two steps back - hartfordinfo.org

24
December 2006 The Brookings Institution Living Cities Census Series 1 “The enduring social and fiscal challenges for cities stemming from high poverty are increasingly shared by their suburbs.” Metropolitan Policy Program Findings An analysis of poverty in cities and suburbs of the nation’s 100 largest metropolitan areas, based on data from the 2005 American Community Survey and Census 2000, indicates that: In 1999 large cities and their suburbs had nearly equal numbers of poor individu- als, but by 2005 the suburban poor outnumbered their city counterparts by at least 1 million. Still, the percentage of all people in poverty rose in both cities and suburbs between 1999 and 2005, following the national trend. In 2005, the poverty rate in large cities (18.8 percent) was twice as high as in suburbs (9.4 percent). Poverty rates rose significantly in Midwestern and Southern metropolitan areas, but remained steady in the West and Northeast. In the Midwest, where job losses were concentrated in the first half of the decade, poverty rates rose in 18 of 20 metro- politan areas. In the West, by contrast, only seven of 23 metro areas experienced poverty rate increases, and poverty actually fell in five. Nearly half of large cities nationwide saw a significant rise in their poverty rates, versus about one-third of their suburbs. Six of the ten cities with the largest poverty- rate increases were located in the Midwest, including Cleveland, Toledo, Detroit, and Columbus. New York City and the Greater Los Angeles area actually experienced small poverty-rate declines over this period. In cities and suburbs where overall poverty rates rose from 1999 to 2005, child poverty rates rose faster. In Midwestern and Southern cities, child poverty rates were up by at least 3 percentage points on average. The cities and suburbs of Houston, Dallas, and Cleveland ranked among those experiencing the greatest increases in child poverty during this period. Economic conditions during the early 2000s brought a rise in poverty nationwide and in many cities and suburbs. Regional impacts, however, have been uneven. These findings emphasize that federal and state labor market supports like the Earned Income Tax Credit and unemployment insurance can act as powerful tools for helping families suffering the effects of economic downturns. At the local level, the enduring social and fiscal challenges for cities that stem from high poverty are increasingly shared by their suburbs. Public and private efforts that give growing suburban poor populations access to economically inte- grated neighborhoods and in-work supports could enhance economic security and mobility for a significant number of Americans. Two Steps Back: City and Suburban Poverty Trends 1999–2005 Alan Berube and Elizabeth Kneebone

Upload: others

Post on 27-Jan-2022

1 views

Category:

Documents


0 download

TRANSCRIPT

December 2006 • The Brookings Institution • Living Cities Census Series 1

“The enduring

social and fiscal

challenges for

cities stemming

from high poverty

are increasingly

shared by their

suburbs.”

M e t r o p o l i t a n P o l i c y P r o g r a m

FindingsAn analysis of poverty in cities and suburbs of the nation’s 100 largest metropolitan areas,based on data from the 2005 American Community Survey and Census 2000, indicatesthat:

■ In 1999 large cities and their suburbs had nearly equal numbers of poor individu-als, but by 2005 the suburban poor outnumbered their city counterparts by at least1 million. Still, the percentage of all people in poverty rose in both cities and suburbsbetween 1999 and 2005, following the national trend. In 2005, the poverty rate in largecities (18.8 percent) was twice as high as in suburbs (9.4 percent).

■ Poverty rates rose significantly in Midwestern and Southern metropolitan areas,but remained steady in the West and Northeast. In the Midwest, where job losseswere concentrated in the first half of the decade, poverty rates rose in 18 of 20 metro-politan areas. In the West, by contrast, only seven of 23 metro areas experienced povertyrate increases, and poverty actually fell in five.

■ Nearly half of large cities nationwide saw a significant rise in their poverty rates,versus about one-third of their suburbs. Six of the ten cities with the largest poverty-rate increases were located in the Midwest, including Cleveland, Toledo, Detroit, andColumbus. New York City and the Greater Los Angeles area actually experienced smallpoverty-rate declines over this period.

■ In cities and suburbs where overall poverty rates rose from 1999 to 2005, childpoverty rates rose faster. In Midwestern and Southern cities, child poverty rates wereup by at least 3 percentage points on average. The cities and suburbs of Houston, Dallas,and Cleveland ranked among those experiencing the greatest increases in child povertyduring this period.

Economic conditions during the early 2000s brought a rise in poverty nationwide and inmany cities and suburbs. Regional impacts, however, have been uneven. These findingsemphasize that federal and state labor market supports like the Earned Income Tax Creditand unemployment insurance can act as powerful tools for helping families suffering theeffects of economic downturns. At the local level, the enduring social and fiscal challengesfor cities that stem from high poverty are increasingly shared by their suburbs. Public andprivate efforts that give growing suburban poor populations access to economically inte-grated neighborhoods and in-work supports could enhance economic security and mobilityfor a significant number of Americans.

Two Steps Back: City and Suburban Poverty Trends1999–2005Alan Berube and Elizabeth Kneebone

Introduction

The first half of the currentdecade brought economicuncertainty and hardship formany Americans. In stark

contrast to the late 1990s, whenemployment and wages were growingat historic rates, the 2000s have beenmarked by an economic recession,stagnant wages for many workers, andjob losses followed by what some havetermed a “jobless recovery.”1

These negative labor market condi-tions took a toll on workers and fami-lies near the bottom of the earningsand skills distribution. Each yearbetween 2000 and 2004, the propor-tion of the U.S. population livingbelow the poverty line—equivalent toan income of $15,753 for a family ofthree in 2005—rose. It leveled off in2005, such that 38 million U.S. resi-dents lived in poverty that year, upfrom 34 million in 1999.2

This paper demonstrates, however,that the impacts of recent economicsluggishness on poverty were unevenacross the country, and coincideclosely with regional economic pat-terns. Population and income dynam-ics meanwhile continue to shift thegeography of poverty within majormetropolitan areas of the UnitedStates. Cities and suburbs have bothexperienced rising poverty rates, butfaster population growth in suburbshas tipped the balance of poor popula-tions towards suburbs. Children havefelt the brunt of the economic down-turn and slow recovery, experiencinglarger increases in poverty than thepopulation at large, in both cities and suburbs.

After reviewing the methodology forthe analysis and the data sources used,we evaluate trends in overall urbanand suburban poverty from 1999 to2005. We then probe the regional pat-terns underlying these movements,paying particular attention to the citiesand suburbs that witnessed the great-est increases (or declines) in their

poverty rates. As part of this, we exam-ine the situation in “first suburbs,” asubset of large metropolitan suburbsthat represent older communitiesclose to the urban core. Finally, ouranalysis sheds light on child povertytrends in large cities and suburbs dur-ing the first half of the decade, extend-ing previous analysis on this subject.3

This report aims primarily todescribe recent poverty trends, andnot to explain their root causes or ulti-mate consequences. Nonetheless, thepatterns revealed indicate a great dealabout the geography of economicgrowth and decline in the currentdecade, and raise important new ques-tions and issues regarding the shiftinglocations of low-income families incontemporary metropolitan America.

Methodology

About the DataThis study uses data from the 2000U.S. Census and the 2005 AmericanCommunity Survey (ACS) to evaluatepoverty trends for the largest metro-politan areas in the United States.4

The ACS is a relatively new nation-wide survey designed to provideannual information on the demo-graphic, social, economic, and housingcharacteristics of U.S. communities. Itwill replace the “long form” in decen-nial censuses in 2010 and beyond. In2005, it was the largest household sur-vey in the nation, and samples approx-imately 3 million addresses on anannual basis.5

In August 2006, the Census Bureaupublished results from the 2005 ACSregarding a range of characteristics forgeographies with populations of65,000 or more. In comparing povertydata from the ACS to poverty datafrom Census 2000, readers shouldbear in mind two differences betweendata from these sources.

First, although both surveys pro-duce income and poverty data for thesame universe of people (individuals

in households), the time periodsreflected in those data are different.Census 2000 was administered inApril of 2000 and collected incomedata for the 1999 calendar year. TheACS surveys a different sample ofpeople each month throughout theyear, and asks recipients for informa-tion about their income in “the last 12 months”. ACS results then averagethe 12 months of income data intoone estimate for the year, adjustingfor inflation that may have occurredfrom month to month. The family’sincome is then compared to the appli-cable poverty threshold for thatperiod, and its members are identifiedas below or above poverty based onthat comparison.

Because the two surveys ask respon-dents somewhat different questionsabout their income, they generateslightly different results for the samepopulations. On this point, CensusBureau tests revealed that medianincome estimates derived from Census2000 slightly exceeded similar esti-mates from an ACS test survey con-ducted that same year. However,poverty estimates from Census 2000and the ACS test survey were statisti-cally the same. Slight variationsemerged when poverty rates were cal-culated for subgroups of individuals(e.g., poverty rates by age group).7

While these conclusions largely facili-tate the present analysis, reportedchanges in child poverty rates shouldbe viewed in the context of these find-ings regarding subgroup estimates.

Second, unlike the Census 2000long form, which surveyed roughly 17 million households nationwide, the2005 ACS surveyed about 3 millionhouseholds. This means that there isless certainty with the ACS than Cen-sus 2000 that the resulting povertyestimates represent the true values for the entire U.S. population, or thepopulation in any given sub-geography.In statistical terms, the smaller samplesize in the ACS generates a larger sampling error.

December 2006 • The Brookings Institution • Living Cities Census Series2

To appropriately reflect the magni-tude of sampling error, the CensusBureau reports margins of error along-side ACS estimates. Cited frequentlyin public opinion poll results, marginsof error suggest that a given value(e.g., the proportion of voters favoringa candidate; the proportion of individ-uals below the poverty line) fallswithin a range, rather than at a spe-cific point. Reported margins of errorin the ACS, and in this report, reflecta 90-percent confidence interval.Roughly interpreted, the 90-percentconfidence interval reflects the rangein which we have 90-percent confi-dence that the true value lies.

In each case where we compare1999 (Census 2000) to 2005 (ACS)poverty estimates, we use standard sta-tistical tests to determine whetherthose estimates are different from oneanother. For instance, the poverty rateestimate for the city of Denver in 2005was 15.3 percent, with a margin oferror of (+/-) 1.6 percentage points.Though the estimate was higher thanthe city’s 1999 poverty rate of 14.3percent, the 1999 rate fell within themargin of error of the 2005 estimate.Thus, we conclude that no changeoccurred in the Denver poverty ratefrom 1999 to 2005.8

In a couple of instances, we high-light cities or suburbs in which thelargest changes in poverty occurredfrom 1999 to 2005. Although the sam-pling errors associated with the 2005ACS make it impossible to know thetrue extent of change, we calculate aconservative measure of change toidentify these places.9

A final note: This report comparespoverty in 1999, near the peak of thelast economic cycle, to poverty in2005, whose place in the current eco-nomic cycle is still undetermined.Thus, the changes viewed here are atleast partly cyclical in nature, and wewould probably derive differentanswers if we compared city and sub-urban poverty in 2005 to that in 1995(pinpointing a similar amount of time

post-recession). However, no datasource is available to estimate city andsuburban poverty in 1995. Moreover,1999 rates provide a useful bench-mark for places that seek to achievecontinuous improvement in the eco-nomic well-being of their residents.

Geographic DefinitionsFor the purposes of this survey, weanalyze the 100 largest metropolitanareas in the United States based ontheir population in the 2000 decennialcensus, and defined as MetropolitanStatistical Areas (MSAs) by the Officeof Management and Budget (OMB) in2003. This new system makes signifi-cant updates to the metropolitan areasfor which Census 2000 data werereported.10 In order to make povertyestimates from 1999 and 2005 geo-graphically comparable, this studysummarizes 1999 poverty data(reported in Census 2000) to corre-spond to the updated metropolitanboundaries.11

The definition of central cityemployed in this report differs from,and is more restrictive than, the “prin-cipal city” concept employed by OMB.For instance, the Los Angeles-LongBeach-Santa Ana, CA MSA contains18 principal cities according to OMBguidelines.12 However, most peopleregard only the largest and most cul-turally important cities in each metroarea to be “central cities.”

To identify these more prominentcities within the 100 largest metropoli-tan areas, we use two main criteria. Acity is designated as a “central city” if:(1) it is listed first in the official met-ropolitan area name; or (2) it is listedafter the primary city in the metropoli-tan name and had a population of atleast 100,000 as of Census 2000.13

As a result, in the Los Angeles exam-ple only the three cities included inthe MSA name are designated as cen-tral cities. In the Cleveland-Elyria-Mentor, OH MSA, only Cleveland istreated as a central city, because nei-ther Elyria nor Mentor had at least

100,000 residents in 2000. We adjustmetropolitan area names to reflectonly those cities that meet our central-city criteria.14

To identify the suburbs of eachmetro area, we subtract central-citytotals from MSA totals to produce a“residual” total for metropolitan sub-urbs. Suburban data thus aggregatedata for whole counties, and portionsof counties lying outside central cities,within individual metropolitan areas.To track differences in poverty trendsacross the U.S., this study alsoassesses metropolitan poverty changesby region, using the four Censusregions: Northeast, Midwest, South,and West.15

A portion of our analysis examinespoverty levels and trends in first sub-urbs based on recent Brookingsresearch that defines these areas.16

These first suburbs lay adjacent tomajor cities in 1950—as either fullcounties or the county remainderwhen the city was excluded—and werepart of metropolitan areas defined atthat time. We analyze these same firstsuburbs, with the exception of threegeographies that are not part of the100 largest metro areas in 2000.17

The Poverty RateThis study details poverty rates forindividuals in the top 100 metropoli-tan areas as well as the central citiesand suburbs located within them. Thepoverty rate represents the proportionof family members and unrelated indi-viduals in a particular place withincomes below the applicable federalpoverty threshold.18 In addition to eval-uating poverty rates for all individualsin the selected geographies, this studyalso examines child poverty rates.Child poverty rates represent the shareof family members and unrelated indi-viduals under 18 in families withincomes below the federal povertylevel.19

Poverty as a statistical concept isoutdated, as most researchers and theAmerican public realize.20 It fails to

December 2006 • The Brookings Institution • Living Cities Census Series 3

consider the level and full range ofexpenses that working families incurtoday, while ignoring some of theimportant efforts that governmentmakes to alleviate poverty (such as theEarned Income Tax Credit). Nor doesit reflect the relative position of thepoor in American society as livingstandards have grown over the past 40years.21 Yet the stability of this measureover time does provide our nation, andits states and cities, with a useful yard-stick that measures the ranks of ourmost economically disadvantaged.

Findings

A. In 1999 large cities and their sub-urbs had nearly equal numbers ofpoor individuals, but by 2005 thesuburban poor outnumbered theircity counterparts by at least 1 million.Traditionally, cities have been viewedas home to poor populations, sur-rounded by middle- and upper-incomesuburbs.

This characterization continues tohold true in certain corners of theUnited States. Increasingly, however, itfails to take account of the rapidgrowth and diversification occurring inAmerican suburbs. It also overlooksthe rise of “Sunbelt” metropolitanareas where the socioeconomic dis-tinctions between cities and suburbshave never loomed as large as they stilldo in older areas of the Northeast andMidwest.

The 1999-to-2005 period, whichbrought economic recession and aslow recovery for workers near the bot-tom, saw a rise in poverty nationwide.At least 4 million more people livedbelow the poverty line in the UnitedStates in 2005 than in 1999 (Figure1). This growth in the poor populationpushed the nation’s poverty rate up byabout 1 percentage point, to 13.3 per-cent in 2005 (Figure 2).

Neither cities nor suburbs wereimmune to the national trend of grow-ing poor populations over the first half

of the 2000s. As Figure 1 shows, the100 largest metropolitan areastogether accounted for about 60 per-cent of the nationwide uptick in thismeasure over the 6-year period, andwere home to roughly 23 million poorin 2005.

Especially notable, though, is theshifting location of the poor withinlarge metropolitan areas. In 1999,these individuals split almost evenlybetween central cities (10.4 million)and their suburbs (10.3 million). Thetotal poor population rose in each geo-graphic category from 1999 to 2005,but at a much faster clip in suburbs.Thus, by 2005, the suburban poor out-numbered their central-city counter-parts by at least 1 million. Together,the poor in suburbs accounted forabout 53 percent of large-metropolitanpoor that year.22

This “tipping” of poor populationsto the suburbs represents a signaldevelopment that upends historicalnotions about who lives in cities and

suburbs. As previous research has sug-gested, the poor—and especially theworking poor—figure prominentlyamong suburban populations today(Berube and Frey, 2002; Berube,2004).23 Subsequent sections of thispaper explore the complicated regionaland intra-metropolitan dynamics thatunderlie this macro-level shift.

Considerably faster growth in thesize of the suburban poor, however,has not altered the fact that city resi-dents on average are much more likelyto be poor than suburban residents. Infact, the overall poverty-rate change incities from 1999 to 2005 mirrored thatin their suburbs, and further mim-icked the national change during thattime (Figure 2). Today, the combinedpoverty rate for all large central cities(18.8 percent) doubles that for theirsuburbs (9.4 percent). Thus, whilepoverty is now more suburban thanurban in one important respect, citiesstill claim a disproportionate share ofthe metropolitan poor.

December 2006 • The Brookings Institution • Living Cities Census Series4

Source: Brookings Institution analysis of Census 2000 and 2005 American Community Survey

Figure 1. Population in Poverty, United States,100 Largest Metropolitan Areas, and their Cities and Suburbs,

1999 and 2005

0

5,000

10,000

15,000

20,000

25,000

30,000

35,000

40,000

2005

1999

Largest Metros-SuburbsLargest Metros-Central Cities100 Largest MetrosNation

Popu

latio

n in

pov

erty

(th

ousa

nds)

33,900

38,232

+/-297

20,629

23,195

+/-152

10,380 10,958

+/-10810,250

12,236

+/-187

B. Poverty rates rose significantly inMidwestern and Southern metropol-itan areas, but remained steady inthe West and actually fell slightly inNortheastern cities.The parallel increases in national,metropolitan, city, and suburbanpoverty rates in recent years mask con-siderable variation across differentregions of the United States. This sec-tion examines recent metropolitanpoverty changes that have occurredwithin these geographies in each ofthe country’s major regions.

Why might region matter? Forstarters, the recent downturn andongoing restructuring of the nationaleconomy have affected some industriesmore than others, with attendant con-sequences for areas that depend onthose industries. For instance, from2000 to 2005, the national economyshed roughly 3 million jobs in manu-facturing and 400,000 in telecommu-nications and internet data processing.Meanwhile, employment in finance/

real estate and government continuedto expand. Differing regional special-izations (e.g., manufacturing in theRust Belt, internet in the Bay Area,finance in New York) thus shape locallabor market conditions. Lower-skilledworkers at risk of poverty need nothave worked, or sought work, in theseparticular industries to have experi-enced the downstream effects of thesedeclines or increases in sectors likeretail or hospitality.

Migration flows between and withinregions can also affect poverty rates, ifthe incomes of those moving in (orout) differ from the incomes of thosemoving out (or in). Several observershave documented a continuing move-ment of middle-class families toSouthern metropolitan areas likeAtlanta, Charlotte, and Orlando, par-ticularly from the North.24 High hous-ing prices on the West Coast,meanwhile, have spurred an inwardmigration of middle-income homebuy-ers in that region.25 These movements

could affect poverty rates in both ori-gin and destination areas. At the sametime, immigration is spreading to awider set of metropolitan areas than indecades past, especially the Southeast,which could alter poverty levels inthose receiving areas.26 Thus, a risingor falling poverty rate in a particularplace may indicate not only changes inthe economic status of existing resi-dents, but changes in the underlyingresident population over time.

Figure 3 depicts the regional povertytrends that coincided with these eco-nomic and demographic shifts. The pic-ture in the Midwest and South looksquite different from that in the North-east and West. In the former regions,poverty rates rose in both cities andsuburbs from 1999 to 2005, and by asomewhat greater degree in their cities.Poverty rates meanwhile held steady inthe West. Northeastern cities actuallywitnessed a small decline in the propor-tion of their residents in poverty—though this largely reflected theoutsized influence of New York City—while suburbs in that region saw a mar-ginal increase.27 Interestingly, whereaspoverty rates in Northeastern cities andsuburbs were historically higher than inthe Midwest, by 2005 the measureswere statistically equivalent betweenthe regions.

One further regional distinction isclear: poverty-rate differences betweencities and suburbs in the West (and toa lesser degree, the South) remainquite muted compared to those in theNortheast and Midwest. Even with thesmall narrowing of the city-suburbanpoverty gap in the Northeast from1999 to 2005, within-metropolitandisparities in these regions clearly out-strip those in the Sunbelt. Amongother factors, this owes to the more“hemmed in” geography of northerncities that inhibits their ability toannex higher-income suburbs; thedeeper legacy of racial and economicdiscrimination in their housing mar-kets; and the heavier out-migration ofmiddle-class families that most

December 2006 • The Brookings Institution • Living Cities Census Series 5

Source: Brookings Institution analysis of Census 2000 and 2005 American Community Survey

Figures in parentheses represent the difference betweenn 19999 rate and 2005 lower-bound

estimate

Figure 2. Poverty Rates for United States,100 Largest Metropolitan Areas, and their Cities and Suburbs,

1999 and 2005

0.0

2.0

4.0

6.0

8.0

10.0

12.0

14.0

16.0

18.0

20.0

(0.7)(0.7)(0.7)(0.8)Largest Metros-SuburbsLargest Metros-Central Cities100 Largest MetrosNation

Perc

ent

in P

over

ty

12.413.3

11.612.3

17.918.8

8.59.4

2005

1999

endured in the 1970s and 1980s.28

The differences by region look evenstarker through the lens of povertytrends in individual metropolitan areas(Figure 4). A little over half—54—ofthe 100 largest metro areas saw theirpoverty rates increase from 1999 to2005, while just seven experienced adecline (changes in the remaining 39areas were statistically insignificant).The Midwest, especially, stands out forits widespread poverty-rate increasesduring that time; the indicator rose infully 18 of its 20 large metropolitanareas. Poverty rates also rose in themajority of large Southern metro areas(21 of 37). By contrast, only seven of23 Western metro areas analyzed sawpoverty increases. Moreover, thatregion contained five of the sevenmetro areas in which overall povertyrates declined. Notably, the 1999-to-2005 regional trend inverts the patternfrom the prior 10 years. In thatdecade, Midwestern and Southerncities and suburbs saw significantdeclines in poverty, while those in theNortheast and California experiencedincreases.29

C. Nearly half of large cities nation-wide saw a significant rise in theirpoverty rates, versus about one-thirdof their suburbs.Big cities experienced mixed results onpoverty across the 1990s. One studyfound that a bare majority of largecities saw poverty rates decline overthe decade, while a little more thanone-third experienced poverty-rateincreases. Results varied for their sub-urbs as well; in about 80 percent ofmetro areas, poverty rates betweencities and suburbs moved in tandem.30

Using results from the ACS intro-duces somewhat greater uncertaintyabout the magnitude and direction ofchange in poverty during the currentdecade, especially for smaller geogra-phies where ACS samples are smaller,and sampling errors are larger.

Nonetheless, we find that povertyrates increased between 1999 and

December 2006 • The Brookings Institution • Living Cities Census Series6

Source: Brookings Institution analysis of Census 2000 and 2005 American Community Survey

Figures in parentheses represent the difference between 1999 rate and 2005 lower-bound estimate

Figure 3. Central-City and Suburban Poverty Rates,100 Largest Metro Areas by Region, 1999 and 2005

(-0.4) (0.3) (2.3) (1.4) (1.6) (1.0) (***) (***)0.0

5.0

10.0

15.0

20.0

25.0

WestSouthMidwestNortheast

Perc

ent

in P

over

ty

21.720.9

7.3 7.9

18.0

20.8

6.2

7.9

17.3

19.2

9.6

10.8

15.8 15.9

10.6 10.4

Central Cities 2005 Suburbs 1999 Suburbs 2005Central Cities 1999

Source: Brookings Institution analysis of Census 2000 and 2005 American Community Survey

Figure 4. Change in Poverty Rates, 100 Largest Metro Areas byRegion, 1999 to 2005

Share of Metro Areas0%

100 Largest Metros

Northeast

Midwest

South

West

10% 20% 30% 40% 50% 60%

Increase Decrease No Change

70% 80% 90% 100%

2005 by a statistically significant mar-gin in 45 of 94 central cities analyzed.In a nearly equal number of cities(43), poverty rates in 2005 were nodifferent statistically than in 1999.31

Only six central cities exhibited a sig-nificant decline in poverty during thisperiod. (Poverty rate trends for all 100metropolitan areas, and their centralcities and suburbs, are displayed inAppendix A.)

As the regional analysis in the previ-ous section suggests, cities experienc-ing poverty-rate increases clustered incertain areas of the country. Map 1shows that among the eleven citiesexperiencing the greatest increases inpoverty rates during this period, six arelocated in the Midwest, includingthree in Ohio and two in Michigan.32

This probably reflects the dispropor-tionate impact of the recent economicdownturn on the manufacturing sec-tor, which figures prominently in boththese states. Between 2000 and 2005,Michigan and Ohio each lost over200,000 manufacturing jobs, equiva-lent to more than 20 percent of alljobs in that sector in each state. Inaddition, three Texas cities rankamong the top ten; all experiencedincreases in unemployment from 2000to 2003, though migration dynamicsin these rapidly growing areas may also have contributed to their poverty-rate climbs.

On the suburban side, poverty-ratechanges tended to be smaller, and1999 poverty rates fell within the mar-gin of error for the 2005 estimate in67 of the 100 suburbs analyzed, indi-cating no statistically significantchange. Among the 33 suburbs wherepoverty rates did change, fully 31experienced an increase, while onlytwo saw their poverty rates decline.Those suburbs experiencing the great-est increases in poverty rates, shownin Map 2, overlap to a great degreewith the central cities that sawincreases. Cleveland, Dallas, Detroit,and Portland (OR) appear on both thecity and suburban “top ten” lists. One

December 2006 • The Brookings Institution • Living Cities Census Series 7

Source: Brookings Institution analysis of Census 2000 and 2005 American Community Survey

Map 1. Central Cities with the Greatest Increases and Decreasesin Poverty Rates, 1999 to 2005

●●●

●●

● ●

● ●

●●

Increases in Poverty

Decreases in Poverty

Portland-Vancouver, OR-WA

Stockton, CA

Los Angeles-Long Beach-Santa Ana, CA

El Paso, TXDallas-Fort Worth-Arlington, TX

Houston, TXNew Orleans-Metairie, LA

Richmond, VA

Washington-Arlington-Alexandria, DC-VA-MD

New York-NorthernNew Jersey, NY-NJ-PA

Omaha, NE

Grand Rapids, MILansing, MIDetroit-Warren-Livonia, MIToledo, OHCleveland, OHColumbus, OH

1 2 3

45

6

1.2.3.4.5.6.

Source: Brookings Institution analysis of Census 2000 and 2005 American Community Survey

Map 2. Suburbs with the Greatest Increases and Decreases inPoverty Rates, 1999 to 2005

●●

●●

●● Increases in Poverty

Decreases inn Poverty

Portland-Vancouver, OR-WA

Los Angeles-Long Beach-Santa Ana, CA

Riverside-San Bernadino-Ontario, CA

Dallas-Fort Worth-Arlingon, TX

McAllen, TX

Atlanta, GA

Greeensboro, NC

Cincinnati, OH-KY-IN

Dayton, OH

Cleveland, OH

Detroit-Warren-Livonia, MI

Portland, ME

difference is the appearance of twoSoutheastern metro areas—Atlantaand Greensboro—on the suburbanroster. In-migration to Atlanta’s sub-urbs from abroad, particularly by LatinAmericans working in construction orseasonal industries, may have con-tributed to increases in workingpoverty in that region’s suburbs.33

As indicated above, only six citiesand two suburbs experienced statisti-cally significant declines in povertyrates from 1999 to 2005 (Maps 1 and2). Their locations indicate thatSouthern California and the MiddleAtlantic seem to have weathered theearly part of the decade well economi-cally. Continued growth in financialservices and government employmenthas benefited New York and Washing-ton, while expanding construction andprofessional/business services employ-ment seems to have buoyed Los Ange-les and its environs throughout theearly 2000s (PPIC 2004).34

The 1999-to-2005 trends providefurther evidence that metropolitaneconomic health is a key determinantof poverty dynamics. Poverty ratestended to move in the same directionbetween cities and suburbs in thesame metropolitan areas. Among the31 suburbs in which poverty ratesincreased by a statistically significantmargin, 20 had central cities that also underwent increases (changeswere statistically insignificant in theother 11).35

Where did the general rise inpoverty in both cities and suburbsleave these places in 2005? Figure 5shows the top and bottom cities andsuburbs by their estimated 2005poverty rate, along with the margin oferror associated with those estimates.A clear Northeast/West divide sepa-rates cities at the top of the povertylist from those at the bottom. Many ofthe high-poverty cities are mid-sizedformer industrial centers of the North-east still dealing with the aftermath ofsignificant job loss, high degrees ofracial and economic segregation, and

middle-class out-migration. Low-poverty Western cities tend to be moregeographically expansive, economicallybuoyant, and in some cases (e.g., SanJose, Santa Rosa, Honolulu) much tooexpensive for poor families to livewithin their borders.

Suburban poverty rankings reflect aregional mirror image of the centralcity rankings. Here, Western andSouthern metro areas that still dependon agriculture—and in many cases,migrant-worker populations—containsignificant poverty in their suburbs.36

Meanwhile, Northeastern and Mid-western suburbs rank among thosewith the lowest poverty rates. Some ofthese are located in economically pros-perous metro areas (e.g., Omaha,Washington, Minneapolis), but thosein other metropolitan areas (e.g., Hart-ford, Allentown, Milwaukee, Bridge-port) have grown and developedprimarily to accommodate middle- andhigher-income homeowners, whiletheir central cities continue to housedisproportionate shares of poor, oftenminority households.37

December 2006 • The Brookings Institution • Living Cities Census Series8

Source: Brookings Institution Analysis of American Community Survey data

Upper and lower bounds represent 90-percent confidence interval for poverty rate estimate

Figure 5. Central Cities and Suburbs with Highest and LowestPoverty Rates, 100 Largest Metro Areas, 2005

Suburbs

Central Cities

Perc

ent

in P

over

tyPe

rcen

t in

Pov

erty

Upper Bound Lower Bound Poverty Rate

Upper Bound Lower Bound Poverty Rate

McA

llen

, TX

El P

aso,

TX

Bak

ersf

ield

, CA

Fres

no,

CA

Alb

uq

uer

qu

e, N

M

Mo

des

to, C

A

New

Orl

ean

s-M

etai

rie, LA

Au

gu

sta-

Ric

hm

on

dC

ou

nty

, GA

-SC

Gre

ensb

oro

, NC

Lake

lan

d, F

L

Om

aha,

NE-

IA

Har

tfo

rd, C

T

Mad

iso

n, W

I

Alle

nto

wn

, PA

-NJ

Was

hin

gto

n-A

rlin

gto

nA

lexa

nd

ria,

DC

-VA

-MD

Bal

tim

ore

, MD

Min

nea

po

lis-S

t. Pa

ul,

MN

-WI

Milw

auke

e, W

I

Bri

dg

epo

rt-S

tam

ford

, CT

Des

Mo

ines

, IA

Cle

vela

nd,

OH

Har

tfo

rd, C

T

Syra

cuse

, NY

Roch

este

r, N

Y

Bat

on

Ro

ug

e, L

A

Spri

ng

field

, MA

Prov

iden

ce, R

I-M

A

Bir

min

gh

am, A

L

Day

ton

, OH

New

Hav

en, C

T

Oxn

ard

-Th

ou

san

d O

aks-

Ven

tura

, CA

Jack

sonv

ille,

FL

Ho

no

lulu

, HI

Bo

ise

Cit

y, ID

Co

lora

do

Sp

rin

gs,

CO

Vir

gin

ia B

each

-No

rfo

lk-

New

po

rt N

ews,

VA-N

C

San

ta R

osa

, CA

Palm

Bay

, FL

Des

Mo

ines

, IA

San

Jo

se-S

un

nyv

ale-

San

ta C

lar,

CA

40.0

35.0

30.0

25.0

20.0

15.0

10.0

5.0

0.0

60.0

50.0

40.0

30.0

20.0

10.0

0.0

To be sure, suburbs themselves rep-resent a diverse lot. In particular,recent research has drawn attention tothe distinct profile of older, inner-ring“first” suburbs that have a more urbanfeel than their newer suburban coun-terparts. Puentes and Warren definethese first suburbs as counties, or por-tions of counties, that surround majorcentral cities and were part of metro-politan areas in 1950. Of the 100 met-ropolitan areas in this analysis, 40contained first suburbs—a total of 61counties (or county remainders net ofcentral cities).

Across all these first suburbs, wefind that the poverty rate rose onlyslightly from 1999 to 2005. Theircombined estimated rate in 2005 was9.3 percent, compared to 8.6 percentin 1999.38 Forty (40) of the 61 firstsuburbs did not experience a statisti-cally significant change in theirpoverty rates during this time, while18 experienced an increase. Thelargest significant increases appearedto occur in Midwestern industrial sub-urbs such as Lake County, IN (outsideChicago); Macomb County, MI (out-side Detroit), Cuyahoga County, OH(outside Cleveland); and MontgomeryCounty, OH (outside Dayton) (Table

1). Not coincidentally, their associatedcities saw poverty increases as well.(Poverty rate trends for all 61 first sub-urbs are displayed in Appendix C.)

Overall, though, first suburbs didnot bear the brunt of increasing subur-ban poverty in the early 2000s.Indeed, their share of the suburbanpoor fell modestly, from 42 percent to40 percent.39 This trend emphasizesthat though the poor remain morelikely to live in and around the urbancore of metropolitan areas, the contin-ued decentralization of metropolitanpopulations to second-tier suburbs and“exurbs” has included low-incomeworkers and their children as well.

D. In cities and suburbs where over-all poverty rates rose from 1999 to2005, child poverty rates rose faster.Children are more likely to live inpoverty than adults. Since the CensusBureau began reporting poverty ratesin the 1960s, the child poverty ratehas typically exceeded that for theU.S. population as a whole by any-where from 3 to 7 percentage points.

Likewise, when economic condi-tions take a turn for the worse, children tend to feel the effects dis-proportionately. Many live in single-

earner households where job orincome losses can plunge a family intopoverty. Large and growing numbersare the children of immigrants, manyof whom have limited formal educa-tion and may be particularly suscepti-ble to income losses during leaneconomic times.40

In 2005, as in 1999, a little overone-third of all poor Americans wereunder age 18. The estimated U.S.child poverty rate during that periodrose from 16.6 percent to 18.5 per-cent.41 As with overall poverty, how-ever, the changes varied substantiallyby region (Figure 6). Child povertyrates rose considerably in Midwesterncities and suburbs, and by smallerthough still-significant amounts in theSouth. In other parts of the nation,child poverty held fairly steady(though again, New York heavily influ-enced aggregate figures for the North-east). Where child poverty-rateincreases did occur, they were on aver-age larger than those occurring for thetotal population. The child povertyrate in Southern cities, for instance,rose by at least 3.4 percentage points,compared to 1.6 percentage points fortheir overall populations. As a result,the already-wide gap between overall

December 2006 • The Brookings Institution • Living Cities Census Series 9

Table 1. First Suburbs Experiencing Greatest Poverty-Rate Increases, 1999 to 2005

Poverty Rate Poverty Rate Margin of Error Change First Suburbs* Metropolitan Area 1999 2005 2005 1999–2005

Lake, IN Chicago-Naperville-Joliet, IL-IN-WI 12.2 16.4 1.6 2.6

Macomb, MI Detroit-Warren-Livonia, MI 5.6 8.6 0.8 2.1

Cuyahoga, OH Cleveland, OH 6.3 9.7 1.8 1.6

Montgomery, OH Dayton, OH 6.6 10.0 2.4 1.0

Camden, NJ Philadelphia, PA-NJ-DE 10.4 12.4 1.3 0.7

Dallas, TX Dallas-Fort Worth-Arlington, TX 8.4 11.6 2.6 0.6

St. Louis, MO St. Louis, MO-IL 6.9 8.2 0.8 0.5

Hennepin, MN Minneapolis-St. Paul, MN-WI 3.9 6.0 1.6 0.5

Harris, TX Houston, TX 9.3 12.2 2.4 0.5

Delaware, PA Philadelphia, PA-NJ-DE 8.0 9.7 1.2 0.4

Source: Brookings Institution analysis of Census 2000 and 2005 American Community Survey data

* First suburbs represent counties and county remainders

** Change from 1999 to 2005 represents difference between 1999 rate and 2005 lower-bound estimate

and child poverty rates (Figures 3 and6), especially in cities, grew evenlarger by 2005.

The child disadvantage on urbanand suburban poverty trends is fur-ther evident from a comparison ofchanges in specific cities and sub-urbs. Table 2 shows the 10 cities and10 suburbs experiencing the largeststatistically significant increases inchild poverty rates from 1999 to2005. With few exceptions, rises intheir proportion of children livingbelow the poverty line exceed similarrises in their overall poverty rates. InHouston, for example, the proportionof children in poverty escalated to anestimated 35 percent in 2005, up atleast 6 percentage points from its1999 rate, and far outpacing the city’scoincident rise in overall poverty.4

(Child poverty rate trends for all 100metropolitan areas, and their centralcities and suburbs, are displayed inAppendix B.)

How “child-centric” the poor popu-lation is in different cities, on theother hand, depends both on thepropensity of children to live in poorfamilies, and on overall populationcharacteristics. Though estimates onthis indicator are less statistically pre-cise, many cities do have high propor-tions of children among their poor(Figure 7), including those with largeHispanic and black populations—groups that are more likely thanwhites to have young children. Thecities near the bottom of the list aremore likely to have students and youngpeople who may be temporarily poor(e.g., Boston, Knoxville, Madison) oraging populations experiencing poverty(e.g. Palm Bay, Pittsburgh).

Conclusion

Poverty has increased nation-ally and in most communitiessince 1999, as results fromthe 2005 American Commu-

nity Survey make clear. The proportion

December 2006 • The Brookings Institution • Living Cities Census Series10

Source: Brookings Institution analysis of Census 2000 and 2005 American Community Survey

Figures in parentheses represent the difference between 1999 rate and 2005 lower-bound estimate

Figure 6. Central-City and Suburban Child Poverty Rates,100 Largest Metro Areas by Region,

1999 to 2005

(***) (0.5) (3.8) (2.1) (3.4) (1.7) (***) (***)0.0

5.0

10.0

15.0

20.0

25.0

30.0

35.0

WestSouthMidwestNortheast WestSouthMidwestNortheast

Perc

ent

in P

over

ty

30.7 30.5

9.310.4

25.8

30.5

7.8

10.5

24.7

28.8

12.6

14.8

22.1 22.8

14.2 14.1

Central Cities 2005 Suburbs 1999 Suburbs 2005Central Cities 1999

Source: Brookings Institution analysis of Census 2000 and 2005 American Community Survey

Upper and lower bounds represent 90-percent confidence intervals for 2005 estimates

Figure 7. Central Cities with Highest and Lowest Proportions ofPoor Population Under Age 18, 100 Largest Metro Areas, 2005

70.0

60.0

50.0

40.0

30.0

20.0

10.0

0.0

Perc

ent

in P

over

ty

Upper Bound Lower Bound Poverty Rate

Au

gu

sta-

Ric

hm

on

d C

ou

nty

,G

A-S

C

Fres

no,

CA

Bak

ersf

ield

, CA

Oxn

ard

-Th

ou

san

d O

aks-

Ven

tura

, CA

Vir

gin

ia B

each

-No

rfo

lk-

New

po

rt N

ews,

VA-N

C

Bir

min

gh

am, A

L

Milw

auke

e, W

I

Lake

lan

d, F

L

Alle

nto

wn

, PA

-NJ

El P

aso,

TX

Bo

ise

Cit

y, ID

Bo

sto

n-C

amb

rid

ge,

MA

-NH

Palm

Bay

, FL

Co

lum

bia

, SC

Pitt

sbu

rgh

, PA

Kn

oxvi

lle, T

N

San

Fra

nci

sco

-Oak

lan

d-

Frem

on

t, C

A

Ho

no

lulu

, HI

Seat

tle-

Taco

ma-

Bel

levu

e,W

A

Mad

iso

n, W

I

December 2006 • The Brookings Institution • Living Cities Census Series 1 1

of children in poverty has risen, too,eroding much of the progress made inthe 1990s. Many cities witnessedincreases in poverty over this period,but the most striking developmentshave affected suburbs, which for thefirst time contain a majority of thenation’s poor population. These trendsare in large part the function of the2001-02 recession and slow wagegrowth thereafter for lower-skilledworkers.

Labor Market Conditions andPovertyThe regional variation in povertytrends from 1999 to 2005 highlightthe important role of regional eco-nomic performance in lifting, andkeeping, families and children out of

poverty. Just as the recession differen-tially affected certain industries andareas of the country, so too werelower-income Americans made worseoff in some places than others.Throughout the Midwest and in mostparts of the South, poverty rates rosein the early part of this decade. Else-where they were mostly stable, thoughvariation existed even within regions(e.g., Los Angeles versus San Fran-cisco; New York vs. Boston.)

Families with children, especiallythose with a single earner, are amongthe groups most likely to live inpoverty in the United States.43 With anunprecedented number of female fam-ily heads in the labor force today, therise in poverty within many U.S.regions in response to weak employ-

ment conditions suggests a need forenhanced labor market insurance poli-cies to help dislocated workers andtheir families weather difficult times.The Earned Income Tax Credit (EITC)is one such policy, as its valueincreases to compensate many lower-income workers who lose hours or areforced to take lower-paying work.44

The number of families claiming theEITC rose quite significantly in theearly 2000s, especially in economicallyhard-hit areas of the country likeDetroit, Cleveland, Memphis, andBaton Rouge. An expanded credit forworkers who maintain at least a mod-est level of earnings could bolster theEITC’s effectiveness in this regard.

Other labor-market policies couldhelp working families from falling into

Table 2. Central Cities and Suburbs with Greatest Increases in Child Poverty Rate, 100 Largest Metro Areas, 1999 to 2005

Child Poverty Rate Overall Poverty RateCentral Cities 2005 Margin of Error Change from 1999* 2005 Margin of Error Change from 1999*

Baton Rouge, LA 47.1 8.1 7.3 29.6 3.8 1.9

Portland-Vancouver, OR-WA 26.2 3.1 6.5 17.6 1.3 3.4

Houston, TX 35.0 2.5 6.2 22.9 1.3 2.5

Lansing, MI 37.2 8.0 5.6 24.4 3.8 3.7

Detroit-Warren-Livonia, MI 38.8 3.0 5.4 26.1 1.6 2.6

Cleveland, OH 47.6 4.3 5.4 32.4 2.3 3.9

Birmingham, AL 49.5 8.6 5.3 28.9 4.0 0.2

Austin, TX 25.7 3.6 5.1 18.1 1.6 2.1

Columbus, OH 27.2 3.0 5.1 18.5 1.3 2.4

Dallas-Fort Worth-Arlington, TX 29.1 2.1 4.7 19.7 1.1 2.6

Suburbs 2005 Margin of Error Change from 1999* 2005 Margin of Error Change from 1999*

Dayton, OH 15.3 3.7 2.8 9.6 1.5 1.1

Greensboro, NC 21.1 5.5 2.5 14.4 2.6 2.1

Houston, TX 17.3 2.5 2.4 12.1 1.4 0.9

Atlanta, GA 13.6 1.4 2.4 10.0 0.6 1.4

Cincinnati, OH-KY-IN 13.2 2.0 2.3 9.4 0.9 1.4

Cleveland, OH 13.1 2.4 2.2 9.2 1.0 1.9

McAllen, TX 55.0 4.2 2.2 43.9 3.0 2.2

Greenville, SC 18.4 2.6 1.5 13.3 1.3 0.4

Detroit-Warren-Livonia, MI 11.6 1.9 1.3 8.4 0.9 1.1

Dallas-Fort Worth-Arlington, TX 12.1 1.8 1.2 9.3 1.0 1.0

Source: Brookings Institution analysis of Census 2000 and 2005 American Community Survey

* Change from 1999 to 2005 represents difference between 1999 rate and 2005 lower-bound estimate

poverty. For instance, only about one-third of the unemployed receivedUnemployment Insurance in 2004, andrates of receipt among low-incomeworking parents are even lower. Certainstate policy decisions, such as deter-mining eligibility based on more recentwork history, and offering benefits toworkers who separated from anemployer due to a good personal reason(such as caring for a sick relative),could improve UI coverage.45 Mean-while, workers in locations experiencingmanufacturing employment declineswould benefit from a combination oftargeted economic, human services,and workforce development interven-tions that lower employers’ costs, raiseproductivity, and provide workers withneeded skills and supports.46

The Changing Location of the PoorThe other major highlight of thisanalysis concerns the continued intra-metropolitan shift in the location ofpoor families. The fact that a majorityof poor individuals in metropolitanareas now live in suburbs signals thelatest stage in the long-run decentral-ization of people and jobs in theUnited States, as well as the growth oflow-wage workers in areas of theSouth and West where city/suburbanincome divides are not as stark. To theextent that this development reflectsan “opening up” of housing andemployment opportunities for low-income families in suburbia, and alessening of overall economic segrega-tion, that is a welcome change.

Yet not all suburbs are createdequally. In many major metropolitanareas of the United States, growth—ofincomes, jobs, and housing wealth—concentrates on one side of the region,while economic stagnation prevails onthe other side. In the Memphis, St.Louis, and Washington, DC regions,this split occurs along an east/westaxis. In Atlanta, Chicago, and Dallas,growth occurs to the north and west,while southern and eastern suburbs

struggle with rising poverty.47 Thespread of poor households from centralcities into the suburbs of these dividedregions does not necessarily enhancetheir access to quality housing, educa-tion, or employment opportunities.

To shrink these regional divides, for-ward-looking state and local leadersare using inclusionary zoning, afford-able housing trust funds, comprehen-sive housing plans, and other policytools that secure residential opportuni-ties for lower-income working familiesin areas of new growth and develop-ment.48 Meanwhile, a growing numberof workforce intermediaries play a crit-ical role in connecting lower-wageworkers and employers to meet metro-politan-wide labor demands.49

The suburbanization of poverty alsopresents new challenges for efforts tolink low-income households to worksupports—especially as the poor movebeyond urbanized “first suburbs” tomore far-flung mature and emergingsuburbs, as this report suggests. Someresearch finds that, for immigrantgroups, take-up of the EITC is lowerin low-density suburban jurisdictionsthan in high-density city neighbor-hoods.50 For the most part, social services providers remain dispropor-tionately located in central-city neigh-borhoods, and lag the movement ofimportant parts of their client base tothe suburbs.51 These shifts make iteven more critical that municipal cam-paigns to connect low-income workersto key public benefits like tax credits,nutritional supports, health insurance,and child care adopt a region-wideapproach that reaches suburban aswell as city families. Similarly, localand regional non-profits should seekopportunities to re-align their servicedelivery to accommodate the changinglocation of families in need.

Tracking Urban and SuburbanPovertyFinally, this study demonstrates theutility of the new American Commu-nity Survey in tracking the well-being

of smaller areas over time. Before itsadvent, researchers were able to exam-ine demographic, social, and economicdata for cities and suburbs only onceper decade, via results from the decen-nial census. With the full-scale ACS,these statistics can be tracked annually,and can inform decision making in thepublic, private, and non-profit spheres.

At the same time, our analysisexposes some of the limitations ofannual ACS data for smaller geogra-phies. The survey’s one-year sample ina mid-sized major-metro city like NewHaven, CT, for instance, yields a sig-nificant margin of error (+/- 4.5 per-centage points) in the poverty estimatethat complicates trend analysis. Begin-ning in 2008, the Census Bureau willpublish three-year averages of ACSdata for geographies of 20,000 ormore people. The combination of datafrom across years will add precision tothe published ACS estimates, andoffer communities a clearer picture oftheir socioeconomic trajectory. Whilemost local leaders will not be able torely on point-in-time estimates ofpoverty and other economic indica-tors, even annual three-year averageswill afford a dramatic improvementover once-a-decade statistics from the decennial census for planning and policy.

How lower-income families andcommunities fare in the latter half ofthe 2000s remains to be seen. The2005-to-2009 trend is not predeter-mined, of course, and this decademight yet see strong economic gains inregions that suffered the greatestlosses in the first half. Whether futuregains might reach down to the bottomof the labor market and lift workersout of poverty, as in the late 1990s, isan open question as well. With assis-tance from the ACS, however,researchers can now closely monitorthe condition of people and placesover time, and assess our collectiveprogress in addressing the challengesfaced by our nation’s most economi-cally disadvantaged.

December 2006 • The Brookings Institution • Living Cities Census Series12

December 2006 • The Brookings Institution • Living Cities Census Series 13

App

endi

x A

. Met

ropo

lita

n, C

entr

al-C

ity,

and

Sub

urba

n P

over

ty R

ates

,10

0 L

arge

st M

etro

Are

as, 1

999

to 2

005

MS

AC

entr

al C

ity

Sub

urbs

1999

1999

1999

to 2

005

to 2

005

to 2

005

Min

imum

Min

imum

Min

imum

1999

2005

MO

EC

hang

e19

9920

05M

OE

Cha

nge

1999

2005

MO

EC

hang

e

NA

TIO

N12

.413

.30.

10.

8

100

LA

RG

ES

T M

ET

RO

AR

EA

S11

.612

.30.

10.

717

.918

.80.

20.

78.

59.

40.

10.

7

MID

WE

ST

(20

Met

ros)

9.8

11.6

0.2

1.6

18.0

20.8

0.4

2.3

6.2

7.9

0.3

1.4

Akr

on, O

H9.

811

.31.

20.

317

.520

.12.

40.

36.

37.

62.

0

Chi

cago

-Nap

ervi

lle-J

olie

t, I

L-I

N-W

I10

.511

.80.

40.

918

.619

.91.

00.

36.

27.

90.

80.

9

Cin

cinn

ati,

OH

-KY-

IN9.

511

.60.

71.

421

.925

.02.

80.

37.

19.

40.

91.

4

Cle

vela

nd, O

H10

.813

.80.

62.

526

.332

.42.

33.

96.

39.

21.

01.

9

Col

umbu

s, O

H9.

912

.10.

71.

514

.818

.51.

32.

46.

07.

61.

5

Day

ton,

OH

10.0

12.7

1.1

1.6

23.0

28.9

4.0

2.0

7.0

9.6

1.5

1.1

Des

Moi

nes,

IA

7.5

6.1

1.0

-0.4

11.4

9.9

2.3

4.8

3.7

2.2

Det

roit

-War

ren-

Liv

onia

, MI

10.6

12.7

0.5

1.6

21.9

26.1

1.6

2.6

6.5

8.4

0.9

1.1

Gra

nd R

apid

s, M

I8.

811

.91.

02.

115

.720

.82.

72.

46.

38.

81.

60.

9

Indi

anap

olis

, IN

8.5

10.5

0.8

1.2

11.9

15.1

1.4

1.9

5.0

6.2

2.0

Kan

sas

Cit

y, M

O-K

S8.

510

.40.

61.

315

.017

.51.

60.

85.

57.

41.

20.

7

Lan

sing

, MI

11.0

15.4

1.2

3.2

16.9

24.4

3.8

3.7

8.8

12.0

2.2

0.9

Mad

ison

, WI

8.9

10.4

0.9

0.6

15.0

17.7

1.9

0.8

4.8

5.8

1.9

Milw

auke

e, W

I10

.612

.50.

71.

221

.324

.91.

71.

93.

65.

11.

5

Min

neap

olis

-St.

Pau

l, M

N-W

I6.

78.

30.

51.

116

.419

.91.

71.

84.

05.

40.

70.

7

Om

aha,

NE

-IA

8.4

10.3

0.9

1.1

11.3

15.3

1.5

2.5

5.3

5.9

2.1

St.

Lou

is, M

O-I

L10

.010

.90.

60.

224

.625

.42.

57.

98.

80.

70.

2

Tole

do, O

H12

.115

.21.

31.

817

.923

.42.

33.

26.

88.

63.

0

Wic

hita

, KS

9.1

12.4

1.2

2.1

11.2

14.8

1.5

2.2

6.0

8.5

4.0

Youn

gsto

wn,

OH

-PA

11.5

11.9

1.0

24.8

24.3

6.0

9.5

10.3

1.4

NO

RT

HE

AS

T (

20 M

etro

s)11

.511

.60.

221

.720

.90.

4-0

.47.

37.

90.

30.

3

Alb

any,

NY

9.3

10.5

1.0

0.3

21.7

26.5

4.6

0.1

7.8

8.9

1.2

Alle

ntow

n, P

A-N

J8.

38.

10.

818

.523

.04.

76.

65.

71.

2

Bos

ton-

Cam

brid

ge, M

A-N

H8.

59.

50.

40.

618

.621

.41.

81.

06.

77.

60.

60.

3

Bri

dgep

ort-

Sta

mfo

rd, C

T6.

97.

30.

913

.613

.62.

44.

24.

81.

6

Buf

falo

, NY

11.9

12.7

0.9

26.6

26.9

2.8

7.0

8.5

1.5

Har

risb

urg,

PA

8.2

9.0

1.0

8.2

9.0

1.0

Har

tfor

d, C

T8.

38.

40.

730

.632

.04.

95.

65.

90.

9

December 2006 • The Brookings Institution • Living Cities Census Series14

App

endi

x A

. Met

ropo

lita

n, C

entr

al-C

ity,

and

Sub

urba

n P

over

ty R

ates

,10

0 L

arge

st M

etro

Are

as, 1

999

to 2

005

(con

tinu

ed)

MS

AC

entr

al C

ity

Sub

urbs

1999

1999

1999

to 2

005

to 2

005

to 2

005

Min

imum

Min

imum

Min

imum

1999

2005

MO

EC

hang

e19

9920

05M

OE

Cha

nge

1999

2005

MO

EC

hang

e

Lan

cast

er, P

A7.

89.

11.

87.

89.

11.

8

New

Hav

en, C

T9.

511

.21.

30.

424

.427

.25.

47.

08.

71.

8

New

Yor

k-N

orth

ern

New

Jer

sey,

NY-

NJ-

PA1

13.6

12.6

0.2

-0.8

21.5

19.2

0.4

-1.8

7.2

7.2

0.6

Phi

lade

lphi

a, P

A-N

J-D

E10

.811

.70.

40.

522

.924

.51.

30.

36.

57.

50.

70.

3

Pit

tsbu

rgh,

PA

10.8

11.4

0.5

20.4

23.2

2.3

0.5

9.4

9.7

0.7

Port

land

, ME

8.0

10.4

1.1

1.2

8.0

10.4

1.1

1.3

Poug

hkee

psie

, NY

9.2

8.7

1.2

9.2

8.7

1.2

Pro

vide

nce,

RI-

MA

11.3

11.6

0.8

29.1

29.4

3.6

9.2

9.6

1.0

Roc

hest

er, N

Y10

.511

.40.

925

.930

.03.

01.

16.

47.

11.

3

Scr

anto

n, P

A10

.813

.31.

60.

915

.022

.96.

41.

510

.211

.92.

0

Spr

ingf

ield

, MA

13.1

15.3

1.2

1.0

23.1

29.6

4.3

2.1

10.1

11.1

2.0

Syr

acus

e, N

Y12

.314

.21.

30.

627

.331

.33.

40.

68.

19.

61.

9

Wor

cest

er, M

A9.

29.

81.

117

.918

.73.

76.

77.

51.

7

SO

UT

H (

37 M

etro

s)12

.013

.30.

11.

117

.319

.20.

31.

69.

610

.80.

31.

0

Atl

anta

, GA

9.5

11.4

0.5

1.4

24.4

26.9

2.5

8.0

10.0

0.6

1.4

Aug

usta

-Ric

hmon

d C

ount

y, G

A-S

C15

.417

.51.

60.

419

.622

.93.

512

.814

.53.

2

Aus

tin,

TX

11.1

13.2

0.9

1.2

14.4

18.1

1.6

2.1

7.4

8.7

2.3

Bal

tim

ore,

MD

9.8

9.5

0.6

22.9

22.6

2.2

5.4

5.5

1.1

Bat

on R

ouge

, LA

17.1

18.2

1.6

24.0

29.6

3.8

1.9

13.9

13.6

2.8

Bir

min

gham

, AL

13.6

13.3

1.2

24.7

28.9

4.0

0.2

10.3

9.2

1.9

Cha

rles

ton,

SC

14.0

14.2

1.3

19.1

19.0

2.9

13.0

13.0

1.7

Cha

rlot

te, N

C-S

C9.

411

.90.

81.

710

.613

.01.

50.

98.

511

.11.

70.

8

Cha

ttan

ooga

, TN

-GA

12.0

13.0

1.5

17.9

19.3

3.2

9.2

10.4

2.5

Col

umbi

a, S

C12

.413

.61.

222

.119

.53.

210

.612

.71.

50.

6

Dal

las-

Fort

Wor

th-A

rlin

gton

, TX

10.8

13.1

0.5

1.8

16.0

19.7

1.1

2.6

7.3

9.3

1.0

1.0

El P

aso,

TX

23.8

29.2

2.0

3.4

22.2

27.2

2.3

2.6

32.0

38.7

15.8

Gre

ensb

oro,

NC

10.6

15.3

1.5

3.2

12.3

17.3

2.8

2.3

9.8

14.4

2.6

2.1

Gre

envi

lle, S

C11

.613

.31.

30.

411

.613

.31.

30.

4

Hou

ston

, TX

13.7

16.1

0.7

1.8

19.2

22.9

1.3

2.5

9.8

12.1

1.4

0.9

Jack

son,

MS

17.0

16.2

1.6

23.5

26.0

3.4

13.2

11.3

2.9

Jack

sonv

ille,

FL

10.7

10.8

1.0

12.2

12.2

1.4

8.0

8.4

3.6

Kno

xvill

e, T

N12

.215

.31.

21.

920

.825

.02.

61.

69.

011

.81.

90.

9

Lak

elan

d, F

L12

.913

.71.

515

.014

.33.

112

.513

.61.

9

December 2006 • The Brookings Institution • Living Cities Census Series 15

App

endi

x A

. Met

ropo

lita

n, C

entr

al-C

ity,

and

Sub

urba

n P

over

ty R

ates

,10

0 L

arge

st M

etro

Are

as, 1

999

to 2

005

(con

tinu

ed)

MS

AC

entr

al C

ity

Sub

urbs

1999

1999

1999

to 2

005

to 2

005

to 2

005

Min

imum

Min

imum

Min

imum

1999

2005

MO

EC

hang

e19

9920

05M

OE

Cha

nge

1999

2005

MO

EC

hang

e

Lit

tle

Roc

k, A

R12

.113

.11.

314

.315

.02.

611

.112

.42.

1

Lou

isvi

lle, K

Y-IN

10.9

11.9

0.8

0.2

12.4

15.3

1.5

1.4

8.8

9.1

2.0

McA

llen,

TX

35.9

41.0

2.3

2.8

23.8

27.0

5.6

38.7

43.9

3.0

2.2

Mem

phis

, TN

-MS

-AR

15.6

17.7

1.2

0.8

20.6

23.6

1.9

1.2

9.9

11.2

3.2

Mia

mi-

Fort

Lau

derd

ale,

FL

14.0

14.0

0.5

25.3

25.0

1.8

12.8

12.9

0.6

Nas

hvill

e, T

N10

.311

.80.

70.

813

.314

.61.

58.

310

.11.

50.

3

New

Orl

eans

-Met

airi

e, L

A18

.317

.81.

423

.421

.21.

8-0

.413

.715

.02.

9

Okl

ahom

a C

ity,

OK

13.5

15.6

1.0

1.0

16.0

18.7

1.8

0.8

11.3

13.0

2.4

Orl

ando

, FL

10.7

11.9

0.8

0.5

15.9

15.1

2.4

10.0

11.5

1.0

0.5

Palm

Bay

, FL

9.5

10.1

1.1

9.5

10.2

3.1

9.5

10.1

1.5

Ral

eigh

, NC

8.9

10.8

0.8

1.1

11.5

15.5

1.7

2.3

7.6

8.4

1.5

Ric

hmon

d, V

A9.

410

.40.

90.

221

.418

.52.

6-0

.26.

88.

91.

10.

9

San

Ant

onio

, TX

15.1

16.2

0.9

0.2

17.3

18.7

1.4

10.7

11.6

3.7

Sar

asot

a, F

L8.

89.

61.

18.

89.

61.

1

Tam

pa-S

t. P

eter

sbur

g-C

lear

wat

er, F

L11

.212

.00.

60.

315

.315

.71.

59.

610

.80.

90.

3

Tuls

a, O

K11

.814

.01.

01.

314

.117

.61.

81.

79.

811

.32.

2

Vir

gini

a B

each

-Nor

folk

-New

port

New

s, V

A-N

C10

.610

.40.

611

.511

.41.

09.

79.

41.

7

Was

hing

ton-

Arl

ingt

on-A

lexa

ndri

a, D

C-V

A-M

D7.

47.

00.

415

.814

.41.

0-0

.45.

55.

60.

5

WE

ST

(23

Met

ros)

12.6

12.5

0.2

15.8

15.9

0.3

10.6

10.4

0.3

Alb

uque

rque

, NM

13.9

14.5

1.1

13.5

13.7

1.4

14.5

15.9

3.7

Bak

ersf

ield

, CA

20.8

21.2

1.6

18.0

18.1

2.9

22.5

23.1

3.2

Boi

se C

ity,

ID

9.4

11.9

1.3

1.3

8.4

11.8

2.4

1.0

10.0

12.0

2.5

Col

orad

o S

prin

gs, C

O7.

910

.71.

31.

58.

711

.71.

51.

56.

38.

94.

7

Den

ver-

Aur

ora,

CO

7.9

9.9

0.6

1.3

12.5

14.5

1.2

0.8

5.2

7.3

1.1

0.9

Fre

sno,

CA

22.9

20.7

1.8

-0.4

26.2

24.3

2.5

19.1

16.2

5.1

Hon

olul

u, H

I9.

99.

40.

911

.812

.01.

58.

57.

61.

9

Las

Veg

as-P

arad

ise,

NV

10.8

11.2

1.0

11.9

12.6

1.6

9.7

10.0

2.1

Los

Ang

eles

-Lon

g B

each

-San

ta A

na, C

A16

.214

.50.

4-1

.222

.019

.80.

8-1

.412

.811

.60.

7-0

.5

Mod

esto

, CA

16.0

14.4

1.9

15.7

13.4

2.9

16.2

15.1

3.8

Oxn

ard-

Tho

usan

d O

aks-

Vent

ura,

CA

9.2

9.9

1.7

10.5

12.5

2.9

7.9

7.0

4.7

Pho

enix

-Mes

a-S

cott

sdal

e, A

Z12

.012

.70.

613

.314

.40.

80.

310

.110

.71.

5

Port

land

-Van

couv

er, O

R-W

A9.

512

.80.

72.

612

.917

.61.

33.

47.

710

.41.

31.

5

Riv

ersi

de-S

an B

erna

rdin

o-O

ntar

io, C

A15

.013

.20.

8-1

.119

.318

.71.

914

.112

.11.

0-1

.0

Sac

ram

ento

, CA

12.7

12.0

0.8

20.0

19.2

2.2

10.6

9.9

1.2

December 2006 • The Brookings Institution • Living Cities Census Series16

App

endi

x A

. Met

ropo

lita

n, C

entr

al-C

ity,

and

Sub

urba

n P

over

ty R

ates

,10

0 L

arge

st M

etro

Are

as, 1

999

to 2

005

(con

tinu

ed)

MS

AC

entr

al C

ity

Sub

urbs

1999

1999

1999

to 2

005

to 2

005

to 2

005

Min

imum

Min

imum

Min

imum

1999

2005

MO

EC

hang

e19

9920

05M

OE

Cha

nge

1999

2005

MO

EC

hang

e

Sal

t L

ake

Cit

y, U

T7.

99.

41.

00.

615

.316

.92.

86.

17.

81.

30.

3

San

Die

go, C

A12

.411

.00.

8-0

.614

.613

.51.

310

.89.

21.

7

San

Fra

ncis

co-O

akla

nd-F

rem

ont,

CA

9.1

9.9

0.4

0.4

12.8

13.1

0.9

7.3

8.4

0.8

0.3

San

Jos

e-S

unny

vale

-San

ta C

lara

, CA

7.6

8.4

0.7

8.3

9.7

0.9

0.4

6.2

5.9

2.7

San

ta R

osa,

CA

8.1

9.1

1.6

8.5

10.9

3.0

7.8

8.2

2.7

Sea

ttle

-Tac

oma-

Bel

levu

e, W

A8.

59.

60.

50.

611

.912

.91.

07.

28.

40.

80.

4

Sto

ckto

n, C

A17

.714

.61.

4-1

.823

.920

.12.

4-1

.413

.010

.43.

1

Tucs

on, A

Z14

.714

.71.

118

.420

.01.

89.

78.

03.

4

1 New

ark

is t

reat

ed a

s a

cent

ral c

ity

in t

he N

ew Y

ork-

Nor

ther

n N

ew J

erse

y M

SA

Off

icia

l OM

B m

etro

are

as n

ames

mod

ifie

d to

ref

lect

onl

y th

ose

citi

es id

enti

fied

as

“cen

tral

cit

ies”

in t

his

anal

ysis

Sou

rce:

Bro

okin

gs I

nsti

tuti

on a

naly

sis

of C

ensu

s 20

00 a

nd 2

005

Am

eric

an C

omm

unit

y S

urve

y da

ta

December 2006 • The Brookings Institution • Living Cities Census Series 17

App

endi

x B

. Met

ropo

lita

n, C

entr

al-C

ity,

and

Sub

urba

n C

hild

Pov

erty

Rat

es,

100

Lar

gest

Met

ro A

reas

, 199

9 to

200

5

MS

AC

entr

al C

ity

Sub

urbs

1999

1999

1999

to 2

005

to 2

005

to 2

005

Min

imum

Min

imum

Min

imum

1999

2005

MO

EC

hang

e19

9920

05M

OE

Cha

nge

1999

2005

MO

EC

hang

e

NA

TIO

N16

.618

.50.

11.

9

100

LA

RG

ES

T M

ET

RO

AR

EA

S15

.717

.30.

21.

525

.327

.60.

41.

811

.212

.80.

31.

3

MID

WE

ST

(20

Met

ros)

13.3

16.3

0.4

2.7

25.8

30.5

0.9

3.8

7.8

10.5

0.7

2.1

Akr

on, O

H13

.514

.92.

926

.027

.15.

47.

79.

54.

9C

hica

go-N

aper

ville

-Jol

iet,

IL

-IN

-WI

14.3

16.3

0.9

1.1

26.7

28.3

2.1

8.0

10.6

1.7

1.0

Cin

cinn

ati,

OH

-KY-

IN12

.416

.31.

42.

532

.535

.46.

08.

813

.22.

02.

3C

leve

land

, OH

15.8

20.9

1.5

3.5

38.0

47.6

4.3

5.4

8.5

13.1

2.4

2.2

Col

umbu

s, O

H12

.416

.61.

52.

719

.027

.23.

05.

17.

69.

33.

3D

ayto

n, O

H13

.520

.02.

54.

132

.343

.29.

21.

68.

815

.33.

72.

8D

es M

oine

s, I

A9.

58.

02.

115

.713

.54.

95.

64.

64.

6D

etro

it-W

arre

n-L

ivon

ia, M

I14

.919

.01.

03.

130

.338

.83.

05.

48.

411

.61.

91.

3G

rand

Rap

ids,

MI

10.7

16.1

1.9

3.5

20.2

29.7

5.7

3.8

7.5

11.3

3.3

0.6

Indi

anap

olis

, IN

10.9

14.8

1.7

2.2

16.7

22.3

3.2

2.4

5.4

8.1

4.3

Kan

sas

Cit

y, M

O-K

S11

.415

.01.

62.

021

.627

.94.

02.

46.

89.

42.

9L

ansi

ng, M

I11

.518

.62.

94.

223

.637

.28.

05.

66.

710

.65.

6M

adis

on, W

I7.

410

.62.

11.

111

.718

.35.

41.

25.

46.

74.

2M

ilwau

kee,

WI

16.0

19.1

1.7

1.3

32.0

38.1

3.6

2.6

4.1

4.9

4.0

Min

neap

olis

-St.

Pau

l, M

N-W

I8.

610

.80.

81.

524

.428

.83.

50.

94.

66.

61.

30.

7O

mah

a, N

E-I

A11

.214

.21.

71.

316

.023

.33.

24.

16.

86.

84.

1S

t. L

ouis

, MO

-IL

14.1

16.1

1.2

0.8

36.9

37.7

5.4

10.9

12.9

1.5

0.5

Tole

do, O

H16

.321

.52.

23.

026

.134

.74.

04.

57.

19.

95.

7W

ichi

ta, K

S11

.815

.52.

61.

214

.918

.23.

00.

47.

511

.28.

4Yo

ungs

tow

n, O

H-P

A17

.718

.42.

637

.739

.115

.214

.315

.43.

8

NO

RT

HE

AS

T (

20 M

etro

s)15

.516

.10.

30.

330

.730

.50.

89.

310

.40.

60.

5A

lban

y, N

Y12

.015

.02.

50.

529

.243

.713

.70.

810

.212

.33.

1A

llent

own,

PA

-NJ

11.5

12.9

2.0

29.9

40.7

9.1

1.8

8.3

8.2

2.9

Bos

ton-

Cam

brid

ge, M

A-N

H10

.312

.00.

90.

724

.831

.25.

01.

48.

29.

31.

2B

ridg

epor

t-S

tam

ford

, CT

8.5

8.8

2.0

18.7

20.9

6.3

4.4

4.3

3.6

Buf

falo

, NY

17.2

17.6

2.1

38.7

37.5

6.2

9.3

10.5

3.7

Har

risb

urg,

PA

11.3

12.3

2.3

11.3

12.3

2.3

Har

tfor

d, C

T11

.111

.71.

441

.342

.57.

86.

67.

52.

1L

anca

ster

, PA

11.2

14.1

4.8

11.2

14.1

4.8

New

Hav

en, C

T13

.317

.73.

01.

432

.639

.812

.09.

814

.54.

00.

7N

ew Y

ork-

Nor

ther

n N

ew J

erse

y, N

Y-N

J-PA

118

.617

.50.

6-0

.530

.527

.80.

9-1

.79.

19.

51.

2P

hila

delp

hia,

PA

-NJ-

DE

14.3

16.6

0.9

1.4

31.6

35.4

2.8

1.0

8.1

9.9

1.6

0.3

December 2006 • The Brookings Institution • Living Cities Census Series18

App

endi

x B

. Met

ropo

lita

n, C

entr

al-C

ity,

and

Sub

urba

n C

hild

Pov

erty

Rat

es,

100

Lar

gest

Met

ro A

reas

, 199

9 to

200

5 (c

onti

nued

)

MS

AC

entr

al C

ity

Sub

urbs

1999

1999

1999

to 2

005

to 2

005

to 2

005

Min

imum

Min

imum

Min

imum

1999

2005

MO

EC

hang

e19

9920

05M

OE

Cha

nge

1999

2005

MO

EC

hang

e

Pit

tsbu

rgh,

PA

14.8

14.5

1.1

27.8

31.4

4.9

12.9

12.5

1.5

Port

land

, ME

10.1

13.9

2.6

1.2

10.1

13.9

2.6

1.2

Poug

hkee

psie

, NY

12.7

13.0

2.4

12.7

13.0

2.4

Pro

vide

nce,

RI-

MA

15.7

18.0

1.9

0.4

40.5

43.4

7.2

12.3

14.6

2.4

Roc

hest

er, N

Y14

.515

.51.

937

.941

.86.

17.

58.

53.

0S

cran

ton,

PA

14.6

20.9

4.0

2.3

19.2

39.7

16.9

3.6

13.9

18.1

5.4

Spr

ingf

ield

, MA

19.3

22.1

2.3

0.5

34.3

46.0

8.3

3.4

13.9

13.5

4.5

Syr

acus

e, N

Y15

.720

.43.

31.

435

.445

.07.

02.

610

.213

.64.

8W

orce

ster

, MA

11.8

13.1

2.8

25.1

26.8

10.2

8.3

9.9

4.2

SO

UT

H (

37 M

etro

s)16

.318

.90.

32.

324

.728

.80.

73.

412

.614

.80.

51.

7A

tlan

ta, G

A12

.316

.11.

22.

639

.349

.15.

44.

59.

913

.61.

42.

4A

ugus

ta-R

ichm

ond

Cou

nty,

GA

-SC

21.3

27.6

3.4

2.9

27.7

38.2

7.5

3.0

17.2

21.2

7.2

Aus

tin,

TX

12.2

17.1

2.0

2.9

17.0

25.7

3.6

5.1

8.0

10.1

4.6

Bal

tim

ore,

MD

12.6

13.4

1.5

31.0

33.3

4.9

6.4

7.1

2.5

Bat

on R

ouge

, LA

21.7

25.7

3.3

0.7

31.7

47.1

8.1

7.3

17.5

17.6

5.6

Bir

min

gham

, AL

18.1

20.3

2.3

35.6

49.5

8.6

5.3

12.9

12.0

3.9

Cha

rles

ton,

SC

19.0

20.5

2.7

24.8

25.7

7.0

18.1

19.3

3.8

Cha

rlot

te, N

C-S

C12

.216

.01.

82.

114

.118

.83.

51.

210

.914

.13.

9C

hatt

anoo

ga, T

N-G

A16

.619

.93.

427

.433

.57.

911

.914

.45.

9C

olum

bia,

SC

16.0

18.7

2.6

0.1

30.2

25.6

8.3

13.7

17.9

3.1

1.0

Dal

las-

Fort

Wor

th-A

rlin

gton

, TX

14.2

18.3

0.9

3.1

22.4

29.1

2.1

4.7

9.1

12.1

1.8

1.2

El P

aso,

TX

31.7

39.9

3.3

4.8

30.1

38.1

4.2

3.8

38.6

47.1

23.4

Gre

ensb

oro,

NC

14.1

22.1

3.2

4.8

16.3

24.4

6.8

1.3

13.1

21.1

5.5

2.5

Gre

envi

lle, S

C14

.318

.42.

61.

514

.318

.42.

61.

5H

oust

on, T

X17

.923

.71.

34.

526

.435

.02.

56.

212

.417

.32.

52.

4Ja

ckso

n, M

S24

.324

.13.

534

.036

.46.

918

.217

.26.

7Ja

ckso

nvill

e, F

L15

.013

.92.

217

.015

.73.

010

.810

.58.

8K

noxv

ille,

TN

15.2

20.8

3.1

2.5

26.5

31.7

6.5

11.5

17.3

4.6

1.1

Lak

elan

d, F

L19

.622

.62.

924

.427

.78.

318

.921

.63.

8L

ittl

e R

ock,

AR

16.8

19.4

3.0

21.5

20.8

6.3

14.9

18.9

4.7

Lou

isvi

lle, K

Y-IN

15.6

16.4

1.8

18.5

22.1

3.2

0.4

11.6

11.3

4.4

McA

llen,

TX

45.7

51.8

3.5

2.6

30.6

34.0

7.9

48.6

55.0

4.2

2.2

Mem

phis

, TN

-MS

-AR

22.3

25.8

2.4

1.2

30.4

34.8

4.0

0.3

13.1

16.2

6.9

Mia

mi-

Fort

Lau

derd

ale,

FL

19.0

19.0

1.2

36.1

36.9

4.4

17.3

17.3

1.4

Nas

hvill

e, T

N13

.416

.01.

90.

619

.823

.54.

59.

711

.63.

9N

ew O

rlea

ns-M

etai

rie,

LA

26.2

26.9

2.9

35.1

33.1

3.8

18.8

22.4

5.8

Okl

ahom

a C

ity,

OK

18.3

21.7

2.0

1.4

23.5

28.2

3.9

0.8

13.9

16.2

5.1

December 2006 • The Brookings Institution • Living Cities Census Series 19

App

endi

x B

. Met

ropo

lita

n, C

entr

al-C

ity,

and

Sub

urba

n C

hild

Pov

erty

Rat

es,

100

Lar

gest

Met

ro A

reas

, 199

9 to

200

5 (c

onti

nued

)

MS

AC

entr

al C

ity

Sub

urbs

1999

1999

1999

to 2

005

to 2

005

to 2

005

Min

imum

Min

imum

Min

imum

1999

2005

MO

EC

hang

e19

9920

05M

OE

Cha

nge

1999

2005

MO

EC

hang

e

Orl

ando

, FL

14.8

16.7

2.0

27.5

22.3

5.4

13.4

15.9

2.4

0.2

Palm

Bay

, FL

13.5

12.8

3.0

11.7

10.5

6.9

14.0

13.4

4.2

Ral

eigh

, NC

10.5

13.7

1.5

1.7

14.3

21.2

4.0

2.9

9.0

10.8

2.7

Ric

hmon

d, V

A12

.814

.71.

933

.426

.86.

2-0

.49.

012

.42.

51.

0S

an A

nton

io, T

X21

.422

.71.

924

.626

.72.

814

.914

.97.

8S

aras

ota,

FL

14.5

13.2

3.0

14.5

13.2

3.0

Tam

pa-S

t. P

eter

sbur

g-C

lear

wat

er, F

L16

.216

.91.

523

.422

.54.

013

.415

.02.

5Tu

lsa,

OK

16.0

21.0

2.1

2.9

20.9

29.2

4.2

4.1

12.3

15.2

4.7

Vir

gini

a B

each

-Nor

folk

-New

port

New

s, V

A-N

C15

.216

.71.

516

.518

.42.

613

.614

.84.

4W

ashi

ngto

n-A

rlin

gton

-Ale

xand

ria,

DC

-VA

-MD

9.1

8.7

0.7

25.3

23.2

2.7

6.6

6.5

0.9

WE

ST

(23

Met

ros)

17.1

17.3

0.4

22.1

22.8

0.7

14.2

14.1

0.7

Alb

uque

rque

, NM

18.7

20.6

2.3

17.9

19.8

3.6

19.6

21.9

7.9

Bak

ersf

ield

, CA

28.2

29.4

3.0

24.8

26.0

5.0

30.2

31.7

5.9

Boi

se C

ity,

ID

11.9

15.0

2.3

0.8

10.0

15.4

5.0

0.4

13.0

14.9

4.0

Col

orad

o S

prin

gs, C

O10

.314

.72.

71.

711

.316

.03.

11.

78.

512

.68.

7D

enve

r-A

uror

a, C

O10

.113

.31.

22.

017

.620

.92.

80.

56.

19.

12.

40.

6F

resn

o, C

A32

.130

.12.

836

.834

.83.

826

.423

.58.

9H

onol

ulu,

HI

12.9

11.3

2.2

15.1

13.9

3.5

11.7

10.1

3.7

Las

Veg

as-P

arad

ise,

NV

14.6

15.2

2.1

15.9

17.5

3.7

13.4

13.5

4.7

Los

Ang

eles

-Lon

g B

each

-San

ta A

na, C

A22

.120

.60.

8-0

.730

.428

.51.

4-0

.517

.516

.41.

4M

odes

to, C

A21

.018

.93.

622

.318

.66.

420

.119

.27.

3O

xnar

d-T

hous

and

Oak

s-Ve

ntur

a, C

A12

.114

.63.

613

.819

.87.

010

.48.

611

.3P

hoen

ix-M

esa-

Sco

ttsd

ale,

AZ

16.4

18.3

1.2

0.7

18.2

21.7

1.8

1.7

13.7

14.4

3.4

Port

land

-Van

couv

er, O

R-W

A11

.817

.31.

73.

816

.626

.23.

16.

59.

713

.62.

71.

2R

iver

side

-San

Ber

nard

ino-

Ont

ario

, CA

20.1

17.7

1.5

-0.9

25.2

25.1

3.4

19.0

16.2

2.0

-0.8

Sac

ram

ento

, CA

17.3

16.7

1.8

29.9

29.1

4.6

13.6

13.1

2.7

Sal

t L

ake

Cit

y, U

T9.

311

.11.

70.

219

.022

.96.

87.

69.

22.

3S

an D

iego

, CA

16.9

15.9

1.8

20.3

19.8

3.2

14.7

13.3

3.7

San

Fra

ncis

co-O

akla

nd-F

rem

ont,

CA

11.3

12.3

1.0

17.9

16.5

2.1

8.7

10.7

1.6

0.5

San

Jos

e-S

unny

vale

-San

ta C

lara

, CA

9.1

10.8

1.4

0.3

10.1

12.5

1.7

0.7

7.2

7.5

5.3

San

ta R

osa,

CA

9.0

11.4

3.5

10.0

15.6

6.3

8.5

9.4

5.9

Sea

ttle

-Tac

oma-

Bel

levu

e, W

A10

.511

.31.

115

.313

.92.

29.

110

.51.

5S

tock

ton,

CA

24.2

19.7

2.5

-1.9

33.2

28.1

3.7

-1.4

16.8

13.5

5.3

Tucs

on, A

Z20

.021

.22.

224

.327

.93.

514

.212

.17.

1

1 New

ark

is t

reat

ed a

s a

cent

ral c

ity

in t

he N

ew Y

ork-

Nor

ther

n N

ew J

erse

y M

SA

Off

icia

l OM

B m

etro

are

as n

ames

mod

ifie

d to

ref

lect

onl

y th

ose

citi

es id

enti

fied

as

“cen

tral

cit

ies”

in t

his

anal

ysis

Sou

rce:

Bro

okin

gs I

nsti

tuti

on a

naly

sis

of C

ensu

s 20

00 a

nd 2

005

Am

eric

an C

omm

unit

y S

urve

y da

ta

December 2006 • The Brookings Institution • Living Cities Census Series20

Appendix C. Poverty Rates in First Suburbs, 100 Largest Metropolitan Areas, 1999 to 2005

First Suburb Counties Metropolitan Area 1999 2005 MOE Minimum Change

ALL FIRST SUBURBS 8.6 9.3 0.2 0.5

MIDWEST (17 First Suburban Counties)Summit, OH Akron, OH 5.0 6.0 2.3Cook, IL Chicago-Naperville-Joliet, IL-IN-WI 6.4 8.1 1.9Lake, IN Chicago-Naperville-Joliet, IL-IN-WI 12.2 16.4 1.6 2.6Hamilton, OH Cincinnati, OH-KY-IN 5.5 7.4 2.5Cuyahoga, OH Cleveland, OH 6.3 9.7 1.8 1.6Franklin, OH Columbus, OH 5.4 7.1 3.5Montgomery, OH Dayton, OH 6.6 10.0 2.4 1.0Macomb, MI Detroit-Warren-Livonia, MI 5.6 8.6 0.8 2.1Oakland, MI Detroit-Warren-Livonia, MI 5.5 6.6 0.6 0.4Wayne, MI Detroit-Warren-Livonia, MI 8.7 11.6 2.6 0.3Marion, IN Indianapolis, IN 6.8 11.2 19.8Milwaukee, WI Milwaukee, WI 4.8 7.2 4.0Hennepin, MN Minneapolis-St. Paul, MN-WI 3.9 6.0 1.6 0.5Madison, IL St. Louis, MO-IL 9.8 10.5 1.9St. Clair, IL St. Louis, MO-IL 14.5 11.8 1.7 -1.0St. Louis, MO St. Louis, MO-IL 6.9 8.2 0.8 0.5Trumbull, OH Youngstown, OH-PA 10.3 10.7 1.7

NORTHEAST (26 First Suburban Counties)Northampton/Lehigh, PA Allentown, PA-NJ 6.5 5.7 1.6Middlesex, MA Boston-Cambridge, MA-NH 6.1 7.1 0.8 0.2Norfolk, MA Boston-Cambridge, MA-NH 4.6 5.2 0.9Fairfield, CT Bridgeport-Stamford, CT 4.2 4.8 1.6Erie, NY Buffalo, NY 5.8 7.8 1.8 0.1Hartford, CT Hartford, CT 5.8 6.0 1.3New Haven, CT New Haven, CT 7.0 8.7 1.8Bergen, NJ New York-Northern New Jersey, NY-NJ-PA 5.0 5.5 0.8Hudson, NJ New York-Northern New Jersey, NY-NJ-PA 15.5 16.5 1.9Middlesex, NJ New York-Northern New Jersey, NY-NJ-PA 6.6 7.8 1.2Nassau, NY New York-Northern New Jersey, NY-NJ-PA 5.2 5.2 0.7Union, NJ New York-Northern New Jersey, NY-NJ-PA 8.4 8.9 1.4Westchester, NY New York-Northern New Jersey, NY-NJ-PA 8.8 7.9 0.9Essex, NJ New York-Northern New Jersey, NY-NJ-PA 9.0 9.7 2.5Bucks, PA Philadelphia, PA-NJ-DE 4.5 5.8 0.8 0.4Burlington, NJ Philadelphia, PA-NJ-DE 4.7 4.6 0.8Camden, NJ Philadelphia, PA-NJ-DE 10.4 12.4 1.3 0.7Delaware, PA Philadelphia, PA-NJ-DE 8.0 9.7 1.2 0.4Montgomery, PA Philadelphia, PA-NJ-DE 4.4 4.9 0.6Allegheny, PA Pittsburgh, PA 8.1 9.1 1.4Providence, RI Providence, RI-MA 10.4 11.5 2.4Monroe, NY Rochester, NY 4.9 6.3 1.9Lackawanna, PA Scranton, PA 8.2 10.3 5.7Hampden, MA Springfield, MA 10.6 10.5 3.1Onondaga, NY Syracuse, NY 12.2 13.7 1.2 0.3Worcester, MA Worcester, MA 6.7 7.5 1.7

Endnotes

1. See, e.g., Lawrence Mishel, Jared Bernstein,and Sylvia Allegretto, The State of WorkingAmerica 2004/2005 (Cornell University Press,2005).

2. Carmen DeNavas-Walt, Bernadette D. Proc-tor, and Cheryl Hill Lee, “Income, Poverty,and Health Insurance Coverage in the UnitedStates: 2005” (Census Bureau, 2006).

3. Brookings Institution and Population Refer-ence Bureau, “Kids in the City: Indicators ofChild Well-Being in Large Cities from the2004 American Community Survey” (Wash-ington: Brookings Institution, 2006).

4. Income and poverty data reported by Census2000 were collected for the 1999 calendaryear. Therefore, this study uses “1999” torefer to estimates from the 2000 decennialcensus.

5. U.S. Census Bureau, “Using Data from the2005 American Community Survey” (2006)(www.census.gov/acs/www/UseData/advance_copy_user_guide.pdf [accessed October2006]).

6. The Census Bureau arrives at the povertythreshold via a somewhat complicated for-mula that adjusts for inflation on a monthlybasis. See www.census.gov/acs/www/Down-loads/2005/usedata/Subject_Definitions.pdf [accessed October 2006].

7. In these cases, the ACS estimates showed aslightly higher percentage of people in povertycompared to Census 2000. For more detail onthe comparability of ACS and census data, seewww.census.gov/acs/www/Downloads/Report05.pdf [accessed October 2006].

8. We use “z-tests” to compare estimates fromCensus 2000 and the 2005 ACS, and con-clude that the estimates are different if theabsolute value of the z-score is greater than1.65. We assume a margin of error of +/- 0.1percentage points for all Census 2000 povertyrate estimates.

9. We use the difference between the 1999poverty rate and the 2005 90-percent confi-dence interval lower-bound (in the case ofpoverty increases) or upper-bound (in the caseof poverty decreases) to identify places under-going the largest changes.

10. William H. Frey and others, “Tracking Ameri-can Trends into the Twenty-First Century: A

Field Guide to the New Metropolitan andMicropolitan Definitions.” In A. Berube, B.Katz, and R. Lang, eds., Redefining Urban andSuburban America: Evidence from Census2000, vol. 3 (Washington: Brookings Institu-tion, 2006).

11. Because of this adjustment, 1999 metropoli-tan poverty figures reported here differ frommetropolitan area data reported by Census2000.

12. OMB designates the city with the largest pop-ulation in the metropolitan area as the princi-pal city and lists it first in the metropolitanname. Other principal cities are determinedbased on criteria including, population size,employment, and the level of their local socialand economic importance. OMB guidelinescan be viewed in their entirety at www.cen-sus.gov/population/www/estimates/about-metro.html [accessed October 2006].

13. There are certain exceptions to these criteria.Newark is not named in the New York-North-ern New Jersey-Long Island, NY-NJ MSAunder the 2003 guidelines, but for this assess-ment it is designated as the central city forNorthern New Jersey based on its populationsize and role as an economic center in thearea. In addition, the 2005 ACS did not

December 2006 • The Brookings Institution • Living Cities Census Series 21

Appendix C. Poverty Rates in First Suburbs, 100 Largest Metropolitan Areas, 1999 to 2005 (continued)

First Suburb Counties Metropolitan Area 1999 2005 MOE Minimum Change

SOUTH (9 First Suburban Counties)Fulton, GA Atlanta, GA 7.2 6.1 3.2Baltimore, MD Baltimore, MD 6.5 6.8 0.8Jefferson, AL Birmingham, AL 9.1 8.7 3.4Dallas, TX Dallas-Fort Worth-Arlington, TX 8.4 11.6 2.6 0.6Harris, TX Houston, TX 9.3 12.2 2.4 0.5Miami-Dade, FL Miami-Fort Lauderdale, FL 16.0 15.9 1.3Hillsborough, FL Tampa-St. Petersburg-Clearwater, FL 10.1 10.9 1.8Montgomery, MD Washington-Arlington-Alexandria, DC-VA-MD 5.4 4.5 0.6 -0.3Prince George’s, MD Washington-Arlington-Alexandria, DC-VA-MD 7.7 8.5 1.3

WEST (9 First Suburban Counties)Los Angeles, CA Los Angeles-Long Beach-Santa Ana, CA 14.6 13.4 1.0 -0.1Orange, CA Los Angeles-Long Beach-Santa Ana, CA 9.1 7.9 0.7 -0.5Maricopa, AZ Phoenix-Mesa-Scottsdale, AZ 9.2 10.1 1.7Sacramento, CA Sacramento, CA 11.2 10.8 2.0San Diego, CA San Diego, CA 10.8 9.2 1.7Alameda, CA San Francisco-Oakland-Fremont, CA 8.4 10.3 1.9San Mateo, CA San Francisco-Oakland-Fremont, CA 5.8 7.4 1.4 0.2King, WA Seattle-Tacoma-Bellevue, WA 6.9 8.1 1.2Pierce, WA Seattle-Tacoma-Bellevue, WA 8.4 8.9 1.7

Official OMB metro areas names modified to reflect only those cities identified as "central cities" in this analysisSource: Brookings Institution analysis of Census 2000 and 2005 American Community Survey data

release estimates for geographies with popula-tions under 65,000. The largest principal cityin six of the 100 largest metropolitan areasdoes not meet the ACS population threshold.Because 2005 poverty data are not currentlyavailable in those cases, these six metropoli-tan areas do not include a central city and areentirely included in the suburban assessment:Greenville, SC; Harrisburg-Carlisle, PA; Lancaster, PA; Portland-South Portland-Bid-deford, ME; Poughkeepsie-Newburgh-Middle-town, NY; Sarasota-Bradenton-Venice, FL.

14. For the six metropolitan areas that do nothave central cities, their names have beenshortened to the first city listed in the MSAname. For 1999 central-city figures in theLouisville, KY-IN MSA, we report those forJefferson County, KY, which roughly equateswith the new Louisville-Jefferson County, KYconsolidated city in 2005.

15. States in the Northeast include: CT, ME, MA,NH, NJ, NY, PA, RI, VT. States in the Mid-west include: IL, IN, IA, KS, MI, MN, MO,NE, ND, OH, SD, WI. States in the Southinclude: AL, AR, DE, DC, FL, GA, KY, LA,MD, MS, NC, OK, SC, TN, TX, VA, WV.States in the West include: AK, AR, CA, CO,HI, ID, MT, NV, NM, OR, UT, WA, WY.

16. Robert Puentes and David Warren, “One-Fifth of America: A Comprehensive Guide toAmerica’s First Suburbs” (Washington: Brook-ings Institution, 2006).

17. Within counties that contain first suburbs, wealso treat as cities (and thus exclude from firstsuburban totals) those defined as centralcities in our analysis. Thus, although Puentesand Warren (ibid) recognize Orange County,CA as a fully first-suburban county, weexclude from its poverty calculations SantaAna, CA, a central city in the Los Angeles-Long Beach-Santa Ana, CA metropolitan area.Likewise, we do not recognize Gary, IN as acity in Lake County, IN, because it does notmeet our criteria to be treated as a central cityin the Chicago-Naperville-Joliet, IL-IN-WImetropolitan area.

18. In 2005, a family of one adult and two chil-dren with an income of $15,735 or less wasconsidered to be poor. The threshold for afamily of two adults and two children was$19,806.

19. Both Census 2000 and the 2005 ACS col-lected income information only for the popu-lation living in households. This means thatindividuals living in group quarters and insti-tutions are not included in poverty measures.Therefore, this study uses the term “popula-tion” to refer to all individuals for whompoverty status is determined.

20. Among its limitations, the federal poverty linedoes not account for regional differences inthe cost of living, nor does it factor in thevalue of tax credits and cash/in-kind govern-ment transfers, such as Food Stamps and theEarned Income Tax Credit. A panel from theNational Academy of Sciences issued a reportin 1995 that recommend changing the way inwhich the poverty threshold is calculated inorder to reflect the following: differences inneed by family size and geography; changes inconsumption patterns and household compo-sition; and changes in labor force patterns. Todate, these changes have not been imple-mented. See National Academy of Sciences,Measuring Poverty: A New Approach (Wash-ington: National Academy Press, 1995).

21. John Cassidy, “Relatively Deprived.” The NewYorker, April 3, 2006.

22. This trend does not owe to the patterns in afew large metropolitan areas, but insteadreflects that more metropolitan areas have amajority of their poor in suburbs than have amajority of their poor in central cities. Of the94 metropolitan areas for which we definecentral cities, 33 have more poor in their sub-urbs than their cities, while 22 have morepoor in their cities than suburbs. In the other39, the number of poor in central cities andsuburbs were not statistically different.

23. This finding hinges in no small part on howwe have chosen to define “central cities” and“suburbs” in our analysis—identifying onlythe large and well-known city or cities withineach metro area as central cities, and focusingon only the 100 largest metro areas in theUnited States. Nonetheless, it is notable thatper this definition a majority of poor Ameri-cans within these metro areas lived in centralcities in 1999, and that the statistic had“flipped” by 2005.

24. William Frey, “The New Great Migration:Black Americans’ Return to the South,1965–2000” (Washington: Brookings Institu-tion, 2004).

25. Hans P. Johnson and Joseph M. Hayes, “TheCentral Valley at a Crossroads: Migration andIts Implications” (San Francisco: Public Pol-icy Institute of California, 2004).

26. Rick Lyman, “New Data Shows Immigrants’Growth and Reach.” The New York Times,August 10, 2006, p. A1.

27. Despite the aggregate pattern in that region,there was little indication from the 1995-to-2005 trends that poverty had been “pushedout” to the suburbs by increasing housingcosts; poverty rates rose by significant marginsin only two Northeastern suburban areas (andacross the whole of the Portland, ME metroarea). However, it is difficult to capture these

underlying dynamics absent more detailedmigration and income data.

28. See, e.g., David Rusk, “Annexation and theFiscal Fate of Cities” (Washington: BrookingsInstitution, 2006); Douglas Massey andNancy Denton, American Apartheid: Segrega-tion and the Making of the Underclass (Har-vard University Press, 1993); William H. Freyand Elaine Fielding, “Changing Urban Popu-lations: Regional Restructuring, Racial Polar-ization, and Poverty Concentration.” Cityscape1(2)(1995): 1–66.

29. Alan Berube and William H. Frey, “A Decadeof Mixed Blessings: Urban and SuburbanPoverty in Census 2000.” In A. Berube, B.Katz, and R. Lang, eds., Redefining Urban andSuburban America: Evidence from Census2000, vol. 2 (Washington: Brookings Institu-tion Press, 2005).

30. Berube and Frey, “A Decade of Mixed Bless-ings.”

31. This does not mean that poverty rates inthese cities did not change over this period,but rather that the 1999 rate lies within thestatistical margin of error associated with the2005 estimate. See Methodology. In 25 of the43 central cities, and 46 of the 67 suburbs,where the 2005 rate was statistically no dif-ferent than the 1999 rate, the 2005 estimatewas at least 0.2 percentage points higherthan the 1999 rate, suggesting that a largersample might have detected more widespreadpoverty increases.

32. The top cities and suburbs for poverty-ratechanges are determined based on the mini-mum amount by which their 2005 estimatesexceed (or trail) their 1999 rates. This mini-mum amount is based on the differencebetween the 1999 rate and the lower boundof the 90-percent confidence interval for the2005 rate in the case of poverty increases,and the upper bound of that interval in thecase of poverty decreases. We cannot be cer-tain that these represent the true cities andsuburbs experiencing the greatest poverty-ratechanges from 1999 to 2005, but the surveyresults suggest these to be among the likeliestlocations. Eleven cities with poverty-rateincreases are shown in Map 1 becauseColumbus and Grand Rapids tied for thetenth spot.

33. See, e.g., Heather A. Smith and Owen J.Furuseth, eds., Latinos in the New South(Hampshire, UK: Aldershot, 2006).

34. New Orleans-Metairie may appear among thecentral cities experiencing poverty-ratedeclines from 1999 to 2005 due to the dispro-portionate out-migration of poor householdsfrom the region subsequent to Hurricane Kat-rina’s impact in August 2005 ; William H.

December 2006 • The Brookings Institution • Living Cities Census Series22

Frey and Audrey Singer, "Katrina and RitaImpacts on Gulf Coast Populations: FirstCensus Findings" (Washington: BrookingsInstitution, 2006)

35. If central city and suburban poverty were trulyunrelated, we would have expected only(0.31)*(0.43)*100 = 13 of the central citiesin metropolitan areas with increasing subur-ban poverty rates to have experiencedincreases in poverty rates as well. In only onemetropolitan area—Richmond, VA—did thecentral-city poverty rate decrease while thesuburban poverty rate increased. No metro-politan areas saw suburban poverty increaseas city poverty decreased.

36. McAllen’s and El Paso’s suburbs featureunderdeveloped border communities that arepredominantly Hispanic and low-income. The90-percent confidence interval for El Paso’ssuburbs is so large because the city of El Pasocomprises the bulk of El Paso County, whichcoincides with the El Paso metropolitan area.

37. Rolf Pendall, Robert Puentes, and JonathanMartin, “From Traditional to Reformed: AReview of Land Use Regulations in theNation’s 50 Largest Metropolitan Areas”(Washington: Brookings Institution, 2006).

38. The margin of error for the 2005 estimate was+/- 0.2 percentage points, indicating that thepoverty rate rose from 1999.

39. The margin of error for the 2005 estimate was+/- 0.9 percentage points, indicating that theproportion declined from 1999.

40. Brookings Institution and Population Refer-ence Bureau, “Kids in the City.”

41. The margin of error for the 2005 estimate was+/- 0.1 percentage points, indicating that theU.S. child poverty rate rose from 1999.

42. It is possible that some portion of thispoverty-rate increase in Houston is attributa-ble to the influx of lower-income evacueesfrom the New Orleans area post-HurricaneKatrina in 2005. However, it is not clear thatmany of these households would havereceived and responded to an ACS survey dur-ing the September-to-December 2005 period,as many lived in temporary accommodations(such as shelters and motels) that probablywere not captured by the ACS.

43. The poverty rate for families with related chil-dren in 2005 was 14.5 percent, compared to5.1 percent for families without children. Forfamilies with children headed by a female, thepoverty rate was 36.2 percent. U.S. CensusBureau, Current Population Survey, 2006Annual Social and Economic Supplement(available at http://pubdb3.census.gov/macro/032006/pov/new04_100_01.htm[accessed October 2006]).

44. Alan Berube, “The New Safety Net: How theTax Code Helped Low-Income Working Fami-lies During the Early 2000s” (Washington:Brookings Institution, 2006).

45. Wayne Vroman, “An Introduction to Unem-ployment and Unemployment Insurance”(Washington: Urban Institute, 2005).

46. For more on these approaches, see HowardWial and Alec Friedhoff, “Bearing the Brunt:Manufacturing Job Loss in the Great LakesRegion, 1995–2005” (Washington: BrookingsInstitution, 2006).

47. Several indicators of these regional dividescan be viewed at the American Research andGeographic Information Systems website,www.ameregis.com [accessed October 2006].

48. See., e.g., David Rusk, “Nine Lessons forInclusionary Zoning.” Keynote remarks to theNational Inclusionary Housing Conference(Washington: October 5, 2005); Mary E.Brooks, Housing Trust Fund Progress Report2002 (Washington: Center for CommunityChange); State of Illinois, Building for Suc-cess: Illinois’ Comprehensive Housing Plan(Chicago: Illinois Housing DevelopmentAuthority, 2005).

49. See, e.g., David Jason Fischer, “WorkforceIntermediaries: Powering Regional Economiesin the New Century” (Baltimore: Annie E.Casey Foundation, 2005); Robert P. Giloth,ed., Workforce Intermediaries for the Twenty-First Century (Temple University Press, 2004).

50. Alan Berube, “¿Tienes EITC? A Study of theEarned Income Tax Credit in ImmigrantCommunities” (Washington: Brookings Insti-tution, 2005).

51. Scott Allard, “Access to Social Services: TheChanging Urban Geography of Poverty andService Provision” (Washington: BrookingsInstitution, 2004).

52. U.S. Census Bureau, “American CommunitySurvey: A Handbook for State and Local Offi-cials” (2004).

December 2006 • The Brookings Institution • Living Cities Census Series 23

For More Information:Alan BerubeResearch DirectorThe Brookings Institution Metropolitan Policy Program(202) [email protected]

Elizabeth KneeboneResearch AnalystThe Brookings Institution Metropolitan Policy Program(202) [email protected]

For General Information:The Brookings Institution Metropolitan Policy Program(202) 797-6139www.brookings.edu/metro

AcknowledgmentsThe authors thank Scott Allard, Matt Fellowes, and Audrey Singer for theirhelpful comments on a draft of this paper, and staff at the Brookings Metro-politan Policy Program for collective guidance and advice on using the Ameri-can Community Survey.

About the Living Cities Census Series

Sweeping U.S. demographic changes alter the context in which urban and metropolitanpolicies take shape. With support from Living Cities: The National CommunityDevelopment Initiative, the Brookings Institution’s Metropolitan Policy Program con-tinues the Living Cities Census Series, a major effort to illustrate how recent demo-graphic, social, and economic trends affect the vitality of cities and metropolitan areas.Building upon prior work focused on the results of Census 2000, the series is payingspecial attention to changes occurring during the current decade.

Living Cities: The National Community Development Initiative is a partnership ofleading foundations, financial institutions, nonprofit organizations, and the federal gov-ernment that is committed to improving the vitality of cities and urban communities.Living Cities supports the work of community development corporations in 23 citiesand uses the lessons of that work to engage in national research and policy develop-ment. Visit Living Cities on the web at www.livingcities.org.

55 West 125th Street, 11th floor • New York, New York 10027Tel: 646-442-2200 • Fax: 646-442-2239

www.livingcities.org

The Brookings Institution

1775 Massachusetts Avenue, NW • Washington D.C. 20036-2188Tel: 202-797-6000 • Fax: 202-797-6004

www.brookings.edu

Direct: 202-797-6139 • Fax/direct: 202-797-2965

www.brookings.edu/metro