -0031

(0012) -235 N

3 Expenditures Log (state expenditures per capita) -0190

(0103) NA -19

plt01 plt005 plt001Note Predicted changes are calculated for the average state with a minimum wage of $725 in 2014

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1943

16 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

According o his model a 10 percen increase in he minimum wage would resul

in a 19 percen reducion in SNAP expendiures Tis resul is reassuringly similar

o Dubersquos resul or povery reducion Te wo enrollmen models are much more

precisely esimaed han he expendiure model Te sae-level enrollmen model

finds ha a 10 percen minimum wage increase in a low-wage sae is associaed wih

a 235 percen decrease in SNAP enrollmens Te amily-level linear probabiliymodel predics a somewha greaer elasiciy or low minimum wage saes an

increase o 10 percen in he ederal minimum would resul in a 317 percen decline

in SNAP enrollmen24 Te differences in hese esimaes sem rom a number o

acors including difference in model uncional orm and daa used We rea his

range o elasiciy esimaes as an upper and lower bound on enrollmen impacs

Harkin-Miller bill National and state-level predicted impacts

Wha would be he prediced change or he SNAP program i he ederalminimum were raised o $1010 as proposed in he Harkin-Miller bill In order o

make his inerence we accoun or he ac ha no all saes are currenly subjec

o he ederal minimum wage a he beginning o 2014 21 saes mainained

higher minimum wages han $725 In hose saes an increase in he ederal

minimum wage may or may no be binding or employers in he sae depending

upon wheher he new ederal minimum exceeds he sae-level minimum Bu

regardless o wheher a minimum wage change is binding he impac on SNAP

aciviy will be lower in high minimum wage saes In order o accoun or his

properly we calculae sae by sae he percenage wage change ha would resul

rom he Harkin-Miller proposal and apply he parameers rom each o he hree

models above o compue he expeced decrease in SNAP aciviy or each sae

In his exercise we use saesrsquo curren (2014) minimum wage levels and assume

as a baseline he 2012 levels o SNAP enrollmen and expendiure as 2012 is he

mos recen year or which SNAP daa are available

able 10 and able 11 in Appendix C repor he esimaed effecs on SNAP

enrollmen and expendiures respecively or each sae under he Harkin-Miller

bill25 An increase o $1010 i enaced oday would represen beween a 393

percen wage increase in a $725 minimum wage sae and an 84 percen increasein Washingon sae which has he highes minimum wage in he naion a $932

as o January 201426 Slighly more han 56 percen o he decrease in expendiures

and abou 59 percen o he decrease in enrollmen would occur in saes wih

presen-day minimum wages o $725

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2043

Results | wwwamericanprogresso

In 2012 hese saes were home o 46 percen o he American populaion

and accouned or a roughly equivalen percenage o oal naional SNAP

expendiures No surprisingly he larges enrollmen decreases come rom

populous saes wih very high SNAP enrollmen raes andor very low minimum

wages Te larges prediced enrollmen reducion991252beween 319000 individuals

and 362000 individuals991252would occur in exas which has a $725 minimum wage In Caliornia which has a 206 percen SNAP paricipaion rae and an $8

per hour minimum wage we anicipae SNAP enrollmen declines o 310000

persons o 371000 persons And in Florida which had a SNAP paricipaion rae

o 166 percen and a $793 minimum wage enrollmen could decline by beween

164000 individuals and 196000 individuals For he our saes ha ogeher

accouned or he greaes amoun o SNAP spending in 2012991252exas Caliornia

Florida and New York respecively991252he combined expendiure reducion rom

he Harkin-Miller bill is prediced o be $14 billion

able 2 summarizes he prediced declines in SNAP aciviy or he naion as a whole ha would resul rom he direc and indirec effecs o he Harkin-Miller

bill Enrollmen would all beween 31 million persons and 36 million persons

represening 75 percen o 87 percen o curren enrollmen Te anicipaed

reducion in program expendiures would be nearly $46 billion or 61 percen o

program expendiures

TABLE 2

Comparison of national SNAP predictions under the Harkin-Miller billrsquos $1010 minimum wage

Model

Enrollment

(persons)

Expenditures

(millions of dollars)

Current (2012) Predicted Change Current (2012) Predicted Chan

Family enrollment (linear probability)

41866195

45489339 -3623144

$74861

NA NA

State enrollment (ordinary least squares) 38745435 -3120759 NA NA

State expenditures (ordinary least squares) NA NA $70305 -$45

Note Calculations use 2014 state minimum wages and the most recent SNAP data from 2012 They assume that per-enrollee expenditures remain constant

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2143

18 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

Tere are o course oher possibiliies or minimum wage legislaion able 3

shows he expeced SNAP changes or he Unied Saes under a variey o wage

scenarios calculaed using he sae-level models I saes were no able o se

heir minimum wages independenly such ha all saes were consrained by

he ederal minimum o $725 SNAP would be received by abou 514000 more

people across he Unied Saes a an addiional program cos o nearly hree-quarers o a billion dollars In conras he effecs o a higher minimum wage

proposal991252a ederal wage floor o $11 per hour991252would decrease enrollmen in

SNAP by more han 10 percen and decrease program coss by 83 percen

TABLE 3

Summary of par ticipation and expenditures under wage scenarios

If all states had

minimum wages of

Enrollment(persons)

Expenditures

(millions of dollars)

Predicted Change Predicted Change

Recent levels (2014) 41866195 $74861

$725 42380520 514326 $75604 $743

$800 41423919 -442276 $74209 -$652

$900 40148451 -1717744 $72350 -$2511

$1000 38872982 -2993212 $70490 -$4371

$1010 38745435 -3120759 $70305 -$4556

$1100 37597514 -4268681 $68631 -$6230

Note Calculations use state-level enrollment model coefficient

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2243

Conclusion | wwwamericanprogresso

Conclusion

An exensive body o lieraure examines employmen effecs o he minimum

wage A much smaller se o sudies examines how he minimum wage affecs

povery and only a handul o papers examine he effecs o he minimum wage

on he EIC Our analysis is he firs o examine he effecs o he minimum

wage on SNAP

Our findings indicae ha increased earnings rom minimum wage changes do

reduce SNAP enrollmens and expendiures We esimae ha he Harkin-Miller bill would save axpayers nearly $46 billion per year equivalen o 61 percen

o SNAP expendiures in 2012 he las year or which daa are available Over a

10-year period he esimaed savings amoun o nearly $46 billion

Our repor is subjec o limiaions ha we expec o overcome in our uure

research Firs he findings do no ake ino accoun possible ineracions among

SNAP he EIC and Medicaid Te eligibiliy cuoffs among hese programs

are quie differen suggesing ha such ineracions may be minor Noneheless

he join effecs can only be deermined by urher research using a causal

model Second i would be useul o know he disribuion o SNAP reducions

along he wage disribuion Using he Congressional Budge Officersquos calculaions

o how much he oal dollar value o wage would increase under he Harkin-

Miller proposal our findings imply ha he decline in overall SNAP spending

equals abou 15 percen o he oal resuling increase in wages Te amoun and

disribuion o his offse are o considerable ineres Minimum wage beneficiaries

who come rom working amilies already well above he povery line would no

see any offse while hose who are currenly considerably below he povery line

will see larger offses Tese issues will also be a subjec or our uure research

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2343

20 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

About the authors

Rachel West is a maser o public policy candidae a he Goldman School

o Public Policy Universiy o Caliornia Berkeley Her research ocuses on

economic policy in he areas o low-wage labor and povery

Michael Reich is proessor o economics and direcor o he Insiue or

Research on Labor and Employmen a he Universiy o Caliornia a Berkeley

His research publicaions cover numerous areas o labor economics including

racial inequaliy labor marke segmenaion high-perormance workplaces

union-managemen cooperaion Japanese labor-managemen sysems living

wages and minimum wages He received his docorae in economics rom

Harvard Universiy

Acknowledgments

We are graeul o Sylvia Allegreto Arindraji Dube Bill Leser Jesse Rohsein

Daniel Tompson and Ben Zipperer or heir valuable suggesions

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2443

References | wwwamericanprogresso

References

Allegreto Sylvia and ohers 2013 ldquoFas Food Povery Wages Te Public Cos o Low-Wage Jobsin he Fas-Food Indusryrdquo Universiy o Caliornia Berkeley Cener or Labor Research andEducaion Available a htplaborcenerberkeleyedupubliccossas_ood_povery_wage

Allegreto Sylvia and ohers 2013 ldquoCredible Research Designs or Minimum Wage Sudiesrdquo

Working Paper 148-13 Universiy o Caliornia Berkeley Insiue or Research on Labor andEmploymen Available a htpwwwirleberkeleyeduworkingpapers148-13pd

Biler Marianne and Hilary Hoynes 2013 ldquo Te More Tings Change he More Tey Say heSame Te Saey Ne Living Arrangemens and Povery in he Grea Recessionrdquo Working Paper19449 Naional Bureau o Economic Research

Congressional Budge Office 2012 ldquoTe Supplemenal Nuriion Assisance Programrdquo Washingon Available a htpwwwcbogovsiesdeaulfilescbofilesatachmens04-19-SNAPpd

991252 991252 991252 2014 ldquoTe Effec o a Minimum-Wage Increase on Employmen and Family Incomerdquo Washingon Available a htpwwwcbogovsiesdeaulfilescbofilesatachmens44995-MinimumWagepd

Dube Arindraji 2013 rdquoMinimum Wages and he Disribuion o Family Incomerdquo Unpublished working paper Available a htpsdldropboxuserconencomu15038936Dube_ MinimumWagesFamilyIncomespd

991252 991252 991252 2014 ldquoTe Povery o Minimum Wage Facsrdquo Arindraji Dube Blog January 22 Availablea htparindubecom20140122he-povery-o-minimum-wage-acs

Lee David and Emmanuel Saez 2012 ldquoOpimal Minimum Wage Policy in Compeiive LaborMarkesrdquo Journal o Public Economics 96 (9) 739ndash749

Neumark David and William Wascher 1992 ldquoEmploymen Effecs o Minimum and Subminimum Wages Panel Daa on Sae Minimum Wage Lawsrdquo Industrial and Labor Relations Review 46 (1)

55ndash81

Neumark David and William Wascher 2011 ldquoDoes a Higher Minimum Wage Enhance heEffeciveness o he Earned Income ax Credirdquo Industrial and Labor Relations Review 64 (4)712ndash746

Page Marianne Joanne Spez and Jane Millar 2005 ldquoDoes he Minimum Wage Affec WelareCaseloadsrdquo Journal o Policy Analysis and Management 24 (2) 273ndash295

Rohsein Jesse 2010 ldquoIs he EIC as Good as an NI Condiional Cash ransers and ax

Incidencerdquo American Economic Journal Economic Policy 2 (1) 177ndash208

Wage and Hours Division 2014 ldquoMinimum Wage Laws in he Saes ndash January 1 2014rdquo USDeparmen o Labor (htpwwwdolgovwhdminwageamericahm [February 2014])

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2543

22 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2643

Appendix A | wwwamericanprogresso

Appendix A

Pre-trend falsification check

Recen minimum wage research27 highlighs a common flaw in previous sudies

ailure o veriy ha he oucome variable is ree o negaive pre-exising rends I

or example SNAP aciviy was already rending down in saes ha raised heir

minimum wages beore hese changes came ino effec our regression analysis could

(misakenly) atribue ha reducion o he minimum wage We check or such

pre-rends by inroducing variables ha represen he prior yearrsquos value or leads o

he minimum wage I he model esimaes he minimum wage o have an effec on

he oucome variable beore he wage change wen ino effec hen an unobservedacor no he minimum wage change caused he change in SNAP aciviy

We es he specificaions above or pre-rends by including a one-year lead in

all hree specificaions We find ha he lead erms are small posiive and no

saisically significan indicaing ha he concurren minimum wage991252no

he wage level in prior periods991252is driving he observed changes in SNAP

oucomes28 In paricular he coefficien (sandard error) on he lead erm in

our preerred amily-level enrollmen regression is 011 and no significan

while he coefficien and sandard error o he conemporaneous minimum

wage is unchanged In he sae-level preerred enrollmen regression he

coefficien o he lead erm is again small (07) and i is no significan Te

corresponding coefficien on he lead erm in he sae-level expendiure

regression is 16 and is no significan Te posiive poin esimaes on hese lead

erms resuls no only rule ou disoring negaive pre-rends Tey also sugges

ha our main resuls may underesimae he rue effecs

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2743

24 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2843

Appendix B | wwwamericanprogresso

Appendix B

Model selection process

For boh he amily-level and sae-level models we es hree mehods o conrol

or unobserved geographic- and ime-varying characerisics as suggesed by he

minimum wage lieraure Firs we include only independen sae-specific fixed

effecs and year-specific fixed effecs Tis specificaion (specificaion 1) implicily

assumes ha amilies in any sae consiue an equally good saisical ldquoconrolrdquo

group or hose in any randomly chosen sae afer accouning or various

characerisics (median income and unemploymen rae among ohers) Similarly

simple ime fixed effecs assume ha amilies surveyed in any year can crediblyserve as a conrol group or amilies surveyed in every oher year o he sample

(1990 hrough 2012)

In oher words specificaion 1 assumes ha a saersquos immediae neighbor provides

no beter a couneracual or he effec o a minimum wage change han does a

sae across he counry We relax his resricive specificaion sequenially in wo

seps In specificaion 2 we replace simple year fixed effecs wih fixed effecs or

each Census divisionyear (capured as an addiional variable in he vecor By

using division-year effecs we remove he resricion ha amilies in each sae

are equally good saisical conrols or all oher amilies Raher we allow or he

possibiliy ha amilies in similar geographic regions (or example he Souh or

he Norheas) may be more similar o one anoher han amilies arher away

Finally in specificaion 3 we add sae-specific linear ime rends o he previous

specificaion Tus specificaion 3 is he mos rigorous model specificaion in ha

i allows or heerogeneiy along hree dimensions Ta is specificaion 3 allows

each sae o have is own ime-varying rends raher han imposing he resricion

ha saes evolve idenically over he 22 years in he sample

We begin building he heoreical specificaion above rom a se o simpleuncondiional models regression o SNAP aciviy (enrollmen or expendiures)

on he log o he minimum wage and a se o geographic- and ime-specific

effecs (specificaions 1 2 and 3 described above) As shown in ables 1ndash3 (or

specificaion 3) we hen add covariaes sequenially o hese models including

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2943

26 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

firs he vecor o amily-level conrols ollowed by each o several sae-level

covariaes in urn he unemploymen rae log o median-amily income and he

employmen-o-populaion raio Comparable resuls or specificaions 2 and 3

will be available in our orhcoming working paper

As expeced he simple uncondiional models indicae ha he relaionship beween he minimum wage and SNAP enrollmen i one exiss is a more complex

one influenced by oher acors In he uncondiional model he coefficien on he

variable o ineres991252he log o he minimum wage991252is small in magniude and no

saisically differen rom zero Once we accoun or he influence o labor marke

condiions and variaion in income levels on program paricipaion (by including

unemploymen rae and median-amily income conrol variables respecively)

he effec o he minimum wage on SNAP enrollmen is precisely esimaed Te

coefficien o he log minimum wage is slighly higher (-0042) in he amily-level

analysis han he coefficien (-031) in he sae-level analysis Te level o precision

is also higher in he amily-level analysis Tis is o be expeced when using 124million observaions compared o 1127

Te second dimension o model choice concerns he effec specificaion ables

7ndash9 compare he primary coefficiens o ineres or he SNAP enrollmen and

expendiure models For boh he enrollmen models he effec sizes are smalles

or specificaion 1 larges or specificaion 2 and inermediae beween hese wo

in specificaion 3 Recall ha Specificaion 3 conains sae-specific linear ime

rends in addiion o he census divisionyear conrols included in specificaion

2 In he amily-level enrollmen model he sandard error o he minimum wage

coefficien is smaller han in he oher wo specificaions Sandard errors on he

oher variables are much smaller in specificaions 2 and 3 han in specificaion

1 On he basis o coefficien significance (join and individual) specificaions 2

and 3 are sricly preerred in boh enrollmen models o specificaion 1 which

conains only sae and year fixed effecs

A concern wih specificaions 2 and 3 is ha rend conrols such as sae linear

rends may incorrecly absorb some o he delayed impac o a minimum wage

When we es his issue by including lagged minimum wages we do no find ha

delayed effecs are significan Anoher concern is ha more sauraed modelsuse less o he saisical variaion which could reduce he saisical power o

he resuls However he sandard errors or our more sauraed models are no

higher and are lower in some cases han or he less sauraed models Overall

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3043

Appendix B | wwwamericanprogresso

his evidence suppors our use o he sauraed model as he preerred model

specificaion Moreover Dubersquos 2013 sudy shows ha more sauraed models

perorm beter han models wih jus sae and ime fixed effecs

Te esimaed enrollmen regressions a boh he amily and sae levels show large

and saisically significan coefficiens Te esimaed minimum wage effec in heexpendiures regressions991252or which we have only sae-level daa991252is also large

and saisically significan

We do no use weighed regression or he sae-level models preerring o keep

analysis o he ldquoreamenrdquo (ha is o say a minimum wage change) appropriae

o he average sae raher han he average amily or individual I insead our

primary ineres were he impac o a minimum wage change on he average amily

or he average individual we migh choose o designae he number o amilies

in each sae or he sae populaion respecively as analyic weighs in order o

obain a coefficien beter suied or such inerence

TABLE 4

SNAP enrollment

Family-level linear probability

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0048 -0047 -0040 -0043 -0042

(0013) (0013) (001) (0008) (0009)

Unemployment rate (100) 0505 0420 0280

(0083) (0086) (0082)

Log median income -0057 -0039

(0011) (0011)

Employment-to-population ratio -0239

(0038)

N 1242022 1242022 1242022 1242022 1242022

plt01 plt005 plt001

Note Robust standard errors are in parentheses Observations are clustered at the state level The outcome variable is binary and equal to one if a family is enrolledin SNAP All models include state fixed effects Census division x-year fixed effects and state-specific linear time trends All specifications except 3a include additional

controls for family size race and marital status of the family head presence of children and presence of an adult maleSource Annual data from the March Current Population Survey for 1990 to 2012 estimation includes Current Population Survey probability weights

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3143

28 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

TABLE 5

SNAP enrollment

State-level linear regression

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0028 -0024 -0026 -0031 -0031

(0014) (0013) (0013) (0012) (0012)

Unemployment rate (100) 0625 0466 0320 0339

(0087) (0088) (0085) (0083)

Log median income -0090 -0065 -0061

(0013) (0013) (0013)

Employment-to-population ratio -0282 -0248

(0037) (0038)

N 1127 1127 1127 1127 1127

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the SNAP enrollment rate All models include state fixed effects Census division x-year fixedeffects and state-specific linear time trends

Source Annual data from the March Current Population Survey for 1990 to 2012

TABLE 6

SNAP expenditures

State-level linear regression

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0181 -0149 -0156 -0153 -0190

(011) (0103) (0102) (0103) (0103)

Unemployment rate (100) 4587 4152 4232 4313

(0622) (0621) (0633) (0628)

Log median income -0246 -0261 -0294

(0075) (0078) (0078)

Employment-to-population ratio 0155 0244

(0237) (024)

N 1127 1127 1127 1127 1127

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the natural log of state SNAP expenditures per capita for 1990 to 2012 All models include state

fixed effects Census division x-year fixed effects and state-specific linear time trends

Source Annual data from Bureau of Economic Analysis national income and product accounts tables for 1990 to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3243

Appendix B | wwwamericanprogresso

TABLE 7

Comparison of specifications SNAP enrollment

Family-level linear probability

(1) (2) (3)

Log minimum wage -0028 -0049 -0042

(0014) (0017) (0009)

Unemployment rate (100) 0275 0297 0280

(0161) (0076) (0082)

Log median income -0077 -0055 -0039

(0014) (0012) (0011)

Employment-to-population ratio -0238 -0250 -0239

(0054) (004) (0038)

N 1242022 1242022 1242022

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses Observations are clustered at the state level The outcomevariable is binary or equal to one if a family is enrolled in SNAP All specifications include additional controls forfamily size race and marital status of the family head presence of children and presence of an adult male

Source Annual data from the March Current Population Survey for 1990 to 2012 estimation includes CurrentPopulation Survey probability weights

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3343

30 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

TABLE 8

Comparison of specifications SNAP enrollment

State-level linear regression

(1) (2) (3)

Log minimum wage -0019 -0035 -0031

(0009) (0012) (0012)

Unemployment rate (100) 0401 0370 0339

(0063) (0077) (0083)

Log median income -0081 -0073 -0061

(0011) (0013) (0013)

Employment-to-population ratio -0183 -0222 -0248

(0039) (0039) (0038)

N 1127 1127 1127

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the SNAP enrollment rate Allregressions include the share of households with children and the racial shares of the population

Source Annual data from the March Current Population Survey for 1990 to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3443

Appendix B | wwwamericanprogresso

TABLE 9

Comparison of specifications SNAP expenditures

State-level linear regression

(1) (2) (3)

Log minimum wage -0121 -0203 -0190

(0075) (0103) (0103)

Unemployment rate (100) 5292 5152 4313

(0464) (0576) (0628)

Log median income -0437 -0417 -0294

(008) (0086) (0078)

Employment-to-population ratio -0040 -0220 0244

(0261) (0260) (0240)

N 1127 1127 1127

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the natural log of state SNAPexpenditures per capita All regressions include the share of households with children and the racial shares ofthe state population

Source Annual data from Bureau of Economic Analysis national income and product accounts tables for 1990to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3543

32 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3643

Appendix C | wwwamericanprogresso

Appendix C Harkin-Miller

policy simulation results

StateMinimum wage

(2014)

SNAP enrollment (2012) Predicted change in SNAP enrollment

RateTotal

(persons)

Model 1 family

enrollment (linearprobability)

Model 2 state

enrollment (linearregression)

Model 3 s

expenditu(linear regre

Alabama $725 164 788682 -66922 -58897 -58906

Alaska $775 120 87436 -8104 -6887 -3288

Arizona $790 201 1319323 -67435 -56738 -64356

Arkansas $725 78 230489 -40977 -36063 -35248

California $800 206 7813680 -371131 -310222 -18223

Colorado $800 164 853155 -50684 -42365 -23926

Connecticut $870 91 326621 -22456 -17975 -13711

Delaware $725 186 170262 -12739 -11211 -10647

District of Columbia $825 133 84009 -5370 -4417 -3632

Florida $793 166 3208026 -195813 -164426 -13046

Georgia $725 160 1586336 -137741 -121224 -11004

Hawaii $725 96 133662 -19310 -16995 -14933

Idaho $725 92 147501 -22165 -19507 -15809

Illinois $825 95 1225084 -109088 -89742 -70955

Indiana $725 125 816233 -90818 -79928 -83985

Iowa $725 155 478011 -42716 -37594 -28556

Kansas $725 135 388269 -40082 -35275 -27461

Kentucky $725 130 568821 -60840 -53544 -52259

Louisiana $725 149 683832 -63929 -56263 -66083

Maine $750 77 101976 -16567 -14323 -15234

Maryland $725 144 846415 -81748 -71946 -38370

Massachusetts $800 130 864721 -64902 -54251 -42913

Michigan $740 146 1439141 -128801 -112140 -11022

Minnesota $725 133 713646 -74730 -65769 -37878

TABLE 10

SNAP enrollments Predicted changes by state under the Harkin-Miller bill

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3743

34 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

StateMinimum wage

(2014)

SNAP enrollment (2012) Predicted change in SNAP enrollment

RateTotal

(persons)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 state

expenditures

(linear regression)

Mississippi $725 129 386501 -41486 -36511 -46467

Missouri $750 172 1036182 -75131 -64952 -56244

Montana $790 132 132452 -10350 -8708 -5846

Nebraska $725 124 230683 -25773 -22683 -12189

Nevada $825 162 446035 -23349 -19209 -11894

New Hampshire $725 127 168404 -18359 -16157 -5735

New Jersey $825 160 1416666 -75175 -61843 -28236

New Mexico $750 149 310896 -25983 -22463 -22512

New York $800 192 3763553 -191193 -159815 -142182

North Carolina $725 174 1697193 -135417 -119179 -113503

North Dakota $725 87 61225 -9743 -8574 -4021

Ohio $795 143 1647345 -115869 -97169 -88580

Oklahoma $725 129 494053 -53006 -46650 -46854

Oregon $910 124 485326 -17036 -13328 -16398

Pennsylvania $725 161 2053643 -177315 -156052 -125586

Rhode Island $800 156 163730 -10258 -8574 -8698

South Carolina $725 94 445277 -65614 -57746 -50304

South Dakota $725 208 173749 -11586 -10197 -7458

Tennessee $725 142 914903 -89667 -78915 -99134

Texas $725 110 2863779 -362018 -318607 -253285

Utah $725 88 251107 -39658 -34902 -19390

Vermont $873 156 97792 -3823 -3055 -2475

Virginia $725 101 829771 -113723 -100086 -58212

Washington $932 72 496934 -23221 -17947 -17756

West Virginia $725 58 107875 -25792 -22699 -21665

Wisconsin $725 75 427822 -79521 -69986 -53210

Wyoming $725 164 94590 -8010 -7050 -3104

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3843

Appendix C | wwwamericanprogresso

StateMinimum wage

(2014)

SNAP expenditures (2012)Predicted change in SNAP expenditures

(millions of dollars)

Per resident

(millions of

dollars)

Total

(millions of

dollars)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 s

expenditu

(linear regre

Alabama $725 $290 $1397 -$1185 -$1043 -$1043

Alaska $775 $253 $185 -$262 -$223 -$106

Arizona $790 $257 $1686 -$935 -$787 -$892

Arkansas $725 $248 $733 -$636 -$560 -$547

California $800 $189 $7164 -$7276 -$6082 -$3573

Colorado $800 $157 $816 -$862 -$721 -$407

Connecticut $870 $191 $686 -$343 -$275 -$210

Delaware $725 $250 $229 -$205 -$180 -$171

District of Columbia $825 $366 $232 -$146 -$120 -$99

Florida $793 $294 $5676 -$4429 -$3719 -$2951

Georgia $725 $317 $3140 -$2936 -$2584 -$2346

Hawaii $725 $335 $465 -$449 -$395 -$347

Idaho $725 $225 $359 -$376 -$331 -$268

Illinois $825 $249 $3200 -$2096 -$1725 -$1364

Indiana $725 $220 $1439 -$1162 -$1023 -$1075

Iowa $725 $192 $589 -$658 -$579 -$440

Kansas $725 $159 $460 -$502 -$441 -$344

Kentucky $725 $298 $1303 -$1133 -$997 -$973

Louisiana $725 $315 $1450 -$1047 -$922 -$1083

Maine $750 $281 $373 -$267 -$231 -$246

Maryland $725 $188 $1109 -$1765 -$1553 -$828

Massachusetts $800 $206 $1366 -$1030 -$861 -$681

Michigan $740 $300 $2963 -$2400 -$2090 -$2054

Minnesota $725 $140 $755 -$1113 -$980 -$564

Mississippi $725 $326 $973 -$649 -$571 -$726

Missouri $750 $241 $1452 -$1278 -$1104 -$956

Montana $790 $190 $191 -$179 -$151 -$101

Nebraska $725 $140 $259 -$409 -$360 -$194

Nevada $825 $191 $527 -$441 -$363 -$225

New Hampshire $725 $126 $167 -$399 -$351 -$125

TABLE 11

SNAP expenditures Predicted changes by state under the Harkin-Miller bill

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3943

36 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

StateMinimum wage

(2014)

SNAP expenditures (2012)Predicted change in SNAP expenditures

(millions of dollars)

Per resident

(millions of

dollars)

Total

(millions of

dollars)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 state

expenditures

(linear regression)

New Jersey $825 $160 $1420 -$1611 -$1325 -$605

New Mexico $750 $324 $675 -$513 -$444 -$445New York $800 $287 $5616 -$3766 -$3148 -$2801

North Carolina $725 $252 $2454 -$2187 -$1925 -$1833

North Dakota $725 $128 $90 -$162 -$143 -$67

Ohio $795 $259 $2995 -$2013 -$1688 -$1539

Oklahoma $725 $248 $945 -$799 -$703 -$706

Oregon $910 $322 $1255 -$272 -$213 -$262

Pennsylvania $725 $218 $2779 -$2930 -$2579 -$2075

Rhode Island $800 $280 $294 -$173 -$144 -$147

South Carolina $725 $291 $1373 -$1337 -$1177 -$1025South Dakota $725 $198 $165 -$192 -$169 -$123

Tennessee $725 $324 $2091 -$1413 -$1243 -$1562

Texas $725 $230 $5997 -$6402 -$5634 -$4479

Utah $725 $141 $402 -$614 -$541 -$300

Vermont $873 $230 $144 -$66 -$53 -$43

Virginia $725 $173 $1413 -$2062 -$1815 -$1056

Washington $932 $244 $1682 -$350 -$270 -$267

West Virginia $725 $273 $508 -$451 -$397 -$379

Wisconsin $725 $204 $1166 -$1302 -$1146 -$871Wyoming $725 $95 $55 -$105 -$93 -$41

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4043

Appendix C | wwwamericanprogresso

FIGURE 2

Raising the minimum wage to $1010would cut taxpayer costs in every state

Predicted decreases in cost and enrollment

in SNAP in 50 states

$200+$51ndash$100

$101ndash$200

0ndash$25

$26ndash$50

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4143

38 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

Endnotes

1 US Department of Agriculture ldquoSupplementalNutrition Assistance Program (SNAP)rdquo available athttpwwwfnsusdagovsnapeligibility (last accessedFebruary 2014)

2 For this initial analysis we do not consider Harkin-Millerrsquos increase in subminimum wages for tippedworkers To do s o would increase the estimated SNAP

savings by an unknown amount

3 The Congressional Budget Office estimates thatworkers currently earning between $1010 and $1150per hour would see their wages rise under the Harkin-Miller proposal Congressional Budget O ffice ldquoTheEffects of a Minimum Wage Increase on Employmentand Family Incomerdquo (2014)

4 Marianne Page Joanne Spetz and Jane Millar ldquoDoesthe Minimum Wage Affect Welfare Caseloadsrdquo Journalof Policy Analysis and Management 24 (2) (2005)273ndash295

5 Marianne Bitler and Hilary Hoynes ldquoThe More ThingsChange the More They Stay the Same The SafetyNet Living Arrangements and Poverty in the GreatRecessionrdquo NBER Working Paper 194 49 2013

6 Sylvia Allegretto and others ldquoFast Food Poverty Wages The Public Cost of Low-Wage Jobs in the Fast-FoodIndustryrdquo (Berkeley California Center for LaborResearch and Education 2013) available at httplaborcenterberkeleyedupubliccostsfast_food_poverty_wage

7 David Neumark and William Wascher ldquoDoes a HigherMinimum Wage Enhance the Effectiveness of theEarned Income Tax Creditrdquo Industrial and LaborRelations Review 64 (4) (2011) 712ndash746

8 David Lee and Emmanuel Saez ldquoOptimal MinimumWage Policy in Competitive Labor Marketsrdquo Journal ofPublic Economics 96 (9) (2012) 739ndash749

9 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family IncomerdquoFebruary 2014

10 Jesse Rothstein ldquoIs the EITC as Good as an NITConditional Cash Transfers and Tax Incidencerdquo AmericanEconomic Journal Economic Policy 2 (1) (2010) 177ndash208

11 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family Incomerdquo

12 Dube Arindrajit 2013 rdquoMinimum Wagesand the Distribution of Family IncomerdquoUnpublished working paper Available at httpsdldropboxusercontentcomu15038936Dube_MinimumWagesFamilyIncomespdf

13 As Dube explains in ldquoThe poverty of Minimum WageFactsrdquo the simulation approach underestimate stemsfrom a number of unwarranted assumptions includingthe range of actual wage increases and the accuracy ofwage data in the Current Population Survey The causal

approach does not make these assumptions

14 Allegretto Sylvia and others 2013 ldquoCredible ResearchDesigns for Minimum Wage Studiesrdquo Working Paper148-13 University of California Berkeley Institutefor Research on Labor and Employment Available athttpwwwirleberkeleyeduworkingpapers148-13pdf

15 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family In comerdquo

16 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo (2012)

17 Ibid

18 Ibid

19 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo

20 US Department of Agriculture ldquoSupplementalNutrition Assistance Program (SNAP)rdquo

21 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo

22 Ibid

23 We deviate from the Census Bureaursquos definition ofa family unit which is ldquotwo people or more (on e ofwhom is the householder) related by bir th marriageor adoption and residing togetherrdquo US Bureau ofthe Census ldquoCurrent Population Survey DefinitionsFamilyrdquo available at httpwwwcensusgovcpsabout

cpsdefhtml (last accessed February 2014) We countas a family unit any individual residing on his or herown two or more persons residing together whodo not belong to a family in the March CPS sampleare constructed as one family in our analysis For thepurposes of food stamp allocations the consumptionresulting from this transfer is probably distributed tofamily members (rather than household members ora single individual within the household) Howeversingle individuals canmdashand domdashreceive SNAPbenefits Excluding them would fail to make theanalysis reflective of the population at large

24 Strictly the family level linear probability modelpredicts the percentage-point decrease in theprobability that an individual family will receive SNAPpayments When applied to a large number of familieshowever we are able to interpret the coefficient asa decrease in the mean of enrollmentmdashthat is a

decrease in the enrollment ratemdashby applying the lawof iterated expectations

25 We generate expenditure predictions from theenrollment modelsmdashand conversely generateenrollment predictions from the expenditure modelmdashby assuming that expenditures per enrolled familyremains the same before and after the minimum wagechange In practice this is likely to be a conservativeestimatemdashthat is to underestimate the decrease inSNAP activity Average SNAP benefits per family willalso decrease as many families that remain eligible forSNAP experience income gains

26 Wage and Hour Division ldquoMinimum Wage Laws inthe States ndash Januar y 1 2014rdquo available at httpwwwdolgovwhdminwageamericahtm (last accessedFebruary 2014)

27 See for example Sylvia Allegretto and others ldquoCredibleResearch Designs for Minimum Wage Studiesrdquo WorkingPaper 148-113 (Berkeley California Institute forResearch on Labor and Employment 2013) available athttpwwwirleberkeleyeduworkingpapers148-13pdf

28 We will report these results in a forthcoming workingpaper

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4243

Endnotes | wwwamericanprogresso

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4343

The Center for American Progress is a nonpartisan research and educational institute

dedicated to promoting a strong just and free America that ensures opportunity

for all We believe that Americans are bound together by a common commitment to

these values and we aspire to ensure that our national policies reflect these values

We work to find progressive and pragmatic solutions to significant domestic and

international problems and develop policy proposals that foster a government that

is ldquoof the people by the people and for the peoplerdquo

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 843

Background | wwwamericanprogress

Background

The federal minimum wage

Te ederal minimum wage was las increased in July o 2009

nearly five years ago During he pas wo decades many saes

have passed legislaion fixing he minimum wage a a higher

level han he ederal minimum Te maps in Figure 1 show ha

while saes in every region o he Unied Saes have adoped

higher minimum wages hey are no disribued randomly bygeography As shown in he 2013 sudy ldquoCredible Research

Designs or Minimum Wage Sudiesrdquo by economiss Sylvia

Allegreto Arindraji Dube Michael Reich and Ben Zipperer

hese saes vary sysemaically rom he oher saes by a number

o characerisics ha affec low-wage employmen rends bu

which are no hemselves relaed o minimum wage policy14

Te nonrandom patern o minimum wage adopion has

imporan implicaions or obaining unbiased esimaes o

minimum wages on employmen In paricular naional panel

sudies ha use sae and ime fixed effec models991252such as

a 1992 sudy by David Neumark and William Wascher991252

spuriously esimae negaive employmen effecs Te reason

or his resul is uncovered using ess or pre-rends Tese

ess find ha low-wage employmen was already declining wo

years beore minimum wages were implemened By making a

saisically large number o local comparisons ha conrol or

heerogeneiy among saes and by ime eliminaes his pre-

rend For his reason we conduc similar ess or our SNAPoucomes and use model specificaions ha include local

comparisons as in he sudy cied above

FIGURE 1

High versus low minimum wage stat

from 1990 to 2012

Means and variances

More than $533

Average minimum wage over 1990ndash2012

Less than or equal to $533

More than $121

Minimum wage variance over 1990ndash2012

Less than or equal to $121

Notes State means and variances were calculated using annual stat

minimum wage data from 1990 to 2012 The shading on the maps

partitions the states into above- and below-median values

Source Sylvia Allegretto and others ldquoCredible Research Designs forMinimum Wage Studiesrdquo Working Paper 148-13 (Institute for Resear

on Labor and Employment 2013) available at httpwwwirleberkeleyeduworkingpapers148-13pdf

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 943

6 Center for American Progress | The Effects of M inim um Wages on SNAP Enro llments and Expenditures

Te CBO recenly projeced ha in 2016 17 million workers will earn less han

he $1010 hourly wage proposed in he Harkin-Miller bill Furhermore he

CBO esimaes ha an addiional 8 million workers earned beween $1010 and

$1150 per hour and were also likely o experience a wage increase15

Supplemental Nutrition Assistance Program

We ocus our inquiry on SNAP Benefis under he program are enirely

ederally unded he program is adminisered by he US Deparmen o

Agriculure ogeher wih sae agencies which share in adminisraion

coss Spending on SNAP has grown in he pas decade reaching $78 bil lion

in 2011 wih SNAP enrollmen increasing o 45 million people abou one-

sevenh o he US populaion16 According o he CBO changes since 1990 in

SNAP spending and enrollmens are primarily he resul o cyclical economic

condiions noably changes in he unemploymen rae and changes in per capiaincome17 Te 2009 American Recovery and Reinvesmen Ac emporarily

increased SNAP benefi amouns by 136 percen as repored by he US

Deparmen o Agriculurersquos Food and Nuriion Service hese higher benefi

levels expired on November 1 201318 Te CBO esimaes ha abou wo-

hirds o he changes in SNAP expendiure are associaed wih changes in he

number o recipiens and one-hird wih changes in he benefis received when

recipiensrsquo incomes change19

In fiscal year 2014 SNAPrsquos maximum monhly benefis are $189 or a single

individual $497 or a amily o hree and $750 or a amily o five Benefis are

reduced by 30 cens per dollar received and phase ou enirely a gross monhly

household incomes o 130 percen o he ederal povery level $1245 or a

single individual $2116 or a amily o hree and $2987 or a amily o five

o deermine benefis SNAP also defines a ne monhly income concep and

ses benefis a 100 percen o he ederal povery level using his concep

Calculaion o ne monhly income can include cerain deducions rom

monhly gross income such as medical expenses and child care coss Alhough

saes are permited some laiude on wha deducions are allowed in pracice

hese vary by very small amouns Our saisical model akes accoun o sae-specific differences in benefis20

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1043

Background | wwwamericanprogress

SNAP imposes an employmen or work-raining requiremen or able-bodied

prime-age aduls991252hose beween he ages o 18 and 50 and wihou disabiliies

or dependen children Such households can receive only hree monhs o benefis

in a hree-year period In recen years abou 85 percen o households receiving

benefis have incomes below he ederal povery level 49 percen have dependen

children 16 percen are age 60 or older 20 percen are disabled and 30 percenrepor some earned income21

A 2012 CBO repor also noes ha ake-up raes among eligible SNAP recipiens

average abou 70 percen wih much lower ake-up among elderly households

Te ake-up rae increases in harder economic imes I also increased when sigma

issues were reduced as SNAP debi cards replaced acual ood samps ake-up

is especially high among hose mos needy Adminisraive spending equaled 91

percen o he poenial spending ha would have occurred i all eligible recipiens

were enrolled Alhough some SNAP policy changes have occurred since 1990

mos were relaively minor and all were naional in scope Te 1996 welare reorm bill eliminaed SNAP eligibiliy or some legal immigrans limied he ime lengh

o eligibiliy or able-bodied childless aduls and reduced maximum benefis

Some o hese resricions were relaxed in 2002 and again in he American

Recovery and Reinvesmen Ac in 200922

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1143

8 Center for American Progress | The Effects of M inim um Wages on SNAP Enro llments and Expenditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1243

Methods and data | wwwamericanprogress

Methods and data

As previously menioned we exploi variaion in minimum wages by sae and ime

o examine heir causal effecs on SNAP enrollmens and expendiures o do so we

merge daa rom 1990 hrough 2012 drawn rom he Annual Social and Economic

Supplemen o he Curren Populaion Survey991252an annual Census Bureau survey

commonly known as he March CPS ha includes23 inormaion on SNAP

enrollmens a he amily level991252wih sae-level daa on minimum wages SNAP

expendiures populaion unemploymen raes and sae median income levels

o conrol or ime-varying heerogeneiy among saes our specificaions includeconrols or sae linear rends and effecs by Census division and ime We esimae

effecs a wo levels allowing or amily variaion and allowing only or sae-level

variaion We also employ a se o sandard demographic conrols such as amily size

and composiion and race and ehnic composiion

Distinguishing causation from correlation

How can we ensure ha our analysis does no pick up a spurious correlaion or

example he endency o more economically vibran saes o implemen higher

minimum wages Disinguishing correlaion and policy endogeneiy rom rue

causal effecs is he primary moivaion or economeric analysis In he ideal

experimen researchers would begin wih wo saes991252ha are alike in every

respec prior o he policy991252 and ldquoreardquo only one o hese saes wih a higher

minimum wage Tey would atemp o shield hese saes rom any influence ha

could obscure heir undersanding o he minimum wagersquos direc effec on SNAP

enrollmen Researchers o course canno conduc such experimens

We can however use saisical mehods o conrol simulaneously heindependen effecs on SNAP o sae unemploymen raes sae income levels

and common rajecories among saes wihin he same Census division By

ensuring similariy along hese dimensions we maximize he likelihood ha

SNAP aciviy in wo saes would have comparable oucomes in he absence

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1343

10 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

o a minimum wage policy change Tus i a new minimum wage policy were

implemened in one sae only he researchers could atribue all o he difference

hey observe in SNAP aciviy o he new minimum wage policy

In oher words we approximae he ideal experimen by using non-experimenal

saisical mehods Te desirable ldquopre-exising similariiesrdquo beween saes ha wehave defined above inorm our choice o conrol variables in a saisical seting

More precisely in our muliple regression models we use median amily income

he unemploymen rae he employmen-o-populaion raio and regional and

ime idenifiers o consruc an appropriae group o peers or each sae on he

eve o a policy change

Data description

wo daa ses include inormaion abou boh income and paricipaion inpublic programs Te Survey o Income and Program Paricipaion or SIPP

which is conduced in inermiten years has he advanage o ollowing he

same individuals over a period o ime In oher words i is a longiudinal daa

se I also has he advanage o conaining monhly daa However he sample

size o he SIPP is no sufficien or analyzing variaions in sae-level minimum

wages Te March CPS has he advanage o a much larger sample size and i is

conduced annually wihou any breaks in ime I has he disadvanage o being

a cross-secional daa se so we canno ollow he same individuals over ime991252

sricly speaking over more han one year On ne he March CPS is much more

suiable or our sudy We examine he empirical relaionship beween minimum

wage policy and ood samp aciviy a wo levels o aggregaion he amily level

and he sae level Family-level daa are drawn rom he March CPS

Te March CPS comprises responses rom he residens o 50000 o 60000

dwelling places surveyed per year and conains deailed inormaion on he

residensrsquo employmen and income including income rom ranser paymens

Te sample or our analysis comprises more han 128 million amily unis during

he period rom 1990 o 2012 (inclusive) Survey weighs allow us o analyze

SNAP paricipaion in a manner ha is represenaive o he US populaion alarge Over all years he share o amilies reporing ood samp receip in he

weighed March CPS sample is 91 percen Te enrollmen rae was a a low o 6

percen in he year 2000 In 2012 he mos recen year in our panel 133 percen

o amilies repored paricipaing in SNAP a some poin during he survey year

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1443

Methods and data | wwwamericanprogresso

Te March CPS also collecs inormaion on he number o SNAP recipiens in

he household he number o monhs he household receives SNAP benefis

and he cash-equivalen value o he SNAP benefis received However he

value o SNAP benefis is severely underrepored among recipiens perhaps

because recipiens are unaware o he exac monhly cash-equivalen value o

benefis hey receive

Our firs empirical sraegy ocuses solely on SNAP enrollmen By using he

amily as he uni o analysis we are able o inser saisical conrols o accoun

or non-wage-relaed acors ha influence any paricular amilyrsquos likelihood o

program paricipaion wih he inenion o isolaing any differences in program

paricipaion ha are due purely o changes in wage policy Tis approach

idenifies he effecs o low-wage labor policy on he exernal margin991252ha is

he effec o he minimum wage on he likelihood ha a amily paricipaes in he

SNAP program a all991252as opposed o he inernal margin or how much SNAP

unding he amily would receive

Our second empirical ramework uses sae-level adminisraive daa Ta is

we aggregae he daa o obain a single daa poin or each saeyear back

o 1990 represening he mean o he oucome or he sae Te sae-level

esimaion serves as a robusness check on he amily-level resuls or SNAP

paricipaion Also using aggregaed daa allows us o esimae direcly he

causal effec o minimum wage changes on SNAP spending Tis is no possible

a he amily level as discussed above daa on cash-equivalen value o ood

samps or SNAP recipiens is very requenly no repored in he March CPS

and when i is repored he inormaion may be unreliable By conras he

Bureau o Economic Analysis publishes aggregae SNAP spending a he sae

level in is Naional Income and Produc Accoun or NIPA ables Tus

while we are unable o observe he heerogeneiy in he cash value o SNAP

or amilies in each sae we are able o calculae average SNAP spending

per residen in each sae per year Supporing covariaes include he annual

unemploymen and employmen daa rom he Bureau o Labor Saisics or

BLS and sae-level populaion series rom he iner-decennial census releases

Minimum wage daa are available rom he BLSrsquos wages and hours division For

sae minimum wage changes enaced a oher imes han he firs o he yearan average value or he year is used

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1543

12 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

Family level model specifications

We firs examine he effec o he minimum wage on paricipaion in public

assisance programs For amily residing in sae and during year we esimae

an equaion o he ollowing orm

(1)

is a binary variable ha is se equal o 1 i a leas one member o amily

received ood samps during he survey year is a se o sae-level

characerisics including annual averages o he unemploymen rae he

employmen-o-populaion raio and he naural log o median amily income

is a vecor o amily atribues including indicaors or he race and marial saus

o he amily head size o he amily he presence o children and he presence oan adul male Sae fixed effecs are capured by o conrol or ime-varying

heerogeneiy our preerred model specificaion also includes year fixed effecs

ha vary by Census division ( ) and sae-level linear ime rends In

Appendix B we jusiy he inclusion o hese las wo erms We also compare he

resuls rom our preerred specificaion wih less sauraed specificaions

Te effec o ineres which is capured by is he expeced change in he

probabiliy o receiving SNAP benefis wih respec o a change in he (log o he)

binding minimum wage in sae during year We repor robus sandard errors

clusered a he sae level We esimae he parameers using linear regression

producing a linear probabiliy model Deails o he model selecion process are

covered in Appendix B below

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1643

Methods and data | wwwamericanprogresso

State-level model specifications

Te sae-level models are similarly specified For sae in year we assume ha

(2)

In his model is now eiher he SNAP enrollmen rae in sae during year

or he naural logarihm o per capia SNAP expendiures in sae during

year is once again a se o sae-level characerisics including he same

sae-level covariaes as in he amily regressions (annual average unemploymen

rae employmen-o-populaion raio naural log o median amily income)

wih he addiion o amily level characerisics averaged across he sae (average

amily size and he shares o populaion consiued by each o five racialehnicgroups) Sae fixed effecs are represened by As above our preerred model

specificaion includes year fixed effecs ha vary by Census division ( ) and

sae-level linear ime rends as elaboraed in Appendix B Te effec o

ineres is capured by

We esimae boh sae-level models (enrollmen and expendiures) using

ordinary leas squares regression Tus he inerpreaion o he coefficien is no

longer ha o a change in probabiliy as in he binary oucome models described

above Raher or he sae-level SNAP enrollmen model represens he

expeced change (in percenage poins) in he saersquos SNAP enrollmen rae ha

is due o a 1 percen change in he minimum wage For he SNAP expendiures

model is simply he elasiciy o SNAP spending wih respec o he minimum

wage991252ha is he percenage change in sae expendiures expeced o resul rom

a 1 percen change in ha saersquos minimum wage For urher deails on model

specificaion reer o Appendix B below

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1743

14 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1843

Results | wwwamericanprogresso

Results

Estimated minimum wage effects on SNAP enrollment and

expenditures

able 1 shows he esimaed parameer o ineres991252he coefficien o he

minimum wage or he preerred model o each ype Coefficiens on he

minimum wage variable are no direcly comparable across models because all

our models have a differen uncional orm o undersand and compare hese

esimaes we compue he change in SNAP aciviy prediced or a paricular wage scenario Te final column in able 1 answers he quesion Wha would be

he expeced change in SNAP aciviy in response o a 10 percen increase in he

minimum wage Te answer o his quesion varies wih he value o he inpu

parameers in he able we calculae he percenage decrease in enrollmen or

expendiures prediced or he average sae wih a minimum wage o $725 in

2014 Te sae-level SNAP expendiure model which is a consan-elasiciy

model conveys elasiciy inormaion direcly or he change in expendiures per

capia in he sae

TABLE 1

Comparison of national SNAP predictions for a 10 percent increase in the federal minimum wage

Model LevelRegression

type

Predicted outcome Coefficient of log

(minimum wage)

(Standard error)

Effect of a 10 percent in

in the minimum wage

VariableForm of

variable

Total

enrollment

Tot

expend

1 Family Linear

probability Enrollment Binary (enrolled=1)

-0042

(0008) -317 N

2

StateLinear regression

(ordinary least

squares)

Enrollment State enrollment rate (percent) -0031

(0012) -235 N

3 Expenditures Log (state expenditures per capita) -0190

(0103) NA -19

plt01 plt005 plt001Note Predicted changes are calculated for the average state with a minimum wage of $725 in 2014

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1943

16 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

According o his model a 10 percen increase in he minimum wage would resul

in a 19 percen reducion in SNAP expendiures Tis resul is reassuringly similar

o Dubersquos resul or povery reducion Te wo enrollmen models are much more

precisely esimaed han he expendiure model Te sae-level enrollmen model

finds ha a 10 percen minimum wage increase in a low-wage sae is associaed wih

a 235 percen decrease in SNAP enrollmens Te amily-level linear probabiliymodel predics a somewha greaer elasiciy or low minimum wage saes an

increase o 10 percen in he ederal minimum would resul in a 317 percen decline

in SNAP enrollmen24 Te differences in hese esimaes sem rom a number o

acors including difference in model uncional orm and daa used We rea his

range o elasiciy esimaes as an upper and lower bound on enrollmen impacs

Harkin-Miller bill National and state-level predicted impacts

Wha would be he prediced change or he SNAP program i he ederalminimum were raised o $1010 as proposed in he Harkin-Miller bill In order o

make his inerence we accoun or he ac ha no all saes are currenly subjec

o he ederal minimum wage a he beginning o 2014 21 saes mainained

higher minimum wages han $725 In hose saes an increase in he ederal

minimum wage may or may no be binding or employers in he sae depending

upon wheher he new ederal minimum exceeds he sae-level minimum Bu

regardless o wheher a minimum wage change is binding he impac on SNAP

aciviy will be lower in high minimum wage saes In order o accoun or his

properly we calculae sae by sae he percenage wage change ha would resul

rom he Harkin-Miller proposal and apply he parameers rom each o he hree

models above o compue he expeced decrease in SNAP aciviy or each sae

In his exercise we use saesrsquo curren (2014) minimum wage levels and assume

as a baseline he 2012 levels o SNAP enrollmen and expendiure as 2012 is he

mos recen year or which SNAP daa are available

able 10 and able 11 in Appendix C repor he esimaed effecs on SNAP

enrollmen and expendiures respecively or each sae under he Harkin-Miller

bill25 An increase o $1010 i enaced oday would represen beween a 393

percen wage increase in a $725 minimum wage sae and an 84 percen increasein Washingon sae which has he highes minimum wage in he naion a $932

as o January 201426 Slighly more han 56 percen o he decrease in expendiures

and abou 59 percen o he decrease in enrollmen would occur in saes wih

presen-day minimum wages o $725

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2043

Results | wwwamericanprogresso

In 2012 hese saes were home o 46 percen o he American populaion

and accouned or a roughly equivalen percenage o oal naional SNAP

expendiures No surprisingly he larges enrollmen decreases come rom

populous saes wih very high SNAP enrollmen raes andor very low minimum

wages Te larges prediced enrollmen reducion991252beween 319000 individuals

and 362000 individuals991252would occur in exas which has a $725 minimum wage In Caliornia which has a 206 percen SNAP paricipaion rae and an $8

per hour minimum wage we anicipae SNAP enrollmen declines o 310000

persons o 371000 persons And in Florida which had a SNAP paricipaion rae

o 166 percen and a $793 minimum wage enrollmen could decline by beween

164000 individuals and 196000 individuals For he our saes ha ogeher

accouned or he greaes amoun o SNAP spending in 2012991252exas Caliornia

Florida and New York respecively991252he combined expendiure reducion rom

he Harkin-Miller bill is prediced o be $14 billion

able 2 summarizes he prediced declines in SNAP aciviy or he naion as a whole ha would resul rom he direc and indirec effecs o he Harkin-Miller

bill Enrollmen would all beween 31 million persons and 36 million persons

represening 75 percen o 87 percen o curren enrollmen Te anicipaed

reducion in program expendiures would be nearly $46 billion or 61 percen o

program expendiures

TABLE 2

Comparison of national SNAP predictions under the Harkin-Miller billrsquos $1010 minimum wage

Model

Enrollment

(persons)

Expenditures

(millions of dollars)

Current (2012) Predicted Change Current (2012) Predicted Chan

Family enrollment (linear probability)

41866195

45489339 -3623144

$74861

NA NA

State enrollment (ordinary least squares) 38745435 -3120759 NA NA

State expenditures (ordinary least squares) NA NA $70305 -$45

Note Calculations use 2014 state minimum wages and the most recent SNAP data from 2012 They assume that per-enrollee expenditures remain constant

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2143

18 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

Tere are o course oher possibiliies or minimum wage legislaion able 3

shows he expeced SNAP changes or he Unied Saes under a variey o wage

scenarios calculaed using he sae-level models I saes were no able o se

heir minimum wages independenly such ha all saes were consrained by

he ederal minimum o $725 SNAP would be received by abou 514000 more

people across he Unied Saes a an addiional program cos o nearly hree-quarers o a billion dollars In conras he effecs o a higher minimum wage

proposal991252a ederal wage floor o $11 per hour991252would decrease enrollmen in

SNAP by more han 10 percen and decrease program coss by 83 percen

TABLE 3

Summary of par ticipation and expenditures under wage scenarios

If all states had

minimum wages of

Enrollment(persons)

Expenditures

(millions of dollars)

Predicted Change Predicted Change

Recent levels (2014) 41866195 $74861

$725 42380520 514326 $75604 $743

$800 41423919 -442276 $74209 -$652

$900 40148451 -1717744 $72350 -$2511

$1000 38872982 -2993212 $70490 -$4371

$1010 38745435 -3120759 $70305 -$4556

$1100 37597514 -4268681 $68631 -$6230

Note Calculations use state-level enrollment model coefficient

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2243

Conclusion | wwwamericanprogresso

Conclusion

An exensive body o lieraure examines employmen effecs o he minimum

wage A much smaller se o sudies examines how he minimum wage affecs

povery and only a handul o papers examine he effecs o he minimum wage

on he EIC Our analysis is he firs o examine he effecs o he minimum

wage on SNAP

Our findings indicae ha increased earnings rom minimum wage changes do

reduce SNAP enrollmens and expendiures We esimae ha he Harkin-Miller bill would save axpayers nearly $46 billion per year equivalen o 61 percen

o SNAP expendiures in 2012 he las year or which daa are available Over a

10-year period he esimaed savings amoun o nearly $46 billion

Our repor is subjec o limiaions ha we expec o overcome in our uure

research Firs he findings do no ake ino accoun possible ineracions among

SNAP he EIC and Medicaid Te eligibiliy cuoffs among hese programs

are quie differen suggesing ha such ineracions may be minor Noneheless

he join effecs can only be deermined by urher research using a causal

model Second i would be useul o know he disribuion o SNAP reducions

along he wage disribuion Using he Congressional Budge Officersquos calculaions

o how much he oal dollar value o wage would increase under he Harkin-

Miller proposal our findings imply ha he decline in overall SNAP spending

equals abou 15 percen o he oal resuling increase in wages Te amoun and

disribuion o his offse are o considerable ineres Minimum wage beneficiaries

who come rom working amilies already well above he povery line would no

see any offse while hose who are currenly considerably below he povery line

will see larger offses Tese issues will also be a subjec or our uure research

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2343

20 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

About the authors

Rachel West is a maser o public policy candidae a he Goldman School

o Public Policy Universiy o Caliornia Berkeley Her research ocuses on

economic policy in he areas o low-wage labor and povery

Michael Reich is proessor o economics and direcor o he Insiue or

Research on Labor and Employmen a he Universiy o Caliornia a Berkeley

His research publicaions cover numerous areas o labor economics including

racial inequaliy labor marke segmenaion high-perormance workplaces

union-managemen cooperaion Japanese labor-managemen sysems living

wages and minimum wages He received his docorae in economics rom

Harvard Universiy

Acknowledgments

We are graeul o Sylvia Allegreto Arindraji Dube Bill Leser Jesse Rohsein

Daniel Tompson and Ben Zipperer or heir valuable suggesions

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2443

References | wwwamericanprogresso

References

Allegreto Sylvia and ohers 2013 ldquoFas Food Povery Wages Te Public Cos o Low-Wage Jobsin he Fas-Food Indusryrdquo Universiy o Caliornia Berkeley Cener or Labor Research andEducaion Available a htplaborcenerberkeleyedupubliccossas_ood_povery_wage

Allegreto Sylvia and ohers 2013 ldquoCredible Research Designs or Minimum Wage Sudiesrdquo

Working Paper 148-13 Universiy o Caliornia Berkeley Insiue or Research on Labor andEmploymen Available a htpwwwirleberkeleyeduworkingpapers148-13pd

Biler Marianne and Hilary Hoynes 2013 ldquo Te More Tings Change he More Tey Say heSame Te Saey Ne Living Arrangemens and Povery in he Grea Recessionrdquo Working Paper19449 Naional Bureau o Economic Research

Congressional Budge Office 2012 ldquoTe Supplemenal Nuriion Assisance Programrdquo Washingon Available a htpwwwcbogovsiesdeaulfilescbofilesatachmens04-19-SNAPpd

991252 991252 991252 2014 ldquoTe Effec o a Minimum-Wage Increase on Employmen and Family Incomerdquo Washingon Available a htpwwwcbogovsiesdeaulfilescbofilesatachmens44995-MinimumWagepd

Dube Arindraji 2013 rdquoMinimum Wages and he Disribuion o Family Incomerdquo Unpublished working paper Available a htpsdldropboxuserconencomu15038936Dube_ MinimumWagesFamilyIncomespd

991252 991252 991252 2014 ldquoTe Povery o Minimum Wage Facsrdquo Arindraji Dube Blog January 22 Availablea htparindubecom20140122he-povery-o-minimum-wage-acs

Lee David and Emmanuel Saez 2012 ldquoOpimal Minimum Wage Policy in Compeiive LaborMarkesrdquo Journal o Public Economics 96 (9) 739ndash749

Neumark David and William Wascher 1992 ldquoEmploymen Effecs o Minimum and Subminimum Wages Panel Daa on Sae Minimum Wage Lawsrdquo Industrial and Labor Relations Review 46 (1)

55ndash81

Neumark David and William Wascher 2011 ldquoDoes a Higher Minimum Wage Enhance heEffeciveness o he Earned Income ax Credirdquo Industrial and Labor Relations Review 64 (4)712ndash746

Page Marianne Joanne Spez and Jane Millar 2005 ldquoDoes he Minimum Wage Affec WelareCaseloadsrdquo Journal o Policy Analysis and Management 24 (2) 273ndash295

Rohsein Jesse 2010 ldquoIs he EIC as Good as an NI Condiional Cash ransers and ax

Incidencerdquo American Economic Journal Economic Policy 2 (1) 177ndash208

Wage and Hours Division 2014 ldquoMinimum Wage Laws in he Saes ndash January 1 2014rdquo USDeparmen o Labor (htpwwwdolgovwhdminwageamericahm [February 2014])

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2543

22 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2643

Appendix A | wwwamericanprogresso

Appendix A

Pre-trend falsification check

Recen minimum wage research27 highlighs a common flaw in previous sudies

ailure o veriy ha he oucome variable is ree o negaive pre-exising rends I

or example SNAP aciviy was already rending down in saes ha raised heir

minimum wages beore hese changes came ino effec our regression analysis could

(misakenly) atribue ha reducion o he minimum wage We check or such

pre-rends by inroducing variables ha represen he prior yearrsquos value or leads o

he minimum wage I he model esimaes he minimum wage o have an effec on

he oucome variable beore he wage change wen ino effec hen an unobservedacor no he minimum wage change caused he change in SNAP aciviy

We es he specificaions above or pre-rends by including a one-year lead in

all hree specificaions We find ha he lead erms are small posiive and no

saisically significan indicaing ha he concurren minimum wage991252no

he wage level in prior periods991252is driving he observed changes in SNAP

oucomes28 In paricular he coefficien (sandard error) on he lead erm in

our preerred amily-level enrollmen regression is 011 and no significan

while he coefficien and sandard error o he conemporaneous minimum

wage is unchanged In he sae-level preerred enrollmen regression he

coefficien o he lead erm is again small (07) and i is no significan Te

corresponding coefficien on he lead erm in he sae-level expendiure

regression is 16 and is no significan Te posiive poin esimaes on hese lead

erms resuls no only rule ou disoring negaive pre-rends Tey also sugges

ha our main resuls may underesimae he rue effecs

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2743

24 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2843

Appendix B | wwwamericanprogresso

Appendix B

Model selection process

For boh he amily-level and sae-level models we es hree mehods o conrol

or unobserved geographic- and ime-varying characerisics as suggesed by he

minimum wage lieraure Firs we include only independen sae-specific fixed

effecs and year-specific fixed effecs Tis specificaion (specificaion 1) implicily

assumes ha amilies in any sae consiue an equally good saisical ldquoconrolrdquo

group or hose in any randomly chosen sae afer accouning or various

characerisics (median income and unemploymen rae among ohers) Similarly

simple ime fixed effecs assume ha amilies surveyed in any year can crediblyserve as a conrol group or amilies surveyed in every oher year o he sample

(1990 hrough 2012)

In oher words specificaion 1 assumes ha a saersquos immediae neighbor provides

no beter a couneracual or he effec o a minimum wage change han does a

sae across he counry We relax his resricive specificaion sequenially in wo

seps In specificaion 2 we replace simple year fixed effecs wih fixed effecs or

each Census divisionyear (capured as an addiional variable in he vecor By

using division-year effecs we remove he resricion ha amilies in each sae

are equally good saisical conrols or all oher amilies Raher we allow or he

possibiliy ha amilies in similar geographic regions (or example he Souh or

he Norheas) may be more similar o one anoher han amilies arher away

Finally in specificaion 3 we add sae-specific linear ime rends o he previous

specificaion Tus specificaion 3 is he mos rigorous model specificaion in ha

i allows or heerogeneiy along hree dimensions Ta is specificaion 3 allows

each sae o have is own ime-varying rends raher han imposing he resricion

ha saes evolve idenically over he 22 years in he sample

We begin building he heoreical specificaion above rom a se o simpleuncondiional models regression o SNAP aciviy (enrollmen or expendiures)

on he log o he minimum wage and a se o geographic- and ime-specific

effecs (specificaions 1 2 and 3 described above) As shown in ables 1ndash3 (or

specificaion 3) we hen add covariaes sequenially o hese models including

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2943

26 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

firs he vecor o amily-level conrols ollowed by each o several sae-level

covariaes in urn he unemploymen rae log o median-amily income and he

employmen-o-populaion raio Comparable resuls or specificaions 2 and 3

will be available in our orhcoming working paper

As expeced he simple uncondiional models indicae ha he relaionship beween he minimum wage and SNAP enrollmen i one exiss is a more complex

one influenced by oher acors In he uncondiional model he coefficien on he

variable o ineres991252he log o he minimum wage991252is small in magniude and no

saisically differen rom zero Once we accoun or he influence o labor marke

condiions and variaion in income levels on program paricipaion (by including

unemploymen rae and median-amily income conrol variables respecively)

he effec o he minimum wage on SNAP enrollmen is precisely esimaed Te

coefficien o he log minimum wage is slighly higher (-0042) in he amily-level

analysis han he coefficien (-031) in he sae-level analysis Te level o precision

is also higher in he amily-level analysis Tis is o be expeced when using 124million observaions compared o 1127

Te second dimension o model choice concerns he effec specificaion ables

7ndash9 compare he primary coefficiens o ineres or he SNAP enrollmen and

expendiure models For boh he enrollmen models he effec sizes are smalles

or specificaion 1 larges or specificaion 2 and inermediae beween hese wo

in specificaion 3 Recall ha Specificaion 3 conains sae-specific linear ime

rends in addiion o he census divisionyear conrols included in specificaion

2 In he amily-level enrollmen model he sandard error o he minimum wage

coefficien is smaller han in he oher wo specificaions Sandard errors on he

oher variables are much smaller in specificaions 2 and 3 han in specificaion

1 On he basis o coefficien significance (join and individual) specificaions 2

and 3 are sricly preerred in boh enrollmen models o specificaion 1 which

conains only sae and year fixed effecs

A concern wih specificaions 2 and 3 is ha rend conrols such as sae linear

rends may incorrecly absorb some o he delayed impac o a minimum wage

When we es his issue by including lagged minimum wages we do no find ha

delayed effecs are significan Anoher concern is ha more sauraed modelsuse less o he saisical variaion which could reduce he saisical power o

he resuls However he sandard errors or our more sauraed models are no

higher and are lower in some cases han or he less sauraed models Overall

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3043

Appendix B | wwwamericanprogresso

his evidence suppors our use o he sauraed model as he preerred model

specificaion Moreover Dubersquos 2013 sudy shows ha more sauraed models

perorm beter han models wih jus sae and ime fixed effecs

Te esimaed enrollmen regressions a boh he amily and sae levels show large

and saisically significan coefficiens Te esimaed minimum wage effec in heexpendiures regressions991252or which we have only sae-level daa991252is also large

and saisically significan

We do no use weighed regression or he sae-level models preerring o keep

analysis o he ldquoreamenrdquo (ha is o say a minimum wage change) appropriae

o he average sae raher han he average amily or individual I insead our

primary ineres were he impac o a minimum wage change on he average amily

or he average individual we migh choose o designae he number o amilies

in each sae or he sae populaion respecively as analyic weighs in order o

obain a coefficien beter suied or such inerence

TABLE 4

SNAP enrollment

Family-level linear probability

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0048 -0047 -0040 -0043 -0042

(0013) (0013) (001) (0008) (0009)

Unemployment rate (100) 0505 0420 0280

(0083) (0086) (0082)

Log median income -0057 -0039

(0011) (0011)

Employment-to-population ratio -0239

(0038)

N 1242022 1242022 1242022 1242022 1242022

plt01 plt005 plt001

Note Robust standard errors are in parentheses Observations are clustered at the state level The outcome variable is binary and equal to one if a family is enrolledin SNAP All models include state fixed effects Census division x-year fixed effects and state-specific linear time trends All specifications except 3a include additional

controls for family size race and marital status of the family head presence of children and presence of an adult maleSource Annual data from the March Current Population Survey for 1990 to 2012 estimation includes Current Population Survey probability weights

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3143

28 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

TABLE 5

SNAP enrollment

State-level linear regression

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0028 -0024 -0026 -0031 -0031

(0014) (0013) (0013) (0012) (0012)

Unemployment rate (100) 0625 0466 0320 0339

(0087) (0088) (0085) (0083)

Log median income -0090 -0065 -0061

(0013) (0013) (0013)

Employment-to-population ratio -0282 -0248

(0037) (0038)

N 1127 1127 1127 1127 1127

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the SNAP enrollment rate All models include state fixed effects Census division x-year fixedeffects and state-specific linear time trends

Source Annual data from the March Current Population Survey for 1990 to 2012

TABLE 6

SNAP expenditures

State-level linear regression

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0181 -0149 -0156 -0153 -0190

(011) (0103) (0102) (0103) (0103)

Unemployment rate (100) 4587 4152 4232 4313

(0622) (0621) (0633) (0628)

Log median income -0246 -0261 -0294

(0075) (0078) (0078)

Employment-to-population ratio 0155 0244

(0237) (024)

N 1127 1127 1127 1127 1127

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the natural log of state SNAP expenditures per capita for 1990 to 2012 All models include state

fixed effects Census division x-year fixed effects and state-specific linear time trends

Source Annual data from Bureau of Economic Analysis national income and product accounts tables for 1990 to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3243

Appendix B | wwwamericanprogresso

TABLE 7

Comparison of specifications SNAP enrollment

Family-level linear probability

(1) (2) (3)

Log minimum wage -0028 -0049 -0042

(0014) (0017) (0009)

Unemployment rate (100) 0275 0297 0280

(0161) (0076) (0082)

Log median income -0077 -0055 -0039

(0014) (0012) (0011)

Employment-to-population ratio -0238 -0250 -0239

(0054) (004) (0038)

N 1242022 1242022 1242022

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses Observations are clustered at the state level The outcomevariable is binary or equal to one if a family is enrolled in SNAP All specifications include additional controls forfamily size race and marital status of the family head presence of children and presence of an adult male

Source Annual data from the March Current Population Survey for 1990 to 2012 estimation includes CurrentPopulation Survey probability weights

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3343

30 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

TABLE 8

Comparison of specifications SNAP enrollment

State-level linear regression

(1) (2) (3)

Log minimum wage -0019 -0035 -0031

(0009) (0012) (0012)

Unemployment rate (100) 0401 0370 0339

(0063) (0077) (0083)

Log median income -0081 -0073 -0061

(0011) (0013) (0013)

Employment-to-population ratio -0183 -0222 -0248

(0039) (0039) (0038)

N 1127 1127 1127

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the SNAP enrollment rate Allregressions include the share of households with children and the racial shares of the population

Source Annual data from the March Current Population Survey for 1990 to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3443

Appendix B | wwwamericanprogresso

TABLE 9

Comparison of specifications SNAP expenditures

State-level linear regression

(1) (2) (3)

Log minimum wage -0121 -0203 -0190

(0075) (0103) (0103)

Unemployment rate (100) 5292 5152 4313

(0464) (0576) (0628)

Log median income -0437 -0417 -0294

(008) (0086) (0078)

Employment-to-population ratio -0040 -0220 0244

(0261) (0260) (0240)

N 1127 1127 1127

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the natural log of state SNAPexpenditures per capita All regressions include the share of households with children and the racial shares ofthe state population

Source Annual data from Bureau of Economic Analysis national income and product accounts tables for 1990to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3543

32 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3643

Appendix C | wwwamericanprogresso

Appendix C Harkin-Miller

policy simulation results

StateMinimum wage

(2014)

SNAP enrollment (2012) Predicted change in SNAP enrollment

RateTotal

(persons)

Model 1 family

enrollment (linearprobability)

Model 2 state

enrollment (linearregression)

Model 3 s

expenditu(linear regre

Alabama $725 164 788682 -66922 -58897 -58906

Alaska $775 120 87436 -8104 -6887 -3288

Arizona $790 201 1319323 -67435 -56738 -64356

Arkansas $725 78 230489 -40977 -36063 -35248

California $800 206 7813680 -371131 -310222 -18223

Colorado $800 164 853155 -50684 -42365 -23926

Connecticut $870 91 326621 -22456 -17975 -13711

Delaware $725 186 170262 -12739 -11211 -10647

District of Columbia $825 133 84009 -5370 -4417 -3632

Florida $793 166 3208026 -195813 -164426 -13046

Georgia $725 160 1586336 -137741 -121224 -11004

Hawaii $725 96 133662 -19310 -16995 -14933

Idaho $725 92 147501 -22165 -19507 -15809

Illinois $825 95 1225084 -109088 -89742 -70955

Indiana $725 125 816233 -90818 -79928 -83985

Iowa $725 155 478011 -42716 -37594 -28556

Kansas $725 135 388269 -40082 -35275 -27461

Kentucky $725 130 568821 -60840 -53544 -52259

Louisiana $725 149 683832 -63929 -56263 -66083

Maine $750 77 101976 -16567 -14323 -15234

Maryland $725 144 846415 -81748 -71946 -38370

Massachusetts $800 130 864721 -64902 -54251 -42913

Michigan $740 146 1439141 -128801 -112140 -11022

Minnesota $725 133 713646 -74730 -65769 -37878

TABLE 10

SNAP enrollments Predicted changes by state under the Harkin-Miller bill

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3743

34 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

StateMinimum wage

(2014)

SNAP enrollment (2012) Predicted change in SNAP enrollment

RateTotal

(persons)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 state

expenditures

(linear regression)

Mississippi $725 129 386501 -41486 -36511 -46467

Missouri $750 172 1036182 -75131 -64952 -56244

Montana $790 132 132452 -10350 -8708 -5846

Nebraska $725 124 230683 -25773 -22683 -12189

Nevada $825 162 446035 -23349 -19209 -11894

New Hampshire $725 127 168404 -18359 -16157 -5735

New Jersey $825 160 1416666 -75175 -61843 -28236

New Mexico $750 149 310896 -25983 -22463 -22512

New York $800 192 3763553 -191193 -159815 -142182

North Carolina $725 174 1697193 -135417 -119179 -113503

North Dakota $725 87 61225 -9743 -8574 -4021

Ohio $795 143 1647345 -115869 -97169 -88580

Oklahoma $725 129 494053 -53006 -46650 -46854

Oregon $910 124 485326 -17036 -13328 -16398

Pennsylvania $725 161 2053643 -177315 -156052 -125586

Rhode Island $800 156 163730 -10258 -8574 -8698

South Carolina $725 94 445277 -65614 -57746 -50304

South Dakota $725 208 173749 -11586 -10197 -7458

Tennessee $725 142 914903 -89667 -78915 -99134

Texas $725 110 2863779 -362018 -318607 -253285

Utah $725 88 251107 -39658 -34902 -19390

Vermont $873 156 97792 -3823 -3055 -2475

Virginia $725 101 829771 -113723 -100086 -58212

Washington $932 72 496934 -23221 -17947 -17756

West Virginia $725 58 107875 -25792 -22699 -21665

Wisconsin $725 75 427822 -79521 -69986 -53210

Wyoming $725 164 94590 -8010 -7050 -3104

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3843

Appendix C | wwwamericanprogresso

StateMinimum wage

(2014)

SNAP expenditures (2012)Predicted change in SNAP expenditures

(millions of dollars)

Per resident

(millions of

dollars)

Total

(millions of

dollars)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 s

expenditu

(linear regre

Alabama $725 $290 $1397 -$1185 -$1043 -$1043

Alaska $775 $253 $185 -$262 -$223 -$106

Arizona $790 $257 $1686 -$935 -$787 -$892

Arkansas $725 $248 $733 -$636 -$560 -$547

California $800 $189 $7164 -$7276 -$6082 -$3573

Colorado $800 $157 $816 -$862 -$721 -$407

Connecticut $870 $191 $686 -$343 -$275 -$210

Delaware $725 $250 $229 -$205 -$180 -$171

District of Columbia $825 $366 $232 -$146 -$120 -$99

Florida $793 $294 $5676 -$4429 -$3719 -$2951

Georgia $725 $317 $3140 -$2936 -$2584 -$2346

Hawaii $725 $335 $465 -$449 -$395 -$347

Idaho $725 $225 $359 -$376 -$331 -$268

Illinois $825 $249 $3200 -$2096 -$1725 -$1364

Indiana $725 $220 $1439 -$1162 -$1023 -$1075

Iowa $725 $192 $589 -$658 -$579 -$440

Kansas $725 $159 $460 -$502 -$441 -$344

Kentucky $725 $298 $1303 -$1133 -$997 -$973

Louisiana $725 $315 $1450 -$1047 -$922 -$1083

Maine $750 $281 $373 -$267 -$231 -$246

Maryland $725 $188 $1109 -$1765 -$1553 -$828

Massachusetts $800 $206 $1366 -$1030 -$861 -$681

Michigan $740 $300 $2963 -$2400 -$2090 -$2054

Minnesota $725 $140 $755 -$1113 -$980 -$564

Mississippi $725 $326 $973 -$649 -$571 -$726

Missouri $750 $241 $1452 -$1278 -$1104 -$956

Montana $790 $190 $191 -$179 -$151 -$101

Nebraska $725 $140 $259 -$409 -$360 -$194

Nevada $825 $191 $527 -$441 -$363 -$225

New Hampshire $725 $126 $167 -$399 -$351 -$125

TABLE 11

SNAP expenditures Predicted changes by state under the Harkin-Miller bill

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3943

36 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

StateMinimum wage

(2014)

SNAP expenditures (2012)Predicted change in SNAP expenditures

(millions of dollars)

Per resident

(millions of

dollars)

Total

(millions of

dollars)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 state

expenditures

(linear regression)

New Jersey $825 $160 $1420 -$1611 -$1325 -$605

New Mexico $750 $324 $675 -$513 -$444 -$445New York $800 $287 $5616 -$3766 -$3148 -$2801

North Carolina $725 $252 $2454 -$2187 -$1925 -$1833

North Dakota $725 $128 $90 -$162 -$143 -$67

Ohio $795 $259 $2995 -$2013 -$1688 -$1539

Oklahoma $725 $248 $945 -$799 -$703 -$706

Oregon $910 $322 $1255 -$272 -$213 -$262

Pennsylvania $725 $218 $2779 -$2930 -$2579 -$2075

Rhode Island $800 $280 $294 -$173 -$144 -$147

South Carolina $725 $291 $1373 -$1337 -$1177 -$1025South Dakota $725 $198 $165 -$192 -$169 -$123

Tennessee $725 $324 $2091 -$1413 -$1243 -$1562

Texas $725 $230 $5997 -$6402 -$5634 -$4479

Utah $725 $141 $402 -$614 -$541 -$300

Vermont $873 $230 $144 -$66 -$53 -$43

Virginia $725 $173 $1413 -$2062 -$1815 -$1056

Washington $932 $244 $1682 -$350 -$270 -$267

West Virginia $725 $273 $508 -$451 -$397 -$379

Wisconsin $725 $204 $1166 -$1302 -$1146 -$871Wyoming $725 $95 $55 -$105 -$93 -$41

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4043

Appendix C | wwwamericanprogresso

FIGURE 2

Raising the minimum wage to $1010would cut taxpayer costs in every state

Predicted decreases in cost and enrollment

in SNAP in 50 states

$200+$51ndash$100

$101ndash$200

0ndash$25

$26ndash$50

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4143

38 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

Endnotes

1 US Department of Agriculture ldquoSupplementalNutrition Assistance Program (SNAP)rdquo available athttpwwwfnsusdagovsnapeligibility (last accessedFebruary 2014)

2 For this initial analysis we do not consider Harkin-Millerrsquos increase in subminimum wages for tippedworkers To do s o would increase the estimated SNAP

savings by an unknown amount

3 The Congressional Budget Office estimates thatworkers currently earning between $1010 and $1150per hour would see their wages rise under the Harkin-Miller proposal Congressional Budget O ffice ldquoTheEffects of a Minimum Wage Increase on Employmentand Family Incomerdquo (2014)

4 Marianne Page Joanne Spetz and Jane Millar ldquoDoesthe Minimum Wage Affect Welfare Caseloadsrdquo Journalof Policy Analysis and Management 24 (2) (2005)273ndash295

5 Marianne Bitler and Hilary Hoynes ldquoThe More ThingsChange the More They Stay the Same The SafetyNet Living Arrangements and Poverty in the GreatRecessionrdquo NBER Working Paper 194 49 2013

6 Sylvia Allegretto and others ldquoFast Food Poverty Wages The Public Cost of Low-Wage Jobs in the Fast-FoodIndustryrdquo (Berkeley California Center for LaborResearch and Education 2013) available at httplaborcenterberkeleyedupubliccostsfast_food_poverty_wage

7 David Neumark and William Wascher ldquoDoes a HigherMinimum Wage Enhance the Effectiveness of theEarned Income Tax Creditrdquo Industrial and LaborRelations Review 64 (4) (2011) 712ndash746

8 David Lee and Emmanuel Saez ldquoOptimal MinimumWage Policy in Competitive Labor Marketsrdquo Journal ofPublic Economics 96 (9) (2012) 739ndash749

9 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family IncomerdquoFebruary 2014

10 Jesse Rothstein ldquoIs the EITC as Good as an NITConditional Cash Transfers and Tax Incidencerdquo AmericanEconomic Journal Economic Policy 2 (1) (2010) 177ndash208

11 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family Incomerdquo

12 Dube Arindrajit 2013 rdquoMinimum Wagesand the Distribution of Family IncomerdquoUnpublished working paper Available at httpsdldropboxusercontentcomu15038936Dube_MinimumWagesFamilyIncomespdf

13 As Dube explains in ldquoThe poverty of Minimum WageFactsrdquo the simulation approach underestimate stemsfrom a number of unwarranted assumptions includingthe range of actual wage increases and the accuracy ofwage data in the Current Population Survey The causal

approach does not make these assumptions

14 Allegretto Sylvia and others 2013 ldquoCredible ResearchDesigns for Minimum Wage Studiesrdquo Working Paper148-13 University of California Berkeley Institutefor Research on Labor and Employment Available athttpwwwirleberkeleyeduworkingpapers148-13pdf

15 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family In comerdquo

16 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo (2012)

17 Ibid

18 Ibid

19 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo

20 US Department of Agriculture ldquoSupplementalNutrition Assistance Program (SNAP)rdquo

21 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo

22 Ibid

23 We deviate from the Census Bureaursquos definition ofa family unit which is ldquotwo people or more (on e ofwhom is the householder) related by bir th marriageor adoption and residing togetherrdquo US Bureau ofthe Census ldquoCurrent Population Survey DefinitionsFamilyrdquo available at httpwwwcensusgovcpsabout

cpsdefhtml (last accessed February 2014) We countas a family unit any individual residing on his or herown two or more persons residing together whodo not belong to a family in the March CPS sampleare constructed as one family in our analysis For thepurposes of food stamp allocations the consumptionresulting from this transfer is probably distributed tofamily members (rather than household members ora single individual within the household) Howeversingle individuals canmdashand domdashreceive SNAPbenefits Excluding them would fail to make theanalysis reflective of the population at large

24 Strictly the family level linear probability modelpredicts the percentage-point decrease in theprobability that an individual family will receive SNAPpayments When applied to a large number of familieshowever we are able to interpret the coefficient asa decrease in the mean of enrollmentmdashthat is a

decrease in the enrollment ratemdashby applying the lawof iterated expectations

25 We generate expenditure predictions from theenrollment modelsmdashand conversely generateenrollment predictions from the expenditure modelmdashby assuming that expenditures per enrolled familyremains the same before and after the minimum wagechange In practice this is likely to be a conservativeestimatemdashthat is to underestimate the decrease inSNAP activity Average SNAP benefits per family willalso decrease as many families that remain eligible forSNAP experience income gains

26 Wage and Hour Division ldquoMinimum Wage Laws inthe States ndash Januar y 1 2014rdquo available at httpwwwdolgovwhdminwageamericahtm (last accessedFebruary 2014)

27 See for example Sylvia Allegretto and others ldquoCredibleResearch Designs for Minimum Wage Studiesrdquo WorkingPaper 148-113 (Berkeley California Institute forResearch on Labor and Employment 2013) available athttpwwwirleberkeleyeduworkingpapers148-13pdf

28 We will report these results in a forthcoming workingpaper

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4243

Endnotes | wwwamericanprogresso

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4343

The Center for American Progress is a nonpartisan research and educational institute

dedicated to promoting a strong just and free America that ensures opportunity

for all We believe that Americans are bound together by a common commitment to

these values and we aspire to ensure that our national policies reflect these values

We work to find progressive and pragmatic solutions to significant domestic and

international problems and develop policy proposals that foster a government that

is ldquoof the people by the people and for the peoplerdquo

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 943

6 Center for American Progress | The Effects of M inim um Wages on SNAP Enro llments and Expenditures

Te CBO recenly projeced ha in 2016 17 million workers will earn less han

he $1010 hourly wage proposed in he Harkin-Miller bill Furhermore he

CBO esimaes ha an addiional 8 million workers earned beween $1010 and

$1150 per hour and were also likely o experience a wage increase15

Supplemental Nutrition Assistance Program

We ocus our inquiry on SNAP Benefis under he program are enirely

ederally unded he program is adminisered by he US Deparmen o

Agriculure ogeher wih sae agencies which share in adminisraion

coss Spending on SNAP has grown in he pas decade reaching $78 bil lion

in 2011 wih SNAP enrollmen increasing o 45 million people abou one-

sevenh o he US populaion16 According o he CBO changes since 1990 in

SNAP spending and enrollmens are primarily he resul o cyclical economic

condiions noably changes in he unemploymen rae and changes in per capiaincome17 Te 2009 American Recovery and Reinvesmen Ac emporarily

increased SNAP benefi amouns by 136 percen as repored by he US

Deparmen o Agriculurersquos Food and Nuriion Service hese higher benefi

levels expired on November 1 201318 Te CBO esimaes ha abou wo-

hirds o he changes in SNAP expendiure are associaed wih changes in he

number o recipiens and one-hird wih changes in he benefis received when

recipiensrsquo incomes change19

In fiscal year 2014 SNAPrsquos maximum monhly benefis are $189 or a single

individual $497 or a amily o hree and $750 or a amily o five Benefis are

reduced by 30 cens per dollar received and phase ou enirely a gross monhly

household incomes o 130 percen o he ederal povery level $1245 or a

single individual $2116 or a amily o hree and $2987 or a amily o five

o deermine benefis SNAP also defines a ne monhly income concep and

ses benefis a 100 percen o he ederal povery level using his concep

Calculaion o ne monhly income can include cerain deducions rom

monhly gross income such as medical expenses and child care coss Alhough

saes are permited some laiude on wha deducions are allowed in pracice

hese vary by very small amouns Our saisical model akes accoun o sae-specific differences in benefis20

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1043

Background | wwwamericanprogress

SNAP imposes an employmen or work-raining requiremen or able-bodied

prime-age aduls991252hose beween he ages o 18 and 50 and wihou disabiliies

or dependen children Such households can receive only hree monhs o benefis

in a hree-year period In recen years abou 85 percen o households receiving

benefis have incomes below he ederal povery level 49 percen have dependen

children 16 percen are age 60 or older 20 percen are disabled and 30 percenrepor some earned income21

A 2012 CBO repor also noes ha ake-up raes among eligible SNAP recipiens

average abou 70 percen wih much lower ake-up among elderly households

Te ake-up rae increases in harder economic imes I also increased when sigma

issues were reduced as SNAP debi cards replaced acual ood samps ake-up

is especially high among hose mos needy Adminisraive spending equaled 91

percen o he poenial spending ha would have occurred i all eligible recipiens

were enrolled Alhough some SNAP policy changes have occurred since 1990

mos were relaively minor and all were naional in scope Te 1996 welare reorm bill eliminaed SNAP eligibiliy or some legal immigrans limied he ime lengh

o eligibiliy or able-bodied childless aduls and reduced maximum benefis

Some o hese resricions were relaxed in 2002 and again in he American

Recovery and Reinvesmen Ac in 200922

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1143

8 Center for American Progress | The Effects of M inim um Wages on SNAP Enro llments and Expenditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1243

Methods and data | wwwamericanprogress

Methods and data

As previously menioned we exploi variaion in minimum wages by sae and ime

o examine heir causal effecs on SNAP enrollmens and expendiures o do so we

merge daa rom 1990 hrough 2012 drawn rom he Annual Social and Economic

Supplemen o he Curren Populaion Survey991252an annual Census Bureau survey

commonly known as he March CPS ha includes23 inormaion on SNAP

enrollmens a he amily level991252wih sae-level daa on minimum wages SNAP

expendiures populaion unemploymen raes and sae median income levels

o conrol or ime-varying heerogeneiy among saes our specificaions includeconrols or sae linear rends and effecs by Census division and ime We esimae

effecs a wo levels allowing or amily variaion and allowing only or sae-level

variaion We also employ a se o sandard demographic conrols such as amily size

and composiion and race and ehnic composiion

Distinguishing causation from correlation

How can we ensure ha our analysis does no pick up a spurious correlaion or

example he endency o more economically vibran saes o implemen higher

minimum wages Disinguishing correlaion and policy endogeneiy rom rue

causal effecs is he primary moivaion or economeric analysis In he ideal

experimen researchers would begin wih wo saes991252ha are alike in every

respec prior o he policy991252 and ldquoreardquo only one o hese saes wih a higher

minimum wage Tey would atemp o shield hese saes rom any influence ha

could obscure heir undersanding o he minimum wagersquos direc effec on SNAP

enrollmen Researchers o course canno conduc such experimens

We can however use saisical mehods o conrol simulaneously heindependen effecs on SNAP o sae unemploymen raes sae income levels

and common rajecories among saes wihin he same Census division By

ensuring similariy along hese dimensions we maximize he likelihood ha

SNAP aciviy in wo saes would have comparable oucomes in he absence

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1343

10 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

o a minimum wage policy change Tus i a new minimum wage policy were

implemened in one sae only he researchers could atribue all o he difference

hey observe in SNAP aciviy o he new minimum wage policy

In oher words we approximae he ideal experimen by using non-experimenal

saisical mehods Te desirable ldquopre-exising similariiesrdquo beween saes ha wehave defined above inorm our choice o conrol variables in a saisical seting

More precisely in our muliple regression models we use median amily income

he unemploymen rae he employmen-o-populaion raio and regional and

ime idenifiers o consruc an appropriae group o peers or each sae on he

eve o a policy change

Data description

wo daa ses include inormaion abou boh income and paricipaion inpublic programs Te Survey o Income and Program Paricipaion or SIPP

which is conduced in inermiten years has he advanage o ollowing he

same individuals over a period o ime In oher words i is a longiudinal daa

se I also has he advanage o conaining monhly daa However he sample

size o he SIPP is no sufficien or analyzing variaions in sae-level minimum

wages Te March CPS has he advanage o a much larger sample size and i is

conduced annually wihou any breaks in ime I has he disadvanage o being

a cross-secional daa se so we canno ollow he same individuals over ime991252

sricly speaking over more han one year On ne he March CPS is much more

suiable or our sudy We examine he empirical relaionship beween minimum

wage policy and ood samp aciviy a wo levels o aggregaion he amily level

and he sae level Family-level daa are drawn rom he March CPS

Te March CPS comprises responses rom he residens o 50000 o 60000

dwelling places surveyed per year and conains deailed inormaion on he

residensrsquo employmen and income including income rom ranser paymens

Te sample or our analysis comprises more han 128 million amily unis during

he period rom 1990 o 2012 (inclusive) Survey weighs allow us o analyze

SNAP paricipaion in a manner ha is represenaive o he US populaion alarge Over all years he share o amilies reporing ood samp receip in he

weighed March CPS sample is 91 percen Te enrollmen rae was a a low o 6

percen in he year 2000 In 2012 he mos recen year in our panel 133 percen

o amilies repored paricipaing in SNAP a some poin during he survey year

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1443

Methods and data | wwwamericanprogresso

Te March CPS also collecs inormaion on he number o SNAP recipiens in

he household he number o monhs he household receives SNAP benefis

and he cash-equivalen value o he SNAP benefis received However he

value o SNAP benefis is severely underrepored among recipiens perhaps

because recipiens are unaware o he exac monhly cash-equivalen value o

benefis hey receive

Our firs empirical sraegy ocuses solely on SNAP enrollmen By using he

amily as he uni o analysis we are able o inser saisical conrols o accoun

or non-wage-relaed acors ha influence any paricular amilyrsquos likelihood o

program paricipaion wih he inenion o isolaing any differences in program

paricipaion ha are due purely o changes in wage policy Tis approach

idenifies he effecs o low-wage labor policy on he exernal margin991252ha is

he effec o he minimum wage on he likelihood ha a amily paricipaes in he

SNAP program a all991252as opposed o he inernal margin or how much SNAP

unding he amily would receive

Our second empirical ramework uses sae-level adminisraive daa Ta is

we aggregae he daa o obain a single daa poin or each saeyear back

o 1990 represening he mean o he oucome or he sae Te sae-level

esimaion serves as a robusness check on he amily-level resuls or SNAP

paricipaion Also using aggregaed daa allows us o esimae direcly he

causal effec o minimum wage changes on SNAP spending Tis is no possible

a he amily level as discussed above daa on cash-equivalen value o ood

samps or SNAP recipiens is very requenly no repored in he March CPS

and when i is repored he inormaion may be unreliable By conras he

Bureau o Economic Analysis publishes aggregae SNAP spending a he sae

level in is Naional Income and Produc Accoun or NIPA ables Tus

while we are unable o observe he heerogeneiy in he cash value o SNAP

or amilies in each sae we are able o calculae average SNAP spending

per residen in each sae per year Supporing covariaes include he annual

unemploymen and employmen daa rom he Bureau o Labor Saisics or

BLS and sae-level populaion series rom he iner-decennial census releases

Minimum wage daa are available rom he BLSrsquos wages and hours division For

sae minimum wage changes enaced a oher imes han he firs o he yearan average value or he year is used

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1543

12 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

Family level model specifications

We firs examine he effec o he minimum wage on paricipaion in public

assisance programs For amily residing in sae and during year we esimae

an equaion o he ollowing orm

(1)

is a binary variable ha is se equal o 1 i a leas one member o amily

received ood samps during he survey year is a se o sae-level

characerisics including annual averages o he unemploymen rae he

employmen-o-populaion raio and he naural log o median amily income

is a vecor o amily atribues including indicaors or he race and marial saus

o he amily head size o he amily he presence o children and he presence oan adul male Sae fixed effecs are capured by o conrol or ime-varying

heerogeneiy our preerred model specificaion also includes year fixed effecs

ha vary by Census division ( ) and sae-level linear ime rends In

Appendix B we jusiy he inclusion o hese las wo erms We also compare he

resuls rom our preerred specificaion wih less sauraed specificaions

Te effec o ineres which is capured by is he expeced change in he

probabiliy o receiving SNAP benefis wih respec o a change in he (log o he)

binding minimum wage in sae during year We repor robus sandard errors

clusered a he sae level We esimae he parameers using linear regression

producing a linear probabiliy model Deails o he model selecion process are

covered in Appendix B below

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1643

Methods and data | wwwamericanprogresso

State-level model specifications

Te sae-level models are similarly specified For sae in year we assume ha

(2)

In his model is now eiher he SNAP enrollmen rae in sae during year

or he naural logarihm o per capia SNAP expendiures in sae during

year is once again a se o sae-level characerisics including he same

sae-level covariaes as in he amily regressions (annual average unemploymen

rae employmen-o-populaion raio naural log o median amily income)

wih he addiion o amily level characerisics averaged across he sae (average

amily size and he shares o populaion consiued by each o five racialehnicgroups) Sae fixed effecs are represened by As above our preerred model

specificaion includes year fixed effecs ha vary by Census division ( ) and

sae-level linear ime rends as elaboraed in Appendix B Te effec o

ineres is capured by

We esimae boh sae-level models (enrollmen and expendiures) using

ordinary leas squares regression Tus he inerpreaion o he coefficien is no

longer ha o a change in probabiliy as in he binary oucome models described

above Raher or he sae-level SNAP enrollmen model represens he

expeced change (in percenage poins) in he saersquos SNAP enrollmen rae ha

is due o a 1 percen change in he minimum wage For he SNAP expendiures

model is simply he elasiciy o SNAP spending wih respec o he minimum

wage991252ha is he percenage change in sae expendiures expeced o resul rom

a 1 percen change in ha saersquos minimum wage For urher deails on model

specificaion reer o Appendix B below

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1743

14 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1843

Results | wwwamericanprogresso

Results

Estimated minimum wage effects on SNAP enrollment and

expenditures

able 1 shows he esimaed parameer o ineres991252he coefficien o he

minimum wage or he preerred model o each ype Coefficiens on he

minimum wage variable are no direcly comparable across models because all

our models have a differen uncional orm o undersand and compare hese

esimaes we compue he change in SNAP aciviy prediced or a paricular wage scenario Te final column in able 1 answers he quesion Wha would be

he expeced change in SNAP aciviy in response o a 10 percen increase in he

minimum wage Te answer o his quesion varies wih he value o he inpu

parameers in he able we calculae he percenage decrease in enrollmen or

expendiures prediced or he average sae wih a minimum wage o $725 in

2014 Te sae-level SNAP expendiure model which is a consan-elasiciy

model conveys elasiciy inormaion direcly or he change in expendiures per

capia in he sae

TABLE 1

Comparison of national SNAP predictions for a 10 percent increase in the federal minimum wage

Model LevelRegression

type

Predicted outcome Coefficient of log

(minimum wage)

(Standard error)

Effect of a 10 percent in

in the minimum wage

VariableForm of

variable

Total

enrollment

Tot

expend

1 Family Linear

probability Enrollment Binary (enrolled=1)

-0042

(0008) -317 N

2

StateLinear regression

(ordinary least

squares)

Enrollment State enrollment rate (percent) -0031

(0012) -235 N

3 Expenditures Log (state expenditures per capita) -0190

(0103) NA -19

plt01 plt005 plt001Note Predicted changes are calculated for the average state with a minimum wage of $725 in 2014

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1943

16 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

According o his model a 10 percen increase in he minimum wage would resul

in a 19 percen reducion in SNAP expendiures Tis resul is reassuringly similar

o Dubersquos resul or povery reducion Te wo enrollmen models are much more

precisely esimaed han he expendiure model Te sae-level enrollmen model

finds ha a 10 percen minimum wage increase in a low-wage sae is associaed wih

a 235 percen decrease in SNAP enrollmens Te amily-level linear probabiliymodel predics a somewha greaer elasiciy or low minimum wage saes an

increase o 10 percen in he ederal minimum would resul in a 317 percen decline

in SNAP enrollmen24 Te differences in hese esimaes sem rom a number o

acors including difference in model uncional orm and daa used We rea his

range o elasiciy esimaes as an upper and lower bound on enrollmen impacs

Harkin-Miller bill National and state-level predicted impacts

Wha would be he prediced change or he SNAP program i he ederalminimum were raised o $1010 as proposed in he Harkin-Miller bill In order o

make his inerence we accoun or he ac ha no all saes are currenly subjec

o he ederal minimum wage a he beginning o 2014 21 saes mainained

higher minimum wages han $725 In hose saes an increase in he ederal

minimum wage may or may no be binding or employers in he sae depending

upon wheher he new ederal minimum exceeds he sae-level minimum Bu

regardless o wheher a minimum wage change is binding he impac on SNAP

aciviy will be lower in high minimum wage saes In order o accoun or his

properly we calculae sae by sae he percenage wage change ha would resul

rom he Harkin-Miller proposal and apply he parameers rom each o he hree

models above o compue he expeced decrease in SNAP aciviy or each sae

In his exercise we use saesrsquo curren (2014) minimum wage levels and assume

as a baseline he 2012 levels o SNAP enrollmen and expendiure as 2012 is he

mos recen year or which SNAP daa are available

able 10 and able 11 in Appendix C repor he esimaed effecs on SNAP

enrollmen and expendiures respecively or each sae under he Harkin-Miller

bill25 An increase o $1010 i enaced oday would represen beween a 393

percen wage increase in a $725 minimum wage sae and an 84 percen increasein Washingon sae which has he highes minimum wage in he naion a $932

as o January 201426 Slighly more han 56 percen o he decrease in expendiures

and abou 59 percen o he decrease in enrollmen would occur in saes wih

presen-day minimum wages o $725

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2043

Results | wwwamericanprogresso

In 2012 hese saes were home o 46 percen o he American populaion

and accouned or a roughly equivalen percenage o oal naional SNAP

expendiures No surprisingly he larges enrollmen decreases come rom

populous saes wih very high SNAP enrollmen raes andor very low minimum

wages Te larges prediced enrollmen reducion991252beween 319000 individuals

and 362000 individuals991252would occur in exas which has a $725 minimum wage In Caliornia which has a 206 percen SNAP paricipaion rae and an $8

per hour minimum wage we anicipae SNAP enrollmen declines o 310000

persons o 371000 persons And in Florida which had a SNAP paricipaion rae

o 166 percen and a $793 minimum wage enrollmen could decline by beween

164000 individuals and 196000 individuals For he our saes ha ogeher

accouned or he greaes amoun o SNAP spending in 2012991252exas Caliornia

Florida and New York respecively991252he combined expendiure reducion rom

he Harkin-Miller bill is prediced o be $14 billion

able 2 summarizes he prediced declines in SNAP aciviy or he naion as a whole ha would resul rom he direc and indirec effecs o he Harkin-Miller

bill Enrollmen would all beween 31 million persons and 36 million persons

represening 75 percen o 87 percen o curren enrollmen Te anicipaed

reducion in program expendiures would be nearly $46 billion or 61 percen o

program expendiures

TABLE 2

Comparison of national SNAP predictions under the Harkin-Miller billrsquos $1010 minimum wage

Model

Enrollment

(persons)

Expenditures

(millions of dollars)

Current (2012) Predicted Change Current (2012) Predicted Chan

Family enrollment (linear probability)

41866195

45489339 -3623144

$74861

NA NA

State enrollment (ordinary least squares) 38745435 -3120759 NA NA

State expenditures (ordinary least squares) NA NA $70305 -$45

Note Calculations use 2014 state minimum wages and the most recent SNAP data from 2012 They assume that per-enrollee expenditures remain constant

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2143

18 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

Tere are o course oher possibiliies or minimum wage legislaion able 3

shows he expeced SNAP changes or he Unied Saes under a variey o wage

scenarios calculaed using he sae-level models I saes were no able o se

heir minimum wages independenly such ha all saes were consrained by

he ederal minimum o $725 SNAP would be received by abou 514000 more

people across he Unied Saes a an addiional program cos o nearly hree-quarers o a billion dollars In conras he effecs o a higher minimum wage

proposal991252a ederal wage floor o $11 per hour991252would decrease enrollmen in

SNAP by more han 10 percen and decrease program coss by 83 percen

TABLE 3

Summary of par ticipation and expenditures under wage scenarios

If all states had

minimum wages of

Enrollment(persons)

Expenditures

(millions of dollars)

Predicted Change Predicted Change

Recent levels (2014) 41866195 $74861

$725 42380520 514326 $75604 $743

$800 41423919 -442276 $74209 -$652

$900 40148451 -1717744 $72350 -$2511

$1000 38872982 -2993212 $70490 -$4371

$1010 38745435 -3120759 $70305 -$4556

$1100 37597514 -4268681 $68631 -$6230

Note Calculations use state-level enrollment model coefficient

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2243

Conclusion | wwwamericanprogresso

Conclusion

An exensive body o lieraure examines employmen effecs o he minimum

wage A much smaller se o sudies examines how he minimum wage affecs

povery and only a handul o papers examine he effecs o he minimum wage

on he EIC Our analysis is he firs o examine he effecs o he minimum

wage on SNAP

Our findings indicae ha increased earnings rom minimum wage changes do

reduce SNAP enrollmens and expendiures We esimae ha he Harkin-Miller bill would save axpayers nearly $46 billion per year equivalen o 61 percen

o SNAP expendiures in 2012 he las year or which daa are available Over a

10-year period he esimaed savings amoun o nearly $46 billion

Our repor is subjec o limiaions ha we expec o overcome in our uure

research Firs he findings do no ake ino accoun possible ineracions among

SNAP he EIC and Medicaid Te eligibiliy cuoffs among hese programs

are quie differen suggesing ha such ineracions may be minor Noneheless

he join effecs can only be deermined by urher research using a causal

model Second i would be useul o know he disribuion o SNAP reducions

along he wage disribuion Using he Congressional Budge Officersquos calculaions

o how much he oal dollar value o wage would increase under he Harkin-

Miller proposal our findings imply ha he decline in overall SNAP spending

equals abou 15 percen o he oal resuling increase in wages Te amoun and

disribuion o his offse are o considerable ineres Minimum wage beneficiaries

who come rom working amilies already well above he povery line would no

see any offse while hose who are currenly considerably below he povery line

will see larger offses Tese issues will also be a subjec or our uure research

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2343

20 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

About the authors

Rachel West is a maser o public policy candidae a he Goldman School

o Public Policy Universiy o Caliornia Berkeley Her research ocuses on

economic policy in he areas o low-wage labor and povery

Michael Reich is proessor o economics and direcor o he Insiue or

Research on Labor and Employmen a he Universiy o Caliornia a Berkeley

His research publicaions cover numerous areas o labor economics including

racial inequaliy labor marke segmenaion high-perormance workplaces

union-managemen cooperaion Japanese labor-managemen sysems living

wages and minimum wages He received his docorae in economics rom

Harvard Universiy

Acknowledgments

We are graeul o Sylvia Allegreto Arindraji Dube Bill Leser Jesse Rohsein

Daniel Tompson and Ben Zipperer or heir valuable suggesions

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2443

References | wwwamericanprogresso

References

Allegreto Sylvia and ohers 2013 ldquoFas Food Povery Wages Te Public Cos o Low-Wage Jobsin he Fas-Food Indusryrdquo Universiy o Caliornia Berkeley Cener or Labor Research andEducaion Available a htplaborcenerberkeleyedupubliccossas_ood_povery_wage

Allegreto Sylvia and ohers 2013 ldquoCredible Research Designs or Minimum Wage Sudiesrdquo

Working Paper 148-13 Universiy o Caliornia Berkeley Insiue or Research on Labor andEmploymen Available a htpwwwirleberkeleyeduworkingpapers148-13pd

Biler Marianne and Hilary Hoynes 2013 ldquo Te More Tings Change he More Tey Say heSame Te Saey Ne Living Arrangemens and Povery in he Grea Recessionrdquo Working Paper19449 Naional Bureau o Economic Research

Congressional Budge Office 2012 ldquoTe Supplemenal Nuriion Assisance Programrdquo Washingon Available a htpwwwcbogovsiesdeaulfilescbofilesatachmens04-19-SNAPpd

991252 991252 991252 2014 ldquoTe Effec o a Minimum-Wage Increase on Employmen and Family Incomerdquo Washingon Available a htpwwwcbogovsiesdeaulfilescbofilesatachmens44995-MinimumWagepd

Dube Arindraji 2013 rdquoMinimum Wages and he Disribuion o Family Incomerdquo Unpublished working paper Available a htpsdldropboxuserconencomu15038936Dube_ MinimumWagesFamilyIncomespd

991252 991252 991252 2014 ldquoTe Povery o Minimum Wage Facsrdquo Arindraji Dube Blog January 22 Availablea htparindubecom20140122he-povery-o-minimum-wage-acs

Lee David and Emmanuel Saez 2012 ldquoOpimal Minimum Wage Policy in Compeiive LaborMarkesrdquo Journal o Public Economics 96 (9) 739ndash749

Neumark David and William Wascher 1992 ldquoEmploymen Effecs o Minimum and Subminimum Wages Panel Daa on Sae Minimum Wage Lawsrdquo Industrial and Labor Relations Review 46 (1)

55ndash81

Neumark David and William Wascher 2011 ldquoDoes a Higher Minimum Wage Enhance heEffeciveness o he Earned Income ax Credirdquo Industrial and Labor Relations Review 64 (4)712ndash746

Page Marianne Joanne Spez and Jane Millar 2005 ldquoDoes he Minimum Wage Affec WelareCaseloadsrdquo Journal o Policy Analysis and Management 24 (2) 273ndash295

Rohsein Jesse 2010 ldquoIs he EIC as Good as an NI Condiional Cash ransers and ax

Incidencerdquo American Economic Journal Economic Policy 2 (1) 177ndash208

Wage and Hours Division 2014 ldquoMinimum Wage Laws in he Saes ndash January 1 2014rdquo USDeparmen o Labor (htpwwwdolgovwhdminwageamericahm [February 2014])

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2543

22 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2643

Appendix A | wwwamericanprogresso

Appendix A

Pre-trend falsification check

Recen minimum wage research27 highlighs a common flaw in previous sudies

ailure o veriy ha he oucome variable is ree o negaive pre-exising rends I

or example SNAP aciviy was already rending down in saes ha raised heir

minimum wages beore hese changes came ino effec our regression analysis could

(misakenly) atribue ha reducion o he minimum wage We check or such

pre-rends by inroducing variables ha represen he prior yearrsquos value or leads o

he minimum wage I he model esimaes he minimum wage o have an effec on

he oucome variable beore he wage change wen ino effec hen an unobservedacor no he minimum wage change caused he change in SNAP aciviy

We es he specificaions above or pre-rends by including a one-year lead in

all hree specificaions We find ha he lead erms are small posiive and no

saisically significan indicaing ha he concurren minimum wage991252no

he wage level in prior periods991252is driving he observed changes in SNAP

oucomes28 In paricular he coefficien (sandard error) on he lead erm in

our preerred amily-level enrollmen regression is 011 and no significan

while he coefficien and sandard error o he conemporaneous minimum

wage is unchanged In he sae-level preerred enrollmen regression he

coefficien o he lead erm is again small (07) and i is no significan Te

corresponding coefficien on he lead erm in he sae-level expendiure

regression is 16 and is no significan Te posiive poin esimaes on hese lead

erms resuls no only rule ou disoring negaive pre-rends Tey also sugges

ha our main resuls may underesimae he rue effecs

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2743

24 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2843

Appendix B | wwwamericanprogresso

Appendix B

Model selection process

For boh he amily-level and sae-level models we es hree mehods o conrol

or unobserved geographic- and ime-varying characerisics as suggesed by he

minimum wage lieraure Firs we include only independen sae-specific fixed

effecs and year-specific fixed effecs Tis specificaion (specificaion 1) implicily

assumes ha amilies in any sae consiue an equally good saisical ldquoconrolrdquo

group or hose in any randomly chosen sae afer accouning or various

characerisics (median income and unemploymen rae among ohers) Similarly

simple ime fixed effecs assume ha amilies surveyed in any year can crediblyserve as a conrol group or amilies surveyed in every oher year o he sample

(1990 hrough 2012)

In oher words specificaion 1 assumes ha a saersquos immediae neighbor provides

no beter a couneracual or he effec o a minimum wage change han does a

sae across he counry We relax his resricive specificaion sequenially in wo

seps In specificaion 2 we replace simple year fixed effecs wih fixed effecs or

each Census divisionyear (capured as an addiional variable in he vecor By

using division-year effecs we remove he resricion ha amilies in each sae

are equally good saisical conrols or all oher amilies Raher we allow or he

possibiliy ha amilies in similar geographic regions (or example he Souh or

he Norheas) may be more similar o one anoher han amilies arher away

Finally in specificaion 3 we add sae-specific linear ime rends o he previous

specificaion Tus specificaion 3 is he mos rigorous model specificaion in ha

i allows or heerogeneiy along hree dimensions Ta is specificaion 3 allows

each sae o have is own ime-varying rends raher han imposing he resricion

ha saes evolve idenically over he 22 years in he sample

We begin building he heoreical specificaion above rom a se o simpleuncondiional models regression o SNAP aciviy (enrollmen or expendiures)

on he log o he minimum wage and a se o geographic- and ime-specific

effecs (specificaions 1 2 and 3 described above) As shown in ables 1ndash3 (or

specificaion 3) we hen add covariaes sequenially o hese models including

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2943

26 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

firs he vecor o amily-level conrols ollowed by each o several sae-level

covariaes in urn he unemploymen rae log o median-amily income and he

employmen-o-populaion raio Comparable resuls or specificaions 2 and 3

will be available in our orhcoming working paper

As expeced he simple uncondiional models indicae ha he relaionship beween he minimum wage and SNAP enrollmen i one exiss is a more complex

one influenced by oher acors In he uncondiional model he coefficien on he

variable o ineres991252he log o he minimum wage991252is small in magniude and no

saisically differen rom zero Once we accoun or he influence o labor marke

condiions and variaion in income levels on program paricipaion (by including

unemploymen rae and median-amily income conrol variables respecively)

he effec o he minimum wage on SNAP enrollmen is precisely esimaed Te

coefficien o he log minimum wage is slighly higher (-0042) in he amily-level

analysis han he coefficien (-031) in he sae-level analysis Te level o precision

is also higher in he amily-level analysis Tis is o be expeced when using 124million observaions compared o 1127

Te second dimension o model choice concerns he effec specificaion ables

7ndash9 compare he primary coefficiens o ineres or he SNAP enrollmen and

expendiure models For boh he enrollmen models he effec sizes are smalles

or specificaion 1 larges or specificaion 2 and inermediae beween hese wo

in specificaion 3 Recall ha Specificaion 3 conains sae-specific linear ime

rends in addiion o he census divisionyear conrols included in specificaion

2 In he amily-level enrollmen model he sandard error o he minimum wage

coefficien is smaller han in he oher wo specificaions Sandard errors on he

oher variables are much smaller in specificaions 2 and 3 han in specificaion

1 On he basis o coefficien significance (join and individual) specificaions 2

and 3 are sricly preerred in boh enrollmen models o specificaion 1 which

conains only sae and year fixed effecs

A concern wih specificaions 2 and 3 is ha rend conrols such as sae linear

rends may incorrecly absorb some o he delayed impac o a minimum wage

When we es his issue by including lagged minimum wages we do no find ha

delayed effecs are significan Anoher concern is ha more sauraed modelsuse less o he saisical variaion which could reduce he saisical power o

he resuls However he sandard errors or our more sauraed models are no

higher and are lower in some cases han or he less sauraed models Overall

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3043

Appendix B | wwwamericanprogresso

his evidence suppors our use o he sauraed model as he preerred model

specificaion Moreover Dubersquos 2013 sudy shows ha more sauraed models

perorm beter han models wih jus sae and ime fixed effecs

Te esimaed enrollmen regressions a boh he amily and sae levels show large

and saisically significan coefficiens Te esimaed minimum wage effec in heexpendiures regressions991252or which we have only sae-level daa991252is also large

and saisically significan

We do no use weighed regression or he sae-level models preerring o keep

analysis o he ldquoreamenrdquo (ha is o say a minimum wage change) appropriae

o he average sae raher han he average amily or individual I insead our

primary ineres were he impac o a minimum wage change on he average amily

or he average individual we migh choose o designae he number o amilies

in each sae or he sae populaion respecively as analyic weighs in order o

obain a coefficien beter suied or such inerence

TABLE 4

SNAP enrollment

Family-level linear probability

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0048 -0047 -0040 -0043 -0042

(0013) (0013) (001) (0008) (0009)

Unemployment rate (100) 0505 0420 0280

(0083) (0086) (0082)

Log median income -0057 -0039

(0011) (0011)

Employment-to-population ratio -0239

(0038)

N 1242022 1242022 1242022 1242022 1242022

plt01 plt005 plt001

Note Robust standard errors are in parentheses Observations are clustered at the state level The outcome variable is binary and equal to one if a family is enrolledin SNAP All models include state fixed effects Census division x-year fixed effects and state-specific linear time trends All specifications except 3a include additional

controls for family size race and marital status of the family head presence of children and presence of an adult maleSource Annual data from the March Current Population Survey for 1990 to 2012 estimation includes Current Population Survey probability weights

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3143

28 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

TABLE 5

SNAP enrollment

State-level linear regression

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0028 -0024 -0026 -0031 -0031

(0014) (0013) (0013) (0012) (0012)

Unemployment rate (100) 0625 0466 0320 0339

(0087) (0088) (0085) (0083)

Log median income -0090 -0065 -0061

(0013) (0013) (0013)

Employment-to-population ratio -0282 -0248

(0037) (0038)

N 1127 1127 1127 1127 1127

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the SNAP enrollment rate All models include state fixed effects Census division x-year fixedeffects and state-specific linear time trends

Source Annual data from the March Current Population Survey for 1990 to 2012

TABLE 6

SNAP expenditures

State-level linear regression

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0181 -0149 -0156 -0153 -0190

(011) (0103) (0102) (0103) (0103)

Unemployment rate (100) 4587 4152 4232 4313

(0622) (0621) (0633) (0628)

Log median income -0246 -0261 -0294

(0075) (0078) (0078)

Employment-to-population ratio 0155 0244

(0237) (024)

N 1127 1127 1127 1127 1127

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the natural log of state SNAP expenditures per capita for 1990 to 2012 All models include state

fixed effects Census division x-year fixed effects and state-specific linear time trends

Source Annual data from Bureau of Economic Analysis national income and product accounts tables for 1990 to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3243

Appendix B | wwwamericanprogresso

TABLE 7

Comparison of specifications SNAP enrollment

Family-level linear probability

(1) (2) (3)

Log minimum wage -0028 -0049 -0042

(0014) (0017) (0009)

Unemployment rate (100) 0275 0297 0280

(0161) (0076) (0082)

Log median income -0077 -0055 -0039

(0014) (0012) (0011)

Employment-to-population ratio -0238 -0250 -0239

(0054) (004) (0038)

N 1242022 1242022 1242022

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses Observations are clustered at the state level The outcomevariable is binary or equal to one if a family is enrolled in SNAP All specifications include additional controls forfamily size race and marital status of the family head presence of children and presence of an adult male

Source Annual data from the March Current Population Survey for 1990 to 2012 estimation includes CurrentPopulation Survey probability weights

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3343

30 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

TABLE 8

Comparison of specifications SNAP enrollment

State-level linear regression

(1) (2) (3)

Log minimum wage -0019 -0035 -0031

(0009) (0012) (0012)

Unemployment rate (100) 0401 0370 0339

(0063) (0077) (0083)

Log median income -0081 -0073 -0061

(0011) (0013) (0013)

Employment-to-population ratio -0183 -0222 -0248

(0039) (0039) (0038)

N 1127 1127 1127

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the SNAP enrollment rate Allregressions include the share of households with children and the racial shares of the population

Source Annual data from the March Current Population Survey for 1990 to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3443

Appendix B | wwwamericanprogresso

TABLE 9

Comparison of specifications SNAP expenditures

State-level linear regression

(1) (2) (3)

Log minimum wage -0121 -0203 -0190

(0075) (0103) (0103)

Unemployment rate (100) 5292 5152 4313

(0464) (0576) (0628)

Log median income -0437 -0417 -0294

(008) (0086) (0078)

Employment-to-population ratio -0040 -0220 0244

(0261) (0260) (0240)

N 1127 1127 1127

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the natural log of state SNAPexpenditures per capita All regressions include the share of households with children and the racial shares ofthe state population

Source Annual data from Bureau of Economic Analysis national income and product accounts tables for 1990to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3543

32 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3643

Appendix C | wwwamericanprogresso

Appendix C Harkin-Miller

policy simulation results

StateMinimum wage

(2014)

SNAP enrollment (2012) Predicted change in SNAP enrollment

RateTotal

(persons)

Model 1 family

enrollment (linearprobability)

Model 2 state

enrollment (linearregression)

Model 3 s

expenditu(linear regre

Alabama $725 164 788682 -66922 -58897 -58906

Alaska $775 120 87436 -8104 -6887 -3288

Arizona $790 201 1319323 -67435 -56738 -64356

Arkansas $725 78 230489 -40977 -36063 -35248

California $800 206 7813680 -371131 -310222 -18223

Colorado $800 164 853155 -50684 -42365 -23926

Connecticut $870 91 326621 -22456 -17975 -13711

Delaware $725 186 170262 -12739 -11211 -10647

District of Columbia $825 133 84009 -5370 -4417 -3632

Florida $793 166 3208026 -195813 -164426 -13046

Georgia $725 160 1586336 -137741 -121224 -11004

Hawaii $725 96 133662 -19310 -16995 -14933

Idaho $725 92 147501 -22165 -19507 -15809

Illinois $825 95 1225084 -109088 -89742 -70955

Indiana $725 125 816233 -90818 -79928 -83985

Iowa $725 155 478011 -42716 -37594 -28556

Kansas $725 135 388269 -40082 -35275 -27461

Kentucky $725 130 568821 -60840 -53544 -52259

Louisiana $725 149 683832 -63929 -56263 -66083

Maine $750 77 101976 -16567 -14323 -15234

Maryland $725 144 846415 -81748 -71946 -38370

Massachusetts $800 130 864721 -64902 -54251 -42913

Michigan $740 146 1439141 -128801 -112140 -11022

Minnesota $725 133 713646 -74730 -65769 -37878

TABLE 10

SNAP enrollments Predicted changes by state under the Harkin-Miller bill

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3743

34 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

StateMinimum wage

(2014)

SNAP enrollment (2012) Predicted change in SNAP enrollment

RateTotal

(persons)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 state

expenditures

(linear regression)

Mississippi $725 129 386501 -41486 -36511 -46467

Missouri $750 172 1036182 -75131 -64952 -56244

Montana $790 132 132452 -10350 -8708 -5846

Nebraska $725 124 230683 -25773 -22683 -12189

Nevada $825 162 446035 -23349 -19209 -11894

New Hampshire $725 127 168404 -18359 -16157 -5735

New Jersey $825 160 1416666 -75175 -61843 -28236

New Mexico $750 149 310896 -25983 -22463 -22512

New York $800 192 3763553 -191193 -159815 -142182

North Carolina $725 174 1697193 -135417 -119179 -113503

North Dakota $725 87 61225 -9743 -8574 -4021

Ohio $795 143 1647345 -115869 -97169 -88580

Oklahoma $725 129 494053 -53006 -46650 -46854

Oregon $910 124 485326 -17036 -13328 -16398

Pennsylvania $725 161 2053643 -177315 -156052 -125586

Rhode Island $800 156 163730 -10258 -8574 -8698

South Carolina $725 94 445277 -65614 -57746 -50304

South Dakota $725 208 173749 -11586 -10197 -7458

Tennessee $725 142 914903 -89667 -78915 -99134

Texas $725 110 2863779 -362018 -318607 -253285

Utah $725 88 251107 -39658 -34902 -19390

Vermont $873 156 97792 -3823 -3055 -2475

Virginia $725 101 829771 -113723 -100086 -58212

Washington $932 72 496934 -23221 -17947 -17756

West Virginia $725 58 107875 -25792 -22699 -21665

Wisconsin $725 75 427822 -79521 -69986 -53210

Wyoming $725 164 94590 -8010 -7050 -3104

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3843

Appendix C | wwwamericanprogresso

StateMinimum wage

(2014)

SNAP expenditures (2012)Predicted change in SNAP expenditures

(millions of dollars)

Per resident

(millions of

dollars)

Total

(millions of

dollars)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 s

expenditu

(linear regre

Alabama $725 $290 $1397 -$1185 -$1043 -$1043

Alaska $775 $253 $185 -$262 -$223 -$106

Arizona $790 $257 $1686 -$935 -$787 -$892

Arkansas $725 $248 $733 -$636 -$560 -$547

California $800 $189 $7164 -$7276 -$6082 -$3573

Colorado $800 $157 $816 -$862 -$721 -$407

Connecticut $870 $191 $686 -$343 -$275 -$210

Delaware $725 $250 $229 -$205 -$180 -$171

District of Columbia $825 $366 $232 -$146 -$120 -$99

Florida $793 $294 $5676 -$4429 -$3719 -$2951

Georgia $725 $317 $3140 -$2936 -$2584 -$2346

Hawaii $725 $335 $465 -$449 -$395 -$347

Idaho $725 $225 $359 -$376 -$331 -$268

Illinois $825 $249 $3200 -$2096 -$1725 -$1364

Indiana $725 $220 $1439 -$1162 -$1023 -$1075

Iowa $725 $192 $589 -$658 -$579 -$440

Kansas $725 $159 $460 -$502 -$441 -$344

Kentucky $725 $298 $1303 -$1133 -$997 -$973

Louisiana $725 $315 $1450 -$1047 -$922 -$1083

Maine $750 $281 $373 -$267 -$231 -$246

Maryland $725 $188 $1109 -$1765 -$1553 -$828

Massachusetts $800 $206 $1366 -$1030 -$861 -$681

Michigan $740 $300 $2963 -$2400 -$2090 -$2054

Minnesota $725 $140 $755 -$1113 -$980 -$564

Mississippi $725 $326 $973 -$649 -$571 -$726

Missouri $750 $241 $1452 -$1278 -$1104 -$956

Montana $790 $190 $191 -$179 -$151 -$101

Nebraska $725 $140 $259 -$409 -$360 -$194

Nevada $825 $191 $527 -$441 -$363 -$225

New Hampshire $725 $126 $167 -$399 -$351 -$125

TABLE 11

SNAP expenditures Predicted changes by state under the Harkin-Miller bill

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3943

36 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

StateMinimum wage

(2014)

SNAP expenditures (2012)Predicted change in SNAP expenditures

(millions of dollars)

Per resident

(millions of

dollars)

Total

(millions of

dollars)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 state

expenditures

(linear regression)

New Jersey $825 $160 $1420 -$1611 -$1325 -$605

New Mexico $750 $324 $675 -$513 -$444 -$445New York $800 $287 $5616 -$3766 -$3148 -$2801

North Carolina $725 $252 $2454 -$2187 -$1925 -$1833

North Dakota $725 $128 $90 -$162 -$143 -$67

Ohio $795 $259 $2995 -$2013 -$1688 -$1539

Oklahoma $725 $248 $945 -$799 -$703 -$706

Oregon $910 $322 $1255 -$272 -$213 -$262

Pennsylvania $725 $218 $2779 -$2930 -$2579 -$2075

Rhode Island $800 $280 $294 -$173 -$144 -$147

South Carolina $725 $291 $1373 -$1337 -$1177 -$1025South Dakota $725 $198 $165 -$192 -$169 -$123

Tennessee $725 $324 $2091 -$1413 -$1243 -$1562

Texas $725 $230 $5997 -$6402 -$5634 -$4479

Utah $725 $141 $402 -$614 -$541 -$300

Vermont $873 $230 $144 -$66 -$53 -$43

Virginia $725 $173 $1413 -$2062 -$1815 -$1056

Washington $932 $244 $1682 -$350 -$270 -$267

West Virginia $725 $273 $508 -$451 -$397 -$379

Wisconsin $725 $204 $1166 -$1302 -$1146 -$871Wyoming $725 $95 $55 -$105 -$93 -$41

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4043

Appendix C | wwwamericanprogresso

FIGURE 2

Raising the minimum wage to $1010would cut taxpayer costs in every state

Predicted decreases in cost and enrollment

in SNAP in 50 states

$200+$51ndash$100

$101ndash$200

0ndash$25

$26ndash$50

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4143

38 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

Endnotes

1 US Department of Agriculture ldquoSupplementalNutrition Assistance Program (SNAP)rdquo available athttpwwwfnsusdagovsnapeligibility (last accessedFebruary 2014)

2 For this initial analysis we do not consider Harkin-Millerrsquos increase in subminimum wages for tippedworkers To do s o would increase the estimated SNAP

savings by an unknown amount

3 The Congressional Budget Office estimates thatworkers currently earning between $1010 and $1150per hour would see their wages rise under the Harkin-Miller proposal Congressional Budget O ffice ldquoTheEffects of a Minimum Wage Increase on Employmentand Family Incomerdquo (2014)

4 Marianne Page Joanne Spetz and Jane Millar ldquoDoesthe Minimum Wage Affect Welfare Caseloadsrdquo Journalof Policy Analysis and Management 24 (2) (2005)273ndash295

5 Marianne Bitler and Hilary Hoynes ldquoThe More ThingsChange the More They Stay the Same The SafetyNet Living Arrangements and Poverty in the GreatRecessionrdquo NBER Working Paper 194 49 2013

6 Sylvia Allegretto and others ldquoFast Food Poverty Wages The Public Cost of Low-Wage Jobs in the Fast-FoodIndustryrdquo (Berkeley California Center for LaborResearch and Education 2013) available at httplaborcenterberkeleyedupubliccostsfast_food_poverty_wage

7 David Neumark and William Wascher ldquoDoes a HigherMinimum Wage Enhance the Effectiveness of theEarned Income Tax Creditrdquo Industrial and LaborRelations Review 64 (4) (2011) 712ndash746

8 David Lee and Emmanuel Saez ldquoOptimal MinimumWage Policy in Competitive Labor Marketsrdquo Journal ofPublic Economics 96 (9) (2012) 739ndash749

9 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family IncomerdquoFebruary 2014

10 Jesse Rothstein ldquoIs the EITC as Good as an NITConditional Cash Transfers and Tax Incidencerdquo AmericanEconomic Journal Economic Policy 2 (1) (2010) 177ndash208

11 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family Incomerdquo

12 Dube Arindrajit 2013 rdquoMinimum Wagesand the Distribution of Family IncomerdquoUnpublished working paper Available at httpsdldropboxusercontentcomu15038936Dube_MinimumWagesFamilyIncomespdf

13 As Dube explains in ldquoThe poverty of Minimum WageFactsrdquo the simulation approach underestimate stemsfrom a number of unwarranted assumptions includingthe range of actual wage increases and the accuracy ofwage data in the Current Population Survey The causal

approach does not make these assumptions

14 Allegretto Sylvia and others 2013 ldquoCredible ResearchDesigns for Minimum Wage Studiesrdquo Working Paper148-13 University of California Berkeley Institutefor Research on Labor and Employment Available athttpwwwirleberkeleyeduworkingpapers148-13pdf

15 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family In comerdquo

16 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo (2012)

17 Ibid

18 Ibid

19 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo

20 US Department of Agriculture ldquoSupplementalNutrition Assistance Program (SNAP)rdquo

21 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo

22 Ibid

23 We deviate from the Census Bureaursquos definition ofa family unit which is ldquotwo people or more (on e ofwhom is the householder) related by bir th marriageor adoption and residing togetherrdquo US Bureau ofthe Census ldquoCurrent Population Survey DefinitionsFamilyrdquo available at httpwwwcensusgovcpsabout

cpsdefhtml (last accessed February 2014) We countas a family unit any individual residing on his or herown two or more persons residing together whodo not belong to a family in the March CPS sampleare constructed as one family in our analysis For thepurposes of food stamp allocations the consumptionresulting from this transfer is probably distributed tofamily members (rather than household members ora single individual within the household) Howeversingle individuals canmdashand domdashreceive SNAPbenefits Excluding them would fail to make theanalysis reflective of the population at large

24 Strictly the family level linear probability modelpredicts the percentage-point decrease in theprobability that an individual family will receive SNAPpayments When applied to a large number of familieshowever we are able to interpret the coefficient asa decrease in the mean of enrollmentmdashthat is a

decrease in the enrollment ratemdashby applying the lawof iterated expectations

25 We generate expenditure predictions from theenrollment modelsmdashand conversely generateenrollment predictions from the expenditure modelmdashby assuming that expenditures per enrolled familyremains the same before and after the minimum wagechange In practice this is likely to be a conservativeestimatemdashthat is to underestimate the decrease inSNAP activity Average SNAP benefits per family willalso decrease as many families that remain eligible forSNAP experience income gains

26 Wage and Hour Division ldquoMinimum Wage Laws inthe States ndash Januar y 1 2014rdquo available at httpwwwdolgovwhdminwageamericahtm (last accessedFebruary 2014)

27 See for example Sylvia Allegretto and others ldquoCredibleResearch Designs for Minimum Wage Studiesrdquo WorkingPaper 148-113 (Berkeley California Institute forResearch on Labor and Employment 2013) available athttpwwwirleberkeleyeduworkingpapers148-13pdf

28 We will report these results in a forthcoming workingpaper

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4243

Endnotes | wwwamericanprogresso

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4343

The Center for American Progress is a nonpartisan research and educational institute

dedicated to promoting a strong just and free America that ensures opportunity

for all We believe that Americans are bound together by a common commitment to

these values and we aspire to ensure that our national policies reflect these values

We work to find progressive and pragmatic solutions to significant domestic and

international problems and develop policy proposals that foster a government that

is ldquoof the people by the people and for the peoplerdquo

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1043

Background | wwwamericanprogress

SNAP imposes an employmen or work-raining requiremen or able-bodied

prime-age aduls991252hose beween he ages o 18 and 50 and wihou disabiliies

or dependen children Such households can receive only hree monhs o benefis

in a hree-year period In recen years abou 85 percen o households receiving

benefis have incomes below he ederal povery level 49 percen have dependen

children 16 percen are age 60 or older 20 percen are disabled and 30 percenrepor some earned income21

A 2012 CBO repor also noes ha ake-up raes among eligible SNAP recipiens

average abou 70 percen wih much lower ake-up among elderly households

Te ake-up rae increases in harder economic imes I also increased when sigma

issues were reduced as SNAP debi cards replaced acual ood samps ake-up

is especially high among hose mos needy Adminisraive spending equaled 91

percen o he poenial spending ha would have occurred i all eligible recipiens

were enrolled Alhough some SNAP policy changes have occurred since 1990

mos were relaively minor and all were naional in scope Te 1996 welare reorm bill eliminaed SNAP eligibiliy or some legal immigrans limied he ime lengh

o eligibiliy or able-bodied childless aduls and reduced maximum benefis

Some o hese resricions were relaxed in 2002 and again in he American

Recovery and Reinvesmen Ac in 200922

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1143

8 Center for American Progress | The Effects of M inim um Wages on SNAP Enro llments and Expenditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1243

Methods and data | wwwamericanprogress

Methods and data

As previously menioned we exploi variaion in minimum wages by sae and ime

o examine heir causal effecs on SNAP enrollmens and expendiures o do so we

merge daa rom 1990 hrough 2012 drawn rom he Annual Social and Economic

Supplemen o he Curren Populaion Survey991252an annual Census Bureau survey

commonly known as he March CPS ha includes23 inormaion on SNAP

enrollmens a he amily level991252wih sae-level daa on minimum wages SNAP

expendiures populaion unemploymen raes and sae median income levels

o conrol or ime-varying heerogeneiy among saes our specificaions includeconrols or sae linear rends and effecs by Census division and ime We esimae

effecs a wo levels allowing or amily variaion and allowing only or sae-level

variaion We also employ a se o sandard demographic conrols such as amily size

and composiion and race and ehnic composiion

Distinguishing causation from correlation

How can we ensure ha our analysis does no pick up a spurious correlaion or

example he endency o more economically vibran saes o implemen higher

minimum wages Disinguishing correlaion and policy endogeneiy rom rue

causal effecs is he primary moivaion or economeric analysis In he ideal

experimen researchers would begin wih wo saes991252ha are alike in every

respec prior o he policy991252 and ldquoreardquo only one o hese saes wih a higher

minimum wage Tey would atemp o shield hese saes rom any influence ha

could obscure heir undersanding o he minimum wagersquos direc effec on SNAP

enrollmen Researchers o course canno conduc such experimens

We can however use saisical mehods o conrol simulaneously heindependen effecs on SNAP o sae unemploymen raes sae income levels

and common rajecories among saes wihin he same Census division By

ensuring similariy along hese dimensions we maximize he likelihood ha

SNAP aciviy in wo saes would have comparable oucomes in he absence

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1343

10 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

o a minimum wage policy change Tus i a new minimum wage policy were

implemened in one sae only he researchers could atribue all o he difference

hey observe in SNAP aciviy o he new minimum wage policy

In oher words we approximae he ideal experimen by using non-experimenal

saisical mehods Te desirable ldquopre-exising similariiesrdquo beween saes ha wehave defined above inorm our choice o conrol variables in a saisical seting

More precisely in our muliple regression models we use median amily income

he unemploymen rae he employmen-o-populaion raio and regional and

ime idenifiers o consruc an appropriae group o peers or each sae on he

eve o a policy change

Data description

wo daa ses include inormaion abou boh income and paricipaion inpublic programs Te Survey o Income and Program Paricipaion or SIPP

which is conduced in inermiten years has he advanage o ollowing he

same individuals over a period o ime In oher words i is a longiudinal daa

se I also has he advanage o conaining monhly daa However he sample

size o he SIPP is no sufficien or analyzing variaions in sae-level minimum

wages Te March CPS has he advanage o a much larger sample size and i is

conduced annually wihou any breaks in ime I has he disadvanage o being

a cross-secional daa se so we canno ollow he same individuals over ime991252

sricly speaking over more han one year On ne he March CPS is much more

suiable or our sudy We examine he empirical relaionship beween minimum

wage policy and ood samp aciviy a wo levels o aggregaion he amily level

and he sae level Family-level daa are drawn rom he March CPS

Te March CPS comprises responses rom he residens o 50000 o 60000

dwelling places surveyed per year and conains deailed inormaion on he

residensrsquo employmen and income including income rom ranser paymens

Te sample or our analysis comprises more han 128 million amily unis during

he period rom 1990 o 2012 (inclusive) Survey weighs allow us o analyze

SNAP paricipaion in a manner ha is represenaive o he US populaion alarge Over all years he share o amilies reporing ood samp receip in he

weighed March CPS sample is 91 percen Te enrollmen rae was a a low o 6

percen in he year 2000 In 2012 he mos recen year in our panel 133 percen

o amilies repored paricipaing in SNAP a some poin during he survey year

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1443

Methods and data | wwwamericanprogresso

Te March CPS also collecs inormaion on he number o SNAP recipiens in

he household he number o monhs he household receives SNAP benefis

and he cash-equivalen value o he SNAP benefis received However he

value o SNAP benefis is severely underrepored among recipiens perhaps

because recipiens are unaware o he exac monhly cash-equivalen value o

benefis hey receive

Our firs empirical sraegy ocuses solely on SNAP enrollmen By using he

amily as he uni o analysis we are able o inser saisical conrols o accoun

or non-wage-relaed acors ha influence any paricular amilyrsquos likelihood o

program paricipaion wih he inenion o isolaing any differences in program

paricipaion ha are due purely o changes in wage policy Tis approach

idenifies he effecs o low-wage labor policy on he exernal margin991252ha is

he effec o he minimum wage on he likelihood ha a amily paricipaes in he

SNAP program a all991252as opposed o he inernal margin or how much SNAP

unding he amily would receive

Our second empirical ramework uses sae-level adminisraive daa Ta is

we aggregae he daa o obain a single daa poin or each saeyear back

o 1990 represening he mean o he oucome or he sae Te sae-level

esimaion serves as a robusness check on he amily-level resuls or SNAP

paricipaion Also using aggregaed daa allows us o esimae direcly he

causal effec o minimum wage changes on SNAP spending Tis is no possible

a he amily level as discussed above daa on cash-equivalen value o ood

samps or SNAP recipiens is very requenly no repored in he March CPS

and when i is repored he inormaion may be unreliable By conras he

Bureau o Economic Analysis publishes aggregae SNAP spending a he sae

level in is Naional Income and Produc Accoun or NIPA ables Tus

while we are unable o observe he heerogeneiy in he cash value o SNAP

or amilies in each sae we are able o calculae average SNAP spending

per residen in each sae per year Supporing covariaes include he annual

unemploymen and employmen daa rom he Bureau o Labor Saisics or

BLS and sae-level populaion series rom he iner-decennial census releases

Minimum wage daa are available rom he BLSrsquos wages and hours division For

sae minimum wage changes enaced a oher imes han he firs o he yearan average value or he year is used

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1543

12 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

Family level model specifications

We firs examine he effec o he minimum wage on paricipaion in public

assisance programs For amily residing in sae and during year we esimae

an equaion o he ollowing orm

(1)

is a binary variable ha is se equal o 1 i a leas one member o amily

received ood samps during he survey year is a se o sae-level

characerisics including annual averages o he unemploymen rae he

employmen-o-populaion raio and he naural log o median amily income

is a vecor o amily atribues including indicaors or he race and marial saus

o he amily head size o he amily he presence o children and he presence oan adul male Sae fixed effecs are capured by o conrol or ime-varying

heerogeneiy our preerred model specificaion also includes year fixed effecs

ha vary by Census division ( ) and sae-level linear ime rends In

Appendix B we jusiy he inclusion o hese las wo erms We also compare he

resuls rom our preerred specificaion wih less sauraed specificaions

Te effec o ineres which is capured by is he expeced change in he

probabiliy o receiving SNAP benefis wih respec o a change in he (log o he)

binding minimum wage in sae during year We repor robus sandard errors

clusered a he sae level We esimae he parameers using linear regression

producing a linear probabiliy model Deails o he model selecion process are

covered in Appendix B below

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1643

Methods and data | wwwamericanprogresso

State-level model specifications

Te sae-level models are similarly specified For sae in year we assume ha

(2)

In his model is now eiher he SNAP enrollmen rae in sae during year

or he naural logarihm o per capia SNAP expendiures in sae during

year is once again a se o sae-level characerisics including he same

sae-level covariaes as in he amily regressions (annual average unemploymen

rae employmen-o-populaion raio naural log o median amily income)

wih he addiion o amily level characerisics averaged across he sae (average

amily size and he shares o populaion consiued by each o five racialehnicgroups) Sae fixed effecs are represened by As above our preerred model

specificaion includes year fixed effecs ha vary by Census division ( ) and

sae-level linear ime rends as elaboraed in Appendix B Te effec o

ineres is capured by

We esimae boh sae-level models (enrollmen and expendiures) using

ordinary leas squares regression Tus he inerpreaion o he coefficien is no

longer ha o a change in probabiliy as in he binary oucome models described

above Raher or he sae-level SNAP enrollmen model represens he

expeced change (in percenage poins) in he saersquos SNAP enrollmen rae ha

is due o a 1 percen change in he minimum wage For he SNAP expendiures

model is simply he elasiciy o SNAP spending wih respec o he minimum

wage991252ha is he percenage change in sae expendiures expeced o resul rom

a 1 percen change in ha saersquos minimum wage For urher deails on model

specificaion reer o Appendix B below

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1743

14 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1843

Results | wwwamericanprogresso

Results

Estimated minimum wage effects on SNAP enrollment and

expenditures

able 1 shows he esimaed parameer o ineres991252he coefficien o he

minimum wage or he preerred model o each ype Coefficiens on he

minimum wage variable are no direcly comparable across models because all

our models have a differen uncional orm o undersand and compare hese

esimaes we compue he change in SNAP aciviy prediced or a paricular wage scenario Te final column in able 1 answers he quesion Wha would be

he expeced change in SNAP aciviy in response o a 10 percen increase in he

minimum wage Te answer o his quesion varies wih he value o he inpu

parameers in he able we calculae he percenage decrease in enrollmen or

expendiures prediced or he average sae wih a minimum wage o $725 in

2014 Te sae-level SNAP expendiure model which is a consan-elasiciy

model conveys elasiciy inormaion direcly or he change in expendiures per

capia in he sae

TABLE 1

Comparison of national SNAP predictions for a 10 percent increase in the federal minimum wage

Model LevelRegression

type

Predicted outcome Coefficient of log

(minimum wage)

(Standard error)

Effect of a 10 percent in

in the minimum wage

VariableForm of

variable

Total

enrollment

Tot

expend

1 Family Linear

probability Enrollment Binary (enrolled=1)

-0042

(0008) -317 N

2

StateLinear regression

(ordinary least

squares)

Enrollment State enrollment rate (percent) -0031

(0012) -235 N

3 Expenditures Log (state expenditures per capita) -0190

(0103) NA -19

plt01 plt005 plt001Note Predicted changes are calculated for the average state with a minimum wage of $725 in 2014

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1943

16 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

According o his model a 10 percen increase in he minimum wage would resul

in a 19 percen reducion in SNAP expendiures Tis resul is reassuringly similar

o Dubersquos resul or povery reducion Te wo enrollmen models are much more

precisely esimaed han he expendiure model Te sae-level enrollmen model

finds ha a 10 percen minimum wage increase in a low-wage sae is associaed wih

a 235 percen decrease in SNAP enrollmens Te amily-level linear probabiliymodel predics a somewha greaer elasiciy or low minimum wage saes an

increase o 10 percen in he ederal minimum would resul in a 317 percen decline

in SNAP enrollmen24 Te differences in hese esimaes sem rom a number o

acors including difference in model uncional orm and daa used We rea his

range o elasiciy esimaes as an upper and lower bound on enrollmen impacs

Harkin-Miller bill National and state-level predicted impacts

Wha would be he prediced change or he SNAP program i he ederalminimum were raised o $1010 as proposed in he Harkin-Miller bill In order o

make his inerence we accoun or he ac ha no all saes are currenly subjec

o he ederal minimum wage a he beginning o 2014 21 saes mainained

higher minimum wages han $725 In hose saes an increase in he ederal

minimum wage may or may no be binding or employers in he sae depending

upon wheher he new ederal minimum exceeds he sae-level minimum Bu

regardless o wheher a minimum wage change is binding he impac on SNAP

aciviy will be lower in high minimum wage saes In order o accoun or his

properly we calculae sae by sae he percenage wage change ha would resul

rom he Harkin-Miller proposal and apply he parameers rom each o he hree

models above o compue he expeced decrease in SNAP aciviy or each sae

In his exercise we use saesrsquo curren (2014) minimum wage levels and assume

as a baseline he 2012 levels o SNAP enrollmen and expendiure as 2012 is he

mos recen year or which SNAP daa are available

able 10 and able 11 in Appendix C repor he esimaed effecs on SNAP

enrollmen and expendiures respecively or each sae under he Harkin-Miller

bill25 An increase o $1010 i enaced oday would represen beween a 393

percen wage increase in a $725 minimum wage sae and an 84 percen increasein Washingon sae which has he highes minimum wage in he naion a $932

as o January 201426 Slighly more han 56 percen o he decrease in expendiures

and abou 59 percen o he decrease in enrollmen would occur in saes wih

presen-day minimum wages o $725

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2043

Results | wwwamericanprogresso

In 2012 hese saes were home o 46 percen o he American populaion

and accouned or a roughly equivalen percenage o oal naional SNAP

expendiures No surprisingly he larges enrollmen decreases come rom

populous saes wih very high SNAP enrollmen raes andor very low minimum

wages Te larges prediced enrollmen reducion991252beween 319000 individuals

and 362000 individuals991252would occur in exas which has a $725 minimum wage In Caliornia which has a 206 percen SNAP paricipaion rae and an $8

per hour minimum wage we anicipae SNAP enrollmen declines o 310000

persons o 371000 persons And in Florida which had a SNAP paricipaion rae

o 166 percen and a $793 minimum wage enrollmen could decline by beween

164000 individuals and 196000 individuals For he our saes ha ogeher

accouned or he greaes amoun o SNAP spending in 2012991252exas Caliornia

Florida and New York respecively991252he combined expendiure reducion rom

he Harkin-Miller bill is prediced o be $14 billion

able 2 summarizes he prediced declines in SNAP aciviy or he naion as a whole ha would resul rom he direc and indirec effecs o he Harkin-Miller

bill Enrollmen would all beween 31 million persons and 36 million persons

represening 75 percen o 87 percen o curren enrollmen Te anicipaed

reducion in program expendiures would be nearly $46 billion or 61 percen o

program expendiures

TABLE 2

Comparison of national SNAP predictions under the Harkin-Miller billrsquos $1010 minimum wage

Model

Enrollment

(persons)

Expenditures

(millions of dollars)

Current (2012) Predicted Change Current (2012) Predicted Chan

Family enrollment (linear probability)

41866195

45489339 -3623144

$74861

NA NA

State enrollment (ordinary least squares) 38745435 -3120759 NA NA

State expenditures (ordinary least squares) NA NA $70305 -$45

Note Calculations use 2014 state minimum wages and the most recent SNAP data from 2012 They assume that per-enrollee expenditures remain constant

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2143

18 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

Tere are o course oher possibiliies or minimum wage legislaion able 3

shows he expeced SNAP changes or he Unied Saes under a variey o wage

scenarios calculaed using he sae-level models I saes were no able o se

heir minimum wages independenly such ha all saes were consrained by

he ederal minimum o $725 SNAP would be received by abou 514000 more

people across he Unied Saes a an addiional program cos o nearly hree-quarers o a billion dollars In conras he effecs o a higher minimum wage

proposal991252a ederal wage floor o $11 per hour991252would decrease enrollmen in

SNAP by more han 10 percen and decrease program coss by 83 percen

TABLE 3

Summary of par ticipation and expenditures under wage scenarios

If all states had

minimum wages of

Enrollment(persons)

Expenditures

(millions of dollars)

Predicted Change Predicted Change

Recent levels (2014) 41866195 $74861

$725 42380520 514326 $75604 $743

$800 41423919 -442276 $74209 -$652

$900 40148451 -1717744 $72350 -$2511

$1000 38872982 -2993212 $70490 -$4371

$1010 38745435 -3120759 $70305 -$4556

$1100 37597514 -4268681 $68631 -$6230

Note Calculations use state-level enrollment model coefficient

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2243

Conclusion | wwwamericanprogresso

Conclusion

An exensive body o lieraure examines employmen effecs o he minimum

wage A much smaller se o sudies examines how he minimum wage affecs

povery and only a handul o papers examine he effecs o he minimum wage

on he EIC Our analysis is he firs o examine he effecs o he minimum

wage on SNAP

Our findings indicae ha increased earnings rom minimum wage changes do

reduce SNAP enrollmens and expendiures We esimae ha he Harkin-Miller bill would save axpayers nearly $46 billion per year equivalen o 61 percen

o SNAP expendiures in 2012 he las year or which daa are available Over a

10-year period he esimaed savings amoun o nearly $46 billion

Our repor is subjec o limiaions ha we expec o overcome in our uure

research Firs he findings do no ake ino accoun possible ineracions among

SNAP he EIC and Medicaid Te eligibiliy cuoffs among hese programs

are quie differen suggesing ha such ineracions may be minor Noneheless

he join effecs can only be deermined by urher research using a causal

model Second i would be useul o know he disribuion o SNAP reducions

along he wage disribuion Using he Congressional Budge Officersquos calculaions

o how much he oal dollar value o wage would increase under he Harkin-

Miller proposal our findings imply ha he decline in overall SNAP spending

equals abou 15 percen o he oal resuling increase in wages Te amoun and

disribuion o his offse are o considerable ineres Minimum wage beneficiaries

who come rom working amilies already well above he povery line would no

see any offse while hose who are currenly considerably below he povery line

will see larger offses Tese issues will also be a subjec or our uure research

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2343

20 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

About the authors

Rachel West is a maser o public policy candidae a he Goldman School

o Public Policy Universiy o Caliornia Berkeley Her research ocuses on

economic policy in he areas o low-wage labor and povery

Michael Reich is proessor o economics and direcor o he Insiue or

Research on Labor and Employmen a he Universiy o Caliornia a Berkeley

His research publicaions cover numerous areas o labor economics including

racial inequaliy labor marke segmenaion high-perormance workplaces

union-managemen cooperaion Japanese labor-managemen sysems living

wages and minimum wages He received his docorae in economics rom

Harvard Universiy

Acknowledgments

We are graeul o Sylvia Allegreto Arindraji Dube Bill Leser Jesse Rohsein

Daniel Tompson and Ben Zipperer or heir valuable suggesions

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2443

References | wwwamericanprogresso

References

Allegreto Sylvia and ohers 2013 ldquoFas Food Povery Wages Te Public Cos o Low-Wage Jobsin he Fas-Food Indusryrdquo Universiy o Caliornia Berkeley Cener or Labor Research andEducaion Available a htplaborcenerberkeleyedupubliccossas_ood_povery_wage

Allegreto Sylvia and ohers 2013 ldquoCredible Research Designs or Minimum Wage Sudiesrdquo

Working Paper 148-13 Universiy o Caliornia Berkeley Insiue or Research on Labor andEmploymen Available a htpwwwirleberkeleyeduworkingpapers148-13pd

Biler Marianne and Hilary Hoynes 2013 ldquo Te More Tings Change he More Tey Say heSame Te Saey Ne Living Arrangemens and Povery in he Grea Recessionrdquo Working Paper19449 Naional Bureau o Economic Research

Congressional Budge Office 2012 ldquoTe Supplemenal Nuriion Assisance Programrdquo Washingon Available a htpwwwcbogovsiesdeaulfilescbofilesatachmens04-19-SNAPpd

991252 991252 991252 2014 ldquoTe Effec o a Minimum-Wage Increase on Employmen and Family Incomerdquo Washingon Available a htpwwwcbogovsiesdeaulfilescbofilesatachmens44995-MinimumWagepd

Dube Arindraji 2013 rdquoMinimum Wages and he Disribuion o Family Incomerdquo Unpublished working paper Available a htpsdldropboxuserconencomu15038936Dube_ MinimumWagesFamilyIncomespd

991252 991252 991252 2014 ldquoTe Povery o Minimum Wage Facsrdquo Arindraji Dube Blog January 22 Availablea htparindubecom20140122he-povery-o-minimum-wage-acs

Lee David and Emmanuel Saez 2012 ldquoOpimal Minimum Wage Policy in Compeiive LaborMarkesrdquo Journal o Public Economics 96 (9) 739ndash749

Neumark David and William Wascher 1992 ldquoEmploymen Effecs o Minimum and Subminimum Wages Panel Daa on Sae Minimum Wage Lawsrdquo Industrial and Labor Relations Review 46 (1)

55ndash81

Neumark David and William Wascher 2011 ldquoDoes a Higher Minimum Wage Enhance heEffeciveness o he Earned Income ax Credirdquo Industrial and Labor Relations Review 64 (4)712ndash746

Page Marianne Joanne Spez and Jane Millar 2005 ldquoDoes he Minimum Wage Affec WelareCaseloadsrdquo Journal o Policy Analysis and Management 24 (2) 273ndash295

Rohsein Jesse 2010 ldquoIs he EIC as Good as an NI Condiional Cash ransers and ax

Incidencerdquo American Economic Journal Economic Policy 2 (1) 177ndash208

Wage and Hours Division 2014 ldquoMinimum Wage Laws in he Saes ndash January 1 2014rdquo USDeparmen o Labor (htpwwwdolgovwhdminwageamericahm [February 2014])

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2543

22 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2643

Appendix A | wwwamericanprogresso

Appendix A

Pre-trend falsification check

Recen minimum wage research27 highlighs a common flaw in previous sudies

ailure o veriy ha he oucome variable is ree o negaive pre-exising rends I

or example SNAP aciviy was already rending down in saes ha raised heir

minimum wages beore hese changes came ino effec our regression analysis could

(misakenly) atribue ha reducion o he minimum wage We check or such

pre-rends by inroducing variables ha represen he prior yearrsquos value or leads o

he minimum wage I he model esimaes he minimum wage o have an effec on

he oucome variable beore he wage change wen ino effec hen an unobservedacor no he minimum wage change caused he change in SNAP aciviy

We es he specificaions above or pre-rends by including a one-year lead in

all hree specificaions We find ha he lead erms are small posiive and no

saisically significan indicaing ha he concurren minimum wage991252no

he wage level in prior periods991252is driving he observed changes in SNAP

oucomes28 In paricular he coefficien (sandard error) on he lead erm in

our preerred amily-level enrollmen regression is 011 and no significan

while he coefficien and sandard error o he conemporaneous minimum

wage is unchanged In he sae-level preerred enrollmen regression he

coefficien o he lead erm is again small (07) and i is no significan Te

corresponding coefficien on he lead erm in he sae-level expendiure

regression is 16 and is no significan Te posiive poin esimaes on hese lead

erms resuls no only rule ou disoring negaive pre-rends Tey also sugges

ha our main resuls may underesimae he rue effecs

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2743

24 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2843

Appendix B | wwwamericanprogresso

Appendix B

Model selection process

For boh he amily-level and sae-level models we es hree mehods o conrol

or unobserved geographic- and ime-varying characerisics as suggesed by he

minimum wage lieraure Firs we include only independen sae-specific fixed

effecs and year-specific fixed effecs Tis specificaion (specificaion 1) implicily

assumes ha amilies in any sae consiue an equally good saisical ldquoconrolrdquo

group or hose in any randomly chosen sae afer accouning or various

characerisics (median income and unemploymen rae among ohers) Similarly

simple ime fixed effecs assume ha amilies surveyed in any year can crediblyserve as a conrol group or amilies surveyed in every oher year o he sample

(1990 hrough 2012)

In oher words specificaion 1 assumes ha a saersquos immediae neighbor provides

no beter a couneracual or he effec o a minimum wage change han does a

sae across he counry We relax his resricive specificaion sequenially in wo

seps In specificaion 2 we replace simple year fixed effecs wih fixed effecs or

each Census divisionyear (capured as an addiional variable in he vecor By

using division-year effecs we remove he resricion ha amilies in each sae

are equally good saisical conrols or all oher amilies Raher we allow or he

possibiliy ha amilies in similar geographic regions (or example he Souh or

he Norheas) may be more similar o one anoher han amilies arher away

Finally in specificaion 3 we add sae-specific linear ime rends o he previous

specificaion Tus specificaion 3 is he mos rigorous model specificaion in ha

i allows or heerogeneiy along hree dimensions Ta is specificaion 3 allows

each sae o have is own ime-varying rends raher han imposing he resricion

ha saes evolve idenically over he 22 years in he sample

We begin building he heoreical specificaion above rom a se o simpleuncondiional models regression o SNAP aciviy (enrollmen or expendiures)

on he log o he minimum wage and a se o geographic- and ime-specific

effecs (specificaions 1 2 and 3 described above) As shown in ables 1ndash3 (or

specificaion 3) we hen add covariaes sequenially o hese models including

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2943

26 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

firs he vecor o amily-level conrols ollowed by each o several sae-level

covariaes in urn he unemploymen rae log o median-amily income and he

employmen-o-populaion raio Comparable resuls or specificaions 2 and 3

will be available in our orhcoming working paper

As expeced he simple uncondiional models indicae ha he relaionship beween he minimum wage and SNAP enrollmen i one exiss is a more complex

one influenced by oher acors In he uncondiional model he coefficien on he

variable o ineres991252he log o he minimum wage991252is small in magniude and no

saisically differen rom zero Once we accoun or he influence o labor marke

condiions and variaion in income levels on program paricipaion (by including

unemploymen rae and median-amily income conrol variables respecively)

he effec o he minimum wage on SNAP enrollmen is precisely esimaed Te

coefficien o he log minimum wage is slighly higher (-0042) in he amily-level

analysis han he coefficien (-031) in he sae-level analysis Te level o precision

is also higher in he amily-level analysis Tis is o be expeced when using 124million observaions compared o 1127

Te second dimension o model choice concerns he effec specificaion ables

7ndash9 compare he primary coefficiens o ineres or he SNAP enrollmen and

expendiure models For boh he enrollmen models he effec sizes are smalles

or specificaion 1 larges or specificaion 2 and inermediae beween hese wo

in specificaion 3 Recall ha Specificaion 3 conains sae-specific linear ime

rends in addiion o he census divisionyear conrols included in specificaion

2 In he amily-level enrollmen model he sandard error o he minimum wage

coefficien is smaller han in he oher wo specificaions Sandard errors on he

oher variables are much smaller in specificaions 2 and 3 han in specificaion

1 On he basis o coefficien significance (join and individual) specificaions 2

and 3 are sricly preerred in boh enrollmen models o specificaion 1 which

conains only sae and year fixed effecs

A concern wih specificaions 2 and 3 is ha rend conrols such as sae linear

rends may incorrecly absorb some o he delayed impac o a minimum wage

When we es his issue by including lagged minimum wages we do no find ha

delayed effecs are significan Anoher concern is ha more sauraed modelsuse less o he saisical variaion which could reduce he saisical power o

he resuls However he sandard errors or our more sauraed models are no

higher and are lower in some cases han or he less sauraed models Overall

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3043

Appendix B | wwwamericanprogresso

his evidence suppors our use o he sauraed model as he preerred model

specificaion Moreover Dubersquos 2013 sudy shows ha more sauraed models

perorm beter han models wih jus sae and ime fixed effecs

Te esimaed enrollmen regressions a boh he amily and sae levels show large

and saisically significan coefficiens Te esimaed minimum wage effec in heexpendiures regressions991252or which we have only sae-level daa991252is also large

and saisically significan

We do no use weighed regression or he sae-level models preerring o keep

analysis o he ldquoreamenrdquo (ha is o say a minimum wage change) appropriae

o he average sae raher han he average amily or individual I insead our

primary ineres were he impac o a minimum wage change on he average amily

or he average individual we migh choose o designae he number o amilies

in each sae or he sae populaion respecively as analyic weighs in order o

obain a coefficien beter suied or such inerence

TABLE 4

SNAP enrollment

Family-level linear probability

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0048 -0047 -0040 -0043 -0042

(0013) (0013) (001) (0008) (0009)

Unemployment rate (100) 0505 0420 0280

(0083) (0086) (0082)

Log median income -0057 -0039

(0011) (0011)

Employment-to-population ratio -0239

(0038)

N 1242022 1242022 1242022 1242022 1242022

plt01 plt005 plt001

Note Robust standard errors are in parentheses Observations are clustered at the state level The outcome variable is binary and equal to one if a family is enrolledin SNAP All models include state fixed effects Census division x-year fixed effects and state-specific linear time trends All specifications except 3a include additional

controls for family size race and marital status of the family head presence of children and presence of an adult maleSource Annual data from the March Current Population Survey for 1990 to 2012 estimation includes Current Population Survey probability weights

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3143

28 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

TABLE 5

SNAP enrollment

State-level linear regression

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0028 -0024 -0026 -0031 -0031

(0014) (0013) (0013) (0012) (0012)

Unemployment rate (100) 0625 0466 0320 0339

(0087) (0088) (0085) (0083)

Log median income -0090 -0065 -0061

(0013) (0013) (0013)

Employment-to-population ratio -0282 -0248

(0037) (0038)

N 1127 1127 1127 1127 1127

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the SNAP enrollment rate All models include state fixed effects Census division x-year fixedeffects and state-specific linear time trends

Source Annual data from the March Current Population Survey for 1990 to 2012

TABLE 6

SNAP expenditures

State-level linear regression

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0181 -0149 -0156 -0153 -0190

(011) (0103) (0102) (0103) (0103)

Unemployment rate (100) 4587 4152 4232 4313

(0622) (0621) (0633) (0628)

Log median income -0246 -0261 -0294

(0075) (0078) (0078)

Employment-to-population ratio 0155 0244

(0237) (024)

N 1127 1127 1127 1127 1127

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the natural log of state SNAP expenditures per capita for 1990 to 2012 All models include state

fixed effects Census division x-year fixed effects and state-specific linear time trends

Source Annual data from Bureau of Economic Analysis national income and product accounts tables for 1990 to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3243

Appendix B | wwwamericanprogresso

TABLE 7

Comparison of specifications SNAP enrollment

Family-level linear probability

(1) (2) (3)

Log minimum wage -0028 -0049 -0042

(0014) (0017) (0009)

Unemployment rate (100) 0275 0297 0280

(0161) (0076) (0082)

Log median income -0077 -0055 -0039

(0014) (0012) (0011)

Employment-to-population ratio -0238 -0250 -0239

(0054) (004) (0038)

N 1242022 1242022 1242022

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses Observations are clustered at the state level The outcomevariable is binary or equal to one if a family is enrolled in SNAP All specifications include additional controls forfamily size race and marital status of the family head presence of children and presence of an adult male

Source Annual data from the March Current Population Survey for 1990 to 2012 estimation includes CurrentPopulation Survey probability weights

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3343

30 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

TABLE 8

Comparison of specifications SNAP enrollment

State-level linear regression

(1) (2) (3)

Log minimum wage -0019 -0035 -0031

(0009) (0012) (0012)

Unemployment rate (100) 0401 0370 0339

(0063) (0077) (0083)

Log median income -0081 -0073 -0061

(0011) (0013) (0013)

Employment-to-population ratio -0183 -0222 -0248

(0039) (0039) (0038)

N 1127 1127 1127

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the SNAP enrollment rate Allregressions include the share of households with children and the racial shares of the population

Source Annual data from the March Current Population Survey for 1990 to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3443

Appendix B | wwwamericanprogresso

TABLE 9

Comparison of specifications SNAP expenditures

State-level linear regression

(1) (2) (3)

Log minimum wage -0121 -0203 -0190

(0075) (0103) (0103)

Unemployment rate (100) 5292 5152 4313

(0464) (0576) (0628)

Log median income -0437 -0417 -0294

(008) (0086) (0078)

Employment-to-population ratio -0040 -0220 0244

(0261) (0260) (0240)

N 1127 1127 1127

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the natural log of state SNAPexpenditures per capita All regressions include the share of households with children and the racial shares ofthe state population

Source Annual data from Bureau of Economic Analysis national income and product accounts tables for 1990to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3543

32 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3643

Appendix C | wwwamericanprogresso

Appendix C Harkin-Miller

policy simulation results

StateMinimum wage

(2014)

SNAP enrollment (2012) Predicted change in SNAP enrollment

RateTotal

(persons)

Model 1 family

enrollment (linearprobability)

Model 2 state

enrollment (linearregression)

Model 3 s

expenditu(linear regre

Alabama $725 164 788682 -66922 -58897 -58906

Alaska $775 120 87436 -8104 -6887 -3288

Arizona $790 201 1319323 -67435 -56738 -64356

Arkansas $725 78 230489 -40977 -36063 -35248

California $800 206 7813680 -371131 -310222 -18223

Colorado $800 164 853155 -50684 -42365 -23926

Connecticut $870 91 326621 -22456 -17975 -13711

Delaware $725 186 170262 -12739 -11211 -10647

District of Columbia $825 133 84009 -5370 -4417 -3632

Florida $793 166 3208026 -195813 -164426 -13046

Georgia $725 160 1586336 -137741 -121224 -11004

Hawaii $725 96 133662 -19310 -16995 -14933

Idaho $725 92 147501 -22165 -19507 -15809

Illinois $825 95 1225084 -109088 -89742 -70955

Indiana $725 125 816233 -90818 -79928 -83985

Iowa $725 155 478011 -42716 -37594 -28556

Kansas $725 135 388269 -40082 -35275 -27461

Kentucky $725 130 568821 -60840 -53544 -52259

Louisiana $725 149 683832 -63929 -56263 -66083

Maine $750 77 101976 -16567 -14323 -15234

Maryland $725 144 846415 -81748 -71946 -38370

Massachusetts $800 130 864721 -64902 -54251 -42913

Michigan $740 146 1439141 -128801 -112140 -11022

Minnesota $725 133 713646 -74730 -65769 -37878

TABLE 10

SNAP enrollments Predicted changes by state under the Harkin-Miller bill

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3743

34 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

StateMinimum wage

(2014)

SNAP enrollment (2012) Predicted change in SNAP enrollment

RateTotal

(persons)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 state

expenditures

(linear regression)

Mississippi $725 129 386501 -41486 -36511 -46467

Missouri $750 172 1036182 -75131 -64952 -56244

Montana $790 132 132452 -10350 -8708 -5846

Nebraska $725 124 230683 -25773 -22683 -12189

Nevada $825 162 446035 -23349 -19209 -11894

New Hampshire $725 127 168404 -18359 -16157 -5735

New Jersey $825 160 1416666 -75175 -61843 -28236

New Mexico $750 149 310896 -25983 -22463 -22512

New York $800 192 3763553 -191193 -159815 -142182

North Carolina $725 174 1697193 -135417 -119179 -113503

North Dakota $725 87 61225 -9743 -8574 -4021

Ohio $795 143 1647345 -115869 -97169 -88580

Oklahoma $725 129 494053 -53006 -46650 -46854

Oregon $910 124 485326 -17036 -13328 -16398

Pennsylvania $725 161 2053643 -177315 -156052 -125586

Rhode Island $800 156 163730 -10258 -8574 -8698

South Carolina $725 94 445277 -65614 -57746 -50304

South Dakota $725 208 173749 -11586 -10197 -7458

Tennessee $725 142 914903 -89667 -78915 -99134

Texas $725 110 2863779 -362018 -318607 -253285

Utah $725 88 251107 -39658 -34902 -19390

Vermont $873 156 97792 -3823 -3055 -2475

Virginia $725 101 829771 -113723 -100086 -58212

Washington $932 72 496934 -23221 -17947 -17756

West Virginia $725 58 107875 -25792 -22699 -21665

Wisconsin $725 75 427822 -79521 -69986 -53210

Wyoming $725 164 94590 -8010 -7050 -3104

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3843

Appendix C | wwwamericanprogresso

StateMinimum wage

(2014)

SNAP expenditures (2012)Predicted change in SNAP expenditures

(millions of dollars)

Per resident

(millions of

dollars)

Total

(millions of

dollars)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 s

expenditu

(linear regre

Alabama $725 $290 $1397 -$1185 -$1043 -$1043

Alaska $775 $253 $185 -$262 -$223 -$106

Arizona $790 $257 $1686 -$935 -$787 -$892

Arkansas $725 $248 $733 -$636 -$560 -$547

California $800 $189 $7164 -$7276 -$6082 -$3573

Colorado $800 $157 $816 -$862 -$721 -$407

Connecticut $870 $191 $686 -$343 -$275 -$210

Delaware $725 $250 $229 -$205 -$180 -$171

District of Columbia $825 $366 $232 -$146 -$120 -$99

Florida $793 $294 $5676 -$4429 -$3719 -$2951

Georgia $725 $317 $3140 -$2936 -$2584 -$2346

Hawaii $725 $335 $465 -$449 -$395 -$347

Idaho $725 $225 $359 -$376 -$331 -$268

Illinois $825 $249 $3200 -$2096 -$1725 -$1364

Indiana $725 $220 $1439 -$1162 -$1023 -$1075

Iowa $725 $192 $589 -$658 -$579 -$440

Kansas $725 $159 $460 -$502 -$441 -$344

Kentucky $725 $298 $1303 -$1133 -$997 -$973

Louisiana $725 $315 $1450 -$1047 -$922 -$1083

Maine $750 $281 $373 -$267 -$231 -$246

Maryland $725 $188 $1109 -$1765 -$1553 -$828

Massachusetts $800 $206 $1366 -$1030 -$861 -$681

Michigan $740 $300 $2963 -$2400 -$2090 -$2054

Minnesota $725 $140 $755 -$1113 -$980 -$564

Mississippi $725 $326 $973 -$649 -$571 -$726

Missouri $750 $241 $1452 -$1278 -$1104 -$956

Montana $790 $190 $191 -$179 -$151 -$101

Nebraska $725 $140 $259 -$409 -$360 -$194

Nevada $825 $191 $527 -$441 -$363 -$225

New Hampshire $725 $126 $167 -$399 -$351 -$125

TABLE 11

SNAP expenditures Predicted changes by state under the Harkin-Miller bill

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3943

36 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

StateMinimum wage

(2014)

SNAP expenditures (2012)Predicted change in SNAP expenditures

(millions of dollars)

Per resident

(millions of

dollars)

Total

(millions of

dollars)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 state

expenditures

(linear regression)

New Jersey $825 $160 $1420 -$1611 -$1325 -$605

New Mexico $750 $324 $675 -$513 -$444 -$445New York $800 $287 $5616 -$3766 -$3148 -$2801

North Carolina $725 $252 $2454 -$2187 -$1925 -$1833

North Dakota $725 $128 $90 -$162 -$143 -$67

Ohio $795 $259 $2995 -$2013 -$1688 -$1539

Oklahoma $725 $248 $945 -$799 -$703 -$706

Oregon $910 $322 $1255 -$272 -$213 -$262

Pennsylvania $725 $218 $2779 -$2930 -$2579 -$2075

Rhode Island $800 $280 $294 -$173 -$144 -$147

South Carolina $725 $291 $1373 -$1337 -$1177 -$1025South Dakota $725 $198 $165 -$192 -$169 -$123

Tennessee $725 $324 $2091 -$1413 -$1243 -$1562

Texas $725 $230 $5997 -$6402 -$5634 -$4479

Utah $725 $141 $402 -$614 -$541 -$300

Vermont $873 $230 $144 -$66 -$53 -$43

Virginia $725 $173 $1413 -$2062 -$1815 -$1056

Washington $932 $244 $1682 -$350 -$270 -$267

West Virginia $725 $273 $508 -$451 -$397 -$379

Wisconsin $725 $204 $1166 -$1302 -$1146 -$871Wyoming $725 $95 $55 -$105 -$93 -$41

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4043

Appendix C | wwwamericanprogresso

FIGURE 2

Raising the minimum wage to $1010would cut taxpayer costs in every state

Predicted decreases in cost and enrollment

in SNAP in 50 states

$200+$51ndash$100

$101ndash$200

0ndash$25

$26ndash$50

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4143

38 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

Endnotes

1 US Department of Agriculture ldquoSupplementalNutrition Assistance Program (SNAP)rdquo available athttpwwwfnsusdagovsnapeligibility (last accessedFebruary 2014)

2 For this initial analysis we do not consider Harkin-Millerrsquos increase in subminimum wages for tippedworkers To do s o would increase the estimated SNAP

savings by an unknown amount

3 The Congressional Budget Office estimates thatworkers currently earning between $1010 and $1150per hour would see their wages rise under the Harkin-Miller proposal Congressional Budget O ffice ldquoTheEffects of a Minimum Wage Increase on Employmentand Family Incomerdquo (2014)

4 Marianne Page Joanne Spetz and Jane Millar ldquoDoesthe Minimum Wage Affect Welfare Caseloadsrdquo Journalof Policy Analysis and Management 24 (2) (2005)273ndash295

5 Marianne Bitler and Hilary Hoynes ldquoThe More ThingsChange the More They Stay the Same The SafetyNet Living Arrangements and Poverty in the GreatRecessionrdquo NBER Working Paper 194 49 2013

6 Sylvia Allegretto and others ldquoFast Food Poverty Wages The Public Cost of Low-Wage Jobs in the Fast-FoodIndustryrdquo (Berkeley California Center for LaborResearch and Education 2013) available at httplaborcenterberkeleyedupubliccostsfast_food_poverty_wage

7 David Neumark and William Wascher ldquoDoes a HigherMinimum Wage Enhance the Effectiveness of theEarned Income Tax Creditrdquo Industrial and LaborRelations Review 64 (4) (2011) 712ndash746

8 David Lee and Emmanuel Saez ldquoOptimal MinimumWage Policy in Competitive Labor Marketsrdquo Journal ofPublic Economics 96 (9) (2012) 739ndash749

9 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family IncomerdquoFebruary 2014

10 Jesse Rothstein ldquoIs the EITC as Good as an NITConditional Cash Transfers and Tax Incidencerdquo AmericanEconomic Journal Economic Policy 2 (1) (2010) 177ndash208

11 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family Incomerdquo

12 Dube Arindrajit 2013 rdquoMinimum Wagesand the Distribution of Family IncomerdquoUnpublished working paper Available at httpsdldropboxusercontentcomu15038936Dube_MinimumWagesFamilyIncomespdf

13 As Dube explains in ldquoThe poverty of Minimum WageFactsrdquo the simulation approach underestimate stemsfrom a number of unwarranted assumptions includingthe range of actual wage increases and the accuracy ofwage data in the Current Population Survey The causal

approach does not make these assumptions

14 Allegretto Sylvia and others 2013 ldquoCredible ResearchDesigns for Minimum Wage Studiesrdquo Working Paper148-13 University of California Berkeley Institutefor Research on Labor and Employment Available athttpwwwirleberkeleyeduworkingpapers148-13pdf

15 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family In comerdquo

16 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo (2012)

17 Ibid

18 Ibid

19 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo

20 US Department of Agriculture ldquoSupplementalNutrition Assistance Program (SNAP)rdquo

21 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo

22 Ibid

23 We deviate from the Census Bureaursquos definition ofa family unit which is ldquotwo people or more (on e ofwhom is the householder) related by bir th marriageor adoption and residing togetherrdquo US Bureau ofthe Census ldquoCurrent Population Survey DefinitionsFamilyrdquo available at httpwwwcensusgovcpsabout

cpsdefhtml (last accessed February 2014) We countas a family unit any individual residing on his or herown two or more persons residing together whodo not belong to a family in the March CPS sampleare constructed as one family in our analysis For thepurposes of food stamp allocations the consumptionresulting from this transfer is probably distributed tofamily members (rather than household members ora single individual within the household) Howeversingle individuals canmdashand domdashreceive SNAPbenefits Excluding them would fail to make theanalysis reflective of the population at large

24 Strictly the family level linear probability modelpredicts the percentage-point decrease in theprobability that an individual family will receive SNAPpayments When applied to a large number of familieshowever we are able to interpret the coefficient asa decrease in the mean of enrollmentmdashthat is a

decrease in the enrollment ratemdashby applying the lawof iterated expectations

25 We generate expenditure predictions from theenrollment modelsmdashand conversely generateenrollment predictions from the expenditure modelmdashby assuming that expenditures per enrolled familyremains the same before and after the minimum wagechange In practice this is likely to be a conservativeestimatemdashthat is to underestimate the decrease inSNAP activity Average SNAP benefits per family willalso decrease as many families that remain eligible forSNAP experience income gains

26 Wage and Hour Division ldquoMinimum Wage Laws inthe States ndash Januar y 1 2014rdquo available at httpwwwdolgovwhdminwageamericahtm (last accessedFebruary 2014)

27 See for example Sylvia Allegretto and others ldquoCredibleResearch Designs for Minimum Wage Studiesrdquo WorkingPaper 148-113 (Berkeley California Institute forResearch on Labor and Employment 2013) available athttpwwwirleberkeleyeduworkingpapers148-13pdf

28 We will report these results in a forthcoming workingpaper

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4243

Endnotes | wwwamericanprogresso

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4343

The Center for American Progress is a nonpartisan research and educational institute

dedicated to promoting a strong just and free America that ensures opportunity

for all We believe that Americans are bound together by a common commitment to

these values and we aspire to ensure that our national policies reflect these values

We work to find progressive and pragmatic solutions to significant domestic and

international problems and develop policy proposals that foster a government that

is ldquoof the people by the people and for the peoplerdquo

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1143

8 Center for American Progress | The Effects of M inim um Wages on SNAP Enro llments and Expenditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1243

Methods and data | wwwamericanprogress

Methods and data

As previously menioned we exploi variaion in minimum wages by sae and ime

o examine heir causal effecs on SNAP enrollmens and expendiures o do so we

merge daa rom 1990 hrough 2012 drawn rom he Annual Social and Economic

Supplemen o he Curren Populaion Survey991252an annual Census Bureau survey

commonly known as he March CPS ha includes23 inormaion on SNAP

enrollmens a he amily level991252wih sae-level daa on minimum wages SNAP

expendiures populaion unemploymen raes and sae median income levels

o conrol or ime-varying heerogeneiy among saes our specificaions includeconrols or sae linear rends and effecs by Census division and ime We esimae

effecs a wo levels allowing or amily variaion and allowing only or sae-level

variaion We also employ a se o sandard demographic conrols such as amily size

and composiion and race and ehnic composiion

Distinguishing causation from correlation

How can we ensure ha our analysis does no pick up a spurious correlaion or

example he endency o more economically vibran saes o implemen higher

minimum wages Disinguishing correlaion and policy endogeneiy rom rue

causal effecs is he primary moivaion or economeric analysis In he ideal

experimen researchers would begin wih wo saes991252ha are alike in every

respec prior o he policy991252 and ldquoreardquo only one o hese saes wih a higher

minimum wage Tey would atemp o shield hese saes rom any influence ha

could obscure heir undersanding o he minimum wagersquos direc effec on SNAP

enrollmen Researchers o course canno conduc such experimens

We can however use saisical mehods o conrol simulaneously heindependen effecs on SNAP o sae unemploymen raes sae income levels

and common rajecories among saes wihin he same Census division By

ensuring similariy along hese dimensions we maximize he likelihood ha

SNAP aciviy in wo saes would have comparable oucomes in he absence

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1343

10 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

o a minimum wage policy change Tus i a new minimum wage policy were

implemened in one sae only he researchers could atribue all o he difference

hey observe in SNAP aciviy o he new minimum wage policy

In oher words we approximae he ideal experimen by using non-experimenal

saisical mehods Te desirable ldquopre-exising similariiesrdquo beween saes ha wehave defined above inorm our choice o conrol variables in a saisical seting

More precisely in our muliple regression models we use median amily income

he unemploymen rae he employmen-o-populaion raio and regional and

ime idenifiers o consruc an appropriae group o peers or each sae on he

eve o a policy change

Data description

wo daa ses include inormaion abou boh income and paricipaion inpublic programs Te Survey o Income and Program Paricipaion or SIPP

which is conduced in inermiten years has he advanage o ollowing he

same individuals over a period o ime In oher words i is a longiudinal daa

se I also has he advanage o conaining monhly daa However he sample

size o he SIPP is no sufficien or analyzing variaions in sae-level minimum

wages Te March CPS has he advanage o a much larger sample size and i is

conduced annually wihou any breaks in ime I has he disadvanage o being

a cross-secional daa se so we canno ollow he same individuals over ime991252

sricly speaking over more han one year On ne he March CPS is much more

suiable or our sudy We examine he empirical relaionship beween minimum

wage policy and ood samp aciviy a wo levels o aggregaion he amily level

and he sae level Family-level daa are drawn rom he March CPS

Te March CPS comprises responses rom he residens o 50000 o 60000

dwelling places surveyed per year and conains deailed inormaion on he

residensrsquo employmen and income including income rom ranser paymens

Te sample or our analysis comprises more han 128 million amily unis during

he period rom 1990 o 2012 (inclusive) Survey weighs allow us o analyze

SNAP paricipaion in a manner ha is represenaive o he US populaion alarge Over all years he share o amilies reporing ood samp receip in he

weighed March CPS sample is 91 percen Te enrollmen rae was a a low o 6

percen in he year 2000 In 2012 he mos recen year in our panel 133 percen

o amilies repored paricipaing in SNAP a some poin during he survey year

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1443

Methods and data | wwwamericanprogresso

Te March CPS also collecs inormaion on he number o SNAP recipiens in

he household he number o monhs he household receives SNAP benefis

and he cash-equivalen value o he SNAP benefis received However he

value o SNAP benefis is severely underrepored among recipiens perhaps

because recipiens are unaware o he exac monhly cash-equivalen value o

benefis hey receive

Our firs empirical sraegy ocuses solely on SNAP enrollmen By using he

amily as he uni o analysis we are able o inser saisical conrols o accoun

or non-wage-relaed acors ha influence any paricular amilyrsquos likelihood o

program paricipaion wih he inenion o isolaing any differences in program

paricipaion ha are due purely o changes in wage policy Tis approach

idenifies he effecs o low-wage labor policy on he exernal margin991252ha is

he effec o he minimum wage on he likelihood ha a amily paricipaes in he

SNAP program a all991252as opposed o he inernal margin or how much SNAP

unding he amily would receive

Our second empirical ramework uses sae-level adminisraive daa Ta is

we aggregae he daa o obain a single daa poin or each saeyear back

o 1990 represening he mean o he oucome or he sae Te sae-level

esimaion serves as a robusness check on he amily-level resuls or SNAP

paricipaion Also using aggregaed daa allows us o esimae direcly he

causal effec o minimum wage changes on SNAP spending Tis is no possible

a he amily level as discussed above daa on cash-equivalen value o ood

samps or SNAP recipiens is very requenly no repored in he March CPS

and when i is repored he inormaion may be unreliable By conras he

Bureau o Economic Analysis publishes aggregae SNAP spending a he sae

level in is Naional Income and Produc Accoun or NIPA ables Tus

while we are unable o observe he heerogeneiy in he cash value o SNAP

or amilies in each sae we are able o calculae average SNAP spending

per residen in each sae per year Supporing covariaes include he annual

unemploymen and employmen daa rom he Bureau o Labor Saisics or

BLS and sae-level populaion series rom he iner-decennial census releases

Minimum wage daa are available rom he BLSrsquos wages and hours division For

sae minimum wage changes enaced a oher imes han he firs o he yearan average value or he year is used

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1543

12 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

Family level model specifications

We firs examine he effec o he minimum wage on paricipaion in public

assisance programs For amily residing in sae and during year we esimae

an equaion o he ollowing orm

(1)

is a binary variable ha is se equal o 1 i a leas one member o amily

received ood samps during he survey year is a se o sae-level

characerisics including annual averages o he unemploymen rae he

employmen-o-populaion raio and he naural log o median amily income

is a vecor o amily atribues including indicaors or he race and marial saus

o he amily head size o he amily he presence o children and he presence oan adul male Sae fixed effecs are capured by o conrol or ime-varying

heerogeneiy our preerred model specificaion also includes year fixed effecs

ha vary by Census division ( ) and sae-level linear ime rends In

Appendix B we jusiy he inclusion o hese las wo erms We also compare he

resuls rom our preerred specificaion wih less sauraed specificaions

Te effec o ineres which is capured by is he expeced change in he

probabiliy o receiving SNAP benefis wih respec o a change in he (log o he)

binding minimum wage in sae during year We repor robus sandard errors

clusered a he sae level We esimae he parameers using linear regression

producing a linear probabiliy model Deails o he model selecion process are

covered in Appendix B below

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1643

Methods and data | wwwamericanprogresso

State-level model specifications

Te sae-level models are similarly specified For sae in year we assume ha

(2)

In his model is now eiher he SNAP enrollmen rae in sae during year

or he naural logarihm o per capia SNAP expendiures in sae during

year is once again a se o sae-level characerisics including he same

sae-level covariaes as in he amily regressions (annual average unemploymen

rae employmen-o-populaion raio naural log o median amily income)

wih he addiion o amily level characerisics averaged across he sae (average

amily size and he shares o populaion consiued by each o five racialehnicgroups) Sae fixed effecs are represened by As above our preerred model

specificaion includes year fixed effecs ha vary by Census division ( ) and

sae-level linear ime rends as elaboraed in Appendix B Te effec o

ineres is capured by

We esimae boh sae-level models (enrollmen and expendiures) using

ordinary leas squares regression Tus he inerpreaion o he coefficien is no

longer ha o a change in probabiliy as in he binary oucome models described

above Raher or he sae-level SNAP enrollmen model represens he

expeced change (in percenage poins) in he saersquos SNAP enrollmen rae ha

is due o a 1 percen change in he minimum wage For he SNAP expendiures

model is simply he elasiciy o SNAP spending wih respec o he minimum

wage991252ha is he percenage change in sae expendiures expeced o resul rom

a 1 percen change in ha saersquos minimum wage For urher deails on model

specificaion reer o Appendix B below

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1743

14 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1843

Results | wwwamericanprogresso

Results

Estimated minimum wage effects on SNAP enrollment and

expenditures

able 1 shows he esimaed parameer o ineres991252he coefficien o he

minimum wage or he preerred model o each ype Coefficiens on he

minimum wage variable are no direcly comparable across models because all

our models have a differen uncional orm o undersand and compare hese

esimaes we compue he change in SNAP aciviy prediced or a paricular wage scenario Te final column in able 1 answers he quesion Wha would be

he expeced change in SNAP aciviy in response o a 10 percen increase in he

minimum wage Te answer o his quesion varies wih he value o he inpu

parameers in he able we calculae he percenage decrease in enrollmen or

expendiures prediced or he average sae wih a minimum wage o $725 in

2014 Te sae-level SNAP expendiure model which is a consan-elasiciy

model conveys elasiciy inormaion direcly or he change in expendiures per

capia in he sae

TABLE 1

Comparison of national SNAP predictions for a 10 percent increase in the federal minimum wage

Model LevelRegression

type

Predicted outcome Coefficient of log

(minimum wage)

(Standard error)

Effect of a 10 percent in

in the minimum wage

VariableForm of

variable

Total

enrollment

Tot

expend

1 Family Linear

probability Enrollment Binary (enrolled=1)

-0042

(0008) -317 N

2

StateLinear regression

(ordinary least

squares)

Enrollment State enrollment rate (percent) -0031

(0012) -235 N

3 Expenditures Log (state expenditures per capita) -0190

(0103) NA -19

plt01 plt005 plt001Note Predicted changes are calculated for the average state with a minimum wage of $725 in 2014

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1943

16 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

According o his model a 10 percen increase in he minimum wage would resul

in a 19 percen reducion in SNAP expendiures Tis resul is reassuringly similar

o Dubersquos resul or povery reducion Te wo enrollmen models are much more

precisely esimaed han he expendiure model Te sae-level enrollmen model

finds ha a 10 percen minimum wage increase in a low-wage sae is associaed wih

a 235 percen decrease in SNAP enrollmens Te amily-level linear probabiliymodel predics a somewha greaer elasiciy or low minimum wage saes an

increase o 10 percen in he ederal minimum would resul in a 317 percen decline

in SNAP enrollmen24 Te differences in hese esimaes sem rom a number o

acors including difference in model uncional orm and daa used We rea his

range o elasiciy esimaes as an upper and lower bound on enrollmen impacs

Harkin-Miller bill National and state-level predicted impacts

Wha would be he prediced change or he SNAP program i he ederalminimum were raised o $1010 as proposed in he Harkin-Miller bill In order o

make his inerence we accoun or he ac ha no all saes are currenly subjec

o he ederal minimum wage a he beginning o 2014 21 saes mainained

higher minimum wages han $725 In hose saes an increase in he ederal

minimum wage may or may no be binding or employers in he sae depending

upon wheher he new ederal minimum exceeds he sae-level minimum Bu

regardless o wheher a minimum wage change is binding he impac on SNAP

aciviy will be lower in high minimum wage saes In order o accoun or his

properly we calculae sae by sae he percenage wage change ha would resul

rom he Harkin-Miller proposal and apply he parameers rom each o he hree

models above o compue he expeced decrease in SNAP aciviy or each sae

In his exercise we use saesrsquo curren (2014) minimum wage levels and assume

as a baseline he 2012 levels o SNAP enrollmen and expendiure as 2012 is he

mos recen year or which SNAP daa are available

able 10 and able 11 in Appendix C repor he esimaed effecs on SNAP

enrollmen and expendiures respecively or each sae under he Harkin-Miller

bill25 An increase o $1010 i enaced oday would represen beween a 393

percen wage increase in a $725 minimum wage sae and an 84 percen increasein Washingon sae which has he highes minimum wage in he naion a $932

as o January 201426 Slighly more han 56 percen o he decrease in expendiures

and abou 59 percen o he decrease in enrollmen would occur in saes wih

presen-day minimum wages o $725

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2043

Results | wwwamericanprogresso

In 2012 hese saes were home o 46 percen o he American populaion

and accouned or a roughly equivalen percenage o oal naional SNAP

expendiures No surprisingly he larges enrollmen decreases come rom

populous saes wih very high SNAP enrollmen raes andor very low minimum

wages Te larges prediced enrollmen reducion991252beween 319000 individuals

and 362000 individuals991252would occur in exas which has a $725 minimum wage In Caliornia which has a 206 percen SNAP paricipaion rae and an $8

per hour minimum wage we anicipae SNAP enrollmen declines o 310000

persons o 371000 persons And in Florida which had a SNAP paricipaion rae

o 166 percen and a $793 minimum wage enrollmen could decline by beween

164000 individuals and 196000 individuals For he our saes ha ogeher

accouned or he greaes amoun o SNAP spending in 2012991252exas Caliornia

Florida and New York respecively991252he combined expendiure reducion rom

he Harkin-Miller bill is prediced o be $14 billion

able 2 summarizes he prediced declines in SNAP aciviy or he naion as a whole ha would resul rom he direc and indirec effecs o he Harkin-Miller

bill Enrollmen would all beween 31 million persons and 36 million persons

represening 75 percen o 87 percen o curren enrollmen Te anicipaed

reducion in program expendiures would be nearly $46 billion or 61 percen o

program expendiures

TABLE 2

Comparison of national SNAP predictions under the Harkin-Miller billrsquos $1010 minimum wage

Model

Enrollment

(persons)

Expenditures

(millions of dollars)

Current (2012) Predicted Change Current (2012) Predicted Chan

Family enrollment (linear probability)

41866195

45489339 -3623144

$74861

NA NA

State enrollment (ordinary least squares) 38745435 -3120759 NA NA

State expenditures (ordinary least squares) NA NA $70305 -$45

Note Calculations use 2014 state minimum wages and the most recent SNAP data from 2012 They assume that per-enrollee expenditures remain constant

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2143

18 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

Tere are o course oher possibiliies or minimum wage legislaion able 3

shows he expeced SNAP changes or he Unied Saes under a variey o wage

scenarios calculaed using he sae-level models I saes were no able o se

heir minimum wages independenly such ha all saes were consrained by

he ederal minimum o $725 SNAP would be received by abou 514000 more

people across he Unied Saes a an addiional program cos o nearly hree-quarers o a billion dollars In conras he effecs o a higher minimum wage

proposal991252a ederal wage floor o $11 per hour991252would decrease enrollmen in

SNAP by more han 10 percen and decrease program coss by 83 percen

TABLE 3

Summary of par ticipation and expenditures under wage scenarios

If all states had

minimum wages of

Enrollment(persons)

Expenditures

(millions of dollars)

Predicted Change Predicted Change

Recent levels (2014) 41866195 $74861

$725 42380520 514326 $75604 $743

$800 41423919 -442276 $74209 -$652

$900 40148451 -1717744 $72350 -$2511

$1000 38872982 -2993212 $70490 -$4371

$1010 38745435 -3120759 $70305 -$4556

$1100 37597514 -4268681 $68631 -$6230

Note Calculations use state-level enrollment model coefficient

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2243

Conclusion | wwwamericanprogresso

Conclusion

An exensive body o lieraure examines employmen effecs o he minimum

wage A much smaller se o sudies examines how he minimum wage affecs

povery and only a handul o papers examine he effecs o he minimum wage

on he EIC Our analysis is he firs o examine he effecs o he minimum

wage on SNAP

Our findings indicae ha increased earnings rom minimum wage changes do

reduce SNAP enrollmens and expendiures We esimae ha he Harkin-Miller bill would save axpayers nearly $46 billion per year equivalen o 61 percen

o SNAP expendiures in 2012 he las year or which daa are available Over a

10-year period he esimaed savings amoun o nearly $46 billion

Our repor is subjec o limiaions ha we expec o overcome in our uure

research Firs he findings do no ake ino accoun possible ineracions among

SNAP he EIC and Medicaid Te eligibiliy cuoffs among hese programs

are quie differen suggesing ha such ineracions may be minor Noneheless

he join effecs can only be deermined by urher research using a causal

model Second i would be useul o know he disribuion o SNAP reducions

along he wage disribuion Using he Congressional Budge Officersquos calculaions

o how much he oal dollar value o wage would increase under he Harkin-

Miller proposal our findings imply ha he decline in overall SNAP spending

equals abou 15 percen o he oal resuling increase in wages Te amoun and

disribuion o his offse are o considerable ineres Minimum wage beneficiaries

who come rom working amilies already well above he povery line would no

see any offse while hose who are currenly considerably below he povery line

will see larger offses Tese issues will also be a subjec or our uure research

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2343

20 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

About the authors

Rachel West is a maser o public policy candidae a he Goldman School

o Public Policy Universiy o Caliornia Berkeley Her research ocuses on

economic policy in he areas o low-wage labor and povery

Michael Reich is proessor o economics and direcor o he Insiue or

Research on Labor and Employmen a he Universiy o Caliornia a Berkeley

His research publicaions cover numerous areas o labor economics including

racial inequaliy labor marke segmenaion high-perormance workplaces

union-managemen cooperaion Japanese labor-managemen sysems living

wages and minimum wages He received his docorae in economics rom

Harvard Universiy

Acknowledgments

We are graeul o Sylvia Allegreto Arindraji Dube Bill Leser Jesse Rohsein

Daniel Tompson and Ben Zipperer or heir valuable suggesions

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2443

References | wwwamericanprogresso

References

Allegreto Sylvia and ohers 2013 ldquoFas Food Povery Wages Te Public Cos o Low-Wage Jobsin he Fas-Food Indusryrdquo Universiy o Caliornia Berkeley Cener or Labor Research andEducaion Available a htplaborcenerberkeleyedupubliccossas_ood_povery_wage

Allegreto Sylvia and ohers 2013 ldquoCredible Research Designs or Minimum Wage Sudiesrdquo

Working Paper 148-13 Universiy o Caliornia Berkeley Insiue or Research on Labor andEmploymen Available a htpwwwirleberkeleyeduworkingpapers148-13pd

Biler Marianne and Hilary Hoynes 2013 ldquo Te More Tings Change he More Tey Say heSame Te Saey Ne Living Arrangemens and Povery in he Grea Recessionrdquo Working Paper19449 Naional Bureau o Economic Research

Congressional Budge Office 2012 ldquoTe Supplemenal Nuriion Assisance Programrdquo Washingon Available a htpwwwcbogovsiesdeaulfilescbofilesatachmens04-19-SNAPpd

991252 991252 991252 2014 ldquoTe Effec o a Minimum-Wage Increase on Employmen and Family Incomerdquo Washingon Available a htpwwwcbogovsiesdeaulfilescbofilesatachmens44995-MinimumWagepd

Dube Arindraji 2013 rdquoMinimum Wages and he Disribuion o Family Incomerdquo Unpublished working paper Available a htpsdldropboxuserconencomu15038936Dube_ MinimumWagesFamilyIncomespd

991252 991252 991252 2014 ldquoTe Povery o Minimum Wage Facsrdquo Arindraji Dube Blog January 22 Availablea htparindubecom20140122he-povery-o-minimum-wage-acs

Lee David and Emmanuel Saez 2012 ldquoOpimal Minimum Wage Policy in Compeiive LaborMarkesrdquo Journal o Public Economics 96 (9) 739ndash749

Neumark David and William Wascher 1992 ldquoEmploymen Effecs o Minimum and Subminimum Wages Panel Daa on Sae Minimum Wage Lawsrdquo Industrial and Labor Relations Review 46 (1)

55ndash81

Neumark David and William Wascher 2011 ldquoDoes a Higher Minimum Wage Enhance heEffeciveness o he Earned Income ax Credirdquo Industrial and Labor Relations Review 64 (4)712ndash746

Page Marianne Joanne Spez and Jane Millar 2005 ldquoDoes he Minimum Wage Affec WelareCaseloadsrdquo Journal o Policy Analysis and Management 24 (2) 273ndash295

Rohsein Jesse 2010 ldquoIs he EIC as Good as an NI Condiional Cash ransers and ax

Incidencerdquo American Economic Journal Economic Policy 2 (1) 177ndash208

Wage and Hours Division 2014 ldquoMinimum Wage Laws in he Saes ndash January 1 2014rdquo USDeparmen o Labor (htpwwwdolgovwhdminwageamericahm [February 2014])

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2543

22 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2643

Appendix A | wwwamericanprogresso

Appendix A

Pre-trend falsification check

Recen minimum wage research27 highlighs a common flaw in previous sudies

ailure o veriy ha he oucome variable is ree o negaive pre-exising rends I

or example SNAP aciviy was already rending down in saes ha raised heir

minimum wages beore hese changes came ino effec our regression analysis could

(misakenly) atribue ha reducion o he minimum wage We check or such

pre-rends by inroducing variables ha represen he prior yearrsquos value or leads o

he minimum wage I he model esimaes he minimum wage o have an effec on

he oucome variable beore he wage change wen ino effec hen an unobservedacor no he minimum wage change caused he change in SNAP aciviy

We es he specificaions above or pre-rends by including a one-year lead in

all hree specificaions We find ha he lead erms are small posiive and no

saisically significan indicaing ha he concurren minimum wage991252no

he wage level in prior periods991252is driving he observed changes in SNAP

oucomes28 In paricular he coefficien (sandard error) on he lead erm in

our preerred amily-level enrollmen regression is 011 and no significan

while he coefficien and sandard error o he conemporaneous minimum

wage is unchanged In he sae-level preerred enrollmen regression he

coefficien o he lead erm is again small (07) and i is no significan Te

corresponding coefficien on he lead erm in he sae-level expendiure

regression is 16 and is no significan Te posiive poin esimaes on hese lead

erms resuls no only rule ou disoring negaive pre-rends Tey also sugges

ha our main resuls may underesimae he rue effecs

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2743

24 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2843

Appendix B | wwwamericanprogresso

Appendix B

Model selection process

For boh he amily-level and sae-level models we es hree mehods o conrol

or unobserved geographic- and ime-varying characerisics as suggesed by he

minimum wage lieraure Firs we include only independen sae-specific fixed

effecs and year-specific fixed effecs Tis specificaion (specificaion 1) implicily

assumes ha amilies in any sae consiue an equally good saisical ldquoconrolrdquo

group or hose in any randomly chosen sae afer accouning or various

characerisics (median income and unemploymen rae among ohers) Similarly

simple ime fixed effecs assume ha amilies surveyed in any year can crediblyserve as a conrol group or amilies surveyed in every oher year o he sample

(1990 hrough 2012)

In oher words specificaion 1 assumes ha a saersquos immediae neighbor provides

no beter a couneracual or he effec o a minimum wage change han does a

sae across he counry We relax his resricive specificaion sequenially in wo

seps In specificaion 2 we replace simple year fixed effecs wih fixed effecs or

each Census divisionyear (capured as an addiional variable in he vecor By

using division-year effecs we remove he resricion ha amilies in each sae

are equally good saisical conrols or all oher amilies Raher we allow or he

possibiliy ha amilies in similar geographic regions (or example he Souh or

he Norheas) may be more similar o one anoher han amilies arher away

Finally in specificaion 3 we add sae-specific linear ime rends o he previous

specificaion Tus specificaion 3 is he mos rigorous model specificaion in ha

i allows or heerogeneiy along hree dimensions Ta is specificaion 3 allows

each sae o have is own ime-varying rends raher han imposing he resricion

ha saes evolve idenically over he 22 years in he sample

We begin building he heoreical specificaion above rom a se o simpleuncondiional models regression o SNAP aciviy (enrollmen or expendiures)

on he log o he minimum wage and a se o geographic- and ime-specific

effecs (specificaions 1 2 and 3 described above) As shown in ables 1ndash3 (or

specificaion 3) we hen add covariaes sequenially o hese models including

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2943

26 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

firs he vecor o amily-level conrols ollowed by each o several sae-level

covariaes in urn he unemploymen rae log o median-amily income and he

employmen-o-populaion raio Comparable resuls or specificaions 2 and 3

will be available in our orhcoming working paper

As expeced he simple uncondiional models indicae ha he relaionship beween he minimum wage and SNAP enrollmen i one exiss is a more complex

one influenced by oher acors In he uncondiional model he coefficien on he

variable o ineres991252he log o he minimum wage991252is small in magniude and no

saisically differen rom zero Once we accoun or he influence o labor marke

condiions and variaion in income levels on program paricipaion (by including

unemploymen rae and median-amily income conrol variables respecively)

he effec o he minimum wage on SNAP enrollmen is precisely esimaed Te

coefficien o he log minimum wage is slighly higher (-0042) in he amily-level

analysis han he coefficien (-031) in he sae-level analysis Te level o precision

is also higher in he amily-level analysis Tis is o be expeced when using 124million observaions compared o 1127

Te second dimension o model choice concerns he effec specificaion ables

7ndash9 compare he primary coefficiens o ineres or he SNAP enrollmen and

expendiure models For boh he enrollmen models he effec sizes are smalles

or specificaion 1 larges or specificaion 2 and inermediae beween hese wo

in specificaion 3 Recall ha Specificaion 3 conains sae-specific linear ime

rends in addiion o he census divisionyear conrols included in specificaion

2 In he amily-level enrollmen model he sandard error o he minimum wage

coefficien is smaller han in he oher wo specificaions Sandard errors on he

oher variables are much smaller in specificaions 2 and 3 han in specificaion

1 On he basis o coefficien significance (join and individual) specificaions 2

and 3 are sricly preerred in boh enrollmen models o specificaion 1 which

conains only sae and year fixed effecs

A concern wih specificaions 2 and 3 is ha rend conrols such as sae linear

rends may incorrecly absorb some o he delayed impac o a minimum wage

When we es his issue by including lagged minimum wages we do no find ha

delayed effecs are significan Anoher concern is ha more sauraed modelsuse less o he saisical variaion which could reduce he saisical power o

he resuls However he sandard errors or our more sauraed models are no

higher and are lower in some cases han or he less sauraed models Overall

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3043

Appendix B | wwwamericanprogresso

his evidence suppors our use o he sauraed model as he preerred model

specificaion Moreover Dubersquos 2013 sudy shows ha more sauraed models

perorm beter han models wih jus sae and ime fixed effecs

Te esimaed enrollmen regressions a boh he amily and sae levels show large

and saisically significan coefficiens Te esimaed minimum wage effec in heexpendiures regressions991252or which we have only sae-level daa991252is also large

and saisically significan

We do no use weighed regression or he sae-level models preerring o keep

analysis o he ldquoreamenrdquo (ha is o say a minimum wage change) appropriae

o he average sae raher han he average amily or individual I insead our

primary ineres were he impac o a minimum wage change on he average amily

or he average individual we migh choose o designae he number o amilies

in each sae or he sae populaion respecively as analyic weighs in order o

obain a coefficien beter suied or such inerence

TABLE 4

SNAP enrollment

Family-level linear probability

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0048 -0047 -0040 -0043 -0042

(0013) (0013) (001) (0008) (0009)

Unemployment rate (100) 0505 0420 0280

(0083) (0086) (0082)

Log median income -0057 -0039

(0011) (0011)

Employment-to-population ratio -0239

(0038)

N 1242022 1242022 1242022 1242022 1242022

plt01 plt005 plt001

Note Robust standard errors are in parentheses Observations are clustered at the state level The outcome variable is binary and equal to one if a family is enrolledin SNAP All models include state fixed effects Census division x-year fixed effects and state-specific linear time trends All specifications except 3a include additional

controls for family size race and marital status of the family head presence of children and presence of an adult maleSource Annual data from the March Current Population Survey for 1990 to 2012 estimation includes Current Population Survey probability weights

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3143

28 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

TABLE 5

SNAP enrollment

State-level linear regression

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0028 -0024 -0026 -0031 -0031

(0014) (0013) (0013) (0012) (0012)

Unemployment rate (100) 0625 0466 0320 0339

(0087) (0088) (0085) (0083)

Log median income -0090 -0065 -0061

(0013) (0013) (0013)

Employment-to-population ratio -0282 -0248

(0037) (0038)

N 1127 1127 1127 1127 1127

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the SNAP enrollment rate All models include state fixed effects Census division x-year fixedeffects and state-specific linear time trends

Source Annual data from the March Current Population Survey for 1990 to 2012

TABLE 6

SNAP expenditures

State-level linear regression

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0181 -0149 -0156 -0153 -0190

(011) (0103) (0102) (0103) (0103)

Unemployment rate (100) 4587 4152 4232 4313

(0622) (0621) (0633) (0628)

Log median income -0246 -0261 -0294

(0075) (0078) (0078)

Employment-to-population ratio 0155 0244

(0237) (024)

N 1127 1127 1127 1127 1127

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the natural log of state SNAP expenditures per capita for 1990 to 2012 All models include state

fixed effects Census division x-year fixed effects and state-specific linear time trends

Source Annual data from Bureau of Economic Analysis national income and product accounts tables for 1990 to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3243

Appendix B | wwwamericanprogresso

TABLE 7

Comparison of specifications SNAP enrollment

Family-level linear probability

(1) (2) (3)

Log minimum wage -0028 -0049 -0042

(0014) (0017) (0009)

Unemployment rate (100) 0275 0297 0280

(0161) (0076) (0082)

Log median income -0077 -0055 -0039

(0014) (0012) (0011)

Employment-to-population ratio -0238 -0250 -0239

(0054) (004) (0038)

N 1242022 1242022 1242022

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses Observations are clustered at the state level The outcomevariable is binary or equal to one if a family is enrolled in SNAP All specifications include additional controls forfamily size race and marital status of the family head presence of children and presence of an adult male

Source Annual data from the March Current Population Survey for 1990 to 2012 estimation includes CurrentPopulation Survey probability weights

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3343

30 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

TABLE 8

Comparison of specifications SNAP enrollment

State-level linear regression

(1) (2) (3)

Log minimum wage -0019 -0035 -0031

(0009) (0012) (0012)

Unemployment rate (100) 0401 0370 0339

(0063) (0077) (0083)

Log median income -0081 -0073 -0061

(0011) (0013) (0013)

Employment-to-population ratio -0183 -0222 -0248

(0039) (0039) (0038)

N 1127 1127 1127

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the SNAP enrollment rate Allregressions include the share of households with children and the racial shares of the population

Source Annual data from the March Current Population Survey for 1990 to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3443

Appendix B | wwwamericanprogresso

TABLE 9

Comparison of specifications SNAP expenditures

State-level linear regression

(1) (2) (3)

Log minimum wage -0121 -0203 -0190

(0075) (0103) (0103)

Unemployment rate (100) 5292 5152 4313

(0464) (0576) (0628)

Log median income -0437 -0417 -0294

(008) (0086) (0078)

Employment-to-population ratio -0040 -0220 0244

(0261) (0260) (0240)

N 1127 1127 1127

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the natural log of state SNAPexpenditures per capita All regressions include the share of households with children and the racial shares ofthe state population

Source Annual data from Bureau of Economic Analysis national income and product accounts tables for 1990to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3543

32 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3643

Appendix C | wwwamericanprogresso

Appendix C Harkin-Miller

policy simulation results

StateMinimum wage

(2014)

SNAP enrollment (2012) Predicted change in SNAP enrollment

RateTotal

(persons)

Model 1 family

enrollment (linearprobability)

Model 2 state

enrollment (linearregression)

Model 3 s

expenditu(linear regre

Alabama $725 164 788682 -66922 -58897 -58906

Alaska $775 120 87436 -8104 -6887 -3288

Arizona $790 201 1319323 -67435 -56738 -64356

Arkansas $725 78 230489 -40977 -36063 -35248

California $800 206 7813680 -371131 -310222 -18223

Colorado $800 164 853155 -50684 -42365 -23926

Connecticut $870 91 326621 -22456 -17975 -13711

Delaware $725 186 170262 -12739 -11211 -10647

District of Columbia $825 133 84009 -5370 -4417 -3632

Florida $793 166 3208026 -195813 -164426 -13046

Georgia $725 160 1586336 -137741 -121224 -11004

Hawaii $725 96 133662 -19310 -16995 -14933

Idaho $725 92 147501 -22165 -19507 -15809

Illinois $825 95 1225084 -109088 -89742 -70955

Indiana $725 125 816233 -90818 -79928 -83985

Iowa $725 155 478011 -42716 -37594 -28556

Kansas $725 135 388269 -40082 -35275 -27461

Kentucky $725 130 568821 -60840 -53544 -52259

Louisiana $725 149 683832 -63929 -56263 -66083

Maine $750 77 101976 -16567 -14323 -15234

Maryland $725 144 846415 -81748 -71946 -38370

Massachusetts $800 130 864721 -64902 -54251 -42913

Michigan $740 146 1439141 -128801 -112140 -11022

Minnesota $725 133 713646 -74730 -65769 -37878

TABLE 10

SNAP enrollments Predicted changes by state under the Harkin-Miller bill

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3743

34 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

StateMinimum wage

(2014)

SNAP enrollment (2012) Predicted change in SNAP enrollment

RateTotal

(persons)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 state

expenditures

(linear regression)

Mississippi $725 129 386501 -41486 -36511 -46467

Missouri $750 172 1036182 -75131 -64952 -56244

Montana $790 132 132452 -10350 -8708 -5846

Nebraska $725 124 230683 -25773 -22683 -12189

Nevada $825 162 446035 -23349 -19209 -11894

New Hampshire $725 127 168404 -18359 -16157 -5735

New Jersey $825 160 1416666 -75175 -61843 -28236

New Mexico $750 149 310896 -25983 -22463 -22512

New York $800 192 3763553 -191193 -159815 -142182

North Carolina $725 174 1697193 -135417 -119179 -113503

North Dakota $725 87 61225 -9743 -8574 -4021

Ohio $795 143 1647345 -115869 -97169 -88580

Oklahoma $725 129 494053 -53006 -46650 -46854

Oregon $910 124 485326 -17036 -13328 -16398

Pennsylvania $725 161 2053643 -177315 -156052 -125586

Rhode Island $800 156 163730 -10258 -8574 -8698

South Carolina $725 94 445277 -65614 -57746 -50304

South Dakota $725 208 173749 -11586 -10197 -7458

Tennessee $725 142 914903 -89667 -78915 -99134

Texas $725 110 2863779 -362018 -318607 -253285

Utah $725 88 251107 -39658 -34902 -19390

Vermont $873 156 97792 -3823 -3055 -2475

Virginia $725 101 829771 -113723 -100086 -58212

Washington $932 72 496934 -23221 -17947 -17756

West Virginia $725 58 107875 -25792 -22699 -21665

Wisconsin $725 75 427822 -79521 -69986 -53210

Wyoming $725 164 94590 -8010 -7050 -3104

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3843

Appendix C | wwwamericanprogresso

StateMinimum wage

(2014)

SNAP expenditures (2012)Predicted change in SNAP expenditures

(millions of dollars)

Per resident

(millions of

dollars)

Total

(millions of

dollars)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 s

expenditu

(linear regre

Alabama $725 $290 $1397 -$1185 -$1043 -$1043

Alaska $775 $253 $185 -$262 -$223 -$106

Arizona $790 $257 $1686 -$935 -$787 -$892

Arkansas $725 $248 $733 -$636 -$560 -$547

California $800 $189 $7164 -$7276 -$6082 -$3573

Colorado $800 $157 $816 -$862 -$721 -$407

Connecticut $870 $191 $686 -$343 -$275 -$210

Delaware $725 $250 $229 -$205 -$180 -$171

District of Columbia $825 $366 $232 -$146 -$120 -$99

Florida $793 $294 $5676 -$4429 -$3719 -$2951

Georgia $725 $317 $3140 -$2936 -$2584 -$2346

Hawaii $725 $335 $465 -$449 -$395 -$347

Idaho $725 $225 $359 -$376 -$331 -$268

Illinois $825 $249 $3200 -$2096 -$1725 -$1364

Indiana $725 $220 $1439 -$1162 -$1023 -$1075

Iowa $725 $192 $589 -$658 -$579 -$440

Kansas $725 $159 $460 -$502 -$441 -$344

Kentucky $725 $298 $1303 -$1133 -$997 -$973

Louisiana $725 $315 $1450 -$1047 -$922 -$1083

Maine $750 $281 $373 -$267 -$231 -$246

Maryland $725 $188 $1109 -$1765 -$1553 -$828

Massachusetts $800 $206 $1366 -$1030 -$861 -$681

Michigan $740 $300 $2963 -$2400 -$2090 -$2054

Minnesota $725 $140 $755 -$1113 -$980 -$564

Mississippi $725 $326 $973 -$649 -$571 -$726

Missouri $750 $241 $1452 -$1278 -$1104 -$956

Montana $790 $190 $191 -$179 -$151 -$101

Nebraska $725 $140 $259 -$409 -$360 -$194

Nevada $825 $191 $527 -$441 -$363 -$225

New Hampshire $725 $126 $167 -$399 -$351 -$125

TABLE 11

SNAP expenditures Predicted changes by state under the Harkin-Miller bill

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3943

36 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

StateMinimum wage

(2014)

SNAP expenditures (2012)Predicted change in SNAP expenditures

(millions of dollars)

Per resident

(millions of

dollars)

Total

(millions of

dollars)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 state

expenditures

(linear regression)

New Jersey $825 $160 $1420 -$1611 -$1325 -$605

New Mexico $750 $324 $675 -$513 -$444 -$445New York $800 $287 $5616 -$3766 -$3148 -$2801

North Carolina $725 $252 $2454 -$2187 -$1925 -$1833

North Dakota $725 $128 $90 -$162 -$143 -$67

Ohio $795 $259 $2995 -$2013 -$1688 -$1539

Oklahoma $725 $248 $945 -$799 -$703 -$706

Oregon $910 $322 $1255 -$272 -$213 -$262

Pennsylvania $725 $218 $2779 -$2930 -$2579 -$2075

Rhode Island $800 $280 $294 -$173 -$144 -$147

South Carolina $725 $291 $1373 -$1337 -$1177 -$1025South Dakota $725 $198 $165 -$192 -$169 -$123

Tennessee $725 $324 $2091 -$1413 -$1243 -$1562

Texas $725 $230 $5997 -$6402 -$5634 -$4479

Utah $725 $141 $402 -$614 -$541 -$300

Vermont $873 $230 $144 -$66 -$53 -$43

Virginia $725 $173 $1413 -$2062 -$1815 -$1056

Washington $932 $244 $1682 -$350 -$270 -$267

West Virginia $725 $273 $508 -$451 -$397 -$379

Wisconsin $725 $204 $1166 -$1302 -$1146 -$871Wyoming $725 $95 $55 -$105 -$93 -$41

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4043

Appendix C | wwwamericanprogresso

FIGURE 2

Raising the minimum wage to $1010would cut taxpayer costs in every state

Predicted decreases in cost and enrollment

in SNAP in 50 states

$200+$51ndash$100

$101ndash$200

0ndash$25

$26ndash$50

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4143

38 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

Endnotes

1 US Department of Agriculture ldquoSupplementalNutrition Assistance Program (SNAP)rdquo available athttpwwwfnsusdagovsnapeligibility (last accessedFebruary 2014)

2 For this initial analysis we do not consider Harkin-Millerrsquos increase in subminimum wages for tippedworkers To do s o would increase the estimated SNAP

savings by an unknown amount

3 The Congressional Budget Office estimates thatworkers currently earning between $1010 and $1150per hour would see their wages rise under the Harkin-Miller proposal Congressional Budget O ffice ldquoTheEffects of a Minimum Wage Increase on Employmentand Family Incomerdquo (2014)

4 Marianne Page Joanne Spetz and Jane Millar ldquoDoesthe Minimum Wage Affect Welfare Caseloadsrdquo Journalof Policy Analysis and Management 24 (2) (2005)273ndash295

5 Marianne Bitler and Hilary Hoynes ldquoThe More ThingsChange the More They Stay the Same The SafetyNet Living Arrangements and Poverty in the GreatRecessionrdquo NBER Working Paper 194 49 2013

6 Sylvia Allegretto and others ldquoFast Food Poverty Wages The Public Cost of Low-Wage Jobs in the Fast-FoodIndustryrdquo (Berkeley California Center for LaborResearch and Education 2013) available at httplaborcenterberkeleyedupubliccostsfast_food_poverty_wage

7 David Neumark and William Wascher ldquoDoes a HigherMinimum Wage Enhance the Effectiveness of theEarned Income Tax Creditrdquo Industrial and LaborRelations Review 64 (4) (2011) 712ndash746

8 David Lee and Emmanuel Saez ldquoOptimal MinimumWage Policy in Competitive Labor Marketsrdquo Journal ofPublic Economics 96 (9) (2012) 739ndash749

9 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family IncomerdquoFebruary 2014

10 Jesse Rothstein ldquoIs the EITC as Good as an NITConditional Cash Transfers and Tax Incidencerdquo AmericanEconomic Journal Economic Policy 2 (1) (2010) 177ndash208

11 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family Incomerdquo

12 Dube Arindrajit 2013 rdquoMinimum Wagesand the Distribution of Family IncomerdquoUnpublished working paper Available at httpsdldropboxusercontentcomu15038936Dube_MinimumWagesFamilyIncomespdf

13 As Dube explains in ldquoThe poverty of Minimum WageFactsrdquo the simulation approach underestimate stemsfrom a number of unwarranted assumptions includingthe range of actual wage increases and the accuracy ofwage data in the Current Population Survey The causal

approach does not make these assumptions

14 Allegretto Sylvia and others 2013 ldquoCredible ResearchDesigns for Minimum Wage Studiesrdquo Working Paper148-13 University of California Berkeley Institutefor Research on Labor and Employment Available athttpwwwirleberkeleyeduworkingpapers148-13pdf

15 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family In comerdquo

16 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo (2012)

17 Ibid

18 Ibid

19 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo

20 US Department of Agriculture ldquoSupplementalNutrition Assistance Program (SNAP)rdquo

21 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo

22 Ibid

23 We deviate from the Census Bureaursquos definition ofa family unit which is ldquotwo people or more (on e ofwhom is the householder) related by bir th marriageor adoption and residing togetherrdquo US Bureau ofthe Census ldquoCurrent Population Survey DefinitionsFamilyrdquo available at httpwwwcensusgovcpsabout

cpsdefhtml (last accessed February 2014) We countas a family unit any individual residing on his or herown two or more persons residing together whodo not belong to a family in the March CPS sampleare constructed as one family in our analysis For thepurposes of food stamp allocations the consumptionresulting from this transfer is probably distributed tofamily members (rather than household members ora single individual within the household) Howeversingle individuals canmdashand domdashreceive SNAPbenefits Excluding them would fail to make theanalysis reflective of the population at large

24 Strictly the family level linear probability modelpredicts the percentage-point decrease in theprobability that an individual family will receive SNAPpayments When applied to a large number of familieshowever we are able to interpret the coefficient asa decrease in the mean of enrollmentmdashthat is a

decrease in the enrollment ratemdashby applying the lawof iterated expectations

25 We generate expenditure predictions from theenrollment modelsmdashand conversely generateenrollment predictions from the expenditure modelmdashby assuming that expenditures per enrolled familyremains the same before and after the minimum wagechange In practice this is likely to be a conservativeestimatemdashthat is to underestimate the decrease inSNAP activity Average SNAP benefits per family willalso decrease as many families that remain eligible forSNAP experience income gains

26 Wage and Hour Division ldquoMinimum Wage Laws inthe States ndash Januar y 1 2014rdquo available at httpwwwdolgovwhdminwageamericahtm (last accessedFebruary 2014)

27 See for example Sylvia Allegretto and others ldquoCredibleResearch Designs for Minimum Wage Studiesrdquo WorkingPaper 148-113 (Berkeley California Institute forResearch on Labor and Employment 2013) available athttpwwwirleberkeleyeduworkingpapers148-13pdf

28 We will report these results in a forthcoming workingpaper

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4243

Endnotes | wwwamericanprogresso

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4343

The Center for American Progress is a nonpartisan research and educational institute

dedicated to promoting a strong just and free America that ensures opportunity

for all We believe that Americans are bound together by a common commitment to

these values and we aspire to ensure that our national policies reflect these values

We work to find progressive and pragmatic solutions to significant domestic and

international problems and develop policy proposals that foster a government that

is ldquoof the people by the people and for the peoplerdquo

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1243

Methods and data | wwwamericanprogress

Methods and data

As previously menioned we exploi variaion in minimum wages by sae and ime

o examine heir causal effecs on SNAP enrollmens and expendiures o do so we

merge daa rom 1990 hrough 2012 drawn rom he Annual Social and Economic

Supplemen o he Curren Populaion Survey991252an annual Census Bureau survey

commonly known as he March CPS ha includes23 inormaion on SNAP

enrollmens a he amily level991252wih sae-level daa on minimum wages SNAP

expendiures populaion unemploymen raes and sae median income levels

o conrol or ime-varying heerogeneiy among saes our specificaions includeconrols or sae linear rends and effecs by Census division and ime We esimae

effecs a wo levels allowing or amily variaion and allowing only or sae-level

variaion We also employ a se o sandard demographic conrols such as amily size

and composiion and race and ehnic composiion

Distinguishing causation from correlation

How can we ensure ha our analysis does no pick up a spurious correlaion or

example he endency o more economically vibran saes o implemen higher

minimum wages Disinguishing correlaion and policy endogeneiy rom rue

causal effecs is he primary moivaion or economeric analysis In he ideal

experimen researchers would begin wih wo saes991252ha are alike in every

respec prior o he policy991252 and ldquoreardquo only one o hese saes wih a higher

minimum wage Tey would atemp o shield hese saes rom any influence ha

could obscure heir undersanding o he minimum wagersquos direc effec on SNAP

enrollmen Researchers o course canno conduc such experimens

We can however use saisical mehods o conrol simulaneously heindependen effecs on SNAP o sae unemploymen raes sae income levels

and common rajecories among saes wihin he same Census division By

ensuring similariy along hese dimensions we maximize he likelihood ha

SNAP aciviy in wo saes would have comparable oucomes in he absence

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1343

10 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

o a minimum wage policy change Tus i a new minimum wage policy were

implemened in one sae only he researchers could atribue all o he difference

hey observe in SNAP aciviy o he new minimum wage policy

In oher words we approximae he ideal experimen by using non-experimenal

saisical mehods Te desirable ldquopre-exising similariiesrdquo beween saes ha wehave defined above inorm our choice o conrol variables in a saisical seting

More precisely in our muliple regression models we use median amily income

he unemploymen rae he employmen-o-populaion raio and regional and

ime idenifiers o consruc an appropriae group o peers or each sae on he

eve o a policy change

Data description

wo daa ses include inormaion abou boh income and paricipaion inpublic programs Te Survey o Income and Program Paricipaion or SIPP

which is conduced in inermiten years has he advanage o ollowing he

same individuals over a period o ime In oher words i is a longiudinal daa

se I also has he advanage o conaining monhly daa However he sample

size o he SIPP is no sufficien or analyzing variaions in sae-level minimum

wages Te March CPS has he advanage o a much larger sample size and i is

conduced annually wihou any breaks in ime I has he disadvanage o being

a cross-secional daa se so we canno ollow he same individuals over ime991252

sricly speaking over more han one year On ne he March CPS is much more

suiable or our sudy We examine he empirical relaionship beween minimum

wage policy and ood samp aciviy a wo levels o aggregaion he amily level

and he sae level Family-level daa are drawn rom he March CPS

Te March CPS comprises responses rom he residens o 50000 o 60000

dwelling places surveyed per year and conains deailed inormaion on he

residensrsquo employmen and income including income rom ranser paymens

Te sample or our analysis comprises more han 128 million amily unis during

he period rom 1990 o 2012 (inclusive) Survey weighs allow us o analyze

SNAP paricipaion in a manner ha is represenaive o he US populaion alarge Over all years he share o amilies reporing ood samp receip in he

weighed March CPS sample is 91 percen Te enrollmen rae was a a low o 6

percen in he year 2000 In 2012 he mos recen year in our panel 133 percen

o amilies repored paricipaing in SNAP a some poin during he survey year

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1443

Methods and data | wwwamericanprogresso

Te March CPS also collecs inormaion on he number o SNAP recipiens in

he household he number o monhs he household receives SNAP benefis

and he cash-equivalen value o he SNAP benefis received However he

value o SNAP benefis is severely underrepored among recipiens perhaps

because recipiens are unaware o he exac monhly cash-equivalen value o

benefis hey receive

Our firs empirical sraegy ocuses solely on SNAP enrollmen By using he

amily as he uni o analysis we are able o inser saisical conrols o accoun

or non-wage-relaed acors ha influence any paricular amilyrsquos likelihood o

program paricipaion wih he inenion o isolaing any differences in program

paricipaion ha are due purely o changes in wage policy Tis approach

idenifies he effecs o low-wage labor policy on he exernal margin991252ha is

he effec o he minimum wage on he likelihood ha a amily paricipaes in he

SNAP program a all991252as opposed o he inernal margin or how much SNAP

unding he amily would receive

Our second empirical ramework uses sae-level adminisraive daa Ta is

we aggregae he daa o obain a single daa poin or each saeyear back

o 1990 represening he mean o he oucome or he sae Te sae-level

esimaion serves as a robusness check on he amily-level resuls or SNAP

paricipaion Also using aggregaed daa allows us o esimae direcly he

causal effec o minimum wage changes on SNAP spending Tis is no possible

a he amily level as discussed above daa on cash-equivalen value o ood

samps or SNAP recipiens is very requenly no repored in he March CPS

and when i is repored he inormaion may be unreliable By conras he

Bureau o Economic Analysis publishes aggregae SNAP spending a he sae

level in is Naional Income and Produc Accoun or NIPA ables Tus

while we are unable o observe he heerogeneiy in he cash value o SNAP

or amilies in each sae we are able o calculae average SNAP spending

per residen in each sae per year Supporing covariaes include he annual

unemploymen and employmen daa rom he Bureau o Labor Saisics or

BLS and sae-level populaion series rom he iner-decennial census releases

Minimum wage daa are available rom he BLSrsquos wages and hours division For

sae minimum wage changes enaced a oher imes han he firs o he yearan average value or he year is used

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1543

12 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

Family level model specifications

We firs examine he effec o he minimum wage on paricipaion in public

assisance programs For amily residing in sae and during year we esimae

an equaion o he ollowing orm

(1)

is a binary variable ha is se equal o 1 i a leas one member o amily

received ood samps during he survey year is a se o sae-level

characerisics including annual averages o he unemploymen rae he

employmen-o-populaion raio and he naural log o median amily income

is a vecor o amily atribues including indicaors or he race and marial saus

o he amily head size o he amily he presence o children and he presence oan adul male Sae fixed effecs are capured by o conrol or ime-varying

heerogeneiy our preerred model specificaion also includes year fixed effecs

ha vary by Census division ( ) and sae-level linear ime rends In

Appendix B we jusiy he inclusion o hese las wo erms We also compare he

resuls rom our preerred specificaion wih less sauraed specificaions

Te effec o ineres which is capured by is he expeced change in he

probabiliy o receiving SNAP benefis wih respec o a change in he (log o he)

binding minimum wage in sae during year We repor robus sandard errors

clusered a he sae level We esimae he parameers using linear regression

producing a linear probabiliy model Deails o he model selecion process are

covered in Appendix B below

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1643

Methods and data | wwwamericanprogresso

State-level model specifications

Te sae-level models are similarly specified For sae in year we assume ha

(2)

In his model is now eiher he SNAP enrollmen rae in sae during year

or he naural logarihm o per capia SNAP expendiures in sae during

year is once again a se o sae-level characerisics including he same

sae-level covariaes as in he amily regressions (annual average unemploymen

rae employmen-o-populaion raio naural log o median amily income)

wih he addiion o amily level characerisics averaged across he sae (average

amily size and he shares o populaion consiued by each o five racialehnicgroups) Sae fixed effecs are represened by As above our preerred model

specificaion includes year fixed effecs ha vary by Census division ( ) and

sae-level linear ime rends as elaboraed in Appendix B Te effec o

ineres is capured by

We esimae boh sae-level models (enrollmen and expendiures) using

ordinary leas squares regression Tus he inerpreaion o he coefficien is no

longer ha o a change in probabiliy as in he binary oucome models described

above Raher or he sae-level SNAP enrollmen model represens he

expeced change (in percenage poins) in he saersquos SNAP enrollmen rae ha

is due o a 1 percen change in he minimum wage For he SNAP expendiures

model is simply he elasiciy o SNAP spending wih respec o he minimum

wage991252ha is he percenage change in sae expendiures expeced o resul rom

a 1 percen change in ha saersquos minimum wage For urher deails on model

specificaion reer o Appendix B below

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1743

14 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1843

Results | wwwamericanprogresso

Results

Estimated minimum wage effects on SNAP enrollment and

expenditures

able 1 shows he esimaed parameer o ineres991252he coefficien o he

minimum wage or he preerred model o each ype Coefficiens on he

minimum wage variable are no direcly comparable across models because all

our models have a differen uncional orm o undersand and compare hese

esimaes we compue he change in SNAP aciviy prediced or a paricular wage scenario Te final column in able 1 answers he quesion Wha would be

he expeced change in SNAP aciviy in response o a 10 percen increase in he

minimum wage Te answer o his quesion varies wih he value o he inpu

parameers in he able we calculae he percenage decrease in enrollmen or

expendiures prediced or he average sae wih a minimum wage o $725 in

2014 Te sae-level SNAP expendiure model which is a consan-elasiciy

model conveys elasiciy inormaion direcly or he change in expendiures per

capia in he sae

TABLE 1

Comparison of national SNAP predictions for a 10 percent increase in the federal minimum wage

Model LevelRegression

type

Predicted outcome Coefficient of log

(minimum wage)

(Standard error)

Effect of a 10 percent in

in the minimum wage

VariableForm of

variable

Total

enrollment

Tot

expend

1 Family Linear

probability Enrollment Binary (enrolled=1)

-0042

(0008) -317 N

2

StateLinear regression

(ordinary least

squares)

Enrollment State enrollment rate (percent) -0031

(0012) -235 N

3 Expenditures Log (state expenditures per capita) -0190

(0103) NA -19

plt01 plt005 plt001Note Predicted changes are calculated for the average state with a minimum wage of $725 in 2014

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1943

16 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

According o his model a 10 percen increase in he minimum wage would resul

in a 19 percen reducion in SNAP expendiures Tis resul is reassuringly similar

o Dubersquos resul or povery reducion Te wo enrollmen models are much more

precisely esimaed han he expendiure model Te sae-level enrollmen model

finds ha a 10 percen minimum wage increase in a low-wage sae is associaed wih

a 235 percen decrease in SNAP enrollmens Te amily-level linear probabiliymodel predics a somewha greaer elasiciy or low minimum wage saes an

increase o 10 percen in he ederal minimum would resul in a 317 percen decline

in SNAP enrollmen24 Te differences in hese esimaes sem rom a number o

acors including difference in model uncional orm and daa used We rea his

range o elasiciy esimaes as an upper and lower bound on enrollmen impacs

Harkin-Miller bill National and state-level predicted impacts

Wha would be he prediced change or he SNAP program i he ederalminimum were raised o $1010 as proposed in he Harkin-Miller bill In order o

make his inerence we accoun or he ac ha no all saes are currenly subjec

o he ederal minimum wage a he beginning o 2014 21 saes mainained

higher minimum wages han $725 In hose saes an increase in he ederal

minimum wage may or may no be binding or employers in he sae depending

upon wheher he new ederal minimum exceeds he sae-level minimum Bu

regardless o wheher a minimum wage change is binding he impac on SNAP

aciviy will be lower in high minimum wage saes In order o accoun or his

properly we calculae sae by sae he percenage wage change ha would resul

rom he Harkin-Miller proposal and apply he parameers rom each o he hree

models above o compue he expeced decrease in SNAP aciviy or each sae

In his exercise we use saesrsquo curren (2014) minimum wage levels and assume

as a baseline he 2012 levels o SNAP enrollmen and expendiure as 2012 is he

mos recen year or which SNAP daa are available

able 10 and able 11 in Appendix C repor he esimaed effecs on SNAP

enrollmen and expendiures respecively or each sae under he Harkin-Miller

bill25 An increase o $1010 i enaced oday would represen beween a 393

percen wage increase in a $725 minimum wage sae and an 84 percen increasein Washingon sae which has he highes minimum wage in he naion a $932

as o January 201426 Slighly more han 56 percen o he decrease in expendiures

and abou 59 percen o he decrease in enrollmen would occur in saes wih

presen-day minimum wages o $725

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2043

Results | wwwamericanprogresso

In 2012 hese saes were home o 46 percen o he American populaion

and accouned or a roughly equivalen percenage o oal naional SNAP

expendiures No surprisingly he larges enrollmen decreases come rom

populous saes wih very high SNAP enrollmen raes andor very low minimum

wages Te larges prediced enrollmen reducion991252beween 319000 individuals

and 362000 individuals991252would occur in exas which has a $725 minimum wage In Caliornia which has a 206 percen SNAP paricipaion rae and an $8

per hour minimum wage we anicipae SNAP enrollmen declines o 310000

persons o 371000 persons And in Florida which had a SNAP paricipaion rae

o 166 percen and a $793 minimum wage enrollmen could decline by beween

164000 individuals and 196000 individuals For he our saes ha ogeher

accouned or he greaes amoun o SNAP spending in 2012991252exas Caliornia

Florida and New York respecively991252he combined expendiure reducion rom

he Harkin-Miller bill is prediced o be $14 billion

able 2 summarizes he prediced declines in SNAP aciviy or he naion as a whole ha would resul rom he direc and indirec effecs o he Harkin-Miller

bill Enrollmen would all beween 31 million persons and 36 million persons

represening 75 percen o 87 percen o curren enrollmen Te anicipaed

reducion in program expendiures would be nearly $46 billion or 61 percen o

program expendiures

TABLE 2

Comparison of national SNAP predictions under the Harkin-Miller billrsquos $1010 minimum wage

Model

Enrollment

(persons)

Expenditures

(millions of dollars)

Current (2012) Predicted Change Current (2012) Predicted Chan

Family enrollment (linear probability)

41866195

45489339 -3623144

$74861

NA NA

State enrollment (ordinary least squares) 38745435 -3120759 NA NA

State expenditures (ordinary least squares) NA NA $70305 -$45

Note Calculations use 2014 state minimum wages and the most recent SNAP data from 2012 They assume that per-enrollee expenditures remain constant

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2143

18 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

Tere are o course oher possibiliies or minimum wage legislaion able 3

shows he expeced SNAP changes or he Unied Saes under a variey o wage

scenarios calculaed using he sae-level models I saes were no able o se

heir minimum wages independenly such ha all saes were consrained by

he ederal minimum o $725 SNAP would be received by abou 514000 more

people across he Unied Saes a an addiional program cos o nearly hree-quarers o a billion dollars In conras he effecs o a higher minimum wage

proposal991252a ederal wage floor o $11 per hour991252would decrease enrollmen in

SNAP by more han 10 percen and decrease program coss by 83 percen

TABLE 3

Summary of par ticipation and expenditures under wage scenarios

If all states had

minimum wages of

Enrollment(persons)

Expenditures

(millions of dollars)

Predicted Change Predicted Change

Recent levels (2014) 41866195 $74861

$725 42380520 514326 $75604 $743

$800 41423919 -442276 $74209 -$652

$900 40148451 -1717744 $72350 -$2511

$1000 38872982 -2993212 $70490 -$4371

$1010 38745435 -3120759 $70305 -$4556

$1100 37597514 -4268681 $68631 -$6230

Note Calculations use state-level enrollment model coefficient

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2243

Conclusion | wwwamericanprogresso

Conclusion

An exensive body o lieraure examines employmen effecs o he minimum

wage A much smaller se o sudies examines how he minimum wage affecs

povery and only a handul o papers examine he effecs o he minimum wage

on he EIC Our analysis is he firs o examine he effecs o he minimum

wage on SNAP

Our findings indicae ha increased earnings rom minimum wage changes do

reduce SNAP enrollmens and expendiures We esimae ha he Harkin-Miller bill would save axpayers nearly $46 billion per year equivalen o 61 percen

o SNAP expendiures in 2012 he las year or which daa are available Over a

10-year period he esimaed savings amoun o nearly $46 billion

Our repor is subjec o limiaions ha we expec o overcome in our uure

research Firs he findings do no ake ino accoun possible ineracions among

SNAP he EIC and Medicaid Te eligibiliy cuoffs among hese programs

are quie differen suggesing ha such ineracions may be minor Noneheless

he join effecs can only be deermined by urher research using a causal

model Second i would be useul o know he disribuion o SNAP reducions

along he wage disribuion Using he Congressional Budge Officersquos calculaions

o how much he oal dollar value o wage would increase under he Harkin-

Miller proposal our findings imply ha he decline in overall SNAP spending

equals abou 15 percen o he oal resuling increase in wages Te amoun and

disribuion o his offse are o considerable ineres Minimum wage beneficiaries

who come rom working amilies already well above he povery line would no

see any offse while hose who are currenly considerably below he povery line

will see larger offses Tese issues will also be a subjec or our uure research

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2343

20 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

About the authors

Rachel West is a maser o public policy candidae a he Goldman School

o Public Policy Universiy o Caliornia Berkeley Her research ocuses on

economic policy in he areas o low-wage labor and povery

Michael Reich is proessor o economics and direcor o he Insiue or

Research on Labor and Employmen a he Universiy o Caliornia a Berkeley

His research publicaions cover numerous areas o labor economics including

racial inequaliy labor marke segmenaion high-perormance workplaces

union-managemen cooperaion Japanese labor-managemen sysems living

wages and minimum wages He received his docorae in economics rom

Harvard Universiy

Acknowledgments

We are graeul o Sylvia Allegreto Arindraji Dube Bill Leser Jesse Rohsein

Daniel Tompson and Ben Zipperer or heir valuable suggesions

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2443

References | wwwamericanprogresso

References

Allegreto Sylvia and ohers 2013 ldquoFas Food Povery Wages Te Public Cos o Low-Wage Jobsin he Fas-Food Indusryrdquo Universiy o Caliornia Berkeley Cener or Labor Research andEducaion Available a htplaborcenerberkeleyedupubliccossas_ood_povery_wage

Allegreto Sylvia and ohers 2013 ldquoCredible Research Designs or Minimum Wage Sudiesrdquo

Working Paper 148-13 Universiy o Caliornia Berkeley Insiue or Research on Labor andEmploymen Available a htpwwwirleberkeleyeduworkingpapers148-13pd

Biler Marianne and Hilary Hoynes 2013 ldquo Te More Tings Change he More Tey Say heSame Te Saey Ne Living Arrangemens and Povery in he Grea Recessionrdquo Working Paper19449 Naional Bureau o Economic Research

Congressional Budge Office 2012 ldquoTe Supplemenal Nuriion Assisance Programrdquo Washingon Available a htpwwwcbogovsiesdeaulfilescbofilesatachmens04-19-SNAPpd

991252 991252 991252 2014 ldquoTe Effec o a Minimum-Wage Increase on Employmen and Family Incomerdquo Washingon Available a htpwwwcbogovsiesdeaulfilescbofilesatachmens44995-MinimumWagepd

Dube Arindraji 2013 rdquoMinimum Wages and he Disribuion o Family Incomerdquo Unpublished working paper Available a htpsdldropboxuserconencomu15038936Dube_ MinimumWagesFamilyIncomespd

991252 991252 991252 2014 ldquoTe Povery o Minimum Wage Facsrdquo Arindraji Dube Blog January 22 Availablea htparindubecom20140122he-povery-o-minimum-wage-acs

Lee David and Emmanuel Saez 2012 ldquoOpimal Minimum Wage Policy in Compeiive LaborMarkesrdquo Journal o Public Economics 96 (9) 739ndash749

Neumark David and William Wascher 1992 ldquoEmploymen Effecs o Minimum and Subminimum Wages Panel Daa on Sae Minimum Wage Lawsrdquo Industrial and Labor Relations Review 46 (1)

55ndash81

Neumark David and William Wascher 2011 ldquoDoes a Higher Minimum Wage Enhance heEffeciveness o he Earned Income ax Credirdquo Industrial and Labor Relations Review 64 (4)712ndash746

Page Marianne Joanne Spez and Jane Millar 2005 ldquoDoes he Minimum Wage Affec WelareCaseloadsrdquo Journal o Policy Analysis and Management 24 (2) 273ndash295

Rohsein Jesse 2010 ldquoIs he EIC as Good as an NI Condiional Cash ransers and ax

Incidencerdquo American Economic Journal Economic Policy 2 (1) 177ndash208

Wage and Hours Division 2014 ldquoMinimum Wage Laws in he Saes ndash January 1 2014rdquo USDeparmen o Labor (htpwwwdolgovwhdminwageamericahm [February 2014])

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2543

22 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2643

Appendix A | wwwamericanprogresso

Appendix A

Pre-trend falsification check

Recen minimum wage research27 highlighs a common flaw in previous sudies

ailure o veriy ha he oucome variable is ree o negaive pre-exising rends I

or example SNAP aciviy was already rending down in saes ha raised heir

minimum wages beore hese changes came ino effec our regression analysis could

(misakenly) atribue ha reducion o he minimum wage We check or such

pre-rends by inroducing variables ha represen he prior yearrsquos value or leads o

he minimum wage I he model esimaes he minimum wage o have an effec on

he oucome variable beore he wage change wen ino effec hen an unobservedacor no he minimum wage change caused he change in SNAP aciviy

We es he specificaions above or pre-rends by including a one-year lead in

all hree specificaions We find ha he lead erms are small posiive and no

saisically significan indicaing ha he concurren minimum wage991252no

he wage level in prior periods991252is driving he observed changes in SNAP

oucomes28 In paricular he coefficien (sandard error) on he lead erm in

our preerred amily-level enrollmen regression is 011 and no significan

while he coefficien and sandard error o he conemporaneous minimum

wage is unchanged In he sae-level preerred enrollmen regression he

coefficien o he lead erm is again small (07) and i is no significan Te

corresponding coefficien on he lead erm in he sae-level expendiure

regression is 16 and is no significan Te posiive poin esimaes on hese lead

erms resuls no only rule ou disoring negaive pre-rends Tey also sugges

ha our main resuls may underesimae he rue effecs

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2743

24 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2843

Appendix B | wwwamericanprogresso

Appendix B

Model selection process

For boh he amily-level and sae-level models we es hree mehods o conrol

or unobserved geographic- and ime-varying characerisics as suggesed by he

minimum wage lieraure Firs we include only independen sae-specific fixed

effecs and year-specific fixed effecs Tis specificaion (specificaion 1) implicily

assumes ha amilies in any sae consiue an equally good saisical ldquoconrolrdquo

group or hose in any randomly chosen sae afer accouning or various

characerisics (median income and unemploymen rae among ohers) Similarly

simple ime fixed effecs assume ha amilies surveyed in any year can crediblyserve as a conrol group or amilies surveyed in every oher year o he sample

(1990 hrough 2012)

In oher words specificaion 1 assumes ha a saersquos immediae neighbor provides

no beter a couneracual or he effec o a minimum wage change han does a

sae across he counry We relax his resricive specificaion sequenially in wo

seps In specificaion 2 we replace simple year fixed effecs wih fixed effecs or

each Census divisionyear (capured as an addiional variable in he vecor By

using division-year effecs we remove he resricion ha amilies in each sae

are equally good saisical conrols or all oher amilies Raher we allow or he

possibiliy ha amilies in similar geographic regions (or example he Souh or

he Norheas) may be more similar o one anoher han amilies arher away

Finally in specificaion 3 we add sae-specific linear ime rends o he previous

specificaion Tus specificaion 3 is he mos rigorous model specificaion in ha

i allows or heerogeneiy along hree dimensions Ta is specificaion 3 allows

each sae o have is own ime-varying rends raher han imposing he resricion

ha saes evolve idenically over he 22 years in he sample

We begin building he heoreical specificaion above rom a se o simpleuncondiional models regression o SNAP aciviy (enrollmen or expendiures)

on he log o he minimum wage and a se o geographic- and ime-specific

effecs (specificaions 1 2 and 3 described above) As shown in ables 1ndash3 (or

specificaion 3) we hen add covariaes sequenially o hese models including

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2943

26 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

firs he vecor o amily-level conrols ollowed by each o several sae-level

covariaes in urn he unemploymen rae log o median-amily income and he

employmen-o-populaion raio Comparable resuls or specificaions 2 and 3

will be available in our orhcoming working paper

As expeced he simple uncondiional models indicae ha he relaionship beween he minimum wage and SNAP enrollmen i one exiss is a more complex

one influenced by oher acors In he uncondiional model he coefficien on he

variable o ineres991252he log o he minimum wage991252is small in magniude and no

saisically differen rom zero Once we accoun or he influence o labor marke

condiions and variaion in income levels on program paricipaion (by including

unemploymen rae and median-amily income conrol variables respecively)

he effec o he minimum wage on SNAP enrollmen is precisely esimaed Te

coefficien o he log minimum wage is slighly higher (-0042) in he amily-level

analysis han he coefficien (-031) in he sae-level analysis Te level o precision

is also higher in he amily-level analysis Tis is o be expeced when using 124million observaions compared o 1127

Te second dimension o model choice concerns he effec specificaion ables

7ndash9 compare he primary coefficiens o ineres or he SNAP enrollmen and

expendiure models For boh he enrollmen models he effec sizes are smalles

or specificaion 1 larges or specificaion 2 and inermediae beween hese wo

in specificaion 3 Recall ha Specificaion 3 conains sae-specific linear ime

rends in addiion o he census divisionyear conrols included in specificaion

2 In he amily-level enrollmen model he sandard error o he minimum wage

coefficien is smaller han in he oher wo specificaions Sandard errors on he

oher variables are much smaller in specificaions 2 and 3 han in specificaion

1 On he basis o coefficien significance (join and individual) specificaions 2

and 3 are sricly preerred in boh enrollmen models o specificaion 1 which

conains only sae and year fixed effecs

A concern wih specificaions 2 and 3 is ha rend conrols such as sae linear

rends may incorrecly absorb some o he delayed impac o a minimum wage

When we es his issue by including lagged minimum wages we do no find ha

delayed effecs are significan Anoher concern is ha more sauraed modelsuse less o he saisical variaion which could reduce he saisical power o

he resuls However he sandard errors or our more sauraed models are no

higher and are lower in some cases han or he less sauraed models Overall

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3043

Appendix B | wwwamericanprogresso

his evidence suppors our use o he sauraed model as he preerred model

specificaion Moreover Dubersquos 2013 sudy shows ha more sauraed models

perorm beter han models wih jus sae and ime fixed effecs

Te esimaed enrollmen regressions a boh he amily and sae levels show large

and saisically significan coefficiens Te esimaed minimum wage effec in heexpendiures regressions991252or which we have only sae-level daa991252is also large

and saisically significan

We do no use weighed regression or he sae-level models preerring o keep

analysis o he ldquoreamenrdquo (ha is o say a minimum wage change) appropriae

o he average sae raher han he average amily or individual I insead our

primary ineres were he impac o a minimum wage change on he average amily

or he average individual we migh choose o designae he number o amilies

in each sae or he sae populaion respecively as analyic weighs in order o

obain a coefficien beter suied or such inerence

TABLE 4

SNAP enrollment

Family-level linear probability

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0048 -0047 -0040 -0043 -0042

(0013) (0013) (001) (0008) (0009)

Unemployment rate (100) 0505 0420 0280

(0083) (0086) (0082)

Log median income -0057 -0039

(0011) (0011)

Employment-to-population ratio -0239

(0038)

N 1242022 1242022 1242022 1242022 1242022

plt01 plt005 plt001

Note Robust standard errors are in parentheses Observations are clustered at the state level The outcome variable is binary and equal to one if a family is enrolledin SNAP All models include state fixed effects Census division x-year fixed effects and state-specific linear time trends All specifications except 3a include additional

controls for family size race and marital status of the family head presence of children and presence of an adult maleSource Annual data from the March Current Population Survey for 1990 to 2012 estimation includes Current Population Survey probability weights

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3143

28 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

TABLE 5

SNAP enrollment

State-level linear regression

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0028 -0024 -0026 -0031 -0031

(0014) (0013) (0013) (0012) (0012)

Unemployment rate (100) 0625 0466 0320 0339

(0087) (0088) (0085) (0083)

Log median income -0090 -0065 -0061

(0013) (0013) (0013)

Employment-to-population ratio -0282 -0248

(0037) (0038)

N 1127 1127 1127 1127 1127

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the SNAP enrollment rate All models include state fixed effects Census division x-year fixedeffects and state-specific linear time trends

Source Annual data from the March Current Population Survey for 1990 to 2012

TABLE 6

SNAP expenditures

State-level linear regression

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0181 -0149 -0156 -0153 -0190

(011) (0103) (0102) (0103) (0103)

Unemployment rate (100) 4587 4152 4232 4313

(0622) (0621) (0633) (0628)

Log median income -0246 -0261 -0294

(0075) (0078) (0078)

Employment-to-population ratio 0155 0244

(0237) (024)

N 1127 1127 1127 1127 1127

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the natural log of state SNAP expenditures per capita for 1990 to 2012 All models include state

fixed effects Census division x-year fixed effects and state-specific linear time trends

Source Annual data from Bureau of Economic Analysis national income and product accounts tables for 1990 to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3243

Appendix B | wwwamericanprogresso

TABLE 7

Comparison of specifications SNAP enrollment

Family-level linear probability

(1) (2) (3)

Log minimum wage -0028 -0049 -0042

(0014) (0017) (0009)

Unemployment rate (100) 0275 0297 0280

(0161) (0076) (0082)

Log median income -0077 -0055 -0039

(0014) (0012) (0011)

Employment-to-population ratio -0238 -0250 -0239

(0054) (004) (0038)

N 1242022 1242022 1242022

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses Observations are clustered at the state level The outcomevariable is binary or equal to one if a family is enrolled in SNAP All specifications include additional controls forfamily size race and marital status of the family head presence of children and presence of an adult male

Source Annual data from the March Current Population Survey for 1990 to 2012 estimation includes CurrentPopulation Survey probability weights

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3343

30 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

TABLE 8

Comparison of specifications SNAP enrollment

State-level linear regression

(1) (2) (3)

Log minimum wage -0019 -0035 -0031

(0009) (0012) (0012)

Unemployment rate (100) 0401 0370 0339

(0063) (0077) (0083)

Log median income -0081 -0073 -0061

(0011) (0013) (0013)

Employment-to-population ratio -0183 -0222 -0248

(0039) (0039) (0038)

N 1127 1127 1127

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the SNAP enrollment rate Allregressions include the share of households with children and the racial shares of the population

Source Annual data from the March Current Population Survey for 1990 to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3443

Appendix B | wwwamericanprogresso

TABLE 9

Comparison of specifications SNAP expenditures

State-level linear regression

(1) (2) (3)

Log minimum wage -0121 -0203 -0190

(0075) (0103) (0103)

Unemployment rate (100) 5292 5152 4313

(0464) (0576) (0628)

Log median income -0437 -0417 -0294

(008) (0086) (0078)

Employment-to-population ratio -0040 -0220 0244

(0261) (0260) (0240)

N 1127 1127 1127

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the natural log of state SNAPexpenditures per capita All regressions include the share of households with children and the racial shares ofthe state population

Source Annual data from Bureau of Economic Analysis national income and product accounts tables for 1990to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3543

32 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3643

Appendix C | wwwamericanprogresso

Appendix C Harkin-Miller

policy simulation results

StateMinimum wage

(2014)

SNAP enrollment (2012) Predicted change in SNAP enrollment

RateTotal

(persons)

Model 1 family

enrollment (linearprobability)

Model 2 state

enrollment (linearregression)

Model 3 s

expenditu(linear regre

Alabama $725 164 788682 -66922 -58897 -58906

Alaska $775 120 87436 -8104 -6887 -3288

Arizona $790 201 1319323 -67435 -56738 -64356

Arkansas $725 78 230489 -40977 -36063 -35248

California $800 206 7813680 -371131 -310222 -18223

Colorado $800 164 853155 -50684 -42365 -23926

Connecticut $870 91 326621 -22456 -17975 -13711

Delaware $725 186 170262 -12739 -11211 -10647

District of Columbia $825 133 84009 -5370 -4417 -3632

Florida $793 166 3208026 -195813 -164426 -13046

Georgia $725 160 1586336 -137741 -121224 -11004

Hawaii $725 96 133662 -19310 -16995 -14933

Idaho $725 92 147501 -22165 -19507 -15809

Illinois $825 95 1225084 -109088 -89742 -70955

Indiana $725 125 816233 -90818 -79928 -83985

Iowa $725 155 478011 -42716 -37594 -28556

Kansas $725 135 388269 -40082 -35275 -27461

Kentucky $725 130 568821 -60840 -53544 -52259

Louisiana $725 149 683832 -63929 -56263 -66083

Maine $750 77 101976 -16567 -14323 -15234

Maryland $725 144 846415 -81748 -71946 -38370

Massachusetts $800 130 864721 -64902 -54251 -42913

Michigan $740 146 1439141 -128801 -112140 -11022

Minnesota $725 133 713646 -74730 -65769 -37878

TABLE 10

SNAP enrollments Predicted changes by state under the Harkin-Miller bill

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3743

34 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

StateMinimum wage

(2014)

SNAP enrollment (2012) Predicted change in SNAP enrollment

RateTotal

(persons)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 state

expenditures

(linear regression)

Mississippi $725 129 386501 -41486 -36511 -46467

Missouri $750 172 1036182 -75131 -64952 -56244

Montana $790 132 132452 -10350 -8708 -5846

Nebraska $725 124 230683 -25773 -22683 -12189

Nevada $825 162 446035 -23349 -19209 -11894

New Hampshire $725 127 168404 -18359 -16157 -5735

New Jersey $825 160 1416666 -75175 -61843 -28236

New Mexico $750 149 310896 -25983 -22463 -22512

New York $800 192 3763553 -191193 -159815 -142182

North Carolina $725 174 1697193 -135417 -119179 -113503

North Dakota $725 87 61225 -9743 -8574 -4021

Ohio $795 143 1647345 -115869 -97169 -88580

Oklahoma $725 129 494053 -53006 -46650 -46854

Oregon $910 124 485326 -17036 -13328 -16398

Pennsylvania $725 161 2053643 -177315 -156052 -125586

Rhode Island $800 156 163730 -10258 -8574 -8698

South Carolina $725 94 445277 -65614 -57746 -50304

South Dakota $725 208 173749 -11586 -10197 -7458

Tennessee $725 142 914903 -89667 -78915 -99134

Texas $725 110 2863779 -362018 -318607 -253285

Utah $725 88 251107 -39658 -34902 -19390

Vermont $873 156 97792 -3823 -3055 -2475

Virginia $725 101 829771 -113723 -100086 -58212

Washington $932 72 496934 -23221 -17947 -17756

West Virginia $725 58 107875 -25792 -22699 -21665

Wisconsin $725 75 427822 -79521 -69986 -53210

Wyoming $725 164 94590 -8010 -7050 -3104

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3843

Appendix C | wwwamericanprogresso

StateMinimum wage

(2014)

SNAP expenditures (2012)Predicted change in SNAP expenditures

(millions of dollars)

Per resident

(millions of

dollars)

Total

(millions of

dollars)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 s

expenditu

(linear regre

Alabama $725 $290 $1397 -$1185 -$1043 -$1043

Alaska $775 $253 $185 -$262 -$223 -$106

Arizona $790 $257 $1686 -$935 -$787 -$892

Arkansas $725 $248 $733 -$636 -$560 -$547

California $800 $189 $7164 -$7276 -$6082 -$3573

Colorado $800 $157 $816 -$862 -$721 -$407

Connecticut $870 $191 $686 -$343 -$275 -$210

Delaware $725 $250 $229 -$205 -$180 -$171

District of Columbia $825 $366 $232 -$146 -$120 -$99

Florida $793 $294 $5676 -$4429 -$3719 -$2951

Georgia $725 $317 $3140 -$2936 -$2584 -$2346

Hawaii $725 $335 $465 -$449 -$395 -$347

Idaho $725 $225 $359 -$376 -$331 -$268

Illinois $825 $249 $3200 -$2096 -$1725 -$1364

Indiana $725 $220 $1439 -$1162 -$1023 -$1075

Iowa $725 $192 $589 -$658 -$579 -$440

Kansas $725 $159 $460 -$502 -$441 -$344

Kentucky $725 $298 $1303 -$1133 -$997 -$973

Louisiana $725 $315 $1450 -$1047 -$922 -$1083

Maine $750 $281 $373 -$267 -$231 -$246

Maryland $725 $188 $1109 -$1765 -$1553 -$828

Massachusetts $800 $206 $1366 -$1030 -$861 -$681

Michigan $740 $300 $2963 -$2400 -$2090 -$2054

Minnesota $725 $140 $755 -$1113 -$980 -$564

Mississippi $725 $326 $973 -$649 -$571 -$726

Missouri $750 $241 $1452 -$1278 -$1104 -$956

Montana $790 $190 $191 -$179 -$151 -$101

Nebraska $725 $140 $259 -$409 -$360 -$194

Nevada $825 $191 $527 -$441 -$363 -$225

New Hampshire $725 $126 $167 -$399 -$351 -$125

TABLE 11

SNAP expenditures Predicted changes by state under the Harkin-Miller bill

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3943

36 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

StateMinimum wage

(2014)

SNAP expenditures (2012)Predicted change in SNAP expenditures

(millions of dollars)

Per resident

(millions of

dollars)

Total

(millions of

dollars)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 state

expenditures

(linear regression)

New Jersey $825 $160 $1420 -$1611 -$1325 -$605

New Mexico $750 $324 $675 -$513 -$444 -$445New York $800 $287 $5616 -$3766 -$3148 -$2801

North Carolina $725 $252 $2454 -$2187 -$1925 -$1833

North Dakota $725 $128 $90 -$162 -$143 -$67

Ohio $795 $259 $2995 -$2013 -$1688 -$1539

Oklahoma $725 $248 $945 -$799 -$703 -$706

Oregon $910 $322 $1255 -$272 -$213 -$262

Pennsylvania $725 $218 $2779 -$2930 -$2579 -$2075

Rhode Island $800 $280 $294 -$173 -$144 -$147

South Carolina $725 $291 $1373 -$1337 -$1177 -$1025South Dakota $725 $198 $165 -$192 -$169 -$123

Tennessee $725 $324 $2091 -$1413 -$1243 -$1562

Texas $725 $230 $5997 -$6402 -$5634 -$4479

Utah $725 $141 $402 -$614 -$541 -$300

Vermont $873 $230 $144 -$66 -$53 -$43

Virginia $725 $173 $1413 -$2062 -$1815 -$1056

Washington $932 $244 $1682 -$350 -$270 -$267

West Virginia $725 $273 $508 -$451 -$397 -$379

Wisconsin $725 $204 $1166 -$1302 -$1146 -$871Wyoming $725 $95 $55 -$105 -$93 -$41

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4043

Appendix C | wwwamericanprogresso

FIGURE 2

Raising the minimum wage to $1010would cut taxpayer costs in every state

Predicted decreases in cost and enrollment

in SNAP in 50 states

$200+$51ndash$100

$101ndash$200

0ndash$25

$26ndash$50

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4143

38 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

Endnotes

1 US Department of Agriculture ldquoSupplementalNutrition Assistance Program (SNAP)rdquo available athttpwwwfnsusdagovsnapeligibility (last accessedFebruary 2014)

2 For this initial analysis we do not consider Harkin-Millerrsquos increase in subminimum wages for tippedworkers To do s o would increase the estimated SNAP

savings by an unknown amount

3 The Congressional Budget Office estimates thatworkers currently earning between $1010 and $1150per hour would see their wages rise under the Harkin-Miller proposal Congressional Budget O ffice ldquoTheEffects of a Minimum Wage Increase on Employmentand Family Incomerdquo (2014)

4 Marianne Page Joanne Spetz and Jane Millar ldquoDoesthe Minimum Wage Affect Welfare Caseloadsrdquo Journalof Policy Analysis and Management 24 (2) (2005)273ndash295

5 Marianne Bitler and Hilary Hoynes ldquoThe More ThingsChange the More They Stay the Same The SafetyNet Living Arrangements and Poverty in the GreatRecessionrdquo NBER Working Paper 194 49 2013

6 Sylvia Allegretto and others ldquoFast Food Poverty Wages The Public Cost of Low-Wage Jobs in the Fast-FoodIndustryrdquo (Berkeley California Center for LaborResearch and Education 2013) available at httplaborcenterberkeleyedupubliccostsfast_food_poverty_wage

7 David Neumark and William Wascher ldquoDoes a HigherMinimum Wage Enhance the Effectiveness of theEarned Income Tax Creditrdquo Industrial and LaborRelations Review 64 (4) (2011) 712ndash746

8 David Lee and Emmanuel Saez ldquoOptimal MinimumWage Policy in Competitive Labor Marketsrdquo Journal ofPublic Economics 96 (9) (2012) 739ndash749

9 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family IncomerdquoFebruary 2014

10 Jesse Rothstein ldquoIs the EITC as Good as an NITConditional Cash Transfers and Tax Incidencerdquo AmericanEconomic Journal Economic Policy 2 (1) (2010) 177ndash208

11 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family Incomerdquo

12 Dube Arindrajit 2013 rdquoMinimum Wagesand the Distribution of Family IncomerdquoUnpublished working paper Available at httpsdldropboxusercontentcomu15038936Dube_MinimumWagesFamilyIncomespdf

13 As Dube explains in ldquoThe poverty of Minimum WageFactsrdquo the simulation approach underestimate stemsfrom a number of unwarranted assumptions includingthe range of actual wage increases and the accuracy ofwage data in the Current Population Survey The causal

approach does not make these assumptions

14 Allegretto Sylvia and others 2013 ldquoCredible ResearchDesigns for Minimum Wage Studiesrdquo Working Paper148-13 University of California Berkeley Institutefor Research on Labor and Employment Available athttpwwwirleberkeleyeduworkingpapers148-13pdf

15 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family In comerdquo

16 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo (2012)

17 Ibid

18 Ibid

19 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo

20 US Department of Agriculture ldquoSupplementalNutrition Assistance Program (SNAP)rdquo

21 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo

22 Ibid

23 We deviate from the Census Bureaursquos definition ofa family unit which is ldquotwo people or more (on e ofwhom is the householder) related by bir th marriageor adoption and residing togetherrdquo US Bureau ofthe Census ldquoCurrent Population Survey DefinitionsFamilyrdquo available at httpwwwcensusgovcpsabout

cpsdefhtml (last accessed February 2014) We countas a family unit any individual residing on his or herown two or more persons residing together whodo not belong to a family in the March CPS sampleare constructed as one family in our analysis For thepurposes of food stamp allocations the consumptionresulting from this transfer is probably distributed tofamily members (rather than household members ora single individual within the household) Howeversingle individuals canmdashand domdashreceive SNAPbenefits Excluding them would fail to make theanalysis reflective of the population at large

24 Strictly the family level linear probability modelpredicts the percentage-point decrease in theprobability that an individual family will receive SNAPpayments When applied to a large number of familieshowever we are able to interpret the coefficient asa decrease in the mean of enrollmentmdashthat is a

decrease in the enrollment ratemdashby applying the lawof iterated expectations

25 We generate expenditure predictions from theenrollment modelsmdashand conversely generateenrollment predictions from the expenditure modelmdashby assuming that expenditures per enrolled familyremains the same before and after the minimum wagechange In practice this is likely to be a conservativeestimatemdashthat is to underestimate the decrease inSNAP activity Average SNAP benefits per family willalso decrease as many families that remain eligible forSNAP experience income gains

26 Wage and Hour Division ldquoMinimum Wage Laws inthe States ndash Januar y 1 2014rdquo available at httpwwwdolgovwhdminwageamericahtm (last accessedFebruary 2014)

27 See for example Sylvia Allegretto and others ldquoCredibleResearch Designs for Minimum Wage Studiesrdquo WorkingPaper 148-113 (Berkeley California Institute forResearch on Labor and Employment 2013) available athttpwwwirleberkeleyeduworkingpapers148-13pdf

28 We will report these results in a forthcoming workingpaper

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4243

Endnotes | wwwamericanprogresso

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4343

The Center for American Progress is a nonpartisan research and educational institute

dedicated to promoting a strong just and free America that ensures opportunity

for all We believe that Americans are bound together by a common commitment to

these values and we aspire to ensure that our national policies reflect these values

We work to find progressive and pragmatic solutions to significant domestic and

international problems and develop policy proposals that foster a government that

is ldquoof the people by the people and for the peoplerdquo

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1343

10 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

o a minimum wage policy change Tus i a new minimum wage policy were

implemened in one sae only he researchers could atribue all o he difference

hey observe in SNAP aciviy o he new minimum wage policy

In oher words we approximae he ideal experimen by using non-experimenal

saisical mehods Te desirable ldquopre-exising similariiesrdquo beween saes ha wehave defined above inorm our choice o conrol variables in a saisical seting

More precisely in our muliple regression models we use median amily income

he unemploymen rae he employmen-o-populaion raio and regional and

ime idenifiers o consruc an appropriae group o peers or each sae on he

eve o a policy change

Data description

wo daa ses include inormaion abou boh income and paricipaion inpublic programs Te Survey o Income and Program Paricipaion or SIPP

which is conduced in inermiten years has he advanage o ollowing he

same individuals over a period o ime In oher words i is a longiudinal daa

se I also has he advanage o conaining monhly daa However he sample

size o he SIPP is no sufficien or analyzing variaions in sae-level minimum

wages Te March CPS has he advanage o a much larger sample size and i is

conduced annually wihou any breaks in ime I has he disadvanage o being

a cross-secional daa se so we canno ollow he same individuals over ime991252

sricly speaking over more han one year On ne he March CPS is much more

suiable or our sudy We examine he empirical relaionship beween minimum

wage policy and ood samp aciviy a wo levels o aggregaion he amily level

and he sae level Family-level daa are drawn rom he March CPS

Te March CPS comprises responses rom he residens o 50000 o 60000

dwelling places surveyed per year and conains deailed inormaion on he

residensrsquo employmen and income including income rom ranser paymens

Te sample or our analysis comprises more han 128 million amily unis during

he period rom 1990 o 2012 (inclusive) Survey weighs allow us o analyze

SNAP paricipaion in a manner ha is represenaive o he US populaion alarge Over all years he share o amilies reporing ood samp receip in he

weighed March CPS sample is 91 percen Te enrollmen rae was a a low o 6

percen in he year 2000 In 2012 he mos recen year in our panel 133 percen

o amilies repored paricipaing in SNAP a some poin during he survey year

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1443

Methods and data | wwwamericanprogresso

Te March CPS also collecs inormaion on he number o SNAP recipiens in

he household he number o monhs he household receives SNAP benefis

and he cash-equivalen value o he SNAP benefis received However he

value o SNAP benefis is severely underrepored among recipiens perhaps

because recipiens are unaware o he exac monhly cash-equivalen value o

benefis hey receive

Our firs empirical sraegy ocuses solely on SNAP enrollmen By using he

amily as he uni o analysis we are able o inser saisical conrols o accoun

or non-wage-relaed acors ha influence any paricular amilyrsquos likelihood o

program paricipaion wih he inenion o isolaing any differences in program

paricipaion ha are due purely o changes in wage policy Tis approach

idenifies he effecs o low-wage labor policy on he exernal margin991252ha is

he effec o he minimum wage on he likelihood ha a amily paricipaes in he

SNAP program a all991252as opposed o he inernal margin or how much SNAP

unding he amily would receive

Our second empirical ramework uses sae-level adminisraive daa Ta is

we aggregae he daa o obain a single daa poin or each saeyear back

o 1990 represening he mean o he oucome or he sae Te sae-level

esimaion serves as a robusness check on he amily-level resuls or SNAP

paricipaion Also using aggregaed daa allows us o esimae direcly he

causal effec o minimum wage changes on SNAP spending Tis is no possible

a he amily level as discussed above daa on cash-equivalen value o ood

samps or SNAP recipiens is very requenly no repored in he March CPS

and when i is repored he inormaion may be unreliable By conras he

Bureau o Economic Analysis publishes aggregae SNAP spending a he sae

level in is Naional Income and Produc Accoun or NIPA ables Tus

while we are unable o observe he heerogeneiy in he cash value o SNAP

or amilies in each sae we are able o calculae average SNAP spending

per residen in each sae per year Supporing covariaes include he annual

unemploymen and employmen daa rom he Bureau o Labor Saisics or

BLS and sae-level populaion series rom he iner-decennial census releases

Minimum wage daa are available rom he BLSrsquos wages and hours division For

sae minimum wage changes enaced a oher imes han he firs o he yearan average value or he year is used

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1543

12 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

Family level model specifications

We firs examine he effec o he minimum wage on paricipaion in public

assisance programs For amily residing in sae and during year we esimae

an equaion o he ollowing orm

(1)

is a binary variable ha is se equal o 1 i a leas one member o amily

received ood samps during he survey year is a se o sae-level

characerisics including annual averages o he unemploymen rae he

employmen-o-populaion raio and he naural log o median amily income

is a vecor o amily atribues including indicaors or he race and marial saus

o he amily head size o he amily he presence o children and he presence oan adul male Sae fixed effecs are capured by o conrol or ime-varying

heerogeneiy our preerred model specificaion also includes year fixed effecs

ha vary by Census division ( ) and sae-level linear ime rends In

Appendix B we jusiy he inclusion o hese las wo erms We also compare he

resuls rom our preerred specificaion wih less sauraed specificaions

Te effec o ineres which is capured by is he expeced change in he

probabiliy o receiving SNAP benefis wih respec o a change in he (log o he)

binding minimum wage in sae during year We repor robus sandard errors

clusered a he sae level We esimae he parameers using linear regression

producing a linear probabiliy model Deails o he model selecion process are

covered in Appendix B below

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1643

Methods and data | wwwamericanprogresso

State-level model specifications

Te sae-level models are similarly specified For sae in year we assume ha

(2)

In his model is now eiher he SNAP enrollmen rae in sae during year

or he naural logarihm o per capia SNAP expendiures in sae during

year is once again a se o sae-level characerisics including he same

sae-level covariaes as in he amily regressions (annual average unemploymen

rae employmen-o-populaion raio naural log o median amily income)

wih he addiion o amily level characerisics averaged across he sae (average

amily size and he shares o populaion consiued by each o five racialehnicgroups) Sae fixed effecs are represened by As above our preerred model

specificaion includes year fixed effecs ha vary by Census division ( ) and

sae-level linear ime rends as elaboraed in Appendix B Te effec o

ineres is capured by

We esimae boh sae-level models (enrollmen and expendiures) using

ordinary leas squares regression Tus he inerpreaion o he coefficien is no

longer ha o a change in probabiliy as in he binary oucome models described

above Raher or he sae-level SNAP enrollmen model represens he

expeced change (in percenage poins) in he saersquos SNAP enrollmen rae ha

is due o a 1 percen change in he minimum wage For he SNAP expendiures

model is simply he elasiciy o SNAP spending wih respec o he minimum

wage991252ha is he percenage change in sae expendiures expeced o resul rom

a 1 percen change in ha saersquos minimum wage For urher deails on model

specificaion reer o Appendix B below

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1743

14 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1843

Results | wwwamericanprogresso

Results

Estimated minimum wage effects on SNAP enrollment and

expenditures

able 1 shows he esimaed parameer o ineres991252he coefficien o he

minimum wage or he preerred model o each ype Coefficiens on he

minimum wage variable are no direcly comparable across models because all

our models have a differen uncional orm o undersand and compare hese

esimaes we compue he change in SNAP aciviy prediced or a paricular wage scenario Te final column in able 1 answers he quesion Wha would be

he expeced change in SNAP aciviy in response o a 10 percen increase in he

minimum wage Te answer o his quesion varies wih he value o he inpu

parameers in he able we calculae he percenage decrease in enrollmen or

expendiures prediced or he average sae wih a minimum wage o $725 in

2014 Te sae-level SNAP expendiure model which is a consan-elasiciy

model conveys elasiciy inormaion direcly or he change in expendiures per

capia in he sae

TABLE 1

Comparison of national SNAP predictions for a 10 percent increase in the federal minimum wage

Model LevelRegression

type

Predicted outcome Coefficient of log

(minimum wage)

(Standard error)

Effect of a 10 percent in

in the minimum wage

VariableForm of

variable

Total

enrollment

Tot

expend

1 Family Linear

probability Enrollment Binary (enrolled=1)

-0042

(0008) -317 N

2

StateLinear regression

(ordinary least

squares)

Enrollment State enrollment rate (percent) -0031

(0012) -235 N

3 Expenditures Log (state expenditures per capita) -0190

(0103) NA -19

plt01 plt005 plt001Note Predicted changes are calculated for the average state with a minimum wage of $725 in 2014

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1943

16 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

According o his model a 10 percen increase in he minimum wage would resul

in a 19 percen reducion in SNAP expendiures Tis resul is reassuringly similar

o Dubersquos resul or povery reducion Te wo enrollmen models are much more

precisely esimaed han he expendiure model Te sae-level enrollmen model

finds ha a 10 percen minimum wage increase in a low-wage sae is associaed wih

a 235 percen decrease in SNAP enrollmens Te amily-level linear probabiliymodel predics a somewha greaer elasiciy or low minimum wage saes an

increase o 10 percen in he ederal minimum would resul in a 317 percen decline

in SNAP enrollmen24 Te differences in hese esimaes sem rom a number o

acors including difference in model uncional orm and daa used We rea his

range o elasiciy esimaes as an upper and lower bound on enrollmen impacs

Harkin-Miller bill National and state-level predicted impacts

Wha would be he prediced change or he SNAP program i he ederalminimum were raised o $1010 as proposed in he Harkin-Miller bill In order o

make his inerence we accoun or he ac ha no all saes are currenly subjec

o he ederal minimum wage a he beginning o 2014 21 saes mainained

higher minimum wages han $725 In hose saes an increase in he ederal

minimum wage may or may no be binding or employers in he sae depending

upon wheher he new ederal minimum exceeds he sae-level minimum Bu

regardless o wheher a minimum wage change is binding he impac on SNAP

aciviy will be lower in high minimum wage saes In order o accoun or his

properly we calculae sae by sae he percenage wage change ha would resul

rom he Harkin-Miller proposal and apply he parameers rom each o he hree

models above o compue he expeced decrease in SNAP aciviy or each sae

In his exercise we use saesrsquo curren (2014) minimum wage levels and assume

as a baseline he 2012 levels o SNAP enrollmen and expendiure as 2012 is he

mos recen year or which SNAP daa are available

able 10 and able 11 in Appendix C repor he esimaed effecs on SNAP

enrollmen and expendiures respecively or each sae under he Harkin-Miller

bill25 An increase o $1010 i enaced oday would represen beween a 393

percen wage increase in a $725 minimum wage sae and an 84 percen increasein Washingon sae which has he highes minimum wage in he naion a $932

as o January 201426 Slighly more han 56 percen o he decrease in expendiures

and abou 59 percen o he decrease in enrollmen would occur in saes wih

presen-day minimum wages o $725

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2043

Results | wwwamericanprogresso

In 2012 hese saes were home o 46 percen o he American populaion

and accouned or a roughly equivalen percenage o oal naional SNAP

expendiures No surprisingly he larges enrollmen decreases come rom

populous saes wih very high SNAP enrollmen raes andor very low minimum

wages Te larges prediced enrollmen reducion991252beween 319000 individuals

and 362000 individuals991252would occur in exas which has a $725 minimum wage In Caliornia which has a 206 percen SNAP paricipaion rae and an $8

per hour minimum wage we anicipae SNAP enrollmen declines o 310000

persons o 371000 persons And in Florida which had a SNAP paricipaion rae

o 166 percen and a $793 minimum wage enrollmen could decline by beween

164000 individuals and 196000 individuals For he our saes ha ogeher

accouned or he greaes amoun o SNAP spending in 2012991252exas Caliornia

Florida and New York respecively991252he combined expendiure reducion rom

he Harkin-Miller bill is prediced o be $14 billion

able 2 summarizes he prediced declines in SNAP aciviy or he naion as a whole ha would resul rom he direc and indirec effecs o he Harkin-Miller

bill Enrollmen would all beween 31 million persons and 36 million persons

represening 75 percen o 87 percen o curren enrollmen Te anicipaed

reducion in program expendiures would be nearly $46 billion or 61 percen o

program expendiures

TABLE 2

Comparison of national SNAP predictions under the Harkin-Miller billrsquos $1010 minimum wage

Model

Enrollment

(persons)

Expenditures

(millions of dollars)

Current (2012) Predicted Change Current (2012) Predicted Chan

Family enrollment (linear probability)

41866195

45489339 -3623144

$74861

NA NA

State enrollment (ordinary least squares) 38745435 -3120759 NA NA

State expenditures (ordinary least squares) NA NA $70305 -$45

Note Calculations use 2014 state minimum wages and the most recent SNAP data from 2012 They assume that per-enrollee expenditures remain constant

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2143

18 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

Tere are o course oher possibiliies or minimum wage legislaion able 3

shows he expeced SNAP changes or he Unied Saes under a variey o wage

scenarios calculaed using he sae-level models I saes were no able o se

heir minimum wages independenly such ha all saes were consrained by

he ederal minimum o $725 SNAP would be received by abou 514000 more

people across he Unied Saes a an addiional program cos o nearly hree-quarers o a billion dollars In conras he effecs o a higher minimum wage

proposal991252a ederal wage floor o $11 per hour991252would decrease enrollmen in

SNAP by more han 10 percen and decrease program coss by 83 percen

TABLE 3

Summary of par ticipation and expenditures under wage scenarios

If all states had

minimum wages of

Enrollment(persons)

Expenditures

(millions of dollars)

Predicted Change Predicted Change

Recent levels (2014) 41866195 $74861

$725 42380520 514326 $75604 $743

$800 41423919 -442276 $74209 -$652

$900 40148451 -1717744 $72350 -$2511

$1000 38872982 -2993212 $70490 -$4371

$1010 38745435 -3120759 $70305 -$4556

$1100 37597514 -4268681 $68631 -$6230

Note Calculations use state-level enrollment model coefficient

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2243

Conclusion | wwwamericanprogresso

Conclusion

An exensive body o lieraure examines employmen effecs o he minimum

wage A much smaller se o sudies examines how he minimum wage affecs

povery and only a handul o papers examine he effecs o he minimum wage

on he EIC Our analysis is he firs o examine he effecs o he minimum

wage on SNAP

Our findings indicae ha increased earnings rom minimum wage changes do

reduce SNAP enrollmens and expendiures We esimae ha he Harkin-Miller bill would save axpayers nearly $46 billion per year equivalen o 61 percen

o SNAP expendiures in 2012 he las year or which daa are available Over a

10-year period he esimaed savings amoun o nearly $46 billion

Our repor is subjec o limiaions ha we expec o overcome in our uure

research Firs he findings do no ake ino accoun possible ineracions among

SNAP he EIC and Medicaid Te eligibiliy cuoffs among hese programs

are quie differen suggesing ha such ineracions may be minor Noneheless

he join effecs can only be deermined by urher research using a causal

model Second i would be useul o know he disribuion o SNAP reducions

along he wage disribuion Using he Congressional Budge Officersquos calculaions

o how much he oal dollar value o wage would increase under he Harkin-

Miller proposal our findings imply ha he decline in overall SNAP spending

equals abou 15 percen o he oal resuling increase in wages Te amoun and

disribuion o his offse are o considerable ineres Minimum wage beneficiaries

who come rom working amilies already well above he povery line would no

see any offse while hose who are currenly considerably below he povery line

will see larger offses Tese issues will also be a subjec or our uure research

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2343

20 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

About the authors

Rachel West is a maser o public policy candidae a he Goldman School

o Public Policy Universiy o Caliornia Berkeley Her research ocuses on

economic policy in he areas o low-wage labor and povery

Michael Reich is proessor o economics and direcor o he Insiue or

Research on Labor and Employmen a he Universiy o Caliornia a Berkeley

His research publicaions cover numerous areas o labor economics including

racial inequaliy labor marke segmenaion high-perormance workplaces

union-managemen cooperaion Japanese labor-managemen sysems living

wages and minimum wages He received his docorae in economics rom

Harvard Universiy

Acknowledgments

We are graeul o Sylvia Allegreto Arindraji Dube Bill Leser Jesse Rohsein

Daniel Tompson and Ben Zipperer or heir valuable suggesions

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2443

References | wwwamericanprogresso

References

Allegreto Sylvia and ohers 2013 ldquoFas Food Povery Wages Te Public Cos o Low-Wage Jobsin he Fas-Food Indusryrdquo Universiy o Caliornia Berkeley Cener or Labor Research andEducaion Available a htplaborcenerberkeleyedupubliccossas_ood_povery_wage

Allegreto Sylvia and ohers 2013 ldquoCredible Research Designs or Minimum Wage Sudiesrdquo

Working Paper 148-13 Universiy o Caliornia Berkeley Insiue or Research on Labor andEmploymen Available a htpwwwirleberkeleyeduworkingpapers148-13pd

Biler Marianne and Hilary Hoynes 2013 ldquo Te More Tings Change he More Tey Say heSame Te Saey Ne Living Arrangemens and Povery in he Grea Recessionrdquo Working Paper19449 Naional Bureau o Economic Research

Congressional Budge Office 2012 ldquoTe Supplemenal Nuriion Assisance Programrdquo Washingon Available a htpwwwcbogovsiesdeaulfilescbofilesatachmens04-19-SNAPpd

991252 991252 991252 2014 ldquoTe Effec o a Minimum-Wage Increase on Employmen and Family Incomerdquo Washingon Available a htpwwwcbogovsiesdeaulfilescbofilesatachmens44995-MinimumWagepd

Dube Arindraji 2013 rdquoMinimum Wages and he Disribuion o Family Incomerdquo Unpublished working paper Available a htpsdldropboxuserconencomu15038936Dube_ MinimumWagesFamilyIncomespd

991252 991252 991252 2014 ldquoTe Povery o Minimum Wage Facsrdquo Arindraji Dube Blog January 22 Availablea htparindubecom20140122he-povery-o-minimum-wage-acs

Lee David and Emmanuel Saez 2012 ldquoOpimal Minimum Wage Policy in Compeiive LaborMarkesrdquo Journal o Public Economics 96 (9) 739ndash749

Neumark David and William Wascher 1992 ldquoEmploymen Effecs o Minimum and Subminimum Wages Panel Daa on Sae Minimum Wage Lawsrdquo Industrial and Labor Relations Review 46 (1)

55ndash81

Neumark David and William Wascher 2011 ldquoDoes a Higher Minimum Wage Enhance heEffeciveness o he Earned Income ax Credirdquo Industrial and Labor Relations Review 64 (4)712ndash746

Page Marianne Joanne Spez and Jane Millar 2005 ldquoDoes he Minimum Wage Affec WelareCaseloadsrdquo Journal o Policy Analysis and Management 24 (2) 273ndash295

Rohsein Jesse 2010 ldquoIs he EIC as Good as an NI Condiional Cash ransers and ax

Incidencerdquo American Economic Journal Economic Policy 2 (1) 177ndash208

Wage and Hours Division 2014 ldquoMinimum Wage Laws in he Saes ndash January 1 2014rdquo USDeparmen o Labor (htpwwwdolgovwhdminwageamericahm [February 2014])

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2543

22 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2643

Appendix A | wwwamericanprogresso

Appendix A

Pre-trend falsification check

Recen minimum wage research27 highlighs a common flaw in previous sudies

ailure o veriy ha he oucome variable is ree o negaive pre-exising rends I

or example SNAP aciviy was already rending down in saes ha raised heir

minimum wages beore hese changes came ino effec our regression analysis could

(misakenly) atribue ha reducion o he minimum wage We check or such

pre-rends by inroducing variables ha represen he prior yearrsquos value or leads o

he minimum wage I he model esimaes he minimum wage o have an effec on

he oucome variable beore he wage change wen ino effec hen an unobservedacor no he minimum wage change caused he change in SNAP aciviy

We es he specificaions above or pre-rends by including a one-year lead in

all hree specificaions We find ha he lead erms are small posiive and no

saisically significan indicaing ha he concurren minimum wage991252no

he wage level in prior periods991252is driving he observed changes in SNAP

oucomes28 In paricular he coefficien (sandard error) on he lead erm in

our preerred amily-level enrollmen regression is 011 and no significan

while he coefficien and sandard error o he conemporaneous minimum

wage is unchanged In he sae-level preerred enrollmen regression he

coefficien o he lead erm is again small (07) and i is no significan Te

corresponding coefficien on he lead erm in he sae-level expendiure

regression is 16 and is no significan Te posiive poin esimaes on hese lead

erms resuls no only rule ou disoring negaive pre-rends Tey also sugges

ha our main resuls may underesimae he rue effecs

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2743

24 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2843

Appendix B | wwwamericanprogresso

Appendix B

Model selection process

For boh he amily-level and sae-level models we es hree mehods o conrol

or unobserved geographic- and ime-varying characerisics as suggesed by he

minimum wage lieraure Firs we include only independen sae-specific fixed

effecs and year-specific fixed effecs Tis specificaion (specificaion 1) implicily

assumes ha amilies in any sae consiue an equally good saisical ldquoconrolrdquo

group or hose in any randomly chosen sae afer accouning or various

characerisics (median income and unemploymen rae among ohers) Similarly

simple ime fixed effecs assume ha amilies surveyed in any year can crediblyserve as a conrol group or amilies surveyed in every oher year o he sample

(1990 hrough 2012)

In oher words specificaion 1 assumes ha a saersquos immediae neighbor provides

no beter a couneracual or he effec o a minimum wage change han does a

sae across he counry We relax his resricive specificaion sequenially in wo

seps In specificaion 2 we replace simple year fixed effecs wih fixed effecs or

each Census divisionyear (capured as an addiional variable in he vecor By

using division-year effecs we remove he resricion ha amilies in each sae

are equally good saisical conrols or all oher amilies Raher we allow or he

possibiliy ha amilies in similar geographic regions (or example he Souh or

he Norheas) may be more similar o one anoher han amilies arher away

Finally in specificaion 3 we add sae-specific linear ime rends o he previous

specificaion Tus specificaion 3 is he mos rigorous model specificaion in ha

i allows or heerogeneiy along hree dimensions Ta is specificaion 3 allows

each sae o have is own ime-varying rends raher han imposing he resricion

ha saes evolve idenically over he 22 years in he sample

We begin building he heoreical specificaion above rom a se o simpleuncondiional models regression o SNAP aciviy (enrollmen or expendiures)

on he log o he minimum wage and a se o geographic- and ime-specific

effecs (specificaions 1 2 and 3 described above) As shown in ables 1ndash3 (or

specificaion 3) we hen add covariaes sequenially o hese models including

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2943

26 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

firs he vecor o amily-level conrols ollowed by each o several sae-level

covariaes in urn he unemploymen rae log o median-amily income and he

employmen-o-populaion raio Comparable resuls or specificaions 2 and 3

will be available in our orhcoming working paper

As expeced he simple uncondiional models indicae ha he relaionship beween he minimum wage and SNAP enrollmen i one exiss is a more complex

one influenced by oher acors In he uncondiional model he coefficien on he

variable o ineres991252he log o he minimum wage991252is small in magniude and no

saisically differen rom zero Once we accoun or he influence o labor marke

condiions and variaion in income levels on program paricipaion (by including

unemploymen rae and median-amily income conrol variables respecively)

he effec o he minimum wage on SNAP enrollmen is precisely esimaed Te

coefficien o he log minimum wage is slighly higher (-0042) in he amily-level

analysis han he coefficien (-031) in he sae-level analysis Te level o precision

is also higher in he amily-level analysis Tis is o be expeced when using 124million observaions compared o 1127

Te second dimension o model choice concerns he effec specificaion ables

7ndash9 compare he primary coefficiens o ineres or he SNAP enrollmen and

expendiure models For boh he enrollmen models he effec sizes are smalles

or specificaion 1 larges or specificaion 2 and inermediae beween hese wo

in specificaion 3 Recall ha Specificaion 3 conains sae-specific linear ime

rends in addiion o he census divisionyear conrols included in specificaion

2 In he amily-level enrollmen model he sandard error o he minimum wage

coefficien is smaller han in he oher wo specificaions Sandard errors on he

oher variables are much smaller in specificaions 2 and 3 han in specificaion

1 On he basis o coefficien significance (join and individual) specificaions 2

and 3 are sricly preerred in boh enrollmen models o specificaion 1 which

conains only sae and year fixed effecs

A concern wih specificaions 2 and 3 is ha rend conrols such as sae linear

rends may incorrecly absorb some o he delayed impac o a minimum wage

When we es his issue by including lagged minimum wages we do no find ha

delayed effecs are significan Anoher concern is ha more sauraed modelsuse less o he saisical variaion which could reduce he saisical power o

he resuls However he sandard errors or our more sauraed models are no

higher and are lower in some cases han or he less sauraed models Overall

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3043

Appendix B | wwwamericanprogresso

his evidence suppors our use o he sauraed model as he preerred model

specificaion Moreover Dubersquos 2013 sudy shows ha more sauraed models

perorm beter han models wih jus sae and ime fixed effecs

Te esimaed enrollmen regressions a boh he amily and sae levels show large

and saisically significan coefficiens Te esimaed minimum wage effec in heexpendiures regressions991252or which we have only sae-level daa991252is also large

and saisically significan

We do no use weighed regression or he sae-level models preerring o keep

analysis o he ldquoreamenrdquo (ha is o say a minimum wage change) appropriae

o he average sae raher han he average amily or individual I insead our

primary ineres were he impac o a minimum wage change on he average amily

or he average individual we migh choose o designae he number o amilies

in each sae or he sae populaion respecively as analyic weighs in order o

obain a coefficien beter suied or such inerence

TABLE 4

SNAP enrollment

Family-level linear probability

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0048 -0047 -0040 -0043 -0042

(0013) (0013) (001) (0008) (0009)

Unemployment rate (100) 0505 0420 0280

(0083) (0086) (0082)

Log median income -0057 -0039

(0011) (0011)

Employment-to-population ratio -0239

(0038)

N 1242022 1242022 1242022 1242022 1242022

plt01 plt005 plt001

Note Robust standard errors are in parentheses Observations are clustered at the state level The outcome variable is binary and equal to one if a family is enrolledin SNAP All models include state fixed effects Census division x-year fixed effects and state-specific linear time trends All specifications except 3a include additional

controls for family size race and marital status of the family head presence of children and presence of an adult maleSource Annual data from the March Current Population Survey for 1990 to 2012 estimation includes Current Population Survey probability weights

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3143

28 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

TABLE 5

SNAP enrollment

State-level linear regression

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0028 -0024 -0026 -0031 -0031

(0014) (0013) (0013) (0012) (0012)

Unemployment rate (100) 0625 0466 0320 0339

(0087) (0088) (0085) (0083)

Log median income -0090 -0065 -0061

(0013) (0013) (0013)

Employment-to-population ratio -0282 -0248

(0037) (0038)

N 1127 1127 1127 1127 1127

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the SNAP enrollment rate All models include state fixed effects Census division x-year fixedeffects and state-specific linear time trends

Source Annual data from the March Current Population Survey for 1990 to 2012

TABLE 6

SNAP expenditures

State-level linear regression

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0181 -0149 -0156 -0153 -0190

(011) (0103) (0102) (0103) (0103)

Unemployment rate (100) 4587 4152 4232 4313

(0622) (0621) (0633) (0628)

Log median income -0246 -0261 -0294

(0075) (0078) (0078)

Employment-to-population ratio 0155 0244

(0237) (024)

N 1127 1127 1127 1127 1127

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the natural log of state SNAP expenditures per capita for 1990 to 2012 All models include state

fixed effects Census division x-year fixed effects and state-specific linear time trends

Source Annual data from Bureau of Economic Analysis national income and product accounts tables for 1990 to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3243

Appendix B | wwwamericanprogresso

TABLE 7

Comparison of specifications SNAP enrollment

Family-level linear probability

(1) (2) (3)

Log minimum wage -0028 -0049 -0042

(0014) (0017) (0009)

Unemployment rate (100) 0275 0297 0280

(0161) (0076) (0082)

Log median income -0077 -0055 -0039

(0014) (0012) (0011)

Employment-to-population ratio -0238 -0250 -0239

(0054) (004) (0038)

N 1242022 1242022 1242022

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses Observations are clustered at the state level The outcomevariable is binary or equal to one if a family is enrolled in SNAP All specifications include additional controls forfamily size race and marital status of the family head presence of children and presence of an adult male

Source Annual data from the March Current Population Survey for 1990 to 2012 estimation includes CurrentPopulation Survey probability weights

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3343

30 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

TABLE 8

Comparison of specifications SNAP enrollment

State-level linear regression

(1) (2) (3)

Log minimum wage -0019 -0035 -0031

(0009) (0012) (0012)

Unemployment rate (100) 0401 0370 0339

(0063) (0077) (0083)

Log median income -0081 -0073 -0061

(0011) (0013) (0013)

Employment-to-population ratio -0183 -0222 -0248

(0039) (0039) (0038)

N 1127 1127 1127

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the SNAP enrollment rate Allregressions include the share of households with children and the racial shares of the population

Source Annual data from the March Current Population Survey for 1990 to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3443

Appendix B | wwwamericanprogresso

TABLE 9

Comparison of specifications SNAP expenditures

State-level linear regression

(1) (2) (3)

Log minimum wage -0121 -0203 -0190

(0075) (0103) (0103)

Unemployment rate (100) 5292 5152 4313

(0464) (0576) (0628)

Log median income -0437 -0417 -0294

(008) (0086) (0078)

Employment-to-population ratio -0040 -0220 0244

(0261) (0260) (0240)

N 1127 1127 1127

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the natural log of state SNAPexpenditures per capita All regressions include the share of households with children and the racial shares ofthe state population

Source Annual data from Bureau of Economic Analysis national income and product accounts tables for 1990to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3543

32 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3643

Appendix C | wwwamericanprogresso

Appendix C Harkin-Miller

policy simulation results

StateMinimum wage

(2014)

SNAP enrollment (2012) Predicted change in SNAP enrollment

RateTotal

(persons)

Model 1 family

enrollment (linearprobability)

Model 2 state

enrollment (linearregression)

Model 3 s

expenditu(linear regre

Alabama $725 164 788682 -66922 -58897 -58906

Alaska $775 120 87436 -8104 -6887 -3288

Arizona $790 201 1319323 -67435 -56738 -64356

Arkansas $725 78 230489 -40977 -36063 -35248

California $800 206 7813680 -371131 -310222 -18223

Colorado $800 164 853155 -50684 -42365 -23926

Connecticut $870 91 326621 -22456 -17975 -13711

Delaware $725 186 170262 -12739 -11211 -10647

District of Columbia $825 133 84009 -5370 -4417 -3632

Florida $793 166 3208026 -195813 -164426 -13046

Georgia $725 160 1586336 -137741 -121224 -11004

Hawaii $725 96 133662 -19310 -16995 -14933

Idaho $725 92 147501 -22165 -19507 -15809

Illinois $825 95 1225084 -109088 -89742 -70955

Indiana $725 125 816233 -90818 -79928 -83985

Iowa $725 155 478011 -42716 -37594 -28556

Kansas $725 135 388269 -40082 -35275 -27461

Kentucky $725 130 568821 -60840 -53544 -52259

Louisiana $725 149 683832 -63929 -56263 -66083

Maine $750 77 101976 -16567 -14323 -15234

Maryland $725 144 846415 -81748 -71946 -38370

Massachusetts $800 130 864721 -64902 -54251 -42913

Michigan $740 146 1439141 -128801 -112140 -11022

Minnesota $725 133 713646 -74730 -65769 -37878

TABLE 10

SNAP enrollments Predicted changes by state under the Harkin-Miller bill

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3743

34 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

StateMinimum wage

(2014)

SNAP enrollment (2012) Predicted change in SNAP enrollment

RateTotal

(persons)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 state

expenditures

(linear regression)

Mississippi $725 129 386501 -41486 -36511 -46467

Missouri $750 172 1036182 -75131 -64952 -56244

Montana $790 132 132452 -10350 -8708 -5846

Nebraska $725 124 230683 -25773 -22683 -12189

Nevada $825 162 446035 -23349 -19209 -11894

New Hampshire $725 127 168404 -18359 -16157 -5735

New Jersey $825 160 1416666 -75175 -61843 -28236

New Mexico $750 149 310896 -25983 -22463 -22512

New York $800 192 3763553 -191193 -159815 -142182

North Carolina $725 174 1697193 -135417 -119179 -113503

North Dakota $725 87 61225 -9743 -8574 -4021

Ohio $795 143 1647345 -115869 -97169 -88580

Oklahoma $725 129 494053 -53006 -46650 -46854

Oregon $910 124 485326 -17036 -13328 -16398

Pennsylvania $725 161 2053643 -177315 -156052 -125586

Rhode Island $800 156 163730 -10258 -8574 -8698

South Carolina $725 94 445277 -65614 -57746 -50304

South Dakota $725 208 173749 -11586 -10197 -7458

Tennessee $725 142 914903 -89667 -78915 -99134

Texas $725 110 2863779 -362018 -318607 -253285

Utah $725 88 251107 -39658 -34902 -19390

Vermont $873 156 97792 -3823 -3055 -2475

Virginia $725 101 829771 -113723 -100086 -58212

Washington $932 72 496934 -23221 -17947 -17756

West Virginia $725 58 107875 -25792 -22699 -21665

Wisconsin $725 75 427822 -79521 -69986 -53210

Wyoming $725 164 94590 -8010 -7050 -3104

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3843

Appendix C | wwwamericanprogresso

StateMinimum wage

(2014)

SNAP expenditures (2012)Predicted change in SNAP expenditures

(millions of dollars)

Per resident

(millions of

dollars)

Total

(millions of

dollars)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 s

expenditu

(linear regre

Alabama $725 $290 $1397 -$1185 -$1043 -$1043

Alaska $775 $253 $185 -$262 -$223 -$106

Arizona $790 $257 $1686 -$935 -$787 -$892

Arkansas $725 $248 $733 -$636 -$560 -$547

California $800 $189 $7164 -$7276 -$6082 -$3573

Colorado $800 $157 $816 -$862 -$721 -$407

Connecticut $870 $191 $686 -$343 -$275 -$210

Delaware $725 $250 $229 -$205 -$180 -$171

District of Columbia $825 $366 $232 -$146 -$120 -$99

Florida $793 $294 $5676 -$4429 -$3719 -$2951

Georgia $725 $317 $3140 -$2936 -$2584 -$2346

Hawaii $725 $335 $465 -$449 -$395 -$347

Idaho $725 $225 $359 -$376 -$331 -$268

Illinois $825 $249 $3200 -$2096 -$1725 -$1364

Indiana $725 $220 $1439 -$1162 -$1023 -$1075

Iowa $725 $192 $589 -$658 -$579 -$440

Kansas $725 $159 $460 -$502 -$441 -$344

Kentucky $725 $298 $1303 -$1133 -$997 -$973

Louisiana $725 $315 $1450 -$1047 -$922 -$1083

Maine $750 $281 $373 -$267 -$231 -$246

Maryland $725 $188 $1109 -$1765 -$1553 -$828

Massachusetts $800 $206 $1366 -$1030 -$861 -$681

Michigan $740 $300 $2963 -$2400 -$2090 -$2054

Minnesota $725 $140 $755 -$1113 -$980 -$564

Mississippi $725 $326 $973 -$649 -$571 -$726

Missouri $750 $241 $1452 -$1278 -$1104 -$956

Montana $790 $190 $191 -$179 -$151 -$101

Nebraska $725 $140 $259 -$409 -$360 -$194

Nevada $825 $191 $527 -$441 -$363 -$225

New Hampshire $725 $126 $167 -$399 -$351 -$125

TABLE 11

SNAP expenditures Predicted changes by state under the Harkin-Miller bill

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3943

36 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

StateMinimum wage

(2014)

SNAP expenditures (2012)Predicted change in SNAP expenditures

(millions of dollars)

Per resident

(millions of

dollars)

Total

(millions of

dollars)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 state

expenditures

(linear regression)

New Jersey $825 $160 $1420 -$1611 -$1325 -$605

New Mexico $750 $324 $675 -$513 -$444 -$445New York $800 $287 $5616 -$3766 -$3148 -$2801

North Carolina $725 $252 $2454 -$2187 -$1925 -$1833

North Dakota $725 $128 $90 -$162 -$143 -$67

Ohio $795 $259 $2995 -$2013 -$1688 -$1539

Oklahoma $725 $248 $945 -$799 -$703 -$706

Oregon $910 $322 $1255 -$272 -$213 -$262

Pennsylvania $725 $218 $2779 -$2930 -$2579 -$2075

Rhode Island $800 $280 $294 -$173 -$144 -$147

South Carolina $725 $291 $1373 -$1337 -$1177 -$1025South Dakota $725 $198 $165 -$192 -$169 -$123

Tennessee $725 $324 $2091 -$1413 -$1243 -$1562

Texas $725 $230 $5997 -$6402 -$5634 -$4479

Utah $725 $141 $402 -$614 -$541 -$300

Vermont $873 $230 $144 -$66 -$53 -$43

Virginia $725 $173 $1413 -$2062 -$1815 -$1056

Washington $932 $244 $1682 -$350 -$270 -$267

West Virginia $725 $273 $508 -$451 -$397 -$379

Wisconsin $725 $204 $1166 -$1302 -$1146 -$871Wyoming $725 $95 $55 -$105 -$93 -$41

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4043

Appendix C | wwwamericanprogresso

FIGURE 2

Raising the minimum wage to $1010would cut taxpayer costs in every state

Predicted decreases in cost and enrollment

in SNAP in 50 states

$200+$51ndash$100

$101ndash$200

0ndash$25

$26ndash$50

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4143

38 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

Endnotes

1 US Department of Agriculture ldquoSupplementalNutrition Assistance Program (SNAP)rdquo available athttpwwwfnsusdagovsnapeligibility (last accessedFebruary 2014)

2 For this initial analysis we do not consider Harkin-Millerrsquos increase in subminimum wages for tippedworkers To do s o would increase the estimated SNAP

savings by an unknown amount

3 The Congressional Budget Office estimates thatworkers currently earning between $1010 and $1150per hour would see their wages rise under the Harkin-Miller proposal Congressional Budget O ffice ldquoTheEffects of a Minimum Wage Increase on Employmentand Family Incomerdquo (2014)

4 Marianne Page Joanne Spetz and Jane Millar ldquoDoesthe Minimum Wage Affect Welfare Caseloadsrdquo Journalof Policy Analysis and Management 24 (2) (2005)273ndash295

5 Marianne Bitler and Hilary Hoynes ldquoThe More ThingsChange the More They Stay the Same The SafetyNet Living Arrangements and Poverty in the GreatRecessionrdquo NBER Working Paper 194 49 2013

6 Sylvia Allegretto and others ldquoFast Food Poverty Wages The Public Cost of Low-Wage Jobs in the Fast-FoodIndustryrdquo (Berkeley California Center for LaborResearch and Education 2013) available at httplaborcenterberkeleyedupubliccostsfast_food_poverty_wage

7 David Neumark and William Wascher ldquoDoes a HigherMinimum Wage Enhance the Effectiveness of theEarned Income Tax Creditrdquo Industrial and LaborRelations Review 64 (4) (2011) 712ndash746

8 David Lee and Emmanuel Saez ldquoOptimal MinimumWage Policy in Competitive Labor Marketsrdquo Journal ofPublic Economics 96 (9) (2012) 739ndash749

9 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family IncomerdquoFebruary 2014

10 Jesse Rothstein ldquoIs the EITC as Good as an NITConditional Cash Transfers and Tax Incidencerdquo AmericanEconomic Journal Economic Policy 2 (1) (2010) 177ndash208

11 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family Incomerdquo

12 Dube Arindrajit 2013 rdquoMinimum Wagesand the Distribution of Family IncomerdquoUnpublished working paper Available at httpsdldropboxusercontentcomu15038936Dube_MinimumWagesFamilyIncomespdf

13 As Dube explains in ldquoThe poverty of Minimum WageFactsrdquo the simulation approach underestimate stemsfrom a number of unwarranted assumptions includingthe range of actual wage increases and the accuracy ofwage data in the Current Population Survey The causal

approach does not make these assumptions

14 Allegretto Sylvia and others 2013 ldquoCredible ResearchDesigns for Minimum Wage Studiesrdquo Working Paper148-13 University of California Berkeley Institutefor Research on Labor and Employment Available athttpwwwirleberkeleyeduworkingpapers148-13pdf

15 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family In comerdquo

16 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo (2012)

17 Ibid

18 Ibid

19 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo

20 US Department of Agriculture ldquoSupplementalNutrition Assistance Program (SNAP)rdquo

21 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo

22 Ibid

23 We deviate from the Census Bureaursquos definition ofa family unit which is ldquotwo people or more (on e ofwhom is the householder) related by bir th marriageor adoption and residing togetherrdquo US Bureau ofthe Census ldquoCurrent Population Survey DefinitionsFamilyrdquo available at httpwwwcensusgovcpsabout

cpsdefhtml (last accessed February 2014) We countas a family unit any individual residing on his or herown two or more persons residing together whodo not belong to a family in the March CPS sampleare constructed as one family in our analysis For thepurposes of food stamp allocations the consumptionresulting from this transfer is probably distributed tofamily members (rather than household members ora single individual within the household) Howeversingle individuals canmdashand domdashreceive SNAPbenefits Excluding them would fail to make theanalysis reflective of the population at large

24 Strictly the family level linear probability modelpredicts the percentage-point decrease in theprobability that an individual family will receive SNAPpayments When applied to a large number of familieshowever we are able to interpret the coefficient asa decrease in the mean of enrollmentmdashthat is a

decrease in the enrollment ratemdashby applying the lawof iterated expectations

25 We generate expenditure predictions from theenrollment modelsmdashand conversely generateenrollment predictions from the expenditure modelmdashby assuming that expenditures per enrolled familyremains the same before and after the minimum wagechange In practice this is likely to be a conservativeestimatemdashthat is to underestimate the decrease inSNAP activity Average SNAP benefits per family willalso decrease as many families that remain eligible forSNAP experience income gains

26 Wage and Hour Division ldquoMinimum Wage Laws inthe States ndash Januar y 1 2014rdquo available at httpwwwdolgovwhdminwageamericahtm (last accessedFebruary 2014)

27 See for example Sylvia Allegretto and others ldquoCredibleResearch Designs for Minimum Wage Studiesrdquo WorkingPaper 148-113 (Berkeley California Institute forResearch on Labor and Employment 2013) available athttpwwwirleberkeleyeduworkingpapers148-13pdf

28 We will report these results in a forthcoming workingpaper

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4243

Endnotes | wwwamericanprogresso

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4343

The Center for American Progress is a nonpartisan research and educational institute

dedicated to promoting a strong just and free America that ensures opportunity

for all We believe that Americans are bound together by a common commitment to

these values and we aspire to ensure that our national policies reflect these values

We work to find progressive and pragmatic solutions to significant domestic and

international problems and develop policy proposals that foster a government that

is ldquoof the people by the people and for the peoplerdquo

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1443

Methods and data | wwwamericanprogresso

Te March CPS also collecs inormaion on he number o SNAP recipiens in

he household he number o monhs he household receives SNAP benefis

and he cash-equivalen value o he SNAP benefis received However he

value o SNAP benefis is severely underrepored among recipiens perhaps

because recipiens are unaware o he exac monhly cash-equivalen value o

benefis hey receive

Our firs empirical sraegy ocuses solely on SNAP enrollmen By using he

amily as he uni o analysis we are able o inser saisical conrols o accoun

or non-wage-relaed acors ha influence any paricular amilyrsquos likelihood o

program paricipaion wih he inenion o isolaing any differences in program

paricipaion ha are due purely o changes in wage policy Tis approach

idenifies he effecs o low-wage labor policy on he exernal margin991252ha is

he effec o he minimum wage on he likelihood ha a amily paricipaes in he

SNAP program a all991252as opposed o he inernal margin or how much SNAP

unding he amily would receive

Our second empirical ramework uses sae-level adminisraive daa Ta is

we aggregae he daa o obain a single daa poin or each saeyear back

o 1990 represening he mean o he oucome or he sae Te sae-level

esimaion serves as a robusness check on he amily-level resuls or SNAP

paricipaion Also using aggregaed daa allows us o esimae direcly he

causal effec o minimum wage changes on SNAP spending Tis is no possible

a he amily level as discussed above daa on cash-equivalen value o ood

samps or SNAP recipiens is very requenly no repored in he March CPS

and when i is repored he inormaion may be unreliable By conras he

Bureau o Economic Analysis publishes aggregae SNAP spending a he sae

level in is Naional Income and Produc Accoun or NIPA ables Tus

while we are unable o observe he heerogeneiy in he cash value o SNAP

or amilies in each sae we are able o calculae average SNAP spending

per residen in each sae per year Supporing covariaes include he annual

unemploymen and employmen daa rom he Bureau o Labor Saisics or

BLS and sae-level populaion series rom he iner-decennial census releases

Minimum wage daa are available rom he BLSrsquos wages and hours division For

sae minimum wage changes enaced a oher imes han he firs o he yearan average value or he year is used

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1543

12 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

Family level model specifications

We firs examine he effec o he minimum wage on paricipaion in public

assisance programs For amily residing in sae and during year we esimae

an equaion o he ollowing orm

(1)

is a binary variable ha is se equal o 1 i a leas one member o amily

received ood samps during he survey year is a se o sae-level

characerisics including annual averages o he unemploymen rae he

employmen-o-populaion raio and he naural log o median amily income

is a vecor o amily atribues including indicaors or he race and marial saus

o he amily head size o he amily he presence o children and he presence oan adul male Sae fixed effecs are capured by o conrol or ime-varying

heerogeneiy our preerred model specificaion also includes year fixed effecs

ha vary by Census division ( ) and sae-level linear ime rends In

Appendix B we jusiy he inclusion o hese las wo erms We also compare he

resuls rom our preerred specificaion wih less sauraed specificaions

Te effec o ineres which is capured by is he expeced change in he

probabiliy o receiving SNAP benefis wih respec o a change in he (log o he)

binding minimum wage in sae during year We repor robus sandard errors

clusered a he sae level We esimae he parameers using linear regression

producing a linear probabiliy model Deails o he model selecion process are

covered in Appendix B below

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1643

Methods and data | wwwamericanprogresso

State-level model specifications

Te sae-level models are similarly specified For sae in year we assume ha

(2)

In his model is now eiher he SNAP enrollmen rae in sae during year

or he naural logarihm o per capia SNAP expendiures in sae during

year is once again a se o sae-level characerisics including he same

sae-level covariaes as in he amily regressions (annual average unemploymen

rae employmen-o-populaion raio naural log o median amily income)

wih he addiion o amily level characerisics averaged across he sae (average

amily size and he shares o populaion consiued by each o five racialehnicgroups) Sae fixed effecs are represened by As above our preerred model

specificaion includes year fixed effecs ha vary by Census division ( ) and

sae-level linear ime rends as elaboraed in Appendix B Te effec o

ineres is capured by

We esimae boh sae-level models (enrollmen and expendiures) using

ordinary leas squares regression Tus he inerpreaion o he coefficien is no

longer ha o a change in probabiliy as in he binary oucome models described

above Raher or he sae-level SNAP enrollmen model represens he

expeced change (in percenage poins) in he saersquos SNAP enrollmen rae ha

is due o a 1 percen change in he minimum wage For he SNAP expendiures

model is simply he elasiciy o SNAP spending wih respec o he minimum

wage991252ha is he percenage change in sae expendiures expeced o resul rom

a 1 percen change in ha saersquos minimum wage For urher deails on model

specificaion reer o Appendix B below

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1743

14 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1843

Results | wwwamericanprogresso

Results

Estimated minimum wage effects on SNAP enrollment and

expenditures

able 1 shows he esimaed parameer o ineres991252he coefficien o he

minimum wage or he preerred model o each ype Coefficiens on he

minimum wage variable are no direcly comparable across models because all

our models have a differen uncional orm o undersand and compare hese

esimaes we compue he change in SNAP aciviy prediced or a paricular wage scenario Te final column in able 1 answers he quesion Wha would be

he expeced change in SNAP aciviy in response o a 10 percen increase in he

minimum wage Te answer o his quesion varies wih he value o he inpu

parameers in he able we calculae he percenage decrease in enrollmen or

expendiures prediced or he average sae wih a minimum wage o $725 in

2014 Te sae-level SNAP expendiure model which is a consan-elasiciy

model conveys elasiciy inormaion direcly or he change in expendiures per

capia in he sae

TABLE 1

Comparison of national SNAP predictions for a 10 percent increase in the federal minimum wage

Model LevelRegression

type

Predicted outcome Coefficient of log

(minimum wage)

(Standard error)

Effect of a 10 percent in

in the minimum wage

VariableForm of

variable

Total

enrollment

Tot

expend

1 Family Linear

probability Enrollment Binary (enrolled=1)

-0042

(0008) -317 N

2

StateLinear regression

(ordinary least

squares)

Enrollment State enrollment rate (percent) -0031

(0012) -235 N

3 Expenditures Log (state expenditures per capita) -0190

(0103) NA -19

plt01 plt005 plt001Note Predicted changes are calculated for the average state with a minimum wage of $725 in 2014

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1943

16 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

According o his model a 10 percen increase in he minimum wage would resul

in a 19 percen reducion in SNAP expendiures Tis resul is reassuringly similar

o Dubersquos resul or povery reducion Te wo enrollmen models are much more

precisely esimaed han he expendiure model Te sae-level enrollmen model

finds ha a 10 percen minimum wage increase in a low-wage sae is associaed wih

a 235 percen decrease in SNAP enrollmens Te amily-level linear probabiliymodel predics a somewha greaer elasiciy or low minimum wage saes an

increase o 10 percen in he ederal minimum would resul in a 317 percen decline

in SNAP enrollmen24 Te differences in hese esimaes sem rom a number o

acors including difference in model uncional orm and daa used We rea his

range o elasiciy esimaes as an upper and lower bound on enrollmen impacs

Harkin-Miller bill National and state-level predicted impacts

Wha would be he prediced change or he SNAP program i he ederalminimum were raised o $1010 as proposed in he Harkin-Miller bill In order o

make his inerence we accoun or he ac ha no all saes are currenly subjec

o he ederal minimum wage a he beginning o 2014 21 saes mainained

higher minimum wages han $725 In hose saes an increase in he ederal

minimum wage may or may no be binding or employers in he sae depending

upon wheher he new ederal minimum exceeds he sae-level minimum Bu

regardless o wheher a minimum wage change is binding he impac on SNAP

aciviy will be lower in high minimum wage saes In order o accoun or his

properly we calculae sae by sae he percenage wage change ha would resul

rom he Harkin-Miller proposal and apply he parameers rom each o he hree

models above o compue he expeced decrease in SNAP aciviy or each sae

In his exercise we use saesrsquo curren (2014) minimum wage levels and assume

as a baseline he 2012 levels o SNAP enrollmen and expendiure as 2012 is he

mos recen year or which SNAP daa are available

able 10 and able 11 in Appendix C repor he esimaed effecs on SNAP

enrollmen and expendiures respecively or each sae under he Harkin-Miller

bill25 An increase o $1010 i enaced oday would represen beween a 393

percen wage increase in a $725 minimum wage sae and an 84 percen increasein Washingon sae which has he highes minimum wage in he naion a $932

as o January 201426 Slighly more han 56 percen o he decrease in expendiures

and abou 59 percen o he decrease in enrollmen would occur in saes wih

presen-day minimum wages o $725

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2043

Results | wwwamericanprogresso

In 2012 hese saes were home o 46 percen o he American populaion

and accouned or a roughly equivalen percenage o oal naional SNAP

expendiures No surprisingly he larges enrollmen decreases come rom

populous saes wih very high SNAP enrollmen raes andor very low minimum

wages Te larges prediced enrollmen reducion991252beween 319000 individuals

and 362000 individuals991252would occur in exas which has a $725 minimum wage In Caliornia which has a 206 percen SNAP paricipaion rae and an $8

per hour minimum wage we anicipae SNAP enrollmen declines o 310000

persons o 371000 persons And in Florida which had a SNAP paricipaion rae

o 166 percen and a $793 minimum wage enrollmen could decline by beween

164000 individuals and 196000 individuals For he our saes ha ogeher

accouned or he greaes amoun o SNAP spending in 2012991252exas Caliornia

Florida and New York respecively991252he combined expendiure reducion rom

he Harkin-Miller bill is prediced o be $14 billion

able 2 summarizes he prediced declines in SNAP aciviy or he naion as a whole ha would resul rom he direc and indirec effecs o he Harkin-Miller

bill Enrollmen would all beween 31 million persons and 36 million persons

represening 75 percen o 87 percen o curren enrollmen Te anicipaed

reducion in program expendiures would be nearly $46 billion or 61 percen o

program expendiures

TABLE 2

Comparison of national SNAP predictions under the Harkin-Miller billrsquos $1010 minimum wage

Model

Enrollment

(persons)

Expenditures

(millions of dollars)

Current (2012) Predicted Change Current (2012) Predicted Chan

Family enrollment (linear probability)

41866195

45489339 -3623144

$74861

NA NA

State enrollment (ordinary least squares) 38745435 -3120759 NA NA

State expenditures (ordinary least squares) NA NA $70305 -$45

Note Calculations use 2014 state minimum wages and the most recent SNAP data from 2012 They assume that per-enrollee expenditures remain constant

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2143

18 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

Tere are o course oher possibiliies or minimum wage legislaion able 3

shows he expeced SNAP changes or he Unied Saes under a variey o wage

scenarios calculaed using he sae-level models I saes were no able o se

heir minimum wages independenly such ha all saes were consrained by

he ederal minimum o $725 SNAP would be received by abou 514000 more

people across he Unied Saes a an addiional program cos o nearly hree-quarers o a billion dollars In conras he effecs o a higher minimum wage

proposal991252a ederal wage floor o $11 per hour991252would decrease enrollmen in

SNAP by more han 10 percen and decrease program coss by 83 percen

TABLE 3

Summary of par ticipation and expenditures under wage scenarios

If all states had

minimum wages of

Enrollment(persons)

Expenditures

(millions of dollars)

Predicted Change Predicted Change

Recent levels (2014) 41866195 $74861

$725 42380520 514326 $75604 $743

$800 41423919 -442276 $74209 -$652

$900 40148451 -1717744 $72350 -$2511

$1000 38872982 -2993212 $70490 -$4371

$1010 38745435 -3120759 $70305 -$4556

$1100 37597514 -4268681 $68631 -$6230

Note Calculations use state-level enrollment model coefficient

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2243

Conclusion | wwwamericanprogresso

Conclusion

An exensive body o lieraure examines employmen effecs o he minimum

wage A much smaller se o sudies examines how he minimum wage affecs

povery and only a handul o papers examine he effecs o he minimum wage

on he EIC Our analysis is he firs o examine he effecs o he minimum

wage on SNAP

Our findings indicae ha increased earnings rom minimum wage changes do

reduce SNAP enrollmens and expendiures We esimae ha he Harkin-Miller bill would save axpayers nearly $46 billion per year equivalen o 61 percen

o SNAP expendiures in 2012 he las year or which daa are available Over a

10-year period he esimaed savings amoun o nearly $46 billion

Our repor is subjec o limiaions ha we expec o overcome in our uure

research Firs he findings do no ake ino accoun possible ineracions among

SNAP he EIC and Medicaid Te eligibiliy cuoffs among hese programs

are quie differen suggesing ha such ineracions may be minor Noneheless

he join effecs can only be deermined by urher research using a causal

model Second i would be useul o know he disribuion o SNAP reducions

along he wage disribuion Using he Congressional Budge Officersquos calculaions

o how much he oal dollar value o wage would increase under he Harkin-

Miller proposal our findings imply ha he decline in overall SNAP spending

equals abou 15 percen o he oal resuling increase in wages Te amoun and

disribuion o his offse are o considerable ineres Minimum wage beneficiaries

who come rom working amilies already well above he povery line would no

see any offse while hose who are currenly considerably below he povery line

will see larger offses Tese issues will also be a subjec or our uure research

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2343

20 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

About the authors

Rachel West is a maser o public policy candidae a he Goldman School

o Public Policy Universiy o Caliornia Berkeley Her research ocuses on

economic policy in he areas o low-wage labor and povery

Michael Reich is proessor o economics and direcor o he Insiue or

Research on Labor and Employmen a he Universiy o Caliornia a Berkeley

His research publicaions cover numerous areas o labor economics including

racial inequaliy labor marke segmenaion high-perormance workplaces

union-managemen cooperaion Japanese labor-managemen sysems living

wages and minimum wages He received his docorae in economics rom

Harvard Universiy

Acknowledgments

We are graeul o Sylvia Allegreto Arindraji Dube Bill Leser Jesse Rohsein

Daniel Tompson and Ben Zipperer or heir valuable suggesions

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2443

References | wwwamericanprogresso

References

Allegreto Sylvia and ohers 2013 ldquoFas Food Povery Wages Te Public Cos o Low-Wage Jobsin he Fas-Food Indusryrdquo Universiy o Caliornia Berkeley Cener or Labor Research andEducaion Available a htplaborcenerberkeleyedupubliccossas_ood_povery_wage

Allegreto Sylvia and ohers 2013 ldquoCredible Research Designs or Minimum Wage Sudiesrdquo

Working Paper 148-13 Universiy o Caliornia Berkeley Insiue or Research on Labor andEmploymen Available a htpwwwirleberkeleyeduworkingpapers148-13pd

Biler Marianne and Hilary Hoynes 2013 ldquo Te More Tings Change he More Tey Say heSame Te Saey Ne Living Arrangemens and Povery in he Grea Recessionrdquo Working Paper19449 Naional Bureau o Economic Research

Congressional Budge Office 2012 ldquoTe Supplemenal Nuriion Assisance Programrdquo Washingon Available a htpwwwcbogovsiesdeaulfilescbofilesatachmens04-19-SNAPpd

991252 991252 991252 2014 ldquoTe Effec o a Minimum-Wage Increase on Employmen and Family Incomerdquo Washingon Available a htpwwwcbogovsiesdeaulfilescbofilesatachmens44995-MinimumWagepd

Dube Arindraji 2013 rdquoMinimum Wages and he Disribuion o Family Incomerdquo Unpublished working paper Available a htpsdldropboxuserconencomu15038936Dube_ MinimumWagesFamilyIncomespd

991252 991252 991252 2014 ldquoTe Povery o Minimum Wage Facsrdquo Arindraji Dube Blog January 22 Availablea htparindubecom20140122he-povery-o-minimum-wage-acs

Lee David and Emmanuel Saez 2012 ldquoOpimal Minimum Wage Policy in Compeiive LaborMarkesrdquo Journal o Public Economics 96 (9) 739ndash749

Neumark David and William Wascher 1992 ldquoEmploymen Effecs o Minimum and Subminimum Wages Panel Daa on Sae Minimum Wage Lawsrdquo Industrial and Labor Relations Review 46 (1)

55ndash81

Neumark David and William Wascher 2011 ldquoDoes a Higher Minimum Wage Enhance heEffeciveness o he Earned Income ax Credirdquo Industrial and Labor Relations Review 64 (4)712ndash746

Page Marianne Joanne Spez and Jane Millar 2005 ldquoDoes he Minimum Wage Affec WelareCaseloadsrdquo Journal o Policy Analysis and Management 24 (2) 273ndash295

Rohsein Jesse 2010 ldquoIs he EIC as Good as an NI Condiional Cash ransers and ax

Incidencerdquo American Economic Journal Economic Policy 2 (1) 177ndash208

Wage and Hours Division 2014 ldquoMinimum Wage Laws in he Saes ndash January 1 2014rdquo USDeparmen o Labor (htpwwwdolgovwhdminwageamericahm [February 2014])

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2543

22 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2643

Appendix A | wwwamericanprogresso

Appendix A

Pre-trend falsification check

Recen minimum wage research27 highlighs a common flaw in previous sudies

ailure o veriy ha he oucome variable is ree o negaive pre-exising rends I

or example SNAP aciviy was already rending down in saes ha raised heir

minimum wages beore hese changes came ino effec our regression analysis could

(misakenly) atribue ha reducion o he minimum wage We check or such

pre-rends by inroducing variables ha represen he prior yearrsquos value or leads o

he minimum wage I he model esimaes he minimum wage o have an effec on

he oucome variable beore he wage change wen ino effec hen an unobservedacor no he minimum wage change caused he change in SNAP aciviy

We es he specificaions above or pre-rends by including a one-year lead in

all hree specificaions We find ha he lead erms are small posiive and no

saisically significan indicaing ha he concurren minimum wage991252no

he wage level in prior periods991252is driving he observed changes in SNAP

oucomes28 In paricular he coefficien (sandard error) on he lead erm in

our preerred amily-level enrollmen regression is 011 and no significan

while he coefficien and sandard error o he conemporaneous minimum

wage is unchanged In he sae-level preerred enrollmen regression he

coefficien o he lead erm is again small (07) and i is no significan Te

corresponding coefficien on he lead erm in he sae-level expendiure

regression is 16 and is no significan Te posiive poin esimaes on hese lead

erms resuls no only rule ou disoring negaive pre-rends Tey also sugges

ha our main resuls may underesimae he rue effecs

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2743

24 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2843

Appendix B | wwwamericanprogresso

Appendix B

Model selection process

For boh he amily-level and sae-level models we es hree mehods o conrol

or unobserved geographic- and ime-varying characerisics as suggesed by he

minimum wage lieraure Firs we include only independen sae-specific fixed

effecs and year-specific fixed effecs Tis specificaion (specificaion 1) implicily

assumes ha amilies in any sae consiue an equally good saisical ldquoconrolrdquo

group or hose in any randomly chosen sae afer accouning or various

characerisics (median income and unemploymen rae among ohers) Similarly

simple ime fixed effecs assume ha amilies surveyed in any year can crediblyserve as a conrol group or amilies surveyed in every oher year o he sample

(1990 hrough 2012)

In oher words specificaion 1 assumes ha a saersquos immediae neighbor provides

no beter a couneracual or he effec o a minimum wage change han does a

sae across he counry We relax his resricive specificaion sequenially in wo

seps In specificaion 2 we replace simple year fixed effecs wih fixed effecs or

each Census divisionyear (capured as an addiional variable in he vecor By

using division-year effecs we remove he resricion ha amilies in each sae

are equally good saisical conrols or all oher amilies Raher we allow or he

possibiliy ha amilies in similar geographic regions (or example he Souh or

he Norheas) may be more similar o one anoher han amilies arher away

Finally in specificaion 3 we add sae-specific linear ime rends o he previous

specificaion Tus specificaion 3 is he mos rigorous model specificaion in ha

i allows or heerogeneiy along hree dimensions Ta is specificaion 3 allows

each sae o have is own ime-varying rends raher han imposing he resricion

ha saes evolve idenically over he 22 years in he sample

We begin building he heoreical specificaion above rom a se o simpleuncondiional models regression o SNAP aciviy (enrollmen or expendiures)

on he log o he minimum wage and a se o geographic- and ime-specific

effecs (specificaions 1 2 and 3 described above) As shown in ables 1ndash3 (or

specificaion 3) we hen add covariaes sequenially o hese models including

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2943

26 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

firs he vecor o amily-level conrols ollowed by each o several sae-level

covariaes in urn he unemploymen rae log o median-amily income and he

employmen-o-populaion raio Comparable resuls or specificaions 2 and 3

will be available in our orhcoming working paper

As expeced he simple uncondiional models indicae ha he relaionship beween he minimum wage and SNAP enrollmen i one exiss is a more complex

one influenced by oher acors In he uncondiional model he coefficien on he

variable o ineres991252he log o he minimum wage991252is small in magniude and no

saisically differen rom zero Once we accoun or he influence o labor marke

condiions and variaion in income levels on program paricipaion (by including

unemploymen rae and median-amily income conrol variables respecively)

he effec o he minimum wage on SNAP enrollmen is precisely esimaed Te

coefficien o he log minimum wage is slighly higher (-0042) in he amily-level

analysis han he coefficien (-031) in he sae-level analysis Te level o precision

is also higher in he amily-level analysis Tis is o be expeced when using 124million observaions compared o 1127

Te second dimension o model choice concerns he effec specificaion ables

7ndash9 compare he primary coefficiens o ineres or he SNAP enrollmen and

expendiure models For boh he enrollmen models he effec sizes are smalles

or specificaion 1 larges or specificaion 2 and inermediae beween hese wo

in specificaion 3 Recall ha Specificaion 3 conains sae-specific linear ime

rends in addiion o he census divisionyear conrols included in specificaion

2 In he amily-level enrollmen model he sandard error o he minimum wage

coefficien is smaller han in he oher wo specificaions Sandard errors on he

oher variables are much smaller in specificaions 2 and 3 han in specificaion

1 On he basis o coefficien significance (join and individual) specificaions 2

and 3 are sricly preerred in boh enrollmen models o specificaion 1 which

conains only sae and year fixed effecs

A concern wih specificaions 2 and 3 is ha rend conrols such as sae linear

rends may incorrecly absorb some o he delayed impac o a minimum wage

When we es his issue by including lagged minimum wages we do no find ha

delayed effecs are significan Anoher concern is ha more sauraed modelsuse less o he saisical variaion which could reduce he saisical power o

he resuls However he sandard errors or our more sauraed models are no

higher and are lower in some cases han or he less sauraed models Overall

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3043

Appendix B | wwwamericanprogresso

his evidence suppors our use o he sauraed model as he preerred model

specificaion Moreover Dubersquos 2013 sudy shows ha more sauraed models

perorm beter han models wih jus sae and ime fixed effecs

Te esimaed enrollmen regressions a boh he amily and sae levels show large

and saisically significan coefficiens Te esimaed minimum wage effec in heexpendiures regressions991252or which we have only sae-level daa991252is also large

and saisically significan

We do no use weighed regression or he sae-level models preerring o keep

analysis o he ldquoreamenrdquo (ha is o say a minimum wage change) appropriae

o he average sae raher han he average amily or individual I insead our

primary ineres were he impac o a minimum wage change on he average amily

or he average individual we migh choose o designae he number o amilies

in each sae or he sae populaion respecively as analyic weighs in order o

obain a coefficien beter suied or such inerence

TABLE 4

SNAP enrollment

Family-level linear probability

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0048 -0047 -0040 -0043 -0042

(0013) (0013) (001) (0008) (0009)

Unemployment rate (100) 0505 0420 0280

(0083) (0086) (0082)

Log median income -0057 -0039

(0011) (0011)

Employment-to-population ratio -0239

(0038)

N 1242022 1242022 1242022 1242022 1242022

plt01 plt005 plt001

Note Robust standard errors are in parentheses Observations are clustered at the state level The outcome variable is binary and equal to one if a family is enrolledin SNAP All models include state fixed effects Census division x-year fixed effects and state-specific linear time trends All specifications except 3a include additional

controls for family size race and marital status of the family head presence of children and presence of an adult maleSource Annual data from the March Current Population Survey for 1990 to 2012 estimation includes Current Population Survey probability weights

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3143

28 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

TABLE 5

SNAP enrollment

State-level linear regression

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0028 -0024 -0026 -0031 -0031

(0014) (0013) (0013) (0012) (0012)

Unemployment rate (100) 0625 0466 0320 0339

(0087) (0088) (0085) (0083)

Log median income -0090 -0065 -0061

(0013) (0013) (0013)

Employment-to-population ratio -0282 -0248

(0037) (0038)

N 1127 1127 1127 1127 1127

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the SNAP enrollment rate All models include state fixed effects Census division x-year fixedeffects and state-specific linear time trends

Source Annual data from the March Current Population Survey for 1990 to 2012

TABLE 6

SNAP expenditures

State-level linear regression

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0181 -0149 -0156 -0153 -0190

(011) (0103) (0102) (0103) (0103)

Unemployment rate (100) 4587 4152 4232 4313

(0622) (0621) (0633) (0628)

Log median income -0246 -0261 -0294

(0075) (0078) (0078)

Employment-to-population ratio 0155 0244

(0237) (024)

N 1127 1127 1127 1127 1127

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the natural log of state SNAP expenditures per capita for 1990 to 2012 All models include state

fixed effects Census division x-year fixed effects and state-specific linear time trends

Source Annual data from Bureau of Economic Analysis national income and product accounts tables for 1990 to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3243

Appendix B | wwwamericanprogresso

TABLE 7

Comparison of specifications SNAP enrollment

Family-level linear probability

(1) (2) (3)

Log minimum wage -0028 -0049 -0042

(0014) (0017) (0009)

Unemployment rate (100) 0275 0297 0280

(0161) (0076) (0082)

Log median income -0077 -0055 -0039

(0014) (0012) (0011)

Employment-to-population ratio -0238 -0250 -0239

(0054) (004) (0038)

N 1242022 1242022 1242022

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses Observations are clustered at the state level The outcomevariable is binary or equal to one if a family is enrolled in SNAP All specifications include additional controls forfamily size race and marital status of the family head presence of children and presence of an adult male

Source Annual data from the March Current Population Survey for 1990 to 2012 estimation includes CurrentPopulation Survey probability weights

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3343

30 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

TABLE 8

Comparison of specifications SNAP enrollment

State-level linear regression

(1) (2) (3)

Log minimum wage -0019 -0035 -0031

(0009) (0012) (0012)

Unemployment rate (100) 0401 0370 0339

(0063) (0077) (0083)

Log median income -0081 -0073 -0061

(0011) (0013) (0013)

Employment-to-population ratio -0183 -0222 -0248

(0039) (0039) (0038)

N 1127 1127 1127

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the SNAP enrollment rate Allregressions include the share of households with children and the racial shares of the population

Source Annual data from the March Current Population Survey for 1990 to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3443

Appendix B | wwwamericanprogresso

TABLE 9

Comparison of specifications SNAP expenditures

State-level linear regression

(1) (2) (3)

Log minimum wage -0121 -0203 -0190

(0075) (0103) (0103)

Unemployment rate (100) 5292 5152 4313

(0464) (0576) (0628)

Log median income -0437 -0417 -0294

(008) (0086) (0078)

Employment-to-population ratio -0040 -0220 0244

(0261) (0260) (0240)

N 1127 1127 1127

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the natural log of state SNAPexpenditures per capita All regressions include the share of households with children and the racial shares ofthe state population

Source Annual data from Bureau of Economic Analysis national income and product accounts tables for 1990to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3543

32 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3643

Appendix C | wwwamericanprogresso

Appendix C Harkin-Miller

policy simulation results

StateMinimum wage

(2014)

SNAP enrollment (2012) Predicted change in SNAP enrollment

RateTotal

(persons)

Model 1 family

enrollment (linearprobability)

Model 2 state

enrollment (linearregression)

Model 3 s

expenditu(linear regre

Alabama $725 164 788682 -66922 -58897 -58906

Alaska $775 120 87436 -8104 -6887 -3288

Arizona $790 201 1319323 -67435 -56738 -64356

Arkansas $725 78 230489 -40977 -36063 -35248

California $800 206 7813680 -371131 -310222 -18223

Colorado $800 164 853155 -50684 -42365 -23926

Connecticut $870 91 326621 -22456 -17975 -13711

Delaware $725 186 170262 -12739 -11211 -10647

District of Columbia $825 133 84009 -5370 -4417 -3632

Florida $793 166 3208026 -195813 -164426 -13046

Georgia $725 160 1586336 -137741 -121224 -11004

Hawaii $725 96 133662 -19310 -16995 -14933

Idaho $725 92 147501 -22165 -19507 -15809

Illinois $825 95 1225084 -109088 -89742 -70955

Indiana $725 125 816233 -90818 -79928 -83985

Iowa $725 155 478011 -42716 -37594 -28556

Kansas $725 135 388269 -40082 -35275 -27461

Kentucky $725 130 568821 -60840 -53544 -52259

Louisiana $725 149 683832 -63929 -56263 -66083

Maine $750 77 101976 -16567 -14323 -15234

Maryland $725 144 846415 -81748 -71946 -38370

Massachusetts $800 130 864721 -64902 -54251 -42913

Michigan $740 146 1439141 -128801 -112140 -11022

Minnesota $725 133 713646 -74730 -65769 -37878

TABLE 10

SNAP enrollments Predicted changes by state under the Harkin-Miller bill

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3743

34 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

StateMinimum wage

(2014)

SNAP enrollment (2012) Predicted change in SNAP enrollment

RateTotal

(persons)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 state

expenditures

(linear regression)

Mississippi $725 129 386501 -41486 -36511 -46467

Missouri $750 172 1036182 -75131 -64952 -56244

Montana $790 132 132452 -10350 -8708 -5846

Nebraska $725 124 230683 -25773 -22683 -12189

Nevada $825 162 446035 -23349 -19209 -11894

New Hampshire $725 127 168404 -18359 -16157 -5735

New Jersey $825 160 1416666 -75175 -61843 -28236

New Mexico $750 149 310896 -25983 -22463 -22512

New York $800 192 3763553 -191193 -159815 -142182

North Carolina $725 174 1697193 -135417 -119179 -113503

North Dakota $725 87 61225 -9743 -8574 -4021

Ohio $795 143 1647345 -115869 -97169 -88580

Oklahoma $725 129 494053 -53006 -46650 -46854

Oregon $910 124 485326 -17036 -13328 -16398

Pennsylvania $725 161 2053643 -177315 -156052 -125586

Rhode Island $800 156 163730 -10258 -8574 -8698

South Carolina $725 94 445277 -65614 -57746 -50304

South Dakota $725 208 173749 -11586 -10197 -7458

Tennessee $725 142 914903 -89667 -78915 -99134

Texas $725 110 2863779 -362018 -318607 -253285

Utah $725 88 251107 -39658 -34902 -19390

Vermont $873 156 97792 -3823 -3055 -2475

Virginia $725 101 829771 -113723 -100086 -58212

Washington $932 72 496934 -23221 -17947 -17756

West Virginia $725 58 107875 -25792 -22699 -21665

Wisconsin $725 75 427822 -79521 -69986 -53210

Wyoming $725 164 94590 -8010 -7050 -3104

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3843

Appendix C | wwwamericanprogresso

StateMinimum wage

(2014)

SNAP expenditures (2012)Predicted change in SNAP expenditures

(millions of dollars)

Per resident

(millions of

dollars)

Total

(millions of

dollars)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 s

expenditu

(linear regre

Alabama $725 $290 $1397 -$1185 -$1043 -$1043

Alaska $775 $253 $185 -$262 -$223 -$106

Arizona $790 $257 $1686 -$935 -$787 -$892

Arkansas $725 $248 $733 -$636 -$560 -$547

California $800 $189 $7164 -$7276 -$6082 -$3573

Colorado $800 $157 $816 -$862 -$721 -$407

Connecticut $870 $191 $686 -$343 -$275 -$210

Delaware $725 $250 $229 -$205 -$180 -$171

District of Columbia $825 $366 $232 -$146 -$120 -$99

Florida $793 $294 $5676 -$4429 -$3719 -$2951

Georgia $725 $317 $3140 -$2936 -$2584 -$2346

Hawaii $725 $335 $465 -$449 -$395 -$347

Idaho $725 $225 $359 -$376 -$331 -$268

Illinois $825 $249 $3200 -$2096 -$1725 -$1364

Indiana $725 $220 $1439 -$1162 -$1023 -$1075

Iowa $725 $192 $589 -$658 -$579 -$440

Kansas $725 $159 $460 -$502 -$441 -$344

Kentucky $725 $298 $1303 -$1133 -$997 -$973

Louisiana $725 $315 $1450 -$1047 -$922 -$1083

Maine $750 $281 $373 -$267 -$231 -$246

Maryland $725 $188 $1109 -$1765 -$1553 -$828

Massachusetts $800 $206 $1366 -$1030 -$861 -$681

Michigan $740 $300 $2963 -$2400 -$2090 -$2054

Minnesota $725 $140 $755 -$1113 -$980 -$564

Mississippi $725 $326 $973 -$649 -$571 -$726

Missouri $750 $241 $1452 -$1278 -$1104 -$956

Montana $790 $190 $191 -$179 -$151 -$101

Nebraska $725 $140 $259 -$409 -$360 -$194

Nevada $825 $191 $527 -$441 -$363 -$225

New Hampshire $725 $126 $167 -$399 -$351 -$125

TABLE 11

SNAP expenditures Predicted changes by state under the Harkin-Miller bill

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3943

36 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

StateMinimum wage

(2014)

SNAP expenditures (2012)Predicted change in SNAP expenditures

(millions of dollars)

Per resident

(millions of

dollars)

Total

(millions of

dollars)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 state

expenditures

(linear regression)

New Jersey $825 $160 $1420 -$1611 -$1325 -$605

New Mexico $750 $324 $675 -$513 -$444 -$445New York $800 $287 $5616 -$3766 -$3148 -$2801

North Carolina $725 $252 $2454 -$2187 -$1925 -$1833

North Dakota $725 $128 $90 -$162 -$143 -$67

Ohio $795 $259 $2995 -$2013 -$1688 -$1539

Oklahoma $725 $248 $945 -$799 -$703 -$706

Oregon $910 $322 $1255 -$272 -$213 -$262

Pennsylvania $725 $218 $2779 -$2930 -$2579 -$2075

Rhode Island $800 $280 $294 -$173 -$144 -$147

South Carolina $725 $291 $1373 -$1337 -$1177 -$1025South Dakota $725 $198 $165 -$192 -$169 -$123

Tennessee $725 $324 $2091 -$1413 -$1243 -$1562

Texas $725 $230 $5997 -$6402 -$5634 -$4479

Utah $725 $141 $402 -$614 -$541 -$300

Vermont $873 $230 $144 -$66 -$53 -$43

Virginia $725 $173 $1413 -$2062 -$1815 -$1056

Washington $932 $244 $1682 -$350 -$270 -$267

West Virginia $725 $273 $508 -$451 -$397 -$379

Wisconsin $725 $204 $1166 -$1302 -$1146 -$871Wyoming $725 $95 $55 -$105 -$93 -$41

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4043

Appendix C | wwwamericanprogresso

FIGURE 2

Raising the minimum wage to $1010would cut taxpayer costs in every state

Predicted decreases in cost and enrollment

in SNAP in 50 states

$200+$51ndash$100

$101ndash$200

0ndash$25

$26ndash$50

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4143

38 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

Endnotes

1 US Department of Agriculture ldquoSupplementalNutrition Assistance Program (SNAP)rdquo available athttpwwwfnsusdagovsnapeligibility (last accessedFebruary 2014)

2 For this initial analysis we do not consider Harkin-Millerrsquos increase in subminimum wages for tippedworkers To do s o would increase the estimated SNAP

savings by an unknown amount

3 The Congressional Budget Office estimates thatworkers currently earning between $1010 and $1150per hour would see their wages rise under the Harkin-Miller proposal Congressional Budget O ffice ldquoTheEffects of a Minimum Wage Increase on Employmentand Family Incomerdquo (2014)

4 Marianne Page Joanne Spetz and Jane Millar ldquoDoesthe Minimum Wage Affect Welfare Caseloadsrdquo Journalof Policy Analysis and Management 24 (2) (2005)273ndash295

5 Marianne Bitler and Hilary Hoynes ldquoThe More ThingsChange the More They Stay the Same The SafetyNet Living Arrangements and Poverty in the GreatRecessionrdquo NBER Working Paper 194 49 2013

6 Sylvia Allegretto and others ldquoFast Food Poverty Wages The Public Cost of Low-Wage Jobs in the Fast-FoodIndustryrdquo (Berkeley California Center for LaborResearch and Education 2013) available at httplaborcenterberkeleyedupubliccostsfast_food_poverty_wage

7 David Neumark and William Wascher ldquoDoes a HigherMinimum Wage Enhance the Effectiveness of theEarned Income Tax Creditrdquo Industrial and LaborRelations Review 64 (4) (2011) 712ndash746

8 David Lee and Emmanuel Saez ldquoOptimal MinimumWage Policy in Competitive Labor Marketsrdquo Journal ofPublic Economics 96 (9) (2012) 739ndash749

9 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family IncomerdquoFebruary 2014

10 Jesse Rothstein ldquoIs the EITC as Good as an NITConditional Cash Transfers and Tax Incidencerdquo AmericanEconomic Journal Economic Policy 2 (1) (2010) 177ndash208

11 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family Incomerdquo

12 Dube Arindrajit 2013 rdquoMinimum Wagesand the Distribution of Family IncomerdquoUnpublished working paper Available at httpsdldropboxusercontentcomu15038936Dube_MinimumWagesFamilyIncomespdf

13 As Dube explains in ldquoThe poverty of Minimum WageFactsrdquo the simulation approach underestimate stemsfrom a number of unwarranted assumptions includingthe range of actual wage increases and the accuracy ofwage data in the Current Population Survey The causal

approach does not make these assumptions

14 Allegretto Sylvia and others 2013 ldquoCredible ResearchDesigns for Minimum Wage Studiesrdquo Working Paper148-13 University of California Berkeley Institutefor Research on Labor and Employment Available athttpwwwirleberkeleyeduworkingpapers148-13pdf

15 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family In comerdquo

16 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo (2012)

17 Ibid

18 Ibid

19 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo

20 US Department of Agriculture ldquoSupplementalNutrition Assistance Program (SNAP)rdquo

21 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo

22 Ibid

23 We deviate from the Census Bureaursquos definition ofa family unit which is ldquotwo people or more (on e ofwhom is the householder) related by bir th marriageor adoption and residing togetherrdquo US Bureau ofthe Census ldquoCurrent Population Survey DefinitionsFamilyrdquo available at httpwwwcensusgovcpsabout

cpsdefhtml (last accessed February 2014) We countas a family unit any individual residing on his or herown two or more persons residing together whodo not belong to a family in the March CPS sampleare constructed as one family in our analysis For thepurposes of food stamp allocations the consumptionresulting from this transfer is probably distributed tofamily members (rather than household members ora single individual within the household) Howeversingle individuals canmdashand domdashreceive SNAPbenefits Excluding them would fail to make theanalysis reflective of the population at large

24 Strictly the family level linear probability modelpredicts the percentage-point decrease in theprobability that an individual family will receive SNAPpayments When applied to a large number of familieshowever we are able to interpret the coefficient asa decrease in the mean of enrollmentmdashthat is a

decrease in the enrollment ratemdashby applying the lawof iterated expectations

25 We generate expenditure predictions from theenrollment modelsmdashand conversely generateenrollment predictions from the expenditure modelmdashby assuming that expenditures per enrolled familyremains the same before and after the minimum wagechange In practice this is likely to be a conservativeestimatemdashthat is to underestimate the decrease inSNAP activity Average SNAP benefits per family willalso decrease as many families that remain eligible forSNAP experience income gains

26 Wage and Hour Division ldquoMinimum Wage Laws inthe States ndash Januar y 1 2014rdquo available at httpwwwdolgovwhdminwageamericahtm (last accessedFebruary 2014)

27 See for example Sylvia Allegretto and others ldquoCredibleResearch Designs for Minimum Wage Studiesrdquo WorkingPaper 148-113 (Berkeley California Institute forResearch on Labor and Employment 2013) available athttpwwwirleberkeleyeduworkingpapers148-13pdf

28 We will report these results in a forthcoming workingpaper

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4243

Endnotes | wwwamericanprogresso

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4343

The Center for American Progress is a nonpartisan research and educational institute

dedicated to promoting a strong just and free America that ensures opportunity

for all We believe that Americans are bound together by a common commitment to

these values and we aspire to ensure that our national policies reflect these values

We work to find progressive and pragmatic solutions to significant domestic and

international problems and develop policy proposals that foster a government that

is ldquoof the people by the people and for the peoplerdquo

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1543

12 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

Family level model specifications

We firs examine he effec o he minimum wage on paricipaion in public

assisance programs For amily residing in sae and during year we esimae

an equaion o he ollowing orm

(1)

is a binary variable ha is se equal o 1 i a leas one member o amily

received ood samps during he survey year is a se o sae-level

characerisics including annual averages o he unemploymen rae he

employmen-o-populaion raio and he naural log o median amily income

is a vecor o amily atribues including indicaors or he race and marial saus

o he amily head size o he amily he presence o children and he presence oan adul male Sae fixed effecs are capured by o conrol or ime-varying

heerogeneiy our preerred model specificaion also includes year fixed effecs

ha vary by Census division ( ) and sae-level linear ime rends In

Appendix B we jusiy he inclusion o hese las wo erms We also compare he

resuls rom our preerred specificaion wih less sauraed specificaions

Te effec o ineres which is capured by is he expeced change in he

probabiliy o receiving SNAP benefis wih respec o a change in he (log o he)

binding minimum wage in sae during year We repor robus sandard errors

clusered a he sae level We esimae he parameers using linear regression

producing a linear probabiliy model Deails o he model selecion process are

covered in Appendix B below

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1643

Methods and data | wwwamericanprogresso

State-level model specifications

Te sae-level models are similarly specified For sae in year we assume ha

(2)

In his model is now eiher he SNAP enrollmen rae in sae during year

or he naural logarihm o per capia SNAP expendiures in sae during

year is once again a se o sae-level characerisics including he same

sae-level covariaes as in he amily regressions (annual average unemploymen

rae employmen-o-populaion raio naural log o median amily income)

wih he addiion o amily level characerisics averaged across he sae (average

amily size and he shares o populaion consiued by each o five racialehnicgroups) Sae fixed effecs are represened by As above our preerred model

specificaion includes year fixed effecs ha vary by Census division ( ) and

sae-level linear ime rends as elaboraed in Appendix B Te effec o

ineres is capured by

We esimae boh sae-level models (enrollmen and expendiures) using

ordinary leas squares regression Tus he inerpreaion o he coefficien is no

longer ha o a change in probabiliy as in he binary oucome models described

above Raher or he sae-level SNAP enrollmen model represens he

expeced change (in percenage poins) in he saersquos SNAP enrollmen rae ha

is due o a 1 percen change in he minimum wage For he SNAP expendiures

model is simply he elasiciy o SNAP spending wih respec o he minimum

wage991252ha is he percenage change in sae expendiures expeced o resul rom

a 1 percen change in ha saersquos minimum wage For urher deails on model

specificaion reer o Appendix B below

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1743

14 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1843

Results | wwwamericanprogresso

Results

Estimated minimum wage effects on SNAP enrollment and

expenditures

able 1 shows he esimaed parameer o ineres991252he coefficien o he

minimum wage or he preerred model o each ype Coefficiens on he

minimum wage variable are no direcly comparable across models because all

our models have a differen uncional orm o undersand and compare hese

esimaes we compue he change in SNAP aciviy prediced or a paricular wage scenario Te final column in able 1 answers he quesion Wha would be

he expeced change in SNAP aciviy in response o a 10 percen increase in he

minimum wage Te answer o his quesion varies wih he value o he inpu

parameers in he able we calculae he percenage decrease in enrollmen or

expendiures prediced or he average sae wih a minimum wage o $725 in

2014 Te sae-level SNAP expendiure model which is a consan-elasiciy

model conveys elasiciy inormaion direcly or he change in expendiures per

capia in he sae

TABLE 1

Comparison of national SNAP predictions for a 10 percent increase in the federal minimum wage

Model LevelRegression

type

Predicted outcome Coefficient of log

(minimum wage)

(Standard error)

Effect of a 10 percent in

in the minimum wage

VariableForm of

variable

Total

enrollment

Tot

expend

1 Family Linear

probability Enrollment Binary (enrolled=1)

-0042

(0008) -317 N

2

StateLinear regression

(ordinary least

squares)

Enrollment State enrollment rate (percent) -0031

(0012) -235 N

3 Expenditures Log (state expenditures per capita) -0190

(0103) NA -19

plt01 plt005 plt001Note Predicted changes are calculated for the average state with a minimum wage of $725 in 2014

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1943

16 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

According o his model a 10 percen increase in he minimum wage would resul

in a 19 percen reducion in SNAP expendiures Tis resul is reassuringly similar

o Dubersquos resul or povery reducion Te wo enrollmen models are much more

precisely esimaed han he expendiure model Te sae-level enrollmen model

finds ha a 10 percen minimum wage increase in a low-wage sae is associaed wih

a 235 percen decrease in SNAP enrollmens Te amily-level linear probabiliymodel predics a somewha greaer elasiciy or low minimum wage saes an

increase o 10 percen in he ederal minimum would resul in a 317 percen decline

in SNAP enrollmen24 Te differences in hese esimaes sem rom a number o

acors including difference in model uncional orm and daa used We rea his

range o elasiciy esimaes as an upper and lower bound on enrollmen impacs

Harkin-Miller bill National and state-level predicted impacts

Wha would be he prediced change or he SNAP program i he ederalminimum were raised o $1010 as proposed in he Harkin-Miller bill In order o

make his inerence we accoun or he ac ha no all saes are currenly subjec

o he ederal minimum wage a he beginning o 2014 21 saes mainained

higher minimum wages han $725 In hose saes an increase in he ederal

minimum wage may or may no be binding or employers in he sae depending

upon wheher he new ederal minimum exceeds he sae-level minimum Bu

regardless o wheher a minimum wage change is binding he impac on SNAP

aciviy will be lower in high minimum wage saes In order o accoun or his

properly we calculae sae by sae he percenage wage change ha would resul

rom he Harkin-Miller proposal and apply he parameers rom each o he hree

models above o compue he expeced decrease in SNAP aciviy or each sae

In his exercise we use saesrsquo curren (2014) minimum wage levels and assume

as a baseline he 2012 levels o SNAP enrollmen and expendiure as 2012 is he

mos recen year or which SNAP daa are available

able 10 and able 11 in Appendix C repor he esimaed effecs on SNAP

enrollmen and expendiures respecively or each sae under he Harkin-Miller

bill25 An increase o $1010 i enaced oday would represen beween a 393

percen wage increase in a $725 minimum wage sae and an 84 percen increasein Washingon sae which has he highes minimum wage in he naion a $932

as o January 201426 Slighly more han 56 percen o he decrease in expendiures

and abou 59 percen o he decrease in enrollmen would occur in saes wih

presen-day minimum wages o $725

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2043

Results | wwwamericanprogresso

In 2012 hese saes were home o 46 percen o he American populaion

and accouned or a roughly equivalen percenage o oal naional SNAP

expendiures No surprisingly he larges enrollmen decreases come rom

populous saes wih very high SNAP enrollmen raes andor very low minimum

wages Te larges prediced enrollmen reducion991252beween 319000 individuals

and 362000 individuals991252would occur in exas which has a $725 minimum wage In Caliornia which has a 206 percen SNAP paricipaion rae and an $8

per hour minimum wage we anicipae SNAP enrollmen declines o 310000

persons o 371000 persons And in Florida which had a SNAP paricipaion rae

o 166 percen and a $793 minimum wage enrollmen could decline by beween

164000 individuals and 196000 individuals For he our saes ha ogeher

accouned or he greaes amoun o SNAP spending in 2012991252exas Caliornia

Florida and New York respecively991252he combined expendiure reducion rom

he Harkin-Miller bill is prediced o be $14 billion

able 2 summarizes he prediced declines in SNAP aciviy or he naion as a whole ha would resul rom he direc and indirec effecs o he Harkin-Miller

bill Enrollmen would all beween 31 million persons and 36 million persons

represening 75 percen o 87 percen o curren enrollmen Te anicipaed

reducion in program expendiures would be nearly $46 billion or 61 percen o

program expendiures

TABLE 2

Comparison of national SNAP predictions under the Harkin-Miller billrsquos $1010 minimum wage

Model

Enrollment

(persons)

Expenditures

(millions of dollars)

Current (2012) Predicted Change Current (2012) Predicted Chan

Family enrollment (linear probability)

41866195

45489339 -3623144

$74861

NA NA

State enrollment (ordinary least squares) 38745435 -3120759 NA NA

State expenditures (ordinary least squares) NA NA $70305 -$45

Note Calculations use 2014 state minimum wages and the most recent SNAP data from 2012 They assume that per-enrollee expenditures remain constant

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2143

18 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

Tere are o course oher possibiliies or minimum wage legislaion able 3

shows he expeced SNAP changes or he Unied Saes under a variey o wage

scenarios calculaed using he sae-level models I saes were no able o se

heir minimum wages independenly such ha all saes were consrained by

he ederal minimum o $725 SNAP would be received by abou 514000 more

people across he Unied Saes a an addiional program cos o nearly hree-quarers o a billion dollars In conras he effecs o a higher minimum wage

proposal991252a ederal wage floor o $11 per hour991252would decrease enrollmen in

SNAP by more han 10 percen and decrease program coss by 83 percen

TABLE 3

Summary of par ticipation and expenditures under wage scenarios

If all states had

minimum wages of

Enrollment(persons)

Expenditures

(millions of dollars)

Predicted Change Predicted Change

Recent levels (2014) 41866195 $74861

$725 42380520 514326 $75604 $743

$800 41423919 -442276 $74209 -$652

$900 40148451 -1717744 $72350 -$2511

$1000 38872982 -2993212 $70490 -$4371

$1010 38745435 -3120759 $70305 -$4556

$1100 37597514 -4268681 $68631 -$6230

Note Calculations use state-level enrollment model coefficient

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2243

Conclusion | wwwamericanprogresso

Conclusion

An exensive body o lieraure examines employmen effecs o he minimum

wage A much smaller se o sudies examines how he minimum wage affecs

povery and only a handul o papers examine he effecs o he minimum wage

on he EIC Our analysis is he firs o examine he effecs o he minimum

wage on SNAP

Our findings indicae ha increased earnings rom minimum wage changes do

reduce SNAP enrollmens and expendiures We esimae ha he Harkin-Miller bill would save axpayers nearly $46 billion per year equivalen o 61 percen

o SNAP expendiures in 2012 he las year or which daa are available Over a

10-year period he esimaed savings amoun o nearly $46 billion

Our repor is subjec o limiaions ha we expec o overcome in our uure

research Firs he findings do no ake ino accoun possible ineracions among

SNAP he EIC and Medicaid Te eligibiliy cuoffs among hese programs

are quie differen suggesing ha such ineracions may be minor Noneheless

he join effecs can only be deermined by urher research using a causal

model Second i would be useul o know he disribuion o SNAP reducions

along he wage disribuion Using he Congressional Budge Officersquos calculaions

o how much he oal dollar value o wage would increase under he Harkin-

Miller proposal our findings imply ha he decline in overall SNAP spending

equals abou 15 percen o he oal resuling increase in wages Te amoun and

disribuion o his offse are o considerable ineres Minimum wage beneficiaries

who come rom working amilies already well above he povery line would no

see any offse while hose who are currenly considerably below he povery line

will see larger offses Tese issues will also be a subjec or our uure research

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2343

20 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

About the authors

Rachel West is a maser o public policy candidae a he Goldman School

o Public Policy Universiy o Caliornia Berkeley Her research ocuses on

economic policy in he areas o low-wage labor and povery

Michael Reich is proessor o economics and direcor o he Insiue or

Research on Labor and Employmen a he Universiy o Caliornia a Berkeley

His research publicaions cover numerous areas o labor economics including

racial inequaliy labor marke segmenaion high-perormance workplaces

union-managemen cooperaion Japanese labor-managemen sysems living

wages and minimum wages He received his docorae in economics rom

Harvard Universiy

Acknowledgments

We are graeul o Sylvia Allegreto Arindraji Dube Bill Leser Jesse Rohsein

Daniel Tompson and Ben Zipperer or heir valuable suggesions

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2443

References | wwwamericanprogresso

References

Allegreto Sylvia and ohers 2013 ldquoFas Food Povery Wages Te Public Cos o Low-Wage Jobsin he Fas-Food Indusryrdquo Universiy o Caliornia Berkeley Cener or Labor Research andEducaion Available a htplaborcenerberkeleyedupubliccossas_ood_povery_wage

Allegreto Sylvia and ohers 2013 ldquoCredible Research Designs or Minimum Wage Sudiesrdquo

Working Paper 148-13 Universiy o Caliornia Berkeley Insiue or Research on Labor andEmploymen Available a htpwwwirleberkeleyeduworkingpapers148-13pd

Biler Marianne and Hilary Hoynes 2013 ldquo Te More Tings Change he More Tey Say heSame Te Saey Ne Living Arrangemens and Povery in he Grea Recessionrdquo Working Paper19449 Naional Bureau o Economic Research

Congressional Budge Office 2012 ldquoTe Supplemenal Nuriion Assisance Programrdquo Washingon Available a htpwwwcbogovsiesdeaulfilescbofilesatachmens04-19-SNAPpd

991252 991252 991252 2014 ldquoTe Effec o a Minimum-Wage Increase on Employmen and Family Incomerdquo Washingon Available a htpwwwcbogovsiesdeaulfilescbofilesatachmens44995-MinimumWagepd

Dube Arindraji 2013 rdquoMinimum Wages and he Disribuion o Family Incomerdquo Unpublished working paper Available a htpsdldropboxuserconencomu15038936Dube_ MinimumWagesFamilyIncomespd

991252 991252 991252 2014 ldquoTe Povery o Minimum Wage Facsrdquo Arindraji Dube Blog January 22 Availablea htparindubecom20140122he-povery-o-minimum-wage-acs

Lee David and Emmanuel Saez 2012 ldquoOpimal Minimum Wage Policy in Compeiive LaborMarkesrdquo Journal o Public Economics 96 (9) 739ndash749

Neumark David and William Wascher 1992 ldquoEmploymen Effecs o Minimum and Subminimum Wages Panel Daa on Sae Minimum Wage Lawsrdquo Industrial and Labor Relations Review 46 (1)

55ndash81

Neumark David and William Wascher 2011 ldquoDoes a Higher Minimum Wage Enhance heEffeciveness o he Earned Income ax Credirdquo Industrial and Labor Relations Review 64 (4)712ndash746

Page Marianne Joanne Spez and Jane Millar 2005 ldquoDoes he Minimum Wage Affec WelareCaseloadsrdquo Journal o Policy Analysis and Management 24 (2) 273ndash295

Rohsein Jesse 2010 ldquoIs he EIC as Good as an NI Condiional Cash ransers and ax

Incidencerdquo American Economic Journal Economic Policy 2 (1) 177ndash208

Wage and Hours Division 2014 ldquoMinimum Wage Laws in he Saes ndash January 1 2014rdquo USDeparmen o Labor (htpwwwdolgovwhdminwageamericahm [February 2014])

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2543

22 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2643

Appendix A | wwwamericanprogresso

Appendix A

Pre-trend falsification check

Recen minimum wage research27 highlighs a common flaw in previous sudies

ailure o veriy ha he oucome variable is ree o negaive pre-exising rends I

or example SNAP aciviy was already rending down in saes ha raised heir

minimum wages beore hese changes came ino effec our regression analysis could

(misakenly) atribue ha reducion o he minimum wage We check or such

pre-rends by inroducing variables ha represen he prior yearrsquos value or leads o

he minimum wage I he model esimaes he minimum wage o have an effec on

he oucome variable beore he wage change wen ino effec hen an unobservedacor no he minimum wage change caused he change in SNAP aciviy

We es he specificaions above or pre-rends by including a one-year lead in

all hree specificaions We find ha he lead erms are small posiive and no

saisically significan indicaing ha he concurren minimum wage991252no

he wage level in prior periods991252is driving he observed changes in SNAP

oucomes28 In paricular he coefficien (sandard error) on he lead erm in

our preerred amily-level enrollmen regression is 011 and no significan

while he coefficien and sandard error o he conemporaneous minimum

wage is unchanged In he sae-level preerred enrollmen regression he

coefficien o he lead erm is again small (07) and i is no significan Te

corresponding coefficien on he lead erm in he sae-level expendiure

regression is 16 and is no significan Te posiive poin esimaes on hese lead

erms resuls no only rule ou disoring negaive pre-rends Tey also sugges

ha our main resuls may underesimae he rue effecs

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2743

24 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2843

Appendix B | wwwamericanprogresso

Appendix B

Model selection process

For boh he amily-level and sae-level models we es hree mehods o conrol

or unobserved geographic- and ime-varying characerisics as suggesed by he

minimum wage lieraure Firs we include only independen sae-specific fixed

effecs and year-specific fixed effecs Tis specificaion (specificaion 1) implicily

assumes ha amilies in any sae consiue an equally good saisical ldquoconrolrdquo

group or hose in any randomly chosen sae afer accouning or various

characerisics (median income and unemploymen rae among ohers) Similarly

simple ime fixed effecs assume ha amilies surveyed in any year can crediblyserve as a conrol group or amilies surveyed in every oher year o he sample

(1990 hrough 2012)

In oher words specificaion 1 assumes ha a saersquos immediae neighbor provides

no beter a couneracual or he effec o a minimum wage change han does a

sae across he counry We relax his resricive specificaion sequenially in wo

seps In specificaion 2 we replace simple year fixed effecs wih fixed effecs or

each Census divisionyear (capured as an addiional variable in he vecor By

using division-year effecs we remove he resricion ha amilies in each sae

are equally good saisical conrols or all oher amilies Raher we allow or he

possibiliy ha amilies in similar geographic regions (or example he Souh or

he Norheas) may be more similar o one anoher han amilies arher away

Finally in specificaion 3 we add sae-specific linear ime rends o he previous

specificaion Tus specificaion 3 is he mos rigorous model specificaion in ha

i allows or heerogeneiy along hree dimensions Ta is specificaion 3 allows

each sae o have is own ime-varying rends raher han imposing he resricion

ha saes evolve idenically over he 22 years in he sample

We begin building he heoreical specificaion above rom a se o simpleuncondiional models regression o SNAP aciviy (enrollmen or expendiures)

on he log o he minimum wage and a se o geographic- and ime-specific

effecs (specificaions 1 2 and 3 described above) As shown in ables 1ndash3 (or

specificaion 3) we hen add covariaes sequenially o hese models including

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2943

26 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

firs he vecor o amily-level conrols ollowed by each o several sae-level

covariaes in urn he unemploymen rae log o median-amily income and he

employmen-o-populaion raio Comparable resuls or specificaions 2 and 3

will be available in our orhcoming working paper

As expeced he simple uncondiional models indicae ha he relaionship beween he minimum wage and SNAP enrollmen i one exiss is a more complex

one influenced by oher acors In he uncondiional model he coefficien on he

variable o ineres991252he log o he minimum wage991252is small in magniude and no

saisically differen rom zero Once we accoun or he influence o labor marke

condiions and variaion in income levels on program paricipaion (by including

unemploymen rae and median-amily income conrol variables respecively)

he effec o he minimum wage on SNAP enrollmen is precisely esimaed Te

coefficien o he log minimum wage is slighly higher (-0042) in he amily-level

analysis han he coefficien (-031) in he sae-level analysis Te level o precision

is also higher in he amily-level analysis Tis is o be expeced when using 124million observaions compared o 1127

Te second dimension o model choice concerns he effec specificaion ables

7ndash9 compare he primary coefficiens o ineres or he SNAP enrollmen and

expendiure models For boh he enrollmen models he effec sizes are smalles

or specificaion 1 larges or specificaion 2 and inermediae beween hese wo

in specificaion 3 Recall ha Specificaion 3 conains sae-specific linear ime

rends in addiion o he census divisionyear conrols included in specificaion

2 In he amily-level enrollmen model he sandard error o he minimum wage

coefficien is smaller han in he oher wo specificaions Sandard errors on he

oher variables are much smaller in specificaions 2 and 3 han in specificaion

1 On he basis o coefficien significance (join and individual) specificaions 2

and 3 are sricly preerred in boh enrollmen models o specificaion 1 which

conains only sae and year fixed effecs

A concern wih specificaions 2 and 3 is ha rend conrols such as sae linear

rends may incorrecly absorb some o he delayed impac o a minimum wage

When we es his issue by including lagged minimum wages we do no find ha

delayed effecs are significan Anoher concern is ha more sauraed modelsuse less o he saisical variaion which could reduce he saisical power o

he resuls However he sandard errors or our more sauraed models are no

higher and are lower in some cases han or he less sauraed models Overall

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3043

Appendix B | wwwamericanprogresso

his evidence suppors our use o he sauraed model as he preerred model

specificaion Moreover Dubersquos 2013 sudy shows ha more sauraed models

perorm beter han models wih jus sae and ime fixed effecs

Te esimaed enrollmen regressions a boh he amily and sae levels show large

and saisically significan coefficiens Te esimaed minimum wage effec in heexpendiures regressions991252or which we have only sae-level daa991252is also large

and saisically significan

We do no use weighed regression or he sae-level models preerring o keep

analysis o he ldquoreamenrdquo (ha is o say a minimum wage change) appropriae

o he average sae raher han he average amily or individual I insead our

primary ineres were he impac o a minimum wage change on he average amily

or he average individual we migh choose o designae he number o amilies

in each sae or he sae populaion respecively as analyic weighs in order o

obain a coefficien beter suied or such inerence

TABLE 4

SNAP enrollment

Family-level linear probability

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0048 -0047 -0040 -0043 -0042

(0013) (0013) (001) (0008) (0009)

Unemployment rate (100) 0505 0420 0280

(0083) (0086) (0082)

Log median income -0057 -0039

(0011) (0011)

Employment-to-population ratio -0239

(0038)

N 1242022 1242022 1242022 1242022 1242022

plt01 plt005 plt001

Note Robust standard errors are in parentheses Observations are clustered at the state level The outcome variable is binary and equal to one if a family is enrolledin SNAP All models include state fixed effects Census division x-year fixed effects and state-specific linear time trends All specifications except 3a include additional

controls for family size race and marital status of the family head presence of children and presence of an adult maleSource Annual data from the March Current Population Survey for 1990 to 2012 estimation includes Current Population Survey probability weights

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3143

28 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

TABLE 5

SNAP enrollment

State-level linear regression

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0028 -0024 -0026 -0031 -0031

(0014) (0013) (0013) (0012) (0012)

Unemployment rate (100) 0625 0466 0320 0339

(0087) (0088) (0085) (0083)

Log median income -0090 -0065 -0061

(0013) (0013) (0013)

Employment-to-population ratio -0282 -0248

(0037) (0038)

N 1127 1127 1127 1127 1127

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the SNAP enrollment rate All models include state fixed effects Census division x-year fixedeffects and state-specific linear time trends

Source Annual data from the March Current Population Survey for 1990 to 2012

TABLE 6

SNAP expenditures

State-level linear regression

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0181 -0149 -0156 -0153 -0190

(011) (0103) (0102) (0103) (0103)

Unemployment rate (100) 4587 4152 4232 4313

(0622) (0621) (0633) (0628)

Log median income -0246 -0261 -0294

(0075) (0078) (0078)

Employment-to-population ratio 0155 0244

(0237) (024)

N 1127 1127 1127 1127 1127

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the natural log of state SNAP expenditures per capita for 1990 to 2012 All models include state

fixed effects Census division x-year fixed effects and state-specific linear time trends

Source Annual data from Bureau of Economic Analysis national income and product accounts tables for 1990 to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3243

Appendix B | wwwamericanprogresso

TABLE 7

Comparison of specifications SNAP enrollment

Family-level linear probability

(1) (2) (3)

Log minimum wage -0028 -0049 -0042

(0014) (0017) (0009)

Unemployment rate (100) 0275 0297 0280

(0161) (0076) (0082)

Log median income -0077 -0055 -0039

(0014) (0012) (0011)

Employment-to-population ratio -0238 -0250 -0239

(0054) (004) (0038)

N 1242022 1242022 1242022

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses Observations are clustered at the state level The outcomevariable is binary or equal to one if a family is enrolled in SNAP All specifications include additional controls forfamily size race and marital status of the family head presence of children and presence of an adult male

Source Annual data from the March Current Population Survey for 1990 to 2012 estimation includes CurrentPopulation Survey probability weights

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3343

30 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

TABLE 8

Comparison of specifications SNAP enrollment

State-level linear regression

(1) (2) (3)

Log minimum wage -0019 -0035 -0031

(0009) (0012) (0012)

Unemployment rate (100) 0401 0370 0339

(0063) (0077) (0083)

Log median income -0081 -0073 -0061

(0011) (0013) (0013)

Employment-to-population ratio -0183 -0222 -0248

(0039) (0039) (0038)

N 1127 1127 1127

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the SNAP enrollment rate Allregressions include the share of households with children and the racial shares of the population

Source Annual data from the March Current Population Survey for 1990 to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3443

Appendix B | wwwamericanprogresso

TABLE 9

Comparison of specifications SNAP expenditures

State-level linear regression

(1) (2) (3)

Log minimum wage -0121 -0203 -0190

(0075) (0103) (0103)

Unemployment rate (100) 5292 5152 4313

(0464) (0576) (0628)

Log median income -0437 -0417 -0294

(008) (0086) (0078)

Employment-to-population ratio -0040 -0220 0244

(0261) (0260) (0240)

N 1127 1127 1127

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the natural log of state SNAPexpenditures per capita All regressions include the share of households with children and the racial shares ofthe state population

Source Annual data from Bureau of Economic Analysis national income and product accounts tables for 1990to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3543

32 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3643

Appendix C | wwwamericanprogresso

Appendix C Harkin-Miller

policy simulation results

StateMinimum wage

(2014)

SNAP enrollment (2012) Predicted change in SNAP enrollment

RateTotal

(persons)

Model 1 family

enrollment (linearprobability)

Model 2 state

enrollment (linearregression)

Model 3 s

expenditu(linear regre

Alabama $725 164 788682 -66922 -58897 -58906

Alaska $775 120 87436 -8104 -6887 -3288

Arizona $790 201 1319323 -67435 -56738 -64356

Arkansas $725 78 230489 -40977 -36063 -35248

California $800 206 7813680 -371131 -310222 -18223

Colorado $800 164 853155 -50684 -42365 -23926

Connecticut $870 91 326621 -22456 -17975 -13711

Delaware $725 186 170262 -12739 -11211 -10647

District of Columbia $825 133 84009 -5370 -4417 -3632

Florida $793 166 3208026 -195813 -164426 -13046

Georgia $725 160 1586336 -137741 -121224 -11004

Hawaii $725 96 133662 -19310 -16995 -14933

Idaho $725 92 147501 -22165 -19507 -15809

Illinois $825 95 1225084 -109088 -89742 -70955

Indiana $725 125 816233 -90818 -79928 -83985

Iowa $725 155 478011 -42716 -37594 -28556

Kansas $725 135 388269 -40082 -35275 -27461

Kentucky $725 130 568821 -60840 -53544 -52259

Louisiana $725 149 683832 -63929 -56263 -66083

Maine $750 77 101976 -16567 -14323 -15234

Maryland $725 144 846415 -81748 -71946 -38370

Massachusetts $800 130 864721 -64902 -54251 -42913

Michigan $740 146 1439141 -128801 -112140 -11022

Minnesota $725 133 713646 -74730 -65769 -37878

TABLE 10

SNAP enrollments Predicted changes by state under the Harkin-Miller bill

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3743

34 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

StateMinimum wage

(2014)

SNAP enrollment (2012) Predicted change in SNAP enrollment

RateTotal

(persons)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 state

expenditures

(linear regression)

Mississippi $725 129 386501 -41486 -36511 -46467

Missouri $750 172 1036182 -75131 -64952 -56244

Montana $790 132 132452 -10350 -8708 -5846

Nebraska $725 124 230683 -25773 -22683 -12189

Nevada $825 162 446035 -23349 -19209 -11894

New Hampshire $725 127 168404 -18359 -16157 -5735

New Jersey $825 160 1416666 -75175 -61843 -28236

New Mexico $750 149 310896 -25983 -22463 -22512

New York $800 192 3763553 -191193 -159815 -142182

North Carolina $725 174 1697193 -135417 -119179 -113503

North Dakota $725 87 61225 -9743 -8574 -4021

Ohio $795 143 1647345 -115869 -97169 -88580

Oklahoma $725 129 494053 -53006 -46650 -46854

Oregon $910 124 485326 -17036 -13328 -16398

Pennsylvania $725 161 2053643 -177315 -156052 -125586

Rhode Island $800 156 163730 -10258 -8574 -8698

South Carolina $725 94 445277 -65614 -57746 -50304

South Dakota $725 208 173749 -11586 -10197 -7458

Tennessee $725 142 914903 -89667 -78915 -99134

Texas $725 110 2863779 -362018 -318607 -253285

Utah $725 88 251107 -39658 -34902 -19390

Vermont $873 156 97792 -3823 -3055 -2475

Virginia $725 101 829771 -113723 -100086 -58212

Washington $932 72 496934 -23221 -17947 -17756

West Virginia $725 58 107875 -25792 -22699 -21665

Wisconsin $725 75 427822 -79521 -69986 -53210

Wyoming $725 164 94590 -8010 -7050 -3104

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3843

Appendix C | wwwamericanprogresso

StateMinimum wage

(2014)

SNAP expenditures (2012)Predicted change in SNAP expenditures

(millions of dollars)

Per resident

(millions of

dollars)

Total

(millions of

dollars)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 s

expenditu

(linear regre

Alabama $725 $290 $1397 -$1185 -$1043 -$1043

Alaska $775 $253 $185 -$262 -$223 -$106

Arizona $790 $257 $1686 -$935 -$787 -$892

Arkansas $725 $248 $733 -$636 -$560 -$547

California $800 $189 $7164 -$7276 -$6082 -$3573

Colorado $800 $157 $816 -$862 -$721 -$407

Connecticut $870 $191 $686 -$343 -$275 -$210

Delaware $725 $250 $229 -$205 -$180 -$171

District of Columbia $825 $366 $232 -$146 -$120 -$99

Florida $793 $294 $5676 -$4429 -$3719 -$2951

Georgia $725 $317 $3140 -$2936 -$2584 -$2346

Hawaii $725 $335 $465 -$449 -$395 -$347

Idaho $725 $225 $359 -$376 -$331 -$268

Illinois $825 $249 $3200 -$2096 -$1725 -$1364

Indiana $725 $220 $1439 -$1162 -$1023 -$1075

Iowa $725 $192 $589 -$658 -$579 -$440

Kansas $725 $159 $460 -$502 -$441 -$344

Kentucky $725 $298 $1303 -$1133 -$997 -$973

Louisiana $725 $315 $1450 -$1047 -$922 -$1083

Maine $750 $281 $373 -$267 -$231 -$246

Maryland $725 $188 $1109 -$1765 -$1553 -$828

Massachusetts $800 $206 $1366 -$1030 -$861 -$681

Michigan $740 $300 $2963 -$2400 -$2090 -$2054

Minnesota $725 $140 $755 -$1113 -$980 -$564

Mississippi $725 $326 $973 -$649 -$571 -$726

Missouri $750 $241 $1452 -$1278 -$1104 -$956

Montana $790 $190 $191 -$179 -$151 -$101

Nebraska $725 $140 $259 -$409 -$360 -$194

Nevada $825 $191 $527 -$441 -$363 -$225

New Hampshire $725 $126 $167 -$399 -$351 -$125

TABLE 11

SNAP expenditures Predicted changes by state under the Harkin-Miller bill

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3943

36 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

StateMinimum wage

(2014)

SNAP expenditures (2012)Predicted change in SNAP expenditures

(millions of dollars)

Per resident

(millions of

dollars)

Total

(millions of

dollars)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 state

expenditures

(linear regression)

New Jersey $825 $160 $1420 -$1611 -$1325 -$605

New Mexico $750 $324 $675 -$513 -$444 -$445New York $800 $287 $5616 -$3766 -$3148 -$2801

North Carolina $725 $252 $2454 -$2187 -$1925 -$1833

North Dakota $725 $128 $90 -$162 -$143 -$67

Ohio $795 $259 $2995 -$2013 -$1688 -$1539

Oklahoma $725 $248 $945 -$799 -$703 -$706

Oregon $910 $322 $1255 -$272 -$213 -$262

Pennsylvania $725 $218 $2779 -$2930 -$2579 -$2075

Rhode Island $800 $280 $294 -$173 -$144 -$147

South Carolina $725 $291 $1373 -$1337 -$1177 -$1025South Dakota $725 $198 $165 -$192 -$169 -$123

Tennessee $725 $324 $2091 -$1413 -$1243 -$1562

Texas $725 $230 $5997 -$6402 -$5634 -$4479

Utah $725 $141 $402 -$614 -$541 -$300

Vermont $873 $230 $144 -$66 -$53 -$43

Virginia $725 $173 $1413 -$2062 -$1815 -$1056

Washington $932 $244 $1682 -$350 -$270 -$267

West Virginia $725 $273 $508 -$451 -$397 -$379

Wisconsin $725 $204 $1166 -$1302 -$1146 -$871Wyoming $725 $95 $55 -$105 -$93 -$41

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4043

Appendix C | wwwamericanprogresso

FIGURE 2

Raising the minimum wage to $1010would cut taxpayer costs in every state

Predicted decreases in cost and enrollment

in SNAP in 50 states

$200+$51ndash$100

$101ndash$200

0ndash$25

$26ndash$50

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4143

38 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

Endnotes

1 US Department of Agriculture ldquoSupplementalNutrition Assistance Program (SNAP)rdquo available athttpwwwfnsusdagovsnapeligibility (last accessedFebruary 2014)

2 For this initial analysis we do not consider Harkin-Millerrsquos increase in subminimum wages for tippedworkers To do s o would increase the estimated SNAP

savings by an unknown amount

3 The Congressional Budget Office estimates thatworkers currently earning between $1010 and $1150per hour would see their wages rise under the Harkin-Miller proposal Congressional Budget O ffice ldquoTheEffects of a Minimum Wage Increase on Employmentand Family Incomerdquo (2014)

4 Marianne Page Joanne Spetz and Jane Millar ldquoDoesthe Minimum Wage Affect Welfare Caseloadsrdquo Journalof Policy Analysis and Management 24 (2) (2005)273ndash295

5 Marianne Bitler and Hilary Hoynes ldquoThe More ThingsChange the More They Stay the Same The SafetyNet Living Arrangements and Poverty in the GreatRecessionrdquo NBER Working Paper 194 49 2013

6 Sylvia Allegretto and others ldquoFast Food Poverty Wages The Public Cost of Low-Wage Jobs in the Fast-FoodIndustryrdquo (Berkeley California Center for LaborResearch and Education 2013) available at httplaborcenterberkeleyedupubliccostsfast_food_poverty_wage

7 David Neumark and William Wascher ldquoDoes a HigherMinimum Wage Enhance the Effectiveness of theEarned Income Tax Creditrdquo Industrial and LaborRelations Review 64 (4) (2011) 712ndash746

8 David Lee and Emmanuel Saez ldquoOptimal MinimumWage Policy in Competitive Labor Marketsrdquo Journal ofPublic Economics 96 (9) (2012) 739ndash749

9 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family IncomerdquoFebruary 2014

10 Jesse Rothstein ldquoIs the EITC as Good as an NITConditional Cash Transfers and Tax Incidencerdquo AmericanEconomic Journal Economic Policy 2 (1) (2010) 177ndash208

11 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family Incomerdquo

12 Dube Arindrajit 2013 rdquoMinimum Wagesand the Distribution of Family IncomerdquoUnpublished working paper Available at httpsdldropboxusercontentcomu15038936Dube_MinimumWagesFamilyIncomespdf

13 As Dube explains in ldquoThe poverty of Minimum WageFactsrdquo the simulation approach underestimate stemsfrom a number of unwarranted assumptions includingthe range of actual wage increases and the accuracy ofwage data in the Current Population Survey The causal

approach does not make these assumptions

14 Allegretto Sylvia and others 2013 ldquoCredible ResearchDesigns for Minimum Wage Studiesrdquo Working Paper148-13 University of California Berkeley Institutefor Research on Labor and Employment Available athttpwwwirleberkeleyeduworkingpapers148-13pdf

15 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family In comerdquo

16 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo (2012)

17 Ibid

18 Ibid

19 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo

20 US Department of Agriculture ldquoSupplementalNutrition Assistance Program (SNAP)rdquo

21 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo

22 Ibid

23 We deviate from the Census Bureaursquos definition ofa family unit which is ldquotwo people or more (on e ofwhom is the householder) related by bir th marriageor adoption and residing togetherrdquo US Bureau ofthe Census ldquoCurrent Population Survey DefinitionsFamilyrdquo available at httpwwwcensusgovcpsabout

cpsdefhtml (last accessed February 2014) We countas a family unit any individual residing on his or herown two or more persons residing together whodo not belong to a family in the March CPS sampleare constructed as one family in our analysis For thepurposes of food stamp allocations the consumptionresulting from this transfer is probably distributed tofamily members (rather than household members ora single individual within the household) Howeversingle individuals canmdashand domdashreceive SNAPbenefits Excluding them would fail to make theanalysis reflective of the population at large

24 Strictly the family level linear probability modelpredicts the percentage-point decrease in theprobability that an individual family will receive SNAPpayments When applied to a large number of familieshowever we are able to interpret the coefficient asa decrease in the mean of enrollmentmdashthat is a

decrease in the enrollment ratemdashby applying the lawof iterated expectations

25 We generate expenditure predictions from theenrollment modelsmdashand conversely generateenrollment predictions from the expenditure modelmdashby assuming that expenditures per enrolled familyremains the same before and after the minimum wagechange In practice this is likely to be a conservativeestimatemdashthat is to underestimate the decrease inSNAP activity Average SNAP benefits per family willalso decrease as many families that remain eligible forSNAP experience income gains

26 Wage and Hour Division ldquoMinimum Wage Laws inthe States ndash Januar y 1 2014rdquo available at httpwwwdolgovwhdminwageamericahtm (last accessedFebruary 2014)

27 See for example Sylvia Allegretto and others ldquoCredibleResearch Designs for Minimum Wage Studiesrdquo WorkingPaper 148-113 (Berkeley California Institute forResearch on Labor and Employment 2013) available athttpwwwirleberkeleyeduworkingpapers148-13pdf

28 We will report these results in a forthcoming workingpaper

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4243

Endnotes | wwwamericanprogresso

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4343

The Center for American Progress is a nonpartisan research and educational institute

dedicated to promoting a strong just and free America that ensures opportunity

for all We believe that Americans are bound together by a common commitment to

these values and we aspire to ensure that our national policies reflect these values

We work to find progressive and pragmatic solutions to significant domestic and

international problems and develop policy proposals that foster a government that

is ldquoof the people by the people and for the peoplerdquo

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1643

Methods and data | wwwamericanprogresso

State-level model specifications

Te sae-level models are similarly specified For sae in year we assume ha

(2)

In his model is now eiher he SNAP enrollmen rae in sae during year

or he naural logarihm o per capia SNAP expendiures in sae during

year is once again a se o sae-level characerisics including he same

sae-level covariaes as in he amily regressions (annual average unemploymen

rae employmen-o-populaion raio naural log o median amily income)

wih he addiion o amily level characerisics averaged across he sae (average

amily size and he shares o populaion consiued by each o five racialehnicgroups) Sae fixed effecs are represened by As above our preerred model

specificaion includes year fixed effecs ha vary by Census division ( ) and

sae-level linear ime rends as elaboraed in Appendix B Te effec o

ineres is capured by

We esimae boh sae-level models (enrollmen and expendiures) using

ordinary leas squares regression Tus he inerpreaion o he coefficien is no

longer ha o a change in probabiliy as in he binary oucome models described

above Raher or he sae-level SNAP enrollmen model represens he

expeced change (in percenage poins) in he saersquos SNAP enrollmen rae ha

is due o a 1 percen change in he minimum wage For he SNAP expendiures

model is simply he elasiciy o SNAP spending wih respec o he minimum

wage991252ha is he percenage change in sae expendiures expeced o resul rom

a 1 percen change in ha saersquos minimum wage For urher deails on model

specificaion reer o Appendix B below

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1743

14 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1843

Results | wwwamericanprogresso

Results

Estimated minimum wage effects on SNAP enrollment and

expenditures

able 1 shows he esimaed parameer o ineres991252he coefficien o he

minimum wage or he preerred model o each ype Coefficiens on he

minimum wage variable are no direcly comparable across models because all

our models have a differen uncional orm o undersand and compare hese

esimaes we compue he change in SNAP aciviy prediced or a paricular wage scenario Te final column in able 1 answers he quesion Wha would be

he expeced change in SNAP aciviy in response o a 10 percen increase in he

minimum wage Te answer o his quesion varies wih he value o he inpu

parameers in he able we calculae he percenage decrease in enrollmen or

expendiures prediced or he average sae wih a minimum wage o $725 in

2014 Te sae-level SNAP expendiure model which is a consan-elasiciy

model conveys elasiciy inormaion direcly or he change in expendiures per

capia in he sae

TABLE 1

Comparison of national SNAP predictions for a 10 percent increase in the federal minimum wage

Model LevelRegression

type

Predicted outcome Coefficient of log

(minimum wage)

(Standard error)

Effect of a 10 percent in

in the minimum wage

VariableForm of

variable

Total

enrollment

Tot

expend

1 Family Linear

probability Enrollment Binary (enrolled=1)

-0042

(0008) -317 N

2

StateLinear regression

(ordinary least

squares)

Enrollment State enrollment rate (percent) -0031

(0012) -235 N

3 Expenditures Log (state expenditures per capita) -0190

(0103) NA -19

plt01 plt005 plt001Note Predicted changes are calculated for the average state with a minimum wage of $725 in 2014

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1943

16 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

According o his model a 10 percen increase in he minimum wage would resul

in a 19 percen reducion in SNAP expendiures Tis resul is reassuringly similar

o Dubersquos resul or povery reducion Te wo enrollmen models are much more

precisely esimaed han he expendiure model Te sae-level enrollmen model

finds ha a 10 percen minimum wage increase in a low-wage sae is associaed wih

a 235 percen decrease in SNAP enrollmens Te amily-level linear probabiliymodel predics a somewha greaer elasiciy or low minimum wage saes an

increase o 10 percen in he ederal minimum would resul in a 317 percen decline

in SNAP enrollmen24 Te differences in hese esimaes sem rom a number o

acors including difference in model uncional orm and daa used We rea his

range o elasiciy esimaes as an upper and lower bound on enrollmen impacs

Harkin-Miller bill National and state-level predicted impacts

Wha would be he prediced change or he SNAP program i he ederalminimum were raised o $1010 as proposed in he Harkin-Miller bill In order o

make his inerence we accoun or he ac ha no all saes are currenly subjec

o he ederal minimum wage a he beginning o 2014 21 saes mainained

higher minimum wages han $725 In hose saes an increase in he ederal

minimum wage may or may no be binding or employers in he sae depending

upon wheher he new ederal minimum exceeds he sae-level minimum Bu

regardless o wheher a minimum wage change is binding he impac on SNAP

aciviy will be lower in high minimum wage saes In order o accoun or his

properly we calculae sae by sae he percenage wage change ha would resul

rom he Harkin-Miller proposal and apply he parameers rom each o he hree

models above o compue he expeced decrease in SNAP aciviy or each sae

In his exercise we use saesrsquo curren (2014) minimum wage levels and assume

as a baseline he 2012 levels o SNAP enrollmen and expendiure as 2012 is he

mos recen year or which SNAP daa are available

able 10 and able 11 in Appendix C repor he esimaed effecs on SNAP

enrollmen and expendiures respecively or each sae under he Harkin-Miller

bill25 An increase o $1010 i enaced oday would represen beween a 393

percen wage increase in a $725 minimum wage sae and an 84 percen increasein Washingon sae which has he highes minimum wage in he naion a $932

as o January 201426 Slighly more han 56 percen o he decrease in expendiures

and abou 59 percen o he decrease in enrollmen would occur in saes wih

presen-day minimum wages o $725

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2043

Results | wwwamericanprogresso

In 2012 hese saes were home o 46 percen o he American populaion

and accouned or a roughly equivalen percenage o oal naional SNAP

expendiures No surprisingly he larges enrollmen decreases come rom

populous saes wih very high SNAP enrollmen raes andor very low minimum

wages Te larges prediced enrollmen reducion991252beween 319000 individuals

and 362000 individuals991252would occur in exas which has a $725 minimum wage In Caliornia which has a 206 percen SNAP paricipaion rae and an $8

per hour minimum wage we anicipae SNAP enrollmen declines o 310000

persons o 371000 persons And in Florida which had a SNAP paricipaion rae

o 166 percen and a $793 minimum wage enrollmen could decline by beween

164000 individuals and 196000 individuals For he our saes ha ogeher

accouned or he greaes amoun o SNAP spending in 2012991252exas Caliornia

Florida and New York respecively991252he combined expendiure reducion rom

he Harkin-Miller bill is prediced o be $14 billion

able 2 summarizes he prediced declines in SNAP aciviy or he naion as a whole ha would resul rom he direc and indirec effecs o he Harkin-Miller

bill Enrollmen would all beween 31 million persons and 36 million persons

represening 75 percen o 87 percen o curren enrollmen Te anicipaed

reducion in program expendiures would be nearly $46 billion or 61 percen o

program expendiures

TABLE 2

Comparison of national SNAP predictions under the Harkin-Miller billrsquos $1010 minimum wage

Model

Enrollment

(persons)

Expenditures

(millions of dollars)

Current (2012) Predicted Change Current (2012) Predicted Chan

Family enrollment (linear probability)

41866195

45489339 -3623144

$74861

NA NA

State enrollment (ordinary least squares) 38745435 -3120759 NA NA

State expenditures (ordinary least squares) NA NA $70305 -$45

Note Calculations use 2014 state minimum wages and the most recent SNAP data from 2012 They assume that per-enrollee expenditures remain constant

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2143

18 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

Tere are o course oher possibiliies or minimum wage legislaion able 3

shows he expeced SNAP changes or he Unied Saes under a variey o wage

scenarios calculaed using he sae-level models I saes were no able o se

heir minimum wages independenly such ha all saes were consrained by

he ederal minimum o $725 SNAP would be received by abou 514000 more

people across he Unied Saes a an addiional program cos o nearly hree-quarers o a billion dollars In conras he effecs o a higher minimum wage

proposal991252a ederal wage floor o $11 per hour991252would decrease enrollmen in

SNAP by more han 10 percen and decrease program coss by 83 percen

TABLE 3

Summary of par ticipation and expenditures under wage scenarios

If all states had

minimum wages of

Enrollment(persons)

Expenditures

(millions of dollars)

Predicted Change Predicted Change

Recent levels (2014) 41866195 $74861

$725 42380520 514326 $75604 $743

$800 41423919 -442276 $74209 -$652

$900 40148451 -1717744 $72350 -$2511

$1000 38872982 -2993212 $70490 -$4371

$1010 38745435 -3120759 $70305 -$4556

$1100 37597514 -4268681 $68631 -$6230

Note Calculations use state-level enrollment model coefficient

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2243

Conclusion | wwwamericanprogresso

Conclusion

An exensive body o lieraure examines employmen effecs o he minimum

wage A much smaller se o sudies examines how he minimum wage affecs

povery and only a handul o papers examine he effecs o he minimum wage

on he EIC Our analysis is he firs o examine he effecs o he minimum

wage on SNAP

Our findings indicae ha increased earnings rom minimum wage changes do

reduce SNAP enrollmens and expendiures We esimae ha he Harkin-Miller bill would save axpayers nearly $46 billion per year equivalen o 61 percen

o SNAP expendiures in 2012 he las year or which daa are available Over a

10-year period he esimaed savings amoun o nearly $46 billion

Our repor is subjec o limiaions ha we expec o overcome in our uure

research Firs he findings do no ake ino accoun possible ineracions among

SNAP he EIC and Medicaid Te eligibiliy cuoffs among hese programs

are quie differen suggesing ha such ineracions may be minor Noneheless

he join effecs can only be deermined by urher research using a causal

model Second i would be useul o know he disribuion o SNAP reducions

along he wage disribuion Using he Congressional Budge Officersquos calculaions

o how much he oal dollar value o wage would increase under he Harkin-

Miller proposal our findings imply ha he decline in overall SNAP spending

equals abou 15 percen o he oal resuling increase in wages Te amoun and

disribuion o his offse are o considerable ineres Minimum wage beneficiaries

who come rom working amilies already well above he povery line would no

see any offse while hose who are currenly considerably below he povery line

will see larger offses Tese issues will also be a subjec or our uure research

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2343

20 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

About the authors

Rachel West is a maser o public policy candidae a he Goldman School

o Public Policy Universiy o Caliornia Berkeley Her research ocuses on

economic policy in he areas o low-wage labor and povery

Michael Reich is proessor o economics and direcor o he Insiue or

Research on Labor and Employmen a he Universiy o Caliornia a Berkeley

His research publicaions cover numerous areas o labor economics including

racial inequaliy labor marke segmenaion high-perormance workplaces

union-managemen cooperaion Japanese labor-managemen sysems living

wages and minimum wages He received his docorae in economics rom

Harvard Universiy

Acknowledgments

We are graeul o Sylvia Allegreto Arindraji Dube Bill Leser Jesse Rohsein

Daniel Tompson and Ben Zipperer or heir valuable suggesions

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2443

References | wwwamericanprogresso

References

Allegreto Sylvia and ohers 2013 ldquoFas Food Povery Wages Te Public Cos o Low-Wage Jobsin he Fas-Food Indusryrdquo Universiy o Caliornia Berkeley Cener or Labor Research andEducaion Available a htplaborcenerberkeleyedupubliccossas_ood_povery_wage

Allegreto Sylvia and ohers 2013 ldquoCredible Research Designs or Minimum Wage Sudiesrdquo

Working Paper 148-13 Universiy o Caliornia Berkeley Insiue or Research on Labor andEmploymen Available a htpwwwirleberkeleyeduworkingpapers148-13pd

Biler Marianne and Hilary Hoynes 2013 ldquo Te More Tings Change he More Tey Say heSame Te Saey Ne Living Arrangemens and Povery in he Grea Recessionrdquo Working Paper19449 Naional Bureau o Economic Research

Congressional Budge Office 2012 ldquoTe Supplemenal Nuriion Assisance Programrdquo Washingon Available a htpwwwcbogovsiesdeaulfilescbofilesatachmens04-19-SNAPpd

991252 991252 991252 2014 ldquoTe Effec o a Minimum-Wage Increase on Employmen and Family Incomerdquo Washingon Available a htpwwwcbogovsiesdeaulfilescbofilesatachmens44995-MinimumWagepd

Dube Arindraji 2013 rdquoMinimum Wages and he Disribuion o Family Incomerdquo Unpublished working paper Available a htpsdldropboxuserconencomu15038936Dube_ MinimumWagesFamilyIncomespd

991252 991252 991252 2014 ldquoTe Povery o Minimum Wage Facsrdquo Arindraji Dube Blog January 22 Availablea htparindubecom20140122he-povery-o-minimum-wage-acs

Lee David and Emmanuel Saez 2012 ldquoOpimal Minimum Wage Policy in Compeiive LaborMarkesrdquo Journal o Public Economics 96 (9) 739ndash749

Neumark David and William Wascher 1992 ldquoEmploymen Effecs o Minimum and Subminimum Wages Panel Daa on Sae Minimum Wage Lawsrdquo Industrial and Labor Relations Review 46 (1)

55ndash81

Neumark David and William Wascher 2011 ldquoDoes a Higher Minimum Wage Enhance heEffeciveness o he Earned Income ax Credirdquo Industrial and Labor Relations Review 64 (4)712ndash746

Page Marianne Joanne Spez and Jane Millar 2005 ldquoDoes he Minimum Wage Affec WelareCaseloadsrdquo Journal o Policy Analysis and Management 24 (2) 273ndash295

Rohsein Jesse 2010 ldquoIs he EIC as Good as an NI Condiional Cash ransers and ax

Incidencerdquo American Economic Journal Economic Policy 2 (1) 177ndash208

Wage and Hours Division 2014 ldquoMinimum Wage Laws in he Saes ndash January 1 2014rdquo USDeparmen o Labor (htpwwwdolgovwhdminwageamericahm [February 2014])

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2543

22 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2643

Appendix A | wwwamericanprogresso

Appendix A

Pre-trend falsification check

Recen minimum wage research27 highlighs a common flaw in previous sudies

ailure o veriy ha he oucome variable is ree o negaive pre-exising rends I

or example SNAP aciviy was already rending down in saes ha raised heir

minimum wages beore hese changes came ino effec our regression analysis could

(misakenly) atribue ha reducion o he minimum wage We check or such

pre-rends by inroducing variables ha represen he prior yearrsquos value or leads o

he minimum wage I he model esimaes he minimum wage o have an effec on

he oucome variable beore he wage change wen ino effec hen an unobservedacor no he minimum wage change caused he change in SNAP aciviy

We es he specificaions above or pre-rends by including a one-year lead in

all hree specificaions We find ha he lead erms are small posiive and no

saisically significan indicaing ha he concurren minimum wage991252no

he wage level in prior periods991252is driving he observed changes in SNAP

oucomes28 In paricular he coefficien (sandard error) on he lead erm in

our preerred amily-level enrollmen regression is 011 and no significan

while he coefficien and sandard error o he conemporaneous minimum

wage is unchanged In he sae-level preerred enrollmen regression he

coefficien o he lead erm is again small (07) and i is no significan Te

corresponding coefficien on he lead erm in he sae-level expendiure

regression is 16 and is no significan Te posiive poin esimaes on hese lead

erms resuls no only rule ou disoring negaive pre-rends Tey also sugges

ha our main resuls may underesimae he rue effecs

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2743

24 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2843

Appendix B | wwwamericanprogresso

Appendix B

Model selection process

For boh he amily-level and sae-level models we es hree mehods o conrol

or unobserved geographic- and ime-varying characerisics as suggesed by he

minimum wage lieraure Firs we include only independen sae-specific fixed

effecs and year-specific fixed effecs Tis specificaion (specificaion 1) implicily

assumes ha amilies in any sae consiue an equally good saisical ldquoconrolrdquo

group or hose in any randomly chosen sae afer accouning or various

characerisics (median income and unemploymen rae among ohers) Similarly

simple ime fixed effecs assume ha amilies surveyed in any year can crediblyserve as a conrol group or amilies surveyed in every oher year o he sample

(1990 hrough 2012)

In oher words specificaion 1 assumes ha a saersquos immediae neighbor provides

no beter a couneracual or he effec o a minimum wage change han does a

sae across he counry We relax his resricive specificaion sequenially in wo

seps In specificaion 2 we replace simple year fixed effecs wih fixed effecs or

each Census divisionyear (capured as an addiional variable in he vecor By

using division-year effecs we remove he resricion ha amilies in each sae

are equally good saisical conrols or all oher amilies Raher we allow or he

possibiliy ha amilies in similar geographic regions (or example he Souh or

he Norheas) may be more similar o one anoher han amilies arher away

Finally in specificaion 3 we add sae-specific linear ime rends o he previous

specificaion Tus specificaion 3 is he mos rigorous model specificaion in ha

i allows or heerogeneiy along hree dimensions Ta is specificaion 3 allows

each sae o have is own ime-varying rends raher han imposing he resricion

ha saes evolve idenically over he 22 years in he sample

We begin building he heoreical specificaion above rom a se o simpleuncondiional models regression o SNAP aciviy (enrollmen or expendiures)

on he log o he minimum wage and a se o geographic- and ime-specific

effecs (specificaions 1 2 and 3 described above) As shown in ables 1ndash3 (or

specificaion 3) we hen add covariaes sequenially o hese models including

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2943

26 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

firs he vecor o amily-level conrols ollowed by each o several sae-level

covariaes in urn he unemploymen rae log o median-amily income and he

employmen-o-populaion raio Comparable resuls or specificaions 2 and 3

will be available in our orhcoming working paper

As expeced he simple uncondiional models indicae ha he relaionship beween he minimum wage and SNAP enrollmen i one exiss is a more complex

one influenced by oher acors In he uncondiional model he coefficien on he

variable o ineres991252he log o he minimum wage991252is small in magniude and no

saisically differen rom zero Once we accoun or he influence o labor marke

condiions and variaion in income levels on program paricipaion (by including

unemploymen rae and median-amily income conrol variables respecively)

he effec o he minimum wage on SNAP enrollmen is precisely esimaed Te

coefficien o he log minimum wage is slighly higher (-0042) in he amily-level

analysis han he coefficien (-031) in he sae-level analysis Te level o precision

is also higher in he amily-level analysis Tis is o be expeced when using 124million observaions compared o 1127

Te second dimension o model choice concerns he effec specificaion ables

7ndash9 compare he primary coefficiens o ineres or he SNAP enrollmen and

expendiure models For boh he enrollmen models he effec sizes are smalles

or specificaion 1 larges or specificaion 2 and inermediae beween hese wo

in specificaion 3 Recall ha Specificaion 3 conains sae-specific linear ime

rends in addiion o he census divisionyear conrols included in specificaion

2 In he amily-level enrollmen model he sandard error o he minimum wage

coefficien is smaller han in he oher wo specificaions Sandard errors on he

oher variables are much smaller in specificaions 2 and 3 han in specificaion

1 On he basis o coefficien significance (join and individual) specificaions 2

and 3 are sricly preerred in boh enrollmen models o specificaion 1 which

conains only sae and year fixed effecs

A concern wih specificaions 2 and 3 is ha rend conrols such as sae linear

rends may incorrecly absorb some o he delayed impac o a minimum wage

When we es his issue by including lagged minimum wages we do no find ha

delayed effecs are significan Anoher concern is ha more sauraed modelsuse less o he saisical variaion which could reduce he saisical power o

he resuls However he sandard errors or our more sauraed models are no

higher and are lower in some cases han or he less sauraed models Overall

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3043

Appendix B | wwwamericanprogresso

his evidence suppors our use o he sauraed model as he preerred model

specificaion Moreover Dubersquos 2013 sudy shows ha more sauraed models

perorm beter han models wih jus sae and ime fixed effecs

Te esimaed enrollmen regressions a boh he amily and sae levels show large

and saisically significan coefficiens Te esimaed minimum wage effec in heexpendiures regressions991252or which we have only sae-level daa991252is also large

and saisically significan

We do no use weighed regression or he sae-level models preerring o keep

analysis o he ldquoreamenrdquo (ha is o say a minimum wage change) appropriae

o he average sae raher han he average amily or individual I insead our

primary ineres were he impac o a minimum wage change on he average amily

or he average individual we migh choose o designae he number o amilies

in each sae or he sae populaion respecively as analyic weighs in order o

obain a coefficien beter suied or such inerence

TABLE 4

SNAP enrollment

Family-level linear probability

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0048 -0047 -0040 -0043 -0042

(0013) (0013) (001) (0008) (0009)

Unemployment rate (100) 0505 0420 0280

(0083) (0086) (0082)

Log median income -0057 -0039

(0011) (0011)

Employment-to-population ratio -0239

(0038)

N 1242022 1242022 1242022 1242022 1242022

plt01 plt005 plt001

Note Robust standard errors are in parentheses Observations are clustered at the state level The outcome variable is binary and equal to one if a family is enrolledin SNAP All models include state fixed effects Census division x-year fixed effects and state-specific linear time trends All specifications except 3a include additional

controls for family size race and marital status of the family head presence of children and presence of an adult maleSource Annual data from the March Current Population Survey for 1990 to 2012 estimation includes Current Population Survey probability weights

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3143

28 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

TABLE 5

SNAP enrollment

State-level linear regression

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0028 -0024 -0026 -0031 -0031

(0014) (0013) (0013) (0012) (0012)

Unemployment rate (100) 0625 0466 0320 0339

(0087) (0088) (0085) (0083)

Log median income -0090 -0065 -0061

(0013) (0013) (0013)

Employment-to-population ratio -0282 -0248

(0037) (0038)

N 1127 1127 1127 1127 1127

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the SNAP enrollment rate All models include state fixed effects Census division x-year fixedeffects and state-specific linear time trends

Source Annual data from the March Current Population Survey for 1990 to 2012

TABLE 6

SNAP expenditures

State-level linear regression

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0181 -0149 -0156 -0153 -0190

(011) (0103) (0102) (0103) (0103)

Unemployment rate (100) 4587 4152 4232 4313

(0622) (0621) (0633) (0628)

Log median income -0246 -0261 -0294

(0075) (0078) (0078)

Employment-to-population ratio 0155 0244

(0237) (024)

N 1127 1127 1127 1127 1127

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the natural log of state SNAP expenditures per capita for 1990 to 2012 All models include state

fixed effects Census division x-year fixed effects and state-specific linear time trends

Source Annual data from Bureau of Economic Analysis national income and product accounts tables for 1990 to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3243

Appendix B | wwwamericanprogresso

TABLE 7

Comparison of specifications SNAP enrollment

Family-level linear probability

(1) (2) (3)

Log minimum wage -0028 -0049 -0042

(0014) (0017) (0009)

Unemployment rate (100) 0275 0297 0280

(0161) (0076) (0082)

Log median income -0077 -0055 -0039

(0014) (0012) (0011)

Employment-to-population ratio -0238 -0250 -0239

(0054) (004) (0038)

N 1242022 1242022 1242022

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses Observations are clustered at the state level The outcomevariable is binary or equal to one if a family is enrolled in SNAP All specifications include additional controls forfamily size race and marital status of the family head presence of children and presence of an adult male

Source Annual data from the March Current Population Survey for 1990 to 2012 estimation includes CurrentPopulation Survey probability weights

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3343

30 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

TABLE 8

Comparison of specifications SNAP enrollment

State-level linear regression

(1) (2) (3)

Log minimum wage -0019 -0035 -0031

(0009) (0012) (0012)

Unemployment rate (100) 0401 0370 0339

(0063) (0077) (0083)

Log median income -0081 -0073 -0061

(0011) (0013) (0013)

Employment-to-population ratio -0183 -0222 -0248

(0039) (0039) (0038)

N 1127 1127 1127

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the SNAP enrollment rate Allregressions include the share of households with children and the racial shares of the population

Source Annual data from the March Current Population Survey for 1990 to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3443

Appendix B | wwwamericanprogresso

TABLE 9

Comparison of specifications SNAP expenditures

State-level linear regression

(1) (2) (3)

Log minimum wage -0121 -0203 -0190

(0075) (0103) (0103)

Unemployment rate (100) 5292 5152 4313

(0464) (0576) (0628)

Log median income -0437 -0417 -0294

(008) (0086) (0078)

Employment-to-population ratio -0040 -0220 0244

(0261) (0260) (0240)

N 1127 1127 1127

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the natural log of state SNAPexpenditures per capita All regressions include the share of households with children and the racial shares ofthe state population

Source Annual data from Bureau of Economic Analysis national income and product accounts tables for 1990to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3543

32 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3643

Appendix C | wwwamericanprogresso

Appendix C Harkin-Miller

policy simulation results

StateMinimum wage

(2014)

SNAP enrollment (2012) Predicted change in SNAP enrollment

RateTotal

(persons)

Model 1 family

enrollment (linearprobability)

Model 2 state

enrollment (linearregression)

Model 3 s

expenditu(linear regre

Alabama $725 164 788682 -66922 -58897 -58906

Alaska $775 120 87436 -8104 -6887 -3288

Arizona $790 201 1319323 -67435 -56738 -64356

Arkansas $725 78 230489 -40977 -36063 -35248

California $800 206 7813680 -371131 -310222 -18223

Colorado $800 164 853155 -50684 -42365 -23926

Connecticut $870 91 326621 -22456 -17975 -13711

Delaware $725 186 170262 -12739 -11211 -10647

District of Columbia $825 133 84009 -5370 -4417 -3632

Florida $793 166 3208026 -195813 -164426 -13046

Georgia $725 160 1586336 -137741 -121224 -11004

Hawaii $725 96 133662 -19310 -16995 -14933

Idaho $725 92 147501 -22165 -19507 -15809

Illinois $825 95 1225084 -109088 -89742 -70955

Indiana $725 125 816233 -90818 -79928 -83985

Iowa $725 155 478011 -42716 -37594 -28556

Kansas $725 135 388269 -40082 -35275 -27461

Kentucky $725 130 568821 -60840 -53544 -52259

Louisiana $725 149 683832 -63929 -56263 -66083

Maine $750 77 101976 -16567 -14323 -15234

Maryland $725 144 846415 -81748 -71946 -38370

Massachusetts $800 130 864721 -64902 -54251 -42913

Michigan $740 146 1439141 -128801 -112140 -11022

Minnesota $725 133 713646 -74730 -65769 -37878

TABLE 10

SNAP enrollments Predicted changes by state under the Harkin-Miller bill

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3743

34 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

StateMinimum wage

(2014)

SNAP enrollment (2012) Predicted change in SNAP enrollment

RateTotal

(persons)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 state

expenditures

(linear regression)

Mississippi $725 129 386501 -41486 -36511 -46467

Missouri $750 172 1036182 -75131 -64952 -56244

Montana $790 132 132452 -10350 -8708 -5846

Nebraska $725 124 230683 -25773 -22683 -12189

Nevada $825 162 446035 -23349 -19209 -11894

New Hampshire $725 127 168404 -18359 -16157 -5735

New Jersey $825 160 1416666 -75175 -61843 -28236

New Mexico $750 149 310896 -25983 -22463 -22512

New York $800 192 3763553 -191193 -159815 -142182

North Carolina $725 174 1697193 -135417 -119179 -113503

North Dakota $725 87 61225 -9743 -8574 -4021

Ohio $795 143 1647345 -115869 -97169 -88580

Oklahoma $725 129 494053 -53006 -46650 -46854

Oregon $910 124 485326 -17036 -13328 -16398

Pennsylvania $725 161 2053643 -177315 -156052 -125586

Rhode Island $800 156 163730 -10258 -8574 -8698

South Carolina $725 94 445277 -65614 -57746 -50304

South Dakota $725 208 173749 -11586 -10197 -7458

Tennessee $725 142 914903 -89667 -78915 -99134

Texas $725 110 2863779 -362018 -318607 -253285

Utah $725 88 251107 -39658 -34902 -19390

Vermont $873 156 97792 -3823 -3055 -2475

Virginia $725 101 829771 -113723 -100086 -58212

Washington $932 72 496934 -23221 -17947 -17756

West Virginia $725 58 107875 -25792 -22699 -21665

Wisconsin $725 75 427822 -79521 -69986 -53210

Wyoming $725 164 94590 -8010 -7050 -3104

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3843

Appendix C | wwwamericanprogresso

StateMinimum wage

(2014)

SNAP expenditures (2012)Predicted change in SNAP expenditures

(millions of dollars)

Per resident

(millions of

dollars)

Total

(millions of

dollars)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 s

expenditu

(linear regre

Alabama $725 $290 $1397 -$1185 -$1043 -$1043

Alaska $775 $253 $185 -$262 -$223 -$106

Arizona $790 $257 $1686 -$935 -$787 -$892

Arkansas $725 $248 $733 -$636 -$560 -$547

California $800 $189 $7164 -$7276 -$6082 -$3573

Colorado $800 $157 $816 -$862 -$721 -$407

Connecticut $870 $191 $686 -$343 -$275 -$210

Delaware $725 $250 $229 -$205 -$180 -$171

District of Columbia $825 $366 $232 -$146 -$120 -$99

Florida $793 $294 $5676 -$4429 -$3719 -$2951

Georgia $725 $317 $3140 -$2936 -$2584 -$2346

Hawaii $725 $335 $465 -$449 -$395 -$347

Idaho $725 $225 $359 -$376 -$331 -$268

Illinois $825 $249 $3200 -$2096 -$1725 -$1364

Indiana $725 $220 $1439 -$1162 -$1023 -$1075

Iowa $725 $192 $589 -$658 -$579 -$440

Kansas $725 $159 $460 -$502 -$441 -$344

Kentucky $725 $298 $1303 -$1133 -$997 -$973

Louisiana $725 $315 $1450 -$1047 -$922 -$1083

Maine $750 $281 $373 -$267 -$231 -$246

Maryland $725 $188 $1109 -$1765 -$1553 -$828

Massachusetts $800 $206 $1366 -$1030 -$861 -$681

Michigan $740 $300 $2963 -$2400 -$2090 -$2054

Minnesota $725 $140 $755 -$1113 -$980 -$564

Mississippi $725 $326 $973 -$649 -$571 -$726

Missouri $750 $241 $1452 -$1278 -$1104 -$956

Montana $790 $190 $191 -$179 -$151 -$101

Nebraska $725 $140 $259 -$409 -$360 -$194

Nevada $825 $191 $527 -$441 -$363 -$225

New Hampshire $725 $126 $167 -$399 -$351 -$125

TABLE 11

SNAP expenditures Predicted changes by state under the Harkin-Miller bill

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3943

36 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

StateMinimum wage

(2014)

SNAP expenditures (2012)Predicted change in SNAP expenditures

(millions of dollars)

Per resident

(millions of

dollars)

Total

(millions of

dollars)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 state

expenditures

(linear regression)

New Jersey $825 $160 $1420 -$1611 -$1325 -$605

New Mexico $750 $324 $675 -$513 -$444 -$445New York $800 $287 $5616 -$3766 -$3148 -$2801

North Carolina $725 $252 $2454 -$2187 -$1925 -$1833

North Dakota $725 $128 $90 -$162 -$143 -$67

Ohio $795 $259 $2995 -$2013 -$1688 -$1539

Oklahoma $725 $248 $945 -$799 -$703 -$706

Oregon $910 $322 $1255 -$272 -$213 -$262

Pennsylvania $725 $218 $2779 -$2930 -$2579 -$2075

Rhode Island $800 $280 $294 -$173 -$144 -$147

South Carolina $725 $291 $1373 -$1337 -$1177 -$1025South Dakota $725 $198 $165 -$192 -$169 -$123

Tennessee $725 $324 $2091 -$1413 -$1243 -$1562

Texas $725 $230 $5997 -$6402 -$5634 -$4479

Utah $725 $141 $402 -$614 -$541 -$300

Vermont $873 $230 $144 -$66 -$53 -$43

Virginia $725 $173 $1413 -$2062 -$1815 -$1056

Washington $932 $244 $1682 -$350 -$270 -$267

West Virginia $725 $273 $508 -$451 -$397 -$379

Wisconsin $725 $204 $1166 -$1302 -$1146 -$871Wyoming $725 $95 $55 -$105 -$93 -$41

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4043

Appendix C | wwwamericanprogresso

FIGURE 2

Raising the minimum wage to $1010would cut taxpayer costs in every state

Predicted decreases in cost and enrollment

in SNAP in 50 states

$200+$51ndash$100

$101ndash$200

0ndash$25

$26ndash$50

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4143

38 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

Endnotes

1 US Department of Agriculture ldquoSupplementalNutrition Assistance Program (SNAP)rdquo available athttpwwwfnsusdagovsnapeligibility (last accessedFebruary 2014)

2 For this initial analysis we do not consider Harkin-Millerrsquos increase in subminimum wages for tippedworkers To do s o would increase the estimated SNAP

savings by an unknown amount

3 The Congressional Budget Office estimates thatworkers currently earning between $1010 and $1150per hour would see their wages rise under the Harkin-Miller proposal Congressional Budget O ffice ldquoTheEffects of a Minimum Wage Increase on Employmentand Family Incomerdquo (2014)

4 Marianne Page Joanne Spetz and Jane Millar ldquoDoesthe Minimum Wage Affect Welfare Caseloadsrdquo Journalof Policy Analysis and Management 24 (2) (2005)273ndash295

5 Marianne Bitler and Hilary Hoynes ldquoThe More ThingsChange the More They Stay the Same The SafetyNet Living Arrangements and Poverty in the GreatRecessionrdquo NBER Working Paper 194 49 2013

6 Sylvia Allegretto and others ldquoFast Food Poverty Wages The Public Cost of Low-Wage Jobs in the Fast-FoodIndustryrdquo (Berkeley California Center for LaborResearch and Education 2013) available at httplaborcenterberkeleyedupubliccostsfast_food_poverty_wage

7 David Neumark and William Wascher ldquoDoes a HigherMinimum Wage Enhance the Effectiveness of theEarned Income Tax Creditrdquo Industrial and LaborRelations Review 64 (4) (2011) 712ndash746

8 David Lee and Emmanuel Saez ldquoOptimal MinimumWage Policy in Competitive Labor Marketsrdquo Journal ofPublic Economics 96 (9) (2012) 739ndash749

9 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family IncomerdquoFebruary 2014

10 Jesse Rothstein ldquoIs the EITC as Good as an NITConditional Cash Transfers and Tax Incidencerdquo AmericanEconomic Journal Economic Policy 2 (1) (2010) 177ndash208

11 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family Incomerdquo

12 Dube Arindrajit 2013 rdquoMinimum Wagesand the Distribution of Family IncomerdquoUnpublished working paper Available at httpsdldropboxusercontentcomu15038936Dube_MinimumWagesFamilyIncomespdf

13 As Dube explains in ldquoThe poverty of Minimum WageFactsrdquo the simulation approach underestimate stemsfrom a number of unwarranted assumptions includingthe range of actual wage increases and the accuracy ofwage data in the Current Population Survey The causal

approach does not make these assumptions

14 Allegretto Sylvia and others 2013 ldquoCredible ResearchDesigns for Minimum Wage Studiesrdquo Working Paper148-13 University of California Berkeley Institutefor Research on Labor and Employment Available athttpwwwirleberkeleyeduworkingpapers148-13pdf

15 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family In comerdquo

16 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo (2012)

17 Ibid

18 Ibid

19 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo

20 US Department of Agriculture ldquoSupplementalNutrition Assistance Program (SNAP)rdquo

21 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo

22 Ibid

23 We deviate from the Census Bureaursquos definition ofa family unit which is ldquotwo people or more (on e ofwhom is the householder) related by bir th marriageor adoption and residing togetherrdquo US Bureau ofthe Census ldquoCurrent Population Survey DefinitionsFamilyrdquo available at httpwwwcensusgovcpsabout

cpsdefhtml (last accessed February 2014) We countas a family unit any individual residing on his or herown two or more persons residing together whodo not belong to a family in the March CPS sampleare constructed as one family in our analysis For thepurposes of food stamp allocations the consumptionresulting from this transfer is probably distributed tofamily members (rather than household members ora single individual within the household) Howeversingle individuals canmdashand domdashreceive SNAPbenefits Excluding them would fail to make theanalysis reflective of the population at large

24 Strictly the family level linear probability modelpredicts the percentage-point decrease in theprobability that an individual family will receive SNAPpayments When applied to a large number of familieshowever we are able to interpret the coefficient asa decrease in the mean of enrollmentmdashthat is a

decrease in the enrollment ratemdashby applying the lawof iterated expectations

25 We generate expenditure predictions from theenrollment modelsmdashand conversely generateenrollment predictions from the expenditure modelmdashby assuming that expenditures per enrolled familyremains the same before and after the minimum wagechange In practice this is likely to be a conservativeestimatemdashthat is to underestimate the decrease inSNAP activity Average SNAP benefits per family willalso decrease as many families that remain eligible forSNAP experience income gains

26 Wage and Hour Division ldquoMinimum Wage Laws inthe States ndash Januar y 1 2014rdquo available at httpwwwdolgovwhdminwageamericahtm (last accessedFebruary 2014)

27 See for example Sylvia Allegretto and others ldquoCredibleResearch Designs for Minimum Wage Studiesrdquo WorkingPaper 148-113 (Berkeley California Institute forResearch on Labor and Employment 2013) available athttpwwwirleberkeleyeduworkingpapers148-13pdf

28 We will report these results in a forthcoming workingpaper

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4243

Endnotes | wwwamericanprogresso

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4343

The Center for American Progress is a nonpartisan research and educational institute

dedicated to promoting a strong just and free America that ensures opportunity

for all We believe that Americans are bound together by a common commitment to

these values and we aspire to ensure that our national policies reflect these values

We work to find progressive and pragmatic solutions to significant domestic and

international problems and develop policy proposals that foster a government that

is ldquoof the people by the people and for the peoplerdquo

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1743

14 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1843

Results | wwwamericanprogresso

Results

Estimated minimum wage effects on SNAP enrollment and

expenditures

able 1 shows he esimaed parameer o ineres991252he coefficien o he

minimum wage or he preerred model o each ype Coefficiens on he

minimum wage variable are no direcly comparable across models because all

our models have a differen uncional orm o undersand and compare hese

esimaes we compue he change in SNAP aciviy prediced or a paricular wage scenario Te final column in able 1 answers he quesion Wha would be

he expeced change in SNAP aciviy in response o a 10 percen increase in he

minimum wage Te answer o his quesion varies wih he value o he inpu

parameers in he able we calculae he percenage decrease in enrollmen or

expendiures prediced or he average sae wih a minimum wage o $725 in

2014 Te sae-level SNAP expendiure model which is a consan-elasiciy

model conveys elasiciy inormaion direcly or he change in expendiures per

capia in he sae

TABLE 1

Comparison of national SNAP predictions for a 10 percent increase in the federal minimum wage

Model LevelRegression

type

Predicted outcome Coefficient of log

(minimum wage)

(Standard error)

Effect of a 10 percent in

in the minimum wage

VariableForm of

variable

Total

enrollment

Tot

expend

1 Family Linear

probability Enrollment Binary (enrolled=1)

-0042

(0008) -317 N

2

StateLinear regression

(ordinary least

squares)

Enrollment State enrollment rate (percent) -0031

(0012) -235 N

3 Expenditures Log (state expenditures per capita) -0190

(0103) NA -19

plt01 plt005 plt001Note Predicted changes are calculated for the average state with a minimum wage of $725 in 2014

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1943

16 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

According o his model a 10 percen increase in he minimum wage would resul

in a 19 percen reducion in SNAP expendiures Tis resul is reassuringly similar

o Dubersquos resul or povery reducion Te wo enrollmen models are much more

precisely esimaed han he expendiure model Te sae-level enrollmen model

finds ha a 10 percen minimum wage increase in a low-wage sae is associaed wih

a 235 percen decrease in SNAP enrollmens Te amily-level linear probabiliymodel predics a somewha greaer elasiciy or low minimum wage saes an

increase o 10 percen in he ederal minimum would resul in a 317 percen decline

in SNAP enrollmen24 Te differences in hese esimaes sem rom a number o

acors including difference in model uncional orm and daa used We rea his

range o elasiciy esimaes as an upper and lower bound on enrollmen impacs

Harkin-Miller bill National and state-level predicted impacts

Wha would be he prediced change or he SNAP program i he ederalminimum were raised o $1010 as proposed in he Harkin-Miller bill In order o

make his inerence we accoun or he ac ha no all saes are currenly subjec

o he ederal minimum wage a he beginning o 2014 21 saes mainained

higher minimum wages han $725 In hose saes an increase in he ederal

minimum wage may or may no be binding or employers in he sae depending

upon wheher he new ederal minimum exceeds he sae-level minimum Bu

regardless o wheher a minimum wage change is binding he impac on SNAP

aciviy will be lower in high minimum wage saes In order o accoun or his

properly we calculae sae by sae he percenage wage change ha would resul

rom he Harkin-Miller proposal and apply he parameers rom each o he hree

models above o compue he expeced decrease in SNAP aciviy or each sae

In his exercise we use saesrsquo curren (2014) minimum wage levels and assume

as a baseline he 2012 levels o SNAP enrollmen and expendiure as 2012 is he

mos recen year or which SNAP daa are available

able 10 and able 11 in Appendix C repor he esimaed effecs on SNAP

enrollmen and expendiures respecively or each sae under he Harkin-Miller

bill25 An increase o $1010 i enaced oday would represen beween a 393

percen wage increase in a $725 minimum wage sae and an 84 percen increasein Washingon sae which has he highes minimum wage in he naion a $932

as o January 201426 Slighly more han 56 percen o he decrease in expendiures

and abou 59 percen o he decrease in enrollmen would occur in saes wih

presen-day minimum wages o $725

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2043

Results | wwwamericanprogresso

In 2012 hese saes were home o 46 percen o he American populaion

and accouned or a roughly equivalen percenage o oal naional SNAP

expendiures No surprisingly he larges enrollmen decreases come rom

populous saes wih very high SNAP enrollmen raes andor very low minimum

wages Te larges prediced enrollmen reducion991252beween 319000 individuals

and 362000 individuals991252would occur in exas which has a $725 minimum wage In Caliornia which has a 206 percen SNAP paricipaion rae and an $8

per hour minimum wage we anicipae SNAP enrollmen declines o 310000

persons o 371000 persons And in Florida which had a SNAP paricipaion rae

o 166 percen and a $793 minimum wage enrollmen could decline by beween

164000 individuals and 196000 individuals For he our saes ha ogeher

accouned or he greaes amoun o SNAP spending in 2012991252exas Caliornia

Florida and New York respecively991252he combined expendiure reducion rom

he Harkin-Miller bill is prediced o be $14 billion

able 2 summarizes he prediced declines in SNAP aciviy or he naion as a whole ha would resul rom he direc and indirec effecs o he Harkin-Miller

bill Enrollmen would all beween 31 million persons and 36 million persons

represening 75 percen o 87 percen o curren enrollmen Te anicipaed

reducion in program expendiures would be nearly $46 billion or 61 percen o

program expendiures

TABLE 2

Comparison of national SNAP predictions under the Harkin-Miller billrsquos $1010 minimum wage

Model

Enrollment

(persons)

Expenditures

(millions of dollars)

Current (2012) Predicted Change Current (2012) Predicted Chan

Family enrollment (linear probability)

41866195

45489339 -3623144

$74861

NA NA

State enrollment (ordinary least squares) 38745435 -3120759 NA NA

State expenditures (ordinary least squares) NA NA $70305 -$45

Note Calculations use 2014 state minimum wages and the most recent SNAP data from 2012 They assume that per-enrollee expenditures remain constant

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2143

18 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

Tere are o course oher possibiliies or minimum wage legislaion able 3

shows he expeced SNAP changes or he Unied Saes under a variey o wage

scenarios calculaed using he sae-level models I saes were no able o se

heir minimum wages independenly such ha all saes were consrained by

he ederal minimum o $725 SNAP would be received by abou 514000 more

people across he Unied Saes a an addiional program cos o nearly hree-quarers o a billion dollars In conras he effecs o a higher minimum wage

proposal991252a ederal wage floor o $11 per hour991252would decrease enrollmen in

SNAP by more han 10 percen and decrease program coss by 83 percen

TABLE 3

Summary of par ticipation and expenditures under wage scenarios

If all states had

minimum wages of

Enrollment(persons)

Expenditures

(millions of dollars)

Predicted Change Predicted Change

Recent levels (2014) 41866195 $74861

$725 42380520 514326 $75604 $743

$800 41423919 -442276 $74209 -$652

$900 40148451 -1717744 $72350 -$2511

$1000 38872982 -2993212 $70490 -$4371

$1010 38745435 -3120759 $70305 -$4556

$1100 37597514 -4268681 $68631 -$6230

Note Calculations use state-level enrollment model coefficient

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2243

Conclusion | wwwamericanprogresso

Conclusion

An exensive body o lieraure examines employmen effecs o he minimum

wage A much smaller se o sudies examines how he minimum wage affecs

povery and only a handul o papers examine he effecs o he minimum wage

on he EIC Our analysis is he firs o examine he effecs o he minimum

wage on SNAP

Our findings indicae ha increased earnings rom minimum wage changes do

reduce SNAP enrollmens and expendiures We esimae ha he Harkin-Miller bill would save axpayers nearly $46 billion per year equivalen o 61 percen

o SNAP expendiures in 2012 he las year or which daa are available Over a

10-year period he esimaed savings amoun o nearly $46 billion

Our repor is subjec o limiaions ha we expec o overcome in our uure

research Firs he findings do no ake ino accoun possible ineracions among

SNAP he EIC and Medicaid Te eligibiliy cuoffs among hese programs

are quie differen suggesing ha such ineracions may be minor Noneheless

he join effecs can only be deermined by urher research using a causal

model Second i would be useul o know he disribuion o SNAP reducions

along he wage disribuion Using he Congressional Budge Officersquos calculaions

o how much he oal dollar value o wage would increase under he Harkin-

Miller proposal our findings imply ha he decline in overall SNAP spending

equals abou 15 percen o he oal resuling increase in wages Te amoun and

disribuion o his offse are o considerable ineres Minimum wage beneficiaries

who come rom working amilies already well above he povery line would no

see any offse while hose who are currenly considerably below he povery line

will see larger offses Tese issues will also be a subjec or our uure research

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2343

20 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

About the authors

Rachel West is a maser o public policy candidae a he Goldman School

o Public Policy Universiy o Caliornia Berkeley Her research ocuses on

economic policy in he areas o low-wage labor and povery

Michael Reich is proessor o economics and direcor o he Insiue or

Research on Labor and Employmen a he Universiy o Caliornia a Berkeley

His research publicaions cover numerous areas o labor economics including

racial inequaliy labor marke segmenaion high-perormance workplaces

union-managemen cooperaion Japanese labor-managemen sysems living

wages and minimum wages He received his docorae in economics rom

Harvard Universiy

Acknowledgments

We are graeul o Sylvia Allegreto Arindraji Dube Bill Leser Jesse Rohsein

Daniel Tompson and Ben Zipperer or heir valuable suggesions

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2443

References | wwwamericanprogresso

References

Allegreto Sylvia and ohers 2013 ldquoFas Food Povery Wages Te Public Cos o Low-Wage Jobsin he Fas-Food Indusryrdquo Universiy o Caliornia Berkeley Cener or Labor Research andEducaion Available a htplaborcenerberkeleyedupubliccossas_ood_povery_wage

Allegreto Sylvia and ohers 2013 ldquoCredible Research Designs or Minimum Wage Sudiesrdquo

Working Paper 148-13 Universiy o Caliornia Berkeley Insiue or Research on Labor andEmploymen Available a htpwwwirleberkeleyeduworkingpapers148-13pd

Biler Marianne and Hilary Hoynes 2013 ldquo Te More Tings Change he More Tey Say heSame Te Saey Ne Living Arrangemens and Povery in he Grea Recessionrdquo Working Paper19449 Naional Bureau o Economic Research

Congressional Budge Office 2012 ldquoTe Supplemenal Nuriion Assisance Programrdquo Washingon Available a htpwwwcbogovsiesdeaulfilescbofilesatachmens04-19-SNAPpd

991252 991252 991252 2014 ldquoTe Effec o a Minimum-Wage Increase on Employmen and Family Incomerdquo Washingon Available a htpwwwcbogovsiesdeaulfilescbofilesatachmens44995-MinimumWagepd

Dube Arindraji 2013 rdquoMinimum Wages and he Disribuion o Family Incomerdquo Unpublished working paper Available a htpsdldropboxuserconencomu15038936Dube_ MinimumWagesFamilyIncomespd

991252 991252 991252 2014 ldquoTe Povery o Minimum Wage Facsrdquo Arindraji Dube Blog January 22 Availablea htparindubecom20140122he-povery-o-minimum-wage-acs

Lee David and Emmanuel Saez 2012 ldquoOpimal Minimum Wage Policy in Compeiive LaborMarkesrdquo Journal o Public Economics 96 (9) 739ndash749

Neumark David and William Wascher 1992 ldquoEmploymen Effecs o Minimum and Subminimum Wages Panel Daa on Sae Minimum Wage Lawsrdquo Industrial and Labor Relations Review 46 (1)

55ndash81

Neumark David and William Wascher 2011 ldquoDoes a Higher Minimum Wage Enhance heEffeciveness o he Earned Income ax Credirdquo Industrial and Labor Relations Review 64 (4)712ndash746

Page Marianne Joanne Spez and Jane Millar 2005 ldquoDoes he Minimum Wage Affec WelareCaseloadsrdquo Journal o Policy Analysis and Management 24 (2) 273ndash295

Rohsein Jesse 2010 ldquoIs he EIC as Good as an NI Condiional Cash ransers and ax

Incidencerdquo American Economic Journal Economic Policy 2 (1) 177ndash208

Wage and Hours Division 2014 ldquoMinimum Wage Laws in he Saes ndash January 1 2014rdquo USDeparmen o Labor (htpwwwdolgovwhdminwageamericahm [February 2014])

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2543

22 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2643

Appendix A | wwwamericanprogresso

Appendix A

Pre-trend falsification check

Recen minimum wage research27 highlighs a common flaw in previous sudies

ailure o veriy ha he oucome variable is ree o negaive pre-exising rends I

or example SNAP aciviy was already rending down in saes ha raised heir

minimum wages beore hese changes came ino effec our regression analysis could

(misakenly) atribue ha reducion o he minimum wage We check or such

pre-rends by inroducing variables ha represen he prior yearrsquos value or leads o

he minimum wage I he model esimaes he minimum wage o have an effec on

he oucome variable beore he wage change wen ino effec hen an unobservedacor no he minimum wage change caused he change in SNAP aciviy

We es he specificaions above or pre-rends by including a one-year lead in

all hree specificaions We find ha he lead erms are small posiive and no

saisically significan indicaing ha he concurren minimum wage991252no

he wage level in prior periods991252is driving he observed changes in SNAP

oucomes28 In paricular he coefficien (sandard error) on he lead erm in

our preerred amily-level enrollmen regression is 011 and no significan

while he coefficien and sandard error o he conemporaneous minimum

wage is unchanged In he sae-level preerred enrollmen regression he

coefficien o he lead erm is again small (07) and i is no significan Te

corresponding coefficien on he lead erm in he sae-level expendiure

regression is 16 and is no significan Te posiive poin esimaes on hese lead

erms resuls no only rule ou disoring negaive pre-rends Tey also sugges

ha our main resuls may underesimae he rue effecs

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2743

24 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2843

Appendix B | wwwamericanprogresso

Appendix B

Model selection process

For boh he amily-level and sae-level models we es hree mehods o conrol

or unobserved geographic- and ime-varying characerisics as suggesed by he

minimum wage lieraure Firs we include only independen sae-specific fixed

effecs and year-specific fixed effecs Tis specificaion (specificaion 1) implicily

assumes ha amilies in any sae consiue an equally good saisical ldquoconrolrdquo

group or hose in any randomly chosen sae afer accouning or various

characerisics (median income and unemploymen rae among ohers) Similarly

simple ime fixed effecs assume ha amilies surveyed in any year can crediblyserve as a conrol group or amilies surveyed in every oher year o he sample

(1990 hrough 2012)

In oher words specificaion 1 assumes ha a saersquos immediae neighbor provides

no beter a couneracual or he effec o a minimum wage change han does a

sae across he counry We relax his resricive specificaion sequenially in wo

seps In specificaion 2 we replace simple year fixed effecs wih fixed effecs or

each Census divisionyear (capured as an addiional variable in he vecor By

using division-year effecs we remove he resricion ha amilies in each sae

are equally good saisical conrols or all oher amilies Raher we allow or he

possibiliy ha amilies in similar geographic regions (or example he Souh or

he Norheas) may be more similar o one anoher han amilies arher away

Finally in specificaion 3 we add sae-specific linear ime rends o he previous

specificaion Tus specificaion 3 is he mos rigorous model specificaion in ha

i allows or heerogeneiy along hree dimensions Ta is specificaion 3 allows

each sae o have is own ime-varying rends raher han imposing he resricion

ha saes evolve idenically over he 22 years in he sample

We begin building he heoreical specificaion above rom a se o simpleuncondiional models regression o SNAP aciviy (enrollmen or expendiures)

on he log o he minimum wage and a se o geographic- and ime-specific

effecs (specificaions 1 2 and 3 described above) As shown in ables 1ndash3 (or

specificaion 3) we hen add covariaes sequenially o hese models including

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2943

26 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

firs he vecor o amily-level conrols ollowed by each o several sae-level

covariaes in urn he unemploymen rae log o median-amily income and he

employmen-o-populaion raio Comparable resuls or specificaions 2 and 3

will be available in our orhcoming working paper

As expeced he simple uncondiional models indicae ha he relaionship beween he minimum wage and SNAP enrollmen i one exiss is a more complex

one influenced by oher acors In he uncondiional model he coefficien on he

variable o ineres991252he log o he minimum wage991252is small in magniude and no

saisically differen rom zero Once we accoun or he influence o labor marke

condiions and variaion in income levels on program paricipaion (by including

unemploymen rae and median-amily income conrol variables respecively)

he effec o he minimum wage on SNAP enrollmen is precisely esimaed Te

coefficien o he log minimum wage is slighly higher (-0042) in he amily-level

analysis han he coefficien (-031) in he sae-level analysis Te level o precision

is also higher in he amily-level analysis Tis is o be expeced when using 124million observaions compared o 1127

Te second dimension o model choice concerns he effec specificaion ables

7ndash9 compare he primary coefficiens o ineres or he SNAP enrollmen and

expendiure models For boh he enrollmen models he effec sizes are smalles

or specificaion 1 larges or specificaion 2 and inermediae beween hese wo

in specificaion 3 Recall ha Specificaion 3 conains sae-specific linear ime

rends in addiion o he census divisionyear conrols included in specificaion

2 In he amily-level enrollmen model he sandard error o he minimum wage

coefficien is smaller han in he oher wo specificaions Sandard errors on he

oher variables are much smaller in specificaions 2 and 3 han in specificaion

1 On he basis o coefficien significance (join and individual) specificaions 2

and 3 are sricly preerred in boh enrollmen models o specificaion 1 which

conains only sae and year fixed effecs

A concern wih specificaions 2 and 3 is ha rend conrols such as sae linear

rends may incorrecly absorb some o he delayed impac o a minimum wage

When we es his issue by including lagged minimum wages we do no find ha

delayed effecs are significan Anoher concern is ha more sauraed modelsuse less o he saisical variaion which could reduce he saisical power o

he resuls However he sandard errors or our more sauraed models are no

higher and are lower in some cases han or he less sauraed models Overall

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3043

Appendix B | wwwamericanprogresso

his evidence suppors our use o he sauraed model as he preerred model

specificaion Moreover Dubersquos 2013 sudy shows ha more sauraed models

perorm beter han models wih jus sae and ime fixed effecs

Te esimaed enrollmen regressions a boh he amily and sae levels show large

and saisically significan coefficiens Te esimaed minimum wage effec in heexpendiures regressions991252or which we have only sae-level daa991252is also large

and saisically significan

We do no use weighed regression or he sae-level models preerring o keep

analysis o he ldquoreamenrdquo (ha is o say a minimum wage change) appropriae

o he average sae raher han he average amily or individual I insead our

primary ineres were he impac o a minimum wage change on he average amily

or he average individual we migh choose o designae he number o amilies

in each sae or he sae populaion respecively as analyic weighs in order o

obain a coefficien beter suied or such inerence

TABLE 4

SNAP enrollment

Family-level linear probability

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0048 -0047 -0040 -0043 -0042

(0013) (0013) (001) (0008) (0009)

Unemployment rate (100) 0505 0420 0280

(0083) (0086) (0082)

Log median income -0057 -0039

(0011) (0011)

Employment-to-population ratio -0239

(0038)

N 1242022 1242022 1242022 1242022 1242022

plt01 plt005 plt001

Note Robust standard errors are in parentheses Observations are clustered at the state level The outcome variable is binary and equal to one if a family is enrolledin SNAP All models include state fixed effects Census division x-year fixed effects and state-specific linear time trends All specifications except 3a include additional

controls for family size race and marital status of the family head presence of children and presence of an adult maleSource Annual data from the March Current Population Survey for 1990 to 2012 estimation includes Current Population Survey probability weights

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3143

28 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

TABLE 5

SNAP enrollment

State-level linear regression

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0028 -0024 -0026 -0031 -0031

(0014) (0013) (0013) (0012) (0012)

Unemployment rate (100) 0625 0466 0320 0339

(0087) (0088) (0085) (0083)

Log median income -0090 -0065 -0061

(0013) (0013) (0013)

Employment-to-population ratio -0282 -0248

(0037) (0038)

N 1127 1127 1127 1127 1127

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the SNAP enrollment rate All models include state fixed effects Census division x-year fixedeffects and state-specific linear time trends

Source Annual data from the March Current Population Survey for 1990 to 2012

TABLE 6

SNAP expenditures

State-level linear regression

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0181 -0149 -0156 -0153 -0190

(011) (0103) (0102) (0103) (0103)

Unemployment rate (100) 4587 4152 4232 4313

(0622) (0621) (0633) (0628)

Log median income -0246 -0261 -0294

(0075) (0078) (0078)

Employment-to-population ratio 0155 0244

(0237) (024)

N 1127 1127 1127 1127 1127

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the natural log of state SNAP expenditures per capita for 1990 to 2012 All models include state

fixed effects Census division x-year fixed effects and state-specific linear time trends

Source Annual data from Bureau of Economic Analysis national income and product accounts tables for 1990 to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3243

Appendix B | wwwamericanprogresso

TABLE 7

Comparison of specifications SNAP enrollment

Family-level linear probability

(1) (2) (3)

Log minimum wage -0028 -0049 -0042

(0014) (0017) (0009)

Unemployment rate (100) 0275 0297 0280

(0161) (0076) (0082)

Log median income -0077 -0055 -0039

(0014) (0012) (0011)

Employment-to-population ratio -0238 -0250 -0239

(0054) (004) (0038)

N 1242022 1242022 1242022

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses Observations are clustered at the state level The outcomevariable is binary or equal to one if a family is enrolled in SNAP All specifications include additional controls forfamily size race and marital status of the family head presence of children and presence of an adult male

Source Annual data from the March Current Population Survey for 1990 to 2012 estimation includes CurrentPopulation Survey probability weights

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3343

30 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

TABLE 8

Comparison of specifications SNAP enrollment

State-level linear regression

(1) (2) (3)

Log minimum wage -0019 -0035 -0031

(0009) (0012) (0012)

Unemployment rate (100) 0401 0370 0339

(0063) (0077) (0083)

Log median income -0081 -0073 -0061

(0011) (0013) (0013)

Employment-to-population ratio -0183 -0222 -0248

(0039) (0039) (0038)

N 1127 1127 1127

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the SNAP enrollment rate Allregressions include the share of households with children and the racial shares of the population

Source Annual data from the March Current Population Survey for 1990 to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3443

Appendix B | wwwamericanprogresso

TABLE 9

Comparison of specifications SNAP expenditures

State-level linear regression

(1) (2) (3)

Log minimum wage -0121 -0203 -0190

(0075) (0103) (0103)

Unemployment rate (100) 5292 5152 4313

(0464) (0576) (0628)

Log median income -0437 -0417 -0294

(008) (0086) (0078)

Employment-to-population ratio -0040 -0220 0244

(0261) (0260) (0240)

N 1127 1127 1127

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the natural log of state SNAPexpenditures per capita All regressions include the share of households with children and the racial shares ofthe state population

Source Annual data from Bureau of Economic Analysis national income and product accounts tables for 1990to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3543

32 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3643

Appendix C | wwwamericanprogresso

Appendix C Harkin-Miller

policy simulation results

StateMinimum wage

(2014)

SNAP enrollment (2012) Predicted change in SNAP enrollment

RateTotal

(persons)

Model 1 family

enrollment (linearprobability)

Model 2 state

enrollment (linearregression)

Model 3 s

expenditu(linear regre

Alabama $725 164 788682 -66922 -58897 -58906

Alaska $775 120 87436 -8104 -6887 -3288

Arizona $790 201 1319323 -67435 -56738 -64356

Arkansas $725 78 230489 -40977 -36063 -35248

California $800 206 7813680 -371131 -310222 -18223

Colorado $800 164 853155 -50684 -42365 -23926

Connecticut $870 91 326621 -22456 -17975 -13711

Delaware $725 186 170262 -12739 -11211 -10647

District of Columbia $825 133 84009 -5370 -4417 -3632

Florida $793 166 3208026 -195813 -164426 -13046

Georgia $725 160 1586336 -137741 -121224 -11004

Hawaii $725 96 133662 -19310 -16995 -14933

Idaho $725 92 147501 -22165 -19507 -15809

Illinois $825 95 1225084 -109088 -89742 -70955

Indiana $725 125 816233 -90818 -79928 -83985

Iowa $725 155 478011 -42716 -37594 -28556

Kansas $725 135 388269 -40082 -35275 -27461

Kentucky $725 130 568821 -60840 -53544 -52259

Louisiana $725 149 683832 -63929 -56263 -66083

Maine $750 77 101976 -16567 -14323 -15234

Maryland $725 144 846415 -81748 -71946 -38370

Massachusetts $800 130 864721 -64902 -54251 -42913

Michigan $740 146 1439141 -128801 -112140 -11022

Minnesota $725 133 713646 -74730 -65769 -37878

TABLE 10

SNAP enrollments Predicted changes by state under the Harkin-Miller bill

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3743

34 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

StateMinimum wage

(2014)

SNAP enrollment (2012) Predicted change in SNAP enrollment

RateTotal

(persons)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 state

expenditures

(linear regression)

Mississippi $725 129 386501 -41486 -36511 -46467

Missouri $750 172 1036182 -75131 -64952 -56244

Montana $790 132 132452 -10350 -8708 -5846

Nebraska $725 124 230683 -25773 -22683 -12189

Nevada $825 162 446035 -23349 -19209 -11894

New Hampshire $725 127 168404 -18359 -16157 -5735

New Jersey $825 160 1416666 -75175 -61843 -28236

New Mexico $750 149 310896 -25983 -22463 -22512

New York $800 192 3763553 -191193 -159815 -142182

North Carolina $725 174 1697193 -135417 -119179 -113503

North Dakota $725 87 61225 -9743 -8574 -4021

Ohio $795 143 1647345 -115869 -97169 -88580

Oklahoma $725 129 494053 -53006 -46650 -46854

Oregon $910 124 485326 -17036 -13328 -16398

Pennsylvania $725 161 2053643 -177315 -156052 -125586

Rhode Island $800 156 163730 -10258 -8574 -8698

South Carolina $725 94 445277 -65614 -57746 -50304

South Dakota $725 208 173749 -11586 -10197 -7458

Tennessee $725 142 914903 -89667 -78915 -99134

Texas $725 110 2863779 -362018 -318607 -253285

Utah $725 88 251107 -39658 -34902 -19390

Vermont $873 156 97792 -3823 -3055 -2475

Virginia $725 101 829771 -113723 -100086 -58212

Washington $932 72 496934 -23221 -17947 -17756

West Virginia $725 58 107875 -25792 -22699 -21665

Wisconsin $725 75 427822 -79521 -69986 -53210

Wyoming $725 164 94590 -8010 -7050 -3104

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3843

Appendix C | wwwamericanprogresso

StateMinimum wage

(2014)

SNAP expenditures (2012)Predicted change in SNAP expenditures

(millions of dollars)

Per resident

(millions of

dollars)

Total

(millions of

dollars)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 s

expenditu

(linear regre

Alabama $725 $290 $1397 -$1185 -$1043 -$1043

Alaska $775 $253 $185 -$262 -$223 -$106

Arizona $790 $257 $1686 -$935 -$787 -$892

Arkansas $725 $248 $733 -$636 -$560 -$547

California $800 $189 $7164 -$7276 -$6082 -$3573

Colorado $800 $157 $816 -$862 -$721 -$407

Connecticut $870 $191 $686 -$343 -$275 -$210

Delaware $725 $250 $229 -$205 -$180 -$171

District of Columbia $825 $366 $232 -$146 -$120 -$99

Florida $793 $294 $5676 -$4429 -$3719 -$2951

Georgia $725 $317 $3140 -$2936 -$2584 -$2346

Hawaii $725 $335 $465 -$449 -$395 -$347

Idaho $725 $225 $359 -$376 -$331 -$268

Illinois $825 $249 $3200 -$2096 -$1725 -$1364

Indiana $725 $220 $1439 -$1162 -$1023 -$1075

Iowa $725 $192 $589 -$658 -$579 -$440

Kansas $725 $159 $460 -$502 -$441 -$344

Kentucky $725 $298 $1303 -$1133 -$997 -$973

Louisiana $725 $315 $1450 -$1047 -$922 -$1083

Maine $750 $281 $373 -$267 -$231 -$246

Maryland $725 $188 $1109 -$1765 -$1553 -$828

Massachusetts $800 $206 $1366 -$1030 -$861 -$681

Michigan $740 $300 $2963 -$2400 -$2090 -$2054

Minnesota $725 $140 $755 -$1113 -$980 -$564

Mississippi $725 $326 $973 -$649 -$571 -$726

Missouri $750 $241 $1452 -$1278 -$1104 -$956

Montana $790 $190 $191 -$179 -$151 -$101

Nebraska $725 $140 $259 -$409 -$360 -$194

Nevada $825 $191 $527 -$441 -$363 -$225

New Hampshire $725 $126 $167 -$399 -$351 -$125

TABLE 11

SNAP expenditures Predicted changes by state under the Harkin-Miller bill

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3943

36 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

StateMinimum wage

(2014)

SNAP expenditures (2012)Predicted change in SNAP expenditures

(millions of dollars)

Per resident

(millions of

dollars)

Total

(millions of

dollars)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 state

expenditures

(linear regression)

New Jersey $825 $160 $1420 -$1611 -$1325 -$605

New Mexico $750 $324 $675 -$513 -$444 -$445New York $800 $287 $5616 -$3766 -$3148 -$2801

North Carolina $725 $252 $2454 -$2187 -$1925 -$1833

North Dakota $725 $128 $90 -$162 -$143 -$67

Ohio $795 $259 $2995 -$2013 -$1688 -$1539

Oklahoma $725 $248 $945 -$799 -$703 -$706

Oregon $910 $322 $1255 -$272 -$213 -$262

Pennsylvania $725 $218 $2779 -$2930 -$2579 -$2075

Rhode Island $800 $280 $294 -$173 -$144 -$147

South Carolina $725 $291 $1373 -$1337 -$1177 -$1025South Dakota $725 $198 $165 -$192 -$169 -$123

Tennessee $725 $324 $2091 -$1413 -$1243 -$1562

Texas $725 $230 $5997 -$6402 -$5634 -$4479

Utah $725 $141 $402 -$614 -$541 -$300

Vermont $873 $230 $144 -$66 -$53 -$43

Virginia $725 $173 $1413 -$2062 -$1815 -$1056

Washington $932 $244 $1682 -$350 -$270 -$267

West Virginia $725 $273 $508 -$451 -$397 -$379

Wisconsin $725 $204 $1166 -$1302 -$1146 -$871Wyoming $725 $95 $55 -$105 -$93 -$41

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4043

Appendix C | wwwamericanprogresso

FIGURE 2

Raising the minimum wage to $1010would cut taxpayer costs in every state

Predicted decreases in cost and enrollment

in SNAP in 50 states

$200+$51ndash$100

$101ndash$200

0ndash$25

$26ndash$50

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4143

38 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

Endnotes

1 US Department of Agriculture ldquoSupplementalNutrition Assistance Program (SNAP)rdquo available athttpwwwfnsusdagovsnapeligibility (last accessedFebruary 2014)

2 For this initial analysis we do not consider Harkin-Millerrsquos increase in subminimum wages for tippedworkers To do s o would increase the estimated SNAP

savings by an unknown amount

3 The Congressional Budget Office estimates thatworkers currently earning between $1010 and $1150per hour would see their wages rise under the Harkin-Miller proposal Congressional Budget O ffice ldquoTheEffects of a Minimum Wage Increase on Employmentand Family Incomerdquo (2014)

4 Marianne Page Joanne Spetz and Jane Millar ldquoDoesthe Minimum Wage Affect Welfare Caseloadsrdquo Journalof Policy Analysis and Management 24 (2) (2005)273ndash295

5 Marianne Bitler and Hilary Hoynes ldquoThe More ThingsChange the More They Stay the Same The SafetyNet Living Arrangements and Poverty in the GreatRecessionrdquo NBER Working Paper 194 49 2013

6 Sylvia Allegretto and others ldquoFast Food Poverty Wages The Public Cost of Low-Wage Jobs in the Fast-FoodIndustryrdquo (Berkeley California Center for LaborResearch and Education 2013) available at httplaborcenterberkeleyedupubliccostsfast_food_poverty_wage

7 David Neumark and William Wascher ldquoDoes a HigherMinimum Wage Enhance the Effectiveness of theEarned Income Tax Creditrdquo Industrial and LaborRelations Review 64 (4) (2011) 712ndash746

8 David Lee and Emmanuel Saez ldquoOptimal MinimumWage Policy in Competitive Labor Marketsrdquo Journal ofPublic Economics 96 (9) (2012) 739ndash749

9 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family IncomerdquoFebruary 2014

10 Jesse Rothstein ldquoIs the EITC as Good as an NITConditional Cash Transfers and Tax Incidencerdquo AmericanEconomic Journal Economic Policy 2 (1) (2010) 177ndash208

11 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family Incomerdquo

12 Dube Arindrajit 2013 rdquoMinimum Wagesand the Distribution of Family IncomerdquoUnpublished working paper Available at httpsdldropboxusercontentcomu15038936Dube_MinimumWagesFamilyIncomespdf

13 As Dube explains in ldquoThe poverty of Minimum WageFactsrdquo the simulation approach underestimate stemsfrom a number of unwarranted assumptions includingthe range of actual wage increases and the accuracy ofwage data in the Current Population Survey The causal

approach does not make these assumptions

14 Allegretto Sylvia and others 2013 ldquoCredible ResearchDesigns for Minimum Wage Studiesrdquo Working Paper148-13 University of California Berkeley Institutefor Research on Labor and Employment Available athttpwwwirleberkeleyeduworkingpapers148-13pdf

15 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family In comerdquo

16 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo (2012)

17 Ibid

18 Ibid

19 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo

20 US Department of Agriculture ldquoSupplementalNutrition Assistance Program (SNAP)rdquo

21 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo

22 Ibid

23 We deviate from the Census Bureaursquos definition ofa family unit which is ldquotwo people or more (on e ofwhom is the householder) related by bir th marriageor adoption and residing togetherrdquo US Bureau ofthe Census ldquoCurrent Population Survey DefinitionsFamilyrdquo available at httpwwwcensusgovcpsabout

cpsdefhtml (last accessed February 2014) We countas a family unit any individual residing on his or herown two or more persons residing together whodo not belong to a family in the March CPS sampleare constructed as one family in our analysis For thepurposes of food stamp allocations the consumptionresulting from this transfer is probably distributed tofamily members (rather than household members ora single individual within the household) Howeversingle individuals canmdashand domdashreceive SNAPbenefits Excluding them would fail to make theanalysis reflective of the population at large

24 Strictly the family level linear probability modelpredicts the percentage-point decrease in theprobability that an individual family will receive SNAPpayments When applied to a large number of familieshowever we are able to interpret the coefficient asa decrease in the mean of enrollmentmdashthat is a

decrease in the enrollment ratemdashby applying the lawof iterated expectations

25 We generate expenditure predictions from theenrollment modelsmdashand conversely generateenrollment predictions from the expenditure modelmdashby assuming that expenditures per enrolled familyremains the same before and after the minimum wagechange In practice this is likely to be a conservativeestimatemdashthat is to underestimate the decrease inSNAP activity Average SNAP benefits per family willalso decrease as many families that remain eligible forSNAP experience income gains

26 Wage and Hour Division ldquoMinimum Wage Laws inthe States ndash Januar y 1 2014rdquo available at httpwwwdolgovwhdminwageamericahtm (last accessedFebruary 2014)

27 See for example Sylvia Allegretto and others ldquoCredibleResearch Designs for Minimum Wage Studiesrdquo WorkingPaper 148-113 (Berkeley California Institute forResearch on Labor and Employment 2013) available athttpwwwirleberkeleyeduworkingpapers148-13pdf

28 We will report these results in a forthcoming workingpaper

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4243

Endnotes | wwwamericanprogresso

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4343

The Center for American Progress is a nonpartisan research and educational institute

dedicated to promoting a strong just and free America that ensures opportunity

for all We believe that Americans are bound together by a common commitment to

these values and we aspire to ensure that our national policies reflect these values

We work to find progressive and pragmatic solutions to significant domestic and

international problems and develop policy proposals that foster a government that

is ldquoof the people by the people and for the peoplerdquo

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1843

Results | wwwamericanprogresso

Results

Estimated minimum wage effects on SNAP enrollment and

expenditures

able 1 shows he esimaed parameer o ineres991252he coefficien o he

minimum wage or he preerred model o each ype Coefficiens on he

minimum wage variable are no direcly comparable across models because all

our models have a differen uncional orm o undersand and compare hese

esimaes we compue he change in SNAP aciviy prediced or a paricular wage scenario Te final column in able 1 answers he quesion Wha would be

he expeced change in SNAP aciviy in response o a 10 percen increase in he

minimum wage Te answer o his quesion varies wih he value o he inpu

parameers in he able we calculae he percenage decrease in enrollmen or

expendiures prediced or he average sae wih a minimum wage o $725 in

2014 Te sae-level SNAP expendiure model which is a consan-elasiciy

model conveys elasiciy inormaion direcly or he change in expendiures per

capia in he sae

TABLE 1

Comparison of national SNAP predictions for a 10 percent increase in the federal minimum wage

Model LevelRegression

type

Predicted outcome Coefficient of log

(minimum wage)

(Standard error)

Effect of a 10 percent in

in the minimum wage

VariableForm of

variable

Total

enrollment

Tot

expend

1 Family Linear

probability Enrollment Binary (enrolled=1)

-0042

(0008) -317 N

2

StateLinear regression

(ordinary least

squares)

Enrollment State enrollment rate (percent) -0031

(0012) -235 N

3 Expenditures Log (state expenditures per capita) -0190

(0103) NA -19

plt01 plt005 plt001Note Predicted changes are calculated for the average state with a minimum wage of $725 in 2014

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1943

16 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

According o his model a 10 percen increase in he minimum wage would resul

in a 19 percen reducion in SNAP expendiures Tis resul is reassuringly similar

o Dubersquos resul or povery reducion Te wo enrollmen models are much more

precisely esimaed han he expendiure model Te sae-level enrollmen model

finds ha a 10 percen minimum wage increase in a low-wage sae is associaed wih

a 235 percen decrease in SNAP enrollmens Te amily-level linear probabiliymodel predics a somewha greaer elasiciy or low minimum wage saes an

increase o 10 percen in he ederal minimum would resul in a 317 percen decline

in SNAP enrollmen24 Te differences in hese esimaes sem rom a number o

acors including difference in model uncional orm and daa used We rea his

range o elasiciy esimaes as an upper and lower bound on enrollmen impacs

Harkin-Miller bill National and state-level predicted impacts

Wha would be he prediced change or he SNAP program i he ederalminimum were raised o $1010 as proposed in he Harkin-Miller bill In order o

make his inerence we accoun or he ac ha no all saes are currenly subjec

o he ederal minimum wage a he beginning o 2014 21 saes mainained

higher minimum wages han $725 In hose saes an increase in he ederal

minimum wage may or may no be binding or employers in he sae depending

upon wheher he new ederal minimum exceeds he sae-level minimum Bu

regardless o wheher a minimum wage change is binding he impac on SNAP

aciviy will be lower in high minimum wage saes In order o accoun or his

properly we calculae sae by sae he percenage wage change ha would resul

rom he Harkin-Miller proposal and apply he parameers rom each o he hree

models above o compue he expeced decrease in SNAP aciviy or each sae

In his exercise we use saesrsquo curren (2014) minimum wage levels and assume

as a baseline he 2012 levels o SNAP enrollmen and expendiure as 2012 is he

mos recen year or which SNAP daa are available

able 10 and able 11 in Appendix C repor he esimaed effecs on SNAP

enrollmen and expendiures respecively or each sae under he Harkin-Miller

bill25 An increase o $1010 i enaced oday would represen beween a 393

percen wage increase in a $725 minimum wage sae and an 84 percen increasein Washingon sae which has he highes minimum wage in he naion a $932

as o January 201426 Slighly more han 56 percen o he decrease in expendiures

and abou 59 percen o he decrease in enrollmen would occur in saes wih

presen-day minimum wages o $725

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2043

Results | wwwamericanprogresso

In 2012 hese saes were home o 46 percen o he American populaion

and accouned or a roughly equivalen percenage o oal naional SNAP

expendiures No surprisingly he larges enrollmen decreases come rom

populous saes wih very high SNAP enrollmen raes andor very low minimum

wages Te larges prediced enrollmen reducion991252beween 319000 individuals

and 362000 individuals991252would occur in exas which has a $725 minimum wage In Caliornia which has a 206 percen SNAP paricipaion rae and an $8

per hour minimum wage we anicipae SNAP enrollmen declines o 310000

persons o 371000 persons And in Florida which had a SNAP paricipaion rae

o 166 percen and a $793 minimum wage enrollmen could decline by beween

164000 individuals and 196000 individuals For he our saes ha ogeher

accouned or he greaes amoun o SNAP spending in 2012991252exas Caliornia

Florida and New York respecively991252he combined expendiure reducion rom

he Harkin-Miller bill is prediced o be $14 billion

able 2 summarizes he prediced declines in SNAP aciviy or he naion as a whole ha would resul rom he direc and indirec effecs o he Harkin-Miller

bill Enrollmen would all beween 31 million persons and 36 million persons

represening 75 percen o 87 percen o curren enrollmen Te anicipaed

reducion in program expendiures would be nearly $46 billion or 61 percen o

program expendiures

TABLE 2

Comparison of national SNAP predictions under the Harkin-Miller billrsquos $1010 minimum wage

Model

Enrollment

(persons)

Expenditures

(millions of dollars)

Current (2012) Predicted Change Current (2012) Predicted Chan

Family enrollment (linear probability)

41866195

45489339 -3623144

$74861

NA NA

State enrollment (ordinary least squares) 38745435 -3120759 NA NA

State expenditures (ordinary least squares) NA NA $70305 -$45

Note Calculations use 2014 state minimum wages and the most recent SNAP data from 2012 They assume that per-enrollee expenditures remain constant

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2143

18 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

Tere are o course oher possibiliies or minimum wage legislaion able 3

shows he expeced SNAP changes or he Unied Saes under a variey o wage

scenarios calculaed using he sae-level models I saes were no able o se

heir minimum wages independenly such ha all saes were consrained by

he ederal minimum o $725 SNAP would be received by abou 514000 more

people across he Unied Saes a an addiional program cos o nearly hree-quarers o a billion dollars In conras he effecs o a higher minimum wage

proposal991252a ederal wage floor o $11 per hour991252would decrease enrollmen in

SNAP by more han 10 percen and decrease program coss by 83 percen

TABLE 3

Summary of par ticipation and expenditures under wage scenarios

If all states had

minimum wages of

Enrollment(persons)

Expenditures

(millions of dollars)

Predicted Change Predicted Change

Recent levels (2014) 41866195 $74861

$725 42380520 514326 $75604 $743

$800 41423919 -442276 $74209 -$652

$900 40148451 -1717744 $72350 -$2511

$1000 38872982 -2993212 $70490 -$4371

$1010 38745435 -3120759 $70305 -$4556

$1100 37597514 -4268681 $68631 -$6230

Note Calculations use state-level enrollment model coefficient

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2243

Conclusion | wwwamericanprogresso

Conclusion

An exensive body o lieraure examines employmen effecs o he minimum

wage A much smaller se o sudies examines how he minimum wage affecs

povery and only a handul o papers examine he effecs o he minimum wage

on he EIC Our analysis is he firs o examine he effecs o he minimum

wage on SNAP

Our findings indicae ha increased earnings rom minimum wage changes do

reduce SNAP enrollmens and expendiures We esimae ha he Harkin-Miller bill would save axpayers nearly $46 billion per year equivalen o 61 percen

o SNAP expendiures in 2012 he las year or which daa are available Over a

10-year period he esimaed savings amoun o nearly $46 billion

Our repor is subjec o limiaions ha we expec o overcome in our uure

research Firs he findings do no ake ino accoun possible ineracions among

SNAP he EIC and Medicaid Te eligibiliy cuoffs among hese programs

are quie differen suggesing ha such ineracions may be minor Noneheless

he join effecs can only be deermined by urher research using a causal

model Second i would be useul o know he disribuion o SNAP reducions

along he wage disribuion Using he Congressional Budge Officersquos calculaions

o how much he oal dollar value o wage would increase under he Harkin-

Miller proposal our findings imply ha he decline in overall SNAP spending

equals abou 15 percen o he oal resuling increase in wages Te amoun and

disribuion o his offse are o considerable ineres Minimum wage beneficiaries

who come rom working amilies already well above he povery line would no

see any offse while hose who are currenly considerably below he povery line

will see larger offses Tese issues will also be a subjec or our uure research

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2343

20 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

About the authors

Rachel West is a maser o public policy candidae a he Goldman School

o Public Policy Universiy o Caliornia Berkeley Her research ocuses on

economic policy in he areas o low-wage labor and povery

Michael Reich is proessor o economics and direcor o he Insiue or

Research on Labor and Employmen a he Universiy o Caliornia a Berkeley

His research publicaions cover numerous areas o labor economics including

racial inequaliy labor marke segmenaion high-perormance workplaces

union-managemen cooperaion Japanese labor-managemen sysems living

wages and minimum wages He received his docorae in economics rom

Harvard Universiy

Acknowledgments

We are graeul o Sylvia Allegreto Arindraji Dube Bill Leser Jesse Rohsein

Daniel Tompson and Ben Zipperer or heir valuable suggesions

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2443

References | wwwamericanprogresso

References

Allegreto Sylvia and ohers 2013 ldquoFas Food Povery Wages Te Public Cos o Low-Wage Jobsin he Fas-Food Indusryrdquo Universiy o Caliornia Berkeley Cener or Labor Research andEducaion Available a htplaborcenerberkeleyedupubliccossas_ood_povery_wage

Allegreto Sylvia and ohers 2013 ldquoCredible Research Designs or Minimum Wage Sudiesrdquo

Working Paper 148-13 Universiy o Caliornia Berkeley Insiue or Research on Labor andEmploymen Available a htpwwwirleberkeleyeduworkingpapers148-13pd

Biler Marianne and Hilary Hoynes 2013 ldquo Te More Tings Change he More Tey Say heSame Te Saey Ne Living Arrangemens and Povery in he Grea Recessionrdquo Working Paper19449 Naional Bureau o Economic Research

Congressional Budge Office 2012 ldquoTe Supplemenal Nuriion Assisance Programrdquo Washingon Available a htpwwwcbogovsiesdeaulfilescbofilesatachmens04-19-SNAPpd

991252 991252 991252 2014 ldquoTe Effec o a Minimum-Wage Increase on Employmen and Family Incomerdquo Washingon Available a htpwwwcbogovsiesdeaulfilescbofilesatachmens44995-MinimumWagepd

Dube Arindraji 2013 rdquoMinimum Wages and he Disribuion o Family Incomerdquo Unpublished working paper Available a htpsdldropboxuserconencomu15038936Dube_ MinimumWagesFamilyIncomespd

991252 991252 991252 2014 ldquoTe Povery o Minimum Wage Facsrdquo Arindraji Dube Blog January 22 Availablea htparindubecom20140122he-povery-o-minimum-wage-acs

Lee David and Emmanuel Saez 2012 ldquoOpimal Minimum Wage Policy in Compeiive LaborMarkesrdquo Journal o Public Economics 96 (9) 739ndash749

Neumark David and William Wascher 1992 ldquoEmploymen Effecs o Minimum and Subminimum Wages Panel Daa on Sae Minimum Wage Lawsrdquo Industrial and Labor Relations Review 46 (1)

55ndash81

Neumark David and William Wascher 2011 ldquoDoes a Higher Minimum Wage Enhance heEffeciveness o he Earned Income ax Credirdquo Industrial and Labor Relations Review 64 (4)712ndash746

Page Marianne Joanne Spez and Jane Millar 2005 ldquoDoes he Minimum Wage Affec WelareCaseloadsrdquo Journal o Policy Analysis and Management 24 (2) 273ndash295

Rohsein Jesse 2010 ldquoIs he EIC as Good as an NI Condiional Cash ransers and ax

Incidencerdquo American Economic Journal Economic Policy 2 (1) 177ndash208

Wage and Hours Division 2014 ldquoMinimum Wage Laws in he Saes ndash January 1 2014rdquo USDeparmen o Labor (htpwwwdolgovwhdminwageamericahm [February 2014])

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2543

22 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2643

Appendix A | wwwamericanprogresso

Appendix A

Pre-trend falsification check

Recen minimum wage research27 highlighs a common flaw in previous sudies

ailure o veriy ha he oucome variable is ree o negaive pre-exising rends I

or example SNAP aciviy was already rending down in saes ha raised heir

minimum wages beore hese changes came ino effec our regression analysis could

(misakenly) atribue ha reducion o he minimum wage We check or such

pre-rends by inroducing variables ha represen he prior yearrsquos value or leads o

he minimum wage I he model esimaes he minimum wage o have an effec on

he oucome variable beore he wage change wen ino effec hen an unobservedacor no he minimum wage change caused he change in SNAP aciviy

We es he specificaions above or pre-rends by including a one-year lead in

all hree specificaions We find ha he lead erms are small posiive and no

saisically significan indicaing ha he concurren minimum wage991252no

he wage level in prior periods991252is driving he observed changes in SNAP

oucomes28 In paricular he coefficien (sandard error) on he lead erm in

our preerred amily-level enrollmen regression is 011 and no significan

while he coefficien and sandard error o he conemporaneous minimum

wage is unchanged In he sae-level preerred enrollmen regression he

coefficien o he lead erm is again small (07) and i is no significan Te

corresponding coefficien on he lead erm in he sae-level expendiure

regression is 16 and is no significan Te posiive poin esimaes on hese lead

erms resuls no only rule ou disoring negaive pre-rends Tey also sugges

ha our main resuls may underesimae he rue effecs

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2743

24 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2843

Appendix B | wwwamericanprogresso

Appendix B

Model selection process

For boh he amily-level and sae-level models we es hree mehods o conrol

or unobserved geographic- and ime-varying characerisics as suggesed by he

minimum wage lieraure Firs we include only independen sae-specific fixed

effecs and year-specific fixed effecs Tis specificaion (specificaion 1) implicily

assumes ha amilies in any sae consiue an equally good saisical ldquoconrolrdquo

group or hose in any randomly chosen sae afer accouning or various

characerisics (median income and unemploymen rae among ohers) Similarly

simple ime fixed effecs assume ha amilies surveyed in any year can crediblyserve as a conrol group or amilies surveyed in every oher year o he sample

(1990 hrough 2012)

In oher words specificaion 1 assumes ha a saersquos immediae neighbor provides

no beter a couneracual or he effec o a minimum wage change han does a

sae across he counry We relax his resricive specificaion sequenially in wo

seps In specificaion 2 we replace simple year fixed effecs wih fixed effecs or

each Census divisionyear (capured as an addiional variable in he vecor By

using division-year effecs we remove he resricion ha amilies in each sae

are equally good saisical conrols or all oher amilies Raher we allow or he

possibiliy ha amilies in similar geographic regions (or example he Souh or

he Norheas) may be more similar o one anoher han amilies arher away

Finally in specificaion 3 we add sae-specific linear ime rends o he previous

specificaion Tus specificaion 3 is he mos rigorous model specificaion in ha

i allows or heerogeneiy along hree dimensions Ta is specificaion 3 allows

each sae o have is own ime-varying rends raher han imposing he resricion

ha saes evolve idenically over he 22 years in he sample

We begin building he heoreical specificaion above rom a se o simpleuncondiional models regression o SNAP aciviy (enrollmen or expendiures)

on he log o he minimum wage and a se o geographic- and ime-specific

effecs (specificaions 1 2 and 3 described above) As shown in ables 1ndash3 (or

specificaion 3) we hen add covariaes sequenially o hese models including

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2943

26 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

firs he vecor o amily-level conrols ollowed by each o several sae-level

covariaes in urn he unemploymen rae log o median-amily income and he

employmen-o-populaion raio Comparable resuls or specificaions 2 and 3

will be available in our orhcoming working paper

As expeced he simple uncondiional models indicae ha he relaionship beween he minimum wage and SNAP enrollmen i one exiss is a more complex

one influenced by oher acors In he uncondiional model he coefficien on he

variable o ineres991252he log o he minimum wage991252is small in magniude and no

saisically differen rom zero Once we accoun or he influence o labor marke

condiions and variaion in income levels on program paricipaion (by including

unemploymen rae and median-amily income conrol variables respecively)

he effec o he minimum wage on SNAP enrollmen is precisely esimaed Te

coefficien o he log minimum wage is slighly higher (-0042) in he amily-level

analysis han he coefficien (-031) in he sae-level analysis Te level o precision

is also higher in he amily-level analysis Tis is o be expeced when using 124million observaions compared o 1127

Te second dimension o model choice concerns he effec specificaion ables

7ndash9 compare he primary coefficiens o ineres or he SNAP enrollmen and

expendiure models For boh he enrollmen models he effec sizes are smalles

or specificaion 1 larges or specificaion 2 and inermediae beween hese wo

in specificaion 3 Recall ha Specificaion 3 conains sae-specific linear ime

rends in addiion o he census divisionyear conrols included in specificaion

2 In he amily-level enrollmen model he sandard error o he minimum wage

coefficien is smaller han in he oher wo specificaions Sandard errors on he

oher variables are much smaller in specificaions 2 and 3 han in specificaion

1 On he basis o coefficien significance (join and individual) specificaions 2

and 3 are sricly preerred in boh enrollmen models o specificaion 1 which

conains only sae and year fixed effecs

A concern wih specificaions 2 and 3 is ha rend conrols such as sae linear

rends may incorrecly absorb some o he delayed impac o a minimum wage

When we es his issue by including lagged minimum wages we do no find ha

delayed effecs are significan Anoher concern is ha more sauraed modelsuse less o he saisical variaion which could reduce he saisical power o

he resuls However he sandard errors or our more sauraed models are no

higher and are lower in some cases han or he less sauraed models Overall

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3043

Appendix B | wwwamericanprogresso

his evidence suppors our use o he sauraed model as he preerred model

specificaion Moreover Dubersquos 2013 sudy shows ha more sauraed models

perorm beter han models wih jus sae and ime fixed effecs

Te esimaed enrollmen regressions a boh he amily and sae levels show large

and saisically significan coefficiens Te esimaed minimum wage effec in heexpendiures regressions991252or which we have only sae-level daa991252is also large

and saisically significan

We do no use weighed regression or he sae-level models preerring o keep

analysis o he ldquoreamenrdquo (ha is o say a minimum wage change) appropriae

o he average sae raher han he average amily or individual I insead our

primary ineres were he impac o a minimum wage change on he average amily

or he average individual we migh choose o designae he number o amilies

in each sae or he sae populaion respecively as analyic weighs in order o

obain a coefficien beter suied or such inerence

TABLE 4

SNAP enrollment

Family-level linear probability

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0048 -0047 -0040 -0043 -0042

(0013) (0013) (001) (0008) (0009)

Unemployment rate (100) 0505 0420 0280

(0083) (0086) (0082)

Log median income -0057 -0039

(0011) (0011)

Employment-to-population ratio -0239

(0038)

N 1242022 1242022 1242022 1242022 1242022

plt01 plt005 plt001

Note Robust standard errors are in parentheses Observations are clustered at the state level The outcome variable is binary and equal to one if a family is enrolledin SNAP All models include state fixed effects Census division x-year fixed effects and state-specific linear time trends All specifications except 3a include additional

controls for family size race and marital status of the family head presence of children and presence of an adult maleSource Annual data from the March Current Population Survey for 1990 to 2012 estimation includes Current Population Survey probability weights

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3143

28 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

TABLE 5

SNAP enrollment

State-level linear regression

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0028 -0024 -0026 -0031 -0031

(0014) (0013) (0013) (0012) (0012)

Unemployment rate (100) 0625 0466 0320 0339

(0087) (0088) (0085) (0083)

Log median income -0090 -0065 -0061

(0013) (0013) (0013)

Employment-to-population ratio -0282 -0248

(0037) (0038)

N 1127 1127 1127 1127 1127

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the SNAP enrollment rate All models include state fixed effects Census division x-year fixedeffects and state-specific linear time trends

Source Annual data from the March Current Population Survey for 1990 to 2012

TABLE 6

SNAP expenditures

State-level linear regression

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0181 -0149 -0156 -0153 -0190

(011) (0103) (0102) (0103) (0103)

Unemployment rate (100) 4587 4152 4232 4313

(0622) (0621) (0633) (0628)

Log median income -0246 -0261 -0294

(0075) (0078) (0078)

Employment-to-population ratio 0155 0244

(0237) (024)

N 1127 1127 1127 1127 1127

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the natural log of state SNAP expenditures per capita for 1990 to 2012 All models include state

fixed effects Census division x-year fixed effects and state-specific linear time trends

Source Annual data from Bureau of Economic Analysis national income and product accounts tables for 1990 to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3243

Appendix B | wwwamericanprogresso

TABLE 7

Comparison of specifications SNAP enrollment

Family-level linear probability

(1) (2) (3)

Log minimum wage -0028 -0049 -0042

(0014) (0017) (0009)

Unemployment rate (100) 0275 0297 0280

(0161) (0076) (0082)

Log median income -0077 -0055 -0039

(0014) (0012) (0011)

Employment-to-population ratio -0238 -0250 -0239

(0054) (004) (0038)

N 1242022 1242022 1242022

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses Observations are clustered at the state level The outcomevariable is binary or equal to one if a family is enrolled in SNAP All specifications include additional controls forfamily size race and marital status of the family head presence of children and presence of an adult male

Source Annual data from the March Current Population Survey for 1990 to 2012 estimation includes CurrentPopulation Survey probability weights

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3343

30 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

TABLE 8

Comparison of specifications SNAP enrollment

State-level linear regression

(1) (2) (3)

Log minimum wage -0019 -0035 -0031

(0009) (0012) (0012)

Unemployment rate (100) 0401 0370 0339

(0063) (0077) (0083)

Log median income -0081 -0073 -0061

(0011) (0013) (0013)

Employment-to-population ratio -0183 -0222 -0248

(0039) (0039) (0038)

N 1127 1127 1127

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the SNAP enrollment rate Allregressions include the share of households with children and the racial shares of the population

Source Annual data from the March Current Population Survey for 1990 to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3443

Appendix B | wwwamericanprogresso

TABLE 9

Comparison of specifications SNAP expenditures

State-level linear regression

(1) (2) (3)

Log minimum wage -0121 -0203 -0190

(0075) (0103) (0103)

Unemployment rate (100) 5292 5152 4313

(0464) (0576) (0628)

Log median income -0437 -0417 -0294

(008) (0086) (0078)

Employment-to-population ratio -0040 -0220 0244

(0261) (0260) (0240)

N 1127 1127 1127

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the natural log of state SNAPexpenditures per capita All regressions include the share of households with children and the racial shares ofthe state population

Source Annual data from Bureau of Economic Analysis national income and product accounts tables for 1990to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3543

32 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3643

Appendix C | wwwamericanprogresso

Appendix C Harkin-Miller

policy simulation results

StateMinimum wage

(2014)

SNAP enrollment (2012) Predicted change in SNAP enrollment

RateTotal

(persons)

Model 1 family

enrollment (linearprobability)

Model 2 state

enrollment (linearregression)

Model 3 s

expenditu(linear regre

Alabama $725 164 788682 -66922 -58897 -58906

Alaska $775 120 87436 -8104 -6887 -3288

Arizona $790 201 1319323 -67435 -56738 -64356

Arkansas $725 78 230489 -40977 -36063 -35248

California $800 206 7813680 -371131 -310222 -18223

Colorado $800 164 853155 -50684 -42365 -23926

Connecticut $870 91 326621 -22456 -17975 -13711

Delaware $725 186 170262 -12739 -11211 -10647

District of Columbia $825 133 84009 -5370 -4417 -3632

Florida $793 166 3208026 -195813 -164426 -13046

Georgia $725 160 1586336 -137741 -121224 -11004

Hawaii $725 96 133662 -19310 -16995 -14933

Idaho $725 92 147501 -22165 -19507 -15809

Illinois $825 95 1225084 -109088 -89742 -70955

Indiana $725 125 816233 -90818 -79928 -83985

Iowa $725 155 478011 -42716 -37594 -28556

Kansas $725 135 388269 -40082 -35275 -27461

Kentucky $725 130 568821 -60840 -53544 -52259

Louisiana $725 149 683832 -63929 -56263 -66083

Maine $750 77 101976 -16567 -14323 -15234

Maryland $725 144 846415 -81748 -71946 -38370

Massachusetts $800 130 864721 -64902 -54251 -42913

Michigan $740 146 1439141 -128801 -112140 -11022

Minnesota $725 133 713646 -74730 -65769 -37878

TABLE 10

SNAP enrollments Predicted changes by state under the Harkin-Miller bill

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3743

34 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

StateMinimum wage

(2014)

SNAP enrollment (2012) Predicted change in SNAP enrollment

RateTotal

(persons)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 state

expenditures

(linear regression)

Mississippi $725 129 386501 -41486 -36511 -46467

Missouri $750 172 1036182 -75131 -64952 -56244

Montana $790 132 132452 -10350 -8708 -5846

Nebraska $725 124 230683 -25773 -22683 -12189

Nevada $825 162 446035 -23349 -19209 -11894

New Hampshire $725 127 168404 -18359 -16157 -5735

New Jersey $825 160 1416666 -75175 -61843 -28236

New Mexico $750 149 310896 -25983 -22463 -22512

New York $800 192 3763553 -191193 -159815 -142182

North Carolina $725 174 1697193 -135417 -119179 -113503

North Dakota $725 87 61225 -9743 -8574 -4021

Ohio $795 143 1647345 -115869 -97169 -88580

Oklahoma $725 129 494053 -53006 -46650 -46854

Oregon $910 124 485326 -17036 -13328 -16398

Pennsylvania $725 161 2053643 -177315 -156052 -125586

Rhode Island $800 156 163730 -10258 -8574 -8698

South Carolina $725 94 445277 -65614 -57746 -50304

South Dakota $725 208 173749 -11586 -10197 -7458

Tennessee $725 142 914903 -89667 -78915 -99134

Texas $725 110 2863779 -362018 -318607 -253285

Utah $725 88 251107 -39658 -34902 -19390

Vermont $873 156 97792 -3823 -3055 -2475

Virginia $725 101 829771 -113723 -100086 -58212

Washington $932 72 496934 -23221 -17947 -17756

West Virginia $725 58 107875 -25792 -22699 -21665

Wisconsin $725 75 427822 -79521 -69986 -53210

Wyoming $725 164 94590 -8010 -7050 -3104

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3843

Appendix C | wwwamericanprogresso

StateMinimum wage

(2014)

SNAP expenditures (2012)Predicted change in SNAP expenditures

(millions of dollars)

Per resident

(millions of

dollars)

Total

(millions of

dollars)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 s

expenditu

(linear regre

Alabama $725 $290 $1397 -$1185 -$1043 -$1043

Alaska $775 $253 $185 -$262 -$223 -$106

Arizona $790 $257 $1686 -$935 -$787 -$892

Arkansas $725 $248 $733 -$636 -$560 -$547

California $800 $189 $7164 -$7276 -$6082 -$3573

Colorado $800 $157 $816 -$862 -$721 -$407

Connecticut $870 $191 $686 -$343 -$275 -$210

Delaware $725 $250 $229 -$205 -$180 -$171

District of Columbia $825 $366 $232 -$146 -$120 -$99

Florida $793 $294 $5676 -$4429 -$3719 -$2951

Georgia $725 $317 $3140 -$2936 -$2584 -$2346

Hawaii $725 $335 $465 -$449 -$395 -$347

Idaho $725 $225 $359 -$376 -$331 -$268

Illinois $825 $249 $3200 -$2096 -$1725 -$1364

Indiana $725 $220 $1439 -$1162 -$1023 -$1075

Iowa $725 $192 $589 -$658 -$579 -$440

Kansas $725 $159 $460 -$502 -$441 -$344

Kentucky $725 $298 $1303 -$1133 -$997 -$973

Louisiana $725 $315 $1450 -$1047 -$922 -$1083

Maine $750 $281 $373 -$267 -$231 -$246

Maryland $725 $188 $1109 -$1765 -$1553 -$828

Massachusetts $800 $206 $1366 -$1030 -$861 -$681

Michigan $740 $300 $2963 -$2400 -$2090 -$2054

Minnesota $725 $140 $755 -$1113 -$980 -$564

Mississippi $725 $326 $973 -$649 -$571 -$726

Missouri $750 $241 $1452 -$1278 -$1104 -$956

Montana $790 $190 $191 -$179 -$151 -$101

Nebraska $725 $140 $259 -$409 -$360 -$194

Nevada $825 $191 $527 -$441 -$363 -$225

New Hampshire $725 $126 $167 -$399 -$351 -$125

TABLE 11

SNAP expenditures Predicted changes by state under the Harkin-Miller bill

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3943

36 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

StateMinimum wage

(2014)

SNAP expenditures (2012)Predicted change in SNAP expenditures

(millions of dollars)

Per resident

(millions of

dollars)

Total

(millions of

dollars)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 state

expenditures

(linear regression)

New Jersey $825 $160 $1420 -$1611 -$1325 -$605

New Mexico $750 $324 $675 -$513 -$444 -$445New York $800 $287 $5616 -$3766 -$3148 -$2801

North Carolina $725 $252 $2454 -$2187 -$1925 -$1833

North Dakota $725 $128 $90 -$162 -$143 -$67

Ohio $795 $259 $2995 -$2013 -$1688 -$1539

Oklahoma $725 $248 $945 -$799 -$703 -$706

Oregon $910 $322 $1255 -$272 -$213 -$262

Pennsylvania $725 $218 $2779 -$2930 -$2579 -$2075

Rhode Island $800 $280 $294 -$173 -$144 -$147

South Carolina $725 $291 $1373 -$1337 -$1177 -$1025South Dakota $725 $198 $165 -$192 -$169 -$123

Tennessee $725 $324 $2091 -$1413 -$1243 -$1562

Texas $725 $230 $5997 -$6402 -$5634 -$4479

Utah $725 $141 $402 -$614 -$541 -$300

Vermont $873 $230 $144 -$66 -$53 -$43

Virginia $725 $173 $1413 -$2062 -$1815 -$1056

Washington $932 $244 $1682 -$350 -$270 -$267

West Virginia $725 $273 $508 -$451 -$397 -$379

Wisconsin $725 $204 $1166 -$1302 -$1146 -$871Wyoming $725 $95 $55 -$105 -$93 -$41

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4043

Appendix C | wwwamericanprogresso

FIGURE 2

Raising the minimum wage to $1010would cut taxpayer costs in every state

Predicted decreases in cost and enrollment

in SNAP in 50 states

$200+$51ndash$100

$101ndash$200

0ndash$25

$26ndash$50

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4143

38 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

Endnotes

1 US Department of Agriculture ldquoSupplementalNutrition Assistance Program (SNAP)rdquo available athttpwwwfnsusdagovsnapeligibility (last accessedFebruary 2014)

2 For this initial analysis we do not consider Harkin-Millerrsquos increase in subminimum wages for tippedworkers To do s o would increase the estimated SNAP

savings by an unknown amount

3 The Congressional Budget Office estimates thatworkers currently earning between $1010 and $1150per hour would see their wages rise under the Harkin-Miller proposal Congressional Budget O ffice ldquoTheEffects of a Minimum Wage Increase on Employmentand Family Incomerdquo (2014)

4 Marianne Page Joanne Spetz and Jane Millar ldquoDoesthe Minimum Wage Affect Welfare Caseloadsrdquo Journalof Policy Analysis and Management 24 (2) (2005)273ndash295

5 Marianne Bitler and Hilary Hoynes ldquoThe More ThingsChange the More They Stay the Same The SafetyNet Living Arrangements and Poverty in the GreatRecessionrdquo NBER Working Paper 194 49 2013

6 Sylvia Allegretto and others ldquoFast Food Poverty Wages The Public Cost of Low-Wage Jobs in the Fast-FoodIndustryrdquo (Berkeley California Center for LaborResearch and Education 2013) available at httplaborcenterberkeleyedupubliccostsfast_food_poverty_wage

7 David Neumark and William Wascher ldquoDoes a HigherMinimum Wage Enhance the Effectiveness of theEarned Income Tax Creditrdquo Industrial and LaborRelations Review 64 (4) (2011) 712ndash746

8 David Lee and Emmanuel Saez ldquoOptimal MinimumWage Policy in Competitive Labor Marketsrdquo Journal ofPublic Economics 96 (9) (2012) 739ndash749

9 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family IncomerdquoFebruary 2014

10 Jesse Rothstein ldquoIs the EITC as Good as an NITConditional Cash Transfers and Tax Incidencerdquo AmericanEconomic Journal Economic Policy 2 (1) (2010) 177ndash208

11 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family Incomerdquo

12 Dube Arindrajit 2013 rdquoMinimum Wagesand the Distribution of Family IncomerdquoUnpublished working paper Available at httpsdldropboxusercontentcomu15038936Dube_MinimumWagesFamilyIncomespdf

13 As Dube explains in ldquoThe poverty of Minimum WageFactsrdquo the simulation approach underestimate stemsfrom a number of unwarranted assumptions includingthe range of actual wage increases and the accuracy ofwage data in the Current Population Survey The causal

approach does not make these assumptions

14 Allegretto Sylvia and others 2013 ldquoCredible ResearchDesigns for Minimum Wage Studiesrdquo Working Paper148-13 University of California Berkeley Institutefor Research on Labor and Employment Available athttpwwwirleberkeleyeduworkingpapers148-13pdf

15 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family In comerdquo

16 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo (2012)

17 Ibid

18 Ibid

19 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo

20 US Department of Agriculture ldquoSupplementalNutrition Assistance Program (SNAP)rdquo

21 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo

22 Ibid

23 We deviate from the Census Bureaursquos definition ofa family unit which is ldquotwo people or more (on e ofwhom is the householder) related by bir th marriageor adoption and residing togetherrdquo US Bureau ofthe Census ldquoCurrent Population Survey DefinitionsFamilyrdquo available at httpwwwcensusgovcpsabout

cpsdefhtml (last accessed February 2014) We countas a family unit any individual residing on his or herown two or more persons residing together whodo not belong to a family in the March CPS sampleare constructed as one family in our analysis For thepurposes of food stamp allocations the consumptionresulting from this transfer is probably distributed tofamily members (rather than household members ora single individual within the household) Howeversingle individuals canmdashand domdashreceive SNAPbenefits Excluding them would fail to make theanalysis reflective of the population at large

24 Strictly the family level linear probability modelpredicts the percentage-point decrease in theprobability that an individual family will receive SNAPpayments When applied to a large number of familieshowever we are able to interpret the coefficient asa decrease in the mean of enrollmentmdashthat is a

decrease in the enrollment ratemdashby applying the lawof iterated expectations

25 We generate expenditure predictions from theenrollment modelsmdashand conversely generateenrollment predictions from the expenditure modelmdashby assuming that expenditures per enrolled familyremains the same before and after the minimum wagechange In practice this is likely to be a conservativeestimatemdashthat is to underestimate the decrease inSNAP activity Average SNAP benefits per family willalso decrease as many families that remain eligible forSNAP experience income gains

26 Wage and Hour Division ldquoMinimum Wage Laws inthe States ndash Januar y 1 2014rdquo available at httpwwwdolgovwhdminwageamericahtm (last accessedFebruary 2014)

27 See for example Sylvia Allegretto and others ldquoCredibleResearch Designs for Minimum Wage Studiesrdquo WorkingPaper 148-113 (Berkeley California Institute forResearch on Labor and Employment 2013) available athttpwwwirleberkeleyeduworkingpapers148-13pdf

28 We will report these results in a forthcoming workingpaper

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4243

Endnotes | wwwamericanprogresso

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4343

The Center for American Progress is a nonpartisan research and educational institute

dedicated to promoting a strong just and free America that ensures opportunity

for all We believe that Americans are bound together by a common commitment to

these values and we aspire to ensure that our national policies reflect these values

We work to find progressive and pragmatic solutions to significant domestic and

international problems and develop policy proposals that foster a government that

is ldquoof the people by the people and for the peoplerdquo

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 1943

16 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

According o his model a 10 percen increase in he minimum wage would resul

in a 19 percen reducion in SNAP expendiures Tis resul is reassuringly similar

o Dubersquos resul or povery reducion Te wo enrollmen models are much more

precisely esimaed han he expendiure model Te sae-level enrollmen model

finds ha a 10 percen minimum wage increase in a low-wage sae is associaed wih

a 235 percen decrease in SNAP enrollmens Te amily-level linear probabiliymodel predics a somewha greaer elasiciy or low minimum wage saes an

increase o 10 percen in he ederal minimum would resul in a 317 percen decline

in SNAP enrollmen24 Te differences in hese esimaes sem rom a number o

acors including difference in model uncional orm and daa used We rea his

range o elasiciy esimaes as an upper and lower bound on enrollmen impacs

Harkin-Miller bill National and state-level predicted impacts

Wha would be he prediced change or he SNAP program i he ederalminimum were raised o $1010 as proposed in he Harkin-Miller bill In order o

make his inerence we accoun or he ac ha no all saes are currenly subjec

o he ederal minimum wage a he beginning o 2014 21 saes mainained

higher minimum wages han $725 In hose saes an increase in he ederal

minimum wage may or may no be binding or employers in he sae depending

upon wheher he new ederal minimum exceeds he sae-level minimum Bu

regardless o wheher a minimum wage change is binding he impac on SNAP

aciviy will be lower in high minimum wage saes In order o accoun or his

properly we calculae sae by sae he percenage wage change ha would resul

rom he Harkin-Miller proposal and apply he parameers rom each o he hree

models above o compue he expeced decrease in SNAP aciviy or each sae

In his exercise we use saesrsquo curren (2014) minimum wage levels and assume

as a baseline he 2012 levels o SNAP enrollmen and expendiure as 2012 is he

mos recen year or which SNAP daa are available

able 10 and able 11 in Appendix C repor he esimaed effecs on SNAP

enrollmen and expendiures respecively or each sae under he Harkin-Miller

bill25 An increase o $1010 i enaced oday would represen beween a 393

percen wage increase in a $725 minimum wage sae and an 84 percen increasein Washingon sae which has he highes minimum wage in he naion a $932

as o January 201426 Slighly more han 56 percen o he decrease in expendiures

and abou 59 percen o he decrease in enrollmen would occur in saes wih

presen-day minimum wages o $725

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2043

Results | wwwamericanprogresso

In 2012 hese saes were home o 46 percen o he American populaion

and accouned or a roughly equivalen percenage o oal naional SNAP

expendiures No surprisingly he larges enrollmen decreases come rom

populous saes wih very high SNAP enrollmen raes andor very low minimum

wages Te larges prediced enrollmen reducion991252beween 319000 individuals

and 362000 individuals991252would occur in exas which has a $725 minimum wage In Caliornia which has a 206 percen SNAP paricipaion rae and an $8

per hour minimum wage we anicipae SNAP enrollmen declines o 310000

persons o 371000 persons And in Florida which had a SNAP paricipaion rae

o 166 percen and a $793 minimum wage enrollmen could decline by beween

164000 individuals and 196000 individuals For he our saes ha ogeher

accouned or he greaes amoun o SNAP spending in 2012991252exas Caliornia

Florida and New York respecively991252he combined expendiure reducion rom

he Harkin-Miller bill is prediced o be $14 billion

able 2 summarizes he prediced declines in SNAP aciviy or he naion as a whole ha would resul rom he direc and indirec effecs o he Harkin-Miller

bill Enrollmen would all beween 31 million persons and 36 million persons

represening 75 percen o 87 percen o curren enrollmen Te anicipaed

reducion in program expendiures would be nearly $46 billion or 61 percen o

program expendiures

TABLE 2

Comparison of national SNAP predictions under the Harkin-Miller billrsquos $1010 minimum wage

Model

Enrollment

(persons)

Expenditures

(millions of dollars)

Current (2012) Predicted Change Current (2012) Predicted Chan

Family enrollment (linear probability)

41866195

45489339 -3623144

$74861

NA NA

State enrollment (ordinary least squares) 38745435 -3120759 NA NA

State expenditures (ordinary least squares) NA NA $70305 -$45

Note Calculations use 2014 state minimum wages and the most recent SNAP data from 2012 They assume that per-enrollee expenditures remain constant

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2143

18 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

Tere are o course oher possibiliies or minimum wage legislaion able 3

shows he expeced SNAP changes or he Unied Saes under a variey o wage

scenarios calculaed using he sae-level models I saes were no able o se

heir minimum wages independenly such ha all saes were consrained by

he ederal minimum o $725 SNAP would be received by abou 514000 more

people across he Unied Saes a an addiional program cos o nearly hree-quarers o a billion dollars In conras he effecs o a higher minimum wage

proposal991252a ederal wage floor o $11 per hour991252would decrease enrollmen in

SNAP by more han 10 percen and decrease program coss by 83 percen

TABLE 3

Summary of par ticipation and expenditures under wage scenarios

If all states had

minimum wages of

Enrollment(persons)

Expenditures

(millions of dollars)

Predicted Change Predicted Change

Recent levels (2014) 41866195 $74861

$725 42380520 514326 $75604 $743

$800 41423919 -442276 $74209 -$652

$900 40148451 -1717744 $72350 -$2511

$1000 38872982 -2993212 $70490 -$4371

$1010 38745435 -3120759 $70305 -$4556

$1100 37597514 -4268681 $68631 -$6230

Note Calculations use state-level enrollment model coefficient

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2243

Conclusion | wwwamericanprogresso

Conclusion

An exensive body o lieraure examines employmen effecs o he minimum

wage A much smaller se o sudies examines how he minimum wage affecs

povery and only a handul o papers examine he effecs o he minimum wage

on he EIC Our analysis is he firs o examine he effecs o he minimum

wage on SNAP

Our findings indicae ha increased earnings rom minimum wage changes do

reduce SNAP enrollmens and expendiures We esimae ha he Harkin-Miller bill would save axpayers nearly $46 billion per year equivalen o 61 percen

o SNAP expendiures in 2012 he las year or which daa are available Over a

10-year period he esimaed savings amoun o nearly $46 billion

Our repor is subjec o limiaions ha we expec o overcome in our uure

research Firs he findings do no ake ino accoun possible ineracions among

SNAP he EIC and Medicaid Te eligibiliy cuoffs among hese programs

are quie differen suggesing ha such ineracions may be minor Noneheless

he join effecs can only be deermined by urher research using a causal

model Second i would be useul o know he disribuion o SNAP reducions

along he wage disribuion Using he Congressional Budge Officersquos calculaions

o how much he oal dollar value o wage would increase under he Harkin-

Miller proposal our findings imply ha he decline in overall SNAP spending

equals abou 15 percen o he oal resuling increase in wages Te amoun and

disribuion o his offse are o considerable ineres Minimum wage beneficiaries

who come rom working amilies already well above he povery line would no

see any offse while hose who are currenly considerably below he povery line

will see larger offses Tese issues will also be a subjec or our uure research

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2343

20 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

About the authors

Rachel West is a maser o public policy candidae a he Goldman School

o Public Policy Universiy o Caliornia Berkeley Her research ocuses on

economic policy in he areas o low-wage labor and povery

Michael Reich is proessor o economics and direcor o he Insiue or

Research on Labor and Employmen a he Universiy o Caliornia a Berkeley

His research publicaions cover numerous areas o labor economics including

racial inequaliy labor marke segmenaion high-perormance workplaces

union-managemen cooperaion Japanese labor-managemen sysems living

wages and minimum wages He received his docorae in economics rom

Harvard Universiy

Acknowledgments

We are graeul o Sylvia Allegreto Arindraji Dube Bill Leser Jesse Rohsein

Daniel Tompson and Ben Zipperer or heir valuable suggesions

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2443

References | wwwamericanprogresso

References

Allegreto Sylvia and ohers 2013 ldquoFas Food Povery Wages Te Public Cos o Low-Wage Jobsin he Fas-Food Indusryrdquo Universiy o Caliornia Berkeley Cener or Labor Research andEducaion Available a htplaborcenerberkeleyedupubliccossas_ood_povery_wage

Allegreto Sylvia and ohers 2013 ldquoCredible Research Designs or Minimum Wage Sudiesrdquo

Working Paper 148-13 Universiy o Caliornia Berkeley Insiue or Research on Labor andEmploymen Available a htpwwwirleberkeleyeduworkingpapers148-13pd

Biler Marianne and Hilary Hoynes 2013 ldquo Te More Tings Change he More Tey Say heSame Te Saey Ne Living Arrangemens and Povery in he Grea Recessionrdquo Working Paper19449 Naional Bureau o Economic Research

Congressional Budge Office 2012 ldquoTe Supplemenal Nuriion Assisance Programrdquo Washingon Available a htpwwwcbogovsiesdeaulfilescbofilesatachmens04-19-SNAPpd

991252 991252 991252 2014 ldquoTe Effec o a Minimum-Wage Increase on Employmen and Family Incomerdquo Washingon Available a htpwwwcbogovsiesdeaulfilescbofilesatachmens44995-MinimumWagepd

Dube Arindraji 2013 rdquoMinimum Wages and he Disribuion o Family Incomerdquo Unpublished working paper Available a htpsdldropboxuserconencomu15038936Dube_ MinimumWagesFamilyIncomespd

991252 991252 991252 2014 ldquoTe Povery o Minimum Wage Facsrdquo Arindraji Dube Blog January 22 Availablea htparindubecom20140122he-povery-o-minimum-wage-acs

Lee David and Emmanuel Saez 2012 ldquoOpimal Minimum Wage Policy in Compeiive LaborMarkesrdquo Journal o Public Economics 96 (9) 739ndash749

Neumark David and William Wascher 1992 ldquoEmploymen Effecs o Minimum and Subminimum Wages Panel Daa on Sae Minimum Wage Lawsrdquo Industrial and Labor Relations Review 46 (1)

55ndash81

Neumark David and William Wascher 2011 ldquoDoes a Higher Minimum Wage Enhance heEffeciveness o he Earned Income ax Credirdquo Industrial and Labor Relations Review 64 (4)712ndash746

Page Marianne Joanne Spez and Jane Millar 2005 ldquoDoes he Minimum Wage Affec WelareCaseloadsrdquo Journal o Policy Analysis and Management 24 (2) 273ndash295

Rohsein Jesse 2010 ldquoIs he EIC as Good as an NI Condiional Cash ransers and ax

Incidencerdquo American Economic Journal Economic Policy 2 (1) 177ndash208

Wage and Hours Division 2014 ldquoMinimum Wage Laws in he Saes ndash January 1 2014rdquo USDeparmen o Labor (htpwwwdolgovwhdminwageamericahm [February 2014])

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2543

22 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2643

Appendix A | wwwamericanprogresso

Appendix A

Pre-trend falsification check

Recen minimum wage research27 highlighs a common flaw in previous sudies

ailure o veriy ha he oucome variable is ree o negaive pre-exising rends I

or example SNAP aciviy was already rending down in saes ha raised heir

minimum wages beore hese changes came ino effec our regression analysis could

(misakenly) atribue ha reducion o he minimum wage We check or such

pre-rends by inroducing variables ha represen he prior yearrsquos value or leads o

he minimum wage I he model esimaes he minimum wage o have an effec on

he oucome variable beore he wage change wen ino effec hen an unobservedacor no he minimum wage change caused he change in SNAP aciviy

We es he specificaions above or pre-rends by including a one-year lead in

all hree specificaions We find ha he lead erms are small posiive and no

saisically significan indicaing ha he concurren minimum wage991252no

he wage level in prior periods991252is driving he observed changes in SNAP

oucomes28 In paricular he coefficien (sandard error) on he lead erm in

our preerred amily-level enrollmen regression is 011 and no significan

while he coefficien and sandard error o he conemporaneous minimum

wage is unchanged In he sae-level preerred enrollmen regression he

coefficien o he lead erm is again small (07) and i is no significan Te

corresponding coefficien on he lead erm in he sae-level expendiure

regression is 16 and is no significan Te posiive poin esimaes on hese lead

erms resuls no only rule ou disoring negaive pre-rends Tey also sugges

ha our main resuls may underesimae he rue effecs

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2743

24 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2843

Appendix B | wwwamericanprogresso

Appendix B

Model selection process

For boh he amily-level and sae-level models we es hree mehods o conrol

or unobserved geographic- and ime-varying characerisics as suggesed by he

minimum wage lieraure Firs we include only independen sae-specific fixed

effecs and year-specific fixed effecs Tis specificaion (specificaion 1) implicily

assumes ha amilies in any sae consiue an equally good saisical ldquoconrolrdquo

group or hose in any randomly chosen sae afer accouning or various

characerisics (median income and unemploymen rae among ohers) Similarly

simple ime fixed effecs assume ha amilies surveyed in any year can crediblyserve as a conrol group or amilies surveyed in every oher year o he sample

(1990 hrough 2012)

In oher words specificaion 1 assumes ha a saersquos immediae neighbor provides

no beter a couneracual or he effec o a minimum wage change han does a

sae across he counry We relax his resricive specificaion sequenially in wo

seps In specificaion 2 we replace simple year fixed effecs wih fixed effecs or

each Census divisionyear (capured as an addiional variable in he vecor By

using division-year effecs we remove he resricion ha amilies in each sae

are equally good saisical conrols or all oher amilies Raher we allow or he

possibiliy ha amilies in similar geographic regions (or example he Souh or

he Norheas) may be more similar o one anoher han amilies arher away

Finally in specificaion 3 we add sae-specific linear ime rends o he previous

specificaion Tus specificaion 3 is he mos rigorous model specificaion in ha

i allows or heerogeneiy along hree dimensions Ta is specificaion 3 allows

each sae o have is own ime-varying rends raher han imposing he resricion

ha saes evolve idenically over he 22 years in he sample

We begin building he heoreical specificaion above rom a se o simpleuncondiional models regression o SNAP aciviy (enrollmen or expendiures)

on he log o he minimum wage and a se o geographic- and ime-specific

effecs (specificaions 1 2 and 3 described above) As shown in ables 1ndash3 (or

specificaion 3) we hen add covariaes sequenially o hese models including

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2943

26 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

firs he vecor o amily-level conrols ollowed by each o several sae-level

covariaes in urn he unemploymen rae log o median-amily income and he

employmen-o-populaion raio Comparable resuls or specificaions 2 and 3

will be available in our orhcoming working paper

As expeced he simple uncondiional models indicae ha he relaionship beween he minimum wage and SNAP enrollmen i one exiss is a more complex

one influenced by oher acors In he uncondiional model he coefficien on he

variable o ineres991252he log o he minimum wage991252is small in magniude and no

saisically differen rom zero Once we accoun or he influence o labor marke

condiions and variaion in income levels on program paricipaion (by including

unemploymen rae and median-amily income conrol variables respecively)

he effec o he minimum wage on SNAP enrollmen is precisely esimaed Te

coefficien o he log minimum wage is slighly higher (-0042) in he amily-level

analysis han he coefficien (-031) in he sae-level analysis Te level o precision

is also higher in he amily-level analysis Tis is o be expeced when using 124million observaions compared o 1127

Te second dimension o model choice concerns he effec specificaion ables

7ndash9 compare he primary coefficiens o ineres or he SNAP enrollmen and

expendiure models For boh he enrollmen models he effec sizes are smalles

or specificaion 1 larges or specificaion 2 and inermediae beween hese wo

in specificaion 3 Recall ha Specificaion 3 conains sae-specific linear ime

rends in addiion o he census divisionyear conrols included in specificaion

2 In he amily-level enrollmen model he sandard error o he minimum wage

coefficien is smaller han in he oher wo specificaions Sandard errors on he

oher variables are much smaller in specificaions 2 and 3 han in specificaion

1 On he basis o coefficien significance (join and individual) specificaions 2

and 3 are sricly preerred in boh enrollmen models o specificaion 1 which

conains only sae and year fixed effecs

A concern wih specificaions 2 and 3 is ha rend conrols such as sae linear

rends may incorrecly absorb some o he delayed impac o a minimum wage

When we es his issue by including lagged minimum wages we do no find ha

delayed effecs are significan Anoher concern is ha more sauraed modelsuse less o he saisical variaion which could reduce he saisical power o

he resuls However he sandard errors or our more sauraed models are no

higher and are lower in some cases han or he less sauraed models Overall

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3043

Appendix B | wwwamericanprogresso

his evidence suppors our use o he sauraed model as he preerred model

specificaion Moreover Dubersquos 2013 sudy shows ha more sauraed models

perorm beter han models wih jus sae and ime fixed effecs

Te esimaed enrollmen regressions a boh he amily and sae levels show large

and saisically significan coefficiens Te esimaed minimum wage effec in heexpendiures regressions991252or which we have only sae-level daa991252is also large

and saisically significan

We do no use weighed regression or he sae-level models preerring o keep

analysis o he ldquoreamenrdquo (ha is o say a minimum wage change) appropriae

o he average sae raher han he average amily or individual I insead our

primary ineres were he impac o a minimum wage change on he average amily

or he average individual we migh choose o designae he number o amilies

in each sae or he sae populaion respecively as analyic weighs in order o

obain a coefficien beter suied or such inerence

TABLE 4

SNAP enrollment

Family-level linear probability

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0048 -0047 -0040 -0043 -0042

(0013) (0013) (001) (0008) (0009)

Unemployment rate (100) 0505 0420 0280

(0083) (0086) (0082)

Log median income -0057 -0039

(0011) (0011)

Employment-to-population ratio -0239

(0038)

N 1242022 1242022 1242022 1242022 1242022

plt01 plt005 plt001

Note Robust standard errors are in parentheses Observations are clustered at the state level The outcome variable is binary and equal to one if a family is enrolledin SNAP All models include state fixed effects Census division x-year fixed effects and state-specific linear time trends All specifications except 3a include additional

controls for family size race and marital status of the family head presence of children and presence of an adult maleSource Annual data from the March Current Population Survey for 1990 to 2012 estimation includes Current Population Survey probability weights

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3143

28 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

TABLE 5

SNAP enrollment

State-level linear regression

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0028 -0024 -0026 -0031 -0031

(0014) (0013) (0013) (0012) (0012)

Unemployment rate (100) 0625 0466 0320 0339

(0087) (0088) (0085) (0083)

Log median income -0090 -0065 -0061

(0013) (0013) (0013)

Employment-to-population ratio -0282 -0248

(0037) (0038)

N 1127 1127 1127 1127 1127

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the SNAP enrollment rate All models include state fixed effects Census division x-year fixedeffects and state-specific linear time trends

Source Annual data from the March Current Population Survey for 1990 to 2012

TABLE 6

SNAP expenditures

State-level linear regression

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0181 -0149 -0156 -0153 -0190

(011) (0103) (0102) (0103) (0103)

Unemployment rate (100) 4587 4152 4232 4313

(0622) (0621) (0633) (0628)

Log median income -0246 -0261 -0294

(0075) (0078) (0078)

Employment-to-population ratio 0155 0244

(0237) (024)

N 1127 1127 1127 1127 1127

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the natural log of state SNAP expenditures per capita for 1990 to 2012 All models include state

fixed effects Census division x-year fixed effects and state-specific linear time trends

Source Annual data from Bureau of Economic Analysis national income and product accounts tables for 1990 to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3243

Appendix B | wwwamericanprogresso

TABLE 7

Comparison of specifications SNAP enrollment

Family-level linear probability

(1) (2) (3)

Log minimum wage -0028 -0049 -0042

(0014) (0017) (0009)

Unemployment rate (100) 0275 0297 0280

(0161) (0076) (0082)

Log median income -0077 -0055 -0039

(0014) (0012) (0011)

Employment-to-population ratio -0238 -0250 -0239

(0054) (004) (0038)

N 1242022 1242022 1242022

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses Observations are clustered at the state level The outcomevariable is binary or equal to one if a family is enrolled in SNAP All specifications include additional controls forfamily size race and marital status of the family head presence of children and presence of an adult male

Source Annual data from the March Current Population Survey for 1990 to 2012 estimation includes CurrentPopulation Survey probability weights

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3343

30 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

TABLE 8

Comparison of specifications SNAP enrollment

State-level linear regression

(1) (2) (3)

Log minimum wage -0019 -0035 -0031

(0009) (0012) (0012)

Unemployment rate (100) 0401 0370 0339

(0063) (0077) (0083)

Log median income -0081 -0073 -0061

(0011) (0013) (0013)

Employment-to-population ratio -0183 -0222 -0248

(0039) (0039) (0038)

N 1127 1127 1127

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the SNAP enrollment rate Allregressions include the share of households with children and the racial shares of the population

Source Annual data from the March Current Population Survey for 1990 to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3443

Appendix B | wwwamericanprogresso

TABLE 9

Comparison of specifications SNAP expenditures

State-level linear regression

(1) (2) (3)

Log minimum wage -0121 -0203 -0190

(0075) (0103) (0103)

Unemployment rate (100) 5292 5152 4313

(0464) (0576) (0628)

Log median income -0437 -0417 -0294

(008) (0086) (0078)

Employment-to-population ratio -0040 -0220 0244

(0261) (0260) (0240)

N 1127 1127 1127

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the natural log of state SNAPexpenditures per capita All regressions include the share of households with children and the racial shares ofthe state population

Source Annual data from Bureau of Economic Analysis national income and product accounts tables for 1990to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3543

32 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3643

Appendix C | wwwamericanprogresso

Appendix C Harkin-Miller

policy simulation results

StateMinimum wage

(2014)

SNAP enrollment (2012) Predicted change in SNAP enrollment

RateTotal

(persons)

Model 1 family

enrollment (linearprobability)

Model 2 state

enrollment (linearregression)

Model 3 s

expenditu(linear regre

Alabama $725 164 788682 -66922 -58897 -58906

Alaska $775 120 87436 -8104 -6887 -3288

Arizona $790 201 1319323 -67435 -56738 -64356

Arkansas $725 78 230489 -40977 -36063 -35248

California $800 206 7813680 -371131 -310222 -18223

Colorado $800 164 853155 -50684 -42365 -23926

Connecticut $870 91 326621 -22456 -17975 -13711

Delaware $725 186 170262 -12739 -11211 -10647

District of Columbia $825 133 84009 -5370 -4417 -3632

Florida $793 166 3208026 -195813 -164426 -13046

Georgia $725 160 1586336 -137741 -121224 -11004

Hawaii $725 96 133662 -19310 -16995 -14933

Idaho $725 92 147501 -22165 -19507 -15809

Illinois $825 95 1225084 -109088 -89742 -70955

Indiana $725 125 816233 -90818 -79928 -83985

Iowa $725 155 478011 -42716 -37594 -28556

Kansas $725 135 388269 -40082 -35275 -27461

Kentucky $725 130 568821 -60840 -53544 -52259

Louisiana $725 149 683832 -63929 -56263 -66083

Maine $750 77 101976 -16567 -14323 -15234

Maryland $725 144 846415 -81748 -71946 -38370

Massachusetts $800 130 864721 -64902 -54251 -42913

Michigan $740 146 1439141 -128801 -112140 -11022

Minnesota $725 133 713646 -74730 -65769 -37878

TABLE 10

SNAP enrollments Predicted changes by state under the Harkin-Miller bill

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3743

34 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

StateMinimum wage

(2014)

SNAP enrollment (2012) Predicted change in SNAP enrollment

RateTotal

(persons)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 state

expenditures

(linear regression)

Mississippi $725 129 386501 -41486 -36511 -46467

Missouri $750 172 1036182 -75131 -64952 -56244

Montana $790 132 132452 -10350 -8708 -5846

Nebraska $725 124 230683 -25773 -22683 -12189

Nevada $825 162 446035 -23349 -19209 -11894

New Hampshire $725 127 168404 -18359 -16157 -5735

New Jersey $825 160 1416666 -75175 -61843 -28236

New Mexico $750 149 310896 -25983 -22463 -22512

New York $800 192 3763553 -191193 -159815 -142182

North Carolina $725 174 1697193 -135417 -119179 -113503

North Dakota $725 87 61225 -9743 -8574 -4021

Ohio $795 143 1647345 -115869 -97169 -88580

Oklahoma $725 129 494053 -53006 -46650 -46854

Oregon $910 124 485326 -17036 -13328 -16398

Pennsylvania $725 161 2053643 -177315 -156052 -125586

Rhode Island $800 156 163730 -10258 -8574 -8698

South Carolina $725 94 445277 -65614 -57746 -50304

South Dakota $725 208 173749 -11586 -10197 -7458

Tennessee $725 142 914903 -89667 -78915 -99134

Texas $725 110 2863779 -362018 -318607 -253285

Utah $725 88 251107 -39658 -34902 -19390

Vermont $873 156 97792 -3823 -3055 -2475

Virginia $725 101 829771 -113723 -100086 -58212

Washington $932 72 496934 -23221 -17947 -17756

West Virginia $725 58 107875 -25792 -22699 -21665

Wisconsin $725 75 427822 -79521 -69986 -53210

Wyoming $725 164 94590 -8010 -7050 -3104

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3843

Appendix C | wwwamericanprogresso

StateMinimum wage

(2014)

SNAP expenditures (2012)Predicted change in SNAP expenditures

(millions of dollars)

Per resident

(millions of

dollars)

Total

(millions of

dollars)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 s

expenditu

(linear regre

Alabama $725 $290 $1397 -$1185 -$1043 -$1043

Alaska $775 $253 $185 -$262 -$223 -$106

Arizona $790 $257 $1686 -$935 -$787 -$892

Arkansas $725 $248 $733 -$636 -$560 -$547

California $800 $189 $7164 -$7276 -$6082 -$3573

Colorado $800 $157 $816 -$862 -$721 -$407

Connecticut $870 $191 $686 -$343 -$275 -$210

Delaware $725 $250 $229 -$205 -$180 -$171

District of Columbia $825 $366 $232 -$146 -$120 -$99

Florida $793 $294 $5676 -$4429 -$3719 -$2951

Georgia $725 $317 $3140 -$2936 -$2584 -$2346

Hawaii $725 $335 $465 -$449 -$395 -$347

Idaho $725 $225 $359 -$376 -$331 -$268

Illinois $825 $249 $3200 -$2096 -$1725 -$1364

Indiana $725 $220 $1439 -$1162 -$1023 -$1075

Iowa $725 $192 $589 -$658 -$579 -$440

Kansas $725 $159 $460 -$502 -$441 -$344

Kentucky $725 $298 $1303 -$1133 -$997 -$973

Louisiana $725 $315 $1450 -$1047 -$922 -$1083

Maine $750 $281 $373 -$267 -$231 -$246

Maryland $725 $188 $1109 -$1765 -$1553 -$828

Massachusetts $800 $206 $1366 -$1030 -$861 -$681

Michigan $740 $300 $2963 -$2400 -$2090 -$2054

Minnesota $725 $140 $755 -$1113 -$980 -$564

Mississippi $725 $326 $973 -$649 -$571 -$726

Missouri $750 $241 $1452 -$1278 -$1104 -$956

Montana $790 $190 $191 -$179 -$151 -$101

Nebraska $725 $140 $259 -$409 -$360 -$194

Nevada $825 $191 $527 -$441 -$363 -$225

New Hampshire $725 $126 $167 -$399 -$351 -$125

TABLE 11

SNAP expenditures Predicted changes by state under the Harkin-Miller bill

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3943

36 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

StateMinimum wage

(2014)

SNAP expenditures (2012)Predicted change in SNAP expenditures

(millions of dollars)

Per resident

(millions of

dollars)

Total

(millions of

dollars)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 state

expenditures

(linear regression)

New Jersey $825 $160 $1420 -$1611 -$1325 -$605

New Mexico $750 $324 $675 -$513 -$444 -$445New York $800 $287 $5616 -$3766 -$3148 -$2801

North Carolina $725 $252 $2454 -$2187 -$1925 -$1833

North Dakota $725 $128 $90 -$162 -$143 -$67

Ohio $795 $259 $2995 -$2013 -$1688 -$1539

Oklahoma $725 $248 $945 -$799 -$703 -$706

Oregon $910 $322 $1255 -$272 -$213 -$262

Pennsylvania $725 $218 $2779 -$2930 -$2579 -$2075

Rhode Island $800 $280 $294 -$173 -$144 -$147

South Carolina $725 $291 $1373 -$1337 -$1177 -$1025South Dakota $725 $198 $165 -$192 -$169 -$123

Tennessee $725 $324 $2091 -$1413 -$1243 -$1562

Texas $725 $230 $5997 -$6402 -$5634 -$4479

Utah $725 $141 $402 -$614 -$541 -$300

Vermont $873 $230 $144 -$66 -$53 -$43

Virginia $725 $173 $1413 -$2062 -$1815 -$1056

Washington $932 $244 $1682 -$350 -$270 -$267

West Virginia $725 $273 $508 -$451 -$397 -$379

Wisconsin $725 $204 $1166 -$1302 -$1146 -$871Wyoming $725 $95 $55 -$105 -$93 -$41

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4043

Appendix C | wwwamericanprogresso

FIGURE 2

Raising the minimum wage to $1010would cut taxpayer costs in every state

Predicted decreases in cost and enrollment

in SNAP in 50 states

$200+$51ndash$100

$101ndash$200

0ndash$25

$26ndash$50

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4143

38 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

Endnotes

1 US Department of Agriculture ldquoSupplementalNutrition Assistance Program (SNAP)rdquo available athttpwwwfnsusdagovsnapeligibility (last accessedFebruary 2014)

2 For this initial analysis we do not consider Harkin-Millerrsquos increase in subminimum wages for tippedworkers To do s o would increase the estimated SNAP

savings by an unknown amount

3 The Congressional Budget Office estimates thatworkers currently earning between $1010 and $1150per hour would see their wages rise under the Harkin-Miller proposal Congressional Budget O ffice ldquoTheEffects of a Minimum Wage Increase on Employmentand Family Incomerdquo (2014)

4 Marianne Page Joanne Spetz and Jane Millar ldquoDoesthe Minimum Wage Affect Welfare Caseloadsrdquo Journalof Policy Analysis and Management 24 (2) (2005)273ndash295

5 Marianne Bitler and Hilary Hoynes ldquoThe More ThingsChange the More They Stay the Same The SafetyNet Living Arrangements and Poverty in the GreatRecessionrdquo NBER Working Paper 194 49 2013

6 Sylvia Allegretto and others ldquoFast Food Poverty Wages The Public Cost of Low-Wage Jobs in the Fast-FoodIndustryrdquo (Berkeley California Center for LaborResearch and Education 2013) available at httplaborcenterberkeleyedupubliccostsfast_food_poverty_wage

7 David Neumark and William Wascher ldquoDoes a HigherMinimum Wage Enhance the Effectiveness of theEarned Income Tax Creditrdquo Industrial and LaborRelations Review 64 (4) (2011) 712ndash746

8 David Lee and Emmanuel Saez ldquoOptimal MinimumWage Policy in Competitive Labor Marketsrdquo Journal ofPublic Economics 96 (9) (2012) 739ndash749

9 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family IncomerdquoFebruary 2014

10 Jesse Rothstein ldquoIs the EITC as Good as an NITConditional Cash Transfers and Tax Incidencerdquo AmericanEconomic Journal Economic Policy 2 (1) (2010) 177ndash208

11 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family Incomerdquo

12 Dube Arindrajit 2013 rdquoMinimum Wagesand the Distribution of Family IncomerdquoUnpublished working paper Available at httpsdldropboxusercontentcomu15038936Dube_MinimumWagesFamilyIncomespdf

13 As Dube explains in ldquoThe poverty of Minimum WageFactsrdquo the simulation approach underestimate stemsfrom a number of unwarranted assumptions includingthe range of actual wage increases and the accuracy ofwage data in the Current Population Survey The causal

approach does not make these assumptions

14 Allegretto Sylvia and others 2013 ldquoCredible ResearchDesigns for Minimum Wage Studiesrdquo Working Paper148-13 University of California Berkeley Institutefor Research on Labor and Employment Available athttpwwwirleberkeleyeduworkingpapers148-13pdf

15 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family In comerdquo

16 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo (2012)

17 Ibid

18 Ibid

19 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo

20 US Department of Agriculture ldquoSupplementalNutrition Assistance Program (SNAP)rdquo

21 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo

22 Ibid

23 We deviate from the Census Bureaursquos definition ofa family unit which is ldquotwo people or more (on e ofwhom is the householder) related by bir th marriageor adoption and residing togetherrdquo US Bureau ofthe Census ldquoCurrent Population Survey DefinitionsFamilyrdquo available at httpwwwcensusgovcpsabout

cpsdefhtml (last accessed February 2014) We countas a family unit any individual residing on his or herown two or more persons residing together whodo not belong to a family in the March CPS sampleare constructed as one family in our analysis For thepurposes of food stamp allocations the consumptionresulting from this transfer is probably distributed tofamily members (rather than household members ora single individual within the household) Howeversingle individuals canmdashand domdashreceive SNAPbenefits Excluding them would fail to make theanalysis reflective of the population at large

24 Strictly the family level linear probability modelpredicts the percentage-point decrease in theprobability that an individual family will receive SNAPpayments When applied to a large number of familieshowever we are able to interpret the coefficient asa decrease in the mean of enrollmentmdashthat is a

decrease in the enrollment ratemdashby applying the lawof iterated expectations

25 We generate expenditure predictions from theenrollment modelsmdashand conversely generateenrollment predictions from the expenditure modelmdashby assuming that expenditures per enrolled familyremains the same before and after the minimum wagechange In practice this is likely to be a conservativeestimatemdashthat is to underestimate the decrease inSNAP activity Average SNAP benefits per family willalso decrease as many families that remain eligible forSNAP experience income gains

26 Wage and Hour Division ldquoMinimum Wage Laws inthe States ndash Januar y 1 2014rdquo available at httpwwwdolgovwhdminwageamericahtm (last accessedFebruary 2014)

27 See for example Sylvia Allegretto and others ldquoCredibleResearch Designs for Minimum Wage Studiesrdquo WorkingPaper 148-113 (Berkeley California Institute forResearch on Labor and Employment 2013) available athttpwwwirleberkeleyeduworkingpapers148-13pdf

28 We will report these results in a forthcoming workingpaper

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4243

Endnotes | wwwamericanprogresso

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4343

The Center for American Progress is a nonpartisan research and educational institute

dedicated to promoting a strong just and free America that ensures opportunity

for all We believe that Americans are bound together by a common commitment to

these values and we aspire to ensure that our national policies reflect these values

We work to find progressive and pragmatic solutions to significant domestic and

international problems and develop policy proposals that foster a government that

is ldquoof the people by the people and for the peoplerdquo

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2043

Results | wwwamericanprogresso

In 2012 hese saes were home o 46 percen o he American populaion

and accouned or a roughly equivalen percenage o oal naional SNAP

expendiures No surprisingly he larges enrollmen decreases come rom

populous saes wih very high SNAP enrollmen raes andor very low minimum

wages Te larges prediced enrollmen reducion991252beween 319000 individuals

and 362000 individuals991252would occur in exas which has a $725 minimum wage In Caliornia which has a 206 percen SNAP paricipaion rae and an $8

per hour minimum wage we anicipae SNAP enrollmen declines o 310000

persons o 371000 persons And in Florida which had a SNAP paricipaion rae

o 166 percen and a $793 minimum wage enrollmen could decline by beween

164000 individuals and 196000 individuals For he our saes ha ogeher

accouned or he greaes amoun o SNAP spending in 2012991252exas Caliornia

Florida and New York respecively991252he combined expendiure reducion rom

he Harkin-Miller bill is prediced o be $14 billion

able 2 summarizes he prediced declines in SNAP aciviy or he naion as a whole ha would resul rom he direc and indirec effecs o he Harkin-Miller

bill Enrollmen would all beween 31 million persons and 36 million persons

represening 75 percen o 87 percen o curren enrollmen Te anicipaed

reducion in program expendiures would be nearly $46 billion or 61 percen o

program expendiures

TABLE 2

Comparison of national SNAP predictions under the Harkin-Miller billrsquos $1010 minimum wage

Model

Enrollment

(persons)

Expenditures

(millions of dollars)

Current (2012) Predicted Change Current (2012) Predicted Chan

Family enrollment (linear probability)

41866195

45489339 -3623144

$74861

NA NA

State enrollment (ordinary least squares) 38745435 -3120759 NA NA

State expenditures (ordinary least squares) NA NA $70305 -$45

Note Calculations use 2014 state minimum wages and the most recent SNAP data from 2012 They assume that per-enrollee expenditures remain constant

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2143

18 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

Tere are o course oher possibiliies or minimum wage legislaion able 3

shows he expeced SNAP changes or he Unied Saes under a variey o wage

scenarios calculaed using he sae-level models I saes were no able o se

heir minimum wages independenly such ha all saes were consrained by

he ederal minimum o $725 SNAP would be received by abou 514000 more

people across he Unied Saes a an addiional program cos o nearly hree-quarers o a billion dollars In conras he effecs o a higher minimum wage

proposal991252a ederal wage floor o $11 per hour991252would decrease enrollmen in

SNAP by more han 10 percen and decrease program coss by 83 percen

TABLE 3

Summary of par ticipation and expenditures under wage scenarios

If all states had

minimum wages of

Enrollment(persons)

Expenditures

(millions of dollars)

Predicted Change Predicted Change

Recent levels (2014) 41866195 $74861

$725 42380520 514326 $75604 $743

$800 41423919 -442276 $74209 -$652

$900 40148451 -1717744 $72350 -$2511

$1000 38872982 -2993212 $70490 -$4371

$1010 38745435 -3120759 $70305 -$4556

$1100 37597514 -4268681 $68631 -$6230

Note Calculations use state-level enrollment model coefficient

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2243

Conclusion | wwwamericanprogresso

Conclusion

An exensive body o lieraure examines employmen effecs o he minimum

wage A much smaller se o sudies examines how he minimum wage affecs

povery and only a handul o papers examine he effecs o he minimum wage

on he EIC Our analysis is he firs o examine he effecs o he minimum

wage on SNAP

Our findings indicae ha increased earnings rom minimum wage changes do

reduce SNAP enrollmens and expendiures We esimae ha he Harkin-Miller bill would save axpayers nearly $46 billion per year equivalen o 61 percen

o SNAP expendiures in 2012 he las year or which daa are available Over a

10-year period he esimaed savings amoun o nearly $46 billion

Our repor is subjec o limiaions ha we expec o overcome in our uure

research Firs he findings do no ake ino accoun possible ineracions among

SNAP he EIC and Medicaid Te eligibiliy cuoffs among hese programs

are quie differen suggesing ha such ineracions may be minor Noneheless

he join effecs can only be deermined by urher research using a causal

model Second i would be useul o know he disribuion o SNAP reducions

along he wage disribuion Using he Congressional Budge Officersquos calculaions

o how much he oal dollar value o wage would increase under he Harkin-

Miller proposal our findings imply ha he decline in overall SNAP spending

equals abou 15 percen o he oal resuling increase in wages Te amoun and

disribuion o his offse are o considerable ineres Minimum wage beneficiaries

who come rom working amilies already well above he povery line would no

see any offse while hose who are currenly considerably below he povery line

will see larger offses Tese issues will also be a subjec or our uure research

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2343

20 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

About the authors

Rachel West is a maser o public policy candidae a he Goldman School

o Public Policy Universiy o Caliornia Berkeley Her research ocuses on

economic policy in he areas o low-wage labor and povery

Michael Reich is proessor o economics and direcor o he Insiue or

Research on Labor and Employmen a he Universiy o Caliornia a Berkeley

His research publicaions cover numerous areas o labor economics including

racial inequaliy labor marke segmenaion high-perormance workplaces

union-managemen cooperaion Japanese labor-managemen sysems living

wages and minimum wages He received his docorae in economics rom

Harvard Universiy

Acknowledgments

We are graeul o Sylvia Allegreto Arindraji Dube Bill Leser Jesse Rohsein

Daniel Tompson and Ben Zipperer or heir valuable suggesions

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2443

References | wwwamericanprogresso

References

Allegreto Sylvia and ohers 2013 ldquoFas Food Povery Wages Te Public Cos o Low-Wage Jobsin he Fas-Food Indusryrdquo Universiy o Caliornia Berkeley Cener or Labor Research andEducaion Available a htplaborcenerberkeleyedupubliccossas_ood_povery_wage

Allegreto Sylvia and ohers 2013 ldquoCredible Research Designs or Minimum Wage Sudiesrdquo

Working Paper 148-13 Universiy o Caliornia Berkeley Insiue or Research on Labor andEmploymen Available a htpwwwirleberkeleyeduworkingpapers148-13pd

Biler Marianne and Hilary Hoynes 2013 ldquo Te More Tings Change he More Tey Say heSame Te Saey Ne Living Arrangemens and Povery in he Grea Recessionrdquo Working Paper19449 Naional Bureau o Economic Research

Congressional Budge Office 2012 ldquoTe Supplemenal Nuriion Assisance Programrdquo Washingon Available a htpwwwcbogovsiesdeaulfilescbofilesatachmens04-19-SNAPpd

991252 991252 991252 2014 ldquoTe Effec o a Minimum-Wage Increase on Employmen and Family Incomerdquo Washingon Available a htpwwwcbogovsiesdeaulfilescbofilesatachmens44995-MinimumWagepd

Dube Arindraji 2013 rdquoMinimum Wages and he Disribuion o Family Incomerdquo Unpublished working paper Available a htpsdldropboxuserconencomu15038936Dube_ MinimumWagesFamilyIncomespd

991252 991252 991252 2014 ldquoTe Povery o Minimum Wage Facsrdquo Arindraji Dube Blog January 22 Availablea htparindubecom20140122he-povery-o-minimum-wage-acs

Lee David and Emmanuel Saez 2012 ldquoOpimal Minimum Wage Policy in Compeiive LaborMarkesrdquo Journal o Public Economics 96 (9) 739ndash749

Neumark David and William Wascher 1992 ldquoEmploymen Effecs o Minimum and Subminimum Wages Panel Daa on Sae Minimum Wage Lawsrdquo Industrial and Labor Relations Review 46 (1)

55ndash81

Neumark David and William Wascher 2011 ldquoDoes a Higher Minimum Wage Enhance heEffeciveness o he Earned Income ax Credirdquo Industrial and Labor Relations Review 64 (4)712ndash746

Page Marianne Joanne Spez and Jane Millar 2005 ldquoDoes he Minimum Wage Affec WelareCaseloadsrdquo Journal o Policy Analysis and Management 24 (2) 273ndash295

Rohsein Jesse 2010 ldquoIs he EIC as Good as an NI Condiional Cash ransers and ax

Incidencerdquo American Economic Journal Economic Policy 2 (1) 177ndash208

Wage and Hours Division 2014 ldquoMinimum Wage Laws in he Saes ndash January 1 2014rdquo USDeparmen o Labor (htpwwwdolgovwhdminwageamericahm [February 2014])

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2543

22 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2643

Appendix A | wwwamericanprogresso

Appendix A

Pre-trend falsification check

Recen minimum wage research27 highlighs a common flaw in previous sudies

ailure o veriy ha he oucome variable is ree o negaive pre-exising rends I

or example SNAP aciviy was already rending down in saes ha raised heir

minimum wages beore hese changes came ino effec our regression analysis could

(misakenly) atribue ha reducion o he minimum wage We check or such

pre-rends by inroducing variables ha represen he prior yearrsquos value or leads o

he minimum wage I he model esimaes he minimum wage o have an effec on

he oucome variable beore he wage change wen ino effec hen an unobservedacor no he minimum wage change caused he change in SNAP aciviy

We es he specificaions above or pre-rends by including a one-year lead in

all hree specificaions We find ha he lead erms are small posiive and no

saisically significan indicaing ha he concurren minimum wage991252no

he wage level in prior periods991252is driving he observed changes in SNAP

oucomes28 In paricular he coefficien (sandard error) on he lead erm in

our preerred amily-level enrollmen regression is 011 and no significan

while he coefficien and sandard error o he conemporaneous minimum

wage is unchanged In he sae-level preerred enrollmen regression he

coefficien o he lead erm is again small (07) and i is no significan Te

corresponding coefficien on he lead erm in he sae-level expendiure

regression is 16 and is no significan Te posiive poin esimaes on hese lead

erms resuls no only rule ou disoring negaive pre-rends Tey also sugges

ha our main resuls may underesimae he rue effecs

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2743

24 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2843

Appendix B | wwwamericanprogresso

Appendix B

Model selection process

For boh he amily-level and sae-level models we es hree mehods o conrol

or unobserved geographic- and ime-varying characerisics as suggesed by he

minimum wage lieraure Firs we include only independen sae-specific fixed

effecs and year-specific fixed effecs Tis specificaion (specificaion 1) implicily

assumes ha amilies in any sae consiue an equally good saisical ldquoconrolrdquo

group or hose in any randomly chosen sae afer accouning or various

characerisics (median income and unemploymen rae among ohers) Similarly

simple ime fixed effecs assume ha amilies surveyed in any year can crediblyserve as a conrol group or amilies surveyed in every oher year o he sample

(1990 hrough 2012)

In oher words specificaion 1 assumes ha a saersquos immediae neighbor provides

no beter a couneracual or he effec o a minimum wage change han does a

sae across he counry We relax his resricive specificaion sequenially in wo

seps In specificaion 2 we replace simple year fixed effecs wih fixed effecs or

each Census divisionyear (capured as an addiional variable in he vecor By

using division-year effecs we remove he resricion ha amilies in each sae

are equally good saisical conrols or all oher amilies Raher we allow or he

possibiliy ha amilies in similar geographic regions (or example he Souh or

he Norheas) may be more similar o one anoher han amilies arher away

Finally in specificaion 3 we add sae-specific linear ime rends o he previous

specificaion Tus specificaion 3 is he mos rigorous model specificaion in ha

i allows or heerogeneiy along hree dimensions Ta is specificaion 3 allows

each sae o have is own ime-varying rends raher han imposing he resricion

ha saes evolve idenically over he 22 years in he sample

We begin building he heoreical specificaion above rom a se o simpleuncondiional models regression o SNAP aciviy (enrollmen or expendiures)

on he log o he minimum wage and a se o geographic- and ime-specific

effecs (specificaions 1 2 and 3 described above) As shown in ables 1ndash3 (or

specificaion 3) we hen add covariaes sequenially o hese models including

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2943

26 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

firs he vecor o amily-level conrols ollowed by each o several sae-level

covariaes in urn he unemploymen rae log o median-amily income and he

employmen-o-populaion raio Comparable resuls or specificaions 2 and 3

will be available in our orhcoming working paper

As expeced he simple uncondiional models indicae ha he relaionship beween he minimum wage and SNAP enrollmen i one exiss is a more complex

one influenced by oher acors In he uncondiional model he coefficien on he

variable o ineres991252he log o he minimum wage991252is small in magniude and no

saisically differen rom zero Once we accoun or he influence o labor marke

condiions and variaion in income levels on program paricipaion (by including

unemploymen rae and median-amily income conrol variables respecively)

he effec o he minimum wage on SNAP enrollmen is precisely esimaed Te

coefficien o he log minimum wage is slighly higher (-0042) in he amily-level

analysis han he coefficien (-031) in he sae-level analysis Te level o precision

is also higher in he amily-level analysis Tis is o be expeced when using 124million observaions compared o 1127

Te second dimension o model choice concerns he effec specificaion ables

7ndash9 compare he primary coefficiens o ineres or he SNAP enrollmen and

expendiure models For boh he enrollmen models he effec sizes are smalles

or specificaion 1 larges or specificaion 2 and inermediae beween hese wo

in specificaion 3 Recall ha Specificaion 3 conains sae-specific linear ime

rends in addiion o he census divisionyear conrols included in specificaion

2 In he amily-level enrollmen model he sandard error o he minimum wage

coefficien is smaller han in he oher wo specificaions Sandard errors on he

oher variables are much smaller in specificaions 2 and 3 han in specificaion

1 On he basis o coefficien significance (join and individual) specificaions 2

and 3 are sricly preerred in boh enrollmen models o specificaion 1 which

conains only sae and year fixed effecs

A concern wih specificaions 2 and 3 is ha rend conrols such as sae linear

rends may incorrecly absorb some o he delayed impac o a minimum wage

When we es his issue by including lagged minimum wages we do no find ha

delayed effecs are significan Anoher concern is ha more sauraed modelsuse less o he saisical variaion which could reduce he saisical power o

he resuls However he sandard errors or our more sauraed models are no

higher and are lower in some cases han or he less sauraed models Overall

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3043

Appendix B | wwwamericanprogresso

his evidence suppors our use o he sauraed model as he preerred model

specificaion Moreover Dubersquos 2013 sudy shows ha more sauraed models

perorm beter han models wih jus sae and ime fixed effecs

Te esimaed enrollmen regressions a boh he amily and sae levels show large

and saisically significan coefficiens Te esimaed minimum wage effec in heexpendiures regressions991252or which we have only sae-level daa991252is also large

and saisically significan

We do no use weighed regression or he sae-level models preerring o keep

analysis o he ldquoreamenrdquo (ha is o say a minimum wage change) appropriae

o he average sae raher han he average amily or individual I insead our

primary ineres were he impac o a minimum wage change on he average amily

or he average individual we migh choose o designae he number o amilies

in each sae or he sae populaion respecively as analyic weighs in order o

obain a coefficien beter suied or such inerence

TABLE 4

SNAP enrollment

Family-level linear probability

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0048 -0047 -0040 -0043 -0042

(0013) (0013) (001) (0008) (0009)

Unemployment rate (100) 0505 0420 0280

(0083) (0086) (0082)

Log median income -0057 -0039

(0011) (0011)

Employment-to-population ratio -0239

(0038)

N 1242022 1242022 1242022 1242022 1242022

plt01 plt005 plt001

Note Robust standard errors are in parentheses Observations are clustered at the state level The outcome variable is binary and equal to one if a family is enrolledin SNAP All models include state fixed effects Census division x-year fixed effects and state-specific linear time trends All specifications except 3a include additional

controls for family size race and marital status of the family head presence of children and presence of an adult maleSource Annual data from the March Current Population Survey for 1990 to 2012 estimation includes Current Population Survey probability weights

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3143

28 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

TABLE 5

SNAP enrollment

State-level linear regression

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0028 -0024 -0026 -0031 -0031

(0014) (0013) (0013) (0012) (0012)

Unemployment rate (100) 0625 0466 0320 0339

(0087) (0088) (0085) (0083)

Log median income -0090 -0065 -0061

(0013) (0013) (0013)

Employment-to-population ratio -0282 -0248

(0037) (0038)

N 1127 1127 1127 1127 1127

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the SNAP enrollment rate All models include state fixed effects Census division x-year fixedeffects and state-specific linear time trends

Source Annual data from the March Current Population Survey for 1990 to 2012

TABLE 6

SNAP expenditures

State-level linear regression

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0181 -0149 -0156 -0153 -0190

(011) (0103) (0102) (0103) (0103)

Unemployment rate (100) 4587 4152 4232 4313

(0622) (0621) (0633) (0628)

Log median income -0246 -0261 -0294

(0075) (0078) (0078)

Employment-to-population ratio 0155 0244

(0237) (024)

N 1127 1127 1127 1127 1127

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the natural log of state SNAP expenditures per capita for 1990 to 2012 All models include state

fixed effects Census division x-year fixed effects and state-specific linear time trends

Source Annual data from Bureau of Economic Analysis national income and product accounts tables for 1990 to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3243

Appendix B | wwwamericanprogresso

TABLE 7

Comparison of specifications SNAP enrollment

Family-level linear probability

(1) (2) (3)

Log minimum wage -0028 -0049 -0042

(0014) (0017) (0009)

Unemployment rate (100) 0275 0297 0280

(0161) (0076) (0082)

Log median income -0077 -0055 -0039

(0014) (0012) (0011)

Employment-to-population ratio -0238 -0250 -0239

(0054) (004) (0038)

N 1242022 1242022 1242022

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses Observations are clustered at the state level The outcomevariable is binary or equal to one if a family is enrolled in SNAP All specifications include additional controls forfamily size race and marital status of the family head presence of children and presence of an adult male

Source Annual data from the March Current Population Survey for 1990 to 2012 estimation includes CurrentPopulation Survey probability weights

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3343

30 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

TABLE 8

Comparison of specifications SNAP enrollment

State-level linear regression

(1) (2) (3)

Log minimum wage -0019 -0035 -0031

(0009) (0012) (0012)

Unemployment rate (100) 0401 0370 0339

(0063) (0077) (0083)

Log median income -0081 -0073 -0061

(0011) (0013) (0013)

Employment-to-population ratio -0183 -0222 -0248

(0039) (0039) (0038)

N 1127 1127 1127

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the SNAP enrollment rate Allregressions include the share of households with children and the racial shares of the population

Source Annual data from the March Current Population Survey for 1990 to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3443

Appendix B | wwwamericanprogresso

TABLE 9

Comparison of specifications SNAP expenditures

State-level linear regression

(1) (2) (3)

Log minimum wage -0121 -0203 -0190

(0075) (0103) (0103)

Unemployment rate (100) 5292 5152 4313

(0464) (0576) (0628)

Log median income -0437 -0417 -0294

(008) (0086) (0078)

Employment-to-population ratio -0040 -0220 0244

(0261) (0260) (0240)

N 1127 1127 1127

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the natural log of state SNAPexpenditures per capita All regressions include the share of households with children and the racial shares ofthe state population

Source Annual data from Bureau of Economic Analysis national income and product accounts tables for 1990to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3543

32 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3643

Appendix C | wwwamericanprogresso

Appendix C Harkin-Miller

policy simulation results

StateMinimum wage

(2014)

SNAP enrollment (2012) Predicted change in SNAP enrollment

RateTotal

(persons)

Model 1 family

enrollment (linearprobability)

Model 2 state

enrollment (linearregression)

Model 3 s

expenditu(linear regre

Alabama $725 164 788682 -66922 -58897 -58906

Alaska $775 120 87436 -8104 -6887 -3288

Arizona $790 201 1319323 -67435 -56738 -64356

Arkansas $725 78 230489 -40977 -36063 -35248

California $800 206 7813680 -371131 -310222 -18223

Colorado $800 164 853155 -50684 -42365 -23926

Connecticut $870 91 326621 -22456 -17975 -13711

Delaware $725 186 170262 -12739 -11211 -10647

District of Columbia $825 133 84009 -5370 -4417 -3632

Florida $793 166 3208026 -195813 -164426 -13046

Georgia $725 160 1586336 -137741 -121224 -11004

Hawaii $725 96 133662 -19310 -16995 -14933

Idaho $725 92 147501 -22165 -19507 -15809

Illinois $825 95 1225084 -109088 -89742 -70955

Indiana $725 125 816233 -90818 -79928 -83985

Iowa $725 155 478011 -42716 -37594 -28556

Kansas $725 135 388269 -40082 -35275 -27461

Kentucky $725 130 568821 -60840 -53544 -52259

Louisiana $725 149 683832 -63929 -56263 -66083

Maine $750 77 101976 -16567 -14323 -15234

Maryland $725 144 846415 -81748 -71946 -38370

Massachusetts $800 130 864721 -64902 -54251 -42913

Michigan $740 146 1439141 -128801 -112140 -11022

Minnesota $725 133 713646 -74730 -65769 -37878

TABLE 10

SNAP enrollments Predicted changes by state under the Harkin-Miller bill

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3743

34 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

StateMinimum wage

(2014)

SNAP enrollment (2012) Predicted change in SNAP enrollment

RateTotal

(persons)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 state

expenditures

(linear regression)

Mississippi $725 129 386501 -41486 -36511 -46467

Missouri $750 172 1036182 -75131 -64952 -56244

Montana $790 132 132452 -10350 -8708 -5846

Nebraska $725 124 230683 -25773 -22683 -12189

Nevada $825 162 446035 -23349 -19209 -11894

New Hampshire $725 127 168404 -18359 -16157 -5735

New Jersey $825 160 1416666 -75175 -61843 -28236

New Mexico $750 149 310896 -25983 -22463 -22512

New York $800 192 3763553 -191193 -159815 -142182

North Carolina $725 174 1697193 -135417 -119179 -113503

North Dakota $725 87 61225 -9743 -8574 -4021

Ohio $795 143 1647345 -115869 -97169 -88580

Oklahoma $725 129 494053 -53006 -46650 -46854

Oregon $910 124 485326 -17036 -13328 -16398

Pennsylvania $725 161 2053643 -177315 -156052 -125586

Rhode Island $800 156 163730 -10258 -8574 -8698

South Carolina $725 94 445277 -65614 -57746 -50304

South Dakota $725 208 173749 -11586 -10197 -7458

Tennessee $725 142 914903 -89667 -78915 -99134

Texas $725 110 2863779 -362018 -318607 -253285

Utah $725 88 251107 -39658 -34902 -19390

Vermont $873 156 97792 -3823 -3055 -2475

Virginia $725 101 829771 -113723 -100086 -58212

Washington $932 72 496934 -23221 -17947 -17756

West Virginia $725 58 107875 -25792 -22699 -21665

Wisconsin $725 75 427822 -79521 -69986 -53210

Wyoming $725 164 94590 -8010 -7050 -3104

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3843

Appendix C | wwwamericanprogresso

StateMinimum wage

(2014)

SNAP expenditures (2012)Predicted change in SNAP expenditures

(millions of dollars)

Per resident

(millions of

dollars)

Total

(millions of

dollars)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 s

expenditu

(linear regre

Alabama $725 $290 $1397 -$1185 -$1043 -$1043

Alaska $775 $253 $185 -$262 -$223 -$106

Arizona $790 $257 $1686 -$935 -$787 -$892

Arkansas $725 $248 $733 -$636 -$560 -$547

California $800 $189 $7164 -$7276 -$6082 -$3573

Colorado $800 $157 $816 -$862 -$721 -$407

Connecticut $870 $191 $686 -$343 -$275 -$210

Delaware $725 $250 $229 -$205 -$180 -$171

District of Columbia $825 $366 $232 -$146 -$120 -$99

Florida $793 $294 $5676 -$4429 -$3719 -$2951

Georgia $725 $317 $3140 -$2936 -$2584 -$2346

Hawaii $725 $335 $465 -$449 -$395 -$347

Idaho $725 $225 $359 -$376 -$331 -$268

Illinois $825 $249 $3200 -$2096 -$1725 -$1364

Indiana $725 $220 $1439 -$1162 -$1023 -$1075

Iowa $725 $192 $589 -$658 -$579 -$440

Kansas $725 $159 $460 -$502 -$441 -$344

Kentucky $725 $298 $1303 -$1133 -$997 -$973

Louisiana $725 $315 $1450 -$1047 -$922 -$1083

Maine $750 $281 $373 -$267 -$231 -$246

Maryland $725 $188 $1109 -$1765 -$1553 -$828

Massachusetts $800 $206 $1366 -$1030 -$861 -$681

Michigan $740 $300 $2963 -$2400 -$2090 -$2054

Minnesota $725 $140 $755 -$1113 -$980 -$564

Mississippi $725 $326 $973 -$649 -$571 -$726

Missouri $750 $241 $1452 -$1278 -$1104 -$956

Montana $790 $190 $191 -$179 -$151 -$101

Nebraska $725 $140 $259 -$409 -$360 -$194

Nevada $825 $191 $527 -$441 -$363 -$225

New Hampshire $725 $126 $167 -$399 -$351 -$125

TABLE 11

SNAP expenditures Predicted changes by state under the Harkin-Miller bill

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3943

36 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

StateMinimum wage

(2014)

SNAP expenditures (2012)Predicted change in SNAP expenditures

(millions of dollars)

Per resident

(millions of

dollars)

Total

(millions of

dollars)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 state

expenditures

(linear regression)

New Jersey $825 $160 $1420 -$1611 -$1325 -$605

New Mexico $750 $324 $675 -$513 -$444 -$445New York $800 $287 $5616 -$3766 -$3148 -$2801

North Carolina $725 $252 $2454 -$2187 -$1925 -$1833

North Dakota $725 $128 $90 -$162 -$143 -$67

Ohio $795 $259 $2995 -$2013 -$1688 -$1539

Oklahoma $725 $248 $945 -$799 -$703 -$706

Oregon $910 $322 $1255 -$272 -$213 -$262

Pennsylvania $725 $218 $2779 -$2930 -$2579 -$2075

Rhode Island $800 $280 $294 -$173 -$144 -$147

South Carolina $725 $291 $1373 -$1337 -$1177 -$1025South Dakota $725 $198 $165 -$192 -$169 -$123

Tennessee $725 $324 $2091 -$1413 -$1243 -$1562

Texas $725 $230 $5997 -$6402 -$5634 -$4479

Utah $725 $141 $402 -$614 -$541 -$300

Vermont $873 $230 $144 -$66 -$53 -$43

Virginia $725 $173 $1413 -$2062 -$1815 -$1056

Washington $932 $244 $1682 -$350 -$270 -$267

West Virginia $725 $273 $508 -$451 -$397 -$379

Wisconsin $725 $204 $1166 -$1302 -$1146 -$871Wyoming $725 $95 $55 -$105 -$93 -$41

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4043

Appendix C | wwwamericanprogresso

FIGURE 2

Raising the minimum wage to $1010would cut taxpayer costs in every state

Predicted decreases in cost and enrollment

in SNAP in 50 states

$200+$51ndash$100

$101ndash$200

0ndash$25

$26ndash$50

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4143

38 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

Endnotes

1 US Department of Agriculture ldquoSupplementalNutrition Assistance Program (SNAP)rdquo available athttpwwwfnsusdagovsnapeligibility (last accessedFebruary 2014)

2 For this initial analysis we do not consider Harkin-Millerrsquos increase in subminimum wages for tippedworkers To do s o would increase the estimated SNAP

savings by an unknown amount

3 The Congressional Budget Office estimates thatworkers currently earning between $1010 and $1150per hour would see their wages rise under the Harkin-Miller proposal Congressional Budget O ffice ldquoTheEffects of a Minimum Wage Increase on Employmentand Family Incomerdquo (2014)

4 Marianne Page Joanne Spetz and Jane Millar ldquoDoesthe Minimum Wage Affect Welfare Caseloadsrdquo Journalof Policy Analysis and Management 24 (2) (2005)273ndash295

5 Marianne Bitler and Hilary Hoynes ldquoThe More ThingsChange the More They Stay the Same The SafetyNet Living Arrangements and Poverty in the GreatRecessionrdquo NBER Working Paper 194 49 2013

6 Sylvia Allegretto and others ldquoFast Food Poverty Wages The Public Cost of Low-Wage Jobs in the Fast-FoodIndustryrdquo (Berkeley California Center for LaborResearch and Education 2013) available at httplaborcenterberkeleyedupubliccostsfast_food_poverty_wage

7 David Neumark and William Wascher ldquoDoes a HigherMinimum Wage Enhance the Effectiveness of theEarned Income Tax Creditrdquo Industrial and LaborRelations Review 64 (4) (2011) 712ndash746

8 David Lee and Emmanuel Saez ldquoOptimal MinimumWage Policy in Competitive Labor Marketsrdquo Journal ofPublic Economics 96 (9) (2012) 739ndash749

9 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family IncomerdquoFebruary 2014

10 Jesse Rothstein ldquoIs the EITC as Good as an NITConditional Cash Transfers and Tax Incidencerdquo AmericanEconomic Journal Economic Policy 2 (1) (2010) 177ndash208

11 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family Incomerdquo

12 Dube Arindrajit 2013 rdquoMinimum Wagesand the Distribution of Family IncomerdquoUnpublished working paper Available at httpsdldropboxusercontentcomu15038936Dube_MinimumWagesFamilyIncomespdf

13 As Dube explains in ldquoThe poverty of Minimum WageFactsrdquo the simulation approach underestimate stemsfrom a number of unwarranted assumptions includingthe range of actual wage increases and the accuracy ofwage data in the Current Population Survey The causal

approach does not make these assumptions

14 Allegretto Sylvia and others 2013 ldquoCredible ResearchDesigns for Minimum Wage Studiesrdquo Working Paper148-13 University of California Berkeley Institutefor Research on Labor and Employment Available athttpwwwirleberkeleyeduworkingpapers148-13pdf

15 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family In comerdquo

16 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo (2012)

17 Ibid

18 Ibid

19 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo

20 US Department of Agriculture ldquoSupplementalNutrition Assistance Program (SNAP)rdquo

21 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo

22 Ibid

23 We deviate from the Census Bureaursquos definition ofa family unit which is ldquotwo people or more (on e ofwhom is the householder) related by bir th marriageor adoption and residing togetherrdquo US Bureau ofthe Census ldquoCurrent Population Survey DefinitionsFamilyrdquo available at httpwwwcensusgovcpsabout

cpsdefhtml (last accessed February 2014) We countas a family unit any individual residing on his or herown two or more persons residing together whodo not belong to a family in the March CPS sampleare constructed as one family in our analysis For thepurposes of food stamp allocations the consumptionresulting from this transfer is probably distributed tofamily members (rather than household members ora single individual within the household) Howeversingle individuals canmdashand domdashreceive SNAPbenefits Excluding them would fail to make theanalysis reflective of the population at large

24 Strictly the family level linear probability modelpredicts the percentage-point decrease in theprobability that an individual family will receive SNAPpayments When applied to a large number of familieshowever we are able to interpret the coefficient asa decrease in the mean of enrollmentmdashthat is a

decrease in the enrollment ratemdashby applying the lawof iterated expectations

25 We generate expenditure predictions from theenrollment modelsmdashand conversely generateenrollment predictions from the expenditure modelmdashby assuming that expenditures per enrolled familyremains the same before and after the minimum wagechange In practice this is likely to be a conservativeestimatemdashthat is to underestimate the decrease inSNAP activity Average SNAP benefits per family willalso decrease as many families that remain eligible forSNAP experience income gains

26 Wage and Hour Division ldquoMinimum Wage Laws inthe States ndash Januar y 1 2014rdquo available at httpwwwdolgovwhdminwageamericahtm (last accessedFebruary 2014)

27 See for example Sylvia Allegretto and others ldquoCredibleResearch Designs for Minimum Wage Studiesrdquo WorkingPaper 148-113 (Berkeley California Institute forResearch on Labor and Employment 2013) available athttpwwwirleberkeleyeduworkingpapers148-13pdf

28 We will report these results in a forthcoming workingpaper

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4243

Endnotes | wwwamericanprogresso

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4343

The Center for American Progress is a nonpartisan research and educational institute

dedicated to promoting a strong just and free America that ensures opportunity

for all We believe that Americans are bound together by a common commitment to

these values and we aspire to ensure that our national policies reflect these values

We work to find progressive and pragmatic solutions to significant domestic and

international problems and develop policy proposals that foster a government that

is ldquoof the people by the people and for the peoplerdquo

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2143

18 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

Tere are o course oher possibiliies or minimum wage legislaion able 3

shows he expeced SNAP changes or he Unied Saes under a variey o wage

scenarios calculaed using he sae-level models I saes were no able o se

heir minimum wages independenly such ha all saes were consrained by

he ederal minimum o $725 SNAP would be received by abou 514000 more

people across he Unied Saes a an addiional program cos o nearly hree-quarers o a billion dollars In conras he effecs o a higher minimum wage

proposal991252a ederal wage floor o $11 per hour991252would decrease enrollmen in

SNAP by more han 10 percen and decrease program coss by 83 percen

TABLE 3

Summary of par ticipation and expenditures under wage scenarios

If all states had

minimum wages of

Enrollment(persons)

Expenditures

(millions of dollars)

Predicted Change Predicted Change

Recent levels (2014) 41866195 $74861

$725 42380520 514326 $75604 $743

$800 41423919 -442276 $74209 -$652

$900 40148451 -1717744 $72350 -$2511

$1000 38872982 -2993212 $70490 -$4371

$1010 38745435 -3120759 $70305 -$4556

$1100 37597514 -4268681 $68631 -$6230

Note Calculations use state-level enrollment model coefficient

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2243

Conclusion | wwwamericanprogresso

Conclusion

An exensive body o lieraure examines employmen effecs o he minimum

wage A much smaller se o sudies examines how he minimum wage affecs

povery and only a handul o papers examine he effecs o he minimum wage

on he EIC Our analysis is he firs o examine he effecs o he minimum

wage on SNAP

Our findings indicae ha increased earnings rom minimum wage changes do

reduce SNAP enrollmens and expendiures We esimae ha he Harkin-Miller bill would save axpayers nearly $46 billion per year equivalen o 61 percen

o SNAP expendiures in 2012 he las year or which daa are available Over a

10-year period he esimaed savings amoun o nearly $46 billion

Our repor is subjec o limiaions ha we expec o overcome in our uure

research Firs he findings do no ake ino accoun possible ineracions among

SNAP he EIC and Medicaid Te eligibiliy cuoffs among hese programs

are quie differen suggesing ha such ineracions may be minor Noneheless

he join effecs can only be deermined by urher research using a causal

model Second i would be useul o know he disribuion o SNAP reducions

along he wage disribuion Using he Congressional Budge Officersquos calculaions

o how much he oal dollar value o wage would increase under he Harkin-

Miller proposal our findings imply ha he decline in overall SNAP spending

equals abou 15 percen o he oal resuling increase in wages Te amoun and

disribuion o his offse are o considerable ineres Minimum wage beneficiaries

who come rom working amilies already well above he povery line would no

see any offse while hose who are currenly considerably below he povery line

will see larger offses Tese issues will also be a subjec or our uure research

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2343

20 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

About the authors

Rachel West is a maser o public policy candidae a he Goldman School

o Public Policy Universiy o Caliornia Berkeley Her research ocuses on

economic policy in he areas o low-wage labor and povery

Michael Reich is proessor o economics and direcor o he Insiue or

Research on Labor and Employmen a he Universiy o Caliornia a Berkeley

His research publicaions cover numerous areas o labor economics including

racial inequaliy labor marke segmenaion high-perormance workplaces

union-managemen cooperaion Japanese labor-managemen sysems living

wages and minimum wages He received his docorae in economics rom

Harvard Universiy

Acknowledgments

We are graeul o Sylvia Allegreto Arindraji Dube Bill Leser Jesse Rohsein

Daniel Tompson and Ben Zipperer or heir valuable suggesions

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2443

References | wwwamericanprogresso

References

Allegreto Sylvia and ohers 2013 ldquoFas Food Povery Wages Te Public Cos o Low-Wage Jobsin he Fas-Food Indusryrdquo Universiy o Caliornia Berkeley Cener or Labor Research andEducaion Available a htplaborcenerberkeleyedupubliccossas_ood_povery_wage

Allegreto Sylvia and ohers 2013 ldquoCredible Research Designs or Minimum Wage Sudiesrdquo

Working Paper 148-13 Universiy o Caliornia Berkeley Insiue or Research on Labor andEmploymen Available a htpwwwirleberkeleyeduworkingpapers148-13pd

Biler Marianne and Hilary Hoynes 2013 ldquo Te More Tings Change he More Tey Say heSame Te Saey Ne Living Arrangemens and Povery in he Grea Recessionrdquo Working Paper19449 Naional Bureau o Economic Research

Congressional Budge Office 2012 ldquoTe Supplemenal Nuriion Assisance Programrdquo Washingon Available a htpwwwcbogovsiesdeaulfilescbofilesatachmens04-19-SNAPpd

991252 991252 991252 2014 ldquoTe Effec o a Minimum-Wage Increase on Employmen and Family Incomerdquo Washingon Available a htpwwwcbogovsiesdeaulfilescbofilesatachmens44995-MinimumWagepd

Dube Arindraji 2013 rdquoMinimum Wages and he Disribuion o Family Incomerdquo Unpublished working paper Available a htpsdldropboxuserconencomu15038936Dube_ MinimumWagesFamilyIncomespd

991252 991252 991252 2014 ldquoTe Povery o Minimum Wage Facsrdquo Arindraji Dube Blog January 22 Availablea htparindubecom20140122he-povery-o-minimum-wage-acs

Lee David and Emmanuel Saez 2012 ldquoOpimal Minimum Wage Policy in Compeiive LaborMarkesrdquo Journal o Public Economics 96 (9) 739ndash749

Neumark David and William Wascher 1992 ldquoEmploymen Effecs o Minimum and Subminimum Wages Panel Daa on Sae Minimum Wage Lawsrdquo Industrial and Labor Relations Review 46 (1)

55ndash81

Neumark David and William Wascher 2011 ldquoDoes a Higher Minimum Wage Enhance heEffeciveness o he Earned Income ax Credirdquo Industrial and Labor Relations Review 64 (4)712ndash746

Page Marianne Joanne Spez and Jane Millar 2005 ldquoDoes he Minimum Wage Affec WelareCaseloadsrdquo Journal o Policy Analysis and Management 24 (2) 273ndash295

Rohsein Jesse 2010 ldquoIs he EIC as Good as an NI Condiional Cash ransers and ax

Incidencerdquo American Economic Journal Economic Policy 2 (1) 177ndash208

Wage and Hours Division 2014 ldquoMinimum Wage Laws in he Saes ndash January 1 2014rdquo USDeparmen o Labor (htpwwwdolgovwhdminwageamericahm [February 2014])

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2543

22 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2643

Appendix A | wwwamericanprogresso

Appendix A

Pre-trend falsification check

Recen minimum wage research27 highlighs a common flaw in previous sudies

ailure o veriy ha he oucome variable is ree o negaive pre-exising rends I

or example SNAP aciviy was already rending down in saes ha raised heir

minimum wages beore hese changes came ino effec our regression analysis could

(misakenly) atribue ha reducion o he minimum wage We check or such

pre-rends by inroducing variables ha represen he prior yearrsquos value or leads o

he minimum wage I he model esimaes he minimum wage o have an effec on

he oucome variable beore he wage change wen ino effec hen an unobservedacor no he minimum wage change caused he change in SNAP aciviy

We es he specificaions above or pre-rends by including a one-year lead in

all hree specificaions We find ha he lead erms are small posiive and no

saisically significan indicaing ha he concurren minimum wage991252no

he wage level in prior periods991252is driving he observed changes in SNAP

oucomes28 In paricular he coefficien (sandard error) on he lead erm in

our preerred amily-level enrollmen regression is 011 and no significan

while he coefficien and sandard error o he conemporaneous minimum

wage is unchanged In he sae-level preerred enrollmen regression he

coefficien o he lead erm is again small (07) and i is no significan Te

corresponding coefficien on he lead erm in he sae-level expendiure

regression is 16 and is no significan Te posiive poin esimaes on hese lead

erms resuls no only rule ou disoring negaive pre-rends Tey also sugges

ha our main resuls may underesimae he rue effecs

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2743

24 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2843

Appendix B | wwwamericanprogresso

Appendix B

Model selection process

For boh he amily-level and sae-level models we es hree mehods o conrol

or unobserved geographic- and ime-varying characerisics as suggesed by he

minimum wage lieraure Firs we include only independen sae-specific fixed

effecs and year-specific fixed effecs Tis specificaion (specificaion 1) implicily

assumes ha amilies in any sae consiue an equally good saisical ldquoconrolrdquo

group or hose in any randomly chosen sae afer accouning or various

characerisics (median income and unemploymen rae among ohers) Similarly

simple ime fixed effecs assume ha amilies surveyed in any year can crediblyserve as a conrol group or amilies surveyed in every oher year o he sample

(1990 hrough 2012)

In oher words specificaion 1 assumes ha a saersquos immediae neighbor provides

no beter a couneracual or he effec o a minimum wage change han does a

sae across he counry We relax his resricive specificaion sequenially in wo

seps In specificaion 2 we replace simple year fixed effecs wih fixed effecs or

each Census divisionyear (capured as an addiional variable in he vecor By

using division-year effecs we remove he resricion ha amilies in each sae

are equally good saisical conrols or all oher amilies Raher we allow or he

possibiliy ha amilies in similar geographic regions (or example he Souh or

he Norheas) may be more similar o one anoher han amilies arher away

Finally in specificaion 3 we add sae-specific linear ime rends o he previous

specificaion Tus specificaion 3 is he mos rigorous model specificaion in ha

i allows or heerogeneiy along hree dimensions Ta is specificaion 3 allows

each sae o have is own ime-varying rends raher han imposing he resricion

ha saes evolve idenically over he 22 years in he sample

We begin building he heoreical specificaion above rom a se o simpleuncondiional models regression o SNAP aciviy (enrollmen or expendiures)

on he log o he minimum wage and a se o geographic- and ime-specific

effecs (specificaions 1 2 and 3 described above) As shown in ables 1ndash3 (or

specificaion 3) we hen add covariaes sequenially o hese models including

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2943

26 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

firs he vecor o amily-level conrols ollowed by each o several sae-level

covariaes in urn he unemploymen rae log o median-amily income and he

employmen-o-populaion raio Comparable resuls or specificaions 2 and 3

will be available in our orhcoming working paper

As expeced he simple uncondiional models indicae ha he relaionship beween he minimum wage and SNAP enrollmen i one exiss is a more complex

one influenced by oher acors In he uncondiional model he coefficien on he

variable o ineres991252he log o he minimum wage991252is small in magniude and no

saisically differen rom zero Once we accoun or he influence o labor marke

condiions and variaion in income levels on program paricipaion (by including

unemploymen rae and median-amily income conrol variables respecively)

he effec o he minimum wage on SNAP enrollmen is precisely esimaed Te

coefficien o he log minimum wage is slighly higher (-0042) in he amily-level

analysis han he coefficien (-031) in he sae-level analysis Te level o precision

is also higher in he amily-level analysis Tis is o be expeced when using 124million observaions compared o 1127

Te second dimension o model choice concerns he effec specificaion ables

7ndash9 compare he primary coefficiens o ineres or he SNAP enrollmen and

expendiure models For boh he enrollmen models he effec sizes are smalles

or specificaion 1 larges or specificaion 2 and inermediae beween hese wo

in specificaion 3 Recall ha Specificaion 3 conains sae-specific linear ime

rends in addiion o he census divisionyear conrols included in specificaion

2 In he amily-level enrollmen model he sandard error o he minimum wage

coefficien is smaller han in he oher wo specificaions Sandard errors on he

oher variables are much smaller in specificaions 2 and 3 han in specificaion

1 On he basis o coefficien significance (join and individual) specificaions 2

and 3 are sricly preerred in boh enrollmen models o specificaion 1 which

conains only sae and year fixed effecs

A concern wih specificaions 2 and 3 is ha rend conrols such as sae linear

rends may incorrecly absorb some o he delayed impac o a minimum wage

When we es his issue by including lagged minimum wages we do no find ha

delayed effecs are significan Anoher concern is ha more sauraed modelsuse less o he saisical variaion which could reduce he saisical power o

he resuls However he sandard errors or our more sauraed models are no

higher and are lower in some cases han or he less sauraed models Overall

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3043

Appendix B | wwwamericanprogresso

his evidence suppors our use o he sauraed model as he preerred model

specificaion Moreover Dubersquos 2013 sudy shows ha more sauraed models

perorm beter han models wih jus sae and ime fixed effecs

Te esimaed enrollmen regressions a boh he amily and sae levels show large

and saisically significan coefficiens Te esimaed minimum wage effec in heexpendiures regressions991252or which we have only sae-level daa991252is also large

and saisically significan

We do no use weighed regression or he sae-level models preerring o keep

analysis o he ldquoreamenrdquo (ha is o say a minimum wage change) appropriae

o he average sae raher han he average amily or individual I insead our

primary ineres were he impac o a minimum wage change on he average amily

or he average individual we migh choose o designae he number o amilies

in each sae or he sae populaion respecively as analyic weighs in order o

obain a coefficien beter suied or such inerence

TABLE 4

SNAP enrollment

Family-level linear probability

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0048 -0047 -0040 -0043 -0042

(0013) (0013) (001) (0008) (0009)

Unemployment rate (100) 0505 0420 0280

(0083) (0086) (0082)

Log median income -0057 -0039

(0011) (0011)

Employment-to-population ratio -0239

(0038)

N 1242022 1242022 1242022 1242022 1242022

plt01 plt005 plt001

Note Robust standard errors are in parentheses Observations are clustered at the state level The outcome variable is binary and equal to one if a family is enrolledin SNAP All models include state fixed effects Census division x-year fixed effects and state-specific linear time trends All specifications except 3a include additional

controls for family size race and marital status of the family head presence of children and presence of an adult maleSource Annual data from the March Current Population Survey for 1990 to 2012 estimation includes Current Population Survey probability weights

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3143

28 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

TABLE 5

SNAP enrollment

State-level linear regression

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0028 -0024 -0026 -0031 -0031

(0014) (0013) (0013) (0012) (0012)

Unemployment rate (100) 0625 0466 0320 0339

(0087) (0088) (0085) (0083)

Log median income -0090 -0065 -0061

(0013) (0013) (0013)

Employment-to-population ratio -0282 -0248

(0037) (0038)

N 1127 1127 1127 1127 1127

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the SNAP enrollment rate All models include state fixed effects Census division x-year fixedeffects and state-specific linear time trends

Source Annual data from the March Current Population Survey for 1990 to 2012

TABLE 6

SNAP expenditures

State-level linear regression

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0181 -0149 -0156 -0153 -0190

(011) (0103) (0102) (0103) (0103)

Unemployment rate (100) 4587 4152 4232 4313

(0622) (0621) (0633) (0628)

Log median income -0246 -0261 -0294

(0075) (0078) (0078)

Employment-to-population ratio 0155 0244

(0237) (024)

N 1127 1127 1127 1127 1127

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the natural log of state SNAP expenditures per capita for 1990 to 2012 All models include state

fixed effects Census division x-year fixed effects and state-specific linear time trends

Source Annual data from Bureau of Economic Analysis national income and product accounts tables for 1990 to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3243

Appendix B | wwwamericanprogresso

TABLE 7

Comparison of specifications SNAP enrollment

Family-level linear probability

(1) (2) (3)

Log minimum wage -0028 -0049 -0042

(0014) (0017) (0009)

Unemployment rate (100) 0275 0297 0280

(0161) (0076) (0082)

Log median income -0077 -0055 -0039

(0014) (0012) (0011)

Employment-to-population ratio -0238 -0250 -0239

(0054) (004) (0038)

N 1242022 1242022 1242022

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses Observations are clustered at the state level The outcomevariable is binary or equal to one if a family is enrolled in SNAP All specifications include additional controls forfamily size race and marital status of the family head presence of children and presence of an adult male

Source Annual data from the March Current Population Survey for 1990 to 2012 estimation includes CurrentPopulation Survey probability weights

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3343

30 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

TABLE 8

Comparison of specifications SNAP enrollment

State-level linear regression

(1) (2) (3)

Log minimum wage -0019 -0035 -0031

(0009) (0012) (0012)

Unemployment rate (100) 0401 0370 0339

(0063) (0077) (0083)

Log median income -0081 -0073 -0061

(0011) (0013) (0013)

Employment-to-population ratio -0183 -0222 -0248

(0039) (0039) (0038)

N 1127 1127 1127

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the SNAP enrollment rate Allregressions include the share of households with children and the racial shares of the population

Source Annual data from the March Current Population Survey for 1990 to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3443

Appendix B | wwwamericanprogresso

TABLE 9

Comparison of specifications SNAP expenditures

State-level linear regression

(1) (2) (3)

Log minimum wage -0121 -0203 -0190

(0075) (0103) (0103)

Unemployment rate (100) 5292 5152 4313

(0464) (0576) (0628)

Log median income -0437 -0417 -0294

(008) (0086) (0078)

Employment-to-population ratio -0040 -0220 0244

(0261) (0260) (0240)

N 1127 1127 1127

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the natural log of state SNAPexpenditures per capita All regressions include the share of households with children and the racial shares ofthe state population

Source Annual data from Bureau of Economic Analysis national income and product accounts tables for 1990to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3543

32 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3643

Appendix C | wwwamericanprogresso

Appendix C Harkin-Miller

policy simulation results

StateMinimum wage

(2014)

SNAP enrollment (2012) Predicted change in SNAP enrollment

RateTotal

(persons)

Model 1 family

enrollment (linearprobability)

Model 2 state

enrollment (linearregression)

Model 3 s

expenditu(linear regre

Alabama $725 164 788682 -66922 -58897 -58906

Alaska $775 120 87436 -8104 -6887 -3288

Arizona $790 201 1319323 -67435 -56738 -64356

Arkansas $725 78 230489 -40977 -36063 -35248

California $800 206 7813680 -371131 -310222 -18223

Colorado $800 164 853155 -50684 -42365 -23926

Connecticut $870 91 326621 -22456 -17975 -13711

Delaware $725 186 170262 -12739 -11211 -10647

District of Columbia $825 133 84009 -5370 -4417 -3632

Florida $793 166 3208026 -195813 -164426 -13046

Georgia $725 160 1586336 -137741 -121224 -11004

Hawaii $725 96 133662 -19310 -16995 -14933

Idaho $725 92 147501 -22165 -19507 -15809

Illinois $825 95 1225084 -109088 -89742 -70955

Indiana $725 125 816233 -90818 -79928 -83985

Iowa $725 155 478011 -42716 -37594 -28556

Kansas $725 135 388269 -40082 -35275 -27461

Kentucky $725 130 568821 -60840 -53544 -52259

Louisiana $725 149 683832 -63929 -56263 -66083

Maine $750 77 101976 -16567 -14323 -15234

Maryland $725 144 846415 -81748 -71946 -38370

Massachusetts $800 130 864721 -64902 -54251 -42913

Michigan $740 146 1439141 -128801 -112140 -11022

Minnesota $725 133 713646 -74730 -65769 -37878

TABLE 10

SNAP enrollments Predicted changes by state under the Harkin-Miller bill

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3743

34 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

StateMinimum wage

(2014)

SNAP enrollment (2012) Predicted change in SNAP enrollment

RateTotal

(persons)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 state

expenditures

(linear regression)

Mississippi $725 129 386501 -41486 -36511 -46467

Missouri $750 172 1036182 -75131 -64952 -56244

Montana $790 132 132452 -10350 -8708 -5846

Nebraska $725 124 230683 -25773 -22683 -12189

Nevada $825 162 446035 -23349 -19209 -11894

New Hampshire $725 127 168404 -18359 -16157 -5735

New Jersey $825 160 1416666 -75175 -61843 -28236

New Mexico $750 149 310896 -25983 -22463 -22512

New York $800 192 3763553 -191193 -159815 -142182

North Carolina $725 174 1697193 -135417 -119179 -113503

North Dakota $725 87 61225 -9743 -8574 -4021

Ohio $795 143 1647345 -115869 -97169 -88580

Oklahoma $725 129 494053 -53006 -46650 -46854

Oregon $910 124 485326 -17036 -13328 -16398

Pennsylvania $725 161 2053643 -177315 -156052 -125586

Rhode Island $800 156 163730 -10258 -8574 -8698

South Carolina $725 94 445277 -65614 -57746 -50304

South Dakota $725 208 173749 -11586 -10197 -7458

Tennessee $725 142 914903 -89667 -78915 -99134

Texas $725 110 2863779 -362018 -318607 -253285

Utah $725 88 251107 -39658 -34902 -19390

Vermont $873 156 97792 -3823 -3055 -2475

Virginia $725 101 829771 -113723 -100086 -58212

Washington $932 72 496934 -23221 -17947 -17756

West Virginia $725 58 107875 -25792 -22699 -21665

Wisconsin $725 75 427822 -79521 -69986 -53210

Wyoming $725 164 94590 -8010 -7050 -3104

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3843

Appendix C | wwwamericanprogresso

StateMinimum wage

(2014)

SNAP expenditures (2012)Predicted change in SNAP expenditures

(millions of dollars)

Per resident

(millions of

dollars)

Total

(millions of

dollars)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 s

expenditu

(linear regre

Alabama $725 $290 $1397 -$1185 -$1043 -$1043

Alaska $775 $253 $185 -$262 -$223 -$106

Arizona $790 $257 $1686 -$935 -$787 -$892

Arkansas $725 $248 $733 -$636 -$560 -$547

California $800 $189 $7164 -$7276 -$6082 -$3573

Colorado $800 $157 $816 -$862 -$721 -$407

Connecticut $870 $191 $686 -$343 -$275 -$210

Delaware $725 $250 $229 -$205 -$180 -$171

District of Columbia $825 $366 $232 -$146 -$120 -$99

Florida $793 $294 $5676 -$4429 -$3719 -$2951

Georgia $725 $317 $3140 -$2936 -$2584 -$2346

Hawaii $725 $335 $465 -$449 -$395 -$347

Idaho $725 $225 $359 -$376 -$331 -$268

Illinois $825 $249 $3200 -$2096 -$1725 -$1364

Indiana $725 $220 $1439 -$1162 -$1023 -$1075

Iowa $725 $192 $589 -$658 -$579 -$440

Kansas $725 $159 $460 -$502 -$441 -$344

Kentucky $725 $298 $1303 -$1133 -$997 -$973

Louisiana $725 $315 $1450 -$1047 -$922 -$1083

Maine $750 $281 $373 -$267 -$231 -$246

Maryland $725 $188 $1109 -$1765 -$1553 -$828

Massachusetts $800 $206 $1366 -$1030 -$861 -$681

Michigan $740 $300 $2963 -$2400 -$2090 -$2054

Minnesota $725 $140 $755 -$1113 -$980 -$564

Mississippi $725 $326 $973 -$649 -$571 -$726

Missouri $750 $241 $1452 -$1278 -$1104 -$956

Montana $790 $190 $191 -$179 -$151 -$101

Nebraska $725 $140 $259 -$409 -$360 -$194

Nevada $825 $191 $527 -$441 -$363 -$225

New Hampshire $725 $126 $167 -$399 -$351 -$125

TABLE 11

SNAP expenditures Predicted changes by state under the Harkin-Miller bill

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3943

36 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

StateMinimum wage

(2014)

SNAP expenditures (2012)Predicted change in SNAP expenditures

(millions of dollars)

Per resident

(millions of

dollars)

Total

(millions of

dollars)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 state

expenditures

(linear regression)

New Jersey $825 $160 $1420 -$1611 -$1325 -$605

New Mexico $750 $324 $675 -$513 -$444 -$445New York $800 $287 $5616 -$3766 -$3148 -$2801

North Carolina $725 $252 $2454 -$2187 -$1925 -$1833

North Dakota $725 $128 $90 -$162 -$143 -$67

Ohio $795 $259 $2995 -$2013 -$1688 -$1539

Oklahoma $725 $248 $945 -$799 -$703 -$706

Oregon $910 $322 $1255 -$272 -$213 -$262

Pennsylvania $725 $218 $2779 -$2930 -$2579 -$2075

Rhode Island $800 $280 $294 -$173 -$144 -$147

South Carolina $725 $291 $1373 -$1337 -$1177 -$1025South Dakota $725 $198 $165 -$192 -$169 -$123

Tennessee $725 $324 $2091 -$1413 -$1243 -$1562

Texas $725 $230 $5997 -$6402 -$5634 -$4479

Utah $725 $141 $402 -$614 -$541 -$300

Vermont $873 $230 $144 -$66 -$53 -$43

Virginia $725 $173 $1413 -$2062 -$1815 -$1056

Washington $932 $244 $1682 -$350 -$270 -$267

West Virginia $725 $273 $508 -$451 -$397 -$379

Wisconsin $725 $204 $1166 -$1302 -$1146 -$871Wyoming $725 $95 $55 -$105 -$93 -$41

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4043

Appendix C | wwwamericanprogresso

FIGURE 2

Raising the minimum wage to $1010would cut taxpayer costs in every state

Predicted decreases in cost and enrollment

in SNAP in 50 states

$200+$51ndash$100

$101ndash$200

0ndash$25

$26ndash$50

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4143

38 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

Endnotes

1 US Department of Agriculture ldquoSupplementalNutrition Assistance Program (SNAP)rdquo available athttpwwwfnsusdagovsnapeligibility (last accessedFebruary 2014)

2 For this initial analysis we do not consider Harkin-Millerrsquos increase in subminimum wages for tippedworkers To do s o would increase the estimated SNAP

savings by an unknown amount

3 The Congressional Budget Office estimates thatworkers currently earning between $1010 and $1150per hour would see their wages rise under the Harkin-Miller proposal Congressional Budget O ffice ldquoTheEffects of a Minimum Wage Increase on Employmentand Family Incomerdquo (2014)

4 Marianne Page Joanne Spetz and Jane Millar ldquoDoesthe Minimum Wage Affect Welfare Caseloadsrdquo Journalof Policy Analysis and Management 24 (2) (2005)273ndash295

5 Marianne Bitler and Hilary Hoynes ldquoThe More ThingsChange the More They Stay the Same The SafetyNet Living Arrangements and Poverty in the GreatRecessionrdquo NBER Working Paper 194 49 2013

6 Sylvia Allegretto and others ldquoFast Food Poverty Wages The Public Cost of Low-Wage Jobs in the Fast-FoodIndustryrdquo (Berkeley California Center for LaborResearch and Education 2013) available at httplaborcenterberkeleyedupubliccostsfast_food_poverty_wage

7 David Neumark and William Wascher ldquoDoes a HigherMinimum Wage Enhance the Effectiveness of theEarned Income Tax Creditrdquo Industrial and LaborRelations Review 64 (4) (2011) 712ndash746

8 David Lee and Emmanuel Saez ldquoOptimal MinimumWage Policy in Competitive Labor Marketsrdquo Journal ofPublic Economics 96 (9) (2012) 739ndash749

9 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family IncomerdquoFebruary 2014

10 Jesse Rothstein ldquoIs the EITC as Good as an NITConditional Cash Transfers and Tax Incidencerdquo AmericanEconomic Journal Economic Policy 2 (1) (2010) 177ndash208

11 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family Incomerdquo

12 Dube Arindrajit 2013 rdquoMinimum Wagesand the Distribution of Family IncomerdquoUnpublished working paper Available at httpsdldropboxusercontentcomu15038936Dube_MinimumWagesFamilyIncomespdf

13 As Dube explains in ldquoThe poverty of Minimum WageFactsrdquo the simulation approach underestimate stemsfrom a number of unwarranted assumptions includingthe range of actual wage increases and the accuracy ofwage data in the Current Population Survey The causal

approach does not make these assumptions

14 Allegretto Sylvia and others 2013 ldquoCredible ResearchDesigns for Minimum Wage Studiesrdquo Working Paper148-13 University of California Berkeley Institutefor Research on Labor and Employment Available athttpwwwirleberkeleyeduworkingpapers148-13pdf

15 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family In comerdquo

16 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo (2012)

17 Ibid

18 Ibid

19 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo

20 US Department of Agriculture ldquoSupplementalNutrition Assistance Program (SNAP)rdquo

21 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo

22 Ibid

23 We deviate from the Census Bureaursquos definition ofa family unit which is ldquotwo people or more (on e ofwhom is the householder) related by bir th marriageor adoption and residing togetherrdquo US Bureau ofthe Census ldquoCurrent Population Survey DefinitionsFamilyrdquo available at httpwwwcensusgovcpsabout

cpsdefhtml (last accessed February 2014) We countas a family unit any individual residing on his or herown two or more persons residing together whodo not belong to a family in the March CPS sampleare constructed as one family in our analysis For thepurposes of food stamp allocations the consumptionresulting from this transfer is probably distributed tofamily members (rather than household members ora single individual within the household) Howeversingle individuals canmdashand domdashreceive SNAPbenefits Excluding them would fail to make theanalysis reflective of the population at large

24 Strictly the family level linear probability modelpredicts the percentage-point decrease in theprobability that an individual family will receive SNAPpayments When applied to a large number of familieshowever we are able to interpret the coefficient asa decrease in the mean of enrollmentmdashthat is a

decrease in the enrollment ratemdashby applying the lawof iterated expectations

25 We generate expenditure predictions from theenrollment modelsmdashand conversely generateenrollment predictions from the expenditure modelmdashby assuming that expenditures per enrolled familyremains the same before and after the minimum wagechange In practice this is likely to be a conservativeestimatemdashthat is to underestimate the decrease inSNAP activity Average SNAP benefits per family willalso decrease as many families that remain eligible forSNAP experience income gains

26 Wage and Hour Division ldquoMinimum Wage Laws inthe States ndash Januar y 1 2014rdquo available at httpwwwdolgovwhdminwageamericahtm (last accessedFebruary 2014)

27 See for example Sylvia Allegretto and others ldquoCredibleResearch Designs for Minimum Wage Studiesrdquo WorkingPaper 148-113 (Berkeley California Institute forResearch on Labor and Employment 2013) available athttpwwwirleberkeleyeduworkingpapers148-13pdf

28 We will report these results in a forthcoming workingpaper

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4243

Endnotes | wwwamericanprogresso

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4343

The Center for American Progress is a nonpartisan research and educational institute

dedicated to promoting a strong just and free America that ensures opportunity

for all We believe that Americans are bound together by a common commitment to

these values and we aspire to ensure that our national policies reflect these values

We work to find progressive and pragmatic solutions to significant domestic and

international problems and develop policy proposals that foster a government that

is ldquoof the people by the people and for the peoplerdquo

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2243

Conclusion | wwwamericanprogresso

Conclusion

An exensive body o lieraure examines employmen effecs o he minimum

wage A much smaller se o sudies examines how he minimum wage affecs

povery and only a handul o papers examine he effecs o he minimum wage

on he EIC Our analysis is he firs o examine he effecs o he minimum

wage on SNAP

Our findings indicae ha increased earnings rom minimum wage changes do

reduce SNAP enrollmens and expendiures We esimae ha he Harkin-Miller bill would save axpayers nearly $46 billion per year equivalen o 61 percen

o SNAP expendiures in 2012 he las year or which daa are available Over a

10-year period he esimaed savings amoun o nearly $46 billion

Our repor is subjec o limiaions ha we expec o overcome in our uure

research Firs he findings do no ake ino accoun possible ineracions among

SNAP he EIC and Medicaid Te eligibiliy cuoffs among hese programs

are quie differen suggesing ha such ineracions may be minor Noneheless

he join effecs can only be deermined by urher research using a causal

model Second i would be useul o know he disribuion o SNAP reducions

along he wage disribuion Using he Congressional Budge Officersquos calculaions

o how much he oal dollar value o wage would increase under he Harkin-

Miller proposal our findings imply ha he decline in overall SNAP spending

equals abou 15 percen o he oal resuling increase in wages Te amoun and

disribuion o his offse are o considerable ineres Minimum wage beneficiaries

who come rom working amilies already well above he povery line would no

see any offse while hose who are currenly considerably below he povery line

will see larger offses Tese issues will also be a subjec or our uure research

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2343

20 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

About the authors

Rachel West is a maser o public policy candidae a he Goldman School

o Public Policy Universiy o Caliornia Berkeley Her research ocuses on

economic policy in he areas o low-wage labor and povery

Michael Reich is proessor o economics and direcor o he Insiue or

Research on Labor and Employmen a he Universiy o Caliornia a Berkeley

His research publicaions cover numerous areas o labor economics including

racial inequaliy labor marke segmenaion high-perormance workplaces

union-managemen cooperaion Japanese labor-managemen sysems living

wages and minimum wages He received his docorae in economics rom

Harvard Universiy

Acknowledgments

We are graeul o Sylvia Allegreto Arindraji Dube Bill Leser Jesse Rohsein

Daniel Tompson and Ben Zipperer or heir valuable suggesions

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2443

References | wwwamericanprogresso

References

Allegreto Sylvia and ohers 2013 ldquoFas Food Povery Wages Te Public Cos o Low-Wage Jobsin he Fas-Food Indusryrdquo Universiy o Caliornia Berkeley Cener or Labor Research andEducaion Available a htplaborcenerberkeleyedupubliccossas_ood_povery_wage

Allegreto Sylvia and ohers 2013 ldquoCredible Research Designs or Minimum Wage Sudiesrdquo

Working Paper 148-13 Universiy o Caliornia Berkeley Insiue or Research on Labor andEmploymen Available a htpwwwirleberkeleyeduworkingpapers148-13pd

Biler Marianne and Hilary Hoynes 2013 ldquo Te More Tings Change he More Tey Say heSame Te Saey Ne Living Arrangemens and Povery in he Grea Recessionrdquo Working Paper19449 Naional Bureau o Economic Research

Congressional Budge Office 2012 ldquoTe Supplemenal Nuriion Assisance Programrdquo Washingon Available a htpwwwcbogovsiesdeaulfilescbofilesatachmens04-19-SNAPpd

991252 991252 991252 2014 ldquoTe Effec o a Minimum-Wage Increase on Employmen and Family Incomerdquo Washingon Available a htpwwwcbogovsiesdeaulfilescbofilesatachmens44995-MinimumWagepd

Dube Arindraji 2013 rdquoMinimum Wages and he Disribuion o Family Incomerdquo Unpublished working paper Available a htpsdldropboxuserconencomu15038936Dube_ MinimumWagesFamilyIncomespd

991252 991252 991252 2014 ldquoTe Povery o Minimum Wage Facsrdquo Arindraji Dube Blog January 22 Availablea htparindubecom20140122he-povery-o-minimum-wage-acs

Lee David and Emmanuel Saez 2012 ldquoOpimal Minimum Wage Policy in Compeiive LaborMarkesrdquo Journal o Public Economics 96 (9) 739ndash749

Neumark David and William Wascher 1992 ldquoEmploymen Effecs o Minimum and Subminimum Wages Panel Daa on Sae Minimum Wage Lawsrdquo Industrial and Labor Relations Review 46 (1)

55ndash81

Neumark David and William Wascher 2011 ldquoDoes a Higher Minimum Wage Enhance heEffeciveness o he Earned Income ax Credirdquo Industrial and Labor Relations Review 64 (4)712ndash746

Page Marianne Joanne Spez and Jane Millar 2005 ldquoDoes he Minimum Wage Affec WelareCaseloadsrdquo Journal o Policy Analysis and Management 24 (2) 273ndash295

Rohsein Jesse 2010 ldquoIs he EIC as Good as an NI Condiional Cash ransers and ax

Incidencerdquo American Economic Journal Economic Policy 2 (1) 177ndash208

Wage and Hours Division 2014 ldquoMinimum Wage Laws in he Saes ndash January 1 2014rdquo USDeparmen o Labor (htpwwwdolgovwhdminwageamericahm [February 2014])

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2543

22 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2643

Appendix A | wwwamericanprogresso

Appendix A

Pre-trend falsification check

Recen minimum wage research27 highlighs a common flaw in previous sudies

ailure o veriy ha he oucome variable is ree o negaive pre-exising rends I

or example SNAP aciviy was already rending down in saes ha raised heir

minimum wages beore hese changes came ino effec our regression analysis could

(misakenly) atribue ha reducion o he minimum wage We check or such

pre-rends by inroducing variables ha represen he prior yearrsquos value or leads o

he minimum wage I he model esimaes he minimum wage o have an effec on

he oucome variable beore he wage change wen ino effec hen an unobservedacor no he minimum wage change caused he change in SNAP aciviy

We es he specificaions above or pre-rends by including a one-year lead in

all hree specificaions We find ha he lead erms are small posiive and no

saisically significan indicaing ha he concurren minimum wage991252no

he wage level in prior periods991252is driving he observed changes in SNAP

oucomes28 In paricular he coefficien (sandard error) on he lead erm in

our preerred amily-level enrollmen regression is 011 and no significan

while he coefficien and sandard error o he conemporaneous minimum

wage is unchanged In he sae-level preerred enrollmen regression he

coefficien o he lead erm is again small (07) and i is no significan Te

corresponding coefficien on he lead erm in he sae-level expendiure

regression is 16 and is no significan Te posiive poin esimaes on hese lead

erms resuls no only rule ou disoring negaive pre-rends Tey also sugges

ha our main resuls may underesimae he rue effecs

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2743

24 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2843

Appendix B | wwwamericanprogresso

Appendix B

Model selection process

For boh he amily-level and sae-level models we es hree mehods o conrol

or unobserved geographic- and ime-varying characerisics as suggesed by he

minimum wage lieraure Firs we include only independen sae-specific fixed

effecs and year-specific fixed effecs Tis specificaion (specificaion 1) implicily

assumes ha amilies in any sae consiue an equally good saisical ldquoconrolrdquo

group or hose in any randomly chosen sae afer accouning or various

characerisics (median income and unemploymen rae among ohers) Similarly

simple ime fixed effecs assume ha amilies surveyed in any year can crediblyserve as a conrol group or amilies surveyed in every oher year o he sample

(1990 hrough 2012)

In oher words specificaion 1 assumes ha a saersquos immediae neighbor provides

no beter a couneracual or he effec o a minimum wage change han does a

sae across he counry We relax his resricive specificaion sequenially in wo

seps In specificaion 2 we replace simple year fixed effecs wih fixed effecs or

each Census divisionyear (capured as an addiional variable in he vecor By

using division-year effecs we remove he resricion ha amilies in each sae

are equally good saisical conrols or all oher amilies Raher we allow or he

possibiliy ha amilies in similar geographic regions (or example he Souh or

he Norheas) may be more similar o one anoher han amilies arher away

Finally in specificaion 3 we add sae-specific linear ime rends o he previous

specificaion Tus specificaion 3 is he mos rigorous model specificaion in ha

i allows or heerogeneiy along hree dimensions Ta is specificaion 3 allows

each sae o have is own ime-varying rends raher han imposing he resricion

ha saes evolve idenically over he 22 years in he sample

We begin building he heoreical specificaion above rom a se o simpleuncondiional models regression o SNAP aciviy (enrollmen or expendiures)

on he log o he minimum wage and a se o geographic- and ime-specific

effecs (specificaions 1 2 and 3 described above) As shown in ables 1ndash3 (or

specificaion 3) we hen add covariaes sequenially o hese models including

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2943

26 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

firs he vecor o amily-level conrols ollowed by each o several sae-level

covariaes in urn he unemploymen rae log o median-amily income and he

employmen-o-populaion raio Comparable resuls or specificaions 2 and 3

will be available in our orhcoming working paper

As expeced he simple uncondiional models indicae ha he relaionship beween he minimum wage and SNAP enrollmen i one exiss is a more complex

one influenced by oher acors In he uncondiional model he coefficien on he

variable o ineres991252he log o he minimum wage991252is small in magniude and no

saisically differen rom zero Once we accoun or he influence o labor marke

condiions and variaion in income levels on program paricipaion (by including

unemploymen rae and median-amily income conrol variables respecively)

he effec o he minimum wage on SNAP enrollmen is precisely esimaed Te

coefficien o he log minimum wage is slighly higher (-0042) in he amily-level

analysis han he coefficien (-031) in he sae-level analysis Te level o precision

is also higher in he amily-level analysis Tis is o be expeced when using 124million observaions compared o 1127

Te second dimension o model choice concerns he effec specificaion ables

7ndash9 compare he primary coefficiens o ineres or he SNAP enrollmen and

expendiure models For boh he enrollmen models he effec sizes are smalles

or specificaion 1 larges or specificaion 2 and inermediae beween hese wo

in specificaion 3 Recall ha Specificaion 3 conains sae-specific linear ime

rends in addiion o he census divisionyear conrols included in specificaion

2 In he amily-level enrollmen model he sandard error o he minimum wage

coefficien is smaller han in he oher wo specificaions Sandard errors on he

oher variables are much smaller in specificaions 2 and 3 han in specificaion

1 On he basis o coefficien significance (join and individual) specificaions 2

and 3 are sricly preerred in boh enrollmen models o specificaion 1 which

conains only sae and year fixed effecs

A concern wih specificaions 2 and 3 is ha rend conrols such as sae linear

rends may incorrecly absorb some o he delayed impac o a minimum wage

When we es his issue by including lagged minimum wages we do no find ha

delayed effecs are significan Anoher concern is ha more sauraed modelsuse less o he saisical variaion which could reduce he saisical power o

he resuls However he sandard errors or our more sauraed models are no

higher and are lower in some cases han or he less sauraed models Overall

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3043

Appendix B | wwwamericanprogresso

his evidence suppors our use o he sauraed model as he preerred model

specificaion Moreover Dubersquos 2013 sudy shows ha more sauraed models

perorm beter han models wih jus sae and ime fixed effecs

Te esimaed enrollmen regressions a boh he amily and sae levels show large

and saisically significan coefficiens Te esimaed minimum wage effec in heexpendiures regressions991252or which we have only sae-level daa991252is also large

and saisically significan

We do no use weighed regression or he sae-level models preerring o keep

analysis o he ldquoreamenrdquo (ha is o say a minimum wage change) appropriae

o he average sae raher han he average amily or individual I insead our

primary ineres were he impac o a minimum wage change on he average amily

or he average individual we migh choose o designae he number o amilies

in each sae or he sae populaion respecively as analyic weighs in order o

obain a coefficien beter suied or such inerence

TABLE 4

SNAP enrollment

Family-level linear probability

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0048 -0047 -0040 -0043 -0042

(0013) (0013) (001) (0008) (0009)

Unemployment rate (100) 0505 0420 0280

(0083) (0086) (0082)

Log median income -0057 -0039

(0011) (0011)

Employment-to-population ratio -0239

(0038)

N 1242022 1242022 1242022 1242022 1242022

plt01 plt005 plt001

Note Robust standard errors are in parentheses Observations are clustered at the state level The outcome variable is binary and equal to one if a family is enrolledin SNAP All models include state fixed effects Census division x-year fixed effects and state-specific linear time trends All specifications except 3a include additional

controls for family size race and marital status of the family head presence of children and presence of an adult maleSource Annual data from the March Current Population Survey for 1990 to 2012 estimation includes Current Population Survey probability weights

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3143

28 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

TABLE 5

SNAP enrollment

State-level linear regression

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0028 -0024 -0026 -0031 -0031

(0014) (0013) (0013) (0012) (0012)

Unemployment rate (100) 0625 0466 0320 0339

(0087) (0088) (0085) (0083)

Log median income -0090 -0065 -0061

(0013) (0013) (0013)

Employment-to-population ratio -0282 -0248

(0037) (0038)

N 1127 1127 1127 1127 1127

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the SNAP enrollment rate All models include state fixed effects Census division x-year fixedeffects and state-specific linear time trends

Source Annual data from the March Current Population Survey for 1990 to 2012

TABLE 6

SNAP expenditures

State-level linear regression

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0181 -0149 -0156 -0153 -0190

(011) (0103) (0102) (0103) (0103)

Unemployment rate (100) 4587 4152 4232 4313

(0622) (0621) (0633) (0628)

Log median income -0246 -0261 -0294

(0075) (0078) (0078)

Employment-to-population ratio 0155 0244

(0237) (024)

N 1127 1127 1127 1127 1127

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the natural log of state SNAP expenditures per capita for 1990 to 2012 All models include state

fixed effects Census division x-year fixed effects and state-specific linear time trends

Source Annual data from Bureau of Economic Analysis national income and product accounts tables for 1990 to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3243

Appendix B | wwwamericanprogresso

TABLE 7

Comparison of specifications SNAP enrollment

Family-level linear probability

(1) (2) (3)

Log minimum wage -0028 -0049 -0042

(0014) (0017) (0009)

Unemployment rate (100) 0275 0297 0280

(0161) (0076) (0082)

Log median income -0077 -0055 -0039

(0014) (0012) (0011)

Employment-to-population ratio -0238 -0250 -0239

(0054) (004) (0038)

N 1242022 1242022 1242022

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses Observations are clustered at the state level The outcomevariable is binary or equal to one if a family is enrolled in SNAP All specifications include additional controls forfamily size race and marital status of the family head presence of children and presence of an adult male

Source Annual data from the March Current Population Survey for 1990 to 2012 estimation includes CurrentPopulation Survey probability weights

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3343

30 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

TABLE 8

Comparison of specifications SNAP enrollment

State-level linear regression

(1) (2) (3)

Log minimum wage -0019 -0035 -0031

(0009) (0012) (0012)

Unemployment rate (100) 0401 0370 0339

(0063) (0077) (0083)

Log median income -0081 -0073 -0061

(0011) (0013) (0013)

Employment-to-population ratio -0183 -0222 -0248

(0039) (0039) (0038)

N 1127 1127 1127

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the SNAP enrollment rate Allregressions include the share of households with children and the racial shares of the population

Source Annual data from the March Current Population Survey for 1990 to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3443

Appendix B | wwwamericanprogresso

TABLE 9

Comparison of specifications SNAP expenditures

State-level linear regression

(1) (2) (3)

Log minimum wage -0121 -0203 -0190

(0075) (0103) (0103)

Unemployment rate (100) 5292 5152 4313

(0464) (0576) (0628)

Log median income -0437 -0417 -0294

(008) (0086) (0078)

Employment-to-population ratio -0040 -0220 0244

(0261) (0260) (0240)

N 1127 1127 1127

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the natural log of state SNAPexpenditures per capita All regressions include the share of households with children and the racial shares ofthe state population

Source Annual data from Bureau of Economic Analysis national income and product accounts tables for 1990to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3543

32 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3643

Appendix C | wwwamericanprogresso

Appendix C Harkin-Miller

policy simulation results

StateMinimum wage

(2014)

SNAP enrollment (2012) Predicted change in SNAP enrollment

RateTotal

(persons)

Model 1 family

enrollment (linearprobability)

Model 2 state

enrollment (linearregression)

Model 3 s

expenditu(linear regre

Alabama $725 164 788682 -66922 -58897 -58906

Alaska $775 120 87436 -8104 -6887 -3288

Arizona $790 201 1319323 -67435 -56738 -64356

Arkansas $725 78 230489 -40977 -36063 -35248

California $800 206 7813680 -371131 -310222 -18223

Colorado $800 164 853155 -50684 -42365 -23926

Connecticut $870 91 326621 -22456 -17975 -13711

Delaware $725 186 170262 -12739 -11211 -10647

District of Columbia $825 133 84009 -5370 -4417 -3632

Florida $793 166 3208026 -195813 -164426 -13046

Georgia $725 160 1586336 -137741 -121224 -11004

Hawaii $725 96 133662 -19310 -16995 -14933

Idaho $725 92 147501 -22165 -19507 -15809

Illinois $825 95 1225084 -109088 -89742 -70955

Indiana $725 125 816233 -90818 -79928 -83985

Iowa $725 155 478011 -42716 -37594 -28556

Kansas $725 135 388269 -40082 -35275 -27461

Kentucky $725 130 568821 -60840 -53544 -52259

Louisiana $725 149 683832 -63929 -56263 -66083

Maine $750 77 101976 -16567 -14323 -15234

Maryland $725 144 846415 -81748 -71946 -38370

Massachusetts $800 130 864721 -64902 -54251 -42913

Michigan $740 146 1439141 -128801 -112140 -11022

Minnesota $725 133 713646 -74730 -65769 -37878

TABLE 10

SNAP enrollments Predicted changes by state under the Harkin-Miller bill

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3743

34 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

StateMinimum wage

(2014)

SNAP enrollment (2012) Predicted change in SNAP enrollment

RateTotal

(persons)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 state

expenditures

(linear regression)

Mississippi $725 129 386501 -41486 -36511 -46467

Missouri $750 172 1036182 -75131 -64952 -56244

Montana $790 132 132452 -10350 -8708 -5846

Nebraska $725 124 230683 -25773 -22683 -12189

Nevada $825 162 446035 -23349 -19209 -11894

New Hampshire $725 127 168404 -18359 -16157 -5735

New Jersey $825 160 1416666 -75175 -61843 -28236

New Mexico $750 149 310896 -25983 -22463 -22512

New York $800 192 3763553 -191193 -159815 -142182

North Carolina $725 174 1697193 -135417 -119179 -113503

North Dakota $725 87 61225 -9743 -8574 -4021

Ohio $795 143 1647345 -115869 -97169 -88580

Oklahoma $725 129 494053 -53006 -46650 -46854

Oregon $910 124 485326 -17036 -13328 -16398

Pennsylvania $725 161 2053643 -177315 -156052 -125586

Rhode Island $800 156 163730 -10258 -8574 -8698

South Carolina $725 94 445277 -65614 -57746 -50304

South Dakota $725 208 173749 -11586 -10197 -7458

Tennessee $725 142 914903 -89667 -78915 -99134

Texas $725 110 2863779 -362018 -318607 -253285

Utah $725 88 251107 -39658 -34902 -19390

Vermont $873 156 97792 -3823 -3055 -2475

Virginia $725 101 829771 -113723 -100086 -58212

Washington $932 72 496934 -23221 -17947 -17756

West Virginia $725 58 107875 -25792 -22699 -21665

Wisconsin $725 75 427822 -79521 -69986 -53210

Wyoming $725 164 94590 -8010 -7050 -3104

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3843

Appendix C | wwwamericanprogresso

StateMinimum wage

(2014)

SNAP expenditures (2012)Predicted change in SNAP expenditures

(millions of dollars)

Per resident

(millions of

dollars)

Total

(millions of

dollars)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 s

expenditu

(linear regre

Alabama $725 $290 $1397 -$1185 -$1043 -$1043

Alaska $775 $253 $185 -$262 -$223 -$106

Arizona $790 $257 $1686 -$935 -$787 -$892

Arkansas $725 $248 $733 -$636 -$560 -$547

California $800 $189 $7164 -$7276 -$6082 -$3573

Colorado $800 $157 $816 -$862 -$721 -$407

Connecticut $870 $191 $686 -$343 -$275 -$210

Delaware $725 $250 $229 -$205 -$180 -$171

District of Columbia $825 $366 $232 -$146 -$120 -$99

Florida $793 $294 $5676 -$4429 -$3719 -$2951

Georgia $725 $317 $3140 -$2936 -$2584 -$2346

Hawaii $725 $335 $465 -$449 -$395 -$347

Idaho $725 $225 $359 -$376 -$331 -$268

Illinois $825 $249 $3200 -$2096 -$1725 -$1364

Indiana $725 $220 $1439 -$1162 -$1023 -$1075

Iowa $725 $192 $589 -$658 -$579 -$440

Kansas $725 $159 $460 -$502 -$441 -$344

Kentucky $725 $298 $1303 -$1133 -$997 -$973

Louisiana $725 $315 $1450 -$1047 -$922 -$1083

Maine $750 $281 $373 -$267 -$231 -$246

Maryland $725 $188 $1109 -$1765 -$1553 -$828

Massachusetts $800 $206 $1366 -$1030 -$861 -$681

Michigan $740 $300 $2963 -$2400 -$2090 -$2054

Minnesota $725 $140 $755 -$1113 -$980 -$564

Mississippi $725 $326 $973 -$649 -$571 -$726

Missouri $750 $241 $1452 -$1278 -$1104 -$956

Montana $790 $190 $191 -$179 -$151 -$101

Nebraska $725 $140 $259 -$409 -$360 -$194

Nevada $825 $191 $527 -$441 -$363 -$225

New Hampshire $725 $126 $167 -$399 -$351 -$125

TABLE 11

SNAP expenditures Predicted changes by state under the Harkin-Miller bill

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3943

36 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

StateMinimum wage

(2014)

SNAP expenditures (2012)Predicted change in SNAP expenditures

(millions of dollars)

Per resident

(millions of

dollars)

Total

(millions of

dollars)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 state

expenditures

(linear regression)

New Jersey $825 $160 $1420 -$1611 -$1325 -$605

New Mexico $750 $324 $675 -$513 -$444 -$445New York $800 $287 $5616 -$3766 -$3148 -$2801

North Carolina $725 $252 $2454 -$2187 -$1925 -$1833

North Dakota $725 $128 $90 -$162 -$143 -$67

Ohio $795 $259 $2995 -$2013 -$1688 -$1539

Oklahoma $725 $248 $945 -$799 -$703 -$706

Oregon $910 $322 $1255 -$272 -$213 -$262

Pennsylvania $725 $218 $2779 -$2930 -$2579 -$2075

Rhode Island $800 $280 $294 -$173 -$144 -$147

South Carolina $725 $291 $1373 -$1337 -$1177 -$1025South Dakota $725 $198 $165 -$192 -$169 -$123

Tennessee $725 $324 $2091 -$1413 -$1243 -$1562

Texas $725 $230 $5997 -$6402 -$5634 -$4479

Utah $725 $141 $402 -$614 -$541 -$300

Vermont $873 $230 $144 -$66 -$53 -$43

Virginia $725 $173 $1413 -$2062 -$1815 -$1056

Washington $932 $244 $1682 -$350 -$270 -$267

West Virginia $725 $273 $508 -$451 -$397 -$379

Wisconsin $725 $204 $1166 -$1302 -$1146 -$871Wyoming $725 $95 $55 -$105 -$93 -$41

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4043

Appendix C | wwwamericanprogresso

FIGURE 2

Raising the minimum wage to $1010would cut taxpayer costs in every state

Predicted decreases in cost and enrollment

in SNAP in 50 states

$200+$51ndash$100

$101ndash$200

0ndash$25

$26ndash$50

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4143

38 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

Endnotes

1 US Department of Agriculture ldquoSupplementalNutrition Assistance Program (SNAP)rdquo available athttpwwwfnsusdagovsnapeligibility (last accessedFebruary 2014)

2 For this initial analysis we do not consider Harkin-Millerrsquos increase in subminimum wages for tippedworkers To do s o would increase the estimated SNAP

savings by an unknown amount

3 The Congressional Budget Office estimates thatworkers currently earning between $1010 and $1150per hour would see their wages rise under the Harkin-Miller proposal Congressional Budget O ffice ldquoTheEffects of a Minimum Wage Increase on Employmentand Family Incomerdquo (2014)

4 Marianne Page Joanne Spetz and Jane Millar ldquoDoesthe Minimum Wage Affect Welfare Caseloadsrdquo Journalof Policy Analysis and Management 24 (2) (2005)273ndash295

5 Marianne Bitler and Hilary Hoynes ldquoThe More ThingsChange the More They Stay the Same The SafetyNet Living Arrangements and Poverty in the GreatRecessionrdquo NBER Working Paper 194 49 2013

6 Sylvia Allegretto and others ldquoFast Food Poverty Wages The Public Cost of Low-Wage Jobs in the Fast-FoodIndustryrdquo (Berkeley California Center for LaborResearch and Education 2013) available at httplaborcenterberkeleyedupubliccostsfast_food_poverty_wage

7 David Neumark and William Wascher ldquoDoes a HigherMinimum Wage Enhance the Effectiveness of theEarned Income Tax Creditrdquo Industrial and LaborRelations Review 64 (4) (2011) 712ndash746

8 David Lee and Emmanuel Saez ldquoOptimal MinimumWage Policy in Competitive Labor Marketsrdquo Journal ofPublic Economics 96 (9) (2012) 739ndash749

9 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family IncomerdquoFebruary 2014

10 Jesse Rothstein ldquoIs the EITC as Good as an NITConditional Cash Transfers and Tax Incidencerdquo AmericanEconomic Journal Economic Policy 2 (1) (2010) 177ndash208

11 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family Incomerdquo

12 Dube Arindrajit 2013 rdquoMinimum Wagesand the Distribution of Family IncomerdquoUnpublished working paper Available at httpsdldropboxusercontentcomu15038936Dube_MinimumWagesFamilyIncomespdf

13 As Dube explains in ldquoThe poverty of Minimum WageFactsrdquo the simulation approach underestimate stemsfrom a number of unwarranted assumptions includingthe range of actual wage increases and the accuracy ofwage data in the Current Population Survey The causal

approach does not make these assumptions

14 Allegretto Sylvia and others 2013 ldquoCredible ResearchDesigns for Minimum Wage Studiesrdquo Working Paper148-13 University of California Berkeley Institutefor Research on Labor and Employment Available athttpwwwirleberkeleyeduworkingpapers148-13pdf

15 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family In comerdquo

16 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo (2012)

17 Ibid

18 Ibid

19 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo

20 US Department of Agriculture ldquoSupplementalNutrition Assistance Program (SNAP)rdquo

21 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo

22 Ibid

23 We deviate from the Census Bureaursquos definition ofa family unit which is ldquotwo people or more (on e ofwhom is the householder) related by bir th marriageor adoption and residing togetherrdquo US Bureau ofthe Census ldquoCurrent Population Survey DefinitionsFamilyrdquo available at httpwwwcensusgovcpsabout

cpsdefhtml (last accessed February 2014) We countas a family unit any individual residing on his or herown two or more persons residing together whodo not belong to a family in the March CPS sampleare constructed as one family in our analysis For thepurposes of food stamp allocations the consumptionresulting from this transfer is probably distributed tofamily members (rather than household members ora single individual within the household) Howeversingle individuals canmdashand domdashreceive SNAPbenefits Excluding them would fail to make theanalysis reflective of the population at large

24 Strictly the family level linear probability modelpredicts the percentage-point decrease in theprobability that an individual family will receive SNAPpayments When applied to a large number of familieshowever we are able to interpret the coefficient asa decrease in the mean of enrollmentmdashthat is a

decrease in the enrollment ratemdashby applying the lawof iterated expectations

25 We generate expenditure predictions from theenrollment modelsmdashand conversely generateenrollment predictions from the expenditure modelmdashby assuming that expenditures per enrolled familyremains the same before and after the minimum wagechange In practice this is likely to be a conservativeestimatemdashthat is to underestimate the decrease inSNAP activity Average SNAP benefits per family willalso decrease as many families that remain eligible forSNAP experience income gains

26 Wage and Hour Division ldquoMinimum Wage Laws inthe States ndash Januar y 1 2014rdquo available at httpwwwdolgovwhdminwageamericahtm (last accessedFebruary 2014)

27 See for example Sylvia Allegretto and others ldquoCredibleResearch Designs for Minimum Wage Studiesrdquo WorkingPaper 148-113 (Berkeley California Institute forResearch on Labor and Employment 2013) available athttpwwwirleberkeleyeduworkingpapers148-13pdf

28 We will report these results in a forthcoming workingpaper

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4243

Endnotes | wwwamericanprogresso

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4343

The Center for American Progress is a nonpartisan research and educational institute

dedicated to promoting a strong just and free America that ensures opportunity

for all We believe that Americans are bound together by a common commitment to

these values and we aspire to ensure that our national policies reflect these values

We work to find progressive and pragmatic solutions to significant domestic and

international problems and develop policy proposals that foster a government that

is ldquoof the people by the people and for the peoplerdquo

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2343

20 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

About the authors

Rachel West is a maser o public policy candidae a he Goldman School

o Public Policy Universiy o Caliornia Berkeley Her research ocuses on

economic policy in he areas o low-wage labor and povery

Michael Reich is proessor o economics and direcor o he Insiue or

Research on Labor and Employmen a he Universiy o Caliornia a Berkeley

His research publicaions cover numerous areas o labor economics including

racial inequaliy labor marke segmenaion high-perormance workplaces

union-managemen cooperaion Japanese labor-managemen sysems living

wages and minimum wages He received his docorae in economics rom

Harvard Universiy

Acknowledgments

We are graeul o Sylvia Allegreto Arindraji Dube Bill Leser Jesse Rohsein

Daniel Tompson and Ben Zipperer or heir valuable suggesions

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2443

References | wwwamericanprogresso

References

Allegreto Sylvia and ohers 2013 ldquoFas Food Povery Wages Te Public Cos o Low-Wage Jobsin he Fas-Food Indusryrdquo Universiy o Caliornia Berkeley Cener or Labor Research andEducaion Available a htplaborcenerberkeleyedupubliccossas_ood_povery_wage

Allegreto Sylvia and ohers 2013 ldquoCredible Research Designs or Minimum Wage Sudiesrdquo

Working Paper 148-13 Universiy o Caliornia Berkeley Insiue or Research on Labor andEmploymen Available a htpwwwirleberkeleyeduworkingpapers148-13pd

Biler Marianne and Hilary Hoynes 2013 ldquo Te More Tings Change he More Tey Say heSame Te Saey Ne Living Arrangemens and Povery in he Grea Recessionrdquo Working Paper19449 Naional Bureau o Economic Research

Congressional Budge Office 2012 ldquoTe Supplemenal Nuriion Assisance Programrdquo Washingon Available a htpwwwcbogovsiesdeaulfilescbofilesatachmens04-19-SNAPpd

991252 991252 991252 2014 ldquoTe Effec o a Minimum-Wage Increase on Employmen and Family Incomerdquo Washingon Available a htpwwwcbogovsiesdeaulfilescbofilesatachmens44995-MinimumWagepd

Dube Arindraji 2013 rdquoMinimum Wages and he Disribuion o Family Incomerdquo Unpublished working paper Available a htpsdldropboxuserconencomu15038936Dube_ MinimumWagesFamilyIncomespd

991252 991252 991252 2014 ldquoTe Povery o Minimum Wage Facsrdquo Arindraji Dube Blog January 22 Availablea htparindubecom20140122he-povery-o-minimum-wage-acs

Lee David and Emmanuel Saez 2012 ldquoOpimal Minimum Wage Policy in Compeiive LaborMarkesrdquo Journal o Public Economics 96 (9) 739ndash749

Neumark David and William Wascher 1992 ldquoEmploymen Effecs o Minimum and Subminimum Wages Panel Daa on Sae Minimum Wage Lawsrdquo Industrial and Labor Relations Review 46 (1)

55ndash81

Neumark David and William Wascher 2011 ldquoDoes a Higher Minimum Wage Enhance heEffeciveness o he Earned Income ax Credirdquo Industrial and Labor Relations Review 64 (4)712ndash746

Page Marianne Joanne Spez and Jane Millar 2005 ldquoDoes he Minimum Wage Affec WelareCaseloadsrdquo Journal o Policy Analysis and Management 24 (2) 273ndash295

Rohsein Jesse 2010 ldquoIs he EIC as Good as an NI Condiional Cash ransers and ax

Incidencerdquo American Economic Journal Economic Policy 2 (1) 177ndash208

Wage and Hours Division 2014 ldquoMinimum Wage Laws in he Saes ndash January 1 2014rdquo USDeparmen o Labor (htpwwwdolgovwhdminwageamericahm [February 2014])

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2543

22 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2643

Appendix A | wwwamericanprogresso

Appendix A

Pre-trend falsification check

Recen minimum wage research27 highlighs a common flaw in previous sudies

ailure o veriy ha he oucome variable is ree o negaive pre-exising rends I

or example SNAP aciviy was already rending down in saes ha raised heir

minimum wages beore hese changes came ino effec our regression analysis could

(misakenly) atribue ha reducion o he minimum wage We check or such

pre-rends by inroducing variables ha represen he prior yearrsquos value or leads o

he minimum wage I he model esimaes he minimum wage o have an effec on

he oucome variable beore he wage change wen ino effec hen an unobservedacor no he minimum wage change caused he change in SNAP aciviy

We es he specificaions above or pre-rends by including a one-year lead in

all hree specificaions We find ha he lead erms are small posiive and no

saisically significan indicaing ha he concurren minimum wage991252no

he wage level in prior periods991252is driving he observed changes in SNAP

oucomes28 In paricular he coefficien (sandard error) on he lead erm in

our preerred amily-level enrollmen regression is 011 and no significan

while he coefficien and sandard error o he conemporaneous minimum

wage is unchanged In he sae-level preerred enrollmen regression he

coefficien o he lead erm is again small (07) and i is no significan Te

corresponding coefficien on he lead erm in he sae-level expendiure

regression is 16 and is no significan Te posiive poin esimaes on hese lead

erms resuls no only rule ou disoring negaive pre-rends Tey also sugges

ha our main resuls may underesimae he rue effecs

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2743

24 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2843

Appendix B | wwwamericanprogresso

Appendix B

Model selection process

For boh he amily-level and sae-level models we es hree mehods o conrol

or unobserved geographic- and ime-varying characerisics as suggesed by he

minimum wage lieraure Firs we include only independen sae-specific fixed

effecs and year-specific fixed effecs Tis specificaion (specificaion 1) implicily

assumes ha amilies in any sae consiue an equally good saisical ldquoconrolrdquo

group or hose in any randomly chosen sae afer accouning or various

characerisics (median income and unemploymen rae among ohers) Similarly

simple ime fixed effecs assume ha amilies surveyed in any year can crediblyserve as a conrol group or amilies surveyed in every oher year o he sample

(1990 hrough 2012)

In oher words specificaion 1 assumes ha a saersquos immediae neighbor provides

no beter a couneracual or he effec o a minimum wage change han does a

sae across he counry We relax his resricive specificaion sequenially in wo

seps In specificaion 2 we replace simple year fixed effecs wih fixed effecs or

each Census divisionyear (capured as an addiional variable in he vecor By

using division-year effecs we remove he resricion ha amilies in each sae

are equally good saisical conrols or all oher amilies Raher we allow or he

possibiliy ha amilies in similar geographic regions (or example he Souh or

he Norheas) may be more similar o one anoher han amilies arher away

Finally in specificaion 3 we add sae-specific linear ime rends o he previous

specificaion Tus specificaion 3 is he mos rigorous model specificaion in ha

i allows or heerogeneiy along hree dimensions Ta is specificaion 3 allows

each sae o have is own ime-varying rends raher han imposing he resricion

ha saes evolve idenically over he 22 years in he sample

We begin building he heoreical specificaion above rom a se o simpleuncondiional models regression o SNAP aciviy (enrollmen or expendiures)

on he log o he minimum wage and a se o geographic- and ime-specific

effecs (specificaions 1 2 and 3 described above) As shown in ables 1ndash3 (or

specificaion 3) we hen add covariaes sequenially o hese models including

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2943

26 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

firs he vecor o amily-level conrols ollowed by each o several sae-level

covariaes in urn he unemploymen rae log o median-amily income and he

employmen-o-populaion raio Comparable resuls or specificaions 2 and 3

will be available in our orhcoming working paper

As expeced he simple uncondiional models indicae ha he relaionship beween he minimum wage and SNAP enrollmen i one exiss is a more complex

one influenced by oher acors In he uncondiional model he coefficien on he

variable o ineres991252he log o he minimum wage991252is small in magniude and no

saisically differen rom zero Once we accoun or he influence o labor marke

condiions and variaion in income levels on program paricipaion (by including

unemploymen rae and median-amily income conrol variables respecively)

he effec o he minimum wage on SNAP enrollmen is precisely esimaed Te

coefficien o he log minimum wage is slighly higher (-0042) in he amily-level

analysis han he coefficien (-031) in he sae-level analysis Te level o precision

is also higher in he amily-level analysis Tis is o be expeced when using 124million observaions compared o 1127

Te second dimension o model choice concerns he effec specificaion ables

7ndash9 compare he primary coefficiens o ineres or he SNAP enrollmen and

expendiure models For boh he enrollmen models he effec sizes are smalles

or specificaion 1 larges or specificaion 2 and inermediae beween hese wo

in specificaion 3 Recall ha Specificaion 3 conains sae-specific linear ime

rends in addiion o he census divisionyear conrols included in specificaion

2 In he amily-level enrollmen model he sandard error o he minimum wage

coefficien is smaller han in he oher wo specificaions Sandard errors on he

oher variables are much smaller in specificaions 2 and 3 han in specificaion

1 On he basis o coefficien significance (join and individual) specificaions 2

and 3 are sricly preerred in boh enrollmen models o specificaion 1 which

conains only sae and year fixed effecs

A concern wih specificaions 2 and 3 is ha rend conrols such as sae linear

rends may incorrecly absorb some o he delayed impac o a minimum wage

When we es his issue by including lagged minimum wages we do no find ha

delayed effecs are significan Anoher concern is ha more sauraed modelsuse less o he saisical variaion which could reduce he saisical power o

he resuls However he sandard errors or our more sauraed models are no

higher and are lower in some cases han or he less sauraed models Overall

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3043

Appendix B | wwwamericanprogresso

his evidence suppors our use o he sauraed model as he preerred model

specificaion Moreover Dubersquos 2013 sudy shows ha more sauraed models

perorm beter han models wih jus sae and ime fixed effecs

Te esimaed enrollmen regressions a boh he amily and sae levels show large

and saisically significan coefficiens Te esimaed minimum wage effec in heexpendiures regressions991252or which we have only sae-level daa991252is also large

and saisically significan

We do no use weighed regression or he sae-level models preerring o keep

analysis o he ldquoreamenrdquo (ha is o say a minimum wage change) appropriae

o he average sae raher han he average amily or individual I insead our

primary ineres were he impac o a minimum wage change on he average amily

or he average individual we migh choose o designae he number o amilies

in each sae or he sae populaion respecively as analyic weighs in order o

obain a coefficien beter suied or such inerence

TABLE 4

SNAP enrollment

Family-level linear probability

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0048 -0047 -0040 -0043 -0042

(0013) (0013) (001) (0008) (0009)

Unemployment rate (100) 0505 0420 0280

(0083) (0086) (0082)

Log median income -0057 -0039

(0011) (0011)

Employment-to-population ratio -0239

(0038)

N 1242022 1242022 1242022 1242022 1242022

plt01 plt005 plt001

Note Robust standard errors are in parentheses Observations are clustered at the state level The outcome variable is binary and equal to one if a family is enrolledin SNAP All models include state fixed effects Census division x-year fixed effects and state-specific linear time trends All specifications except 3a include additional

controls for family size race and marital status of the family head presence of children and presence of an adult maleSource Annual data from the March Current Population Survey for 1990 to 2012 estimation includes Current Population Survey probability weights

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3143

28 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

TABLE 5

SNAP enrollment

State-level linear regression

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0028 -0024 -0026 -0031 -0031

(0014) (0013) (0013) (0012) (0012)

Unemployment rate (100) 0625 0466 0320 0339

(0087) (0088) (0085) (0083)

Log median income -0090 -0065 -0061

(0013) (0013) (0013)

Employment-to-population ratio -0282 -0248

(0037) (0038)

N 1127 1127 1127 1127 1127

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the SNAP enrollment rate All models include state fixed effects Census division x-year fixedeffects and state-specific linear time trends

Source Annual data from the March Current Population Survey for 1990 to 2012

TABLE 6

SNAP expenditures

State-level linear regression

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0181 -0149 -0156 -0153 -0190

(011) (0103) (0102) (0103) (0103)

Unemployment rate (100) 4587 4152 4232 4313

(0622) (0621) (0633) (0628)

Log median income -0246 -0261 -0294

(0075) (0078) (0078)

Employment-to-population ratio 0155 0244

(0237) (024)

N 1127 1127 1127 1127 1127

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the natural log of state SNAP expenditures per capita for 1990 to 2012 All models include state

fixed effects Census division x-year fixed effects and state-specific linear time trends

Source Annual data from Bureau of Economic Analysis national income and product accounts tables for 1990 to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3243

Appendix B | wwwamericanprogresso

TABLE 7

Comparison of specifications SNAP enrollment

Family-level linear probability

(1) (2) (3)

Log minimum wage -0028 -0049 -0042

(0014) (0017) (0009)

Unemployment rate (100) 0275 0297 0280

(0161) (0076) (0082)

Log median income -0077 -0055 -0039

(0014) (0012) (0011)

Employment-to-population ratio -0238 -0250 -0239

(0054) (004) (0038)

N 1242022 1242022 1242022

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses Observations are clustered at the state level The outcomevariable is binary or equal to one if a family is enrolled in SNAP All specifications include additional controls forfamily size race and marital status of the family head presence of children and presence of an adult male

Source Annual data from the March Current Population Survey for 1990 to 2012 estimation includes CurrentPopulation Survey probability weights

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3343

30 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

TABLE 8

Comparison of specifications SNAP enrollment

State-level linear regression

(1) (2) (3)

Log minimum wage -0019 -0035 -0031

(0009) (0012) (0012)

Unemployment rate (100) 0401 0370 0339

(0063) (0077) (0083)

Log median income -0081 -0073 -0061

(0011) (0013) (0013)

Employment-to-population ratio -0183 -0222 -0248

(0039) (0039) (0038)

N 1127 1127 1127

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the SNAP enrollment rate Allregressions include the share of households with children and the racial shares of the population

Source Annual data from the March Current Population Survey for 1990 to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3443

Appendix B | wwwamericanprogresso

TABLE 9

Comparison of specifications SNAP expenditures

State-level linear regression

(1) (2) (3)

Log minimum wage -0121 -0203 -0190

(0075) (0103) (0103)

Unemployment rate (100) 5292 5152 4313

(0464) (0576) (0628)

Log median income -0437 -0417 -0294

(008) (0086) (0078)

Employment-to-population ratio -0040 -0220 0244

(0261) (0260) (0240)

N 1127 1127 1127

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the natural log of state SNAPexpenditures per capita All regressions include the share of households with children and the racial shares ofthe state population

Source Annual data from Bureau of Economic Analysis national income and product accounts tables for 1990to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3543

32 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3643

Appendix C | wwwamericanprogresso

Appendix C Harkin-Miller

policy simulation results

StateMinimum wage

(2014)

SNAP enrollment (2012) Predicted change in SNAP enrollment

RateTotal

(persons)

Model 1 family

enrollment (linearprobability)

Model 2 state

enrollment (linearregression)

Model 3 s

expenditu(linear regre

Alabama $725 164 788682 -66922 -58897 -58906

Alaska $775 120 87436 -8104 -6887 -3288

Arizona $790 201 1319323 -67435 -56738 -64356

Arkansas $725 78 230489 -40977 -36063 -35248

California $800 206 7813680 -371131 -310222 -18223

Colorado $800 164 853155 -50684 -42365 -23926

Connecticut $870 91 326621 -22456 -17975 -13711

Delaware $725 186 170262 -12739 -11211 -10647

District of Columbia $825 133 84009 -5370 -4417 -3632

Florida $793 166 3208026 -195813 -164426 -13046

Georgia $725 160 1586336 -137741 -121224 -11004

Hawaii $725 96 133662 -19310 -16995 -14933

Idaho $725 92 147501 -22165 -19507 -15809

Illinois $825 95 1225084 -109088 -89742 -70955

Indiana $725 125 816233 -90818 -79928 -83985

Iowa $725 155 478011 -42716 -37594 -28556

Kansas $725 135 388269 -40082 -35275 -27461

Kentucky $725 130 568821 -60840 -53544 -52259

Louisiana $725 149 683832 -63929 -56263 -66083

Maine $750 77 101976 -16567 -14323 -15234

Maryland $725 144 846415 -81748 -71946 -38370

Massachusetts $800 130 864721 -64902 -54251 -42913

Michigan $740 146 1439141 -128801 -112140 -11022

Minnesota $725 133 713646 -74730 -65769 -37878

TABLE 10

SNAP enrollments Predicted changes by state under the Harkin-Miller bill

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3743

34 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

StateMinimum wage

(2014)

SNAP enrollment (2012) Predicted change in SNAP enrollment

RateTotal

(persons)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 state

expenditures

(linear regression)

Mississippi $725 129 386501 -41486 -36511 -46467

Missouri $750 172 1036182 -75131 -64952 -56244

Montana $790 132 132452 -10350 -8708 -5846

Nebraska $725 124 230683 -25773 -22683 -12189

Nevada $825 162 446035 -23349 -19209 -11894

New Hampshire $725 127 168404 -18359 -16157 -5735

New Jersey $825 160 1416666 -75175 -61843 -28236

New Mexico $750 149 310896 -25983 -22463 -22512

New York $800 192 3763553 -191193 -159815 -142182

North Carolina $725 174 1697193 -135417 -119179 -113503

North Dakota $725 87 61225 -9743 -8574 -4021

Ohio $795 143 1647345 -115869 -97169 -88580

Oklahoma $725 129 494053 -53006 -46650 -46854

Oregon $910 124 485326 -17036 -13328 -16398

Pennsylvania $725 161 2053643 -177315 -156052 -125586

Rhode Island $800 156 163730 -10258 -8574 -8698

South Carolina $725 94 445277 -65614 -57746 -50304

South Dakota $725 208 173749 -11586 -10197 -7458

Tennessee $725 142 914903 -89667 -78915 -99134

Texas $725 110 2863779 -362018 -318607 -253285

Utah $725 88 251107 -39658 -34902 -19390

Vermont $873 156 97792 -3823 -3055 -2475

Virginia $725 101 829771 -113723 -100086 -58212

Washington $932 72 496934 -23221 -17947 -17756

West Virginia $725 58 107875 -25792 -22699 -21665

Wisconsin $725 75 427822 -79521 -69986 -53210

Wyoming $725 164 94590 -8010 -7050 -3104

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3843

Appendix C | wwwamericanprogresso

StateMinimum wage

(2014)

SNAP expenditures (2012)Predicted change in SNAP expenditures

(millions of dollars)

Per resident

(millions of

dollars)

Total

(millions of

dollars)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 s

expenditu

(linear regre

Alabama $725 $290 $1397 -$1185 -$1043 -$1043

Alaska $775 $253 $185 -$262 -$223 -$106

Arizona $790 $257 $1686 -$935 -$787 -$892

Arkansas $725 $248 $733 -$636 -$560 -$547

California $800 $189 $7164 -$7276 -$6082 -$3573

Colorado $800 $157 $816 -$862 -$721 -$407

Connecticut $870 $191 $686 -$343 -$275 -$210

Delaware $725 $250 $229 -$205 -$180 -$171

District of Columbia $825 $366 $232 -$146 -$120 -$99

Florida $793 $294 $5676 -$4429 -$3719 -$2951

Georgia $725 $317 $3140 -$2936 -$2584 -$2346

Hawaii $725 $335 $465 -$449 -$395 -$347

Idaho $725 $225 $359 -$376 -$331 -$268

Illinois $825 $249 $3200 -$2096 -$1725 -$1364

Indiana $725 $220 $1439 -$1162 -$1023 -$1075

Iowa $725 $192 $589 -$658 -$579 -$440

Kansas $725 $159 $460 -$502 -$441 -$344

Kentucky $725 $298 $1303 -$1133 -$997 -$973

Louisiana $725 $315 $1450 -$1047 -$922 -$1083

Maine $750 $281 $373 -$267 -$231 -$246

Maryland $725 $188 $1109 -$1765 -$1553 -$828

Massachusetts $800 $206 $1366 -$1030 -$861 -$681

Michigan $740 $300 $2963 -$2400 -$2090 -$2054

Minnesota $725 $140 $755 -$1113 -$980 -$564

Mississippi $725 $326 $973 -$649 -$571 -$726

Missouri $750 $241 $1452 -$1278 -$1104 -$956

Montana $790 $190 $191 -$179 -$151 -$101

Nebraska $725 $140 $259 -$409 -$360 -$194

Nevada $825 $191 $527 -$441 -$363 -$225

New Hampshire $725 $126 $167 -$399 -$351 -$125

TABLE 11

SNAP expenditures Predicted changes by state under the Harkin-Miller bill

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3943

36 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

StateMinimum wage

(2014)

SNAP expenditures (2012)Predicted change in SNAP expenditures

(millions of dollars)

Per resident

(millions of

dollars)

Total

(millions of

dollars)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 state

expenditures

(linear regression)

New Jersey $825 $160 $1420 -$1611 -$1325 -$605

New Mexico $750 $324 $675 -$513 -$444 -$445New York $800 $287 $5616 -$3766 -$3148 -$2801

North Carolina $725 $252 $2454 -$2187 -$1925 -$1833

North Dakota $725 $128 $90 -$162 -$143 -$67

Ohio $795 $259 $2995 -$2013 -$1688 -$1539

Oklahoma $725 $248 $945 -$799 -$703 -$706

Oregon $910 $322 $1255 -$272 -$213 -$262

Pennsylvania $725 $218 $2779 -$2930 -$2579 -$2075

Rhode Island $800 $280 $294 -$173 -$144 -$147

South Carolina $725 $291 $1373 -$1337 -$1177 -$1025South Dakota $725 $198 $165 -$192 -$169 -$123

Tennessee $725 $324 $2091 -$1413 -$1243 -$1562

Texas $725 $230 $5997 -$6402 -$5634 -$4479

Utah $725 $141 $402 -$614 -$541 -$300

Vermont $873 $230 $144 -$66 -$53 -$43

Virginia $725 $173 $1413 -$2062 -$1815 -$1056

Washington $932 $244 $1682 -$350 -$270 -$267

West Virginia $725 $273 $508 -$451 -$397 -$379

Wisconsin $725 $204 $1166 -$1302 -$1146 -$871Wyoming $725 $95 $55 -$105 -$93 -$41

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4043

Appendix C | wwwamericanprogresso

FIGURE 2

Raising the minimum wage to $1010would cut taxpayer costs in every state

Predicted decreases in cost and enrollment

in SNAP in 50 states

$200+$51ndash$100

$101ndash$200

0ndash$25

$26ndash$50

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4143

38 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

Endnotes

1 US Department of Agriculture ldquoSupplementalNutrition Assistance Program (SNAP)rdquo available athttpwwwfnsusdagovsnapeligibility (last accessedFebruary 2014)

2 For this initial analysis we do not consider Harkin-Millerrsquos increase in subminimum wages for tippedworkers To do s o would increase the estimated SNAP

savings by an unknown amount

3 The Congressional Budget Office estimates thatworkers currently earning between $1010 and $1150per hour would see their wages rise under the Harkin-Miller proposal Congressional Budget O ffice ldquoTheEffects of a Minimum Wage Increase on Employmentand Family Incomerdquo (2014)

4 Marianne Page Joanne Spetz and Jane Millar ldquoDoesthe Minimum Wage Affect Welfare Caseloadsrdquo Journalof Policy Analysis and Management 24 (2) (2005)273ndash295

5 Marianne Bitler and Hilary Hoynes ldquoThe More ThingsChange the More They Stay the Same The SafetyNet Living Arrangements and Poverty in the GreatRecessionrdquo NBER Working Paper 194 49 2013

6 Sylvia Allegretto and others ldquoFast Food Poverty Wages The Public Cost of Low-Wage Jobs in the Fast-FoodIndustryrdquo (Berkeley California Center for LaborResearch and Education 2013) available at httplaborcenterberkeleyedupubliccostsfast_food_poverty_wage

7 David Neumark and William Wascher ldquoDoes a HigherMinimum Wage Enhance the Effectiveness of theEarned Income Tax Creditrdquo Industrial and LaborRelations Review 64 (4) (2011) 712ndash746

8 David Lee and Emmanuel Saez ldquoOptimal MinimumWage Policy in Competitive Labor Marketsrdquo Journal ofPublic Economics 96 (9) (2012) 739ndash749

9 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family IncomerdquoFebruary 2014

10 Jesse Rothstein ldquoIs the EITC as Good as an NITConditional Cash Transfers and Tax Incidencerdquo AmericanEconomic Journal Economic Policy 2 (1) (2010) 177ndash208

11 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family Incomerdquo

12 Dube Arindrajit 2013 rdquoMinimum Wagesand the Distribution of Family IncomerdquoUnpublished working paper Available at httpsdldropboxusercontentcomu15038936Dube_MinimumWagesFamilyIncomespdf

13 As Dube explains in ldquoThe poverty of Minimum WageFactsrdquo the simulation approach underestimate stemsfrom a number of unwarranted assumptions includingthe range of actual wage increases and the accuracy ofwage data in the Current Population Survey The causal

approach does not make these assumptions

14 Allegretto Sylvia and others 2013 ldquoCredible ResearchDesigns for Minimum Wage Studiesrdquo Working Paper148-13 University of California Berkeley Institutefor Research on Labor and Employment Available athttpwwwirleberkeleyeduworkingpapers148-13pdf

15 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family In comerdquo

16 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo (2012)

17 Ibid

18 Ibid

19 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo

20 US Department of Agriculture ldquoSupplementalNutrition Assistance Program (SNAP)rdquo

21 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo

22 Ibid

23 We deviate from the Census Bureaursquos definition ofa family unit which is ldquotwo people or more (on e ofwhom is the householder) related by bir th marriageor adoption and residing togetherrdquo US Bureau ofthe Census ldquoCurrent Population Survey DefinitionsFamilyrdquo available at httpwwwcensusgovcpsabout

cpsdefhtml (last accessed February 2014) We countas a family unit any individual residing on his or herown two or more persons residing together whodo not belong to a family in the March CPS sampleare constructed as one family in our analysis For thepurposes of food stamp allocations the consumptionresulting from this transfer is probably distributed tofamily members (rather than household members ora single individual within the household) Howeversingle individuals canmdashand domdashreceive SNAPbenefits Excluding them would fail to make theanalysis reflective of the population at large

24 Strictly the family level linear probability modelpredicts the percentage-point decrease in theprobability that an individual family will receive SNAPpayments When applied to a large number of familieshowever we are able to interpret the coefficient asa decrease in the mean of enrollmentmdashthat is a

decrease in the enrollment ratemdashby applying the lawof iterated expectations

25 We generate expenditure predictions from theenrollment modelsmdashand conversely generateenrollment predictions from the expenditure modelmdashby assuming that expenditures per enrolled familyremains the same before and after the minimum wagechange In practice this is likely to be a conservativeestimatemdashthat is to underestimate the decrease inSNAP activity Average SNAP benefits per family willalso decrease as many families that remain eligible forSNAP experience income gains

26 Wage and Hour Division ldquoMinimum Wage Laws inthe States ndash Januar y 1 2014rdquo available at httpwwwdolgovwhdminwageamericahtm (last accessedFebruary 2014)

27 See for example Sylvia Allegretto and others ldquoCredibleResearch Designs for Minimum Wage Studiesrdquo WorkingPaper 148-113 (Berkeley California Institute forResearch on Labor and Employment 2013) available athttpwwwirleberkeleyeduworkingpapers148-13pdf

28 We will report these results in a forthcoming workingpaper

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4243

Endnotes | wwwamericanprogresso

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4343

The Center for American Progress is a nonpartisan research and educational institute

dedicated to promoting a strong just and free America that ensures opportunity

for all We believe that Americans are bound together by a common commitment to

these values and we aspire to ensure that our national policies reflect these values

We work to find progressive and pragmatic solutions to significant domestic and

international problems and develop policy proposals that foster a government that

is ldquoof the people by the people and for the peoplerdquo

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2443

References | wwwamericanprogresso

References

Allegreto Sylvia and ohers 2013 ldquoFas Food Povery Wages Te Public Cos o Low-Wage Jobsin he Fas-Food Indusryrdquo Universiy o Caliornia Berkeley Cener or Labor Research andEducaion Available a htplaborcenerberkeleyedupubliccossas_ood_povery_wage

Allegreto Sylvia and ohers 2013 ldquoCredible Research Designs or Minimum Wage Sudiesrdquo

Working Paper 148-13 Universiy o Caliornia Berkeley Insiue or Research on Labor andEmploymen Available a htpwwwirleberkeleyeduworkingpapers148-13pd

Biler Marianne and Hilary Hoynes 2013 ldquo Te More Tings Change he More Tey Say heSame Te Saey Ne Living Arrangemens and Povery in he Grea Recessionrdquo Working Paper19449 Naional Bureau o Economic Research

Congressional Budge Office 2012 ldquoTe Supplemenal Nuriion Assisance Programrdquo Washingon Available a htpwwwcbogovsiesdeaulfilescbofilesatachmens04-19-SNAPpd

991252 991252 991252 2014 ldquoTe Effec o a Minimum-Wage Increase on Employmen and Family Incomerdquo Washingon Available a htpwwwcbogovsiesdeaulfilescbofilesatachmens44995-MinimumWagepd

Dube Arindraji 2013 rdquoMinimum Wages and he Disribuion o Family Incomerdquo Unpublished working paper Available a htpsdldropboxuserconencomu15038936Dube_ MinimumWagesFamilyIncomespd

991252 991252 991252 2014 ldquoTe Povery o Minimum Wage Facsrdquo Arindraji Dube Blog January 22 Availablea htparindubecom20140122he-povery-o-minimum-wage-acs

Lee David and Emmanuel Saez 2012 ldquoOpimal Minimum Wage Policy in Compeiive LaborMarkesrdquo Journal o Public Economics 96 (9) 739ndash749

Neumark David and William Wascher 1992 ldquoEmploymen Effecs o Minimum and Subminimum Wages Panel Daa on Sae Minimum Wage Lawsrdquo Industrial and Labor Relations Review 46 (1)

55ndash81

Neumark David and William Wascher 2011 ldquoDoes a Higher Minimum Wage Enhance heEffeciveness o he Earned Income ax Credirdquo Industrial and Labor Relations Review 64 (4)712ndash746

Page Marianne Joanne Spez and Jane Millar 2005 ldquoDoes he Minimum Wage Affec WelareCaseloadsrdquo Journal o Policy Analysis and Management 24 (2) 273ndash295

Rohsein Jesse 2010 ldquoIs he EIC as Good as an NI Condiional Cash ransers and ax

Incidencerdquo American Economic Journal Economic Policy 2 (1) 177ndash208

Wage and Hours Division 2014 ldquoMinimum Wage Laws in he Saes ndash January 1 2014rdquo USDeparmen o Labor (htpwwwdolgovwhdminwageamericahm [February 2014])

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2543

22 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2643

Appendix A | wwwamericanprogresso

Appendix A

Pre-trend falsification check

Recen minimum wage research27 highlighs a common flaw in previous sudies

ailure o veriy ha he oucome variable is ree o negaive pre-exising rends I

or example SNAP aciviy was already rending down in saes ha raised heir

minimum wages beore hese changes came ino effec our regression analysis could

(misakenly) atribue ha reducion o he minimum wage We check or such

pre-rends by inroducing variables ha represen he prior yearrsquos value or leads o

he minimum wage I he model esimaes he minimum wage o have an effec on

he oucome variable beore he wage change wen ino effec hen an unobservedacor no he minimum wage change caused he change in SNAP aciviy

We es he specificaions above or pre-rends by including a one-year lead in

all hree specificaions We find ha he lead erms are small posiive and no

saisically significan indicaing ha he concurren minimum wage991252no

he wage level in prior periods991252is driving he observed changes in SNAP

oucomes28 In paricular he coefficien (sandard error) on he lead erm in

our preerred amily-level enrollmen regression is 011 and no significan

while he coefficien and sandard error o he conemporaneous minimum

wage is unchanged In he sae-level preerred enrollmen regression he

coefficien o he lead erm is again small (07) and i is no significan Te

corresponding coefficien on he lead erm in he sae-level expendiure

regression is 16 and is no significan Te posiive poin esimaes on hese lead

erms resuls no only rule ou disoring negaive pre-rends Tey also sugges

ha our main resuls may underesimae he rue effecs

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2743

24 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2843

Appendix B | wwwamericanprogresso

Appendix B

Model selection process

For boh he amily-level and sae-level models we es hree mehods o conrol

or unobserved geographic- and ime-varying characerisics as suggesed by he

minimum wage lieraure Firs we include only independen sae-specific fixed

effecs and year-specific fixed effecs Tis specificaion (specificaion 1) implicily

assumes ha amilies in any sae consiue an equally good saisical ldquoconrolrdquo

group or hose in any randomly chosen sae afer accouning or various

characerisics (median income and unemploymen rae among ohers) Similarly

simple ime fixed effecs assume ha amilies surveyed in any year can crediblyserve as a conrol group or amilies surveyed in every oher year o he sample

(1990 hrough 2012)

In oher words specificaion 1 assumes ha a saersquos immediae neighbor provides

no beter a couneracual or he effec o a minimum wage change han does a

sae across he counry We relax his resricive specificaion sequenially in wo

seps In specificaion 2 we replace simple year fixed effecs wih fixed effecs or

each Census divisionyear (capured as an addiional variable in he vecor By

using division-year effecs we remove he resricion ha amilies in each sae

are equally good saisical conrols or all oher amilies Raher we allow or he

possibiliy ha amilies in similar geographic regions (or example he Souh or

he Norheas) may be more similar o one anoher han amilies arher away

Finally in specificaion 3 we add sae-specific linear ime rends o he previous

specificaion Tus specificaion 3 is he mos rigorous model specificaion in ha

i allows or heerogeneiy along hree dimensions Ta is specificaion 3 allows

each sae o have is own ime-varying rends raher han imposing he resricion

ha saes evolve idenically over he 22 years in he sample

We begin building he heoreical specificaion above rom a se o simpleuncondiional models regression o SNAP aciviy (enrollmen or expendiures)

on he log o he minimum wage and a se o geographic- and ime-specific

effecs (specificaions 1 2 and 3 described above) As shown in ables 1ndash3 (or

specificaion 3) we hen add covariaes sequenially o hese models including

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2943

26 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

firs he vecor o amily-level conrols ollowed by each o several sae-level

covariaes in urn he unemploymen rae log o median-amily income and he

employmen-o-populaion raio Comparable resuls or specificaions 2 and 3

will be available in our orhcoming working paper

As expeced he simple uncondiional models indicae ha he relaionship beween he minimum wage and SNAP enrollmen i one exiss is a more complex

one influenced by oher acors In he uncondiional model he coefficien on he

variable o ineres991252he log o he minimum wage991252is small in magniude and no

saisically differen rom zero Once we accoun or he influence o labor marke

condiions and variaion in income levels on program paricipaion (by including

unemploymen rae and median-amily income conrol variables respecively)

he effec o he minimum wage on SNAP enrollmen is precisely esimaed Te

coefficien o he log minimum wage is slighly higher (-0042) in he amily-level

analysis han he coefficien (-031) in he sae-level analysis Te level o precision

is also higher in he amily-level analysis Tis is o be expeced when using 124million observaions compared o 1127

Te second dimension o model choice concerns he effec specificaion ables

7ndash9 compare he primary coefficiens o ineres or he SNAP enrollmen and

expendiure models For boh he enrollmen models he effec sizes are smalles

or specificaion 1 larges or specificaion 2 and inermediae beween hese wo

in specificaion 3 Recall ha Specificaion 3 conains sae-specific linear ime

rends in addiion o he census divisionyear conrols included in specificaion

2 In he amily-level enrollmen model he sandard error o he minimum wage

coefficien is smaller han in he oher wo specificaions Sandard errors on he

oher variables are much smaller in specificaions 2 and 3 han in specificaion

1 On he basis o coefficien significance (join and individual) specificaions 2

and 3 are sricly preerred in boh enrollmen models o specificaion 1 which

conains only sae and year fixed effecs

A concern wih specificaions 2 and 3 is ha rend conrols such as sae linear

rends may incorrecly absorb some o he delayed impac o a minimum wage

When we es his issue by including lagged minimum wages we do no find ha

delayed effecs are significan Anoher concern is ha more sauraed modelsuse less o he saisical variaion which could reduce he saisical power o

he resuls However he sandard errors or our more sauraed models are no

higher and are lower in some cases han or he less sauraed models Overall

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3043

Appendix B | wwwamericanprogresso

his evidence suppors our use o he sauraed model as he preerred model

specificaion Moreover Dubersquos 2013 sudy shows ha more sauraed models

perorm beter han models wih jus sae and ime fixed effecs

Te esimaed enrollmen regressions a boh he amily and sae levels show large

and saisically significan coefficiens Te esimaed minimum wage effec in heexpendiures regressions991252or which we have only sae-level daa991252is also large

and saisically significan

We do no use weighed regression or he sae-level models preerring o keep

analysis o he ldquoreamenrdquo (ha is o say a minimum wage change) appropriae

o he average sae raher han he average amily or individual I insead our

primary ineres were he impac o a minimum wage change on he average amily

or he average individual we migh choose o designae he number o amilies

in each sae or he sae populaion respecively as analyic weighs in order o

obain a coefficien beter suied or such inerence

TABLE 4

SNAP enrollment

Family-level linear probability

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0048 -0047 -0040 -0043 -0042

(0013) (0013) (001) (0008) (0009)

Unemployment rate (100) 0505 0420 0280

(0083) (0086) (0082)

Log median income -0057 -0039

(0011) (0011)

Employment-to-population ratio -0239

(0038)

N 1242022 1242022 1242022 1242022 1242022

plt01 plt005 plt001

Note Robust standard errors are in parentheses Observations are clustered at the state level The outcome variable is binary and equal to one if a family is enrolledin SNAP All models include state fixed effects Census division x-year fixed effects and state-specific linear time trends All specifications except 3a include additional

controls for family size race and marital status of the family head presence of children and presence of an adult maleSource Annual data from the March Current Population Survey for 1990 to 2012 estimation includes Current Population Survey probability weights

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3143

28 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

TABLE 5

SNAP enrollment

State-level linear regression

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0028 -0024 -0026 -0031 -0031

(0014) (0013) (0013) (0012) (0012)

Unemployment rate (100) 0625 0466 0320 0339

(0087) (0088) (0085) (0083)

Log median income -0090 -0065 -0061

(0013) (0013) (0013)

Employment-to-population ratio -0282 -0248

(0037) (0038)

N 1127 1127 1127 1127 1127

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the SNAP enrollment rate All models include state fixed effects Census division x-year fixedeffects and state-specific linear time trends

Source Annual data from the March Current Population Survey for 1990 to 2012

TABLE 6

SNAP expenditures

State-level linear regression

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0181 -0149 -0156 -0153 -0190

(011) (0103) (0102) (0103) (0103)

Unemployment rate (100) 4587 4152 4232 4313

(0622) (0621) (0633) (0628)

Log median income -0246 -0261 -0294

(0075) (0078) (0078)

Employment-to-population ratio 0155 0244

(0237) (024)

N 1127 1127 1127 1127 1127

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the natural log of state SNAP expenditures per capita for 1990 to 2012 All models include state

fixed effects Census division x-year fixed effects and state-specific linear time trends

Source Annual data from Bureau of Economic Analysis national income and product accounts tables for 1990 to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3243

Appendix B | wwwamericanprogresso

TABLE 7

Comparison of specifications SNAP enrollment

Family-level linear probability

(1) (2) (3)

Log minimum wage -0028 -0049 -0042

(0014) (0017) (0009)

Unemployment rate (100) 0275 0297 0280

(0161) (0076) (0082)

Log median income -0077 -0055 -0039

(0014) (0012) (0011)

Employment-to-population ratio -0238 -0250 -0239

(0054) (004) (0038)

N 1242022 1242022 1242022

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses Observations are clustered at the state level The outcomevariable is binary or equal to one if a family is enrolled in SNAP All specifications include additional controls forfamily size race and marital status of the family head presence of children and presence of an adult male

Source Annual data from the March Current Population Survey for 1990 to 2012 estimation includes CurrentPopulation Survey probability weights

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3343

30 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

TABLE 8

Comparison of specifications SNAP enrollment

State-level linear regression

(1) (2) (3)

Log minimum wage -0019 -0035 -0031

(0009) (0012) (0012)

Unemployment rate (100) 0401 0370 0339

(0063) (0077) (0083)

Log median income -0081 -0073 -0061

(0011) (0013) (0013)

Employment-to-population ratio -0183 -0222 -0248

(0039) (0039) (0038)

N 1127 1127 1127

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the SNAP enrollment rate Allregressions include the share of households with children and the racial shares of the population

Source Annual data from the March Current Population Survey for 1990 to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3443

Appendix B | wwwamericanprogresso

TABLE 9

Comparison of specifications SNAP expenditures

State-level linear regression

(1) (2) (3)

Log minimum wage -0121 -0203 -0190

(0075) (0103) (0103)

Unemployment rate (100) 5292 5152 4313

(0464) (0576) (0628)

Log median income -0437 -0417 -0294

(008) (0086) (0078)

Employment-to-population ratio -0040 -0220 0244

(0261) (0260) (0240)

N 1127 1127 1127

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the natural log of state SNAPexpenditures per capita All regressions include the share of households with children and the racial shares ofthe state population

Source Annual data from Bureau of Economic Analysis national income and product accounts tables for 1990to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3543

32 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3643

Appendix C | wwwamericanprogresso

Appendix C Harkin-Miller

policy simulation results

StateMinimum wage

(2014)

SNAP enrollment (2012) Predicted change in SNAP enrollment

RateTotal

(persons)

Model 1 family

enrollment (linearprobability)

Model 2 state

enrollment (linearregression)

Model 3 s

expenditu(linear regre

Alabama $725 164 788682 -66922 -58897 -58906

Alaska $775 120 87436 -8104 -6887 -3288

Arizona $790 201 1319323 -67435 -56738 -64356

Arkansas $725 78 230489 -40977 -36063 -35248

California $800 206 7813680 -371131 -310222 -18223

Colorado $800 164 853155 -50684 -42365 -23926

Connecticut $870 91 326621 -22456 -17975 -13711

Delaware $725 186 170262 -12739 -11211 -10647

District of Columbia $825 133 84009 -5370 -4417 -3632

Florida $793 166 3208026 -195813 -164426 -13046

Georgia $725 160 1586336 -137741 -121224 -11004

Hawaii $725 96 133662 -19310 -16995 -14933

Idaho $725 92 147501 -22165 -19507 -15809

Illinois $825 95 1225084 -109088 -89742 -70955

Indiana $725 125 816233 -90818 -79928 -83985

Iowa $725 155 478011 -42716 -37594 -28556

Kansas $725 135 388269 -40082 -35275 -27461

Kentucky $725 130 568821 -60840 -53544 -52259

Louisiana $725 149 683832 -63929 -56263 -66083

Maine $750 77 101976 -16567 -14323 -15234

Maryland $725 144 846415 -81748 -71946 -38370

Massachusetts $800 130 864721 -64902 -54251 -42913

Michigan $740 146 1439141 -128801 -112140 -11022

Minnesota $725 133 713646 -74730 -65769 -37878

TABLE 10

SNAP enrollments Predicted changes by state under the Harkin-Miller bill

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3743

34 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

StateMinimum wage

(2014)

SNAP enrollment (2012) Predicted change in SNAP enrollment

RateTotal

(persons)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 state

expenditures

(linear regression)

Mississippi $725 129 386501 -41486 -36511 -46467

Missouri $750 172 1036182 -75131 -64952 -56244

Montana $790 132 132452 -10350 -8708 -5846

Nebraska $725 124 230683 -25773 -22683 -12189

Nevada $825 162 446035 -23349 -19209 -11894

New Hampshire $725 127 168404 -18359 -16157 -5735

New Jersey $825 160 1416666 -75175 -61843 -28236

New Mexico $750 149 310896 -25983 -22463 -22512

New York $800 192 3763553 -191193 -159815 -142182

North Carolina $725 174 1697193 -135417 -119179 -113503

North Dakota $725 87 61225 -9743 -8574 -4021

Ohio $795 143 1647345 -115869 -97169 -88580

Oklahoma $725 129 494053 -53006 -46650 -46854

Oregon $910 124 485326 -17036 -13328 -16398

Pennsylvania $725 161 2053643 -177315 -156052 -125586

Rhode Island $800 156 163730 -10258 -8574 -8698

South Carolina $725 94 445277 -65614 -57746 -50304

South Dakota $725 208 173749 -11586 -10197 -7458

Tennessee $725 142 914903 -89667 -78915 -99134

Texas $725 110 2863779 -362018 -318607 -253285

Utah $725 88 251107 -39658 -34902 -19390

Vermont $873 156 97792 -3823 -3055 -2475

Virginia $725 101 829771 -113723 -100086 -58212

Washington $932 72 496934 -23221 -17947 -17756

West Virginia $725 58 107875 -25792 -22699 -21665

Wisconsin $725 75 427822 -79521 -69986 -53210

Wyoming $725 164 94590 -8010 -7050 -3104

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3843

Appendix C | wwwamericanprogresso

StateMinimum wage

(2014)

SNAP expenditures (2012)Predicted change in SNAP expenditures

(millions of dollars)

Per resident

(millions of

dollars)

Total

(millions of

dollars)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 s

expenditu

(linear regre

Alabama $725 $290 $1397 -$1185 -$1043 -$1043

Alaska $775 $253 $185 -$262 -$223 -$106

Arizona $790 $257 $1686 -$935 -$787 -$892

Arkansas $725 $248 $733 -$636 -$560 -$547

California $800 $189 $7164 -$7276 -$6082 -$3573

Colorado $800 $157 $816 -$862 -$721 -$407

Connecticut $870 $191 $686 -$343 -$275 -$210

Delaware $725 $250 $229 -$205 -$180 -$171

District of Columbia $825 $366 $232 -$146 -$120 -$99

Florida $793 $294 $5676 -$4429 -$3719 -$2951

Georgia $725 $317 $3140 -$2936 -$2584 -$2346

Hawaii $725 $335 $465 -$449 -$395 -$347

Idaho $725 $225 $359 -$376 -$331 -$268

Illinois $825 $249 $3200 -$2096 -$1725 -$1364

Indiana $725 $220 $1439 -$1162 -$1023 -$1075

Iowa $725 $192 $589 -$658 -$579 -$440

Kansas $725 $159 $460 -$502 -$441 -$344

Kentucky $725 $298 $1303 -$1133 -$997 -$973

Louisiana $725 $315 $1450 -$1047 -$922 -$1083

Maine $750 $281 $373 -$267 -$231 -$246

Maryland $725 $188 $1109 -$1765 -$1553 -$828

Massachusetts $800 $206 $1366 -$1030 -$861 -$681

Michigan $740 $300 $2963 -$2400 -$2090 -$2054

Minnesota $725 $140 $755 -$1113 -$980 -$564

Mississippi $725 $326 $973 -$649 -$571 -$726

Missouri $750 $241 $1452 -$1278 -$1104 -$956

Montana $790 $190 $191 -$179 -$151 -$101

Nebraska $725 $140 $259 -$409 -$360 -$194

Nevada $825 $191 $527 -$441 -$363 -$225

New Hampshire $725 $126 $167 -$399 -$351 -$125

TABLE 11

SNAP expenditures Predicted changes by state under the Harkin-Miller bill

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3943

36 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

StateMinimum wage

(2014)

SNAP expenditures (2012)Predicted change in SNAP expenditures

(millions of dollars)

Per resident

(millions of

dollars)

Total

(millions of

dollars)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 state

expenditures

(linear regression)

New Jersey $825 $160 $1420 -$1611 -$1325 -$605

New Mexico $750 $324 $675 -$513 -$444 -$445New York $800 $287 $5616 -$3766 -$3148 -$2801

North Carolina $725 $252 $2454 -$2187 -$1925 -$1833

North Dakota $725 $128 $90 -$162 -$143 -$67

Ohio $795 $259 $2995 -$2013 -$1688 -$1539

Oklahoma $725 $248 $945 -$799 -$703 -$706

Oregon $910 $322 $1255 -$272 -$213 -$262

Pennsylvania $725 $218 $2779 -$2930 -$2579 -$2075

Rhode Island $800 $280 $294 -$173 -$144 -$147

South Carolina $725 $291 $1373 -$1337 -$1177 -$1025South Dakota $725 $198 $165 -$192 -$169 -$123

Tennessee $725 $324 $2091 -$1413 -$1243 -$1562

Texas $725 $230 $5997 -$6402 -$5634 -$4479

Utah $725 $141 $402 -$614 -$541 -$300

Vermont $873 $230 $144 -$66 -$53 -$43

Virginia $725 $173 $1413 -$2062 -$1815 -$1056

Washington $932 $244 $1682 -$350 -$270 -$267

West Virginia $725 $273 $508 -$451 -$397 -$379

Wisconsin $725 $204 $1166 -$1302 -$1146 -$871Wyoming $725 $95 $55 -$105 -$93 -$41

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4043

Appendix C | wwwamericanprogresso

FIGURE 2

Raising the minimum wage to $1010would cut taxpayer costs in every state

Predicted decreases in cost and enrollment

in SNAP in 50 states

$200+$51ndash$100

$101ndash$200

0ndash$25

$26ndash$50

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4143

38 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

Endnotes

1 US Department of Agriculture ldquoSupplementalNutrition Assistance Program (SNAP)rdquo available athttpwwwfnsusdagovsnapeligibility (last accessedFebruary 2014)

2 For this initial analysis we do not consider Harkin-Millerrsquos increase in subminimum wages for tippedworkers To do s o would increase the estimated SNAP

savings by an unknown amount

3 The Congressional Budget Office estimates thatworkers currently earning between $1010 and $1150per hour would see their wages rise under the Harkin-Miller proposal Congressional Budget O ffice ldquoTheEffects of a Minimum Wage Increase on Employmentand Family Incomerdquo (2014)

4 Marianne Page Joanne Spetz and Jane Millar ldquoDoesthe Minimum Wage Affect Welfare Caseloadsrdquo Journalof Policy Analysis and Management 24 (2) (2005)273ndash295

5 Marianne Bitler and Hilary Hoynes ldquoThe More ThingsChange the More They Stay the Same The SafetyNet Living Arrangements and Poverty in the GreatRecessionrdquo NBER Working Paper 194 49 2013

6 Sylvia Allegretto and others ldquoFast Food Poverty Wages The Public Cost of Low-Wage Jobs in the Fast-FoodIndustryrdquo (Berkeley California Center for LaborResearch and Education 2013) available at httplaborcenterberkeleyedupubliccostsfast_food_poverty_wage

7 David Neumark and William Wascher ldquoDoes a HigherMinimum Wage Enhance the Effectiveness of theEarned Income Tax Creditrdquo Industrial and LaborRelations Review 64 (4) (2011) 712ndash746

8 David Lee and Emmanuel Saez ldquoOptimal MinimumWage Policy in Competitive Labor Marketsrdquo Journal ofPublic Economics 96 (9) (2012) 739ndash749

9 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family IncomerdquoFebruary 2014

10 Jesse Rothstein ldquoIs the EITC as Good as an NITConditional Cash Transfers and Tax Incidencerdquo AmericanEconomic Journal Economic Policy 2 (1) (2010) 177ndash208

11 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family Incomerdquo

12 Dube Arindrajit 2013 rdquoMinimum Wagesand the Distribution of Family IncomerdquoUnpublished working paper Available at httpsdldropboxusercontentcomu15038936Dube_MinimumWagesFamilyIncomespdf

13 As Dube explains in ldquoThe poverty of Minimum WageFactsrdquo the simulation approach underestimate stemsfrom a number of unwarranted assumptions includingthe range of actual wage increases and the accuracy ofwage data in the Current Population Survey The causal

approach does not make these assumptions

14 Allegretto Sylvia and others 2013 ldquoCredible ResearchDesigns for Minimum Wage Studiesrdquo Working Paper148-13 University of California Berkeley Institutefor Research on Labor and Employment Available athttpwwwirleberkeleyeduworkingpapers148-13pdf

15 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family In comerdquo

16 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo (2012)

17 Ibid

18 Ibid

19 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo

20 US Department of Agriculture ldquoSupplementalNutrition Assistance Program (SNAP)rdquo

21 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo

22 Ibid

23 We deviate from the Census Bureaursquos definition ofa family unit which is ldquotwo people or more (on e ofwhom is the householder) related by bir th marriageor adoption and residing togetherrdquo US Bureau ofthe Census ldquoCurrent Population Survey DefinitionsFamilyrdquo available at httpwwwcensusgovcpsabout

cpsdefhtml (last accessed February 2014) We countas a family unit any individual residing on his or herown two or more persons residing together whodo not belong to a family in the March CPS sampleare constructed as one family in our analysis For thepurposes of food stamp allocations the consumptionresulting from this transfer is probably distributed tofamily members (rather than household members ora single individual within the household) Howeversingle individuals canmdashand domdashreceive SNAPbenefits Excluding them would fail to make theanalysis reflective of the population at large

24 Strictly the family level linear probability modelpredicts the percentage-point decrease in theprobability that an individual family will receive SNAPpayments When applied to a large number of familieshowever we are able to interpret the coefficient asa decrease in the mean of enrollmentmdashthat is a

decrease in the enrollment ratemdashby applying the lawof iterated expectations

25 We generate expenditure predictions from theenrollment modelsmdashand conversely generateenrollment predictions from the expenditure modelmdashby assuming that expenditures per enrolled familyremains the same before and after the minimum wagechange In practice this is likely to be a conservativeestimatemdashthat is to underestimate the decrease inSNAP activity Average SNAP benefits per family willalso decrease as many families that remain eligible forSNAP experience income gains

26 Wage and Hour Division ldquoMinimum Wage Laws inthe States ndash Januar y 1 2014rdquo available at httpwwwdolgovwhdminwageamericahtm (last accessedFebruary 2014)

27 See for example Sylvia Allegretto and others ldquoCredibleResearch Designs for Minimum Wage Studiesrdquo WorkingPaper 148-113 (Berkeley California Institute forResearch on Labor and Employment 2013) available athttpwwwirleberkeleyeduworkingpapers148-13pdf

28 We will report these results in a forthcoming workingpaper

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4243

Endnotes | wwwamericanprogresso

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4343

The Center for American Progress is a nonpartisan research and educational institute

dedicated to promoting a strong just and free America that ensures opportunity

for all We believe that Americans are bound together by a common commitment to

these values and we aspire to ensure that our national policies reflect these values

We work to find progressive and pragmatic solutions to significant domestic and

international problems and develop policy proposals that foster a government that

is ldquoof the people by the people and for the peoplerdquo

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2543

22 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2643

Appendix A | wwwamericanprogresso

Appendix A

Pre-trend falsification check

Recen minimum wage research27 highlighs a common flaw in previous sudies

ailure o veriy ha he oucome variable is ree o negaive pre-exising rends I

or example SNAP aciviy was already rending down in saes ha raised heir

minimum wages beore hese changes came ino effec our regression analysis could

(misakenly) atribue ha reducion o he minimum wage We check or such

pre-rends by inroducing variables ha represen he prior yearrsquos value or leads o

he minimum wage I he model esimaes he minimum wage o have an effec on

he oucome variable beore he wage change wen ino effec hen an unobservedacor no he minimum wage change caused he change in SNAP aciviy

We es he specificaions above or pre-rends by including a one-year lead in

all hree specificaions We find ha he lead erms are small posiive and no

saisically significan indicaing ha he concurren minimum wage991252no

he wage level in prior periods991252is driving he observed changes in SNAP

oucomes28 In paricular he coefficien (sandard error) on he lead erm in

our preerred amily-level enrollmen regression is 011 and no significan

while he coefficien and sandard error o he conemporaneous minimum

wage is unchanged In he sae-level preerred enrollmen regression he

coefficien o he lead erm is again small (07) and i is no significan Te

corresponding coefficien on he lead erm in he sae-level expendiure

regression is 16 and is no significan Te posiive poin esimaes on hese lead

erms resuls no only rule ou disoring negaive pre-rends Tey also sugges

ha our main resuls may underesimae he rue effecs

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2743

24 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2843

Appendix B | wwwamericanprogresso

Appendix B

Model selection process

For boh he amily-level and sae-level models we es hree mehods o conrol

or unobserved geographic- and ime-varying characerisics as suggesed by he

minimum wage lieraure Firs we include only independen sae-specific fixed

effecs and year-specific fixed effecs Tis specificaion (specificaion 1) implicily

assumes ha amilies in any sae consiue an equally good saisical ldquoconrolrdquo

group or hose in any randomly chosen sae afer accouning or various

characerisics (median income and unemploymen rae among ohers) Similarly

simple ime fixed effecs assume ha amilies surveyed in any year can crediblyserve as a conrol group or amilies surveyed in every oher year o he sample

(1990 hrough 2012)

In oher words specificaion 1 assumes ha a saersquos immediae neighbor provides

no beter a couneracual or he effec o a minimum wage change han does a

sae across he counry We relax his resricive specificaion sequenially in wo

seps In specificaion 2 we replace simple year fixed effecs wih fixed effecs or

each Census divisionyear (capured as an addiional variable in he vecor By

using division-year effecs we remove he resricion ha amilies in each sae

are equally good saisical conrols or all oher amilies Raher we allow or he

possibiliy ha amilies in similar geographic regions (or example he Souh or

he Norheas) may be more similar o one anoher han amilies arher away

Finally in specificaion 3 we add sae-specific linear ime rends o he previous

specificaion Tus specificaion 3 is he mos rigorous model specificaion in ha

i allows or heerogeneiy along hree dimensions Ta is specificaion 3 allows

each sae o have is own ime-varying rends raher han imposing he resricion

ha saes evolve idenically over he 22 years in he sample

We begin building he heoreical specificaion above rom a se o simpleuncondiional models regression o SNAP aciviy (enrollmen or expendiures)

on he log o he minimum wage and a se o geographic- and ime-specific

effecs (specificaions 1 2 and 3 described above) As shown in ables 1ndash3 (or

specificaion 3) we hen add covariaes sequenially o hese models including

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2943

26 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

firs he vecor o amily-level conrols ollowed by each o several sae-level

covariaes in urn he unemploymen rae log o median-amily income and he

employmen-o-populaion raio Comparable resuls or specificaions 2 and 3

will be available in our orhcoming working paper

As expeced he simple uncondiional models indicae ha he relaionship beween he minimum wage and SNAP enrollmen i one exiss is a more complex

one influenced by oher acors In he uncondiional model he coefficien on he

variable o ineres991252he log o he minimum wage991252is small in magniude and no

saisically differen rom zero Once we accoun or he influence o labor marke

condiions and variaion in income levels on program paricipaion (by including

unemploymen rae and median-amily income conrol variables respecively)

he effec o he minimum wage on SNAP enrollmen is precisely esimaed Te

coefficien o he log minimum wage is slighly higher (-0042) in he amily-level

analysis han he coefficien (-031) in he sae-level analysis Te level o precision

is also higher in he amily-level analysis Tis is o be expeced when using 124million observaions compared o 1127

Te second dimension o model choice concerns he effec specificaion ables

7ndash9 compare he primary coefficiens o ineres or he SNAP enrollmen and

expendiure models For boh he enrollmen models he effec sizes are smalles

or specificaion 1 larges or specificaion 2 and inermediae beween hese wo

in specificaion 3 Recall ha Specificaion 3 conains sae-specific linear ime

rends in addiion o he census divisionyear conrols included in specificaion

2 In he amily-level enrollmen model he sandard error o he minimum wage

coefficien is smaller han in he oher wo specificaions Sandard errors on he

oher variables are much smaller in specificaions 2 and 3 han in specificaion

1 On he basis o coefficien significance (join and individual) specificaions 2

and 3 are sricly preerred in boh enrollmen models o specificaion 1 which

conains only sae and year fixed effecs

A concern wih specificaions 2 and 3 is ha rend conrols such as sae linear

rends may incorrecly absorb some o he delayed impac o a minimum wage

When we es his issue by including lagged minimum wages we do no find ha

delayed effecs are significan Anoher concern is ha more sauraed modelsuse less o he saisical variaion which could reduce he saisical power o

he resuls However he sandard errors or our more sauraed models are no

higher and are lower in some cases han or he less sauraed models Overall

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3043

Appendix B | wwwamericanprogresso

his evidence suppors our use o he sauraed model as he preerred model

specificaion Moreover Dubersquos 2013 sudy shows ha more sauraed models

perorm beter han models wih jus sae and ime fixed effecs

Te esimaed enrollmen regressions a boh he amily and sae levels show large

and saisically significan coefficiens Te esimaed minimum wage effec in heexpendiures regressions991252or which we have only sae-level daa991252is also large

and saisically significan

We do no use weighed regression or he sae-level models preerring o keep

analysis o he ldquoreamenrdquo (ha is o say a minimum wage change) appropriae

o he average sae raher han he average amily or individual I insead our

primary ineres were he impac o a minimum wage change on he average amily

or he average individual we migh choose o designae he number o amilies

in each sae or he sae populaion respecively as analyic weighs in order o

obain a coefficien beter suied or such inerence

TABLE 4

SNAP enrollment

Family-level linear probability

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0048 -0047 -0040 -0043 -0042

(0013) (0013) (001) (0008) (0009)

Unemployment rate (100) 0505 0420 0280

(0083) (0086) (0082)

Log median income -0057 -0039

(0011) (0011)

Employment-to-population ratio -0239

(0038)

N 1242022 1242022 1242022 1242022 1242022

plt01 plt005 plt001

Note Robust standard errors are in parentheses Observations are clustered at the state level The outcome variable is binary and equal to one if a family is enrolledin SNAP All models include state fixed effects Census division x-year fixed effects and state-specific linear time trends All specifications except 3a include additional

controls for family size race and marital status of the family head presence of children and presence of an adult maleSource Annual data from the March Current Population Survey for 1990 to 2012 estimation includes Current Population Survey probability weights

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3143

28 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

TABLE 5

SNAP enrollment

State-level linear regression

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0028 -0024 -0026 -0031 -0031

(0014) (0013) (0013) (0012) (0012)

Unemployment rate (100) 0625 0466 0320 0339

(0087) (0088) (0085) (0083)

Log median income -0090 -0065 -0061

(0013) (0013) (0013)

Employment-to-population ratio -0282 -0248

(0037) (0038)

N 1127 1127 1127 1127 1127

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the SNAP enrollment rate All models include state fixed effects Census division x-year fixedeffects and state-specific linear time trends

Source Annual data from the March Current Population Survey for 1990 to 2012

TABLE 6

SNAP expenditures

State-level linear regression

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0181 -0149 -0156 -0153 -0190

(011) (0103) (0102) (0103) (0103)

Unemployment rate (100) 4587 4152 4232 4313

(0622) (0621) (0633) (0628)

Log median income -0246 -0261 -0294

(0075) (0078) (0078)

Employment-to-population ratio 0155 0244

(0237) (024)

N 1127 1127 1127 1127 1127

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the natural log of state SNAP expenditures per capita for 1990 to 2012 All models include state

fixed effects Census division x-year fixed effects and state-specific linear time trends

Source Annual data from Bureau of Economic Analysis national income and product accounts tables for 1990 to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3243

Appendix B | wwwamericanprogresso

TABLE 7

Comparison of specifications SNAP enrollment

Family-level linear probability

(1) (2) (3)

Log minimum wage -0028 -0049 -0042

(0014) (0017) (0009)

Unemployment rate (100) 0275 0297 0280

(0161) (0076) (0082)

Log median income -0077 -0055 -0039

(0014) (0012) (0011)

Employment-to-population ratio -0238 -0250 -0239

(0054) (004) (0038)

N 1242022 1242022 1242022

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses Observations are clustered at the state level The outcomevariable is binary or equal to one if a family is enrolled in SNAP All specifications include additional controls forfamily size race and marital status of the family head presence of children and presence of an adult male

Source Annual data from the March Current Population Survey for 1990 to 2012 estimation includes CurrentPopulation Survey probability weights

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3343

30 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

TABLE 8

Comparison of specifications SNAP enrollment

State-level linear regression

(1) (2) (3)

Log minimum wage -0019 -0035 -0031

(0009) (0012) (0012)

Unemployment rate (100) 0401 0370 0339

(0063) (0077) (0083)

Log median income -0081 -0073 -0061

(0011) (0013) (0013)

Employment-to-population ratio -0183 -0222 -0248

(0039) (0039) (0038)

N 1127 1127 1127

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the SNAP enrollment rate Allregressions include the share of households with children and the racial shares of the population

Source Annual data from the March Current Population Survey for 1990 to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3443

Appendix B | wwwamericanprogresso

TABLE 9

Comparison of specifications SNAP expenditures

State-level linear regression

(1) (2) (3)

Log minimum wage -0121 -0203 -0190

(0075) (0103) (0103)

Unemployment rate (100) 5292 5152 4313

(0464) (0576) (0628)

Log median income -0437 -0417 -0294

(008) (0086) (0078)

Employment-to-population ratio -0040 -0220 0244

(0261) (0260) (0240)

N 1127 1127 1127

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the natural log of state SNAPexpenditures per capita All regressions include the share of households with children and the racial shares ofthe state population

Source Annual data from Bureau of Economic Analysis national income and product accounts tables for 1990to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3543

32 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3643

Appendix C | wwwamericanprogresso

Appendix C Harkin-Miller

policy simulation results

StateMinimum wage

(2014)

SNAP enrollment (2012) Predicted change in SNAP enrollment

RateTotal

(persons)

Model 1 family

enrollment (linearprobability)

Model 2 state

enrollment (linearregression)

Model 3 s

expenditu(linear regre

Alabama $725 164 788682 -66922 -58897 -58906

Alaska $775 120 87436 -8104 -6887 -3288

Arizona $790 201 1319323 -67435 -56738 -64356

Arkansas $725 78 230489 -40977 -36063 -35248

California $800 206 7813680 -371131 -310222 -18223

Colorado $800 164 853155 -50684 -42365 -23926

Connecticut $870 91 326621 -22456 -17975 -13711

Delaware $725 186 170262 -12739 -11211 -10647

District of Columbia $825 133 84009 -5370 -4417 -3632

Florida $793 166 3208026 -195813 -164426 -13046

Georgia $725 160 1586336 -137741 -121224 -11004

Hawaii $725 96 133662 -19310 -16995 -14933

Idaho $725 92 147501 -22165 -19507 -15809

Illinois $825 95 1225084 -109088 -89742 -70955

Indiana $725 125 816233 -90818 -79928 -83985

Iowa $725 155 478011 -42716 -37594 -28556

Kansas $725 135 388269 -40082 -35275 -27461

Kentucky $725 130 568821 -60840 -53544 -52259

Louisiana $725 149 683832 -63929 -56263 -66083

Maine $750 77 101976 -16567 -14323 -15234

Maryland $725 144 846415 -81748 -71946 -38370

Massachusetts $800 130 864721 -64902 -54251 -42913

Michigan $740 146 1439141 -128801 -112140 -11022

Minnesota $725 133 713646 -74730 -65769 -37878

TABLE 10

SNAP enrollments Predicted changes by state under the Harkin-Miller bill

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3743

34 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

StateMinimum wage

(2014)

SNAP enrollment (2012) Predicted change in SNAP enrollment

RateTotal

(persons)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 state

expenditures

(linear regression)

Mississippi $725 129 386501 -41486 -36511 -46467

Missouri $750 172 1036182 -75131 -64952 -56244

Montana $790 132 132452 -10350 -8708 -5846

Nebraska $725 124 230683 -25773 -22683 -12189

Nevada $825 162 446035 -23349 -19209 -11894

New Hampshire $725 127 168404 -18359 -16157 -5735

New Jersey $825 160 1416666 -75175 -61843 -28236

New Mexico $750 149 310896 -25983 -22463 -22512

New York $800 192 3763553 -191193 -159815 -142182

North Carolina $725 174 1697193 -135417 -119179 -113503

North Dakota $725 87 61225 -9743 -8574 -4021

Ohio $795 143 1647345 -115869 -97169 -88580

Oklahoma $725 129 494053 -53006 -46650 -46854

Oregon $910 124 485326 -17036 -13328 -16398

Pennsylvania $725 161 2053643 -177315 -156052 -125586

Rhode Island $800 156 163730 -10258 -8574 -8698

South Carolina $725 94 445277 -65614 -57746 -50304

South Dakota $725 208 173749 -11586 -10197 -7458

Tennessee $725 142 914903 -89667 -78915 -99134

Texas $725 110 2863779 -362018 -318607 -253285

Utah $725 88 251107 -39658 -34902 -19390

Vermont $873 156 97792 -3823 -3055 -2475

Virginia $725 101 829771 -113723 -100086 -58212

Washington $932 72 496934 -23221 -17947 -17756

West Virginia $725 58 107875 -25792 -22699 -21665

Wisconsin $725 75 427822 -79521 -69986 -53210

Wyoming $725 164 94590 -8010 -7050 -3104

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3843

Appendix C | wwwamericanprogresso

StateMinimum wage

(2014)

SNAP expenditures (2012)Predicted change in SNAP expenditures

(millions of dollars)

Per resident

(millions of

dollars)

Total

(millions of

dollars)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 s

expenditu

(linear regre

Alabama $725 $290 $1397 -$1185 -$1043 -$1043

Alaska $775 $253 $185 -$262 -$223 -$106

Arizona $790 $257 $1686 -$935 -$787 -$892

Arkansas $725 $248 $733 -$636 -$560 -$547

California $800 $189 $7164 -$7276 -$6082 -$3573

Colorado $800 $157 $816 -$862 -$721 -$407

Connecticut $870 $191 $686 -$343 -$275 -$210

Delaware $725 $250 $229 -$205 -$180 -$171

District of Columbia $825 $366 $232 -$146 -$120 -$99

Florida $793 $294 $5676 -$4429 -$3719 -$2951

Georgia $725 $317 $3140 -$2936 -$2584 -$2346

Hawaii $725 $335 $465 -$449 -$395 -$347

Idaho $725 $225 $359 -$376 -$331 -$268

Illinois $825 $249 $3200 -$2096 -$1725 -$1364

Indiana $725 $220 $1439 -$1162 -$1023 -$1075

Iowa $725 $192 $589 -$658 -$579 -$440

Kansas $725 $159 $460 -$502 -$441 -$344

Kentucky $725 $298 $1303 -$1133 -$997 -$973

Louisiana $725 $315 $1450 -$1047 -$922 -$1083

Maine $750 $281 $373 -$267 -$231 -$246

Maryland $725 $188 $1109 -$1765 -$1553 -$828

Massachusetts $800 $206 $1366 -$1030 -$861 -$681

Michigan $740 $300 $2963 -$2400 -$2090 -$2054

Minnesota $725 $140 $755 -$1113 -$980 -$564

Mississippi $725 $326 $973 -$649 -$571 -$726

Missouri $750 $241 $1452 -$1278 -$1104 -$956

Montana $790 $190 $191 -$179 -$151 -$101

Nebraska $725 $140 $259 -$409 -$360 -$194

Nevada $825 $191 $527 -$441 -$363 -$225

New Hampshire $725 $126 $167 -$399 -$351 -$125

TABLE 11

SNAP expenditures Predicted changes by state under the Harkin-Miller bill

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3943

36 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

StateMinimum wage

(2014)

SNAP expenditures (2012)Predicted change in SNAP expenditures

(millions of dollars)

Per resident

(millions of

dollars)

Total

(millions of

dollars)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 state

expenditures

(linear regression)

New Jersey $825 $160 $1420 -$1611 -$1325 -$605

New Mexico $750 $324 $675 -$513 -$444 -$445New York $800 $287 $5616 -$3766 -$3148 -$2801

North Carolina $725 $252 $2454 -$2187 -$1925 -$1833

North Dakota $725 $128 $90 -$162 -$143 -$67

Ohio $795 $259 $2995 -$2013 -$1688 -$1539

Oklahoma $725 $248 $945 -$799 -$703 -$706

Oregon $910 $322 $1255 -$272 -$213 -$262

Pennsylvania $725 $218 $2779 -$2930 -$2579 -$2075

Rhode Island $800 $280 $294 -$173 -$144 -$147

South Carolina $725 $291 $1373 -$1337 -$1177 -$1025South Dakota $725 $198 $165 -$192 -$169 -$123

Tennessee $725 $324 $2091 -$1413 -$1243 -$1562

Texas $725 $230 $5997 -$6402 -$5634 -$4479

Utah $725 $141 $402 -$614 -$541 -$300

Vermont $873 $230 $144 -$66 -$53 -$43

Virginia $725 $173 $1413 -$2062 -$1815 -$1056

Washington $932 $244 $1682 -$350 -$270 -$267

West Virginia $725 $273 $508 -$451 -$397 -$379

Wisconsin $725 $204 $1166 -$1302 -$1146 -$871Wyoming $725 $95 $55 -$105 -$93 -$41

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4043

Appendix C | wwwamericanprogresso

FIGURE 2

Raising the minimum wage to $1010would cut taxpayer costs in every state

Predicted decreases in cost and enrollment

in SNAP in 50 states

$200+$51ndash$100

$101ndash$200

0ndash$25

$26ndash$50

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4143

38 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

Endnotes

1 US Department of Agriculture ldquoSupplementalNutrition Assistance Program (SNAP)rdquo available athttpwwwfnsusdagovsnapeligibility (last accessedFebruary 2014)

2 For this initial analysis we do not consider Harkin-Millerrsquos increase in subminimum wages for tippedworkers To do s o would increase the estimated SNAP

savings by an unknown amount

3 The Congressional Budget Office estimates thatworkers currently earning between $1010 and $1150per hour would see their wages rise under the Harkin-Miller proposal Congressional Budget O ffice ldquoTheEffects of a Minimum Wage Increase on Employmentand Family Incomerdquo (2014)

4 Marianne Page Joanne Spetz and Jane Millar ldquoDoesthe Minimum Wage Affect Welfare Caseloadsrdquo Journalof Policy Analysis and Management 24 (2) (2005)273ndash295

5 Marianne Bitler and Hilary Hoynes ldquoThe More ThingsChange the More They Stay the Same The SafetyNet Living Arrangements and Poverty in the GreatRecessionrdquo NBER Working Paper 194 49 2013

6 Sylvia Allegretto and others ldquoFast Food Poverty Wages The Public Cost of Low-Wage Jobs in the Fast-FoodIndustryrdquo (Berkeley California Center for LaborResearch and Education 2013) available at httplaborcenterberkeleyedupubliccostsfast_food_poverty_wage

7 David Neumark and William Wascher ldquoDoes a HigherMinimum Wage Enhance the Effectiveness of theEarned Income Tax Creditrdquo Industrial and LaborRelations Review 64 (4) (2011) 712ndash746

8 David Lee and Emmanuel Saez ldquoOptimal MinimumWage Policy in Competitive Labor Marketsrdquo Journal ofPublic Economics 96 (9) (2012) 739ndash749

9 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family IncomerdquoFebruary 2014

10 Jesse Rothstein ldquoIs the EITC as Good as an NITConditional Cash Transfers and Tax Incidencerdquo AmericanEconomic Journal Economic Policy 2 (1) (2010) 177ndash208

11 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family Incomerdquo

12 Dube Arindrajit 2013 rdquoMinimum Wagesand the Distribution of Family IncomerdquoUnpublished working paper Available at httpsdldropboxusercontentcomu15038936Dube_MinimumWagesFamilyIncomespdf

13 As Dube explains in ldquoThe poverty of Minimum WageFactsrdquo the simulation approach underestimate stemsfrom a number of unwarranted assumptions includingthe range of actual wage increases and the accuracy ofwage data in the Current Population Survey The causal

approach does not make these assumptions

14 Allegretto Sylvia and others 2013 ldquoCredible ResearchDesigns for Minimum Wage Studiesrdquo Working Paper148-13 University of California Berkeley Institutefor Research on Labor and Employment Available athttpwwwirleberkeleyeduworkingpapers148-13pdf

15 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family In comerdquo

16 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo (2012)

17 Ibid

18 Ibid

19 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo

20 US Department of Agriculture ldquoSupplementalNutrition Assistance Program (SNAP)rdquo

21 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo

22 Ibid

23 We deviate from the Census Bureaursquos definition ofa family unit which is ldquotwo people or more (on e ofwhom is the householder) related by bir th marriageor adoption and residing togetherrdquo US Bureau ofthe Census ldquoCurrent Population Survey DefinitionsFamilyrdquo available at httpwwwcensusgovcpsabout

cpsdefhtml (last accessed February 2014) We countas a family unit any individual residing on his or herown two or more persons residing together whodo not belong to a family in the March CPS sampleare constructed as one family in our analysis For thepurposes of food stamp allocations the consumptionresulting from this transfer is probably distributed tofamily members (rather than household members ora single individual within the household) Howeversingle individuals canmdashand domdashreceive SNAPbenefits Excluding them would fail to make theanalysis reflective of the population at large

24 Strictly the family level linear probability modelpredicts the percentage-point decrease in theprobability that an individual family will receive SNAPpayments When applied to a large number of familieshowever we are able to interpret the coefficient asa decrease in the mean of enrollmentmdashthat is a

decrease in the enrollment ratemdashby applying the lawof iterated expectations

25 We generate expenditure predictions from theenrollment modelsmdashand conversely generateenrollment predictions from the expenditure modelmdashby assuming that expenditures per enrolled familyremains the same before and after the minimum wagechange In practice this is likely to be a conservativeestimatemdashthat is to underestimate the decrease inSNAP activity Average SNAP benefits per family willalso decrease as many families that remain eligible forSNAP experience income gains

26 Wage and Hour Division ldquoMinimum Wage Laws inthe States ndash Januar y 1 2014rdquo available at httpwwwdolgovwhdminwageamericahtm (last accessedFebruary 2014)

27 See for example Sylvia Allegretto and others ldquoCredibleResearch Designs for Minimum Wage Studiesrdquo WorkingPaper 148-113 (Berkeley California Institute forResearch on Labor and Employment 2013) available athttpwwwirleberkeleyeduworkingpapers148-13pdf

28 We will report these results in a forthcoming workingpaper

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4243

Endnotes | wwwamericanprogresso

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4343

The Center for American Progress is a nonpartisan research and educational institute

dedicated to promoting a strong just and free America that ensures opportunity

for all We believe that Americans are bound together by a common commitment to

these values and we aspire to ensure that our national policies reflect these values

We work to find progressive and pragmatic solutions to significant domestic and

international problems and develop policy proposals that foster a government that

is ldquoof the people by the people and for the peoplerdquo

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2643

Appendix A | wwwamericanprogresso

Appendix A

Pre-trend falsification check

Recen minimum wage research27 highlighs a common flaw in previous sudies

ailure o veriy ha he oucome variable is ree o negaive pre-exising rends I

or example SNAP aciviy was already rending down in saes ha raised heir

minimum wages beore hese changes came ino effec our regression analysis could

(misakenly) atribue ha reducion o he minimum wage We check or such

pre-rends by inroducing variables ha represen he prior yearrsquos value or leads o

he minimum wage I he model esimaes he minimum wage o have an effec on

he oucome variable beore he wage change wen ino effec hen an unobservedacor no he minimum wage change caused he change in SNAP aciviy

We es he specificaions above or pre-rends by including a one-year lead in

all hree specificaions We find ha he lead erms are small posiive and no

saisically significan indicaing ha he concurren minimum wage991252no

he wage level in prior periods991252is driving he observed changes in SNAP

oucomes28 In paricular he coefficien (sandard error) on he lead erm in

our preerred amily-level enrollmen regression is 011 and no significan

while he coefficien and sandard error o he conemporaneous minimum

wage is unchanged In he sae-level preerred enrollmen regression he

coefficien o he lead erm is again small (07) and i is no significan Te

corresponding coefficien on he lead erm in he sae-level expendiure

regression is 16 and is no significan Te posiive poin esimaes on hese lead

erms resuls no only rule ou disoring negaive pre-rends Tey also sugges

ha our main resuls may underesimae he rue effecs

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2743

24 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2843

Appendix B | wwwamericanprogresso

Appendix B

Model selection process

For boh he amily-level and sae-level models we es hree mehods o conrol

or unobserved geographic- and ime-varying characerisics as suggesed by he

minimum wage lieraure Firs we include only independen sae-specific fixed

effecs and year-specific fixed effecs Tis specificaion (specificaion 1) implicily

assumes ha amilies in any sae consiue an equally good saisical ldquoconrolrdquo

group or hose in any randomly chosen sae afer accouning or various

characerisics (median income and unemploymen rae among ohers) Similarly

simple ime fixed effecs assume ha amilies surveyed in any year can crediblyserve as a conrol group or amilies surveyed in every oher year o he sample

(1990 hrough 2012)

In oher words specificaion 1 assumes ha a saersquos immediae neighbor provides

no beter a couneracual or he effec o a minimum wage change han does a

sae across he counry We relax his resricive specificaion sequenially in wo

seps In specificaion 2 we replace simple year fixed effecs wih fixed effecs or

each Census divisionyear (capured as an addiional variable in he vecor By

using division-year effecs we remove he resricion ha amilies in each sae

are equally good saisical conrols or all oher amilies Raher we allow or he

possibiliy ha amilies in similar geographic regions (or example he Souh or

he Norheas) may be more similar o one anoher han amilies arher away

Finally in specificaion 3 we add sae-specific linear ime rends o he previous

specificaion Tus specificaion 3 is he mos rigorous model specificaion in ha

i allows or heerogeneiy along hree dimensions Ta is specificaion 3 allows

each sae o have is own ime-varying rends raher han imposing he resricion

ha saes evolve idenically over he 22 years in he sample

We begin building he heoreical specificaion above rom a se o simpleuncondiional models regression o SNAP aciviy (enrollmen or expendiures)

on he log o he minimum wage and a se o geographic- and ime-specific

effecs (specificaions 1 2 and 3 described above) As shown in ables 1ndash3 (or

specificaion 3) we hen add covariaes sequenially o hese models including

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2943

26 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

firs he vecor o amily-level conrols ollowed by each o several sae-level

covariaes in urn he unemploymen rae log o median-amily income and he

employmen-o-populaion raio Comparable resuls or specificaions 2 and 3

will be available in our orhcoming working paper

As expeced he simple uncondiional models indicae ha he relaionship beween he minimum wage and SNAP enrollmen i one exiss is a more complex

one influenced by oher acors In he uncondiional model he coefficien on he

variable o ineres991252he log o he minimum wage991252is small in magniude and no

saisically differen rom zero Once we accoun or he influence o labor marke

condiions and variaion in income levels on program paricipaion (by including

unemploymen rae and median-amily income conrol variables respecively)

he effec o he minimum wage on SNAP enrollmen is precisely esimaed Te

coefficien o he log minimum wage is slighly higher (-0042) in he amily-level

analysis han he coefficien (-031) in he sae-level analysis Te level o precision

is also higher in he amily-level analysis Tis is o be expeced when using 124million observaions compared o 1127

Te second dimension o model choice concerns he effec specificaion ables

7ndash9 compare he primary coefficiens o ineres or he SNAP enrollmen and

expendiure models For boh he enrollmen models he effec sizes are smalles

or specificaion 1 larges or specificaion 2 and inermediae beween hese wo

in specificaion 3 Recall ha Specificaion 3 conains sae-specific linear ime

rends in addiion o he census divisionyear conrols included in specificaion

2 In he amily-level enrollmen model he sandard error o he minimum wage

coefficien is smaller han in he oher wo specificaions Sandard errors on he

oher variables are much smaller in specificaions 2 and 3 han in specificaion

1 On he basis o coefficien significance (join and individual) specificaions 2

and 3 are sricly preerred in boh enrollmen models o specificaion 1 which

conains only sae and year fixed effecs

A concern wih specificaions 2 and 3 is ha rend conrols such as sae linear

rends may incorrecly absorb some o he delayed impac o a minimum wage

When we es his issue by including lagged minimum wages we do no find ha

delayed effecs are significan Anoher concern is ha more sauraed modelsuse less o he saisical variaion which could reduce he saisical power o

he resuls However he sandard errors or our more sauraed models are no

higher and are lower in some cases han or he less sauraed models Overall

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3043

Appendix B | wwwamericanprogresso

his evidence suppors our use o he sauraed model as he preerred model

specificaion Moreover Dubersquos 2013 sudy shows ha more sauraed models

perorm beter han models wih jus sae and ime fixed effecs

Te esimaed enrollmen regressions a boh he amily and sae levels show large

and saisically significan coefficiens Te esimaed minimum wage effec in heexpendiures regressions991252or which we have only sae-level daa991252is also large

and saisically significan

We do no use weighed regression or he sae-level models preerring o keep

analysis o he ldquoreamenrdquo (ha is o say a minimum wage change) appropriae

o he average sae raher han he average amily or individual I insead our

primary ineres were he impac o a minimum wage change on he average amily

or he average individual we migh choose o designae he number o amilies

in each sae or he sae populaion respecively as analyic weighs in order o

obain a coefficien beter suied or such inerence

TABLE 4

SNAP enrollment

Family-level linear probability

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0048 -0047 -0040 -0043 -0042

(0013) (0013) (001) (0008) (0009)

Unemployment rate (100) 0505 0420 0280

(0083) (0086) (0082)

Log median income -0057 -0039

(0011) (0011)

Employment-to-population ratio -0239

(0038)

N 1242022 1242022 1242022 1242022 1242022

plt01 plt005 plt001

Note Robust standard errors are in parentheses Observations are clustered at the state level The outcome variable is binary and equal to one if a family is enrolledin SNAP All models include state fixed effects Census division x-year fixed effects and state-specific linear time trends All specifications except 3a include additional

controls for family size race and marital status of the family head presence of children and presence of an adult maleSource Annual data from the March Current Population Survey for 1990 to 2012 estimation includes Current Population Survey probability weights

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3143

28 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

TABLE 5

SNAP enrollment

State-level linear regression

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0028 -0024 -0026 -0031 -0031

(0014) (0013) (0013) (0012) (0012)

Unemployment rate (100) 0625 0466 0320 0339

(0087) (0088) (0085) (0083)

Log median income -0090 -0065 -0061

(0013) (0013) (0013)

Employment-to-population ratio -0282 -0248

(0037) (0038)

N 1127 1127 1127 1127 1127

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the SNAP enrollment rate All models include state fixed effects Census division x-year fixedeffects and state-specific linear time trends

Source Annual data from the March Current Population Survey for 1990 to 2012

TABLE 6

SNAP expenditures

State-level linear regression

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0181 -0149 -0156 -0153 -0190

(011) (0103) (0102) (0103) (0103)

Unemployment rate (100) 4587 4152 4232 4313

(0622) (0621) (0633) (0628)

Log median income -0246 -0261 -0294

(0075) (0078) (0078)

Employment-to-population ratio 0155 0244

(0237) (024)

N 1127 1127 1127 1127 1127

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the natural log of state SNAP expenditures per capita for 1990 to 2012 All models include state

fixed effects Census division x-year fixed effects and state-specific linear time trends

Source Annual data from Bureau of Economic Analysis national income and product accounts tables for 1990 to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3243

Appendix B | wwwamericanprogresso

TABLE 7

Comparison of specifications SNAP enrollment

Family-level linear probability

(1) (2) (3)

Log minimum wage -0028 -0049 -0042

(0014) (0017) (0009)

Unemployment rate (100) 0275 0297 0280

(0161) (0076) (0082)

Log median income -0077 -0055 -0039

(0014) (0012) (0011)

Employment-to-population ratio -0238 -0250 -0239

(0054) (004) (0038)

N 1242022 1242022 1242022

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses Observations are clustered at the state level The outcomevariable is binary or equal to one if a family is enrolled in SNAP All specifications include additional controls forfamily size race and marital status of the family head presence of children and presence of an adult male

Source Annual data from the March Current Population Survey for 1990 to 2012 estimation includes CurrentPopulation Survey probability weights

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3343

30 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

TABLE 8

Comparison of specifications SNAP enrollment

State-level linear regression

(1) (2) (3)

Log minimum wage -0019 -0035 -0031

(0009) (0012) (0012)

Unemployment rate (100) 0401 0370 0339

(0063) (0077) (0083)

Log median income -0081 -0073 -0061

(0011) (0013) (0013)

Employment-to-population ratio -0183 -0222 -0248

(0039) (0039) (0038)

N 1127 1127 1127

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the SNAP enrollment rate Allregressions include the share of households with children and the racial shares of the population

Source Annual data from the March Current Population Survey for 1990 to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3443

Appendix B | wwwamericanprogresso

TABLE 9

Comparison of specifications SNAP expenditures

State-level linear regression

(1) (2) (3)

Log minimum wage -0121 -0203 -0190

(0075) (0103) (0103)

Unemployment rate (100) 5292 5152 4313

(0464) (0576) (0628)

Log median income -0437 -0417 -0294

(008) (0086) (0078)

Employment-to-population ratio -0040 -0220 0244

(0261) (0260) (0240)

N 1127 1127 1127

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the natural log of state SNAPexpenditures per capita All regressions include the share of households with children and the racial shares ofthe state population

Source Annual data from Bureau of Economic Analysis national income and product accounts tables for 1990to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3543

32 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3643

Appendix C | wwwamericanprogresso

Appendix C Harkin-Miller

policy simulation results

StateMinimum wage

(2014)

SNAP enrollment (2012) Predicted change in SNAP enrollment

RateTotal

(persons)

Model 1 family

enrollment (linearprobability)

Model 2 state

enrollment (linearregression)

Model 3 s

expenditu(linear regre

Alabama $725 164 788682 -66922 -58897 -58906

Alaska $775 120 87436 -8104 -6887 -3288

Arizona $790 201 1319323 -67435 -56738 -64356

Arkansas $725 78 230489 -40977 -36063 -35248

California $800 206 7813680 -371131 -310222 -18223

Colorado $800 164 853155 -50684 -42365 -23926

Connecticut $870 91 326621 -22456 -17975 -13711

Delaware $725 186 170262 -12739 -11211 -10647

District of Columbia $825 133 84009 -5370 -4417 -3632

Florida $793 166 3208026 -195813 -164426 -13046

Georgia $725 160 1586336 -137741 -121224 -11004

Hawaii $725 96 133662 -19310 -16995 -14933

Idaho $725 92 147501 -22165 -19507 -15809

Illinois $825 95 1225084 -109088 -89742 -70955

Indiana $725 125 816233 -90818 -79928 -83985

Iowa $725 155 478011 -42716 -37594 -28556

Kansas $725 135 388269 -40082 -35275 -27461

Kentucky $725 130 568821 -60840 -53544 -52259

Louisiana $725 149 683832 -63929 -56263 -66083

Maine $750 77 101976 -16567 -14323 -15234

Maryland $725 144 846415 -81748 -71946 -38370

Massachusetts $800 130 864721 -64902 -54251 -42913

Michigan $740 146 1439141 -128801 -112140 -11022

Minnesota $725 133 713646 -74730 -65769 -37878

TABLE 10

SNAP enrollments Predicted changes by state under the Harkin-Miller bill

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3743

34 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

StateMinimum wage

(2014)

SNAP enrollment (2012) Predicted change in SNAP enrollment

RateTotal

(persons)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 state

expenditures

(linear regression)

Mississippi $725 129 386501 -41486 -36511 -46467

Missouri $750 172 1036182 -75131 -64952 -56244

Montana $790 132 132452 -10350 -8708 -5846

Nebraska $725 124 230683 -25773 -22683 -12189

Nevada $825 162 446035 -23349 -19209 -11894

New Hampshire $725 127 168404 -18359 -16157 -5735

New Jersey $825 160 1416666 -75175 -61843 -28236

New Mexico $750 149 310896 -25983 -22463 -22512

New York $800 192 3763553 -191193 -159815 -142182

North Carolina $725 174 1697193 -135417 -119179 -113503

North Dakota $725 87 61225 -9743 -8574 -4021

Ohio $795 143 1647345 -115869 -97169 -88580

Oklahoma $725 129 494053 -53006 -46650 -46854

Oregon $910 124 485326 -17036 -13328 -16398

Pennsylvania $725 161 2053643 -177315 -156052 -125586

Rhode Island $800 156 163730 -10258 -8574 -8698

South Carolina $725 94 445277 -65614 -57746 -50304

South Dakota $725 208 173749 -11586 -10197 -7458

Tennessee $725 142 914903 -89667 -78915 -99134

Texas $725 110 2863779 -362018 -318607 -253285

Utah $725 88 251107 -39658 -34902 -19390

Vermont $873 156 97792 -3823 -3055 -2475

Virginia $725 101 829771 -113723 -100086 -58212

Washington $932 72 496934 -23221 -17947 -17756

West Virginia $725 58 107875 -25792 -22699 -21665

Wisconsin $725 75 427822 -79521 -69986 -53210

Wyoming $725 164 94590 -8010 -7050 -3104

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3843

Appendix C | wwwamericanprogresso

StateMinimum wage

(2014)

SNAP expenditures (2012)Predicted change in SNAP expenditures

(millions of dollars)

Per resident

(millions of

dollars)

Total

(millions of

dollars)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 s

expenditu

(linear regre

Alabama $725 $290 $1397 -$1185 -$1043 -$1043

Alaska $775 $253 $185 -$262 -$223 -$106

Arizona $790 $257 $1686 -$935 -$787 -$892

Arkansas $725 $248 $733 -$636 -$560 -$547

California $800 $189 $7164 -$7276 -$6082 -$3573

Colorado $800 $157 $816 -$862 -$721 -$407

Connecticut $870 $191 $686 -$343 -$275 -$210

Delaware $725 $250 $229 -$205 -$180 -$171

District of Columbia $825 $366 $232 -$146 -$120 -$99

Florida $793 $294 $5676 -$4429 -$3719 -$2951

Georgia $725 $317 $3140 -$2936 -$2584 -$2346

Hawaii $725 $335 $465 -$449 -$395 -$347

Idaho $725 $225 $359 -$376 -$331 -$268

Illinois $825 $249 $3200 -$2096 -$1725 -$1364

Indiana $725 $220 $1439 -$1162 -$1023 -$1075

Iowa $725 $192 $589 -$658 -$579 -$440

Kansas $725 $159 $460 -$502 -$441 -$344

Kentucky $725 $298 $1303 -$1133 -$997 -$973

Louisiana $725 $315 $1450 -$1047 -$922 -$1083

Maine $750 $281 $373 -$267 -$231 -$246

Maryland $725 $188 $1109 -$1765 -$1553 -$828

Massachusetts $800 $206 $1366 -$1030 -$861 -$681

Michigan $740 $300 $2963 -$2400 -$2090 -$2054

Minnesota $725 $140 $755 -$1113 -$980 -$564

Mississippi $725 $326 $973 -$649 -$571 -$726

Missouri $750 $241 $1452 -$1278 -$1104 -$956

Montana $790 $190 $191 -$179 -$151 -$101

Nebraska $725 $140 $259 -$409 -$360 -$194

Nevada $825 $191 $527 -$441 -$363 -$225

New Hampshire $725 $126 $167 -$399 -$351 -$125

TABLE 11

SNAP expenditures Predicted changes by state under the Harkin-Miller bill

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3943

36 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

StateMinimum wage

(2014)

SNAP expenditures (2012)Predicted change in SNAP expenditures

(millions of dollars)

Per resident

(millions of

dollars)

Total

(millions of

dollars)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 state

expenditures

(linear regression)

New Jersey $825 $160 $1420 -$1611 -$1325 -$605

New Mexico $750 $324 $675 -$513 -$444 -$445New York $800 $287 $5616 -$3766 -$3148 -$2801

North Carolina $725 $252 $2454 -$2187 -$1925 -$1833

North Dakota $725 $128 $90 -$162 -$143 -$67

Ohio $795 $259 $2995 -$2013 -$1688 -$1539

Oklahoma $725 $248 $945 -$799 -$703 -$706

Oregon $910 $322 $1255 -$272 -$213 -$262

Pennsylvania $725 $218 $2779 -$2930 -$2579 -$2075

Rhode Island $800 $280 $294 -$173 -$144 -$147

South Carolina $725 $291 $1373 -$1337 -$1177 -$1025South Dakota $725 $198 $165 -$192 -$169 -$123

Tennessee $725 $324 $2091 -$1413 -$1243 -$1562

Texas $725 $230 $5997 -$6402 -$5634 -$4479

Utah $725 $141 $402 -$614 -$541 -$300

Vermont $873 $230 $144 -$66 -$53 -$43

Virginia $725 $173 $1413 -$2062 -$1815 -$1056

Washington $932 $244 $1682 -$350 -$270 -$267

West Virginia $725 $273 $508 -$451 -$397 -$379

Wisconsin $725 $204 $1166 -$1302 -$1146 -$871Wyoming $725 $95 $55 -$105 -$93 -$41

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4043

Appendix C | wwwamericanprogresso

FIGURE 2

Raising the minimum wage to $1010would cut taxpayer costs in every state

Predicted decreases in cost and enrollment

in SNAP in 50 states

$200+$51ndash$100

$101ndash$200

0ndash$25

$26ndash$50

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4143

38 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

Endnotes

1 US Department of Agriculture ldquoSupplementalNutrition Assistance Program (SNAP)rdquo available athttpwwwfnsusdagovsnapeligibility (last accessedFebruary 2014)

2 For this initial analysis we do not consider Harkin-Millerrsquos increase in subminimum wages for tippedworkers To do s o would increase the estimated SNAP

savings by an unknown amount

3 The Congressional Budget Office estimates thatworkers currently earning between $1010 and $1150per hour would see their wages rise under the Harkin-Miller proposal Congressional Budget O ffice ldquoTheEffects of a Minimum Wage Increase on Employmentand Family Incomerdquo (2014)

4 Marianne Page Joanne Spetz and Jane Millar ldquoDoesthe Minimum Wage Affect Welfare Caseloadsrdquo Journalof Policy Analysis and Management 24 (2) (2005)273ndash295

5 Marianne Bitler and Hilary Hoynes ldquoThe More ThingsChange the More They Stay the Same The SafetyNet Living Arrangements and Poverty in the GreatRecessionrdquo NBER Working Paper 194 49 2013

6 Sylvia Allegretto and others ldquoFast Food Poverty Wages The Public Cost of Low-Wage Jobs in the Fast-FoodIndustryrdquo (Berkeley California Center for LaborResearch and Education 2013) available at httplaborcenterberkeleyedupubliccostsfast_food_poverty_wage

7 David Neumark and William Wascher ldquoDoes a HigherMinimum Wage Enhance the Effectiveness of theEarned Income Tax Creditrdquo Industrial and LaborRelations Review 64 (4) (2011) 712ndash746

8 David Lee and Emmanuel Saez ldquoOptimal MinimumWage Policy in Competitive Labor Marketsrdquo Journal ofPublic Economics 96 (9) (2012) 739ndash749

9 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family IncomerdquoFebruary 2014

10 Jesse Rothstein ldquoIs the EITC as Good as an NITConditional Cash Transfers and Tax Incidencerdquo AmericanEconomic Journal Economic Policy 2 (1) (2010) 177ndash208

11 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family Incomerdquo

12 Dube Arindrajit 2013 rdquoMinimum Wagesand the Distribution of Family IncomerdquoUnpublished working paper Available at httpsdldropboxusercontentcomu15038936Dube_MinimumWagesFamilyIncomespdf

13 As Dube explains in ldquoThe poverty of Minimum WageFactsrdquo the simulation approach underestimate stemsfrom a number of unwarranted assumptions includingthe range of actual wage increases and the accuracy ofwage data in the Current Population Survey The causal

approach does not make these assumptions

14 Allegretto Sylvia and others 2013 ldquoCredible ResearchDesigns for Minimum Wage Studiesrdquo Working Paper148-13 University of California Berkeley Institutefor Research on Labor and Employment Available athttpwwwirleberkeleyeduworkingpapers148-13pdf

15 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family In comerdquo

16 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo (2012)

17 Ibid

18 Ibid

19 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo

20 US Department of Agriculture ldquoSupplementalNutrition Assistance Program (SNAP)rdquo

21 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo

22 Ibid

23 We deviate from the Census Bureaursquos definition ofa family unit which is ldquotwo people or more (on e ofwhom is the householder) related by bir th marriageor adoption and residing togetherrdquo US Bureau ofthe Census ldquoCurrent Population Survey DefinitionsFamilyrdquo available at httpwwwcensusgovcpsabout

cpsdefhtml (last accessed February 2014) We countas a family unit any individual residing on his or herown two or more persons residing together whodo not belong to a family in the March CPS sampleare constructed as one family in our analysis For thepurposes of food stamp allocations the consumptionresulting from this transfer is probably distributed tofamily members (rather than household members ora single individual within the household) Howeversingle individuals canmdashand domdashreceive SNAPbenefits Excluding them would fail to make theanalysis reflective of the population at large

24 Strictly the family level linear probability modelpredicts the percentage-point decrease in theprobability that an individual family will receive SNAPpayments When applied to a large number of familieshowever we are able to interpret the coefficient asa decrease in the mean of enrollmentmdashthat is a

decrease in the enrollment ratemdashby applying the lawof iterated expectations

25 We generate expenditure predictions from theenrollment modelsmdashand conversely generateenrollment predictions from the expenditure modelmdashby assuming that expenditures per enrolled familyremains the same before and after the minimum wagechange In practice this is likely to be a conservativeestimatemdashthat is to underestimate the decrease inSNAP activity Average SNAP benefits per family willalso decrease as many families that remain eligible forSNAP experience income gains

26 Wage and Hour Division ldquoMinimum Wage Laws inthe States ndash Januar y 1 2014rdquo available at httpwwwdolgovwhdminwageamericahtm (last accessedFebruary 2014)

27 See for example Sylvia Allegretto and others ldquoCredibleResearch Designs for Minimum Wage Studiesrdquo WorkingPaper 148-113 (Berkeley California Institute forResearch on Labor and Employment 2013) available athttpwwwirleberkeleyeduworkingpapers148-13pdf

28 We will report these results in a forthcoming workingpaper

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4243

Endnotes | wwwamericanprogresso

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4343

The Center for American Progress is a nonpartisan research and educational institute

dedicated to promoting a strong just and free America that ensures opportunity

for all We believe that Americans are bound together by a common commitment to

these values and we aspire to ensure that our national policies reflect these values

We work to find progressive and pragmatic solutions to significant domestic and

international problems and develop policy proposals that foster a government that

is ldquoof the people by the people and for the peoplerdquo

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2743

24 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2843

Appendix B | wwwamericanprogresso

Appendix B

Model selection process

For boh he amily-level and sae-level models we es hree mehods o conrol

or unobserved geographic- and ime-varying characerisics as suggesed by he

minimum wage lieraure Firs we include only independen sae-specific fixed

effecs and year-specific fixed effecs Tis specificaion (specificaion 1) implicily

assumes ha amilies in any sae consiue an equally good saisical ldquoconrolrdquo

group or hose in any randomly chosen sae afer accouning or various

characerisics (median income and unemploymen rae among ohers) Similarly

simple ime fixed effecs assume ha amilies surveyed in any year can crediblyserve as a conrol group or amilies surveyed in every oher year o he sample

(1990 hrough 2012)

In oher words specificaion 1 assumes ha a saersquos immediae neighbor provides

no beter a couneracual or he effec o a minimum wage change han does a

sae across he counry We relax his resricive specificaion sequenially in wo

seps In specificaion 2 we replace simple year fixed effecs wih fixed effecs or

each Census divisionyear (capured as an addiional variable in he vecor By

using division-year effecs we remove he resricion ha amilies in each sae

are equally good saisical conrols or all oher amilies Raher we allow or he

possibiliy ha amilies in similar geographic regions (or example he Souh or

he Norheas) may be more similar o one anoher han amilies arher away

Finally in specificaion 3 we add sae-specific linear ime rends o he previous

specificaion Tus specificaion 3 is he mos rigorous model specificaion in ha

i allows or heerogeneiy along hree dimensions Ta is specificaion 3 allows

each sae o have is own ime-varying rends raher han imposing he resricion

ha saes evolve idenically over he 22 years in he sample

We begin building he heoreical specificaion above rom a se o simpleuncondiional models regression o SNAP aciviy (enrollmen or expendiures)

on he log o he minimum wage and a se o geographic- and ime-specific

effecs (specificaions 1 2 and 3 described above) As shown in ables 1ndash3 (or

specificaion 3) we hen add covariaes sequenially o hese models including

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2943

26 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

firs he vecor o amily-level conrols ollowed by each o several sae-level

covariaes in urn he unemploymen rae log o median-amily income and he

employmen-o-populaion raio Comparable resuls or specificaions 2 and 3

will be available in our orhcoming working paper

As expeced he simple uncondiional models indicae ha he relaionship beween he minimum wage and SNAP enrollmen i one exiss is a more complex

one influenced by oher acors In he uncondiional model he coefficien on he

variable o ineres991252he log o he minimum wage991252is small in magniude and no

saisically differen rom zero Once we accoun or he influence o labor marke

condiions and variaion in income levels on program paricipaion (by including

unemploymen rae and median-amily income conrol variables respecively)

he effec o he minimum wage on SNAP enrollmen is precisely esimaed Te

coefficien o he log minimum wage is slighly higher (-0042) in he amily-level

analysis han he coefficien (-031) in he sae-level analysis Te level o precision

is also higher in he amily-level analysis Tis is o be expeced when using 124million observaions compared o 1127

Te second dimension o model choice concerns he effec specificaion ables

7ndash9 compare he primary coefficiens o ineres or he SNAP enrollmen and

expendiure models For boh he enrollmen models he effec sizes are smalles

or specificaion 1 larges or specificaion 2 and inermediae beween hese wo

in specificaion 3 Recall ha Specificaion 3 conains sae-specific linear ime

rends in addiion o he census divisionyear conrols included in specificaion

2 In he amily-level enrollmen model he sandard error o he minimum wage

coefficien is smaller han in he oher wo specificaions Sandard errors on he

oher variables are much smaller in specificaions 2 and 3 han in specificaion

1 On he basis o coefficien significance (join and individual) specificaions 2

and 3 are sricly preerred in boh enrollmen models o specificaion 1 which

conains only sae and year fixed effecs

A concern wih specificaions 2 and 3 is ha rend conrols such as sae linear

rends may incorrecly absorb some o he delayed impac o a minimum wage

When we es his issue by including lagged minimum wages we do no find ha

delayed effecs are significan Anoher concern is ha more sauraed modelsuse less o he saisical variaion which could reduce he saisical power o

he resuls However he sandard errors or our more sauraed models are no

higher and are lower in some cases han or he less sauraed models Overall

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3043

Appendix B | wwwamericanprogresso

his evidence suppors our use o he sauraed model as he preerred model

specificaion Moreover Dubersquos 2013 sudy shows ha more sauraed models

perorm beter han models wih jus sae and ime fixed effecs

Te esimaed enrollmen regressions a boh he amily and sae levels show large

and saisically significan coefficiens Te esimaed minimum wage effec in heexpendiures regressions991252or which we have only sae-level daa991252is also large

and saisically significan

We do no use weighed regression or he sae-level models preerring o keep

analysis o he ldquoreamenrdquo (ha is o say a minimum wage change) appropriae

o he average sae raher han he average amily or individual I insead our

primary ineres were he impac o a minimum wage change on he average amily

or he average individual we migh choose o designae he number o amilies

in each sae or he sae populaion respecively as analyic weighs in order o

obain a coefficien beter suied or such inerence

TABLE 4

SNAP enrollment

Family-level linear probability

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0048 -0047 -0040 -0043 -0042

(0013) (0013) (001) (0008) (0009)

Unemployment rate (100) 0505 0420 0280

(0083) (0086) (0082)

Log median income -0057 -0039

(0011) (0011)

Employment-to-population ratio -0239

(0038)

N 1242022 1242022 1242022 1242022 1242022

plt01 plt005 plt001

Note Robust standard errors are in parentheses Observations are clustered at the state level The outcome variable is binary and equal to one if a family is enrolledin SNAP All models include state fixed effects Census division x-year fixed effects and state-specific linear time trends All specifications except 3a include additional

controls for family size race and marital status of the family head presence of children and presence of an adult maleSource Annual data from the March Current Population Survey for 1990 to 2012 estimation includes Current Population Survey probability weights

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3143

28 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

TABLE 5

SNAP enrollment

State-level linear regression

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0028 -0024 -0026 -0031 -0031

(0014) (0013) (0013) (0012) (0012)

Unemployment rate (100) 0625 0466 0320 0339

(0087) (0088) (0085) (0083)

Log median income -0090 -0065 -0061

(0013) (0013) (0013)

Employment-to-population ratio -0282 -0248

(0037) (0038)

N 1127 1127 1127 1127 1127

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the SNAP enrollment rate All models include state fixed effects Census division x-year fixedeffects and state-specific linear time trends

Source Annual data from the March Current Population Survey for 1990 to 2012

TABLE 6

SNAP expenditures

State-level linear regression

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0181 -0149 -0156 -0153 -0190

(011) (0103) (0102) (0103) (0103)

Unemployment rate (100) 4587 4152 4232 4313

(0622) (0621) (0633) (0628)

Log median income -0246 -0261 -0294

(0075) (0078) (0078)

Employment-to-population ratio 0155 0244

(0237) (024)

N 1127 1127 1127 1127 1127

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the natural log of state SNAP expenditures per capita for 1990 to 2012 All models include state

fixed effects Census division x-year fixed effects and state-specific linear time trends

Source Annual data from Bureau of Economic Analysis national income and product accounts tables for 1990 to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3243

Appendix B | wwwamericanprogresso

TABLE 7

Comparison of specifications SNAP enrollment

Family-level linear probability

(1) (2) (3)

Log minimum wage -0028 -0049 -0042

(0014) (0017) (0009)

Unemployment rate (100) 0275 0297 0280

(0161) (0076) (0082)

Log median income -0077 -0055 -0039

(0014) (0012) (0011)

Employment-to-population ratio -0238 -0250 -0239

(0054) (004) (0038)

N 1242022 1242022 1242022

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses Observations are clustered at the state level The outcomevariable is binary or equal to one if a family is enrolled in SNAP All specifications include additional controls forfamily size race and marital status of the family head presence of children and presence of an adult male

Source Annual data from the March Current Population Survey for 1990 to 2012 estimation includes CurrentPopulation Survey probability weights

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3343

30 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

TABLE 8

Comparison of specifications SNAP enrollment

State-level linear regression

(1) (2) (3)

Log minimum wage -0019 -0035 -0031

(0009) (0012) (0012)

Unemployment rate (100) 0401 0370 0339

(0063) (0077) (0083)

Log median income -0081 -0073 -0061

(0011) (0013) (0013)

Employment-to-population ratio -0183 -0222 -0248

(0039) (0039) (0038)

N 1127 1127 1127

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the SNAP enrollment rate Allregressions include the share of households with children and the racial shares of the population

Source Annual data from the March Current Population Survey for 1990 to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3443

Appendix B | wwwamericanprogresso

TABLE 9

Comparison of specifications SNAP expenditures

State-level linear regression

(1) (2) (3)

Log minimum wage -0121 -0203 -0190

(0075) (0103) (0103)

Unemployment rate (100) 5292 5152 4313

(0464) (0576) (0628)

Log median income -0437 -0417 -0294

(008) (0086) (0078)

Employment-to-population ratio -0040 -0220 0244

(0261) (0260) (0240)

N 1127 1127 1127

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the natural log of state SNAPexpenditures per capita All regressions include the share of households with children and the racial shares ofthe state population

Source Annual data from Bureau of Economic Analysis national income and product accounts tables for 1990to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3543

32 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3643

Appendix C | wwwamericanprogresso

Appendix C Harkin-Miller

policy simulation results

StateMinimum wage

(2014)

SNAP enrollment (2012) Predicted change in SNAP enrollment

RateTotal

(persons)

Model 1 family

enrollment (linearprobability)

Model 2 state

enrollment (linearregression)

Model 3 s

expenditu(linear regre

Alabama $725 164 788682 -66922 -58897 -58906

Alaska $775 120 87436 -8104 -6887 -3288

Arizona $790 201 1319323 -67435 -56738 -64356

Arkansas $725 78 230489 -40977 -36063 -35248

California $800 206 7813680 -371131 -310222 -18223

Colorado $800 164 853155 -50684 -42365 -23926

Connecticut $870 91 326621 -22456 -17975 -13711

Delaware $725 186 170262 -12739 -11211 -10647

District of Columbia $825 133 84009 -5370 -4417 -3632

Florida $793 166 3208026 -195813 -164426 -13046

Georgia $725 160 1586336 -137741 -121224 -11004

Hawaii $725 96 133662 -19310 -16995 -14933

Idaho $725 92 147501 -22165 -19507 -15809

Illinois $825 95 1225084 -109088 -89742 -70955

Indiana $725 125 816233 -90818 -79928 -83985

Iowa $725 155 478011 -42716 -37594 -28556

Kansas $725 135 388269 -40082 -35275 -27461

Kentucky $725 130 568821 -60840 -53544 -52259

Louisiana $725 149 683832 -63929 -56263 -66083

Maine $750 77 101976 -16567 -14323 -15234

Maryland $725 144 846415 -81748 -71946 -38370

Massachusetts $800 130 864721 -64902 -54251 -42913

Michigan $740 146 1439141 -128801 -112140 -11022

Minnesota $725 133 713646 -74730 -65769 -37878

TABLE 10

SNAP enrollments Predicted changes by state under the Harkin-Miller bill

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3743

34 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

StateMinimum wage

(2014)

SNAP enrollment (2012) Predicted change in SNAP enrollment

RateTotal

(persons)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 state

expenditures

(linear regression)

Mississippi $725 129 386501 -41486 -36511 -46467

Missouri $750 172 1036182 -75131 -64952 -56244

Montana $790 132 132452 -10350 -8708 -5846

Nebraska $725 124 230683 -25773 -22683 -12189

Nevada $825 162 446035 -23349 -19209 -11894

New Hampshire $725 127 168404 -18359 -16157 -5735

New Jersey $825 160 1416666 -75175 -61843 -28236

New Mexico $750 149 310896 -25983 -22463 -22512

New York $800 192 3763553 -191193 -159815 -142182

North Carolina $725 174 1697193 -135417 -119179 -113503

North Dakota $725 87 61225 -9743 -8574 -4021

Ohio $795 143 1647345 -115869 -97169 -88580

Oklahoma $725 129 494053 -53006 -46650 -46854

Oregon $910 124 485326 -17036 -13328 -16398

Pennsylvania $725 161 2053643 -177315 -156052 -125586

Rhode Island $800 156 163730 -10258 -8574 -8698

South Carolina $725 94 445277 -65614 -57746 -50304

South Dakota $725 208 173749 -11586 -10197 -7458

Tennessee $725 142 914903 -89667 -78915 -99134

Texas $725 110 2863779 -362018 -318607 -253285

Utah $725 88 251107 -39658 -34902 -19390

Vermont $873 156 97792 -3823 -3055 -2475

Virginia $725 101 829771 -113723 -100086 -58212

Washington $932 72 496934 -23221 -17947 -17756

West Virginia $725 58 107875 -25792 -22699 -21665

Wisconsin $725 75 427822 -79521 -69986 -53210

Wyoming $725 164 94590 -8010 -7050 -3104

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3843

Appendix C | wwwamericanprogresso

StateMinimum wage

(2014)

SNAP expenditures (2012)Predicted change in SNAP expenditures

(millions of dollars)

Per resident

(millions of

dollars)

Total

(millions of

dollars)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 s

expenditu

(linear regre

Alabama $725 $290 $1397 -$1185 -$1043 -$1043

Alaska $775 $253 $185 -$262 -$223 -$106

Arizona $790 $257 $1686 -$935 -$787 -$892

Arkansas $725 $248 $733 -$636 -$560 -$547

California $800 $189 $7164 -$7276 -$6082 -$3573

Colorado $800 $157 $816 -$862 -$721 -$407

Connecticut $870 $191 $686 -$343 -$275 -$210

Delaware $725 $250 $229 -$205 -$180 -$171

District of Columbia $825 $366 $232 -$146 -$120 -$99

Florida $793 $294 $5676 -$4429 -$3719 -$2951

Georgia $725 $317 $3140 -$2936 -$2584 -$2346

Hawaii $725 $335 $465 -$449 -$395 -$347

Idaho $725 $225 $359 -$376 -$331 -$268

Illinois $825 $249 $3200 -$2096 -$1725 -$1364

Indiana $725 $220 $1439 -$1162 -$1023 -$1075

Iowa $725 $192 $589 -$658 -$579 -$440

Kansas $725 $159 $460 -$502 -$441 -$344

Kentucky $725 $298 $1303 -$1133 -$997 -$973

Louisiana $725 $315 $1450 -$1047 -$922 -$1083

Maine $750 $281 $373 -$267 -$231 -$246

Maryland $725 $188 $1109 -$1765 -$1553 -$828

Massachusetts $800 $206 $1366 -$1030 -$861 -$681

Michigan $740 $300 $2963 -$2400 -$2090 -$2054

Minnesota $725 $140 $755 -$1113 -$980 -$564

Mississippi $725 $326 $973 -$649 -$571 -$726

Missouri $750 $241 $1452 -$1278 -$1104 -$956

Montana $790 $190 $191 -$179 -$151 -$101

Nebraska $725 $140 $259 -$409 -$360 -$194

Nevada $825 $191 $527 -$441 -$363 -$225

New Hampshire $725 $126 $167 -$399 -$351 -$125

TABLE 11

SNAP expenditures Predicted changes by state under the Harkin-Miller bill

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3943

36 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

StateMinimum wage

(2014)

SNAP expenditures (2012)Predicted change in SNAP expenditures

(millions of dollars)

Per resident

(millions of

dollars)

Total

(millions of

dollars)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 state

expenditures

(linear regression)

New Jersey $825 $160 $1420 -$1611 -$1325 -$605

New Mexico $750 $324 $675 -$513 -$444 -$445New York $800 $287 $5616 -$3766 -$3148 -$2801

North Carolina $725 $252 $2454 -$2187 -$1925 -$1833

North Dakota $725 $128 $90 -$162 -$143 -$67

Ohio $795 $259 $2995 -$2013 -$1688 -$1539

Oklahoma $725 $248 $945 -$799 -$703 -$706

Oregon $910 $322 $1255 -$272 -$213 -$262

Pennsylvania $725 $218 $2779 -$2930 -$2579 -$2075

Rhode Island $800 $280 $294 -$173 -$144 -$147

South Carolina $725 $291 $1373 -$1337 -$1177 -$1025South Dakota $725 $198 $165 -$192 -$169 -$123

Tennessee $725 $324 $2091 -$1413 -$1243 -$1562

Texas $725 $230 $5997 -$6402 -$5634 -$4479

Utah $725 $141 $402 -$614 -$541 -$300

Vermont $873 $230 $144 -$66 -$53 -$43

Virginia $725 $173 $1413 -$2062 -$1815 -$1056

Washington $932 $244 $1682 -$350 -$270 -$267

West Virginia $725 $273 $508 -$451 -$397 -$379

Wisconsin $725 $204 $1166 -$1302 -$1146 -$871Wyoming $725 $95 $55 -$105 -$93 -$41

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4043

Appendix C | wwwamericanprogresso

FIGURE 2

Raising the minimum wage to $1010would cut taxpayer costs in every state

Predicted decreases in cost and enrollment

in SNAP in 50 states

$200+$51ndash$100

$101ndash$200

0ndash$25

$26ndash$50

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4143

38 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

Endnotes

1 US Department of Agriculture ldquoSupplementalNutrition Assistance Program (SNAP)rdquo available athttpwwwfnsusdagovsnapeligibility (last accessedFebruary 2014)

2 For this initial analysis we do not consider Harkin-Millerrsquos increase in subminimum wages for tippedworkers To do s o would increase the estimated SNAP

savings by an unknown amount

3 The Congressional Budget Office estimates thatworkers currently earning between $1010 and $1150per hour would see their wages rise under the Harkin-Miller proposal Congressional Budget O ffice ldquoTheEffects of a Minimum Wage Increase on Employmentand Family Incomerdquo (2014)

4 Marianne Page Joanne Spetz and Jane Millar ldquoDoesthe Minimum Wage Affect Welfare Caseloadsrdquo Journalof Policy Analysis and Management 24 (2) (2005)273ndash295

5 Marianne Bitler and Hilary Hoynes ldquoThe More ThingsChange the More They Stay the Same The SafetyNet Living Arrangements and Poverty in the GreatRecessionrdquo NBER Working Paper 194 49 2013

6 Sylvia Allegretto and others ldquoFast Food Poverty Wages The Public Cost of Low-Wage Jobs in the Fast-FoodIndustryrdquo (Berkeley California Center for LaborResearch and Education 2013) available at httplaborcenterberkeleyedupubliccostsfast_food_poverty_wage

7 David Neumark and William Wascher ldquoDoes a HigherMinimum Wage Enhance the Effectiveness of theEarned Income Tax Creditrdquo Industrial and LaborRelations Review 64 (4) (2011) 712ndash746

8 David Lee and Emmanuel Saez ldquoOptimal MinimumWage Policy in Competitive Labor Marketsrdquo Journal ofPublic Economics 96 (9) (2012) 739ndash749

9 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family IncomerdquoFebruary 2014

10 Jesse Rothstein ldquoIs the EITC as Good as an NITConditional Cash Transfers and Tax Incidencerdquo AmericanEconomic Journal Economic Policy 2 (1) (2010) 177ndash208

11 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family Incomerdquo

12 Dube Arindrajit 2013 rdquoMinimum Wagesand the Distribution of Family IncomerdquoUnpublished working paper Available at httpsdldropboxusercontentcomu15038936Dube_MinimumWagesFamilyIncomespdf

13 As Dube explains in ldquoThe poverty of Minimum WageFactsrdquo the simulation approach underestimate stemsfrom a number of unwarranted assumptions includingthe range of actual wage increases and the accuracy ofwage data in the Current Population Survey The causal

approach does not make these assumptions

14 Allegretto Sylvia and others 2013 ldquoCredible ResearchDesigns for Minimum Wage Studiesrdquo Working Paper148-13 University of California Berkeley Institutefor Research on Labor and Employment Available athttpwwwirleberkeleyeduworkingpapers148-13pdf

15 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family In comerdquo

16 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo (2012)

17 Ibid

18 Ibid

19 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo

20 US Department of Agriculture ldquoSupplementalNutrition Assistance Program (SNAP)rdquo

21 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo

22 Ibid

23 We deviate from the Census Bureaursquos definition ofa family unit which is ldquotwo people or more (on e ofwhom is the householder) related by bir th marriageor adoption and residing togetherrdquo US Bureau ofthe Census ldquoCurrent Population Survey DefinitionsFamilyrdquo available at httpwwwcensusgovcpsabout

cpsdefhtml (last accessed February 2014) We countas a family unit any individual residing on his or herown two or more persons residing together whodo not belong to a family in the March CPS sampleare constructed as one family in our analysis For thepurposes of food stamp allocations the consumptionresulting from this transfer is probably distributed tofamily members (rather than household members ora single individual within the household) Howeversingle individuals canmdashand domdashreceive SNAPbenefits Excluding them would fail to make theanalysis reflective of the population at large

24 Strictly the family level linear probability modelpredicts the percentage-point decrease in theprobability that an individual family will receive SNAPpayments When applied to a large number of familieshowever we are able to interpret the coefficient asa decrease in the mean of enrollmentmdashthat is a

decrease in the enrollment ratemdashby applying the lawof iterated expectations

25 We generate expenditure predictions from theenrollment modelsmdashand conversely generateenrollment predictions from the expenditure modelmdashby assuming that expenditures per enrolled familyremains the same before and after the minimum wagechange In practice this is likely to be a conservativeestimatemdashthat is to underestimate the decrease inSNAP activity Average SNAP benefits per family willalso decrease as many families that remain eligible forSNAP experience income gains

26 Wage and Hour Division ldquoMinimum Wage Laws inthe States ndash Januar y 1 2014rdquo available at httpwwwdolgovwhdminwageamericahtm (last accessedFebruary 2014)

27 See for example Sylvia Allegretto and others ldquoCredibleResearch Designs for Minimum Wage Studiesrdquo WorkingPaper 148-113 (Berkeley California Institute forResearch on Labor and Employment 2013) available athttpwwwirleberkeleyeduworkingpapers148-13pdf

28 We will report these results in a forthcoming workingpaper

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4243

Endnotes | wwwamericanprogresso

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4343

The Center for American Progress is a nonpartisan research and educational institute

dedicated to promoting a strong just and free America that ensures opportunity

for all We believe that Americans are bound together by a common commitment to

these values and we aspire to ensure that our national policies reflect these values

We work to find progressive and pragmatic solutions to significant domestic and

international problems and develop policy proposals that foster a government that

is ldquoof the people by the people and for the peoplerdquo

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2843

Appendix B | wwwamericanprogresso

Appendix B

Model selection process

For boh he amily-level and sae-level models we es hree mehods o conrol

or unobserved geographic- and ime-varying characerisics as suggesed by he

minimum wage lieraure Firs we include only independen sae-specific fixed

effecs and year-specific fixed effecs Tis specificaion (specificaion 1) implicily

assumes ha amilies in any sae consiue an equally good saisical ldquoconrolrdquo

group or hose in any randomly chosen sae afer accouning or various

characerisics (median income and unemploymen rae among ohers) Similarly

simple ime fixed effecs assume ha amilies surveyed in any year can crediblyserve as a conrol group or amilies surveyed in every oher year o he sample

(1990 hrough 2012)

In oher words specificaion 1 assumes ha a saersquos immediae neighbor provides

no beter a couneracual or he effec o a minimum wage change han does a

sae across he counry We relax his resricive specificaion sequenially in wo

seps In specificaion 2 we replace simple year fixed effecs wih fixed effecs or

each Census divisionyear (capured as an addiional variable in he vecor By

using division-year effecs we remove he resricion ha amilies in each sae

are equally good saisical conrols or all oher amilies Raher we allow or he

possibiliy ha amilies in similar geographic regions (or example he Souh or

he Norheas) may be more similar o one anoher han amilies arher away

Finally in specificaion 3 we add sae-specific linear ime rends o he previous

specificaion Tus specificaion 3 is he mos rigorous model specificaion in ha

i allows or heerogeneiy along hree dimensions Ta is specificaion 3 allows

each sae o have is own ime-varying rends raher han imposing he resricion

ha saes evolve idenically over he 22 years in he sample

We begin building he heoreical specificaion above rom a se o simpleuncondiional models regression o SNAP aciviy (enrollmen or expendiures)

on he log o he minimum wage and a se o geographic- and ime-specific

effecs (specificaions 1 2 and 3 described above) As shown in ables 1ndash3 (or

specificaion 3) we hen add covariaes sequenially o hese models including

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2943

26 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

firs he vecor o amily-level conrols ollowed by each o several sae-level

covariaes in urn he unemploymen rae log o median-amily income and he

employmen-o-populaion raio Comparable resuls or specificaions 2 and 3

will be available in our orhcoming working paper

As expeced he simple uncondiional models indicae ha he relaionship beween he minimum wage and SNAP enrollmen i one exiss is a more complex

one influenced by oher acors In he uncondiional model he coefficien on he

variable o ineres991252he log o he minimum wage991252is small in magniude and no

saisically differen rom zero Once we accoun or he influence o labor marke

condiions and variaion in income levels on program paricipaion (by including

unemploymen rae and median-amily income conrol variables respecively)

he effec o he minimum wage on SNAP enrollmen is precisely esimaed Te

coefficien o he log minimum wage is slighly higher (-0042) in he amily-level

analysis han he coefficien (-031) in he sae-level analysis Te level o precision

is also higher in he amily-level analysis Tis is o be expeced when using 124million observaions compared o 1127

Te second dimension o model choice concerns he effec specificaion ables

7ndash9 compare he primary coefficiens o ineres or he SNAP enrollmen and

expendiure models For boh he enrollmen models he effec sizes are smalles

or specificaion 1 larges or specificaion 2 and inermediae beween hese wo

in specificaion 3 Recall ha Specificaion 3 conains sae-specific linear ime

rends in addiion o he census divisionyear conrols included in specificaion

2 In he amily-level enrollmen model he sandard error o he minimum wage

coefficien is smaller han in he oher wo specificaions Sandard errors on he

oher variables are much smaller in specificaions 2 and 3 han in specificaion

1 On he basis o coefficien significance (join and individual) specificaions 2

and 3 are sricly preerred in boh enrollmen models o specificaion 1 which

conains only sae and year fixed effecs

A concern wih specificaions 2 and 3 is ha rend conrols such as sae linear

rends may incorrecly absorb some o he delayed impac o a minimum wage

When we es his issue by including lagged minimum wages we do no find ha

delayed effecs are significan Anoher concern is ha more sauraed modelsuse less o he saisical variaion which could reduce he saisical power o

he resuls However he sandard errors or our more sauraed models are no

higher and are lower in some cases han or he less sauraed models Overall

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3043

Appendix B | wwwamericanprogresso

his evidence suppors our use o he sauraed model as he preerred model

specificaion Moreover Dubersquos 2013 sudy shows ha more sauraed models

perorm beter han models wih jus sae and ime fixed effecs

Te esimaed enrollmen regressions a boh he amily and sae levels show large

and saisically significan coefficiens Te esimaed minimum wage effec in heexpendiures regressions991252or which we have only sae-level daa991252is also large

and saisically significan

We do no use weighed regression or he sae-level models preerring o keep

analysis o he ldquoreamenrdquo (ha is o say a minimum wage change) appropriae

o he average sae raher han he average amily or individual I insead our

primary ineres were he impac o a minimum wage change on he average amily

or he average individual we migh choose o designae he number o amilies

in each sae or he sae populaion respecively as analyic weighs in order o

obain a coefficien beter suied or such inerence

TABLE 4

SNAP enrollment

Family-level linear probability

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0048 -0047 -0040 -0043 -0042

(0013) (0013) (001) (0008) (0009)

Unemployment rate (100) 0505 0420 0280

(0083) (0086) (0082)

Log median income -0057 -0039

(0011) (0011)

Employment-to-population ratio -0239

(0038)

N 1242022 1242022 1242022 1242022 1242022

plt01 plt005 plt001

Note Robust standard errors are in parentheses Observations are clustered at the state level The outcome variable is binary and equal to one if a family is enrolledin SNAP All models include state fixed effects Census division x-year fixed effects and state-specific linear time trends All specifications except 3a include additional

controls for family size race and marital status of the family head presence of children and presence of an adult maleSource Annual data from the March Current Population Survey for 1990 to 2012 estimation includes Current Population Survey probability weights

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3143

28 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

TABLE 5

SNAP enrollment

State-level linear regression

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0028 -0024 -0026 -0031 -0031

(0014) (0013) (0013) (0012) (0012)

Unemployment rate (100) 0625 0466 0320 0339

(0087) (0088) (0085) (0083)

Log median income -0090 -0065 -0061

(0013) (0013) (0013)

Employment-to-population ratio -0282 -0248

(0037) (0038)

N 1127 1127 1127 1127 1127

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the SNAP enrollment rate All models include state fixed effects Census division x-year fixedeffects and state-specific linear time trends

Source Annual data from the March Current Population Survey for 1990 to 2012

TABLE 6

SNAP expenditures

State-level linear regression

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0181 -0149 -0156 -0153 -0190

(011) (0103) (0102) (0103) (0103)

Unemployment rate (100) 4587 4152 4232 4313

(0622) (0621) (0633) (0628)

Log median income -0246 -0261 -0294

(0075) (0078) (0078)

Employment-to-population ratio 0155 0244

(0237) (024)

N 1127 1127 1127 1127 1127

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the natural log of state SNAP expenditures per capita for 1990 to 2012 All models include state

fixed effects Census division x-year fixed effects and state-specific linear time trends

Source Annual data from Bureau of Economic Analysis national income and product accounts tables for 1990 to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3243

Appendix B | wwwamericanprogresso

TABLE 7

Comparison of specifications SNAP enrollment

Family-level linear probability

(1) (2) (3)

Log minimum wage -0028 -0049 -0042

(0014) (0017) (0009)

Unemployment rate (100) 0275 0297 0280

(0161) (0076) (0082)

Log median income -0077 -0055 -0039

(0014) (0012) (0011)

Employment-to-population ratio -0238 -0250 -0239

(0054) (004) (0038)

N 1242022 1242022 1242022

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses Observations are clustered at the state level The outcomevariable is binary or equal to one if a family is enrolled in SNAP All specifications include additional controls forfamily size race and marital status of the family head presence of children and presence of an adult male

Source Annual data from the March Current Population Survey for 1990 to 2012 estimation includes CurrentPopulation Survey probability weights

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3343

30 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

TABLE 8

Comparison of specifications SNAP enrollment

State-level linear regression

(1) (2) (3)

Log minimum wage -0019 -0035 -0031

(0009) (0012) (0012)

Unemployment rate (100) 0401 0370 0339

(0063) (0077) (0083)

Log median income -0081 -0073 -0061

(0011) (0013) (0013)

Employment-to-population ratio -0183 -0222 -0248

(0039) (0039) (0038)

N 1127 1127 1127

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the SNAP enrollment rate Allregressions include the share of households with children and the racial shares of the population

Source Annual data from the March Current Population Survey for 1990 to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3443

Appendix B | wwwamericanprogresso

TABLE 9

Comparison of specifications SNAP expenditures

State-level linear regression

(1) (2) (3)

Log minimum wage -0121 -0203 -0190

(0075) (0103) (0103)

Unemployment rate (100) 5292 5152 4313

(0464) (0576) (0628)

Log median income -0437 -0417 -0294

(008) (0086) (0078)

Employment-to-population ratio -0040 -0220 0244

(0261) (0260) (0240)

N 1127 1127 1127

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the natural log of state SNAPexpenditures per capita All regressions include the share of households with children and the racial shares ofthe state population

Source Annual data from Bureau of Economic Analysis national income and product accounts tables for 1990to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3543

32 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3643

Appendix C | wwwamericanprogresso

Appendix C Harkin-Miller

policy simulation results

StateMinimum wage

(2014)

SNAP enrollment (2012) Predicted change in SNAP enrollment

RateTotal

(persons)

Model 1 family

enrollment (linearprobability)

Model 2 state

enrollment (linearregression)

Model 3 s

expenditu(linear regre

Alabama $725 164 788682 -66922 -58897 -58906

Alaska $775 120 87436 -8104 -6887 -3288

Arizona $790 201 1319323 -67435 -56738 -64356

Arkansas $725 78 230489 -40977 -36063 -35248

California $800 206 7813680 -371131 -310222 -18223

Colorado $800 164 853155 -50684 -42365 -23926

Connecticut $870 91 326621 -22456 -17975 -13711

Delaware $725 186 170262 -12739 -11211 -10647

District of Columbia $825 133 84009 -5370 -4417 -3632

Florida $793 166 3208026 -195813 -164426 -13046

Georgia $725 160 1586336 -137741 -121224 -11004

Hawaii $725 96 133662 -19310 -16995 -14933

Idaho $725 92 147501 -22165 -19507 -15809

Illinois $825 95 1225084 -109088 -89742 -70955

Indiana $725 125 816233 -90818 -79928 -83985

Iowa $725 155 478011 -42716 -37594 -28556

Kansas $725 135 388269 -40082 -35275 -27461

Kentucky $725 130 568821 -60840 -53544 -52259

Louisiana $725 149 683832 -63929 -56263 -66083

Maine $750 77 101976 -16567 -14323 -15234

Maryland $725 144 846415 -81748 -71946 -38370

Massachusetts $800 130 864721 -64902 -54251 -42913

Michigan $740 146 1439141 -128801 -112140 -11022

Minnesota $725 133 713646 -74730 -65769 -37878

TABLE 10

SNAP enrollments Predicted changes by state under the Harkin-Miller bill

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3743

34 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

StateMinimum wage

(2014)

SNAP enrollment (2012) Predicted change in SNAP enrollment

RateTotal

(persons)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 state

expenditures

(linear regression)

Mississippi $725 129 386501 -41486 -36511 -46467

Missouri $750 172 1036182 -75131 -64952 -56244

Montana $790 132 132452 -10350 -8708 -5846

Nebraska $725 124 230683 -25773 -22683 -12189

Nevada $825 162 446035 -23349 -19209 -11894

New Hampshire $725 127 168404 -18359 -16157 -5735

New Jersey $825 160 1416666 -75175 -61843 -28236

New Mexico $750 149 310896 -25983 -22463 -22512

New York $800 192 3763553 -191193 -159815 -142182

North Carolina $725 174 1697193 -135417 -119179 -113503

North Dakota $725 87 61225 -9743 -8574 -4021

Ohio $795 143 1647345 -115869 -97169 -88580

Oklahoma $725 129 494053 -53006 -46650 -46854

Oregon $910 124 485326 -17036 -13328 -16398

Pennsylvania $725 161 2053643 -177315 -156052 -125586

Rhode Island $800 156 163730 -10258 -8574 -8698

South Carolina $725 94 445277 -65614 -57746 -50304

South Dakota $725 208 173749 -11586 -10197 -7458

Tennessee $725 142 914903 -89667 -78915 -99134

Texas $725 110 2863779 -362018 -318607 -253285

Utah $725 88 251107 -39658 -34902 -19390

Vermont $873 156 97792 -3823 -3055 -2475

Virginia $725 101 829771 -113723 -100086 -58212

Washington $932 72 496934 -23221 -17947 -17756

West Virginia $725 58 107875 -25792 -22699 -21665

Wisconsin $725 75 427822 -79521 -69986 -53210

Wyoming $725 164 94590 -8010 -7050 -3104

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3843

Appendix C | wwwamericanprogresso

StateMinimum wage

(2014)

SNAP expenditures (2012)Predicted change in SNAP expenditures

(millions of dollars)

Per resident

(millions of

dollars)

Total

(millions of

dollars)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 s

expenditu

(linear regre

Alabama $725 $290 $1397 -$1185 -$1043 -$1043

Alaska $775 $253 $185 -$262 -$223 -$106

Arizona $790 $257 $1686 -$935 -$787 -$892

Arkansas $725 $248 $733 -$636 -$560 -$547

California $800 $189 $7164 -$7276 -$6082 -$3573

Colorado $800 $157 $816 -$862 -$721 -$407

Connecticut $870 $191 $686 -$343 -$275 -$210

Delaware $725 $250 $229 -$205 -$180 -$171

District of Columbia $825 $366 $232 -$146 -$120 -$99

Florida $793 $294 $5676 -$4429 -$3719 -$2951

Georgia $725 $317 $3140 -$2936 -$2584 -$2346

Hawaii $725 $335 $465 -$449 -$395 -$347

Idaho $725 $225 $359 -$376 -$331 -$268

Illinois $825 $249 $3200 -$2096 -$1725 -$1364

Indiana $725 $220 $1439 -$1162 -$1023 -$1075

Iowa $725 $192 $589 -$658 -$579 -$440

Kansas $725 $159 $460 -$502 -$441 -$344

Kentucky $725 $298 $1303 -$1133 -$997 -$973

Louisiana $725 $315 $1450 -$1047 -$922 -$1083

Maine $750 $281 $373 -$267 -$231 -$246

Maryland $725 $188 $1109 -$1765 -$1553 -$828

Massachusetts $800 $206 $1366 -$1030 -$861 -$681

Michigan $740 $300 $2963 -$2400 -$2090 -$2054

Minnesota $725 $140 $755 -$1113 -$980 -$564

Mississippi $725 $326 $973 -$649 -$571 -$726

Missouri $750 $241 $1452 -$1278 -$1104 -$956

Montana $790 $190 $191 -$179 -$151 -$101

Nebraska $725 $140 $259 -$409 -$360 -$194

Nevada $825 $191 $527 -$441 -$363 -$225

New Hampshire $725 $126 $167 -$399 -$351 -$125

TABLE 11

SNAP expenditures Predicted changes by state under the Harkin-Miller bill

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3943

36 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

StateMinimum wage

(2014)

SNAP expenditures (2012)Predicted change in SNAP expenditures

(millions of dollars)

Per resident

(millions of

dollars)

Total

(millions of

dollars)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 state

expenditures

(linear regression)

New Jersey $825 $160 $1420 -$1611 -$1325 -$605

New Mexico $750 $324 $675 -$513 -$444 -$445New York $800 $287 $5616 -$3766 -$3148 -$2801

North Carolina $725 $252 $2454 -$2187 -$1925 -$1833

North Dakota $725 $128 $90 -$162 -$143 -$67

Ohio $795 $259 $2995 -$2013 -$1688 -$1539

Oklahoma $725 $248 $945 -$799 -$703 -$706

Oregon $910 $322 $1255 -$272 -$213 -$262

Pennsylvania $725 $218 $2779 -$2930 -$2579 -$2075

Rhode Island $800 $280 $294 -$173 -$144 -$147

South Carolina $725 $291 $1373 -$1337 -$1177 -$1025South Dakota $725 $198 $165 -$192 -$169 -$123

Tennessee $725 $324 $2091 -$1413 -$1243 -$1562

Texas $725 $230 $5997 -$6402 -$5634 -$4479

Utah $725 $141 $402 -$614 -$541 -$300

Vermont $873 $230 $144 -$66 -$53 -$43

Virginia $725 $173 $1413 -$2062 -$1815 -$1056

Washington $932 $244 $1682 -$350 -$270 -$267

West Virginia $725 $273 $508 -$451 -$397 -$379

Wisconsin $725 $204 $1166 -$1302 -$1146 -$871Wyoming $725 $95 $55 -$105 -$93 -$41

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4043

Appendix C | wwwamericanprogresso

FIGURE 2

Raising the minimum wage to $1010would cut taxpayer costs in every state

Predicted decreases in cost and enrollment

in SNAP in 50 states

$200+$51ndash$100

$101ndash$200

0ndash$25

$26ndash$50

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4143

38 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

Endnotes

1 US Department of Agriculture ldquoSupplementalNutrition Assistance Program (SNAP)rdquo available athttpwwwfnsusdagovsnapeligibility (last accessedFebruary 2014)

2 For this initial analysis we do not consider Harkin-Millerrsquos increase in subminimum wages for tippedworkers To do s o would increase the estimated SNAP

savings by an unknown amount

3 The Congressional Budget Office estimates thatworkers currently earning between $1010 and $1150per hour would see their wages rise under the Harkin-Miller proposal Congressional Budget O ffice ldquoTheEffects of a Minimum Wage Increase on Employmentand Family Incomerdquo (2014)

4 Marianne Page Joanne Spetz and Jane Millar ldquoDoesthe Minimum Wage Affect Welfare Caseloadsrdquo Journalof Policy Analysis and Management 24 (2) (2005)273ndash295

5 Marianne Bitler and Hilary Hoynes ldquoThe More ThingsChange the More They Stay the Same The SafetyNet Living Arrangements and Poverty in the GreatRecessionrdquo NBER Working Paper 194 49 2013

6 Sylvia Allegretto and others ldquoFast Food Poverty Wages The Public Cost of Low-Wage Jobs in the Fast-FoodIndustryrdquo (Berkeley California Center for LaborResearch and Education 2013) available at httplaborcenterberkeleyedupubliccostsfast_food_poverty_wage

7 David Neumark and William Wascher ldquoDoes a HigherMinimum Wage Enhance the Effectiveness of theEarned Income Tax Creditrdquo Industrial and LaborRelations Review 64 (4) (2011) 712ndash746

8 David Lee and Emmanuel Saez ldquoOptimal MinimumWage Policy in Competitive Labor Marketsrdquo Journal ofPublic Economics 96 (9) (2012) 739ndash749

9 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family IncomerdquoFebruary 2014

10 Jesse Rothstein ldquoIs the EITC as Good as an NITConditional Cash Transfers and Tax Incidencerdquo AmericanEconomic Journal Economic Policy 2 (1) (2010) 177ndash208

11 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family Incomerdquo

12 Dube Arindrajit 2013 rdquoMinimum Wagesand the Distribution of Family IncomerdquoUnpublished working paper Available at httpsdldropboxusercontentcomu15038936Dube_MinimumWagesFamilyIncomespdf

13 As Dube explains in ldquoThe poverty of Minimum WageFactsrdquo the simulation approach underestimate stemsfrom a number of unwarranted assumptions includingthe range of actual wage increases and the accuracy ofwage data in the Current Population Survey The causal

approach does not make these assumptions

14 Allegretto Sylvia and others 2013 ldquoCredible ResearchDesigns for Minimum Wage Studiesrdquo Working Paper148-13 University of California Berkeley Institutefor Research on Labor and Employment Available athttpwwwirleberkeleyeduworkingpapers148-13pdf

15 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family In comerdquo

16 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo (2012)

17 Ibid

18 Ibid

19 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo

20 US Department of Agriculture ldquoSupplementalNutrition Assistance Program (SNAP)rdquo

21 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo

22 Ibid

23 We deviate from the Census Bureaursquos definition ofa family unit which is ldquotwo people or more (on e ofwhom is the householder) related by bir th marriageor adoption and residing togetherrdquo US Bureau ofthe Census ldquoCurrent Population Survey DefinitionsFamilyrdquo available at httpwwwcensusgovcpsabout

cpsdefhtml (last accessed February 2014) We countas a family unit any individual residing on his or herown two or more persons residing together whodo not belong to a family in the March CPS sampleare constructed as one family in our analysis For thepurposes of food stamp allocations the consumptionresulting from this transfer is probably distributed tofamily members (rather than household members ora single individual within the household) Howeversingle individuals canmdashand domdashreceive SNAPbenefits Excluding them would fail to make theanalysis reflective of the population at large

24 Strictly the family level linear probability modelpredicts the percentage-point decrease in theprobability that an individual family will receive SNAPpayments When applied to a large number of familieshowever we are able to interpret the coefficient asa decrease in the mean of enrollmentmdashthat is a

decrease in the enrollment ratemdashby applying the lawof iterated expectations

25 We generate expenditure predictions from theenrollment modelsmdashand conversely generateenrollment predictions from the expenditure modelmdashby assuming that expenditures per enrolled familyremains the same before and after the minimum wagechange In practice this is likely to be a conservativeestimatemdashthat is to underestimate the decrease inSNAP activity Average SNAP benefits per family willalso decrease as many families that remain eligible forSNAP experience income gains

26 Wage and Hour Division ldquoMinimum Wage Laws inthe States ndash Januar y 1 2014rdquo available at httpwwwdolgovwhdminwageamericahtm (last accessedFebruary 2014)

27 See for example Sylvia Allegretto and others ldquoCredibleResearch Designs for Minimum Wage Studiesrdquo WorkingPaper 148-113 (Berkeley California Institute forResearch on Labor and Employment 2013) available athttpwwwirleberkeleyeduworkingpapers148-13pdf

28 We will report these results in a forthcoming workingpaper

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4243

Endnotes | wwwamericanprogresso

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4343

The Center for American Progress is a nonpartisan research and educational institute

dedicated to promoting a strong just and free America that ensures opportunity

for all We believe that Americans are bound together by a common commitment to

these values and we aspire to ensure that our national policies reflect these values

We work to find progressive and pragmatic solutions to significant domestic and

international problems and develop policy proposals that foster a government that

is ldquoof the people by the people and for the peoplerdquo

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 2943

26 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

firs he vecor o amily-level conrols ollowed by each o several sae-level

covariaes in urn he unemploymen rae log o median-amily income and he

employmen-o-populaion raio Comparable resuls or specificaions 2 and 3

will be available in our orhcoming working paper

As expeced he simple uncondiional models indicae ha he relaionship beween he minimum wage and SNAP enrollmen i one exiss is a more complex

one influenced by oher acors In he uncondiional model he coefficien on he

variable o ineres991252he log o he minimum wage991252is small in magniude and no

saisically differen rom zero Once we accoun or he influence o labor marke

condiions and variaion in income levels on program paricipaion (by including

unemploymen rae and median-amily income conrol variables respecively)

he effec o he minimum wage on SNAP enrollmen is precisely esimaed Te

coefficien o he log minimum wage is slighly higher (-0042) in he amily-level

analysis han he coefficien (-031) in he sae-level analysis Te level o precision

is also higher in he amily-level analysis Tis is o be expeced when using 124million observaions compared o 1127

Te second dimension o model choice concerns he effec specificaion ables

7ndash9 compare he primary coefficiens o ineres or he SNAP enrollmen and

expendiure models For boh he enrollmen models he effec sizes are smalles

or specificaion 1 larges or specificaion 2 and inermediae beween hese wo

in specificaion 3 Recall ha Specificaion 3 conains sae-specific linear ime

rends in addiion o he census divisionyear conrols included in specificaion

2 In he amily-level enrollmen model he sandard error o he minimum wage

coefficien is smaller han in he oher wo specificaions Sandard errors on he

oher variables are much smaller in specificaions 2 and 3 han in specificaion

1 On he basis o coefficien significance (join and individual) specificaions 2

and 3 are sricly preerred in boh enrollmen models o specificaion 1 which

conains only sae and year fixed effecs

A concern wih specificaions 2 and 3 is ha rend conrols such as sae linear

rends may incorrecly absorb some o he delayed impac o a minimum wage

When we es his issue by including lagged minimum wages we do no find ha

delayed effecs are significan Anoher concern is ha more sauraed modelsuse less o he saisical variaion which could reduce he saisical power o

he resuls However he sandard errors or our more sauraed models are no

higher and are lower in some cases han or he less sauraed models Overall

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3043

Appendix B | wwwamericanprogresso

his evidence suppors our use o he sauraed model as he preerred model

specificaion Moreover Dubersquos 2013 sudy shows ha more sauraed models

perorm beter han models wih jus sae and ime fixed effecs

Te esimaed enrollmen regressions a boh he amily and sae levels show large

and saisically significan coefficiens Te esimaed minimum wage effec in heexpendiures regressions991252or which we have only sae-level daa991252is also large

and saisically significan

We do no use weighed regression or he sae-level models preerring o keep

analysis o he ldquoreamenrdquo (ha is o say a minimum wage change) appropriae

o he average sae raher han he average amily or individual I insead our

primary ineres were he impac o a minimum wage change on he average amily

or he average individual we migh choose o designae he number o amilies

in each sae or he sae populaion respecively as analyic weighs in order o

obain a coefficien beter suied or such inerence

TABLE 4

SNAP enrollment

Family-level linear probability

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0048 -0047 -0040 -0043 -0042

(0013) (0013) (001) (0008) (0009)

Unemployment rate (100) 0505 0420 0280

(0083) (0086) (0082)

Log median income -0057 -0039

(0011) (0011)

Employment-to-population ratio -0239

(0038)

N 1242022 1242022 1242022 1242022 1242022

plt01 plt005 plt001

Note Robust standard errors are in parentheses Observations are clustered at the state level The outcome variable is binary and equal to one if a family is enrolledin SNAP All models include state fixed effects Census division x-year fixed effects and state-specific linear time trends All specifications except 3a include additional

controls for family size race and marital status of the family head presence of children and presence of an adult maleSource Annual data from the March Current Population Survey for 1990 to 2012 estimation includes Current Population Survey probability weights

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3143

28 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

TABLE 5

SNAP enrollment

State-level linear regression

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0028 -0024 -0026 -0031 -0031

(0014) (0013) (0013) (0012) (0012)

Unemployment rate (100) 0625 0466 0320 0339

(0087) (0088) (0085) (0083)

Log median income -0090 -0065 -0061

(0013) (0013) (0013)

Employment-to-population ratio -0282 -0248

(0037) (0038)

N 1127 1127 1127 1127 1127

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the SNAP enrollment rate All models include state fixed effects Census division x-year fixedeffects and state-specific linear time trends

Source Annual data from the March Current Population Survey for 1990 to 2012

TABLE 6

SNAP expenditures

State-level linear regression

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0181 -0149 -0156 -0153 -0190

(011) (0103) (0102) (0103) (0103)

Unemployment rate (100) 4587 4152 4232 4313

(0622) (0621) (0633) (0628)

Log median income -0246 -0261 -0294

(0075) (0078) (0078)

Employment-to-population ratio 0155 0244

(0237) (024)

N 1127 1127 1127 1127 1127

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the natural log of state SNAP expenditures per capita for 1990 to 2012 All models include state

fixed effects Census division x-year fixed effects and state-specific linear time trends

Source Annual data from Bureau of Economic Analysis national income and product accounts tables for 1990 to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3243

Appendix B | wwwamericanprogresso

TABLE 7

Comparison of specifications SNAP enrollment

Family-level linear probability

(1) (2) (3)

Log minimum wage -0028 -0049 -0042

(0014) (0017) (0009)

Unemployment rate (100) 0275 0297 0280

(0161) (0076) (0082)

Log median income -0077 -0055 -0039

(0014) (0012) (0011)

Employment-to-population ratio -0238 -0250 -0239

(0054) (004) (0038)

N 1242022 1242022 1242022

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses Observations are clustered at the state level The outcomevariable is binary or equal to one if a family is enrolled in SNAP All specifications include additional controls forfamily size race and marital status of the family head presence of children and presence of an adult male

Source Annual data from the March Current Population Survey for 1990 to 2012 estimation includes CurrentPopulation Survey probability weights

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3343

30 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

TABLE 8

Comparison of specifications SNAP enrollment

State-level linear regression

(1) (2) (3)

Log minimum wage -0019 -0035 -0031

(0009) (0012) (0012)

Unemployment rate (100) 0401 0370 0339

(0063) (0077) (0083)

Log median income -0081 -0073 -0061

(0011) (0013) (0013)

Employment-to-population ratio -0183 -0222 -0248

(0039) (0039) (0038)

N 1127 1127 1127

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the SNAP enrollment rate Allregressions include the share of households with children and the racial shares of the population

Source Annual data from the March Current Population Survey for 1990 to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3443

Appendix B | wwwamericanprogresso

TABLE 9

Comparison of specifications SNAP expenditures

State-level linear regression

(1) (2) (3)

Log minimum wage -0121 -0203 -0190

(0075) (0103) (0103)

Unemployment rate (100) 5292 5152 4313

(0464) (0576) (0628)

Log median income -0437 -0417 -0294

(008) (0086) (0078)

Employment-to-population ratio -0040 -0220 0244

(0261) (0260) (0240)

N 1127 1127 1127

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the natural log of state SNAPexpenditures per capita All regressions include the share of households with children and the racial shares ofthe state population

Source Annual data from Bureau of Economic Analysis national income and product accounts tables for 1990to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3543

32 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3643

Appendix C | wwwamericanprogresso

Appendix C Harkin-Miller

policy simulation results

StateMinimum wage

(2014)

SNAP enrollment (2012) Predicted change in SNAP enrollment

RateTotal

(persons)

Model 1 family

enrollment (linearprobability)

Model 2 state

enrollment (linearregression)

Model 3 s

expenditu(linear regre

Alabama $725 164 788682 -66922 -58897 -58906

Alaska $775 120 87436 -8104 -6887 -3288

Arizona $790 201 1319323 -67435 -56738 -64356

Arkansas $725 78 230489 -40977 -36063 -35248

California $800 206 7813680 -371131 -310222 -18223

Colorado $800 164 853155 -50684 -42365 -23926

Connecticut $870 91 326621 -22456 -17975 -13711

Delaware $725 186 170262 -12739 -11211 -10647

District of Columbia $825 133 84009 -5370 -4417 -3632

Florida $793 166 3208026 -195813 -164426 -13046

Georgia $725 160 1586336 -137741 -121224 -11004

Hawaii $725 96 133662 -19310 -16995 -14933

Idaho $725 92 147501 -22165 -19507 -15809

Illinois $825 95 1225084 -109088 -89742 -70955

Indiana $725 125 816233 -90818 -79928 -83985

Iowa $725 155 478011 -42716 -37594 -28556

Kansas $725 135 388269 -40082 -35275 -27461

Kentucky $725 130 568821 -60840 -53544 -52259

Louisiana $725 149 683832 -63929 -56263 -66083

Maine $750 77 101976 -16567 -14323 -15234

Maryland $725 144 846415 -81748 -71946 -38370

Massachusetts $800 130 864721 -64902 -54251 -42913

Michigan $740 146 1439141 -128801 -112140 -11022

Minnesota $725 133 713646 -74730 -65769 -37878

TABLE 10

SNAP enrollments Predicted changes by state under the Harkin-Miller bill

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3743

34 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

StateMinimum wage

(2014)

SNAP enrollment (2012) Predicted change in SNAP enrollment

RateTotal

(persons)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 state

expenditures

(linear regression)

Mississippi $725 129 386501 -41486 -36511 -46467

Missouri $750 172 1036182 -75131 -64952 -56244

Montana $790 132 132452 -10350 -8708 -5846

Nebraska $725 124 230683 -25773 -22683 -12189

Nevada $825 162 446035 -23349 -19209 -11894

New Hampshire $725 127 168404 -18359 -16157 -5735

New Jersey $825 160 1416666 -75175 -61843 -28236

New Mexico $750 149 310896 -25983 -22463 -22512

New York $800 192 3763553 -191193 -159815 -142182

North Carolina $725 174 1697193 -135417 -119179 -113503

North Dakota $725 87 61225 -9743 -8574 -4021

Ohio $795 143 1647345 -115869 -97169 -88580

Oklahoma $725 129 494053 -53006 -46650 -46854

Oregon $910 124 485326 -17036 -13328 -16398

Pennsylvania $725 161 2053643 -177315 -156052 -125586

Rhode Island $800 156 163730 -10258 -8574 -8698

South Carolina $725 94 445277 -65614 -57746 -50304

South Dakota $725 208 173749 -11586 -10197 -7458

Tennessee $725 142 914903 -89667 -78915 -99134

Texas $725 110 2863779 -362018 -318607 -253285

Utah $725 88 251107 -39658 -34902 -19390

Vermont $873 156 97792 -3823 -3055 -2475

Virginia $725 101 829771 -113723 -100086 -58212

Washington $932 72 496934 -23221 -17947 -17756

West Virginia $725 58 107875 -25792 -22699 -21665

Wisconsin $725 75 427822 -79521 -69986 -53210

Wyoming $725 164 94590 -8010 -7050 -3104

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3843

Appendix C | wwwamericanprogresso

StateMinimum wage

(2014)

SNAP expenditures (2012)Predicted change in SNAP expenditures

(millions of dollars)

Per resident

(millions of

dollars)

Total

(millions of

dollars)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 s

expenditu

(linear regre

Alabama $725 $290 $1397 -$1185 -$1043 -$1043

Alaska $775 $253 $185 -$262 -$223 -$106

Arizona $790 $257 $1686 -$935 -$787 -$892

Arkansas $725 $248 $733 -$636 -$560 -$547

California $800 $189 $7164 -$7276 -$6082 -$3573

Colorado $800 $157 $816 -$862 -$721 -$407

Connecticut $870 $191 $686 -$343 -$275 -$210

Delaware $725 $250 $229 -$205 -$180 -$171

District of Columbia $825 $366 $232 -$146 -$120 -$99

Florida $793 $294 $5676 -$4429 -$3719 -$2951

Georgia $725 $317 $3140 -$2936 -$2584 -$2346

Hawaii $725 $335 $465 -$449 -$395 -$347

Idaho $725 $225 $359 -$376 -$331 -$268

Illinois $825 $249 $3200 -$2096 -$1725 -$1364

Indiana $725 $220 $1439 -$1162 -$1023 -$1075

Iowa $725 $192 $589 -$658 -$579 -$440

Kansas $725 $159 $460 -$502 -$441 -$344

Kentucky $725 $298 $1303 -$1133 -$997 -$973

Louisiana $725 $315 $1450 -$1047 -$922 -$1083

Maine $750 $281 $373 -$267 -$231 -$246

Maryland $725 $188 $1109 -$1765 -$1553 -$828

Massachusetts $800 $206 $1366 -$1030 -$861 -$681

Michigan $740 $300 $2963 -$2400 -$2090 -$2054

Minnesota $725 $140 $755 -$1113 -$980 -$564

Mississippi $725 $326 $973 -$649 -$571 -$726

Missouri $750 $241 $1452 -$1278 -$1104 -$956

Montana $790 $190 $191 -$179 -$151 -$101

Nebraska $725 $140 $259 -$409 -$360 -$194

Nevada $825 $191 $527 -$441 -$363 -$225

New Hampshire $725 $126 $167 -$399 -$351 -$125

TABLE 11

SNAP expenditures Predicted changes by state under the Harkin-Miller bill

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3943

36 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

StateMinimum wage

(2014)

SNAP expenditures (2012)Predicted change in SNAP expenditures

(millions of dollars)

Per resident

(millions of

dollars)

Total

(millions of

dollars)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 state

expenditures

(linear regression)

New Jersey $825 $160 $1420 -$1611 -$1325 -$605

New Mexico $750 $324 $675 -$513 -$444 -$445New York $800 $287 $5616 -$3766 -$3148 -$2801

North Carolina $725 $252 $2454 -$2187 -$1925 -$1833

North Dakota $725 $128 $90 -$162 -$143 -$67

Ohio $795 $259 $2995 -$2013 -$1688 -$1539

Oklahoma $725 $248 $945 -$799 -$703 -$706

Oregon $910 $322 $1255 -$272 -$213 -$262

Pennsylvania $725 $218 $2779 -$2930 -$2579 -$2075

Rhode Island $800 $280 $294 -$173 -$144 -$147

South Carolina $725 $291 $1373 -$1337 -$1177 -$1025South Dakota $725 $198 $165 -$192 -$169 -$123

Tennessee $725 $324 $2091 -$1413 -$1243 -$1562

Texas $725 $230 $5997 -$6402 -$5634 -$4479

Utah $725 $141 $402 -$614 -$541 -$300

Vermont $873 $230 $144 -$66 -$53 -$43

Virginia $725 $173 $1413 -$2062 -$1815 -$1056

Washington $932 $244 $1682 -$350 -$270 -$267

West Virginia $725 $273 $508 -$451 -$397 -$379

Wisconsin $725 $204 $1166 -$1302 -$1146 -$871Wyoming $725 $95 $55 -$105 -$93 -$41

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4043

Appendix C | wwwamericanprogresso

FIGURE 2

Raising the minimum wage to $1010would cut taxpayer costs in every state

Predicted decreases in cost and enrollment

in SNAP in 50 states

$200+$51ndash$100

$101ndash$200

0ndash$25

$26ndash$50

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4143

38 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

Endnotes

1 US Department of Agriculture ldquoSupplementalNutrition Assistance Program (SNAP)rdquo available athttpwwwfnsusdagovsnapeligibility (last accessedFebruary 2014)

2 For this initial analysis we do not consider Harkin-Millerrsquos increase in subminimum wages for tippedworkers To do s o would increase the estimated SNAP

savings by an unknown amount

3 The Congressional Budget Office estimates thatworkers currently earning between $1010 and $1150per hour would see their wages rise under the Harkin-Miller proposal Congressional Budget O ffice ldquoTheEffects of a Minimum Wage Increase on Employmentand Family Incomerdquo (2014)

4 Marianne Page Joanne Spetz and Jane Millar ldquoDoesthe Minimum Wage Affect Welfare Caseloadsrdquo Journalof Policy Analysis and Management 24 (2) (2005)273ndash295

5 Marianne Bitler and Hilary Hoynes ldquoThe More ThingsChange the More They Stay the Same The SafetyNet Living Arrangements and Poverty in the GreatRecessionrdquo NBER Working Paper 194 49 2013

6 Sylvia Allegretto and others ldquoFast Food Poverty Wages The Public Cost of Low-Wage Jobs in the Fast-FoodIndustryrdquo (Berkeley California Center for LaborResearch and Education 2013) available at httplaborcenterberkeleyedupubliccostsfast_food_poverty_wage

7 David Neumark and William Wascher ldquoDoes a HigherMinimum Wage Enhance the Effectiveness of theEarned Income Tax Creditrdquo Industrial and LaborRelations Review 64 (4) (2011) 712ndash746

8 David Lee and Emmanuel Saez ldquoOptimal MinimumWage Policy in Competitive Labor Marketsrdquo Journal ofPublic Economics 96 (9) (2012) 739ndash749

9 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family IncomerdquoFebruary 2014

10 Jesse Rothstein ldquoIs the EITC as Good as an NITConditional Cash Transfers and Tax Incidencerdquo AmericanEconomic Journal Economic Policy 2 (1) (2010) 177ndash208

11 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family Incomerdquo

12 Dube Arindrajit 2013 rdquoMinimum Wagesand the Distribution of Family IncomerdquoUnpublished working paper Available at httpsdldropboxusercontentcomu15038936Dube_MinimumWagesFamilyIncomespdf

13 As Dube explains in ldquoThe poverty of Minimum WageFactsrdquo the simulation approach underestimate stemsfrom a number of unwarranted assumptions includingthe range of actual wage increases and the accuracy ofwage data in the Current Population Survey The causal

approach does not make these assumptions

14 Allegretto Sylvia and others 2013 ldquoCredible ResearchDesigns for Minimum Wage Studiesrdquo Working Paper148-13 University of California Berkeley Institutefor Research on Labor and Employment Available athttpwwwirleberkeleyeduworkingpapers148-13pdf

15 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family In comerdquo

16 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo (2012)

17 Ibid

18 Ibid

19 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo

20 US Department of Agriculture ldquoSupplementalNutrition Assistance Program (SNAP)rdquo

21 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo

22 Ibid

23 We deviate from the Census Bureaursquos definition ofa family unit which is ldquotwo people or more (on e ofwhom is the householder) related by bir th marriageor adoption and residing togetherrdquo US Bureau ofthe Census ldquoCurrent Population Survey DefinitionsFamilyrdquo available at httpwwwcensusgovcpsabout

cpsdefhtml (last accessed February 2014) We countas a family unit any individual residing on his or herown two or more persons residing together whodo not belong to a family in the March CPS sampleare constructed as one family in our analysis For thepurposes of food stamp allocations the consumptionresulting from this transfer is probably distributed tofamily members (rather than household members ora single individual within the household) Howeversingle individuals canmdashand domdashreceive SNAPbenefits Excluding them would fail to make theanalysis reflective of the population at large

24 Strictly the family level linear probability modelpredicts the percentage-point decrease in theprobability that an individual family will receive SNAPpayments When applied to a large number of familieshowever we are able to interpret the coefficient asa decrease in the mean of enrollmentmdashthat is a

decrease in the enrollment ratemdashby applying the lawof iterated expectations

25 We generate expenditure predictions from theenrollment modelsmdashand conversely generateenrollment predictions from the expenditure modelmdashby assuming that expenditures per enrolled familyremains the same before and after the minimum wagechange In practice this is likely to be a conservativeestimatemdashthat is to underestimate the decrease inSNAP activity Average SNAP benefits per family willalso decrease as many families that remain eligible forSNAP experience income gains

26 Wage and Hour Division ldquoMinimum Wage Laws inthe States ndash Januar y 1 2014rdquo available at httpwwwdolgovwhdminwageamericahtm (last accessedFebruary 2014)

27 See for example Sylvia Allegretto and others ldquoCredibleResearch Designs for Minimum Wage Studiesrdquo WorkingPaper 148-113 (Berkeley California Institute forResearch on Labor and Employment 2013) available athttpwwwirleberkeleyeduworkingpapers148-13pdf

28 We will report these results in a forthcoming workingpaper

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4243

Endnotes | wwwamericanprogresso

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4343

The Center for American Progress is a nonpartisan research and educational institute

dedicated to promoting a strong just and free America that ensures opportunity

for all We believe that Americans are bound together by a common commitment to

these values and we aspire to ensure that our national policies reflect these values

We work to find progressive and pragmatic solutions to significant domestic and

international problems and develop policy proposals that foster a government that

is ldquoof the people by the people and for the peoplerdquo

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3043

Appendix B | wwwamericanprogresso

his evidence suppors our use o he sauraed model as he preerred model

specificaion Moreover Dubersquos 2013 sudy shows ha more sauraed models

perorm beter han models wih jus sae and ime fixed effecs

Te esimaed enrollmen regressions a boh he amily and sae levels show large

and saisically significan coefficiens Te esimaed minimum wage effec in heexpendiures regressions991252or which we have only sae-level daa991252is also large

and saisically significan

We do no use weighed regression or he sae-level models preerring o keep

analysis o he ldquoreamenrdquo (ha is o say a minimum wage change) appropriae

o he average sae raher han he average amily or individual I insead our

primary ineres were he impac o a minimum wage change on he average amily

or he average individual we migh choose o designae he number o amilies

in each sae or he sae populaion respecively as analyic weighs in order o

obain a coefficien beter suied or such inerence

TABLE 4

SNAP enrollment

Family-level linear probability

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0048 -0047 -0040 -0043 -0042

(0013) (0013) (001) (0008) (0009)

Unemployment rate (100) 0505 0420 0280

(0083) (0086) (0082)

Log median income -0057 -0039

(0011) (0011)

Employment-to-population ratio -0239

(0038)

N 1242022 1242022 1242022 1242022 1242022

plt01 plt005 plt001

Note Robust standard errors are in parentheses Observations are clustered at the state level The outcome variable is binary and equal to one if a family is enrolledin SNAP All models include state fixed effects Census division x-year fixed effects and state-specific linear time trends All specifications except 3a include additional

controls for family size race and marital status of the family head presence of children and presence of an adult maleSource Annual data from the March Current Population Survey for 1990 to 2012 estimation includes Current Population Survey probability weights

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3143

28 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

TABLE 5

SNAP enrollment

State-level linear regression

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0028 -0024 -0026 -0031 -0031

(0014) (0013) (0013) (0012) (0012)

Unemployment rate (100) 0625 0466 0320 0339

(0087) (0088) (0085) (0083)

Log median income -0090 -0065 -0061

(0013) (0013) (0013)

Employment-to-population ratio -0282 -0248

(0037) (0038)

N 1127 1127 1127 1127 1127

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the SNAP enrollment rate All models include state fixed effects Census division x-year fixedeffects and state-specific linear time trends

Source Annual data from the March Current Population Survey for 1990 to 2012

TABLE 6

SNAP expenditures

State-level linear regression

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0181 -0149 -0156 -0153 -0190

(011) (0103) (0102) (0103) (0103)

Unemployment rate (100) 4587 4152 4232 4313

(0622) (0621) (0633) (0628)

Log median income -0246 -0261 -0294

(0075) (0078) (0078)

Employment-to-population ratio 0155 0244

(0237) (024)

N 1127 1127 1127 1127 1127

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the natural log of state SNAP expenditures per capita for 1990 to 2012 All models include state

fixed effects Census division x-year fixed effects and state-specific linear time trends

Source Annual data from Bureau of Economic Analysis national income and product accounts tables for 1990 to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3243

Appendix B | wwwamericanprogresso

TABLE 7

Comparison of specifications SNAP enrollment

Family-level linear probability

(1) (2) (3)

Log minimum wage -0028 -0049 -0042

(0014) (0017) (0009)

Unemployment rate (100) 0275 0297 0280

(0161) (0076) (0082)

Log median income -0077 -0055 -0039

(0014) (0012) (0011)

Employment-to-population ratio -0238 -0250 -0239

(0054) (004) (0038)

N 1242022 1242022 1242022

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses Observations are clustered at the state level The outcomevariable is binary or equal to one if a family is enrolled in SNAP All specifications include additional controls forfamily size race and marital status of the family head presence of children and presence of an adult male

Source Annual data from the March Current Population Survey for 1990 to 2012 estimation includes CurrentPopulation Survey probability weights

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3343

30 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

TABLE 8

Comparison of specifications SNAP enrollment

State-level linear regression

(1) (2) (3)

Log minimum wage -0019 -0035 -0031

(0009) (0012) (0012)

Unemployment rate (100) 0401 0370 0339

(0063) (0077) (0083)

Log median income -0081 -0073 -0061

(0011) (0013) (0013)

Employment-to-population ratio -0183 -0222 -0248

(0039) (0039) (0038)

N 1127 1127 1127

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the SNAP enrollment rate Allregressions include the share of households with children and the racial shares of the population

Source Annual data from the March Current Population Survey for 1990 to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3443

Appendix B | wwwamericanprogresso

TABLE 9

Comparison of specifications SNAP expenditures

State-level linear regression

(1) (2) (3)

Log minimum wage -0121 -0203 -0190

(0075) (0103) (0103)

Unemployment rate (100) 5292 5152 4313

(0464) (0576) (0628)

Log median income -0437 -0417 -0294

(008) (0086) (0078)

Employment-to-population ratio -0040 -0220 0244

(0261) (0260) (0240)

N 1127 1127 1127

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the natural log of state SNAPexpenditures per capita All regressions include the share of households with children and the racial shares ofthe state population

Source Annual data from Bureau of Economic Analysis national income and product accounts tables for 1990to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3543

32 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3643

Appendix C | wwwamericanprogresso

Appendix C Harkin-Miller

policy simulation results

StateMinimum wage

(2014)

SNAP enrollment (2012) Predicted change in SNAP enrollment

RateTotal

(persons)

Model 1 family

enrollment (linearprobability)

Model 2 state

enrollment (linearregression)

Model 3 s

expenditu(linear regre

Alabama $725 164 788682 -66922 -58897 -58906

Alaska $775 120 87436 -8104 -6887 -3288

Arizona $790 201 1319323 -67435 -56738 -64356

Arkansas $725 78 230489 -40977 -36063 -35248

California $800 206 7813680 -371131 -310222 -18223

Colorado $800 164 853155 -50684 -42365 -23926

Connecticut $870 91 326621 -22456 -17975 -13711

Delaware $725 186 170262 -12739 -11211 -10647

District of Columbia $825 133 84009 -5370 -4417 -3632

Florida $793 166 3208026 -195813 -164426 -13046

Georgia $725 160 1586336 -137741 -121224 -11004

Hawaii $725 96 133662 -19310 -16995 -14933

Idaho $725 92 147501 -22165 -19507 -15809

Illinois $825 95 1225084 -109088 -89742 -70955

Indiana $725 125 816233 -90818 -79928 -83985

Iowa $725 155 478011 -42716 -37594 -28556

Kansas $725 135 388269 -40082 -35275 -27461

Kentucky $725 130 568821 -60840 -53544 -52259

Louisiana $725 149 683832 -63929 -56263 -66083

Maine $750 77 101976 -16567 -14323 -15234

Maryland $725 144 846415 -81748 -71946 -38370

Massachusetts $800 130 864721 -64902 -54251 -42913

Michigan $740 146 1439141 -128801 -112140 -11022

Minnesota $725 133 713646 -74730 -65769 -37878

TABLE 10

SNAP enrollments Predicted changes by state under the Harkin-Miller bill

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3743

34 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

StateMinimum wage

(2014)

SNAP enrollment (2012) Predicted change in SNAP enrollment

RateTotal

(persons)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 state

expenditures

(linear regression)

Mississippi $725 129 386501 -41486 -36511 -46467

Missouri $750 172 1036182 -75131 -64952 -56244

Montana $790 132 132452 -10350 -8708 -5846

Nebraska $725 124 230683 -25773 -22683 -12189

Nevada $825 162 446035 -23349 -19209 -11894

New Hampshire $725 127 168404 -18359 -16157 -5735

New Jersey $825 160 1416666 -75175 -61843 -28236

New Mexico $750 149 310896 -25983 -22463 -22512

New York $800 192 3763553 -191193 -159815 -142182

North Carolina $725 174 1697193 -135417 -119179 -113503

North Dakota $725 87 61225 -9743 -8574 -4021

Ohio $795 143 1647345 -115869 -97169 -88580

Oklahoma $725 129 494053 -53006 -46650 -46854

Oregon $910 124 485326 -17036 -13328 -16398

Pennsylvania $725 161 2053643 -177315 -156052 -125586

Rhode Island $800 156 163730 -10258 -8574 -8698

South Carolina $725 94 445277 -65614 -57746 -50304

South Dakota $725 208 173749 -11586 -10197 -7458

Tennessee $725 142 914903 -89667 -78915 -99134

Texas $725 110 2863779 -362018 -318607 -253285

Utah $725 88 251107 -39658 -34902 -19390

Vermont $873 156 97792 -3823 -3055 -2475

Virginia $725 101 829771 -113723 -100086 -58212

Washington $932 72 496934 -23221 -17947 -17756

West Virginia $725 58 107875 -25792 -22699 -21665

Wisconsin $725 75 427822 -79521 -69986 -53210

Wyoming $725 164 94590 -8010 -7050 -3104

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3843

Appendix C | wwwamericanprogresso

StateMinimum wage

(2014)

SNAP expenditures (2012)Predicted change in SNAP expenditures

(millions of dollars)

Per resident

(millions of

dollars)

Total

(millions of

dollars)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 s

expenditu

(linear regre

Alabama $725 $290 $1397 -$1185 -$1043 -$1043

Alaska $775 $253 $185 -$262 -$223 -$106

Arizona $790 $257 $1686 -$935 -$787 -$892

Arkansas $725 $248 $733 -$636 -$560 -$547

California $800 $189 $7164 -$7276 -$6082 -$3573

Colorado $800 $157 $816 -$862 -$721 -$407

Connecticut $870 $191 $686 -$343 -$275 -$210

Delaware $725 $250 $229 -$205 -$180 -$171

District of Columbia $825 $366 $232 -$146 -$120 -$99

Florida $793 $294 $5676 -$4429 -$3719 -$2951

Georgia $725 $317 $3140 -$2936 -$2584 -$2346

Hawaii $725 $335 $465 -$449 -$395 -$347

Idaho $725 $225 $359 -$376 -$331 -$268

Illinois $825 $249 $3200 -$2096 -$1725 -$1364

Indiana $725 $220 $1439 -$1162 -$1023 -$1075

Iowa $725 $192 $589 -$658 -$579 -$440

Kansas $725 $159 $460 -$502 -$441 -$344

Kentucky $725 $298 $1303 -$1133 -$997 -$973

Louisiana $725 $315 $1450 -$1047 -$922 -$1083

Maine $750 $281 $373 -$267 -$231 -$246

Maryland $725 $188 $1109 -$1765 -$1553 -$828

Massachusetts $800 $206 $1366 -$1030 -$861 -$681

Michigan $740 $300 $2963 -$2400 -$2090 -$2054

Minnesota $725 $140 $755 -$1113 -$980 -$564

Mississippi $725 $326 $973 -$649 -$571 -$726

Missouri $750 $241 $1452 -$1278 -$1104 -$956

Montana $790 $190 $191 -$179 -$151 -$101

Nebraska $725 $140 $259 -$409 -$360 -$194

Nevada $825 $191 $527 -$441 -$363 -$225

New Hampshire $725 $126 $167 -$399 -$351 -$125

TABLE 11

SNAP expenditures Predicted changes by state under the Harkin-Miller bill

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3943

36 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

StateMinimum wage

(2014)

SNAP expenditures (2012)Predicted change in SNAP expenditures

(millions of dollars)

Per resident

(millions of

dollars)

Total

(millions of

dollars)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 state

expenditures

(linear regression)

New Jersey $825 $160 $1420 -$1611 -$1325 -$605

New Mexico $750 $324 $675 -$513 -$444 -$445New York $800 $287 $5616 -$3766 -$3148 -$2801

North Carolina $725 $252 $2454 -$2187 -$1925 -$1833

North Dakota $725 $128 $90 -$162 -$143 -$67

Ohio $795 $259 $2995 -$2013 -$1688 -$1539

Oklahoma $725 $248 $945 -$799 -$703 -$706

Oregon $910 $322 $1255 -$272 -$213 -$262

Pennsylvania $725 $218 $2779 -$2930 -$2579 -$2075

Rhode Island $800 $280 $294 -$173 -$144 -$147

South Carolina $725 $291 $1373 -$1337 -$1177 -$1025South Dakota $725 $198 $165 -$192 -$169 -$123

Tennessee $725 $324 $2091 -$1413 -$1243 -$1562

Texas $725 $230 $5997 -$6402 -$5634 -$4479

Utah $725 $141 $402 -$614 -$541 -$300

Vermont $873 $230 $144 -$66 -$53 -$43

Virginia $725 $173 $1413 -$2062 -$1815 -$1056

Washington $932 $244 $1682 -$350 -$270 -$267

West Virginia $725 $273 $508 -$451 -$397 -$379

Wisconsin $725 $204 $1166 -$1302 -$1146 -$871Wyoming $725 $95 $55 -$105 -$93 -$41

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4043

Appendix C | wwwamericanprogresso

FIGURE 2

Raising the minimum wage to $1010would cut taxpayer costs in every state

Predicted decreases in cost and enrollment

in SNAP in 50 states

$200+$51ndash$100

$101ndash$200

0ndash$25

$26ndash$50

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4143

38 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

Endnotes

1 US Department of Agriculture ldquoSupplementalNutrition Assistance Program (SNAP)rdquo available athttpwwwfnsusdagovsnapeligibility (last accessedFebruary 2014)

2 For this initial analysis we do not consider Harkin-Millerrsquos increase in subminimum wages for tippedworkers To do s o would increase the estimated SNAP

savings by an unknown amount

3 The Congressional Budget Office estimates thatworkers currently earning between $1010 and $1150per hour would see their wages rise under the Harkin-Miller proposal Congressional Budget O ffice ldquoTheEffects of a Minimum Wage Increase on Employmentand Family Incomerdquo (2014)

4 Marianne Page Joanne Spetz and Jane Millar ldquoDoesthe Minimum Wage Affect Welfare Caseloadsrdquo Journalof Policy Analysis and Management 24 (2) (2005)273ndash295

5 Marianne Bitler and Hilary Hoynes ldquoThe More ThingsChange the More They Stay the Same The SafetyNet Living Arrangements and Poverty in the GreatRecessionrdquo NBER Working Paper 194 49 2013

6 Sylvia Allegretto and others ldquoFast Food Poverty Wages The Public Cost of Low-Wage Jobs in the Fast-FoodIndustryrdquo (Berkeley California Center for LaborResearch and Education 2013) available at httplaborcenterberkeleyedupubliccostsfast_food_poverty_wage

7 David Neumark and William Wascher ldquoDoes a HigherMinimum Wage Enhance the Effectiveness of theEarned Income Tax Creditrdquo Industrial and LaborRelations Review 64 (4) (2011) 712ndash746

8 David Lee and Emmanuel Saez ldquoOptimal MinimumWage Policy in Competitive Labor Marketsrdquo Journal ofPublic Economics 96 (9) (2012) 739ndash749

9 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family IncomerdquoFebruary 2014

10 Jesse Rothstein ldquoIs the EITC as Good as an NITConditional Cash Transfers and Tax Incidencerdquo AmericanEconomic Journal Economic Policy 2 (1) (2010) 177ndash208

11 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family Incomerdquo

12 Dube Arindrajit 2013 rdquoMinimum Wagesand the Distribution of Family IncomerdquoUnpublished working paper Available at httpsdldropboxusercontentcomu15038936Dube_MinimumWagesFamilyIncomespdf

13 As Dube explains in ldquoThe poverty of Minimum WageFactsrdquo the simulation approach underestimate stemsfrom a number of unwarranted assumptions includingthe range of actual wage increases and the accuracy ofwage data in the Current Population Survey The causal

approach does not make these assumptions

14 Allegretto Sylvia and others 2013 ldquoCredible ResearchDesigns for Minimum Wage Studiesrdquo Working Paper148-13 University of California Berkeley Institutefor Research on Labor and Employment Available athttpwwwirleberkeleyeduworkingpapers148-13pdf

15 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family In comerdquo

16 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo (2012)

17 Ibid

18 Ibid

19 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo

20 US Department of Agriculture ldquoSupplementalNutrition Assistance Program (SNAP)rdquo

21 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo

22 Ibid

23 We deviate from the Census Bureaursquos definition ofa family unit which is ldquotwo people or more (on e ofwhom is the householder) related by bir th marriageor adoption and residing togetherrdquo US Bureau ofthe Census ldquoCurrent Population Survey DefinitionsFamilyrdquo available at httpwwwcensusgovcpsabout

cpsdefhtml (last accessed February 2014) We countas a family unit any individual residing on his or herown two or more persons residing together whodo not belong to a family in the March CPS sampleare constructed as one family in our analysis For thepurposes of food stamp allocations the consumptionresulting from this transfer is probably distributed tofamily members (rather than household members ora single individual within the household) Howeversingle individuals canmdashand domdashreceive SNAPbenefits Excluding them would fail to make theanalysis reflective of the population at large

24 Strictly the family level linear probability modelpredicts the percentage-point decrease in theprobability that an individual family will receive SNAPpayments When applied to a large number of familieshowever we are able to interpret the coefficient asa decrease in the mean of enrollmentmdashthat is a

decrease in the enrollment ratemdashby applying the lawof iterated expectations

25 We generate expenditure predictions from theenrollment modelsmdashand conversely generateenrollment predictions from the expenditure modelmdashby assuming that expenditures per enrolled familyremains the same before and after the minimum wagechange In practice this is likely to be a conservativeestimatemdashthat is to underestimate the decrease inSNAP activity Average SNAP benefits per family willalso decrease as many families that remain eligible forSNAP experience income gains

26 Wage and Hour Division ldquoMinimum Wage Laws inthe States ndash Januar y 1 2014rdquo available at httpwwwdolgovwhdminwageamericahtm (last accessedFebruary 2014)

27 See for example Sylvia Allegretto and others ldquoCredibleResearch Designs for Minimum Wage Studiesrdquo WorkingPaper 148-113 (Berkeley California Institute forResearch on Labor and Employment 2013) available athttpwwwirleberkeleyeduworkingpapers148-13pdf

28 We will report these results in a forthcoming workingpaper

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4243

Endnotes | wwwamericanprogresso

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4343

The Center for American Progress is a nonpartisan research and educational institute

dedicated to promoting a strong just and free America that ensures opportunity

for all We believe that Americans are bound together by a common commitment to

these values and we aspire to ensure that our national policies reflect these values

We work to find progressive and pragmatic solutions to significant domestic and

international problems and develop policy proposals that foster a government that

is ldquoof the people by the people and for the peoplerdquo

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3143

28 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

TABLE 5

SNAP enrollment

State-level linear regression

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0028 -0024 -0026 -0031 -0031

(0014) (0013) (0013) (0012) (0012)

Unemployment rate (100) 0625 0466 0320 0339

(0087) (0088) (0085) (0083)

Log median income -0090 -0065 -0061

(0013) (0013) (0013)

Employment-to-population ratio -0282 -0248

(0037) (0038)

N 1127 1127 1127 1127 1127

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the SNAP enrollment rate All models include state fixed effects Census division x-year fixedeffects and state-specific linear time trends

Source Annual data from the March Current Population Survey for 1990 to 2012

TABLE 6

SNAP expenditures

State-level linear regression

(3a) (3b) (3c) (3d) (3e)

Log minimum wage -0181 -0149 -0156 -0153 -0190

(011) (0103) (0102) (0103) (0103)

Unemployment rate (100) 4587 4152 4232 4313

(0622) (0621) (0633) (0628)

Log median income -0246 -0261 -0294

(0075) (0078) (0078)

Employment-to-population ratio 0155 0244

(0237) (024)

N 1127 1127 1127 1127 1127

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the natural log of state SNAP expenditures per capita for 1990 to 2012 All models include state

fixed effects Census division x-year fixed effects and state-specific linear time trends

Source Annual data from Bureau of Economic Analysis national income and product accounts tables for 1990 to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3243

Appendix B | wwwamericanprogresso

TABLE 7

Comparison of specifications SNAP enrollment

Family-level linear probability

(1) (2) (3)

Log minimum wage -0028 -0049 -0042

(0014) (0017) (0009)

Unemployment rate (100) 0275 0297 0280

(0161) (0076) (0082)

Log median income -0077 -0055 -0039

(0014) (0012) (0011)

Employment-to-population ratio -0238 -0250 -0239

(0054) (004) (0038)

N 1242022 1242022 1242022

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses Observations are clustered at the state level The outcomevariable is binary or equal to one if a family is enrolled in SNAP All specifications include additional controls forfamily size race and marital status of the family head presence of children and presence of an adult male

Source Annual data from the March Current Population Survey for 1990 to 2012 estimation includes CurrentPopulation Survey probability weights

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3343

30 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

TABLE 8

Comparison of specifications SNAP enrollment

State-level linear regression

(1) (2) (3)

Log minimum wage -0019 -0035 -0031

(0009) (0012) (0012)

Unemployment rate (100) 0401 0370 0339

(0063) (0077) (0083)

Log median income -0081 -0073 -0061

(0011) (0013) (0013)

Employment-to-population ratio -0183 -0222 -0248

(0039) (0039) (0038)

N 1127 1127 1127

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the SNAP enrollment rate Allregressions include the share of households with children and the racial shares of the population

Source Annual data from the March Current Population Survey for 1990 to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3443

Appendix B | wwwamericanprogresso

TABLE 9

Comparison of specifications SNAP expenditures

State-level linear regression

(1) (2) (3)

Log minimum wage -0121 -0203 -0190

(0075) (0103) (0103)

Unemployment rate (100) 5292 5152 4313

(0464) (0576) (0628)

Log median income -0437 -0417 -0294

(008) (0086) (0078)

Employment-to-population ratio -0040 -0220 0244

(0261) (0260) (0240)

N 1127 1127 1127

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the natural log of state SNAPexpenditures per capita All regressions include the share of households with children and the racial shares ofthe state population

Source Annual data from Bureau of Economic Analysis national income and product accounts tables for 1990to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3543

32 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3643

Appendix C | wwwamericanprogresso

Appendix C Harkin-Miller

policy simulation results

StateMinimum wage

(2014)

SNAP enrollment (2012) Predicted change in SNAP enrollment

RateTotal

(persons)

Model 1 family

enrollment (linearprobability)

Model 2 state

enrollment (linearregression)

Model 3 s

expenditu(linear regre

Alabama $725 164 788682 -66922 -58897 -58906

Alaska $775 120 87436 -8104 -6887 -3288

Arizona $790 201 1319323 -67435 -56738 -64356

Arkansas $725 78 230489 -40977 -36063 -35248

California $800 206 7813680 -371131 -310222 -18223

Colorado $800 164 853155 -50684 -42365 -23926

Connecticut $870 91 326621 -22456 -17975 -13711

Delaware $725 186 170262 -12739 -11211 -10647

District of Columbia $825 133 84009 -5370 -4417 -3632

Florida $793 166 3208026 -195813 -164426 -13046

Georgia $725 160 1586336 -137741 -121224 -11004

Hawaii $725 96 133662 -19310 -16995 -14933

Idaho $725 92 147501 -22165 -19507 -15809

Illinois $825 95 1225084 -109088 -89742 -70955

Indiana $725 125 816233 -90818 -79928 -83985

Iowa $725 155 478011 -42716 -37594 -28556

Kansas $725 135 388269 -40082 -35275 -27461

Kentucky $725 130 568821 -60840 -53544 -52259

Louisiana $725 149 683832 -63929 -56263 -66083

Maine $750 77 101976 -16567 -14323 -15234

Maryland $725 144 846415 -81748 -71946 -38370

Massachusetts $800 130 864721 -64902 -54251 -42913

Michigan $740 146 1439141 -128801 -112140 -11022

Minnesota $725 133 713646 -74730 -65769 -37878

TABLE 10

SNAP enrollments Predicted changes by state under the Harkin-Miller bill

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3743

34 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

StateMinimum wage

(2014)

SNAP enrollment (2012) Predicted change in SNAP enrollment

RateTotal

(persons)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 state

expenditures

(linear regression)

Mississippi $725 129 386501 -41486 -36511 -46467

Missouri $750 172 1036182 -75131 -64952 -56244

Montana $790 132 132452 -10350 -8708 -5846

Nebraska $725 124 230683 -25773 -22683 -12189

Nevada $825 162 446035 -23349 -19209 -11894

New Hampshire $725 127 168404 -18359 -16157 -5735

New Jersey $825 160 1416666 -75175 -61843 -28236

New Mexico $750 149 310896 -25983 -22463 -22512

New York $800 192 3763553 -191193 -159815 -142182

North Carolina $725 174 1697193 -135417 -119179 -113503

North Dakota $725 87 61225 -9743 -8574 -4021

Ohio $795 143 1647345 -115869 -97169 -88580

Oklahoma $725 129 494053 -53006 -46650 -46854

Oregon $910 124 485326 -17036 -13328 -16398

Pennsylvania $725 161 2053643 -177315 -156052 -125586

Rhode Island $800 156 163730 -10258 -8574 -8698

South Carolina $725 94 445277 -65614 -57746 -50304

South Dakota $725 208 173749 -11586 -10197 -7458

Tennessee $725 142 914903 -89667 -78915 -99134

Texas $725 110 2863779 -362018 -318607 -253285

Utah $725 88 251107 -39658 -34902 -19390

Vermont $873 156 97792 -3823 -3055 -2475

Virginia $725 101 829771 -113723 -100086 -58212

Washington $932 72 496934 -23221 -17947 -17756

West Virginia $725 58 107875 -25792 -22699 -21665

Wisconsin $725 75 427822 -79521 -69986 -53210

Wyoming $725 164 94590 -8010 -7050 -3104

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3843

Appendix C | wwwamericanprogresso

StateMinimum wage

(2014)

SNAP expenditures (2012)Predicted change in SNAP expenditures

(millions of dollars)

Per resident

(millions of

dollars)

Total

(millions of

dollars)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 s

expenditu

(linear regre

Alabama $725 $290 $1397 -$1185 -$1043 -$1043

Alaska $775 $253 $185 -$262 -$223 -$106

Arizona $790 $257 $1686 -$935 -$787 -$892

Arkansas $725 $248 $733 -$636 -$560 -$547

California $800 $189 $7164 -$7276 -$6082 -$3573

Colorado $800 $157 $816 -$862 -$721 -$407

Connecticut $870 $191 $686 -$343 -$275 -$210

Delaware $725 $250 $229 -$205 -$180 -$171

District of Columbia $825 $366 $232 -$146 -$120 -$99

Florida $793 $294 $5676 -$4429 -$3719 -$2951

Georgia $725 $317 $3140 -$2936 -$2584 -$2346

Hawaii $725 $335 $465 -$449 -$395 -$347

Idaho $725 $225 $359 -$376 -$331 -$268

Illinois $825 $249 $3200 -$2096 -$1725 -$1364

Indiana $725 $220 $1439 -$1162 -$1023 -$1075

Iowa $725 $192 $589 -$658 -$579 -$440

Kansas $725 $159 $460 -$502 -$441 -$344

Kentucky $725 $298 $1303 -$1133 -$997 -$973

Louisiana $725 $315 $1450 -$1047 -$922 -$1083

Maine $750 $281 $373 -$267 -$231 -$246

Maryland $725 $188 $1109 -$1765 -$1553 -$828

Massachusetts $800 $206 $1366 -$1030 -$861 -$681

Michigan $740 $300 $2963 -$2400 -$2090 -$2054

Minnesota $725 $140 $755 -$1113 -$980 -$564

Mississippi $725 $326 $973 -$649 -$571 -$726

Missouri $750 $241 $1452 -$1278 -$1104 -$956

Montana $790 $190 $191 -$179 -$151 -$101

Nebraska $725 $140 $259 -$409 -$360 -$194

Nevada $825 $191 $527 -$441 -$363 -$225

New Hampshire $725 $126 $167 -$399 -$351 -$125

TABLE 11

SNAP expenditures Predicted changes by state under the Harkin-Miller bill

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3943

36 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

StateMinimum wage

(2014)

SNAP expenditures (2012)Predicted change in SNAP expenditures

(millions of dollars)

Per resident

(millions of

dollars)

Total

(millions of

dollars)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 state

expenditures

(linear regression)

New Jersey $825 $160 $1420 -$1611 -$1325 -$605

New Mexico $750 $324 $675 -$513 -$444 -$445New York $800 $287 $5616 -$3766 -$3148 -$2801

North Carolina $725 $252 $2454 -$2187 -$1925 -$1833

North Dakota $725 $128 $90 -$162 -$143 -$67

Ohio $795 $259 $2995 -$2013 -$1688 -$1539

Oklahoma $725 $248 $945 -$799 -$703 -$706

Oregon $910 $322 $1255 -$272 -$213 -$262

Pennsylvania $725 $218 $2779 -$2930 -$2579 -$2075

Rhode Island $800 $280 $294 -$173 -$144 -$147

South Carolina $725 $291 $1373 -$1337 -$1177 -$1025South Dakota $725 $198 $165 -$192 -$169 -$123

Tennessee $725 $324 $2091 -$1413 -$1243 -$1562

Texas $725 $230 $5997 -$6402 -$5634 -$4479

Utah $725 $141 $402 -$614 -$541 -$300

Vermont $873 $230 $144 -$66 -$53 -$43

Virginia $725 $173 $1413 -$2062 -$1815 -$1056

Washington $932 $244 $1682 -$350 -$270 -$267

West Virginia $725 $273 $508 -$451 -$397 -$379

Wisconsin $725 $204 $1166 -$1302 -$1146 -$871Wyoming $725 $95 $55 -$105 -$93 -$41

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4043

Appendix C | wwwamericanprogresso

FIGURE 2

Raising the minimum wage to $1010would cut taxpayer costs in every state

Predicted decreases in cost and enrollment

in SNAP in 50 states

$200+$51ndash$100

$101ndash$200

0ndash$25

$26ndash$50

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4143

38 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

Endnotes

1 US Department of Agriculture ldquoSupplementalNutrition Assistance Program (SNAP)rdquo available athttpwwwfnsusdagovsnapeligibility (last accessedFebruary 2014)

2 For this initial analysis we do not consider Harkin-Millerrsquos increase in subminimum wages for tippedworkers To do s o would increase the estimated SNAP

savings by an unknown amount

3 The Congressional Budget Office estimates thatworkers currently earning between $1010 and $1150per hour would see their wages rise under the Harkin-Miller proposal Congressional Budget O ffice ldquoTheEffects of a Minimum Wage Increase on Employmentand Family Incomerdquo (2014)

4 Marianne Page Joanne Spetz and Jane Millar ldquoDoesthe Minimum Wage Affect Welfare Caseloadsrdquo Journalof Policy Analysis and Management 24 (2) (2005)273ndash295

5 Marianne Bitler and Hilary Hoynes ldquoThe More ThingsChange the More They Stay the Same The SafetyNet Living Arrangements and Poverty in the GreatRecessionrdquo NBER Working Paper 194 49 2013

6 Sylvia Allegretto and others ldquoFast Food Poverty Wages The Public Cost of Low-Wage Jobs in the Fast-FoodIndustryrdquo (Berkeley California Center for LaborResearch and Education 2013) available at httplaborcenterberkeleyedupubliccostsfast_food_poverty_wage

7 David Neumark and William Wascher ldquoDoes a HigherMinimum Wage Enhance the Effectiveness of theEarned Income Tax Creditrdquo Industrial and LaborRelations Review 64 (4) (2011) 712ndash746

8 David Lee and Emmanuel Saez ldquoOptimal MinimumWage Policy in Competitive Labor Marketsrdquo Journal ofPublic Economics 96 (9) (2012) 739ndash749

9 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family IncomerdquoFebruary 2014

10 Jesse Rothstein ldquoIs the EITC as Good as an NITConditional Cash Transfers and Tax Incidencerdquo AmericanEconomic Journal Economic Policy 2 (1) (2010) 177ndash208

11 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family Incomerdquo

12 Dube Arindrajit 2013 rdquoMinimum Wagesand the Distribution of Family IncomerdquoUnpublished working paper Available at httpsdldropboxusercontentcomu15038936Dube_MinimumWagesFamilyIncomespdf

13 As Dube explains in ldquoThe poverty of Minimum WageFactsrdquo the simulation approach underestimate stemsfrom a number of unwarranted assumptions includingthe range of actual wage increases and the accuracy ofwage data in the Current Population Survey The causal

approach does not make these assumptions

14 Allegretto Sylvia and others 2013 ldquoCredible ResearchDesigns for Minimum Wage Studiesrdquo Working Paper148-13 University of California Berkeley Institutefor Research on Labor and Employment Available athttpwwwirleberkeleyeduworkingpapers148-13pdf

15 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family In comerdquo

16 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo (2012)

17 Ibid

18 Ibid

19 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo

20 US Department of Agriculture ldquoSupplementalNutrition Assistance Program (SNAP)rdquo

21 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo

22 Ibid

23 We deviate from the Census Bureaursquos definition ofa family unit which is ldquotwo people or more (on e ofwhom is the householder) related by bir th marriageor adoption and residing togetherrdquo US Bureau ofthe Census ldquoCurrent Population Survey DefinitionsFamilyrdquo available at httpwwwcensusgovcpsabout

cpsdefhtml (last accessed February 2014) We countas a family unit any individual residing on his or herown two or more persons residing together whodo not belong to a family in the March CPS sampleare constructed as one family in our analysis For thepurposes of food stamp allocations the consumptionresulting from this transfer is probably distributed tofamily members (rather than household members ora single individual within the household) Howeversingle individuals canmdashand domdashreceive SNAPbenefits Excluding them would fail to make theanalysis reflective of the population at large

24 Strictly the family level linear probability modelpredicts the percentage-point decrease in theprobability that an individual family will receive SNAPpayments When applied to a large number of familieshowever we are able to interpret the coefficient asa decrease in the mean of enrollmentmdashthat is a

decrease in the enrollment ratemdashby applying the lawof iterated expectations

25 We generate expenditure predictions from theenrollment modelsmdashand conversely generateenrollment predictions from the expenditure modelmdashby assuming that expenditures per enrolled familyremains the same before and after the minimum wagechange In practice this is likely to be a conservativeestimatemdashthat is to underestimate the decrease inSNAP activity Average SNAP benefits per family willalso decrease as many families that remain eligible forSNAP experience income gains

26 Wage and Hour Division ldquoMinimum Wage Laws inthe States ndash Januar y 1 2014rdquo available at httpwwwdolgovwhdminwageamericahtm (last accessedFebruary 2014)

27 See for example Sylvia Allegretto and others ldquoCredibleResearch Designs for Minimum Wage Studiesrdquo WorkingPaper 148-113 (Berkeley California Institute forResearch on Labor and Employment 2013) available athttpwwwirleberkeleyeduworkingpapers148-13pdf

28 We will report these results in a forthcoming workingpaper

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4243

Endnotes | wwwamericanprogresso

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4343

The Center for American Progress is a nonpartisan research and educational institute

dedicated to promoting a strong just and free America that ensures opportunity

for all We believe that Americans are bound together by a common commitment to

these values and we aspire to ensure that our national policies reflect these values

We work to find progressive and pragmatic solutions to significant domestic and

international problems and develop policy proposals that foster a government that

is ldquoof the people by the people and for the peoplerdquo

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3243

Appendix B | wwwamericanprogresso

TABLE 7

Comparison of specifications SNAP enrollment

Family-level linear probability

(1) (2) (3)

Log minimum wage -0028 -0049 -0042

(0014) (0017) (0009)

Unemployment rate (100) 0275 0297 0280

(0161) (0076) (0082)

Log median income -0077 -0055 -0039

(0014) (0012) (0011)

Employment-to-population ratio -0238 -0250 -0239

(0054) (004) (0038)

N 1242022 1242022 1242022

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses Observations are clustered at the state level The outcomevariable is binary or equal to one if a family is enrolled in SNAP All specifications include additional controls forfamily size race and marital status of the family head presence of children and presence of an adult male

Source Annual data from the March Current Population Survey for 1990 to 2012 estimation includes CurrentPopulation Survey probability weights

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3343

30 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

TABLE 8

Comparison of specifications SNAP enrollment

State-level linear regression

(1) (2) (3)

Log minimum wage -0019 -0035 -0031

(0009) (0012) (0012)

Unemployment rate (100) 0401 0370 0339

(0063) (0077) (0083)

Log median income -0081 -0073 -0061

(0011) (0013) (0013)

Employment-to-population ratio -0183 -0222 -0248

(0039) (0039) (0038)

N 1127 1127 1127

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the SNAP enrollment rate Allregressions include the share of households with children and the racial shares of the population

Source Annual data from the March Current Population Survey for 1990 to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3443

Appendix B | wwwamericanprogresso

TABLE 9

Comparison of specifications SNAP expenditures

State-level linear regression

(1) (2) (3)

Log minimum wage -0121 -0203 -0190

(0075) (0103) (0103)

Unemployment rate (100) 5292 5152 4313

(0464) (0576) (0628)

Log median income -0437 -0417 -0294

(008) (0086) (0078)

Employment-to-population ratio -0040 -0220 0244

(0261) (0260) (0240)

N 1127 1127 1127

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the natural log of state SNAPexpenditures per capita All regressions include the share of households with children and the racial shares ofthe state population

Source Annual data from Bureau of Economic Analysis national income and product accounts tables for 1990to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3543

32 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3643

Appendix C | wwwamericanprogresso

Appendix C Harkin-Miller

policy simulation results

StateMinimum wage

(2014)

SNAP enrollment (2012) Predicted change in SNAP enrollment

RateTotal

(persons)

Model 1 family

enrollment (linearprobability)

Model 2 state

enrollment (linearregression)

Model 3 s

expenditu(linear regre

Alabama $725 164 788682 -66922 -58897 -58906

Alaska $775 120 87436 -8104 -6887 -3288

Arizona $790 201 1319323 -67435 -56738 -64356

Arkansas $725 78 230489 -40977 -36063 -35248

California $800 206 7813680 -371131 -310222 -18223

Colorado $800 164 853155 -50684 -42365 -23926

Connecticut $870 91 326621 -22456 -17975 -13711

Delaware $725 186 170262 -12739 -11211 -10647

District of Columbia $825 133 84009 -5370 -4417 -3632

Florida $793 166 3208026 -195813 -164426 -13046

Georgia $725 160 1586336 -137741 -121224 -11004

Hawaii $725 96 133662 -19310 -16995 -14933

Idaho $725 92 147501 -22165 -19507 -15809

Illinois $825 95 1225084 -109088 -89742 -70955

Indiana $725 125 816233 -90818 -79928 -83985

Iowa $725 155 478011 -42716 -37594 -28556

Kansas $725 135 388269 -40082 -35275 -27461

Kentucky $725 130 568821 -60840 -53544 -52259

Louisiana $725 149 683832 -63929 -56263 -66083

Maine $750 77 101976 -16567 -14323 -15234

Maryland $725 144 846415 -81748 -71946 -38370

Massachusetts $800 130 864721 -64902 -54251 -42913

Michigan $740 146 1439141 -128801 -112140 -11022

Minnesota $725 133 713646 -74730 -65769 -37878

TABLE 10

SNAP enrollments Predicted changes by state under the Harkin-Miller bill

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3743

34 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

StateMinimum wage

(2014)

SNAP enrollment (2012) Predicted change in SNAP enrollment

RateTotal

(persons)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 state

expenditures

(linear regression)

Mississippi $725 129 386501 -41486 -36511 -46467

Missouri $750 172 1036182 -75131 -64952 -56244

Montana $790 132 132452 -10350 -8708 -5846

Nebraska $725 124 230683 -25773 -22683 -12189

Nevada $825 162 446035 -23349 -19209 -11894

New Hampshire $725 127 168404 -18359 -16157 -5735

New Jersey $825 160 1416666 -75175 -61843 -28236

New Mexico $750 149 310896 -25983 -22463 -22512

New York $800 192 3763553 -191193 -159815 -142182

North Carolina $725 174 1697193 -135417 -119179 -113503

North Dakota $725 87 61225 -9743 -8574 -4021

Ohio $795 143 1647345 -115869 -97169 -88580

Oklahoma $725 129 494053 -53006 -46650 -46854

Oregon $910 124 485326 -17036 -13328 -16398

Pennsylvania $725 161 2053643 -177315 -156052 -125586

Rhode Island $800 156 163730 -10258 -8574 -8698

South Carolina $725 94 445277 -65614 -57746 -50304

South Dakota $725 208 173749 -11586 -10197 -7458

Tennessee $725 142 914903 -89667 -78915 -99134

Texas $725 110 2863779 -362018 -318607 -253285

Utah $725 88 251107 -39658 -34902 -19390

Vermont $873 156 97792 -3823 -3055 -2475

Virginia $725 101 829771 -113723 -100086 -58212

Washington $932 72 496934 -23221 -17947 -17756

West Virginia $725 58 107875 -25792 -22699 -21665

Wisconsin $725 75 427822 -79521 -69986 -53210

Wyoming $725 164 94590 -8010 -7050 -3104

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3843

Appendix C | wwwamericanprogresso

StateMinimum wage

(2014)

SNAP expenditures (2012)Predicted change in SNAP expenditures

(millions of dollars)

Per resident

(millions of

dollars)

Total

(millions of

dollars)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 s

expenditu

(linear regre

Alabama $725 $290 $1397 -$1185 -$1043 -$1043

Alaska $775 $253 $185 -$262 -$223 -$106

Arizona $790 $257 $1686 -$935 -$787 -$892

Arkansas $725 $248 $733 -$636 -$560 -$547

California $800 $189 $7164 -$7276 -$6082 -$3573

Colorado $800 $157 $816 -$862 -$721 -$407

Connecticut $870 $191 $686 -$343 -$275 -$210

Delaware $725 $250 $229 -$205 -$180 -$171

District of Columbia $825 $366 $232 -$146 -$120 -$99

Florida $793 $294 $5676 -$4429 -$3719 -$2951

Georgia $725 $317 $3140 -$2936 -$2584 -$2346

Hawaii $725 $335 $465 -$449 -$395 -$347

Idaho $725 $225 $359 -$376 -$331 -$268

Illinois $825 $249 $3200 -$2096 -$1725 -$1364

Indiana $725 $220 $1439 -$1162 -$1023 -$1075

Iowa $725 $192 $589 -$658 -$579 -$440

Kansas $725 $159 $460 -$502 -$441 -$344

Kentucky $725 $298 $1303 -$1133 -$997 -$973

Louisiana $725 $315 $1450 -$1047 -$922 -$1083

Maine $750 $281 $373 -$267 -$231 -$246

Maryland $725 $188 $1109 -$1765 -$1553 -$828

Massachusetts $800 $206 $1366 -$1030 -$861 -$681

Michigan $740 $300 $2963 -$2400 -$2090 -$2054

Minnesota $725 $140 $755 -$1113 -$980 -$564

Mississippi $725 $326 $973 -$649 -$571 -$726

Missouri $750 $241 $1452 -$1278 -$1104 -$956

Montana $790 $190 $191 -$179 -$151 -$101

Nebraska $725 $140 $259 -$409 -$360 -$194

Nevada $825 $191 $527 -$441 -$363 -$225

New Hampshire $725 $126 $167 -$399 -$351 -$125

TABLE 11

SNAP expenditures Predicted changes by state under the Harkin-Miller bill

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3943

36 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

StateMinimum wage

(2014)

SNAP expenditures (2012)Predicted change in SNAP expenditures

(millions of dollars)

Per resident

(millions of

dollars)

Total

(millions of

dollars)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 state

expenditures

(linear regression)

New Jersey $825 $160 $1420 -$1611 -$1325 -$605

New Mexico $750 $324 $675 -$513 -$444 -$445New York $800 $287 $5616 -$3766 -$3148 -$2801

North Carolina $725 $252 $2454 -$2187 -$1925 -$1833

North Dakota $725 $128 $90 -$162 -$143 -$67

Ohio $795 $259 $2995 -$2013 -$1688 -$1539

Oklahoma $725 $248 $945 -$799 -$703 -$706

Oregon $910 $322 $1255 -$272 -$213 -$262

Pennsylvania $725 $218 $2779 -$2930 -$2579 -$2075

Rhode Island $800 $280 $294 -$173 -$144 -$147

South Carolina $725 $291 $1373 -$1337 -$1177 -$1025South Dakota $725 $198 $165 -$192 -$169 -$123

Tennessee $725 $324 $2091 -$1413 -$1243 -$1562

Texas $725 $230 $5997 -$6402 -$5634 -$4479

Utah $725 $141 $402 -$614 -$541 -$300

Vermont $873 $230 $144 -$66 -$53 -$43

Virginia $725 $173 $1413 -$2062 -$1815 -$1056

Washington $932 $244 $1682 -$350 -$270 -$267

West Virginia $725 $273 $508 -$451 -$397 -$379

Wisconsin $725 $204 $1166 -$1302 -$1146 -$871Wyoming $725 $95 $55 -$105 -$93 -$41

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4043

Appendix C | wwwamericanprogresso

FIGURE 2

Raising the minimum wage to $1010would cut taxpayer costs in every state

Predicted decreases in cost and enrollment

in SNAP in 50 states

$200+$51ndash$100

$101ndash$200

0ndash$25

$26ndash$50

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4143

38 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

Endnotes

1 US Department of Agriculture ldquoSupplementalNutrition Assistance Program (SNAP)rdquo available athttpwwwfnsusdagovsnapeligibility (last accessedFebruary 2014)

2 For this initial analysis we do not consider Harkin-Millerrsquos increase in subminimum wages for tippedworkers To do s o would increase the estimated SNAP

savings by an unknown amount

3 The Congressional Budget Office estimates thatworkers currently earning between $1010 and $1150per hour would see their wages rise under the Harkin-Miller proposal Congressional Budget O ffice ldquoTheEffects of a Minimum Wage Increase on Employmentand Family Incomerdquo (2014)

4 Marianne Page Joanne Spetz and Jane Millar ldquoDoesthe Minimum Wage Affect Welfare Caseloadsrdquo Journalof Policy Analysis and Management 24 (2) (2005)273ndash295

5 Marianne Bitler and Hilary Hoynes ldquoThe More ThingsChange the More They Stay the Same The SafetyNet Living Arrangements and Poverty in the GreatRecessionrdquo NBER Working Paper 194 49 2013

6 Sylvia Allegretto and others ldquoFast Food Poverty Wages The Public Cost of Low-Wage Jobs in the Fast-FoodIndustryrdquo (Berkeley California Center for LaborResearch and Education 2013) available at httplaborcenterberkeleyedupubliccostsfast_food_poverty_wage

7 David Neumark and William Wascher ldquoDoes a HigherMinimum Wage Enhance the Effectiveness of theEarned Income Tax Creditrdquo Industrial and LaborRelations Review 64 (4) (2011) 712ndash746

8 David Lee and Emmanuel Saez ldquoOptimal MinimumWage Policy in Competitive Labor Marketsrdquo Journal ofPublic Economics 96 (9) (2012) 739ndash749

9 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family IncomerdquoFebruary 2014

10 Jesse Rothstein ldquoIs the EITC as Good as an NITConditional Cash Transfers and Tax Incidencerdquo AmericanEconomic Journal Economic Policy 2 (1) (2010) 177ndash208

11 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family Incomerdquo

12 Dube Arindrajit 2013 rdquoMinimum Wagesand the Distribution of Family IncomerdquoUnpublished working paper Available at httpsdldropboxusercontentcomu15038936Dube_MinimumWagesFamilyIncomespdf

13 As Dube explains in ldquoThe poverty of Minimum WageFactsrdquo the simulation approach underestimate stemsfrom a number of unwarranted assumptions includingthe range of actual wage increases and the accuracy ofwage data in the Current Population Survey The causal

approach does not make these assumptions

14 Allegretto Sylvia and others 2013 ldquoCredible ResearchDesigns for Minimum Wage Studiesrdquo Working Paper148-13 University of California Berkeley Institutefor Research on Labor and Employment Available athttpwwwirleberkeleyeduworkingpapers148-13pdf

15 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family In comerdquo

16 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo (2012)

17 Ibid

18 Ibid

19 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo

20 US Department of Agriculture ldquoSupplementalNutrition Assistance Program (SNAP)rdquo

21 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo

22 Ibid

23 We deviate from the Census Bureaursquos definition ofa family unit which is ldquotwo people or more (on e ofwhom is the householder) related by bir th marriageor adoption and residing togetherrdquo US Bureau ofthe Census ldquoCurrent Population Survey DefinitionsFamilyrdquo available at httpwwwcensusgovcpsabout

cpsdefhtml (last accessed February 2014) We countas a family unit any individual residing on his or herown two or more persons residing together whodo not belong to a family in the March CPS sampleare constructed as one family in our analysis For thepurposes of food stamp allocations the consumptionresulting from this transfer is probably distributed tofamily members (rather than household members ora single individual within the household) Howeversingle individuals canmdashand domdashreceive SNAPbenefits Excluding them would fail to make theanalysis reflective of the population at large

24 Strictly the family level linear probability modelpredicts the percentage-point decrease in theprobability that an individual family will receive SNAPpayments When applied to a large number of familieshowever we are able to interpret the coefficient asa decrease in the mean of enrollmentmdashthat is a

decrease in the enrollment ratemdashby applying the lawof iterated expectations

25 We generate expenditure predictions from theenrollment modelsmdashand conversely generateenrollment predictions from the expenditure modelmdashby assuming that expenditures per enrolled familyremains the same before and after the minimum wagechange In practice this is likely to be a conservativeestimatemdashthat is to underestimate the decrease inSNAP activity Average SNAP benefits per family willalso decrease as many families that remain eligible forSNAP experience income gains

26 Wage and Hour Division ldquoMinimum Wage Laws inthe States ndash Januar y 1 2014rdquo available at httpwwwdolgovwhdminwageamericahtm (last accessedFebruary 2014)

27 See for example Sylvia Allegretto and others ldquoCredibleResearch Designs for Minimum Wage Studiesrdquo WorkingPaper 148-113 (Berkeley California Institute forResearch on Labor and Employment 2013) available athttpwwwirleberkeleyeduworkingpapers148-13pdf

28 We will report these results in a forthcoming workingpaper

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4243

Endnotes | wwwamericanprogresso

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4343

The Center for American Progress is a nonpartisan research and educational institute

dedicated to promoting a strong just and free America that ensures opportunity

for all We believe that Americans are bound together by a common commitment to

these values and we aspire to ensure that our national policies reflect these values

We work to find progressive and pragmatic solutions to significant domestic and

international problems and develop policy proposals that foster a government that

is ldquoof the people by the people and for the peoplerdquo

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3343

30 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

TABLE 8

Comparison of specifications SNAP enrollment

State-level linear regression

(1) (2) (3)

Log minimum wage -0019 -0035 -0031

(0009) (0012) (0012)

Unemployment rate (100) 0401 0370 0339

(0063) (0077) (0083)

Log median income -0081 -0073 -0061

(0011) (0013) (0013)

Employment-to-population ratio -0183 -0222 -0248

(0039) (0039) (0038)

N 1127 1127 1127

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the SNAP enrollment rate Allregressions include the share of households with children and the racial shares of the population

Source Annual data from the March Current Population Survey for 1990 to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3443

Appendix B | wwwamericanprogresso

TABLE 9

Comparison of specifications SNAP expenditures

State-level linear regression

(1) (2) (3)

Log minimum wage -0121 -0203 -0190

(0075) (0103) (0103)

Unemployment rate (100) 5292 5152 4313

(0464) (0576) (0628)

Log median income -0437 -0417 -0294

(008) (0086) (0078)

Employment-to-population ratio -0040 -0220 0244

(0261) (0260) (0240)

N 1127 1127 1127

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the natural log of state SNAPexpenditures per capita All regressions include the share of households with children and the racial shares ofthe state population

Source Annual data from Bureau of Economic Analysis national income and product accounts tables for 1990to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3543

32 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3643

Appendix C | wwwamericanprogresso

Appendix C Harkin-Miller

policy simulation results

StateMinimum wage

(2014)

SNAP enrollment (2012) Predicted change in SNAP enrollment

RateTotal

(persons)

Model 1 family

enrollment (linearprobability)

Model 2 state

enrollment (linearregression)

Model 3 s

expenditu(linear regre

Alabama $725 164 788682 -66922 -58897 -58906

Alaska $775 120 87436 -8104 -6887 -3288

Arizona $790 201 1319323 -67435 -56738 -64356

Arkansas $725 78 230489 -40977 -36063 -35248

California $800 206 7813680 -371131 -310222 -18223

Colorado $800 164 853155 -50684 -42365 -23926

Connecticut $870 91 326621 -22456 -17975 -13711

Delaware $725 186 170262 -12739 -11211 -10647

District of Columbia $825 133 84009 -5370 -4417 -3632

Florida $793 166 3208026 -195813 -164426 -13046

Georgia $725 160 1586336 -137741 -121224 -11004

Hawaii $725 96 133662 -19310 -16995 -14933

Idaho $725 92 147501 -22165 -19507 -15809

Illinois $825 95 1225084 -109088 -89742 -70955

Indiana $725 125 816233 -90818 -79928 -83985

Iowa $725 155 478011 -42716 -37594 -28556

Kansas $725 135 388269 -40082 -35275 -27461

Kentucky $725 130 568821 -60840 -53544 -52259

Louisiana $725 149 683832 -63929 -56263 -66083

Maine $750 77 101976 -16567 -14323 -15234

Maryland $725 144 846415 -81748 -71946 -38370

Massachusetts $800 130 864721 -64902 -54251 -42913

Michigan $740 146 1439141 -128801 -112140 -11022

Minnesota $725 133 713646 -74730 -65769 -37878

TABLE 10

SNAP enrollments Predicted changes by state under the Harkin-Miller bill

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3743

34 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

StateMinimum wage

(2014)

SNAP enrollment (2012) Predicted change in SNAP enrollment

RateTotal

(persons)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 state

expenditures

(linear regression)

Mississippi $725 129 386501 -41486 -36511 -46467

Missouri $750 172 1036182 -75131 -64952 -56244

Montana $790 132 132452 -10350 -8708 -5846

Nebraska $725 124 230683 -25773 -22683 -12189

Nevada $825 162 446035 -23349 -19209 -11894

New Hampshire $725 127 168404 -18359 -16157 -5735

New Jersey $825 160 1416666 -75175 -61843 -28236

New Mexico $750 149 310896 -25983 -22463 -22512

New York $800 192 3763553 -191193 -159815 -142182

North Carolina $725 174 1697193 -135417 -119179 -113503

North Dakota $725 87 61225 -9743 -8574 -4021

Ohio $795 143 1647345 -115869 -97169 -88580

Oklahoma $725 129 494053 -53006 -46650 -46854

Oregon $910 124 485326 -17036 -13328 -16398

Pennsylvania $725 161 2053643 -177315 -156052 -125586

Rhode Island $800 156 163730 -10258 -8574 -8698

South Carolina $725 94 445277 -65614 -57746 -50304

South Dakota $725 208 173749 -11586 -10197 -7458

Tennessee $725 142 914903 -89667 -78915 -99134

Texas $725 110 2863779 -362018 -318607 -253285

Utah $725 88 251107 -39658 -34902 -19390

Vermont $873 156 97792 -3823 -3055 -2475

Virginia $725 101 829771 -113723 -100086 -58212

Washington $932 72 496934 -23221 -17947 -17756

West Virginia $725 58 107875 -25792 -22699 -21665

Wisconsin $725 75 427822 -79521 -69986 -53210

Wyoming $725 164 94590 -8010 -7050 -3104

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3843

Appendix C | wwwamericanprogresso

StateMinimum wage

(2014)

SNAP expenditures (2012)Predicted change in SNAP expenditures

(millions of dollars)

Per resident

(millions of

dollars)

Total

(millions of

dollars)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 s

expenditu

(linear regre

Alabama $725 $290 $1397 -$1185 -$1043 -$1043

Alaska $775 $253 $185 -$262 -$223 -$106

Arizona $790 $257 $1686 -$935 -$787 -$892

Arkansas $725 $248 $733 -$636 -$560 -$547

California $800 $189 $7164 -$7276 -$6082 -$3573

Colorado $800 $157 $816 -$862 -$721 -$407

Connecticut $870 $191 $686 -$343 -$275 -$210

Delaware $725 $250 $229 -$205 -$180 -$171

District of Columbia $825 $366 $232 -$146 -$120 -$99

Florida $793 $294 $5676 -$4429 -$3719 -$2951

Georgia $725 $317 $3140 -$2936 -$2584 -$2346

Hawaii $725 $335 $465 -$449 -$395 -$347

Idaho $725 $225 $359 -$376 -$331 -$268

Illinois $825 $249 $3200 -$2096 -$1725 -$1364

Indiana $725 $220 $1439 -$1162 -$1023 -$1075

Iowa $725 $192 $589 -$658 -$579 -$440

Kansas $725 $159 $460 -$502 -$441 -$344

Kentucky $725 $298 $1303 -$1133 -$997 -$973

Louisiana $725 $315 $1450 -$1047 -$922 -$1083

Maine $750 $281 $373 -$267 -$231 -$246

Maryland $725 $188 $1109 -$1765 -$1553 -$828

Massachusetts $800 $206 $1366 -$1030 -$861 -$681

Michigan $740 $300 $2963 -$2400 -$2090 -$2054

Minnesota $725 $140 $755 -$1113 -$980 -$564

Mississippi $725 $326 $973 -$649 -$571 -$726

Missouri $750 $241 $1452 -$1278 -$1104 -$956

Montana $790 $190 $191 -$179 -$151 -$101

Nebraska $725 $140 $259 -$409 -$360 -$194

Nevada $825 $191 $527 -$441 -$363 -$225

New Hampshire $725 $126 $167 -$399 -$351 -$125

TABLE 11

SNAP expenditures Predicted changes by state under the Harkin-Miller bill

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3943

36 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

StateMinimum wage

(2014)

SNAP expenditures (2012)Predicted change in SNAP expenditures

(millions of dollars)

Per resident

(millions of

dollars)

Total

(millions of

dollars)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 state

expenditures

(linear regression)

New Jersey $825 $160 $1420 -$1611 -$1325 -$605

New Mexico $750 $324 $675 -$513 -$444 -$445New York $800 $287 $5616 -$3766 -$3148 -$2801

North Carolina $725 $252 $2454 -$2187 -$1925 -$1833

North Dakota $725 $128 $90 -$162 -$143 -$67

Ohio $795 $259 $2995 -$2013 -$1688 -$1539

Oklahoma $725 $248 $945 -$799 -$703 -$706

Oregon $910 $322 $1255 -$272 -$213 -$262

Pennsylvania $725 $218 $2779 -$2930 -$2579 -$2075

Rhode Island $800 $280 $294 -$173 -$144 -$147

South Carolina $725 $291 $1373 -$1337 -$1177 -$1025South Dakota $725 $198 $165 -$192 -$169 -$123

Tennessee $725 $324 $2091 -$1413 -$1243 -$1562

Texas $725 $230 $5997 -$6402 -$5634 -$4479

Utah $725 $141 $402 -$614 -$541 -$300

Vermont $873 $230 $144 -$66 -$53 -$43

Virginia $725 $173 $1413 -$2062 -$1815 -$1056

Washington $932 $244 $1682 -$350 -$270 -$267

West Virginia $725 $273 $508 -$451 -$397 -$379

Wisconsin $725 $204 $1166 -$1302 -$1146 -$871Wyoming $725 $95 $55 -$105 -$93 -$41

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4043

Appendix C | wwwamericanprogresso

FIGURE 2

Raising the minimum wage to $1010would cut taxpayer costs in every state

Predicted decreases in cost and enrollment

in SNAP in 50 states

$200+$51ndash$100

$101ndash$200

0ndash$25

$26ndash$50

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4143

38 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

Endnotes

1 US Department of Agriculture ldquoSupplementalNutrition Assistance Program (SNAP)rdquo available athttpwwwfnsusdagovsnapeligibility (last accessedFebruary 2014)

2 For this initial analysis we do not consider Harkin-Millerrsquos increase in subminimum wages for tippedworkers To do s o would increase the estimated SNAP

savings by an unknown amount

3 The Congressional Budget Office estimates thatworkers currently earning between $1010 and $1150per hour would see their wages rise under the Harkin-Miller proposal Congressional Budget O ffice ldquoTheEffects of a Minimum Wage Increase on Employmentand Family Incomerdquo (2014)

4 Marianne Page Joanne Spetz and Jane Millar ldquoDoesthe Minimum Wage Affect Welfare Caseloadsrdquo Journalof Policy Analysis and Management 24 (2) (2005)273ndash295

5 Marianne Bitler and Hilary Hoynes ldquoThe More ThingsChange the More They Stay the Same The SafetyNet Living Arrangements and Poverty in the GreatRecessionrdquo NBER Working Paper 194 49 2013

6 Sylvia Allegretto and others ldquoFast Food Poverty Wages The Public Cost of Low-Wage Jobs in the Fast-FoodIndustryrdquo (Berkeley California Center for LaborResearch and Education 2013) available at httplaborcenterberkeleyedupubliccostsfast_food_poverty_wage

7 David Neumark and William Wascher ldquoDoes a HigherMinimum Wage Enhance the Effectiveness of theEarned Income Tax Creditrdquo Industrial and LaborRelations Review 64 (4) (2011) 712ndash746

8 David Lee and Emmanuel Saez ldquoOptimal MinimumWage Policy in Competitive Labor Marketsrdquo Journal ofPublic Economics 96 (9) (2012) 739ndash749

9 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family IncomerdquoFebruary 2014

10 Jesse Rothstein ldquoIs the EITC as Good as an NITConditional Cash Transfers and Tax Incidencerdquo AmericanEconomic Journal Economic Policy 2 (1) (2010) 177ndash208

11 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family Incomerdquo

12 Dube Arindrajit 2013 rdquoMinimum Wagesand the Distribution of Family IncomerdquoUnpublished working paper Available at httpsdldropboxusercontentcomu15038936Dube_MinimumWagesFamilyIncomespdf

13 As Dube explains in ldquoThe poverty of Minimum WageFactsrdquo the simulation approach underestimate stemsfrom a number of unwarranted assumptions includingthe range of actual wage increases and the accuracy ofwage data in the Current Population Survey The causal

approach does not make these assumptions

14 Allegretto Sylvia and others 2013 ldquoCredible ResearchDesigns for Minimum Wage Studiesrdquo Working Paper148-13 University of California Berkeley Institutefor Research on Labor and Employment Available athttpwwwirleberkeleyeduworkingpapers148-13pdf

15 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family In comerdquo

16 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo (2012)

17 Ibid

18 Ibid

19 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo

20 US Department of Agriculture ldquoSupplementalNutrition Assistance Program (SNAP)rdquo

21 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo

22 Ibid

23 We deviate from the Census Bureaursquos definition ofa family unit which is ldquotwo people or more (on e ofwhom is the householder) related by bir th marriageor adoption and residing togetherrdquo US Bureau ofthe Census ldquoCurrent Population Survey DefinitionsFamilyrdquo available at httpwwwcensusgovcpsabout

cpsdefhtml (last accessed February 2014) We countas a family unit any individual residing on his or herown two or more persons residing together whodo not belong to a family in the March CPS sampleare constructed as one family in our analysis For thepurposes of food stamp allocations the consumptionresulting from this transfer is probably distributed tofamily members (rather than household members ora single individual within the household) Howeversingle individuals canmdashand domdashreceive SNAPbenefits Excluding them would fail to make theanalysis reflective of the population at large

24 Strictly the family level linear probability modelpredicts the percentage-point decrease in theprobability that an individual family will receive SNAPpayments When applied to a large number of familieshowever we are able to interpret the coefficient asa decrease in the mean of enrollmentmdashthat is a

decrease in the enrollment ratemdashby applying the lawof iterated expectations

25 We generate expenditure predictions from theenrollment modelsmdashand conversely generateenrollment predictions from the expenditure modelmdashby assuming that expenditures per enrolled familyremains the same before and after the minimum wagechange In practice this is likely to be a conservativeestimatemdashthat is to underestimate the decrease inSNAP activity Average SNAP benefits per family willalso decrease as many families that remain eligible forSNAP experience income gains

26 Wage and Hour Division ldquoMinimum Wage Laws inthe States ndash Januar y 1 2014rdquo available at httpwwwdolgovwhdminwageamericahtm (last accessedFebruary 2014)

27 See for example Sylvia Allegretto and others ldquoCredibleResearch Designs for Minimum Wage Studiesrdquo WorkingPaper 148-113 (Berkeley California Institute forResearch on Labor and Employment 2013) available athttpwwwirleberkeleyeduworkingpapers148-13pdf

28 We will report these results in a forthcoming workingpaper

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4243

Endnotes | wwwamericanprogresso

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4343

The Center for American Progress is a nonpartisan research and educational institute

dedicated to promoting a strong just and free America that ensures opportunity

for all We believe that Americans are bound together by a common commitment to

these values and we aspire to ensure that our national policies reflect these values

We work to find progressive and pragmatic solutions to significant domestic and

international problems and develop policy proposals that foster a government that

is ldquoof the people by the people and for the peoplerdquo

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3443

Appendix B | wwwamericanprogresso

TABLE 9

Comparison of specifications SNAP expenditures

State-level linear regression

(1) (2) (3)

Log minimum wage -0121 -0203 -0190

(0075) (0103) (0103)

Unemployment rate (100) 5292 5152 4313

(0464) (0576) (0628)

Log median income -0437 -0417 -0294

(008) (0086) (0078)

Employment-to-population ratio -0040 -0220 0244

(0261) (0260) (0240)

N 1127 1127 1127

State fixed effects Y Y Y

Year fixed effects Y

Division x-year fixed effects Y Y

State-specific linear trends Y

plt01 plt005 plt001

Note Robust standard errors are in parentheses The outcome variable is the natural log of state SNAPexpenditures per capita All regressions include the share of households with children and the racial shares ofthe state population

Source Annual data from Bureau of Economic Analysis national income and product accounts tables for 1990to 2012

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3543

32 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3643

Appendix C | wwwamericanprogresso

Appendix C Harkin-Miller

policy simulation results

StateMinimum wage

(2014)

SNAP enrollment (2012) Predicted change in SNAP enrollment

RateTotal

(persons)

Model 1 family

enrollment (linearprobability)

Model 2 state

enrollment (linearregression)

Model 3 s

expenditu(linear regre

Alabama $725 164 788682 -66922 -58897 -58906

Alaska $775 120 87436 -8104 -6887 -3288

Arizona $790 201 1319323 -67435 -56738 -64356

Arkansas $725 78 230489 -40977 -36063 -35248

California $800 206 7813680 -371131 -310222 -18223

Colorado $800 164 853155 -50684 -42365 -23926

Connecticut $870 91 326621 -22456 -17975 -13711

Delaware $725 186 170262 -12739 -11211 -10647

District of Columbia $825 133 84009 -5370 -4417 -3632

Florida $793 166 3208026 -195813 -164426 -13046

Georgia $725 160 1586336 -137741 -121224 -11004

Hawaii $725 96 133662 -19310 -16995 -14933

Idaho $725 92 147501 -22165 -19507 -15809

Illinois $825 95 1225084 -109088 -89742 -70955

Indiana $725 125 816233 -90818 -79928 -83985

Iowa $725 155 478011 -42716 -37594 -28556

Kansas $725 135 388269 -40082 -35275 -27461

Kentucky $725 130 568821 -60840 -53544 -52259

Louisiana $725 149 683832 -63929 -56263 -66083

Maine $750 77 101976 -16567 -14323 -15234

Maryland $725 144 846415 -81748 -71946 -38370

Massachusetts $800 130 864721 -64902 -54251 -42913

Michigan $740 146 1439141 -128801 -112140 -11022

Minnesota $725 133 713646 -74730 -65769 -37878

TABLE 10

SNAP enrollments Predicted changes by state under the Harkin-Miller bill

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3743

34 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

StateMinimum wage

(2014)

SNAP enrollment (2012) Predicted change in SNAP enrollment

RateTotal

(persons)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 state

expenditures

(linear regression)

Mississippi $725 129 386501 -41486 -36511 -46467

Missouri $750 172 1036182 -75131 -64952 -56244

Montana $790 132 132452 -10350 -8708 -5846

Nebraska $725 124 230683 -25773 -22683 -12189

Nevada $825 162 446035 -23349 -19209 -11894

New Hampshire $725 127 168404 -18359 -16157 -5735

New Jersey $825 160 1416666 -75175 -61843 -28236

New Mexico $750 149 310896 -25983 -22463 -22512

New York $800 192 3763553 -191193 -159815 -142182

North Carolina $725 174 1697193 -135417 -119179 -113503

North Dakota $725 87 61225 -9743 -8574 -4021

Ohio $795 143 1647345 -115869 -97169 -88580

Oklahoma $725 129 494053 -53006 -46650 -46854

Oregon $910 124 485326 -17036 -13328 -16398

Pennsylvania $725 161 2053643 -177315 -156052 -125586

Rhode Island $800 156 163730 -10258 -8574 -8698

South Carolina $725 94 445277 -65614 -57746 -50304

South Dakota $725 208 173749 -11586 -10197 -7458

Tennessee $725 142 914903 -89667 -78915 -99134

Texas $725 110 2863779 -362018 -318607 -253285

Utah $725 88 251107 -39658 -34902 -19390

Vermont $873 156 97792 -3823 -3055 -2475

Virginia $725 101 829771 -113723 -100086 -58212

Washington $932 72 496934 -23221 -17947 -17756

West Virginia $725 58 107875 -25792 -22699 -21665

Wisconsin $725 75 427822 -79521 -69986 -53210

Wyoming $725 164 94590 -8010 -7050 -3104

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3843

Appendix C | wwwamericanprogresso

StateMinimum wage

(2014)

SNAP expenditures (2012)Predicted change in SNAP expenditures

(millions of dollars)

Per resident

(millions of

dollars)

Total

(millions of

dollars)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 s

expenditu

(linear regre

Alabama $725 $290 $1397 -$1185 -$1043 -$1043

Alaska $775 $253 $185 -$262 -$223 -$106

Arizona $790 $257 $1686 -$935 -$787 -$892

Arkansas $725 $248 $733 -$636 -$560 -$547

California $800 $189 $7164 -$7276 -$6082 -$3573

Colorado $800 $157 $816 -$862 -$721 -$407

Connecticut $870 $191 $686 -$343 -$275 -$210

Delaware $725 $250 $229 -$205 -$180 -$171

District of Columbia $825 $366 $232 -$146 -$120 -$99

Florida $793 $294 $5676 -$4429 -$3719 -$2951

Georgia $725 $317 $3140 -$2936 -$2584 -$2346

Hawaii $725 $335 $465 -$449 -$395 -$347

Idaho $725 $225 $359 -$376 -$331 -$268

Illinois $825 $249 $3200 -$2096 -$1725 -$1364

Indiana $725 $220 $1439 -$1162 -$1023 -$1075

Iowa $725 $192 $589 -$658 -$579 -$440

Kansas $725 $159 $460 -$502 -$441 -$344

Kentucky $725 $298 $1303 -$1133 -$997 -$973

Louisiana $725 $315 $1450 -$1047 -$922 -$1083

Maine $750 $281 $373 -$267 -$231 -$246

Maryland $725 $188 $1109 -$1765 -$1553 -$828

Massachusetts $800 $206 $1366 -$1030 -$861 -$681

Michigan $740 $300 $2963 -$2400 -$2090 -$2054

Minnesota $725 $140 $755 -$1113 -$980 -$564

Mississippi $725 $326 $973 -$649 -$571 -$726

Missouri $750 $241 $1452 -$1278 -$1104 -$956

Montana $790 $190 $191 -$179 -$151 -$101

Nebraska $725 $140 $259 -$409 -$360 -$194

Nevada $825 $191 $527 -$441 -$363 -$225

New Hampshire $725 $126 $167 -$399 -$351 -$125

TABLE 11

SNAP expenditures Predicted changes by state under the Harkin-Miller bill

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3943

36 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

StateMinimum wage

(2014)

SNAP expenditures (2012)Predicted change in SNAP expenditures

(millions of dollars)

Per resident

(millions of

dollars)

Total

(millions of

dollars)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 state

expenditures

(linear regression)

New Jersey $825 $160 $1420 -$1611 -$1325 -$605

New Mexico $750 $324 $675 -$513 -$444 -$445New York $800 $287 $5616 -$3766 -$3148 -$2801

North Carolina $725 $252 $2454 -$2187 -$1925 -$1833

North Dakota $725 $128 $90 -$162 -$143 -$67

Ohio $795 $259 $2995 -$2013 -$1688 -$1539

Oklahoma $725 $248 $945 -$799 -$703 -$706

Oregon $910 $322 $1255 -$272 -$213 -$262

Pennsylvania $725 $218 $2779 -$2930 -$2579 -$2075

Rhode Island $800 $280 $294 -$173 -$144 -$147

South Carolina $725 $291 $1373 -$1337 -$1177 -$1025South Dakota $725 $198 $165 -$192 -$169 -$123

Tennessee $725 $324 $2091 -$1413 -$1243 -$1562

Texas $725 $230 $5997 -$6402 -$5634 -$4479

Utah $725 $141 $402 -$614 -$541 -$300

Vermont $873 $230 $144 -$66 -$53 -$43

Virginia $725 $173 $1413 -$2062 -$1815 -$1056

Washington $932 $244 $1682 -$350 -$270 -$267

West Virginia $725 $273 $508 -$451 -$397 -$379

Wisconsin $725 $204 $1166 -$1302 -$1146 -$871Wyoming $725 $95 $55 -$105 -$93 -$41

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4043

Appendix C | wwwamericanprogresso

FIGURE 2

Raising the minimum wage to $1010would cut taxpayer costs in every state

Predicted decreases in cost and enrollment

in SNAP in 50 states

$200+$51ndash$100

$101ndash$200

0ndash$25

$26ndash$50

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4143

38 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

Endnotes

1 US Department of Agriculture ldquoSupplementalNutrition Assistance Program (SNAP)rdquo available athttpwwwfnsusdagovsnapeligibility (last accessedFebruary 2014)

2 For this initial analysis we do not consider Harkin-Millerrsquos increase in subminimum wages for tippedworkers To do s o would increase the estimated SNAP

savings by an unknown amount

3 The Congressional Budget Office estimates thatworkers currently earning between $1010 and $1150per hour would see their wages rise under the Harkin-Miller proposal Congressional Budget O ffice ldquoTheEffects of a Minimum Wage Increase on Employmentand Family Incomerdquo (2014)

4 Marianne Page Joanne Spetz and Jane Millar ldquoDoesthe Minimum Wage Affect Welfare Caseloadsrdquo Journalof Policy Analysis and Management 24 (2) (2005)273ndash295

5 Marianne Bitler and Hilary Hoynes ldquoThe More ThingsChange the More They Stay the Same The SafetyNet Living Arrangements and Poverty in the GreatRecessionrdquo NBER Working Paper 194 49 2013

6 Sylvia Allegretto and others ldquoFast Food Poverty Wages The Public Cost of Low-Wage Jobs in the Fast-FoodIndustryrdquo (Berkeley California Center for LaborResearch and Education 2013) available at httplaborcenterberkeleyedupubliccostsfast_food_poverty_wage

7 David Neumark and William Wascher ldquoDoes a HigherMinimum Wage Enhance the Effectiveness of theEarned Income Tax Creditrdquo Industrial and LaborRelations Review 64 (4) (2011) 712ndash746

8 David Lee and Emmanuel Saez ldquoOptimal MinimumWage Policy in Competitive Labor Marketsrdquo Journal ofPublic Economics 96 (9) (2012) 739ndash749

9 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family IncomerdquoFebruary 2014

10 Jesse Rothstein ldquoIs the EITC as Good as an NITConditional Cash Transfers and Tax Incidencerdquo AmericanEconomic Journal Economic Policy 2 (1) (2010) 177ndash208

11 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family Incomerdquo

12 Dube Arindrajit 2013 rdquoMinimum Wagesand the Distribution of Family IncomerdquoUnpublished working paper Available at httpsdldropboxusercontentcomu15038936Dube_MinimumWagesFamilyIncomespdf

13 As Dube explains in ldquoThe poverty of Minimum WageFactsrdquo the simulation approach underestimate stemsfrom a number of unwarranted assumptions includingthe range of actual wage increases and the accuracy ofwage data in the Current Population Survey The causal

approach does not make these assumptions

14 Allegretto Sylvia and others 2013 ldquoCredible ResearchDesigns for Minimum Wage Studiesrdquo Working Paper148-13 University of California Berkeley Institutefor Research on Labor and Employment Available athttpwwwirleberkeleyeduworkingpapers148-13pdf

15 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family In comerdquo

16 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo (2012)

17 Ibid

18 Ibid

19 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo

20 US Department of Agriculture ldquoSupplementalNutrition Assistance Program (SNAP)rdquo

21 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo

22 Ibid

23 We deviate from the Census Bureaursquos definition ofa family unit which is ldquotwo people or more (on e ofwhom is the householder) related by bir th marriageor adoption and residing togetherrdquo US Bureau ofthe Census ldquoCurrent Population Survey DefinitionsFamilyrdquo available at httpwwwcensusgovcpsabout

cpsdefhtml (last accessed February 2014) We countas a family unit any individual residing on his or herown two or more persons residing together whodo not belong to a family in the March CPS sampleare constructed as one family in our analysis For thepurposes of food stamp allocations the consumptionresulting from this transfer is probably distributed tofamily members (rather than household members ora single individual within the household) Howeversingle individuals canmdashand domdashreceive SNAPbenefits Excluding them would fail to make theanalysis reflective of the population at large

24 Strictly the family level linear probability modelpredicts the percentage-point decrease in theprobability that an individual family will receive SNAPpayments When applied to a large number of familieshowever we are able to interpret the coefficient asa decrease in the mean of enrollmentmdashthat is a

decrease in the enrollment ratemdashby applying the lawof iterated expectations

25 We generate expenditure predictions from theenrollment modelsmdashand conversely generateenrollment predictions from the expenditure modelmdashby assuming that expenditures per enrolled familyremains the same before and after the minimum wagechange In practice this is likely to be a conservativeestimatemdashthat is to underestimate the decrease inSNAP activity Average SNAP benefits per family willalso decrease as many families that remain eligible forSNAP experience income gains

26 Wage and Hour Division ldquoMinimum Wage Laws inthe States ndash Januar y 1 2014rdquo available at httpwwwdolgovwhdminwageamericahtm (last accessedFebruary 2014)

27 See for example Sylvia Allegretto and others ldquoCredibleResearch Designs for Minimum Wage Studiesrdquo WorkingPaper 148-113 (Berkeley California Institute forResearch on Labor and Employment 2013) available athttpwwwirleberkeleyeduworkingpapers148-13pdf

28 We will report these results in a forthcoming workingpaper

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4243

Endnotes | wwwamericanprogresso

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4343

The Center for American Progress is a nonpartisan research and educational institute

dedicated to promoting a strong just and free America that ensures opportunity

for all We believe that Americans are bound together by a common commitment to

these values and we aspire to ensure that our national policies reflect these values

We work to find progressive and pragmatic solutions to significant domestic and

international problems and develop policy proposals that foster a government that

is ldquoof the people by the people and for the peoplerdquo

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3543

32 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3643

Appendix C | wwwamericanprogresso

Appendix C Harkin-Miller

policy simulation results

StateMinimum wage

(2014)

SNAP enrollment (2012) Predicted change in SNAP enrollment

RateTotal

(persons)

Model 1 family

enrollment (linearprobability)

Model 2 state

enrollment (linearregression)

Model 3 s

expenditu(linear regre

Alabama $725 164 788682 -66922 -58897 -58906

Alaska $775 120 87436 -8104 -6887 -3288

Arizona $790 201 1319323 -67435 -56738 -64356

Arkansas $725 78 230489 -40977 -36063 -35248

California $800 206 7813680 -371131 -310222 -18223

Colorado $800 164 853155 -50684 -42365 -23926

Connecticut $870 91 326621 -22456 -17975 -13711

Delaware $725 186 170262 -12739 -11211 -10647

District of Columbia $825 133 84009 -5370 -4417 -3632

Florida $793 166 3208026 -195813 -164426 -13046

Georgia $725 160 1586336 -137741 -121224 -11004

Hawaii $725 96 133662 -19310 -16995 -14933

Idaho $725 92 147501 -22165 -19507 -15809

Illinois $825 95 1225084 -109088 -89742 -70955

Indiana $725 125 816233 -90818 -79928 -83985

Iowa $725 155 478011 -42716 -37594 -28556

Kansas $725 135 388269 -40082 -35275 -27461

Kentucky $725 130 568821 -60840 -53544 -52259

Louisiana $725 149 683832 -63929 -56263 -66083

Maine $750 77 101976 -16567 -14323 -15234

Maryland $725 144 846415 -81748 -71946 -38370

Massachusetts $800 130 864721 -64902 -54251 -42913

Michigan $740 146 1439141 -128801 -112140 -11022

Minnesota $725 133 713646 -74730 -65769 -37878

TABLE 10

SNAP enrollments Predicted changes by state under the Harkin-Miller bill

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3743

34 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

StateMinimum wage

(2014)

SNAP enrollment (2012) Predicted change in SNAP enrollment

RateTotal

(persons)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 state

expenditures

(linear regression)

Mississippi $725 129 386501 -41486 -36511 -46467

Missouri $750 172 1036182 -75131 -64952 -56244

Montana $790 132 132452 -10350 -8708 -5846

Nebraska $725 124 230683 -25773 -22683 -12189

Nevada $825 162 446035 -23349 -19209 -11894

New Hampshire $725 127 168404 -18359 -16157 -5735

New Jersey $825 160 1416666 -75175 -61843 -28236

New Mexico $750 149 310896 -25983 -22463 -22512

New York $800 192 3763553 -191193 -159815 -142182

North Carolina $725 174 1697193 -135417 -119179 -113503

North Dakota $725 87 61225 -9743 -8574 -4021

Ohio $795 143 1647345 -115869 -97169 -88580

Oklahoma $725 129 494053 -53006 -46650 -46854

Oregon $910 124 485326 -17036 -13328 -16398

Pennsylvania $725 161 2053643 -177315 -156052 -125586

Rhode Island $800 156 163730 -10258 -8574 -8698

South Carolina $725 94 445277 -65614 -57746 -50304

South Dakota $725 208 173749 -11586 -10197 -7458

Tennessee $725 142 914903 -89667 -78915 -99134

Texas $725 110 2863779 -362018 -318607 -253285

Utah $725 88 251107 -39658 -34902 -19390

Vermont $873 156 97792 -3823 -3055 -2475

Virginia $725 101 829771 -113723 -100086 -58212

Washington $932 72 496934 -23221 -17947 -17756

West Virginia $725 58 107875 -25792 -22699 -21665

Wisconsin $725 75 427822 -79521 -69986 -53210

Wyoming $725 164 94590 -8010 -7050 -3104

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3843

Appendix C | wwwamericanprogresso

StateMinimum wage

(2014)

SNAP expenditures (2012)Predicted change in SNAP expenditures

(millions of dollars)

Per resident

(millions of

dollars)

Total

(millions of

dollars)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 s

expenditu

(linear regre

Alabama $725 $290 $1397 -$1185 -$1043 -$1043

Alaska $775 $253 $185 -$262 -$223 -$106

Arizona $790 $257 $1686 -$935 -$787 -$892

Arkansas $725 $248 $733 -$636 -$560 -$547

California $800 $189 $7164 -$7276 -$6082 -$3573

Colorado $800 $157 $816 -$862 -$721 -$407

Connecticut $870 $191 $686 -$343 -$275 -$210

Delaware $725 $250 $229 -$205 -$180 -$171

District of Columbia $825 $366 $232 -$146 -$120 -$99

Florida $793 $294 $5676 -$4429 -$3719 -$2951

Georgia $725 $317 $3140 -$2936 -$2584 -$2346

Hawaii $725 $335 $465 -$449 -$395 -$347

Idaho $725 $225 $359 -$376 -$331 -$268

Illinois $825 $249 $3200 -$2096 -$1725 -$1364

Indiana $725 $220 $1439 -$1162 -$1023 -$1075

Iowa $725 $192 $589 -$658 -$579 -$440

Kansas $725 $159 $460 -$502 -$441 -$344

Kentucky $725 $298 $1303 -$1133 -$997 -$973

Louisiana $725 $315 $1450 -$1047 -$922 -$1083

Maine $750 $281 $373 -$267 -$231 -$246

Maryland $725 $188 $1109 -$1765 -$1553 -$828

Massachusetts $800 $206 $1366 -$1030 -$861 -$681

Michigan $740 $300 $2963 -$2400 -$2090 -$2054

Minnesota $725 $140 $755 -$1113 -$980 -$564

Mississippi $725 $326 $973 -$649 -$571 -$726

Missouri $750 $241 $1452 -$1278 -$1104 -$956

Montana $790 $190 $191 -$179 -$151 -$101

Nebraska $725 $140 $259 -$409 -$360 -$194

Nevada $825 $191 $527 -$441 -$363 -$225

New Hampshire $725 $126 $167 -$399 -$351 -$125

TABLE 11

SNAP expenditures Predicted changes by state under the Harkin-Miller bill

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3943

36 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

StateMinimum wage

(2014)

SNAP expenditures (2012)Predicted change in SNAP expenditures

(millions of dollars)

Per resident

(millions of

dollars)

Total

(millions of

dollars)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 state

expenditures

(linear regression)

New Jersey $825 $160 $1420 -$1611 -$1325 -$605

New Mexico $750 $324 $675 -$513 -$444 -$445New York $800 $287 $5616 -$3766 -$3148 -$2801

North Carolina $725 $252 $2454 -$2187 -$1925 -$1833

North Dakota $725 $128 $90 -$162 -$143 -$67

Ohio $795 $259 $2995 -$2013 -$1688 -$1539

Oklahoma $725 $248 $945 -$799 -$703 -$706

Oregon $910 $322 $1255 -$272 -$213 -$262

Pennsylvania $725 $218 $2779 -$2930 -$2579 -$2075

Rhode Island $800 $280 $294 -$173 -$144 -$147

South Carolina $725 $291 $1373 -$1337 -$1177 -$1025South Dakota $725 $198 $165 -$192 -$169 -$123

Tennessee $725 $324 $2091 -$1413 -$1243 -$1562

Texas $725 $230 $5997 -$6402 -$5634 -$4479

Utah $725 $141 $402 -$614 -$541 -$300

Vermont $873 $230 $144 -$66 -$53 -$43

Virginia $725 $173 $1413 -$2062 -$1815 -$1056

Washington $932 $244 $1682 -$350 -$270 -$267

West Virginia $725 $273 $508 -$451 -$397 -$379

Wisconsin $725 $204 $1166 -$1302 -$1146 -$871Wyoming $725 $95 $55 -$105 -$93 -$41

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4043

Appendix C | wwwamericanprogresso

FIGURE 2

Raising the minimum wage to $1010would cut taxpayer costs in every state

Predicted decreases in cost and enrollment

in SNAP in 50 states

$200+$51ndash$100

$101ndash$200

0ndash$25

$26ndash$50

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4143

38 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

Endnotes

1 US Department of Agriculture ldquoSupplementalNutrition Assistance Program (SNAP)rdquo available athttpwwwfnsusdagovsnapeligibility (last accessedFebruary 2014)

2 For this initial analysis we do not consider Harkin-Millerrsquos increase in subminimum wages for tippedworkers To do s o would increase the estimated SNAP

savings by an unknown amount

3 The Congressional Budget Office estimates thatworkers currently earning between $1010 and $1150per hour would see their wages rise under the Harkin-Miller proposal Congressional Budget O ffice ldquoTheEffects of a Minimum Wage Increase on Employmentand Family Incomerdquo (2014)

4 Marianne Page Joanne Spetz and Jane Millar ldquoDoesthe Minimum Wage Affect Welfare Caseloadsrdquo Journalof Policy Analysis and Management 24 (2) (2005)273ndash295

5 Marianne Bitler and Hilary Hoynes ldquoThe More ThingsChange the More They Stay the Same The SafetyNet Living Arrangements and Poverty in the GreatRecessionrdquo NBER Working Paper 194 49 2013

6 Sylvia Allegretto and others ldquoFast Food Poverty Wages The Public Cost of Low-Wage Jobs in the Fast-FoodIndustryrdquo (Berkeley California Center for LaborResearch and Education 2013) available at httplaborcenterberkeleyedupubliccostsfast_food_poverty_wage

7 David Neumark and William Wascher ldquoDoes a HigherMinimum Wage Enhance the Effectiveness of theEarned Income Tax Creditrdquo Industrial and LaborRelations Review 64 (4) (2011) 712ndash746

8 David Lee and Emmanuel Saez ldquoOptimal MinimumWage Policy in Competitive Labor Marketsrdquo Journal ofPublic Economics 96 (9) (2012) 739ndash749

9 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family IncomerdquoFebruary 2014

10 Jesse Rothstein ldquoIs the EITC as Good as an NITConditional Cash Transfers and Tax Incidencerdquo AmericanEconomic Journal Economic Policy 2 (1) (2010) 177ndash208

11 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family Incomerdquo

12 Dube Arindrajit 2013 rdquoMinimum Wagesand the Distribution of Family IncomerdquoUnpublished working paper Available at httpsdldropboxusercontentcomu15038936Dube_MinimumWagesFamilyIncomespdf

13 As Dube explains in ldquoThe poverty of Minimum WageFactsrdquo the simulation approach underestimate stemsfrom a number of unwarranted assumptions includingthe range of actual wage increases and the accuracy ofwage data in the Current Population Survey The causal

approach does not make these assumptions

14 Allegretto Sylvia and others 2013 ldquoCredible ResearchDesigns for Minimum Wage Studiesrdquo Working Paper148-13 University of California Berkeley Institutefor Research on Labor and Employment Available athttpwwwirleberkeleyeduworkingpapers148-13pdf

15 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family In comerdquo

16 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo (2012)

17 Ibid

18 Ibid

19 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo

20 US Department of Agriculture ldquoSupplementalNutrition Assistance Program (SNAP)rdquo

21 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo

22 Ibid

23 We deviate from the Census Bureaursquos definition ofa family unit which is ldquotwo people or more (on e ofwhom is the householder) related by bir th marriageor adoption and residing togetherrdquo US Bureau ofthe Census ldquoCurrent Population Survey DefinitionsFamilyrdquo available at httpwwwcensusgovcpsabout

cpsdefhtml (last accessed February 2014) We countas a family unit any individual residing on his or herown two or more persons residing together whodo not belong to a family in the March CPS sampleare constructed as one family in our analysis For thepurposes of food stamp allocations the consumptionresulting from this transfer is probably distributed tofamily members (rather than household members ora single individual within the household) Howeversingle individuals canmdashand domdashreceive SNAPbenefits Excluding them would fail to make theanalysis reflective of the population at large

24 Strictly the family level linear probability modelpredicts the percentage-point decrease in theprobability that an individual family will receive SNAPpayments When applied to a large number of familieshowever we are able to interpret the coefficient asa decrease in the mean of enrollmentmdashthat is a

decrease in the enrollment ratemdashby applying the lawof iterated expectations

25 We generate expenditure predictions from theenrollment modelsmdashand conversely generateenrollment predictions from the expenditure modelmdashby assuming that expenditures per enrolled familyremains the same before and after the minimum wagechange In practice this is likely to be a conservativeestimatemdashthat is to underestimate the decrease inSNAP activity Average SNAP benefits per family willalso decrease as many families that remain eligible forSNAP experience income gains

26 Wage and Hour Division ldquoMinimum Wage Laws inthe States ndash Januar y 1 2014rdquo available at httpwwwdolgovwhdminwageamericahtm (last accessedFebruary 2014)

27 See for example Sylvia Allegretto and others ldquoCredibleResearch Designs for Minimum Wage Studiesrdquo WorkingPaper 148-113 (Berkeley California Institute forResearch on Labor and Employment 2013) available athttpwwwirleberkeleyeduworkingpapers148-13pdf

28 We will report these results in a forthcoming workingpaper

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4243

Endnotes | wwwamericanprogresso

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4343

The Center for American Progress is a nonpartisan research and educational institute

dedicated to promoting a strong just and free America that ensures opportunity

for all We believe that Americans are bound together by a common commitment to

these values and we aspire to ensure that our national policies reflect these values

We work to find progressive and pragmatic solutions to significant domestic and

international problems and develop policy proposals that foster a government that

is ldquoof the people by the people and for the peoplerdquo

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3643

Appendix C | wwwamericanprogresso

Appendix C Harkin-Miller

policy simulation results

StateMinimum wage

(2014)

SNAP enrollment (2012) Predicted change in SNAP enrollment

RateTotal

(persons)

Model 1 family

enrollment (linearprobability)

Model 2 state

enrollment (linearregression)

Model 3 s

expenditu(linear regre

Alabama $725 164 788682 -66922 -58897 -58906

Alaska $775 120 87436 -8104 -6887 -3288

Arizona $790 201 1319323 -67435 -56738 -64356

Arkansas $725 78 230489 -40977 -36063 -35248

California $800 206 7813680 -371131 -310222 -18223

Colorado $800 164 853155 -50684 -42365 -23926

Connecticut $870 91 326621 -22456 -17975 -13711

Delaware $725 186 170262 -12739 -11211 -10647

District of Columbia $825 133 84009 -5370 -4417 -3632

Florida $793 166 3208026 -195813 -164426 -13046

Georgia $725 160 1586336 -137741 -121224 -11004

Hawaii $725 96 133662 -19310 -16995 -14933

Idaho $725 92 147501 -22165 -19507 -15809

Illinois $825 95 1225084 -109088 -89742 -70955

Indiana $725 125 816233 -90818 -79928 -83985

Iowa $725 155 478011 -42716 -37594 -28556

Kansas $725 135 388269 -40082 -35275 -27461

Kentucky $725 130 568821 -60840 -53544 -52259

Louisiana $725 149 683832 -63929 -56263 -66083

Maine $750 77 101976 -16567 -14323 -15234

Maryland $725 144 846415 -81748 -71946 -38370

Massachusetts $800 130 864721 -64902 -54251 -42913

Michigan $740 146 1439141 -128801 -112140 -11022

Minnesota $725 133 713646 -74730 -65769 -37878

TABLE 10

SNAP enrollments Predicted changes by state under the Harkin-Miller bill

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3743

34 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

StateMinimum wage

(2014)

SNAP enrollment (2012) Predicted change in SNAP enrollment

RateTotal

(persons)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 state

expenditures

(linear regression)

Mississippi $725 129 386501 -41486 -36511 -46467

Missouri $750 172 1036182 -75131 -64952 -56244

Montana $790 132 132452 -10350 -8708 -5846

Nebraska $725 124 230683 -25773 -22683 -12189

Nevada $825 162 446035 -23349 -19209 -11894

New Hampshire $725 127 168404 -18359 -16157 -5735

New Jersey $825 160 1416666 -75175 -61843 -28236

New Mexico $750 149 310896 -25983 -22463 -22512

New York $800 192 3763553 -191193 -159815 -142182

North Carolina $725 174 1697193 -135417 -119179 -113503

North Dakota $725 87 61225 -9743 -8574 -4021

Ohio $795 143 1647345 -115869 -97169 -88580

Oklahoma $725 129 494053 -53006 -46650 -46854

Oregon $910 124 485326 -17036 -13328 -16398

Pennsylvania $725 161 2053643 -177315 -156052 -125586

Rhode Island $800 156 163730 -10258 -8574 -8698

South Carolina $725 94 445277 -65614 -57746 -50304

South Dakota $725 208 173749 -11586 -10197 -7458

Tennessee $725 142 914903 -89667 -78915 -99134

Texas $725 110 2863779 -362018 -318607 -253285

Utah $725 88 251107 -39658 -34902 -19390

Vermont $873 156 97792 -3823 -3055 -2475

Virginia $725 101 829771 -113723 -100086 -58212

Washington $932 72 496934 -23221 -17947 -17756

West Virginia $725 58 107875 -25792 -22699 -21665

Wisconsin $725 75 427822 -79521 -69986 -53210

Wyoming $725 164 94590 -8010 -7050 -3104

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3843

Appendix C | wwwamericanprogresso

StateMinimum wage

(2014)

SNAP expenditures (2012)Predicted change in SNAP expenditures

(millions of dollars)

Per resident

(millions of

dollars)

Total

(millions of

dollars)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 s

expenditu

(linear regre

Alabama $725 $290 $1397 -$1185 -$1043 -$1043

Alaska $775 $253 $185 -$262 -$223 -$106

Arizona $790 $257 $1686 -$935 -$787 -$892

Arkansas $725 $248 $733 -$636 -$560 -$547

California $800 $189 $7164 -$7276 -$6082 -$3573

Colorado $800 $157 $816 -$862 -$721 -$407

Connecticut $870 $191 $686 -$343 -$275 -$210

Delaware $725 $250 $229 -$205 -$180 -$171

District of Columbia $825 $366 $232 -$146 -$120 -$99

Florida $793 $294 $5676 -$4429 -$3719 -$2951

Georgia $725 $317 $3140 -$2936 -$2584 -$2346

Hawaii $725 $335 $465 -$449 -$395 -$347

Idaho $725 $225 $359 -$376 -$331 -$268

Illinois $825 $249 $3200 -$2096 -$1725 -$1364

Indiana $725 $220 $1439 -$1162 -$1023 -$1075

Iowa $725 $192 $589 -$658 -$579 -$440

Kansas $725 $159 $460 -$502 -$441 -$344

Kentucky $725 $298 $1303 -$1133 -$997 -$973

Louisiana $725 $315 $1450 -$1047 -$922 -$1083

Maine $750 $281 $373 -$267 -$231 -$246

Maryland $725 $188 $1109 -$1765 -$1553 -$828

Massachusetts $800 $206 $1366 -$1030 -$861 -$681

Michigan $740 $300 $2963 -$2400 -$2090 -$2054

Minnesota $725 $140 $755 -$1113 -$980 -$564

Mississippi $725 $326 $973 -$649 -$571 -$726

Missouri $750 $241 $1452 -$1278 -$1104 -$956

Montana $790 $190 $191 -$179 -$151 -$101

Nebraska $725 $140 $259 -$409 -$360 -$194

Nevada $825 $191 $527 -$441 -$363 -$225

New Hampshire $725 $126 $167 -$399 -$351 -$125

TABLE 11

SNAP expenditures Predicted changes by state under the Harkin-Miller bill

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3943

36 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

StateMinimum wage

(2014)

SNAP expenditures (2012)Predicted change in SNAP expenditures

(millions of dollars)

Per resident

(millions of

dollars)

Total

(millions of

dollars)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 state

expenditures

(linear regression)

New Jersey $825 $160 $1420 -$1611 -$1325 -$605

New Mexico $750 $324 $675 -$513 -$444 -$445New York $800 $287 $5616 -$3766 -$3148 -$2801

North Carolina $725 $252 $2454 -$2187 -$1925 -$1833

North Dakota $725 $128 $90 -$162 -$143 -$67

Ohio $795 $259 $2995 -$2013 -$1688 -$1539

Oklahoma $725 $248 $945 -$799 -$703 -$706

Oregon $910 $322 $1255 -$272 -$213 -$262

Pennsylvania $725 $218 $2779 -$2930 -$2579 -$2075

Rhode Island $800 $280 $294 -$173 -$144 -$147

South Carolina $725 $291 $1373 -$1337 -$1177 -$1025South Dakota $725 $198 $165 -$192 -$169 -$123

Tennessee $725 $324 $2091 -$1413 -$1243 -$1562

Texas $725 $230 $5997 -$6402 -$5634 -$4479

Utah $725 $141 $402 -$614 -$541 -$300

Vermont $873 $230 $144 -$66 -$53 -$43

Virginia $725 $173 $1413 -$2062 -$1815 -$1056

Washington $932 $244 $1682 -$350 -$270 -$267

West Virginia $725 $273 $508 -$451 -$397 -$379

Wisconsin $725 $204 $1166 -$1302 -$1146 -$871Wyoming $725 $95 $55 -$105 -$93 -$41

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4043

Appendix C | wwwamericanprogresso

FIGURE 2

Raising the minimum wage to $1010would cut taxpayer costs in every state

Predicted decreases in cost and enrollment

in SNAP in 50 states

$200+$51ndash$100

$101ndash$200

0ndash$25

$26ndash$50

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4143

38 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

Endnotes

1 US Department of Agriculture ldquoSupplementalNutrition Assistance Program (SNAP)rdquo available athttpwwwfnsusdagovsnapeligibility (last accessedFebruary 2014)

2 For this initial analysis we do not consider Harkin-Millerrsquos increase in subminimum wages for tippedworkers To do s o would increase the estimated SNAP

savings by an unknown amount

3 The Congressional Budget Office estimates thatworkers currently earning between $1010 and $1150per hour would see their wages rise under the Harkin-Miller proposal Congressional Budget O ffice ldquoTheEffects of a Minimum Wage Increase on Employmentand Family Incomerdquo (2014)

4 Marianne Page Joanne Spetz and Jane Millar ldquoDoesthe Minimum Wage Affect Welfare Caseloadsrdquo Journalof Policy Analysis and Management 24 (2) (2005)273ndash295

5 Marianne Bitler and Hilary Hoynes ldquoThe More ThingsChange the More They Stay the Same The SafetyNet Living Arrangements and Poverty in the GreatRecessionrdquo NBER Working Paper 194 49 2013

6 Sylvia Allegretto and others ldquoFast Food Poverty Wages The Public Cost of Low-Wage Jobs in the Fast-FoodIndustryrdquo (Berkeley California Center for LaborResearch and Education 2013) available at httplaborcenterberkeleyedupubliccostsfast_food_poverty_wage

7 David Neumark and William Wascher ldquoDoes a HigherMinimum Wage Enhance the Effectiveness of theEarned Income Tax Creditrdquo Industrial and LaborRelations Review 64 (4) (2011) 712ndash746

8 David Lee and Emmanuel Saez ldquoOptimal MinimumWage Policy in Competitive Labor Marketsrdquo Journal ofPublic Economics 96 (9) (2012) 739ndash749

9 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family IncomerdquoFebruary 2014

10 Jesse Rothstein ldquoIs the EITC as Good as an NITConditional Cash Transfers and Tax Incidencerdquo AmericanEconomic Journal Economic Policy 2 (1) (2010) 177ndash208

11 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family Incomerdquo

12 Dube Arindrajit 2013 rdquoMinimum Wagesand the Distribution of Family IncomerdquoUnpublished working paper Available at httpsdldropboxusercontentcomu15038936Dube_MinimumWagesFamilyIncomespdf

13 As Dube explains in ldquoThe poverty of Minimum WageFactsrdquo the simulation approach underestimate stemsfrom a number of unwarranted assumptions includingthe range of actual wage increases and the accuracy ofwage data in the Current Population Survey The causal

approach does not make these assumptions

14 Allegretto Sylvia and others 2013 ldquoCredible ResearchDesigns for Minimum Wage Studiesrdquo Working Paper148-13 University of California Berkeley Institutefor Research on Labor and Employment Available athttpwwwirleberkeleyeduworkingpapers148-13pdf

15 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family In comerdquo

16 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo (2012)

17 Ibid

18 Ibid

19 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo

20 US Department of Agriculture ldquoSupplementalNutrition Assistance Program (SNAP)rdquo

21 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo

22 Ibid

23 We deviate from the Census Bureaursquos definition ofa family unit which is ldquotwo people or more (on e ofwhom is the householder) related by bir th marriageor adoption and residing togetherrdquo US Bureau ofthe Census ldquoCurrent Population Survey DefinitionsFamilyrdquo available at httpwwwcensusgovcpsabout

cpsdefhtml (last accessed February 2014) We countas a family unit any individual residing on his or herown two or more persons residing together whodo not belong to a family in the March CPS sampleare constructed as one family in our analysis For thepurposes of food stamp allocations the consumptionresulting from this transfer is probably distributed tofamily members (rather than household members ora single individual within the household) Howeversingle individuals canmdashand domdashreceive SNAPbenefits Excluding them would fail to make theanalysis reflective of the population at large

24 Strictly the family level linear probability modelpredicts the percentage-point decrease in theprobability that an individual family will receive SNAPpayments When applied to a large number of familieshowever we are able to interpret the coefficient asa decrease in the mean of enrollmentmdashthat is a

decrease in the enrollment ratemdashby applying the lawof iterated expectations

25 We generate expenditure predictions from theenrollment modelsmdashand conversely generateenrollment predictions from the expenditure modelmdashby assuming that expenditures per enrolled familyremains the same before and after the minimum wagechange In practice this is likely to be a conservativeestimatemdashthat is to underestimate the decrease inSNAP activity Average SNAP benefits per family willalso decrease as many families that remain eligible forSNAP experience income gains

26 Wage and Hour Division ldquoMinimum Wage Laws inthe States ndash Januar y 1 2014rdquo available at httpwwwdolgovwhdminwageamericahtm (last accessedFebruary 2014)

27 See for example Sylvia Allegretto and others ldquoCredibleResearch Designs for Minimum Wage Studiesrdquo WorkingPaper 148-113 (Berkeley California Institute forResearch on Labor and Employment 2013) available athttpwwwirleberkeleyeduworkingpapers148-13pdf

28 We will report these results in a forthcoming workingpaper

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4243

Endnotes | wwwamericanprogresso

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4343

The Center for American Progress is a nonpartisan research and educational institute

dedicated to promoting a strong just and free America that ensures opportunity

for all We believe that Americans are bound together by a common commitment to

these values and we aspire to ensure that our national policies reflect these values

We work to find progressive and pragmatic solutions to significant domestic and

international problems and develop policy proposals that foster a government that

is ldquoof the people by the people and for the peoplerdquo

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3743

34 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

StateMinimum wage

(2014)

SNAP enrollment (2012) Predicted change in SNAP enrollment

RateTotal

(persons)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 state

expenditures

(linear regression)

Mississippi $725 129 386501 -41486 -36511 -46467

Missouri $750 172 1036182 -75131 -64952 -56244

Montana $790 132 132452 -10350 -8708 -5846

Nebraska $725 124 230683 -25773 -22683 -12189

Nevada $825 162 446035 -23349 -19209 -11894

New Hampshire $725 127 168404 -18359 -16157 -5735

New Jersey $825 160 1416666 -75175 -61843 -28236

New Mexico $750 149 310896 -25983 -22463 -22512

New York $800 192 3763553 -191193 -159815 -142182

North Carolina $725 174 1697193 -135417 -119179 -113503

North Dakota $725 87 61225 -9743 -8574 -4021

Ohio $795 143 1647345 -115869 -97169 -88580

Oklahoma $725 129 494053 -53006 -46650 -46854

Oregon $910 124 485326 -17036 -13328 -16398

Pennsylvania $725 161 2053643 -177315 -156052 -125586

Rhode Island $800 156 163730 -10258 -8574 -8698

South Carolina $725 94 445277 -65614 -57746 -50304

South Dakota $725 208 173749 -11586 -10197 -7458

Tennessee $725 142 914903 -89667 -78915 -99134

Texas $725 110 2863779 -362018 -318607 -253285

Utah $725 88 251107 -39658 -34902 -19390

Vermont $873 156 97792 -3823 -3055 -2475

Virginia $725 101 829771 -113723 -100086 -58212

Washington $932 72 496934 -23221 -17947 -17756

West Virginia $725 58 107875 -25792 -22699 -21665

Wisconsin $725 75 427822 -79521 -69986 -53210

Wyoming $725 164 94590 -8010 -7050 -3104

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3843

Appendix C | wwwamericanprogresso

StateMinimum wage

(2014)

SNAP expenditures (2012)Predicted change in SNAP expenditures

(millions of dollars)

Per resident

(millions of

dollars)

Total

(millions of

dollars)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 s

expenditu

(linear regre

Alabama $725 $290 $1397 -$1185 -$1043 -$1043

Alaska $775 $253 $185 -$262 -$223 -$106

Arizona $790 $257 $1686 -$935 -$787 -$892

Arkansas $725 $248 $733 -$636 -$560 -$547

California $800 $189 $7164 -$7276 -$6082 -$3573

Colorado $800 $157 $816 -$862 -$721 -$407

Connecticut $870 $191 $686 -$343 -$275 -$210

Delaware $725 $250 $229 -$205 -$180 -$171

District of Columbia $825 $366 $232 -$146 -$120 -$99

Florida $793 $294 $5676 -$4429 -$3719 -$2951

Georgia $725 $317 $3140 -$2936 -$2584 -$2346

Hawaii $725 $335 $465 -$449 -$395 -$347

Idaho $725 $225 $359 -$376 -$331 -$268

Illinois $825 $249 $3200 -$2096 -$1725 -$1364

Indiana $725 $220 $1439 -$1162 -$1023 -$1075

Iowa $725 $192 $589 -$658 -$579 -$440

Kansas $725 $159 $460 -$502 -$441 -$344

Kentucky $725 $298 $1303 -$1133 -$997 -$973

Louisiana $725 $315 $1450 -$1047 -$922 -$1083

Maine $750 $281 $373 -$267 -$231 -$246

Maryland $725 $188 $1109 -$1765 -$1553 -$828

Massachusetts $800 $206 $1366 -$1030 -$861 -$681

Michigan $740 $300 $2963 -$2400 -$2090 -$2054

Minnesota $725 $140 $755 -$1113 -$980 -$564

Mississippi $725 $326 $973 -$649 -$571 -$726

Missouri $750 $241 $1452 -$1278 -$1104 -$956

Montana $790 $190 $191 -$179 -$151 -$101

Nebraska $725 $140 $259 -$409 -$360 -$194

Nevada $825 $191 $527 -$441 -$363 -$225

New Hampshire $725 $126 $167 -$399 -$351 -$125

TABLE 11

SNAP expenditures Predicted changes by state under the Harkin-Miller bill

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3943

36 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

StateMinimum wage

(2014)

SNAP expenditures (2012)Predicted change in SNAP expenditures

(millions of dollars)

Per resident

(millions of

dollars)

Total

(millions of

dollars)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 state

expenditures

(linear regression)

New Jersey $825 $160 $1420 -$1611 -$1325 -$605

New Mexico $750 $324 $675 -$513 -$444 -$445New York $800 $287 $5616 -$3766 -$3148 -$2801

North Carolina $725 $252 $2454 -$2187 -$1925 -$1833

North Dakota $725 $128 $90 -$162 -$143 -$67

Ohio $795 $259 $2995 -$2013 -$1688 -$1539

Oklahoma $725 $248 $945 -$799 -$703 -$706

Oregon $910 $322 $1255 -$272 -$213 -$262

Pennsylvania $725 $218 $2779 -$2930 -$2579 -$2075

Rhode Island $800 $280 $294 -$173 -$144 -$147

South Carolina $725 $291 $1373 -$1337 -$1177 -$1025South Dakota $725 $198 $165 -$192 -$169 -$123

Tennessee $725 $324 $2091 -$1413 -$1243 -$1562

Texas $725 $230 $5997 -$6402 -$5634 -$4479

Utah $725 $141 $402 -$614 -$541 -$300

Vermont $873 $230 $144 -$66 -$53 -$43

Virginia $725 $173 $1413 -$2062 -$1815 -$1056

Washington $932 $244 $1682 -$350 -$270 -$267

West Virginia $725 $273 $508 -$451 -$397 -$379

Wisconsin $725 $204 $1166 -$1302 -$1146 -$871Wyoming $725 $95 $55 -$105 -$93 -$41

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4043

Appendix C | wwwamericanprogresso

FIGURE 2

Raising the minimum wage to $1010would cut taxpayer costs in every state

Predicted decreases in cost and enrollment

in SNAP in 50 states

$200+$51ndash$100

$101ndash$200

0ndash$25

$26ndash$50

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4143

38 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

Endnotes

1 US Department of Agriculture ldquoSupplementalNutrition Assistance Program (SNAP)rdquo available athttpwwwfnsusdagovsnapeligibility (last accessedFebruary 2014)

2 For this initial analysis we do not consider Harkin-Millerrsquos increase in subminimum wages for tippedworkers To do s o would increase the estimated SNAP

savings by an unknown amount

3 The Congressional Budget Office estimates thatworkers currently earning between $1010 and $1150per hour would see their wages rise under the Harkin-Miller proposal Congressional Budget O ffice ldquoTheEffects of a Minimum Wage Increase on Employmentand Family Incomerdquo (2014)

4 Marianne Page Joanne Spetz and Jane Millar ldquoDoesthe Minimum Wage Affect Welfare Caseloadsrdquo Journalof Policy Analysis and Management 24 (2) (2005)273ndash295

5 Marianne Bitler and Hilary Hoynes ldquoThe More ThingsChange the More They Stay the Same The SafetyNet Living Arrangements and Poverty in the GreatRecessionrdquo NBER Working Paper 194 49 2013

6 Sylvia Allegretto and others ldquoFast Food Poverty Wages The Public Cost of Low-Wage Jobs in the Fast-FoodIndustryrdquo (Berkeley California Center for LaborResearch and Education 2013) available at httplaborcenterberkeleyedupubliccostsfast_food_poverty_wage

7 David Neumark and William Wascher ldquoDoes a HigherMinimum Wage Enhance the Effectiveness of theEarned Income Tax Creditrdquo Industrial and LaborRelations Review 64 (4) (2011) 712ndash746

8 David Lee and Emmanuel Saez ldquoOptimal MinimumWage Policy in Competitive Labor Marketsrdquo Journal ofPublic Economics 96 (9) (2012) 739ndash749

9 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family IncomerdquoFebruary 2014

10 Jesse Rothstein ldquoIs the EITC as Good as an NITConditional Cash Transfers and Tax Incidencerdquo AmericanEconomic Journal Economic Policy 2 (1) (2010) 177ndash208

11 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family Incomerdquo

12 Dube Arindrajit 2013 rdquoMinimum Wagesand the Distribution of Family IncomerdquoUnpublished working paper Available at httpsdldropboxusercontentcomu15038936Dube_MinimumWagesFamilyIncomespdf

13 As Dube explains in ldquoThe poverty of Minimum WageFactsrdquo the simulation approach underestimate stemsfrom a number of unwarranted assumptions includingthe range of actual wage increases and the accuracy ofwage data in the Current Population Survey The causal

approach does not make these assumptions

14 Allegretto Sylvia and others 2013 ldquoCredible ResearchDesigns for Minimum Wage Studiesrdquo Working Paper148-13 University of California Berkeley Institutefor Research on Labor and Employment Available athttpwwwirleberkeleyeduworkingpapers148-13pdf

15 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family In comerdquo

16 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo (2012)

17 Ibid

18 Ibid

19 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo

20 US Department of Agriculture ldquoSupplementalNutrition Assistance Program (SNAP)rdquo

21 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo

22 Ibid

23 We deviate from the Census Bureaursquos definition ofa family unit which is ldquotwo people or more (on e ofwhom is the householder) related by bir th marriageor adoption and residing togetherrdquo US Bureau ofthe Census ldquoCurrent Population Survey DefinitionsFamilyrdquo available at httpwwwcensusgovcpsabout

cpsdefhtml (last accessed February 2014) We countas a family unit any individual residing on his or herown two or more persons residing together whodo not belong to a family in the March CPS sampleare constructed as one family in our analysis For thepurposes of food stamp allocations the consumptionresulting from this transfer is probably distributed tofamily members (rather than household members ora single individual within the household) Howeversingle individuals canmdashand domdashreceive SNAPbenefits Excluding them would fail to make theanalysis reflective of the population at large

24 Strictly the family level linear probability modelpredicts the percentage-point decrease in theprobability that an individual family will receive SNAPpayments When applied to a large number of familieshowever we are able to interpret the coefficient asa decrease in the mean of enrollmentmdashthat is a

decrease in the enrollment ratemdashby applying the lawof iterated expectations

25 We generate expenditure predictions from theenrollment modelsmdashand conversely generateenrollment predictions from the expenditure modelmdashby assuming that expenditures per enrolled familyremains the same before and after the minimum wagechange In practice this is likely to be a conservativeestimatemdashthat is to underestimate the decrease inSNAP activity Average SNAP benefits per family willalso decrease as many families that remain eligible forSNAP experience income gains

26 Wage and Hour Division ldquoMinimum Wage Laws inthe States ndash Januar y 1 2014rdquo available at httpwwwdolgovwhdminwageamericahtm (last accessedFebruary 2014)

27 See for example Sylvia Allegretto and others ldquoCredibleResearch Designs for Minimum Wage Studiesrdquo WorkingPaper 148-113 (Berkeley California Institute forResearch on Labor and Employment 2013) available athttpwwwirleberkeleyeduworkingpapers148-13pdf

28 We will report these results in a forthcoming workingpaper

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4243

Endnotes | wwwamericanprogresso

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4343

The Center for American Progress is a nonpartisan research and educational institute

dedicated to promoting a strong just and free America that ensures opportunity

for all We believe that Americans are bound together by a common commitment to

these values and we aspire to ensure that our national policies reflect these values

We work to find progressive and pragmatic solutions to significant domestic and

international problems and develop policy proposals that foster a government that

is ldquoof the people by the people and for the peoplerdquo

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3843

Appendix C | wwwamericanprogresso

StateMinimum wage

(2014)

SNAP expenditures (2012)Predicted change in SNAP expenditures

(millions of dollars)

Per resident

(millions of

dollars)

Total

(millions of

dollars)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 s

expenditu

(linear regre

Alabama $725 $290 $1397 -$1185 -$1043 -$1043

Alaska $775 $253 $185 -$262 -$223 -$106

Arizona $790 $257 $1686 -$935 -$787 -$892

Arkansas $725 $248 $733 -$636 -$560 -$547

California $800 $189 $7164 -$7276 -$6082 -$3573

Colorado $800 $157 $816 -$862 -$721 -$407

Connecticut $870 $191 $686 -$343 -$275 -$210

Delaware $725 $250 $229 -$205 -$180 -$171

District of Columbia $825 $366 $232 -$146 -$120 -$99

Florida $793 $294 $5676 -$4429 -$3719 -$2951

Georgia $725 $317 $3140 -$2936 -$2584 -$2346

Hawaii $725 $335 $465 -$449 -$395 -$347

Idaho $725 $225 $359 -$376 -$331 -$268

Illinois $825 $249 $3200 -$2096 -$1725 -$1364

Indiana $725 $220 $1439 -$1162 -$1023 -$1075

Iowa $725 $192 $589 -$658 -$579 -$440

Kansas $725 $159 $460 -$502 -$441 -$344

Kentucky $725 $298 $1303 -$1133 -$997 -$973

Louisiana $725 $315 $1450 -$1047 -$922 -$1083

Maine $750 $281 $373 -$267 -$231 -$246

Maryland $725 $188 $1109 -$1765 -$1553 -$828

Massachusetts $800 $206 $1366 -$1030 -$861 -$681

Michigan $740 $300 $2963 -$2400 -$2090 -$2054

Minnesota $725 $140 $755 -$1113 -$980 -$564

Mississippi $725 $326 $973 -$649 -$571 -$726

Missouri $750 $241 $1452 -$1278 -$1104 -$956

Montana $790 $190 $191 -$179 -$151 -$101

Nebraska $725 $140 $259 -$409 -$360 -$194

Nevada $825 $191 $527 -$441 -$363 -$225

New Hampshire $725 $126 $167 -$399 -$351 -$125

TABLE 11

SNAP expenditures Predicted changes by state under the Harkin-Miller bill

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3943

36 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

StateMinimum wage

(2014)

SNAP expenditures (2012)Predicted change in SNAP expenditures

(millions of dollars)

Per resident

(millions of

dollars)

Total

(millions of

dollars)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 state

expenditures

(linear regression)

New Jersey $825 $160 $1420 -$1611 -$1325 -$605

New Mexico $750 $324 $675 -$513 -$444 -$445New York $800 $287 $5616 -$3766 -$3148 -$2801

North Carolina $725 $252 $2454 -$2187 -$1925 -$1833

North Dakota $725 $128 $90 -$162 -$143 -$67

Ohio $795 $259 $2995 -$2013 -$1688 -$1539

Oklahoma $725 $248 $945 -$799 -$703 -$706

Oregon $910 $322 $1255 -$272 -$213 -$262

Pennsylvania $725 $218 $2779 -$2930 -$2579 -$2075

Rhode Island $800 $280 $294 -$173 -$144 -$147

South Carolina $725 $291 $1373 -$1337 -$1177 -$1025South Dakota $725 $198 $165 -$192 -$169 -$123

Tennessee $725 $324 $2091 -$1413 -$1243 -$1562

Texas $725 $230 $5997 -$6402 -$5634 -$4479

Utah $725 $141 $402 -$614 -$541 -$300

Vermont $873 $230 $144 -$66 -$53 -$43

Virginia $725 $173 $1413 -$2062 -$1815 -$1056

Washington $932 $244 $1682 -$350 -$270 -$267

West Virginia $725 $273 $508 -$451 -$397 -$379

Wisconsin $725 $204 $1166 -$1302 -$1146 -$871Wyoming $725 $95 $55 -$105 -$93 -$41

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4043

Appendix C | wwwamericanprogresso

FIGURE 2

Raising the minimum wage to $1010would cut taxpayer costs in every state

Predicted decreases in cost and enrollment

in SNAP in 50 states

$200+$51ndash$100

$101ndash$200

0ndash$25

$26ndash$50

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4143

38 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

Endnotes

1 US Department of Agriculture ldquoSupplementalNutrition Assistance Program (SNAP)rdquo available athttpwwwfnsusdagovsnapeligibility (last accessedFebruary 2014)

2 For this initial analysis we do not consider Harkin-Millerrsquos increase in subminimum wages for tippedworkers To do s o would increase the estimated SNAP

savings by an unknown amount

3 The Congressional Budget Office estimates thatworkers currently earning between $1010 and $1150per hour would see their wages rise under the Harkin-Miller proposal Congressional Budget O ffice ldquoTheEffects of a Minimum Wage Increase on Employmentand Family Incomerdquo (2014)

4 Marianne Page Joanne Spetz and Jane Millar ldquoDoesthe Minimum Wage Affect Welfare Caseloadsrdquo Journalof Policy Analysis and Management 24 (2) (2005)273ndash295

5 Marianne Bitler and Hilary Hoynes ldquoThe More ThingsChange the More They Stay the Same The SafetyNet Living Arrangements and Poverty in the GreatRecessionrdquo NBER Working Paper 194 49 2013

6 Sylvia Allegretto and others ldquoFast Food Poverty Wages The Public Cost of Low-Wage Jobs in the Fast-FoodIndustryrdquo (Berkeley California Center for LaborResearch and Education 2013) available at httplaborcenterberkeleyedupubliccostsfast_food_poverty_wage

7 David Neumark and William Wascher ldquoDoes a HigherMinimum Wage Enhance the Effectiveness of theEarned Income Tax Creditrdquo Industrial and LaborRelations Review 64 (4) (2011) 712ndash746

8 David Lee and Emmanuel Saez ldquoOptimal MinimumWage Policy in Competitive Labor Marketsrdquo Journal ofPublic Economics 96 (9) (2012) 739ndash749

9 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family IncomerdquoFebruary 2014

10 Jesse Rothstein ldquoIs the EITC as Good as an NITConditional Cash Transfers and Tax Incidencerdquo AmericanEconomic Journal Economic Policy 2 (1) (2010) 177ndash208

11 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family Incomerdquo

12 Dube Arindrajit 2013 rdquoMinimum Wagesand the Distribution of Family IncomerdquoUnpublished working paper Available at httpsdldropboxusercontentcomu15038936Dube_MinimumWagesFamilyIncomespdf

13 As Dube explains in ldquoThe poverty of Minimum WageFactsrdquo the simulation approach underestimate stemsfrom a number of unwarranted assumptions includingthe range of actual wage increases and the accuracy ofwage data in the Current Population Survey The causal

approach does not make these assumptions

14 Allegretto Sylvia and others 2013 ldquoCredible ResearchDesigns for Minimum Wage Studiesrdquo Working Paper148-13 University of California Berkeley Institutefor Research on Labor and Employment Available athttpwwwirleberkeleyeduworkingpapers148-13pdf

15 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family In comerdquo

16 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo (2012)

17 Ibid

18 Ibid

19 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo

20 US Department of Agriculture ldquoSupplementalNutrition Assistance Program (SNAP)rdquo

21 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo

22 Ibid

23 We deviate from the Census Bureaursquos definition ofa family unit which is ldquotwo people or more (on e ofwhom is the householder) related by bir th marriageor adoption and residing togetherrdquo US Bureau ofthe Census ldquoCurrent Population Survey DefinitionsFamilyrdquo available at httpwwwcensusgovcpsabout

cpsdefhtml (last accessed February 2014) We countas a family unit any individual residing on his or herown two or more persons residing together whodo not belong to a family in the March CPS sampleare constructed as one family in our analysis For thepurposes of food stamp allocations the consumptionresulting from this transfer is probably distributed tofamily members (rather than household members ora single individual within the household) Howeversingle individuals canmdashand domdashreceive SNAPbenefits Excluding them would fail to make theanalysis reflective of the population at large

24 Strictly the family level linear probability modelpredicts the percentage-point decrease in theprobability that an individual family will receive SNAPpayments When applied to a large number of familieshowever we are able to interpret the coefficient asa decrease in the mean of enrollmentmdashthat is a

decrease in the enrollment ratemdashby applying the lawof iterated expectations

25 We generate expenditure predictions from theenrollment modelsmdashand conversely generateenrollment predictions from the expenditure modelmdashby assuming that expenditures per enrolled familyremains the same before and after the minimum wagechange In practice this is likely to be a conservativeestimatemdashthat is to underestimate the decrease inSNAP activity Average SNAP benefits per family willalso decrease as many families that remain eligible forSNAP experience income gains

26 Wage and Hour Division ldquoMinimum Wage Laws inthe States ndash Januar y 1 2014rdquo available at httpwwwdolgovwhdminwageamericahtm (last accessedFebruary 2014)

27 See for example Sylvia Allegretto and others ldquoCredibleResearch Designs for Minimum Wage Studiesrdquo WorkingPaper 148-113 (Berkeley California Institute forResearch on Labor and Employment 2013) available athttpwwwirleberkeleyeduworkingpapers148-13pdf

28 We will report these results in a forthcoming workingpaper

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4243

Endnotes | wwwamericanprogresso

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4343

The Center for American Progress is a nonpartisan research and educational institute

dedicated to promoting a strong just and free America that ensures opportunity

for all We believe that Americans are bound together by a common commitment to

these values and we aspire to ensure that our national policies reflect these values

We work to find progressive and pragmatic solutions to significant domestic and

international problems and develop policy proposals that foster a government that

is ldquoof the people by the people and for the peoplerdquo

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 3943

36 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

StateMinimum wage

(2014)

SNAP expenditures (2012)Predicted change in SNAP expenditures

(millions of dollars)

Per resident

(millions of

dollars)

Total

(millions of

dollars)

Model 1 family

enrollment (linear

probability)

Model 2 state

enrollment (linear

regression)

Model 3 state

expenditures

(linear regression)

New Jersey $825 $160 $1420 -$1611 -$1325 -$605

New Mexico $750 $324 $675 -$513 -$444 -$445New York $800 $287 $5616 -$3766 -$3148 -$2801

North Carolina $725 $252 $2454 -$2187 -$1925 -$1833

North Dakota $725 $128 $90 -$162 -$143 -$67

Ohio $795 $259 $2995 -$2013 -$1688 -$1539

Oklahoma $725 $248 $945 -$799 -$703 -$706

Oregon $910 $322 $1255 -$272 -$213 -$262

Pennsylvania $725 $218 $2779 -$2930 -$2579 -$2075

Rhode Island $800 $280 $294 -$173 -$144 -$147

South Carolina $725 $291 $1373 -$1337 -$1177 -$1025South Dakota $725 $198 $165 -$192 -$169 -$123

Tennessee $725 $324 $2091 -$1413 -$1243 -$1562

Texas $725 $230 $5997 -$6402 -$5634 -$4479

Utah $725 $141 $402 -$614 -$541 -$300

Vermont $873 $230 $144 -$66 -$53 -$43

Virginia $725 $173 $1413 -$2062 -$1815 -$1056

Washington $932 $244 $1682 -$350 -$270 -$267

West Virginia $725 $273 $508 -$451 -$397 -$379

Wisconsin $725 $204 $1166 -$1302 -$1146 -$871Wyoming $725 $95 $55 -$105 -$93 -$41

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4043

Appendix C | wwwamericanprogresso

FIGURE 2

Raising the minimum wage to $1010would cut taxpayer costs in every state

Predicted decreases in cost and enrollment

in SNAP in 50 states

$200+$51ndash$100

$101ndash$200

0ndash$25

$26ndash$50

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4143

38 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

Endnotes

1 US Department of Agriculture ldquoSupplementalNutrition Assistance Program (SNAP)rdquo available athttpwwwfnsusdagovsnapeligibility (last accessedFebruary 2014)

2 For this initial analysis we do not consider Harkin-Millerrsquos increase in subminimum wages for tippedworkers To do s o would increase the estimated SNAP

savings by an unknown amount

3 The Congressional Budget Office estimates thatworkers currently earning between $1010 and $1150per hour would see their wages rise under the Harkin-Miller proposal Congressional Budget O ffice ldquoTheEffects of a Minimum Wage Increase on Employmentand Family Incomerdquo (2014)

4 Marianne Page Joanne Spetz and Jane Millar ldquoDoesthe Minimum Wage Affect Welfare Caseloadsrdquo Journalof Policy Analysis and Management 24 (2) (2005)273ndash295

5 Marianne Bitler and Hilary Hoynes ldquoThe More ThingsChange the More They Stay the Same The SafetyNet Living Arrangements and Poverty in the GreatRecessionrdquo NBER Working Paper 194 49 2013

6 Sylvia Allegretto and others ldquoFast Food Poverty Wages The Public Cost of Low-Wage Jobs in the Fast-FoodIndustryrdquo (Berkeley California Center for LaborResearch and Education 2013) available at httplaborcenterberkeleyedupubliccostsfast_food_poverty_wage

7 David Neumark and William Wascher ldquoDoes a HigherMinimum Wage Enhance the Effectiveness of theEarned Income Tax Creditrdquo Industrial and LaborRelations Review 64 (4) (2011) 712ndash746

8 David Lee and Emmanuel Saez ldquoOptimal MinimumWage Policy in Competitive Labor Marketsrdquo Journal ofPublic Economics 96 (9) (2012) 739ndash749

9 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family IncomerdquoFebruary 2014

10 Jesse Rothstein ldquoIs the EITC as Good as an NITConditional Cash Transfers and Tax Incidencerdquo AmericanEconomic Journal Economic Policy 2 (1) (2010) 177ndash208

11 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family Incomerdquo

12 Dube Arindrajit 2013 rdquoMinimum Wagesand the Distribution of Family IncomerdquoUnpublished working paper Available at httpsdldropboxusercontentcomu15038936Dube_MinimumWagesFamilyIncomespdf

13 As Dube explains in ldquoThe poverty of Minimum WageFactsrdquo the simulation approach underestimate stemsfrom a number of unwarranted assumptions includingthe range of actual wage increases and the accuracy ofwage data in the Current Population Survey The causal

approach does not make these assumptions

14 Allegretto Sylvia and others 2013 ldquoCredible ResearchDesigns for Minimum Wage Studiesrdquo Working Paper148-13 University of California Berkeley Institutefor Research on Labor and Employment Available athttpwwwirleberkeleyeduworkingpapers148-13pdf

15 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family In comerdquo

16 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo (2012)

17 Ibid

18 Ibid

19 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo

20 US Department of Agriculture ldquoSupplementalNutrition Assistance Program (SNAP)rdquo

21 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo

22 Ibid

23 We deviate from the Census Bureaursquos definition ofa family unit which is ldquotwo people or more (on e ofwhom is the householder) related by bir th marriageor adoption and residing togetherrdquo US Bureau ofthe Census ldquoCurrent Population Survey DefinitionsFamilyrdquo available at httpwwwcensusgovcpsabout

cpsdefhtml (last accessed February 2014) We countas a family unit any individual residing on his or herown two or more persons residing together whodo not belong to a family in the March CPS sampleare constructed as one family in our analysis For thepurposes of food stamp allocations the consumptionresulting from this transfer is probably distributed tofamily members (rather than household members ora single individual within the household) Howeversingle individuals canmdashand domdashreceive SNAPbenefits Excluding them would fail to make theanalysis reflective of the population at large

24 Strictly the family level linear probability modelpredicts the percentage-point decrease in theprobability that an individual family will receive SNAPpayments When applied to a large number of familieshowever we are able to interpret the coefficient asa decrease in the mean of enrollmentmdashthat is a

decrease in the enrollment ratemdashby applying the lawof iterated expectations

25 We generate expenditure predictions from theenrollment modelsmdashand conversely generateenrollment predictions from the expenditure modelmdashby assuming that expenditures per enrolled familyremains the same before and after the minimum wagechange In practice this is likely to be a conservativeestimatemdashthat is to underestimate the decrease inSNAP activity Average SNAP benefits per family willalso decrease as many families that remain eligible forSNAP experience income gains

26 Wage and Hour Division ldquoMinimum Wage Laws inthe States ndash Januar y 1 2014rdquo available at httpwwwdolgovwhdminwageamericahtm (last accessedFebruary 2014)

27 See for example Sylvia Allegretto and others ldquoCredibleResearch Designs for Minimum Wage Studiesrdquo WorkingPaper 148-113 (Berkeley California Institute forResearch on Labor and Employment 2013) available athttpwwwirleberkeleyeduworkingpapers148-13pdf

28 We will report these results in a forthcoming workingpaper

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4243

Endnotes | wwwamericanprogresso

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4343

The Center for American Progress is a nonpartisan research and educational institute

dedicated to promoting a strong just and free America that ensures opportunity

for all We believe that Americans are bound together by a common commitment to

these values and we aspire to ensure that our national policies reflect these values

We work to find progressive and pragmatic solutions to significant domestic and

international problems and develop policy proposals that foster a government that

is ldquoof the people by the people and for the peoplerdquo

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4043

Appendix C | wwwamericanprogresso

FIGURE 2

Raising the minimum wage to $1010would cut taxpayer costs in every state

Predicted decreases in cost and enrollment

in SNAP in 50 states

$200+$51ndash$100

$101ndash$200

0ndash$25

$26ndash$50

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4143

38 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

Endnotes

1 US Department of Agriculture ldquoSupplementalNutrition Assistance Program (SNAP)rdquo available athttpwwwfnsusdagovsnapeligibility (last accessedFebruary 2014)

2 For this initial analysis we do not consider Harkin-Millerrsquos increase in subminimum wages for tippedworkers To do s o would increase the estimated SNAP

savings by an unknown amount

3 The Congressional Budget Office estimates thatworkers currently earning between $1010 and $1150per hour would see their wages rise under the Harkin-Miller proposal Congressional Budget O ffice ldquoTheEffects of a Minimum Wage Increase on Employmentand Family Incomerdquo (2014)

4 Marianne Page Joanne Spetz and Jane Millar ldquoDoesthe Minimum Wage Affect Welfare Caseloadsrdquo Journalof Policy Analysis and Management 24 (2) (2005)273ndash295

5 Marianne Bitler and Hilary Hoynes ldquoThe More ThingsChange the More They Stay the Same The SafetyNet Living Arrangements and Poverty in the GreatRecessionrdquo NBER Working Paper 194 49 2013

6 Sylvia Allegretto and others ldquoFast Food Poverty Wages The Public Cost of Low-Wage Jobs in the Fast-FoodIndustryrdquo (Berkeley California Center for LaborResearch and Education 2013) available at httplaborcenterberkeleyedupubliccostsfast_food_poverty_wage

7 David Neumark and William Wascher ldquoDoes a HigherMinimum Wage Enhance the Effectiveness of theEarned Income Tax Creditrdquo Industrial and LaborRelations Review 64 (4) (2011) 712ndash746

8 David Lee and Emmanuel Saez ldquoOptimal MinimumWage Policy in Competitive Labor Marketsrdquo Journal ofPublic Economics 96 (9) (2012) 739ndash749

9 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family IncomerdquoFebruary 2014

10 Jesse Rothstein ldquoIs the EITC as Good as an NITConditional Cash Transfers and Tax Incidencerdquo AmericanEconomic Journal Economic Policy 2 (1) (2010) 177ndash208

11 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family Incomerdquo

12 Dube Arindrajit 2013 rdquoMinimum Wagesand the Distribution of Family IncomerdquoUnpublished working paper Available at httpsdldropboxusercontentcomu15038936Dube_MinimumWagesFamilyIncomespdf

13 As Dube explains in ldquoThe poverty of Minimum WageFactsrdquo the simulation approach underestimate stemsfrom a number of unwarranted assumptions includingthe range of actual wage increases and the accuracy ofwage data in the Current Population Survey The causal

approach does not make these assumptions

14 Allegretto Sylvia and others 2013 ldquoCredible ResearchDesigns for Minimum Wage Studiesrdquo Working Paper148-13 University of California Berkeley Institutefor Research on Labor and Employment Available athttpwwwirleberkeleyeduworkingpapers148-13pdf

15 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family In comerdquo

16 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo (2012)

17 Ibid

18 Ibid

19 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo

20 US Department of Agriculture ldquoSupplementalNutrition Assistance Program (SNAP)rdquo

21 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo

22 Ibid

23 We deviate from the Census Bureaursquos definition ofa family unit which is ldquotwo people or more (on e ofwhom is the householder) related by bir th marriageor adoption and residing togetherrdquo US Bureau ofthe Census ldquoCurrent Population Survey DefinitionsFamilyrdquo available at httpwwwcensusgovcpsabout

cpsdefhtml (last accessed February 2014) We countas a family unit any individual residing on his or herown two or more persons residing together whodo not belong to a family in the March CPS sampleare constructed as one family in our analysis For thepurposes of food stamp allocations the consumptionresulting from this transfer is probably distributed tofamily members (rather than household members ora single individual within the household) Howeversingle individuals canmdashand domdashreceive SNAPbenefits Excluding them would fail to make theanalysis reflective of the population at large

24 Strictly the family level linear probability modelpredicts the percentage-point decrease in theprobability that an individual family will receive SNAPpayments When applied to a large number of familieshowever we are able to interpret the coefficient asa decrease in the mean of enrollmentmdashthat is a

decrease in the enrollment ratemdashby applying the lawof iterated expectations

25 We generate expenditure predictions from theenrollment modelsmdashand conversely generateenrollment predictions from the expenditure modelmdashby assuming that expenditures per enrolled familyremains the same before and after the minimum wagechange In practice this is likely to be a conservativeestimatemdashthat is to underestimate the decrease inSNAP activity Average SNAP benefits per family willalso decrease as many families that remain eligible forSNAP experience income gains

26 Wage and Hour Division ldquoMinimum Wage Laws inthe States ndash Januar y 1 2014rdquo available at httpwwwdolgovwhdminwageamericahtm (last accessedFebruary 2014)

27 See for example Sylvia Allegretto and others ldquoCredibleResearch Designs for Minimum Wage Studiesrdquo WorkingPaper 148-113 (Berkeley California Institute forResearch on Labor and Employment 2013) available athttpwwwirleberkeleyeduworkingpapers148-13pdf

28 We will report these results in a forthcoming workingpaper

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4243

Endnotes | wwwamericanprogresso

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4343

The Center for American Progress is a nonpartisan research and educational institute

dedicated to promoting a strong just and free America that ensures opportunity

for all We believe that Americans are bound together by a common commitment to

these values and we aspire to ensure that our national policies reflect these values

We work to find progressive and pragmatic solutions to significant domestic and

international problems and develop policy proposals that foster a government that

is ldquoof the people by the people and for the peoplerdquo

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4143

38 Center for American Progress | The Effec ts of M inim um Wages on SNAP Enro llments and Expe nditures

Endnotes

1 US Department of Agriculture ldquoSupplementalNutrition Assistance Program (SNAP)rdquo available athttpwwwfnsusdagovsnapeligibility (last accessedFebruary 2014)

2 For this initial analysis we do not consider Harkin-Millerrsquos increase in subminimum wages for tippedworkers To do s o would increase the estimated SNAP

savings by an unknown amount

3 The Congressional Budget Office estimates thatworkers currently earning between $1010 and $1150per hour would see their wages rise under the Harkin-Miller proposal Congressional Budget O ffice ldquoTheEffects of a Minimum Wage Increase on Employmentand Family Incomerdquo (2014)

4 Marianne Page Joanne Spetz and Jane Millar ldquoDoesthe Minimum Wage Affect Welfare Caseloadsrdquo Journalof Policy Analysis and Management 24 (2) (2005)273ndash295

5 Marianne Bitler and Hilary Hoynes ldquoThe More ThingsChange the More They Stay the Same The SafetyNet Living Arrangements and Poverty in the GreatRecessionrdquo NBER Working Paper 194 49 2013

6 Sylvia Allegretto and others ldquoFast Food Poverty Wages The Public Cost of Low-Wage Jobs in the Fast-FoodIndustryrdquo (Berkeley California Center for LaborResearch and Education 2013) available at httplaborcenterberkeleyedupubliccostsfast_food_poverty_wage

7 David Neumark and William Wascher ldquoDoes a HigherMinimum Wage Enhance the Effectiveness of theEarned Income Tax Creditrdquo Industrial and LaborRelations Review 64 (4) (2011) 712ndash746

8 David Lee and Emmanuel Saez ldquoOptimal MinimumWage Policy in Competitive Labor Marketsrdquo Journal ofPublic Economics 96 (9) (2012) 739ndash749

9 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family IncomerdquoFebruary 2014

10 Jesse Rothstein ldquoIs the EITC as Good as an NITConditional Cash Transfers and Tax Incidencerdquo AmericanEconomic Journal Economic Policy 2 (1) (2010) 177ndash208

11 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family Incomerdquo

12 Dube Arindrajit 2013 rdquoMinimum Wagesand the Distribution of Family IncomerdquoUnpublished working paper Available at httpsdldropboxusercontentcomu15038936Dube_MinimumWagesFamilyIncomespdf

13 As Dube explains in ldquoThe poverty of Minimum WageFactsrdquo the simulation approach underestimate stemsfrom a number of unwarranted assumptions includingthe range of actual wage increases and the accuracy ofwage data in the Current Population Survey The causal

approach does not make these assumptions

14 Allegretto Sylvia and others 2013 ldquoCredible ResearchDesigns for Minimum Wage Studiesrdquo Working Paper148-13 University of California Berkeley Institutefor Research on Labor and Employment Available athttpwwwirleberkeleyeduworkingpapers148-13pdf

15 Congressional Budget Office ldquoThe Effects of a MinimumWage Increase on Employment and Family In comerdquo

16 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo (2012)

17 Ibid

18 Ibid

19 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo

20 US Department of Agriculture ldquoSupplementalNutrition Assistance Program (SNAP)rdquo

21 Congressional Budget Office ldquoThe SupplementalNutrition Assistance Programrdquo

22 Ibid

23 We deviate from the Census Bureaursquos definition ofa family unit which is ldquotwo people or more (on e ofwhom is the householder) related by bir th marriageor adoption and residing togetherrdquo US Bureau ofthe Census ldquoCurrent Population Survey DefinitionsFamilyrdquo available at httpwwwcensusgovcpsabout

cpsdefhtml (last accessed February 2014) We countas a family unit any individual residing on his or herown two or more persons residing together whodo not belong to a family in the March CPS sampleare constructed as one family in our analysis For thepurposes of food stamp allocations the consumptionresulting from this transfer is probably distributed tofamily members (rather than household members ora single individual within the household) Howeversingle individuals canmdashand domdashreceive SNAPbenefits Excluding them would fail to make theanalysis reflective of the population at large

24 Strictly the family level linear probability modelpredicts the percentage-point decrease in theprobability that an individual family will receive SNAPpayments When applied to a large number of familieshowever we are able to interpret the coefficient asa decrease in the mean of enrollmentmdashthat is a

decrease in the enrollment ratemdashby applying the lawof iterated expectations

25 We generate expenditure predictions from theenrollment modelsmdashand conversely generateenrollment predictions from the expenditure modelmdashby assuming that expenditures per enrolled familyremains the same before and after the minimum wagechange In practice this is likely to be a conservativeestimatemdashthat is to underestimate the decrease inSNAP activity Average SNAP benefits per family willalso decrease as many families that remain eligible forSNAP experience income gains

26 Wage and Hour Division ldquoMinimum Wage Laws inthe States ndash Januar y 1 2014rdquo available at httpwwwdolgovwhdminwageamericahtm (last accessedFebruary 2014)

27 See for example Sylvia Allegretto and others ldquoCredibleResearch Designs for Minimum Wage Studiesrdquo WorkingPaper 148-113 (Berkeley California Institute forResearch on Labor and Employment 2013) available athttpwwwirleberkeleyeduworkingpapers148-13pdf

28 We will report these results in a forthcoming workingpaper

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4243

Endnotes | wwwamericanprogresso

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4343

The Center for American Progress is a nonpartisan research and educational institute

dedicated to promoting a strong just and free America that ensures opportunity

for all We believe that Americans are bound together by a common commitment to

these values and we aspire to ensure that our national policies reflect these values

We work to find progressive and pragmatic solutions to significant domestic and

international problems and develop policy proposals that foster a government that

is ldquoof the people by the people and for the peoplerdquo

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4243

Endnotes | wwwamericanprogresso

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4343

The Center for American Progress is a nonpartisan research and educational institute

dedicated to promoting a strong just and free America that ensures opportunity

for all We believe that Americans are bound together by a common commitment to

these values and we aspire to ensure that our national policies reflect these values

We work to find progressive and pragmatic solutions to significant domestic and

international problems and develop policy proposals that foster a government that

is ldquoof the people by the people and for the peoplerdquo

8122019 The Effects of Minimum Wages on SNAP Enrollments and Expenditures

httpslidepdfcomreaderfullthe-effects-of-minimum-wages-on-snap-enrollments-and-expenditures 4343

The Center for American Progress is a nonpartisan research and educational institute

dedicated to promoting a strong just and free America that ensures opportunity

for all We believe that Americans are bound together by a common commitment to

these values and we aspire to ensure that our national policies reflect these values

We work to find progressive and pragmatic solutions to significant domestic and

international problems and develop policy proposals that foster a government that

is ldquoof the people by the people and for the peoplerdquo