savings interventions and behaviors - prosperity now...evaluating tax time savings interventions and...

JANUARY 2020

Evaluating Tax Time Savings Interventions and Behaviors

EVALUATING TAX TIME SAVINGS INTERVENTIONS AND BEHAVIORS

2

Evaluating Tax Time Savings Interventions and

Behaviors

AUTHORS

Rebecca Thompson, Kasey Wiedrich, Mindy Hernandez, Stephen Roll, Yung Chun, Tim Lucas

About Prosperity Now Prosperity Now (formerly CFED) believes that everyone deserves a chance to prosper. Since 1979, we have

helped make it possible for millions of people, especially people of color and those of limited incomes, to achieve

financial security, stability and, ultimately, prosperity. We offer a unique combination of scalable practical

solutions, in-depth research and proven policy solutions, all aimed at building wealth for those who need it most.

To learn more about Prosperity Now, visit prosperitynow.org.

About Social Policy Institute at Washington University in St. Louis Combining lessons from areas such as economics, business, medicine, public policy, social work and sociology,

the Social Policy Institute (SPI) at Washington University provides a systems-level understanding of policy

problems and promotes systems-level solutions.

The Institute aims to build upon existing international collaborations and establish new partnerships, further

enhancing Washington University’s reputation as a global research leader. SPI values the robust array of its

partnerships—ranging from local St. Louis nonprofits, to national government agencies in the U.S. and abroad,

to corporations that affect the daily lives of people in communities.

About SaverLife (formerly EARN) SaverLife is a nonprofit on a mission — to inspire, inform and reward the millions of Americans who need help

saving money. Through engaging technologies and strategic partnerships, SaverLife gives working people the

methods and motivation to take control of their financial future.

SaverLife.org is designed for one purpose - increasing the savings rate for low- income Americans. On average,

savings rates increase 2.3x after joining SaverLife and 64% of members deposit $500 in their first six months.

Despite average household incomes of $30,000, SaverLife members save 11% of their income, compared to the

national average of 7.7%.

Acknowledgments We extend gratitude and thanks to the individuals and organizations who contributed to this research and report

for their efforts, participation and commitment to advancing the national discourse related to tax time savings

and the financial well-being of low-and-moderate income households. Thanks to the staff and volunteers at the

six participating VITA programs for graciously administering the pre- and post- surveys to their clients. And to

the survey respondents, who, by sharing insights into their financial lives, have enabled us to learn more than

we’ve ever known about tax time savings behaviors and financial well-being of LMI households.

Special thanks to our funder, Intuit Financial Freedom Foundation, for their unending support of the VITA field,

tax time savings, prosperity and the financial well-being of the most economically challenged in our society.


3


4

Table of Contents Executive Summary .............................................................................................................................5

Introduction........................................................................................................................................9

PART I: Background and Methodology ............................................................................................... 10

The Importance of Tax Time .......................................................................................................... 10

Research on Tax-Time Saving Supports .......................................................................................... 11

Overview of Partners/Tax Filing Platforms ....................................................................................... 14

Volunteer Income Tax Assistance (VITA) ..................................................................................... 14

Refund to Savings Initiative & Turbo Tax Free File Product ........................................................... 15

SaverLife ................................................................................................................................... 16

Methodology ................................................................................................................................. 16

Data Sources ............................................................................................................................. 16

Analytical Approach .................................................................................................................... 17

Demographic and Financial Characteristics Across Platforms ............................................................. 17

PART II. Savings Behaviors and Tax-Time Savings Interventions ........................................................ 20

Refund Savings Behaviors ........................................................................................................... 20

How Tax Filers Use Refunds ........................................................................................................ 24

What Happens after the Refund is Deposited? SaverLife Transactions............................................ 28

Examining Tax Time Savings Supports at VITA Sites..................................................................... 30

PART III: The Relationship between Refund Savings and Financial Well-Being ..................................... 33

Material Hardship ....................................................................................................................... 33

Dealing with Unexpected Expenses.............................................................................................. 35

Financial Well-Being ................................................................................................................... 37

Works Cited ...................................................................................................................................... 43

Appendix A: Literature Review .......................................................................................................... 45

Using Financial Incentives and Savings Matches to Promote Tax Time Savings ............................... 45

Using Behavioral Economics Nudges to Promote Refund Savings ................................................... 47

Promoting Different Savings Vehicles at Tax Time ........................................................................ 50

Appendix B: Enhanced Data Collection Project Tax-Time Survey .......................................................... 52

Appendix C: Comparison of Questions Included in VITA, R2S and SaverLife Data Sources .................... 53

Appendix D: VITA Programs included in the VITA Tax-Time Research Project ....................................... 55


5

Executive Summary

In this report, we explore the behaviors and outcomes related to savings and financial well-being of low-

and moderate-income (LMI) tax filers in the United States. We present findings from research conducted

by Prosperity Now, the Social Policy Institute at Washington University in St. Louis and SaverLife (formerly

EARN) during the 2019 tax season. Through surveys administered at or near the moment of filing taxes

and then three to six months later, we examined how tax filers used their tax refund and the role that the

refund plays in LMI filers’ financial lives and financial well-being. This analysis is unique in that it compares

tax filers’ outcomes over time across three different tax-filing and savings program platforms: volunteer

income tax assistance (VITA) sites, online tax filing through the Turbo Tax Free File Product (TTTFP) and

SaverLife’s saving program.

Overall, we find that LMI tax filers across all platforms are saving a portion of their tax refund at a high

rate and that this savings rate is substantially higher than recorded with previous measures of savings at

tax time. We also saw a relationship between saving and financial well-being: taxpayers who both intended

to save and successfully saved some of their tax refund experienced fewer financial hardships and increased

financial well-being in the months after tax filing. We also find that savings is just one of the ways that tax

filers are using their refund and it is often not the most important use of the refund. In fact, many filers

are using their tax refunds to “catch up” more often than to “get ahead” (i.e., pay bills, purchase household

necessities and pay down debt).

Participants in SaverLife reported the highest rates of saving, which we believe demonstrates that

participating in a savings program can support successful saving for those tax filers that are predisposed

to save. Despite these findings, as well as previous research on the effectiveness of tax-time savings

interventions, we found that savings and other financial capability interventions were not being

implemented systematically across VITA sites. In a survey of VITA Program Managers and staff, we found

a number of barriers to integrating savings and financial capability supports into VITA sites, including

prioritizing completing accurate tax returns over supporting savings and the perception from staff that VITA

filers are not interested hearing about financial products or services.

In presenting this analysis, we aim to provide useful insights to the tax preparation field on the barriers for

saving for LMI tax filers and suggestions on how these barriers can be overcome. While we do find high

rates of successful savings, there are a number of filers who intended to save their refund but did not save

once they received it. There is an opportunity to bridge this gap between intentions and action. We also

explore the ways in which a tax refund can impact families’ lives beyond its role in building savings. This

work thus intends to provide a foundation for a broader conversation around how the tax moment can be

used to support many different components of financial well-being.

Finding 1: LMI tax filers are saving a portion of their refund at higher rates than previously

thought or measured, but a number of filers who intend to save are not successful when they

get their refund.

We asked tax filers if they intended to save their refund and followed up to ask if they saved any of it after

they received the refund. When measured in this way, we found savings rates significantly higher than


6

previously measured through data on split refunds or deposits into savings accounts. Approximately 60%

of VITA and TTFFP filers intended to save their refund and roughly two-thirds of those with savings

intentions went on to successfully save. Savings rates were higher for SaverLife participants, with 93%

intending to save and 80% of those filers successfully going on to save a portion of their refund. However,

the rates of unsuccessful savers—those who intended to save but did not—were more similar across

platforms and varied from 14% to 21% of filers.

REFUND SAVING INTENTIONS VS BEHAVIORS, BY PLATFORM

Finding 2: Tax filers who successfully saved experienced fewer hardships and higher financial

well-being after filing their taxes

In the months after filing taxes, it was relatively common for tax filers to experience material hardships.

For example, 38% of TTFFP filers reported food insecurity, 31% of SaverLife participants skipped essential

bills and 23% of VITA filers skipped medical care or expenses. In addition, 22% of VITA filers reported that

they couldn’t manage a modest $400 emergency expense. We also found that tax filers who both intended

to save and successfully saved some of their tax refund experienced fewer financial hardships and increased

financial well-being in the months subsequent to filing their tax return. After controlling for financial well-

being at tax filing, we found that saving a portion of the refund and intending to save for longer than six

months were correlated with positive financial well-being changes in the months after tax filing.

39.3%

6.5%

40.2%

20.5%

14.3%

17.4%

40.2%

79.2%

42.4%

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

VITA SaverLife TTTFP

Res

po

nd

ents

(%

)

Intended to Save and Did Save

Intended to Save but Did Not Save

No Saving Intentions


7

REFUND SAVINGS INTENTION OR ACTION AND CHANGES IN FINANCIAL

WELL-BEING SCORE, BY PLATFORM (REGRESSION RESULTS)

*** p<0.01, ** p<0.05, * p<0.1

Finding 3: Savings is just one of the uses of the refund and not the most important use for

many LMI tax filers.

Tax filers reporting using their tax refund in a variety of ways. While savings for both emergencies and

long-term goals were among the top uses for filers in each platform, VITA and TTFFP filers were more

likely to report using the refund to purchase household necessities and pay down debt. Paying down debt

was also the second most common use for SaverLife participants. We find that this suggests that tax filers

are often using their tax refund to “catch up” on their finances rather than focusing on future needs and

“getting ahead.” When asked to rate the importance of these different uses of the refund, savings was

rated lower or at similar rates to other uses.

TOP USES OF REFUND, BY PLATFORM

0 1 2 3 4 5 6 7 8 9

VITA

SaverLife

TTFFP

Change in Financial Well-Being Score

Refund Saved at Follow-Up

Intend to Save > 6 mo

0% 10% 20% 30% 40% 50% 60%

VITA

SaverLife

TTFFP

Respondents (%)

Purchase household necessities(rent, bills groceries)

Pay down debt (credit card, pay day,past due bills)

Saving for emergencies orunexpected needs

Save for specific future purchase(education, home, vacation)

**

**

*

**

**


8

Finding 4: Despite evidence that interventions can support successful savings, there are many

barriers to integrating savings programs and other financial capability services into VITA sites.

We found that savings and other financial capability interventions were not being implemented

systematically across VITA programs and sites. Only three of 17 VITA sites surveyed through this research

were implementing financial capability services or providing access to financial products on site. Program

Managers and Site Coordinators reported a number of barriers to offering financial services, the most

prominent of which was a lack of training. Almost two-thirds (63%) of Site Coordinators reported that they

did not feel prepared to train volunteers on financial products and services. Tax-time savings were also not

a priority for Site Coordinators compared to preparing accurate tax returns for VITA filers. This lack of

supply of financial products and services at VITA sites does not match the demand for these services from

VITA filers. More than half of VITA filers expressed interest in some kind of financial service at tax time.

Additional income supports, emergency savings and help managing credit and debt were the three most

requested services.

HIGHEST PRIORITIES, VITA SITE COORDINATORS

0 2 4 6 8 10 12 14 16 18

Connecting to benefits

Increasing savings/asset buildng

Emergency savings

Increasing # of returns

Complex returns

Keeping up w/demand

Respect to clients

Accurate returns

No. of Responses


9

Introduction

The processes of filing a tax return and receiving a tax refund are important moments in households’

financial lives and have emerged as promising opportunities to promote savings and financial capability,

particularly in low- and moderate-income (LMI) households. The majority (roughly 80 percent) of US

households receive a tax refund (Internal Revenue Service, 2019). For many LMI tax filers, the tax refund

represents the largest single payment they may receive in a year (Roll, Davison, Grinstein-Weiss, Despard,

& Bufe, 2018), providing tremendous untapped potential to leverage the tax-time moment to improve LMI

taxpayers’ overall financial well-being and, at a minimum, use some of their refund to build emergency

savings.

A lack of emergency savings has become a cause for concern among American households. Many

households, especially those of limited means and people of color, could not cover even a modest $400

expense (Board of Governors of the Federal Reserve System, 2016). The purpose of this report is to better

understand the role that the tax refund and the tax moment more broadly can play in helping households

build emergency savings and make real changes in the financial well-being of economically vulnerable tax

filers. Through a collaboration across large-scale tax preparation and savings data platforms, we sought

answers to key savings research questions, to provide deeper insights into the financial well-being of LMI

households and to understand the barriers to building savings at tax time and how they might be overcome.

Specifically, this report explores the following key research questions:

➢ What are the savings behaviors of LMI tax filers and how are they supported by tax-time savings

interventions and supports?

➢ What is the impact of tax time savings on the financial well-being of LMI filers?

With support from the Intuit Financial Freedom Foundation, Prosperity Now, the Social Policy Institute (SPI)

at Washington University in St. Louis and SaverLife (formerly EARN) partnered to conduct research during

and after the 2019 filing season to answer these research questions and learn more about the savings

behaviors and financial well-being of LMI taxpayers. Through this partnership, we surveyed taxpayers who

filed their annual tax returns across several platforms: VITA sites, specifically member programs in

Prosperity Now’s Taxpayer Opportunity Network; TurboTax Free File Product (TTFFP), through SPI’s Refund

to Savings Initiative; and SaverLife.org, whose members filed their taxes using their preferred method,

including paid preparers. With data from these three populations, we were able to analyze and compare

tax time behaviors, savings and financial well-being from different cross-sections of the LMI tax-filing

population of the United States.

Through this research, we have gained insight into and highlight, several key findings, some of which dispel

previously held myths regarding savings behaviors of LMI populations and the benefits of the tax refund.

Other findings shed light on new or emerging opportunities for further exploration into ways to transform

the tax moment into a more meaningful and valuable experience for LMI households.


10

Several key takeaways from this research include:

1. Low-and-moderate income tax filers are saving at higher rates than previously measured.

2. Households use their tax refund in a variety of ways and for numerous reasons. Savings is just

one of the uses of the refund and not the most important one for many filers.

3. Savings interventions and supports offered at VITA sites are not being consistently implemented

across the field.

4. Taxpayers who both intended to save and, subsequently, successfully saved some of their tax

refund experienced fewer financial hardships and increased financial well-being in the months

after tax filing.

This report is divided into three sections. First, we share background information to set the stage for the

importance and value of this conversation, an overview of each of the participating partners in this research

and our research methodology. Next, Part Two deals with understanding the extent to which tax time

savings interventions and supports impact taxpayer savings behaviors and balances. In this section, we

share relevant historical research on the efficacy of tax time savings interventions, as well as several key

data points that point to some of our key findings and implications.

Part Three deals with furthering our understanding of the relationship between tax time savings and the

financial well-being of LMI taxpayers. In this section, we focus primarily on the results of our research on

self-reported taxpayer financial well-being as it relates to savings behaviors. We also extend this focus

beyond savings by examining other factors deemed significant (as reported by taxpayers) in influencing

the importance and use of their tax refund and its impact on household financial well-being.

Finally, we share a summary of our key findings and the implications for how the knowledge gained through

this research can be applied to further strengthen the value of the tax time moment for low-and-moderate

income households.

PART I: Background and Methodology

The Importance of Tax Time

Tax filing is a nearly universal experience in the U.S., with 150+ million individual tax returns filed in 2018

(Internal Revenue Service, 2019). This makes it one of the few touchpoints available to policymakers,

practitioners and researchers to reach a majority of U.S. households. As a result, even modest changes to

the tax code or the systems through which households file their taxes can have an impact on the finances

of millions of households.

The tax refund is very valuable to many American households and LMI households in particular. Through

a combination of over-withheld income and refundable tax credits like the Earned Income Tax Credit and

the Additional Child Tax Credit, the federal tax refund can be the single largest payment many LMI

households receive all year: A typical EITC recipient’s tax refund can account for one-fifth of their annual

income, or about three months of earnings (Sykes, Križ, Edin, & Halpern-Meekin, 2015). We also know the


11

tax refund is important to tax filers because of how well they remember the refund amount they received.

In 2017, SPI researchers asked TTFFP filers to recall how much their refund was six months after they filed

their taxes and compared those answers to their actual refund amount in administrative tax data. LMI tax

filers were very accurate in remembering their actual refund value.1 Indeed, 30% of tax filers were able to

remember the exact dollar amount of their refund six months later and almost 60% of filers recalled their

refund amount to within $100 of the actual amount.

The relative size and lump-sum delivery of the tax refund create an opportunity in which LMI households

can engage in a variety of financial behaviors that they cannot typically engage in throughout the rest of

the year. These behaviors include improving the household balance sheet through building their savings

and paying down debt (Grinstein-Weiss et al., 2015), as well as taking care of essential obligations like

managing overdue bills, home and car repairs, or making large purchases like appliances (Mendenhall et

al., 2012; Shaefer, Song, & Williams Shanks, 2013; Sykes, Križ, Edin, & Halpern-Meekin, 2015; Tach &

Greene, 2014). As such, interventions that increase the value of the tax refund to households (such as

providing financial incentives for saving the refund or providing access to financial products and services

at VITA sites) or that help households better manage and allocate their tax refund can lead to meaningful

improvements in the financial security and well-being of refund recipients.

There is also extensive evidence showing that providing financial support at tax time through credits like

the EITC is associated with an array of improved household outcomes, including:

➢ Increases in employment and reductions in poverty rates (Eissa & Liebman, 1996; Gundersen &

Ziliak, 2004; Meyer, 2010);

➢ Increases in savings and reductions in debt (Jones & Michelmore, 2018; Shaefer et al., 2013);

➢ Improvements in health and nutrition (Evans & Garthwaite, 2014; Hoynes, Miller, & Simon, 2015;

McGranahan & Schanzenbach, 2013); and

➢ Improvements in child achievement and educational measures (Dahl & Lochner, 2012; Manoli &

Turner, 2018).

From this, we can see that households use large tax refunds to address a wide array of needs in their lives,

ranging from the purely financial (building savings, debt management) to the types of investments that will

pay off over the course of generations (improved child health and educational outcomes).

Research on Tax-Time Saving Supports

The existing research shared in this section has been instrumental in informing and establishing a

foundation for the work done through this project. Utilizing the tax moment to promote financial capability

is still a relatively new area of study, but we know that interventions aimed at improving financial capability

through tax filing can work, even though much of the existing research is somewhat limited in scope. For

example, most of the research on maximizing the tax refund to help LMI households has happened in the

last decade and has typically focused on tax filing dynamics in a single setting (e.g., H&R Block, individual

1 The correlation between the actual refund and refund recollection six months later was extremely high (r=0.91).


12

VITA sites, TurboTax Freedom Edition, etc.) and has often examined a single savings outcome, such as

IRA contributions (Duflo, Gale, Liebman, Orszag, & Saez, 2006), savings bond purchases (Tufano, 2011)

and deposits into savings accounts to buffer against emergencies (e.g., Grinstein-Weiss et al., 2017; Roll,

Russell, Perantie, & Grinstein-Weiss, 2019).

The research and practice agenda emerging from this foundational work has led to several advancements

in the field. These include the Consumer Financial Protection Bureau’s establishment of tax time savings

cohorts and the development of “Ten Promising Practices” for VITA sites to increase savings at tax time

(Consumer Financial Protection Bureau, 2015), the development of a compendium of behavioral

interventions that can nudge savings at tax time through the Refund to Savings Initiative (Davison,

Covington, Kondratjeva, Roll, & Grinstein‐Weiss, 2018), the incorporation of savings interventions into

online tax filing platforms and the establishment of nonprofit programs aimed at facilitating tax time

savings—including SaverLife’s saving platform and Commonwealth’s SaveYourRefund initiative. Though

many of these innovations have focused on promoting savings in general, they often focus specifically on

building emergency savings through the use of tax refunds.

This focus on emergency savings is an important, if narrow, one. Why do we care about emergency

savings? Emergency savings is often a households’ first line of defense against the economic volatility that

disproportionately impacts LMI households and is associated with a variety of negative household

experiences (e.g., Leete & Bania, 2010). Conversely, even small amounts of emergency savings are

associated with a variety of improved household outcomes including reductions in hardships like food

insecurity, missing essential bill payments and skipping needed medical care (Babiarz & Robb, 2014;

Gjertson, 2016; McKernan, Ratcliffe, & Vinopal, 2009). Specifically focusing on tax time, refund savings

itself has also been linked with reductions in an array of hardships (Grinstein-Weiss et al., 2016). Given the

severe lack of liquid savings in LMI households and the demonstrated positive benefits of holding these

savings, the first priority of many tax-time financial capability programs is helping their populations save

for emergencies (Babiarz & Robb, 2014; Gjertson, 2016; McKernan, Ratcliffe, & Vinopal, 2009).

In Appendix A, we provide a detailed overview of the research on tax-time savings efforts and outcomes,

organized by the two primary means of promoting tax refund savings: the use of financial incentives or

matches to promote refund savings and the use of low-touch behavioral interventions to “nudge” tax filers

to save. We discuss some of the key lessons from this research here.

Matches for tax refund savings deposits provide a direct incentive for tax-time saving. For example, a

program may offer to match 50% of the refund amount still held in a qualified savings account for a set

period of time (e.g., receiving a $50 match for saving $100 of the refund for six months) or a refund

deposited into a qualified restricted investment account like an IRA (e.g., a $1,000 refund deposit to an

IRA at tax filing would receive an additional $500 deposit to that account). For example, the $aveNYC and

$aveUSA programs were major initiatives to test the effectiveness of savings matches at tax-time in

partnership with a variety of community organizations, including VITA sites, between 2009 and 2013. The

savings matches offered ranged between 50% and 80% and led to very encouraging results. For example,

80% of participants in the $aveNYC program saved for the full year required to receive the match and, for

most participants, it was their first time saving any of their refund. In the $aveUSA program, those who


13

received the savings match had over $500 more in savings by the end of the program than a control group

(Azurdia & Freeman, 2016; Key, Tucker, Grinstein-Weiss, & Comer, 2015).

LESSONS LEARNED FROM TAX TIME SAVINGS RESEARCH

The growing body of work on tax time savings research offers a number of lessons that can inform the

development and implementation of savings initiatives in tax filing organizations:

Lesson 1: Matched savings programs can work well at promoting tax-time savings, but the

match needs to be simple

Many of the matched savings programs that have been studied were very effective at driving both

short-term and long-term savings deposits, but one of the defining features of successful matched

savings initiatives appears to be their simplicity. Offering tax filers a simple match on a saved refund

(e.g. matching 50% of an IRA deposit) may be more attractive than tying the match to complex

programs (such as the Saver’s Credit) or by only offering a match if tax filers engage with some other

product (such as opening a savings account).

Lesson 2: Interventions that encourage splitting the refund into savings can be effective,

but overall splitting rates will likely remain low

Several of the interventions outlined in Appendix A showed that it is possible to increase the rate of

splitting the refund into savings using Form 8888. However, even in the most successful interventions,

the overall rate of refund savings remained low. This is because splitting the tax refund across multiple

accounts does not seem to appeal to the vast majority of tax filers; less than 1% of tax filers tend to

split their refund using Form 8888 in a given year (Treasury Inspector General for Tax Administration,

2015). As such, it may be beneficial to promote tax refund savings in ways that do not rely on Form

8888.

Lesson 3: Interventions that change the structure of decisions tend to be more effective

than interventions that change the environment

Interventions that change the environment experienced by tax filers—such as using targeted messaging

or providing general savings cues—can be effective. However, the effects of these interventions tend to

be relatively small in comparison to interventions that actually alter the structure of savings options. For

example, the Refund to Savings Initiative found that emergency savings messaging was incrementally

more effective at promoting savings deposits than other types of messaging. However, the positive

effects of emergency savings messaging were much smaller than the positive effects of making the

savings deposit option the first deposit option shown to tax filers. Similarly, the Common Cents Lab

found that providing table tents and altering intake forms to promote savings had small effects on

savings deposits, but actually getting tax filers to commit to saving their refund well in advance of

receiving it was much more effective. These findings indicate that changes made at the moment people

make decisions—either in shifting the decision to a more opportune moment or altering the way options

are presented—may be among the most effective tools available for promoting refund savings.


14

The second major approach to promoting tax time savings uses the techniques of behavioral economics to

motivate increased rates of refund savings through low-touch “nudges.” Behavioral economics draws on

both the fields of economics and psychology to better account for the complexity and nuances of human

decision-making than more traditional approaches. As such, it can help us understand how both external

environmental cues and unconscious mental processes can lead to a variety of different issues in financial

behavior—including issues like saving too little and taking on high-cost debt—which can be particularly

harmful to LMI households. Behavioral economics approaches may be especially attractive in the context

of promoting savings at tax time generally and at VITA sites specifically because they are designed to be

integrated into existing processes and practices without incurring substantial additional costs. Table A2 in

Appendix A outlines the array of behavioral interventions that have been tested in the context of promoting

savings at tax time. These interventions range from providing suggested savings amounts to encourage

filers to split their refund, to using different types of reminders to encourage additional EITC claiming, to

encouraging tax filers to pre-commit to save their refund before they even know the amount of their refund.

For example, the R2S Initiative has embedded behavioral interventions to promote savings into TTFFP for

several years. While many of the interventions were effective, the most effective approach by far was

restructuring refund deposit decisions to make savings deposits the first or “default” option. In one year,

this approach led to a 58% increase in refund savings deposits (Grinstein-Weiss et al., 2017)

Overview of Partners/Tax Filing Platforms

In this section, we provide brief overviews of the three platforms under study in this report: Volunteer

Income Tax Assistance sites, TurboTax Free File Product and SaverLife.

Volunteer Income Tax Assistance (VITA)

VITA programs provide free tax preparation and filing services to LMI households through a network of

community partners and volunteers. In 2019, thousands of IRS-certified volunteers working at the more

than 3,700 VITA sites throughout the United States provided free tax help to more than 1.5 million

households who generally made $56,000 or less, persons with disabilities and limited English-speaking

taxpayers who needed assistance in preparing their tax returns. Prosperity Now manages a national

coalition of VITA programs, the Taxpayer Opportunity Network, which has more than 4,500 members

representing more than 1,000 different organizations.

In 2019, Prosperity Now led a coalition of organizations to pilot consistent data collection on VITA tax filers

savings behaviors and financial well-being across VITA sites. Through this Enhanced Data Collection Project,

14 VITA programs administered a short survey to over 14,000 VITA clients during the tax filing process.

Prosperity Now invited six of the programs from this pilot to participate in further research and administer

follow-up surveys to their clients. In addition to surveying the tax filers at their VITA sites, Program

Managers and Site Coordinators from each participating VITA program were asked to complete a short

online survey to capture the challenges and opportunities of tax-time savings from the perspective of people

on the ground in VITA sites.


15

The six (6) VITA programs representing six states participating in this tax-time research are:2

1. Metro St. Louis Community Tax Coalition (St. Louis, MO)

2. Arkansas Asset Builders (Little Rock, AR)

3. Campaign for Working Families (Daytona Beach, FL)

4. Community Development Corporation of Brownsville (Brownsville, TX)

5. Accounting Aid Society (Detroit, MI)

6. Northern Virginia CASH (Annandale, VA)

Their participation and client responses have led to a greater understanding of the opportunities and

challenges involved in converting tax-time refunds into savings from the perspective of VITA clients and

VITA staff and use those insights to target best practices and suggest promising changes to VITA sites’

tax-time offerings. More information detailing the interactions, findings and results from VITA sites will be

shared in a report to be released in early 2020.

Refund to Savings Initiative & Turbo Tax Free File Product

Intuit’s TurboTax Free File Product (TTFFP)—formerly known as TurboTax Freedom Edition—is free tax

filing software offered to LMI tax filers as part of the IRS’ Free File Alliance. While the specific criteria to

qualify for this filing software varies slightly from year to year, it is typically offered to filers who either

make less than $33,000 in adjusted gross income per year or who qualify for the Earned Income Tax Credit,

with looser qualifications for households with active-duty military members.

Since 2012, the Refund to Savings (R2S) Initiative has worked with Intuit to embed behavioral interventions

promoting increased refund savings in this version of TurboTax. These interventions have ranged from

showing different types of messaging around the importance of savings to pre-selecting savings deposit

options for tax filers who pre-commit to save the refund. Many of these interventions have been successful

in driving increased savings deposits and the R2S Initiative has been instrumental in making tax time

savings a policy and program priority for many different stakeholders, including the Consumer Financial

Protection Bureau.

A key component of the R2S Initiative is the Household Financial Survey (HFS), which is a primary data

source for this report and the basis for the surveys administered to the VITA filers and SaverLife participants

featured in this report. There are two waves of the HFS administered every year. For the first wave of the

survey, TurboTax filers are invited to take the survey immediately after they file their taxes. These filers

are then invited back for a second wave of the survey six months later. The indicators collected as part of

the HFS change each year but, in general, the survey captures detailed information pertaining to tax filers’

refund usage, assets, liabilities, demographic traits, financial behaviors and economic circumstances.

2 For a complete list of all VITA sites and the populations served at each, please see Appendix C.


16

SaverLife

SaverLife is an innovative online platform that helps low-income households achieve financial security.

SaverLife leverages the power of technology to make savings available to anyone, anywhere, anytime.

Based on principles of human-centered design and behavioral economics, SaverLife increases the financial

stability of low-income households by providing:

➢ A prize-based savings program that rewards users for saving regularly ➢ Opportunities to win prizes for saving with initiatives like “Savers Win” during tax-time and

“Scratch and Save” where Savers can win $5 for saving $5 each week ➢ Financial education content developed for low-income families by a financial coach ➢ Access to quality financial products and services ➢ SMS/Text notifications to keep savings top of mind

Anyone can sign up for SaverLife at saverlife.org. Participants typically come to SaverLife through a variety

of marketing channels such as affiliate partnerships, organic web traffic, partnerships with direct service

provider organizations and employers.

Methodology

Data Sources

Each partnering organization in this project surveyed tax filers at or near the time that they filed their 2018

tax returns and administered follow-up surveys to tax filers three to six months after the respondents filed

their taxes. See Table 1 for an overview of the similarities and differences in the surveys administered by

each organization. The survey administered during or near tax time at VITA sites and by SaverLife was a

ten-question survey developed for the Enhanced Data Collection Project, which covered their refund savings

intentions, aspects of their financial circumstances and behaviors and their financial well-being as measured

through the CFPB’s five-item financial well-being scale (see Appendix B). Respondents at VITA sites

completed the survey in person while filing their taxes and SaverLife members completed the survey online.

The survey for TTFFP filers was administered in a somewhat different way. After filing their taxes through

TTFFP, a third of tax filers were randomly invited to take the Household Financial Survey (HFS) online. This

survey was much more extensive than the surveys offered at the VITA and SaverLife platforms, although

it included the same questions found in the shorter surveys used by the other two platforms.

The survey administered in the second wave of data collection to VITA tax filers and SaverLife members

was designed jointly by Prosperity Now, SPI and SaverLife and again was adapted from the HFS (see

Appendix C). In the months after tax filing, respondents to the first wave of the survey in each platform

were invited to complete a second survey. SaverLife and TTFFP filers were contacted on a rolling basis

roughly six months after they completed their taxes and completed the survey online. However, to minimize

the administrative burden on individual VITA sites, all VITA filers were contacted within the same time

period in July. As such, VITA tax filers who filed in January were contacted roughly six months after filing,

while VITA filers who filed in April were contacted roughly three months after filing. On average, VITA filers

were contacted around four months after filing and the surveys were administered by phone or online.

Taxpayers responding to the survey across any of the platforms received a $5 gift card for participating.

http://www.saverlife.org/


17

In addition to surveying its members, SaverLife also analyzed bank account transaction-level data of a

select group of its members to understand changes to spending and savings habits3. SaverLife is able to

access this data through linkages created to bank accounts which SaverLife uses to monitor savings rates

and provide financial and other incentives. We use this data set to supplement our core analysis and provide

additional detail on the refund usage behaviors of tax filers.

TABLE 1 | Survey Administration and Sample Overview

Platform VITA SaverLife TTFFP

Wave 1

Sample Size 579 653 14,928

Timing During tax filing Near tax time Following tax filing

Mode In-person Online Online

Wave 2

Sample Size 171 106 2,615

Timing 3-6 months after filing 6 months after filing 6 months after filing

Mode Phone/online Online Online

Analytical Approach

Most of the analyses in this report rely on simple descriptive statistics to illustrate the characteristics of tax

filers and the patterns we observe in tax filing behaviors both during and after tax filing. However, to

investigate the relationship between tax refund savings behaviors and financial well-being, we use

regression approaches that allow us to better estimate the relationships we want to explore. Specifically,

we use ordinary least squares regression to estimate the relationship between tax refund savings behaviors

and financial well-being at wave 2 of the survey, while controlling for financial well-being at wave 1.

Demographic and Financial Characteristics Across Platforms

Table 2 examines the demographic characteristics of tax filers in each platform. VITA filers tend to be older

and have less educational attainment than filers in other platforms, while TTFFP filers tend to be single at

higher rates and exhibit more gender balance than filers in the other platforms. VITA filers and SaverLife

members are majority people of color, while TTFFP filers are almost three-fourths White. Both SaverLife

and TTFFP filers have higher educational attainment than VITA filers, with over 40% having at least a

bachelor’s degree, compared to VITA’s 24%.

3 These members are not necessarily the same members who received the survey though their population statistics are largely

similar to what is reported later in this report.


18

TABLE 2 | Demographic characteristics, by platform


mean/pct mean/pct mean/pct

Female 68.3% 77.1% 50.8%

Age 51.6 37.5 35.1

Race

White, Non-Hispanic 39.0% 43.4% 72.3%

Black, Non-Hispanic 21.4% 30.1% 5.4%

Asian, Non-Hispanic 0.5% 5.8% 6.5%

Latino 37.1% 9.7% 9.8%

Other 1.9% 11.1% 6.1%

Marital Status

Single 37.2% 41.2% 72.4%

Married 25.1% 37.7% 10.6%

Separated 3.4% 2.6% 1.2%

Divorced 25.1% 14.0% 12.9%

Widowed 9.2% 4.4% 2.9%

Live with Partner 29.3% N/A 16.3%

Highest Education

Less than high school degree or

GED

8.7% 1.8% 2.5%

High school degree or GED 37.5% 13.2% 12.5%

Certificate or Associate degree 29.8% 8.8% 12.4%

Bachelor's degree 15.9% 22.8% 25.5%

Graduate or professional school 8.2% 19.3% 18.4%

Observations 1,016 653 14,928

Table 3 examines the financial characteristics of filers in each platform. More than half of filers in each

platform made $30,000 or less in the prior year. By comparison, the median household income in the US

in 2018 was $63, 179 (Rothbaum & Edwards, 2019). SaverLife filers were much more likely than filers in


19

other platforms to report very high and very low incomes, while the incomes of VITA and TTFFP filers were

more evenly distributed.

TABLE 3 | Financial characteristics, by platform


mean/pct mean/pct mean/pct

Income

0 to $5K 7.5% 31.9% 16.8%

$5K+1 to $10K 8.5% 4.5% 11.6%

$10K+1 to $15K 14.6% 10.7% 12.6%

$15K+1 to $20K 13.6% 6.8% 12.3%

$20K+1 to $30K 23.6% 12.0% 23.9%

$30K+1 to $40K 19.1% 8.6% 12.2%

>$40k+1 13.0% 25.5% 10.5%

Own home 48.0% N/A N/A

Income Volatility

No Volatility 76.7% 61.4% 62.3%

Some Volatility 12.7% 22.9% 25.6%

High Volatility 10.6% 15.7% 12.1%

Have any credit card 63.3% 75.0% 64.0%

Paid credit card in full 38.5% 45.4% 43.0%

Any alternative finance

services 6.6% 18.7% 19.5%

Observations 1,016 653 14,928

VITA filers were the most likely to report no volatility in their incomes whereas SaverLife filers were the

most likely to report high levels of income volatility. Majorities of tax filers across all platforms reported

credit card ownership and around 40% of credit card owners across platforms reported paying their credit

card in full every month, indicating that the majority of filers typically accrue at least some interest charges

in a given month. Finally, the rate of alternative financial service use was much higher among filers in

SaverLife and TTFFP platforms (almost 20%) compared to those in the VITA platform (around 7%).


20

PART II. Savings Behaviors and Tax-Time Savings

Interventions

In this section, we examine the refund savings intentions and behaviors of tax filers across platforms, the

extent to which tax time savings interventions and supports impact taxpayer savings behaviors and

balances and the varied use and relative importance of the tax refund to LMI households for specific

purposes including, but not limited to, building savings.

Refund Savings Behaviors

Table 4 outlines the differences in refund savings behaviors among filers in the three platforms. The

overwhelming majority of tax filers in each platform expected to receive a refund, ranging from 81% of

TTFFP filers expecting a refund to 88% of SaverLife filers expecting a refund. Refund savings intentions

(measured at tax filing) of tax filers vary substantially across platforms. Unsurprisingly, given its function

as a savings program, SaverLife filers were the most likely to intend to save; only 7% reported no refund

savings intentions whatsoever. Savings intentions of VITA and TTFFP filers were much more similar, with

61% of VITA and 60% of TTFFP filers expressing plans to save at least some of their refund. When filers

did intend to save, they were much more likely to plan to save for longer periods. Around three-fourths of

those who intended to save a portion of their refund planned to do so for more than six months.

Actual refund savings (measured at follow-up) also varied by platform. We define the savings rate as the

percentage of tax filers who reported receiving a refund to answered positively to the question: “Upon

receiving your tax refund, did you save any of it?” As with savings intentions, the saving rate for SaverLife

KEY FINDING

LMI tax filers are saving at higher rates than data have shown

previously

Contrary to previous research and popular understanding, our research

has shown that LMI filers, irrespective of income, want to, and do save a

portion of their tax refund. Traditional measures of savings, including

measuring rates of refund splitting (through IRS Form 8888) and savings

deposits, only capture a small fraction of overall refund savings

behaviors.


21

members was considerably higher at 84%, but over half of VITA and TTFFP filers reported saving a portion

of their refund.

TABLE 4 | Refund characteristics, by platform

VITA SaverLife TTFFP

Tax Time

Expect refund 85.1% 87.5% 81.1%

Refund saving intentions

No saving 39.3% 6.5% 40.2%

Intend to Save 60.7% 93.5% 59.8%

Save but spend < 6

months

17.3% 19.0% 12.7%

Save > 6 months 43.4% 74.5% 47.1%

Refund saving locations

Cash 4.2% 10.3% 5.7%

Checking 39.5% 14.9% 33.4%

Savings 49.6% 63.2% 66.6%

Pre-Paid 1.2% 6.2% 1.2%

Retirement 1.2% 5.4% 5.2%

Other 4.2% N/A 3.3%

Follow-up (3 – 6 months post-filing)

Refund saving rate

(actual) 53.2% 84.0% 55.0%

Observations* 1,012

(171)

655

(106)

14,814

(2,625)

* Observations for the follow-up survey in parentheses.

There are also interesting patterns in the locations tax filers decide to place their savings. Notably, we see

that VITA filers are more likely to save outside of a traditional savings account. Whereas about two-thirds

of SaverLife and TTFFP filers saved their refund in a savings account, less than half of VITA filers did so.

These filers were much more likely to save their refund in a checking account than filers in the other

platforms. This finding is important because one of the most common ways of measuring “savings” at VITA

sites is through savings account deposits, but here we see evidence that this metric may be severely

undercounting the actual rate of refund savings at these sites.

These results also speak to the risks and limitations of using refund “splitting”—or depositing part of the

refund into a savings account and part of the refund in some other account—as a measure of savings,


22

which is a common practice in VITA sites. Analyses of IRS data have found that the overall refund splitting

rates among VITA clients in the U.S. is extremely low, at 0.55% in 2015 (Treasury Inspector General for

Tax Administration, 2015). The gap between the split rate and the refund savings rate (<1% versus 53.2%

in this study) may indicate that traditional measures of refund savings are significantly undercounting

savings rates.

These results also highlight that another key refund savings metric—the depositing of the refund into a

savings account—may also be substantially undercounting the actual rate of refund savings among filers.

For example, 9.3% of TTFFP users deposited any of their refund into savings in 2017 (Roll et al., 2019).

By contrast, our results show that 60% of TTFFP users intended to save the refund at tax filing and 55%

reported still having some savings at follow-up. The gap between the observed rate of refund savings

deposits and actual savings intentions and behaviors means that the reliance on refund savings deposits

as a measure of savings may exclude the vast majority of refund savers.

We next examine the relationship between savings intentions (measured at tax filing) and actual savings

(measured 3-6 months later, at follow-up). Figure 1 shows the likelihood of tax filers falling into one of

three groups in each tax filing platform: Those who had no savings intentions at tax filing, those who

intended to save at tax filing and reported no savings at follow-up (unsuccessful savers) and those who

intended to save and did so (successful savers).

FIGURE 1. REFUND SAVING INTENTIONS VS BEHAVIORS, BY PLATFORM

Sources: VITA Tax Time Survey (N=825); SaverLife Tax Time Survey (N=573); HFS 2019 (N=12,277)

As with savings intentions, SaverLife filers were more likely to be successful savers, with almost 80% of

filers following through on their savings behaviors. Savings behaviors among filers at TTFFP and VITA

39.3%

6.5%

40.2%

20.5%

14.3%

17.4%

40.2%

79.2%

42.4%

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

VITA SaverLife TTTFP

Res

po

nd

ents

(%

)

Intended to Save and Did Save

Intended to Save but Did Not Save

No Saving Intentions


23

platforms were similar to each other, with 40% of VITA filers and 42% of TTFFP filers successfully saved,

which translates to roughly two-thirds of those who intended to save tax filing successfully doing so. Put

differently, both TTFFP and VITA filers were as likely to successfully save their refund as they were to not

save the refund at all, in contrast to what would observe if we only counted refund splitting or refund

savings deposits as “saving.” Across all platforms, the rates of unsuccessful savers were fairly similar: from

14% to 21% who intended to save at tax time failed to do so.

Table 5 examines the characteristics that predict tax filers successfully following through on their savings

intentions in each platform. Generally, speaking, successful savers were those that had relatively stable

financial circumstances in the months after tax filing, who earmarked the refund specifically for different

savings purposes and who did not report relying on high-cost sources of credit like payday loans.

TABLE 5 | Characteristics of a successful saver, by platform

A successful refund saver at

VITA sites…

A successful refund saver at

SaverLife…

A successful refund saver in

TTFFP…

Did not experience high

levels of income volatility

Was older on average Was slightly more likely to

be Non-Hispanic White or

Asian

Explicitly saved their refund

for emergencies


for emergencies or for

specific purposes like

education

Was slightly more likely to

be married and had at least

a bachelor's degree

Did not report using credit-

based alternative financial

services

Did not report using their

refund for household

necessities or healthcare

expenses

Saved their refund in a

savings or retirement

account

Did not save their refund on

a pre-paid card


for emergencies or for

specific purposes like

education

Did not experience any

hardship in the months after

tax filing

Had higher income and less

income volatility



services

Could already manage an

emergency expense prior to

receiving the refund

Did not experience any

hardship in the months after

tax filing


24



services

How Tax Filers Use Refunds

Figures 2 and 3 demonstrate the ways in which tax filers reported using their tax refund beyond just

savings, as well as the importance to which they assigned a given use of the refund. At both TTFFP and

VITA platforms, the most commonly reported use of the refund was to pay for household necessities,

followed by paying off debts. Though using the refund to save for emergencies, specific goals, or large

purchases was common at both TTFFP and VITA platforms, the rate of saving for these purposes was much

higher for TTFFP filers than for VITA filers. Roughly 10-15% of VITA and TTFFP filers reported using the

refund for more niche purposes like healthcare expenses or making advance payments on their expenses.

The most common use of the refund by SaverLife filers, by contrast, was for emergency savings, potentially

indicating that these filers may have fewer urgent needs than VITA and TTFFP filers and can thus dedicate

more of their refund to building a savings buffer.

KEY FINDING

Savings is just one of the uses of the refund, and not the most

important one for some of the populations

Households use their tax refund in a variety of ways and for numerous

reasons which can lead to higher financial well-being scores. Though

both short- and long-term savings of the refund was a common usage,

these purposes were not the most common nor, in many cases, the most

important usage of the refund. We found that many LMI households are

using their tax refund to "catch up" rather than "get ahead".


25

FIGURE 2. REFUND USAGE BY PLATFORM

Sources: VITA Follow-up Survey (N=173); SaverLife Follow-up Survey (N=106); HFS 2019 (N=2,616)

Immediately after filers were asked about the different purposes for which they used their refund, they

were asked about the importance of each use they identified. Figure 3 illustrates the average importance

assigned to each usage, where a score of 0 represents “not at all important” and a score of 4 represents

“very important”. These results show that VITA filers consistently rated each use of the refund as being

similarly important—their scores indicate a perceived importance somewhere between “important” and

“very important.” There was more diversity in TTFFP and SaverLife filers. TTFFP filers rated paying down

debt as the most important use of the refund (a score of 3.37) and saving for emergencies as among the

least important uses of the refund (a score of 2.61). SaverLife filers, by contrast, rated “other” uses of the

refund as the most important use to them and covering healthcare expenses as the least important use

(though only six SaverLife filers in total selected indicated they used the refund for healthcare).

0% 10% 20% 30% 40% 50% 60%

VITA

SaverLife

TTFFP

Respondents (%)

Purchase household necessities (rent,bills groceries)

Pay down debt (credit card, pay day,past due bills)

Saving for emergencies or unexpectedneeds

Save for specific future purchase(education, home, vacation)

Make large purchase (phone, vehicle,car repair)

Cover healthcare expenses (Doctorvisits, prescriptions)

Make payments in advance (futurerent or bills)

Others


26

FIGURE 3. REFUND IMPORTANCE BY PLATFORM


We also asked filers to tell us what their most important use of the refund was in their own words. Figures

4a, 4b and 4c present these results for filers in each platform. Across all three platforms, we observe some

commonalities: general uses like bills, savings and debt, are prominently reported as the most important

usage of the refund across all sites, as is the usage of the refund for housing expenses. However, we also

see several interesting differences between the sites. For TTFFP filers, shown in Figure 4c, we see that

general savings is as prominent as any other usage, while savings is less prominent for SaverLife and VITA

filers. Conversely, paying bills appear to be less commonly reported as the most important usage of the

refund for TTFFP filers. VITA filers also commonly reported healthcare as the most important usage of the

refund, possibly due to the fact that VITA filers tended to be older. Paying off property taxes was also

commonly reported as the most important usage among SaverLife and VITA filers.

0 0.5 1 1.5 2 2.5 3 3.5 4

Others

Make payments in advance (future rent or bills)

Cover healthcare expenses (doctor visits, prescriptions)

Make large purchase (phone, vehicle, car repair)

Save for specific future purchase (education, home,vacation)

Saving for emergencies or unexpected needs

Pay down debt (credit card, pay day, past due bills)

Purchase household necessities (rent, bills, groceries)

Importance (0: Not at all importance; 4: Very important)

TTFFP

SaverLife

VITA


27

FIGURE 4C. THE MOST IMPORTANT

USAGE OF THE TAX REFUND, TTFFP

FIGURE 4B. THE MOST IMPORTANT

USAGE OF THE TAX REFUND, SAVERLIFE

FIGURE 4A. THE MOST IMPORTANT

USAGE OF THE TAX REFUND, VITA


28

What Happens after the Refund is Deposited? SaverLife Transactions

One advantage of SaverLife data is that it allows us a rare look into tax filers’ bank accounts after the

refund is deposited. Using these data, we can not only see the rate of savings deposits and withdrawals

for households receiving tax refunds, we can also see how spending changes for these households after

they receive the refund. The results of these analyses are shown in Figure 5.

SaverLife has observed that about half of refunds are placed in savings accounts upon receipt. SaverLife

members then tend to make net withdrawals from savings accounts in the months following saving their

refunds. This corroborates the data showing SaverLife members use refunds for emergencies, as savings

balances that were inflated from the receipt of refunds act as a lifeline to ward off emergency expenses or

income dips in the months following the receipt of a refund.

FIGURE 5. CHANGES IN SAVINGS BALANCES AFTER RECEIVING THE REFUND

AMONG SAVERLIFE FILERS

For refund savers receiving refunds in February, the combined amount of saving and spending increased

95% versus only two percent for the entire SaverLife member population. In fact, transfers to savings

accounts increased 191% during the month, an average of $1,296 per saver, or 48% of refunds received.

February refund recipients also used the influx of funds to “catch up.” Comparing February spending to the

average of the prior three months, the data show a 53% increase in utility payments, a 61% increase in

credit card payments and a 43% increase in telecommunications payments. Payments to municipal

government departments, such as the DMV or city revenue departments, increased by 87%. And while

spending on automobile maintenance only increased by 10%, this is significantly different from the overall

population, where spending in the same category fell by 20%.

$0

$500

$1,000

$1,500

$2,000

$2,500

$3,000

$3,500

Jan Feb Mar Apr May Jun

Savings Deposits

Savings Withdrawals

Net Change

Changes in Savings

Balances Since 1/31 for

People Receiving

Refunds in February


29

FIGURE 6. REFUND SPENDING AMONG SAVERLIFE FILERS: FEBRUARY

SPENDING VS. PRIOR THREE-MONTH AVERAGE

-20%

3%

6%

-7%

14%

-11%

10%

43%

53%

61%

87%

98%

Automobiles

Cellular Services

Utility Payments

Credit Card

Government

Healthcare

February Refund Recipients

Total Population

KEY FINDING

Savings interventions and supports are not being consistently

offered or implemented at VITA sites.

VITA programs across the board do not systematically promote savings or

offer additional products and services at their sites. Despite client interest in

learning more about additional financial tools and resources, each site’s

offering of savings interventions and supports is dependent primarily on the

willingness and capability of the Site Coordinator and volunteers, and the

relative importance they place on those interventions. Program Managers

and Site Coordinators alike generally lack confidence in their ability to train

volunteers and/or talk with taxpayers about savings or other financial

supports at tax time.


30

Examining Tax Time Savings Supports at VITA Sites

Research on tax-time savings—referenced on page 12 and Appendix A of this report—and by the CFPB and

others posits that offering financial programs and services at VITA sites in the right ways (see the CFPB’s

Ten Promising Practices) should result in increased clients’ savings. We explore this hypothesis by looking

at whether clients who file taxes at VITA sites that offer financial services are more successful savers.

However, we find no correlation between products and services offered at VITA sites and corresponding

savings by VITA clients. This is likely due to the small number of VITA sites offering financial products and

services in a systemic way.

As detailed in Figure 7, only three of the seventeen sites that responded to our survey are offering bank

accounts, IDA accounts and financial education systematically.

FIGURE 7. FINANCIAL PROGRAMS/PRODUCTS OFFERED AT VITA SITES

Source: Survey of VITA Site Coordinators (N=17)

The lack of supply does not reflect a lack of demand. We find a significant and unmet, interest in financial

products and services among VITA clients. Our research indicates that more than half of all VITA clients

are interested in some kind of financial services with support for monthly income, emergency savings and

help managing credit and debt coming in as the three most requested services.

0 2 4 6 8 10 12 14 16 18

529 accounts/children savings

Small business

Assist w/public benefits

Credit counseling

Credit reports

Financial coaching

Financial education

IDAs

Checking/savings accounts

No. of Responses

Offer systematically Offered on-site or by partner on-site Not offered in systemic way


31

FIGURE 8. VITA CLIENTS’ INTEREST IN FINANCIAL PRODUCTS AND

SERVICES AT TAX TIME

Source: VITA Follow-up Survey (N=173)

VITA staff report significant barriers that keep them from offering financial services at tax time despite

client interest and good research on the benefits of offering such services:

• Site Coordinators and volunteers are not prepared or trained to discuss financial products or

services. A surprising 94% of Site Coordinators only feel somewhat prepared or not at all

prepared to train volunteers on financial products and services, with 63% reporting feeling not all

prepared. Similarly, when asked about challenges to implementing the CFPB’s promising

practices, Site Coordinators report a lack of “trained dedicated staff” and “lack of training to know

what to offer” as the top two barriers.

• Tax-time savings is not viewed as a priority for Site Coordinators. As detailed in Figure 9, filling

out accurate returns and showing respect to clients are most cited by Site Coordinators as their

“top priorities”. Site Coordinators most commonly reported that using tax time to build

emergency savings, connecting clients to financial services and using refunds to help clients build

savings were a “low priority or not possible at this time”.

• Site Coordinators and Program Managers incorrectly perceive that VITA clients are uninterested

in tax-time savings. All Program Managers surveyed reported that a “huge barrier” to offering

financial products/services is a lack of interest from their clients.

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

Current Financial Goals

Long-term Financial Goals

Manage Credit and Debt

Emergency Savings

Support Monthly Income

Very interested Somewhat interested Not interested


32

FIGURE 9. VITA SITE COORDINATOR PRIORITIES AT TAX TIME

Source: Survey of VITA Site Coordinators (N=17)

For VITA clients, the two most often cited barriers to learning more about savings and other financial

products and services include the stress of tax time and the public nature of VITA sites. Importantly, while

VITA staff viewed client time as a key barrier, roughly 60% of clients report that time is not a key barrier.

Finally, we find that there are promising and yet unrealized, ways of offering these services at VITA sites

in the future. For example, when we look at interest in financial products by savings intention, we see that

VITA clients with an intention to save are significantly more interested in every type of financial service. In

addition, we find that VITA clients expecting a refund are more comfortable talking about financial products

and services and clients planning to save find tax-time much less stressful. Therefore, VITA sites might be

able to maximize their limited staff by focusing services on those clients (those with an intention to save

and refund expectations) who are most likely to be interested in financial services.

Specific types of support would also help staff offer financial programs at VITA sites and apply the

evidenced-based, tax-time savings suggestions detailed in Part 1 of this report. Notably, volunteers

dedicated to tax-time savings, trainings on different financial products and services and reliable partners

would help VITA staff offer financial products and services successfully. Site Coordinators and Program

Managers list “dedicated volunteers just for this (savings) purpose” and “dedicated reliable partners” as

the top two specific supports that would make offering products/services at tax time more successful. For

Site Coordinators, “more training” is the third most requested support.

0 2 4 6 8 10 12 14 16 18

Connecting to benefits

Increasing savings/asset buildng

Emergency savings

Increasing # of returns

Complex returns

Keeping up w/demand

Respect to clients

Accurate returns

No. of Responses


33

PART III: The Relationship between Refund

Savings and Financial Well-Being

In this section, we examine the relationships between refund savings behaviors and three different

measures of “financial well-being”: the experience of different types of hardship, the ability to access $400

in an emergency and the financial well-being score developed by the Consumer Financial Protection Bureau.

Material Hardship

The experience of hardships is fairly common among filers across all platforms, particularly given the fact

that our hardship questions only ask about hardships experienced since households filed their taxes (over

the prior 3-6 months). Since filing their taxes, it was relatively common for these households to experience

food insecurity, overdraft accounts or borrow using alternative financial services, skip medical care, or skip

payments on essential bills. For example, over 30% of SaverLife filers reported skipping essential bills since

tax filing and around 20% of VITA and TTFFP filers reported skipping essential bills and skipping medical

care since filing taxes. While skipping rent or mortgage payments was less common among tax filers across

all platforms, it was not a rare occurrence. In particular, over 10% of SaverLife filers reported skipping a

KEY FINDING

Those who successfully saved their refund experienced fewer

hardships and increased financial well-being after filing their

taxes

Taxpayers who intended to save, and successfully saved some of their

tax refund experienced fewer financial hardships and increased financial

well-being in the months subsequent to filing their tax return. While

savings intentions and actions are correlated with lower rates of hardship

and improved financial well-being, hardships are a common experience

for many households in these platforms in the months after tax filing.

For example, over a third of TTFFP filers reported food insecurity, and

roughly a third of SaverLife filers reported skipping essential bills.

However, VITA filers had the highest rates of reporting that they could

not manage a modest $400 emergency expense, but they also had the

highest reported rates of financial well-being.


34

housing payment since filing their taxes. TTFFP filers also reported very high levels of food hardship—for

example, skipping meals or reducing food consumption due to affordability—relative to filers at the other

platforms. Almost 40% of TTFFP filers reported this hardship, roughly double that reported by filers at

other platforms.

FIGURE 10. THE EXPERIENCE OF HARDSHIPS, BY PLATFORM


FIGURE 11

Figure examines the relationship between refund savings behaviors (both savings intentions and savings

actions) and the experience of any hardship. We find that both intending to save the refund and actually

saving the refund were generally associated with lower levels of hardship in the months after tax filing.

This is to some degree unsurprising, as households who have the facility to save at least some of their

refund are likely going to be those households who have relatively stable, sustainable financial situations

and may thus experience lower rates of hardship.

However, we also observe that both VITA and TTFFP filers with short-term savings intentions (<6 months)

were as likely to experience hardship as those filers with no savings intentions. It may be that certain tax

filers anticipate future economic volatility (and thus the potential for hardship) and so only plan to save for

the short-term, which may result in short-term savers exhibiting similar hardship patterns as non-savers.

0% 5% 10% 15% 20% 25% 30% 35% 40%

VITA

SaverLife

TTFFP

Respondents (%)

Overdrafted a bank account

Not affordable essential food

Skipped medical care

Skipped bills

Skipped rent


35

FIGURE 11. THE RELATIONSHIP BETWEEN REFUND SAVINGS BEHAVIORS

AND HARDSHIP

Sources: VITA Follow-up Surveys (N=119); SaverLife Follow-up Surveys (N=102); HFS 2019, (N=2,404)

Dealing with Unexpected Expenses

We also asked tax filers about the ways in which they would manage a $400 emergency expense at both

tax time and in the follow-up survey, as shown in Figure 12. Encouragingly, the most common way tax

filers reported managing this small expense at both tax time and follow-up was through resources that

would not require them to take on debt, such as liquid assets (for example, cash or money held in a

checking or savings account); though VITA filers were the least likely to report being able to manage an

emergency in these ways. VITA filers also exhibit a concerning lack of resilience to emergencies— roughly

20% of VITA filers could not manage a $400 expense if it occurred at tax time, which is much higher than

their TTFFP counterparts. Additionally, SaverLife filers are slightly more likely to report not being able to

manage this expense at follow-up than they were at tax time, possibly implying that a portion of SaverLife

filers had drained the resources provided by the refund by the time of the follow-up survey.

20%

30%

40%

50%

60%

70%

80%

No Savings Saving < 6 mo Saving > 6 mo No Saved Saved

Intention Action

Exp

erie

nce

d A

ny

Har

dsh

ip (

%)



36

FIGURE 12. ACCESS TO $400 IN AN EMERGENCY, BY PLATFORM (TAX TIME

AND FOLLOW-UP)

Sources: VITA Surveys, Tax Time (N=941) and Follow-up (N=216); SaverLife Surveys, Tax Time (N=593) and Follow-

up (N=132); HFS 2019, Tax Time (N=12,758) and Follow-up (N=2,944)

In Figure 13, we further explore tax filers’ emergency resources by examining the relationship between

refund savings intentions and being able to manage a $400 emergency expense. The primary takeaway

from this figure is that there is a notable correlation between refund savings intentions and the inability to

manage $400 in an emergency. For example, 30% of VITA filers with no savings intentions reported being

unable to manage a $400 expense, which was almost double the rate for VITA filers who intended to save

the refund for six months or more. At the same time, refund savings intentions do not appear to be strongly

predictive of managing emergency expenses in the future; the rates of being unable to manage an

emergency expense are similar at tax time and follow-up for TTFFP and VITA platforms, as well as for

SaverLife filers who planned to save for six months or more.

0%

10%

20%

30%

40%

50%

60%

70%

80%

Tax Time Follow-up Tax Time Follow-up Tax Time Follow-up

VITA SaverLife TTFE

Res

po

nd

ents

(%

)

Manage Expense Without Borrowing Manage Expense With Borrowing Could Not Manage Expense


37


AND ACCESS TO EMERGENCY RESOURCES, BY PLATFORM

Sources: VITA Surveys, Tax Time (N=807) and Follow-up (N=169); SaverLife Surveys, Tax Time (N=522) and Follow-up (N=106); HFS 2019, Tax Time (N=10,767) and Follow-up (N= 2,625)

Financial Well-Being

FIGURE Figure 14 presents the financial well-being scores of tax filers—as measured by the CFPB’s

financial well-being scale at both tax time and follow-up—based on the refund savings intentions of those

tax filers. The first thing we observe is that there is a strong correlation between refund savings intentions

and financial well-being scores at tax filing. For example, VITA filers who had no savings intentions had a

financial well-being score of 50 at tax time, while those who planned to save for more than six months had

a financial well-being score of 55. The second thing we observe is that SaverLife and VITA filers exhibit

similar financial well-being patterns across survey waves. Households who did not plan to save or who

planned to save for the short-term exhibit declines in financial well-being in the months after tax filing,

while households who planned to save for more than six months saw a slight uptick in financial well-being.

Interestingly, financial well-being for TTFFP filers was largely flat across survey waves, regardless of savings

intentions.

0%

10%

20%

30%

40%

50%

60%

No Savings Saving < 6mo

Saving > 6mo


Saving > 6mo


Saving > 6mo

Intention Intention Intention

VITA SaverLife TTFP

Res

po

nd

ents

(%

)

Could not manage expenses in an emergency (Tax Time)

Could not manage expense in an emergency (Follow-up)


38


AND FINANCIAL WELL-BEING SCORE, BY PLATFORM

Sources: VITA Surveys, Tax Time (N=817) and Follow-up (N=170); SaverLife Surveys, Tax Time (N=522) and Follow-

up (N=106); HFS 2019, Tax Time (N=10,408) and Follow-up (N=2,621)

One issue with examining the relationship between tax refund savings decisions and changes in financial

well-being is that savings intentions or savings actions may be correlated with an individual’s level of

financial well-being when they make savings decisions. This dynamic makes it difficult to assess the degree

to which these decisions may be related to subsequent changes in financial well-being. For example, an

individual with low initial levels of financial well-being may be less likely to intend to save the refund and

also may exhibit lower levels of financial well-being in the future.

To account for this, Figures 15a and 15b present the results of a regression examining the relationship

between different refund savings decisions and changes in financial well-being while controlling for tax

filers’ initial levels of financial well-being. From the figures, we see that both savings intentions and savings

actions are generally correlated with positive financial well-being changes in the months after tax filing. In

particular, intending to save the refund for more than six months at tax time is associated with large and

often significant increases in financial well-being after tax filing. For SaverLife and TTFFP filers, we also see

that actually having some of the refund still in savings at follow-up is also associated with significant and

substantial increases in financial well-being.

30

35

40

45

50

55

60

No Savings Savings < 6mo

Savings > 6mo


Savings > 6mo


Savings > 6mo


Fin

anci

al W

ell-

Bei

ng

Sco

re

FWB, Tax Time FWB, Follow-Up


39

FIGURE 15A. THE RELATIONSHIP BETWEEN REFUND SAVINGS INTENTION

AND CHANGES IN FINANCIAL WELL-BEING SCORE, BY PLATFORM

(REGRESSION RESULTS)

* p<0.1

FIGURE 15B. THE RELATIONSHIP BETWEEN REFUND SAVINGS ACTION AND

CHANGES IN FINANCIAL WELL-BEING SCORE, BY PLATFORM (REGRESSION

RESULTS)

** p<0.05, * p<0.1

0 1 2 3 4 5 6 7 8 9

VITA

SaverLife

TTFFP


Intend to Save > 6 mo

Intend to Save < 6 mo

0 1 2 3 4 5 6 7 8 9

VITA

SaverLife

TTFFP


Refund Saved at Follow-Up

Sources: VITA Tax Time and Follow-up Survey (N=134); SaverLife Tax Time and Follow-up Survey (N=111); HFS 2019 (N=2,538)

Sources: VITA Tax Time and Follow-up Survey (N=123); SaverLife Tax Time

and Follow-up Survey (N=101); HFS 2019 (N=2,470)

*

*

*

*

*

*

*

*

*


40

Summary and Implications

Through our research, we have uncovered several “hidden truths” about the tax time savings behaviors,

the interventions that work well in supporting and encouraging tax time savings and the overall financial

well-being of LMI households. This research has also provided supporting data to dispel previously held

myths about the savings intentions and capabilities of LMI households and helped to identify gaps in service

delivery that, if addressed, could lead to even better financial outcomes for LMI households at tax time and

beyond.

SUMMARY FINDING: LMI TAXPAYERS ARE SAVING SOME OF THEIR

REFUND AT HIGHER RATES THAN PREVIOUSLY THOUGHT OR MEASURED

In contrast with previous data reports and popular understanding, our research has shown that LMI filers,

irrespective of income, want to and do save a portion of their tax refund. Traditional measures of

savings, including measuring rates of refund splitting (through IRS Form 8888) and savings deposits, only

capture a small fraction of overall refund savings behaviors. We found that LMI taxpayers had a plan in

mind for how they would allocate their refund towards their financial goals at the time of tax filing and

the majority of those who were expecting a refund had intentions to save at least a portion. This finding

indicates that previously held paternalistic views that LMI filers need to be told to save or how to handle

their money may be antiquated and worth revisiting, as we find that LMI filers often have to work harder

at achieving financial security and well-being than most. However, we did find that a significant number

of filers (20%- VITA, 14% - SaverLife and 17% - TTTFP) who intended to save were not successful in

doing so.

➢ Implication: There is an opportunity for the field to bridge the savings intention-

action gap.

VITA programs and other service providers can offer additional support for the financial well-being

of LMI filers by:

Facilitating access to savings programs, financial counseling and other supports to help open

doors for taxpayers wishing to improve their financial situation.

Influencing the “refund planning process”, when LMI taxpayers mentally start allocating their

refund by sharing tax time savings messages often and well in the months before the filing

season, at a time when LMI filers are mentally allocating their refund.

SUMMARY FINDING: LMI HOUSEHOLDS ARE USING THEIR TAX REFUND TO

“CATCH UP” RATHER THAN TO “GET AHEAD”

Savings is just one of the uses of the refund and not the most important one for some of the populations.

Though both short- and long-term savings of the refund was a common usage, the most common uses for

the refund were to cover household necessities (rent, bills, groceries) and paying down debt (credit cards,

past due bills).


41

➢ Implication: Tax-time savings is varied and the field may need to enlarge its view of

"savings" and “tax-time financial capability” beyond just deposits into savings

accounts and opening other accounts.

This may include extending the field’s idea of the value of the tax time moment for LMI households

into activities like making advance purchases of household necessities to offset potential income

shortfalls, saving interest over time by paying down debt sooner and other potentially well-being

enhancing uses of the refund. It also implies that offering supports at VITA sites like referrals to

credit counseling, debt management organizations or support for supplementing monthly incomes

by providing information on job trainings and/or social supports like LIHEAP and SNAP among

others would be beneficial to clients over the course of the year.

SUMMARY FINDING: SAVINGS INTERVENTIONS AND SUPPORTS ARE NOT

BEING CONSISTENTLY OFFERED OR IMPLEMENTED AT VITA SITES.

VITA programs across the board do not systematically promote savings or offer additional products and

services at their sites. Despite client interest in learning more about additional financial tools and resources,

each site’s offering of savings interventions and supports is dependent primarily on the willingness and

capability of the Site Coordinator and volunteers and the relative importance they place on those

interventions. Program Managers and Site Coordinators alike generally lack confidence in their ability to train

volunteers and/or talk with taxpayers about savings or other financial supports at tax time.

➢ Implication: There is an opportunity for standardization and improvement in the

implementation of the most successful practices for promoting savings and financial

capability supports at VITA sites.

Through our VITA-specific research, we identified several barriers to successful implementation of

the CFPB’s Promising Practices for Promoting Savings at Tax Time, resulting in limited success of

tax time savings interventions. There is an opportunity, particularly in the VITA field, for technical

assistance and support to translate the “promising practices” into action. There are three keys that

may lead to successfully closing the “savings intention-action gap”:

Focus on consistency in VITA services and volunteer promotion of savings and other financial

well-being services.

Focus on when the taxpayer starts mentally paying their refund and message about savings

opportunities at tax time early and often.

View the tax moment as one opportunity among many that occur throughout the year to help

taxpayers build financial well-being and for VITA programs to collaborate with other services

like financial coaching and credit counseling programs to offer support throughout the year.

SUMMARY FINDING: HAVING SAVINGS ON HAND TO HELP GUARD AGAINST

FINANCIAL HARDSHIP CONTRIBUTES TO A HIGHER SENSE OF FINANCIAL

WELL-BEING.

Taxpayers who intended to save and successfully saved some of their tax refund experienced fewer

financial hardships and increased financial well-being in the months subsequent to filing their tax return.


42

While savings intentions and actions are correlated with lower rates of hardship and improved financial

well-being, hardship is a common experience for many households in these platforms in the months after

tax filing. For example, over a third of TTFFP filers reported food insecurity and roughly a third of

SaverLife filers reported skipping essential bills. However, VITA filers had the highest rates of reporting

that they could not manage a modest $400 emergency expense, but they also had the highest reported

rates of financial well-being.

➢ Implication: There is an opportunity for VITA and other platforms to offer services

that can help households maximize the use of the refund for other obligations, such as essential bills or paying down debt. These may include facilitating access to benefits like SNAP or LIHEAP; or additional services like

credit counseling and/or coaching; secured credit cards; access to subsidized health insurance,

etc. Many VITA programs are already engaging in efforts to provide these types of services.

There is an opportunity for additional tools, information, resources and technical assistance to

advance these efforts and make them more widespread.


43

Works Cited Azurdia, G., & Freedman, S. (2016). Encouraging Nonretirement Savings at Tax Time. Retrieved from Washington, D.C.:

MDRC website: https://www.mdrc.org/sites/default/files/SaveUSA_FinalReport 2015.pdf

Babiarz, P., & Robb, C. A. (2014). Financial literacy and emergency saving. Journal of Family and Economic Issues, 35(1), 40–50. https://doi.org/10.1007/s10834-013-9369-9

Beverly, S. G., Tescher, J., & Romich, J. L. (2004). Linking tax refunds and low-cost bank accounts: Early lessons for program design and evaluation. Journal of Consumer Affairs, 38(2), 332–341. https://doi.org/10.1111/j.1745-6606.2004.tb00872.x

Bhargava, S., & Manoli, D. (2015). Psychological frictions and the incomplete take-up of social benefits: Evidence from an IRS field experiment. American Economic Review, 105(11), 3489–3529. https://doi.org/10.1257/aer.20121493

Board of Governors of the Federal Reserve System. (2016). Report on the Economic Well-Being of U.S. Households in 2015. Washington D.C.

Bronchetti, E. T., Dee, T. S., Huffman, D. B., & Magenheim, E. (2013). When a nudge isn’t enough: defaults and saving among low-income tax filers. National Tax Journal, 66(3), 609–634. https://doi.org/10.17310/ntj.2013.3.04

Common Cents Lab. (2016). Common Cents Lab End of Year 2016 Report. Retrieved from http://advanced-hindsight.com/wp-content/uploads/2017/02/ Common-Cents-Lab-End-Of-Year-Report-2016_Rev-15.pdf.

Common Cents Lab. (2017). Common Cents Lab End of Year 2017 Report. Retrieved from https://advanced-hindsight.com/wp-content/uploads/2018/05/2017-Common-Cents-Lab-Report-509.pdf

Consumer Financial Protection Bureau. (2015). Increasing Saving at Tax Time and Promising Practices for the Field. Washington D.C.: Consumer Financial Protection Bureau.

Dahl, G. B., & Lochner, L. (2012). The impact of family income on child achievement: evidence from the earned income tax credi. The American Economic Review.

Davison, G., Covington, M., Kondratjeva, O., Roll, S. P., & Grinstein‐Weiss, M. (2018). A Toolkit for Expanding Financial Capability at Tax Time. St. Louis, MO: Center for Social Development.

Despard, M., Grinstein-Weiss, M., DeRuyter, A., Guo, S., Oliphant, J. E., & Friedline, T. (2018). Effects of a randomized tax-time savings intervention on savings account ownership among low- and moderate-income households. Journal of Financial Counseling and Planning, 29(2), 219–233. https://doi.org/10.1891/1052-3073.29.2.219

Duflo, E., Gale, W., Liebman, J., Orszag, P., & Saez, E. (2006). Saving incentives for low- and middle-income families: Evidence

from a field experiment with H&R block. Quarterly Journal of Economics, 121(4), 1311–1346. https://doi.org/10.1162/qjec.121.4.1311

Eissa, N., & Liebman, J. B. (1996). Labor supply response to the Earned Income Tax Credit. The Quarterly Journal of Economics. https://doi.org/10.2307/2946689

Evans, W. N., & Garthwaite, C. L. (2014). Giving mom a break: The impact of higher EITC payments on maternal health. American Economic Journal: Economic Policy. https://doi.org/10.1257/pol.6.2.258

Gjertson, L. (2016). Emergency saving and household hardship. Journal of Family and Economic Issues, 37(1), 1–17. https://doi.org/10.1007/s10834-014-9434-z

Grinstein-Weiss, M., Despard, M., Guo, S., Russell, B., Key, C., & Raghavan, R. (2016). Do tax-time savings deposits reduce hardship among low-income filers? A propensity score analysis. Journal of the Society for Social Work and Research, 7(4), 707–728. https://doi.org/10.1086/689357

Grinstein-Weiss, M., Perantie, D. C., Russell, B. D., Comer, K., Taylor, S. H., Luo, L., … Ariely, D. (2015). Refund to Savings 2013: Comprehensive Report on a Large-Scale Tax-Time Saving Program. https://doi.org/https://doi.org/10.7936/K7280749

Grinstein-Weiss, M., Russell, B. D., Gale, W. G., Key, C., & Ariely, D. (2017). Behavioral interventions to increase tax-time saving: evidence from a national randomized trial. Journal of Consumer Affairs, 51(1), 3–26. https://doi.org/10.1111/joca.12114

Gundersen, C., & Ziliak, J. P. (2004). Poverty and macroeconomic performance across space, race and family structure.


44

Demography, 41(1), 61–86. Retrieved from https://link.springer.com/content/pdf/10.1353%2Fdem.2004.0004.pdf

Hoynes, H., Miller, D., & Simon, D. (2015). Income, the earned income tax credit and infant health. American Economic Journal: Economic Policy. https://doi.org/10.1257/pol.20120179

Internal Revenue Service. (2019). Internal Revenue Service Data Book, 2018. Washington D.C.: Internal Revenue Service.

Jones, L. E., & Michelmore, K. (2018). The impact of the earned income tax credit on household finances. Journal of Policy Analysis and Management. https://doi.org/10.1002/pam.22062

Key, C., Tucker, J. N., Grinstein-Weiss, M., & Comer, K. (2015). Tax-time savings among low-income households in the $aveNYC program. Journal of Consumer Affairs, 49(3), 489–518. https://doi.org/10.1111/joca.12070

Leete, L., & Bania, N. (2010). The effect of income shocks on food insufficiency. Review of Economics of the Household. https://doi.org/10.1007/s11150-009-9075-4

Manoli, D., & Turner, N. (2018). Cash-on-hand and college enrollment: Evidence from population tax data and the earned Income Tax Credit. American Economic Journal: Economic Policy. https://doi.org/10.1257/pol.20160298

McGranahan, L., & Schanzenbach, D. W. (2013). The Earned Income Tax Credit and food consumption patterns. SSRN Electronic Journal. https://doi.org/10.2139/ssrn.2366846

McKernan, S., Ratcliffe, C., & Vinopal, K. (2009). Do Assets Help Families Cope with Adverse Events? Washington, DC: Urban Institute.

Mendenhall, R., Edin, K., Crowley, S., Sykes, J., Tach, L., Kriz, K., & Kling, J. R. (2012). The role of earned income tax credit in the budgets of low-income households. Social Service Review, 86(3), 367–400. https://doi.org/10.1086/667972

Meyer, B. D. (2010). The Effects of the Earned Income Tax Credit and Recent Reforms The Effects of the Earned Income Tax Credit and Recent Reforms Executive Summary. In Tax Policy and the Economy (Vol. 24, pp. 153–180). Retrieved from www.nber.org/chapters/c11973

Roll, S. P., Davison, G., Grinstein‐Weiss, M., Despard, M. R., & Bufe, S. (2018). Refund to Savings 2015–2016: Field Experiments to Promote Tax-Time Saving in Low- and Moderate-Income Households. https://doi.org/https://doi.org/10.7936/K72J6BD5

Roll, S. P., Grinstein‐Weiss, M., Gallagher, E. A., & Cryder, C. E. (2019). Can Pre-Commitment Increase Savings Deposits? Evidence from a Tax Time Field Experiment. https://doi.org/http://dx.doi.org/10.2139/ssrn.3464634

Roll, S. P., Russell, B. D., Perantie, D. C., & Grinstein-Weiss, M. (2019). Encouraging tax-time savings with a low-touch, large-scale intervention: evidence from the Refund to Savings experiment. Journal of Consumer Affairs, 53(1), 87–125. https://doi.org/10.1111/joca.12194

Rothbaum, J., & Edwards, A. (2019). U.S. Median Household Income was $63,179 in 2018, Not Significantly Different from 2017. Retrieved December 20, 2019, from https://www.census.gov/library/stories/2019/09/us-median-household-income-not-significantly-different-from-2017.html

Shaefer, H. L., Song, X., & Williams Shanks, T. R. (2013). Do single mothers in the United States use the Earned Income Tax Credit to reduce unsecured debt? Review of Economics of the Household, 11(4), 659–680. https://doi.org/10.1007/s11150-012-9144-y

Sykes, J., Križ, K., Edin, K., & Halpern-Meekin, S. (2015). Dignity and dreams: What the Earned Income Tax Credit (EITC) means to low-income families. American Sociological Review, 80(2), 243–267.

https://doi.org/10.1177/0003122414551552

Tach, L. M., & Greene, S. S. (2014). Robbing Peter to pay Paul: economic and cultural explanations for how lower-income families manage debt. Social Problems. https://doi.org/10.1525/sp.2013.11262

Treasury Inspector General for Tax Administration. (2015). Results of the 2015 Filing Season. Retrieved from https://www.treasury.gov/tigta/auditreports/2015reports/201540080fr.html

Tufano, P. (2011). Just keep my money! Supporting tax-time savings with US Savings Bonds. American Economic Journal: Economic Policy, 3(4), 172–200. https://doi.org/10.1257/pol.3.4.172


45

Appendix A: Literature Review

In this section, we outline some of the key rigorous research that can help us identify the promising

approaches to encouraging savings at tax time, as well as the lessons learned from this growing body of

research. Many of these approaches were tested experimentally and most focused on building short-term

or emergency savings. Tables A1 and A2 provided a detailed overview of all the studies covered in this

section.

Using Financial Incentives and Savings Matches to Promote Tax Time Savings

A common approach to promoting tax time savings directly incentivizes tax filers to save their refund by

providing different types of savings matches for tax refund savings deposits. For example, a program may

offer to match 50% of the refund amount still held in a qualified savings account for a set period of time

(e.g., receiving a $50 match for saving $100 of the refund for six months). This approach may be more

effective at promoting short-term or emergency savings. Alternately, a program may offer to match a

refund deposited to a qualified restricted investment account like an IRA (e.g., a $1,000 refund deposit to

an IRA at tax filing would receive an additional $500 deposit to that account). This approach may be more

effective at promoting long-term savings, like saving for college or retirement. Table A1 provides an

overview of the research that has been done on using savings matches to incentivize tax refund savings.4

TABLE A1 | Matched Tax Time Savings Programs

What Did the Program Do? What Did Researchers Find?

Increasing match rates in the Saver's Credit (Duflo et al., 2006)

Researchers partnered with an H&R Block tax site to

experimentally provide an array of different matches

for depositing their refund into Individual

Retirement Accounts (IRAs). One of the savings

matches they offered was an increase in the match

rate on the Saver's Credit—a non-refundable tax

credit meant to incentivize long-term savings—from

25% to 100%.

Changing the match rate in the Saver's Credit

from 25% to 100% only increased participation in

the credit by 1.3%. However, the authors note

that the Saver's Credit is complex, difficult to

understand and nonrefundable, which may limit

its appeal.

4 Notably, one type of matched savings approach not covered in this overview is prize-linked savings which uses gaming

or lottery designs to incentivize savings. While prize-liked savings programs exist and are sometimes used among VITA

sites (e.g., SaveYourRefund which both offers a $25,000 grand prize and smaller $100 prizes for filers who split some

of their refund into a savings account), there is relatively little rigorous evidence for their impact on tax refund savings

behaviors so we do not include them in our overview.

https://saveyourrefund.com/home/


46

Providing a match for IRA deposits at tax time (Duflo et al., 2006)

In the same H&R Block experiment as the one

above, researchers randomly offered tax filers a 0%,

20%, or 50% match for depositing their tax refund

into an IRA, up to a match limit of $1,000. For

example, a $1,000 refund deposit into an IRA would

receive a $500 match.

Unlike increasing the match rate on the Saver's

Credit, providing a savings match for depositing

the tax refund into an IRA was very effective.

Without a savings match, the IRA deposit rate was

3%. With a 20% and 50% match, the IRA deposit

rates were 8% and 14%.

Pairing low-cost account opening with a savings match (Beverly, Tescher, & Romich,

2004)

Researchers partnered with a community

development financial institution (CDFI) to pilot the

Extra Credit Savings Program, which offered VITA

filers the opportunity to open no-fee/no minimum

balance savings accounts and deposit their tax

refund into these accounts. These accounts also

featured a savings incentive of a 10% bonus on any

funds remaining in the account at the end of the

year.

Overall, there was very little evidence that the

savings matches this program incentivized

additional savings in the new accounts. The

program appears to have helped unbanked

participants develop relationships with financial

institutions, but only 8% of unbanked participants

identified the match as the most important feature

of the program.

Offering a large savings match for keeping refunds in savings accounts (Azurdia &

Freeman, 2016; Key, Tucker, Grinstein-Weiss, & Comer, 2015)

Both the $aveNYC and $aveUSA programs

partnered with VITA sites to use savings matches to

encourage LMI tax filers to save their tax refund. In

both cases, the programs offered tax filers a 50%

savings match if they saved some of their refund in

the months after tax filing.

The experimental evaluation of the $aveUSA

program demonstrated an array of positive

effects: $aveUSA participants accumulated $365

in matching contributions and had accumulated

$522 more in savings than control participants;

$aveUSA participants were also more likely to

have a savings goal at the end of the evaluation

period.


47

A Promise Made, A Savings Account Paid? Lessons From a Tax Time Nudge (SaverLife)

In 2018, SaverLife and CommonWealth launched

Savers Win, a campaign to encourage low-income

families to save their tax refunds. To win cash

prizes, Savers pledged to save their tax refunds by

direct depositing a portion into an account linked to

SaverLife.

Over 9,000 individuals, many of whom were new

to SaverLife, pledged to save over $2.4M in tax

refunds. Of survey respondents, 65% said the

pledge influenced what they did with their refund

and 54% said the chance to win a prize influenced

what they did with their refund.

Transactional data confirm SaverLife Member

claims. 23% of Members who made a pledge to

save diverted a part of their refund directly into a

savings account, versus only 9% for the total

SaverLife population.

Using Behavioral Economics Nudges to Promote Refund Savings

The second major approach to promoting tax time savings uses the techniques of behavioral economics to

motivate increased rates of refund savings through low-touch “nudges.” Behavioral economics draws on

both the fields of economics and psychology to better account for the complexity and nuances of human

decision-making than more traditional approaches. As such, it can help us understand the ways in which

both external environmental cues and unconscious mental processes can lead to a variety of different issues

in financial behavior—including issues like saving too little and taking on high-cost debt—which can be

particularly harmful to LMI households.

Behavioral economics approaches may be especially attractive in the context of promoting savings at tax

time generally and at VITA sites specifically. This is because many behavioral interventions are designed

to be integrated into existing processes and practices without incurring substantial additional costs. This

class of interventions may thus be useful for resource-constrained organizations that aim to promote

savings behaviors but do not have the financial capacity to implement, for example, their own matched

savings program.

Table A2 outlines the array of behavioral interventions that have been tested in the context of promoting

savings at tax time. These interventions range from providing suggested savings amounts to encourage

filers to split their refund, to using different types of reminders to encourage additional EITC claiming, to

encouraging tax filers to pre-commit to saving their refund before they even know the amount of their

refund.


48

TABLE A2 | Using Behavioral Economics to Promote Savings at Tax Time

What Did the Program Do? What Did Researchers Find?

Suggesting tax refund savings amounts (Roll et al., 2019)

Refund to Savings researchers used TurboTax

Freedom Edition to test the impact of different

suggested savings amounts (savings "anchors") on

encouraging LMI tax filers to split their refund into

a savings account. Suggested savings amounts

included 25%, 50%, 75%, $100 and $250 of the

refund.

Offering suggested savings amounts was

associated with higher rates of refund savings and

larger savings suggestions were more effective

than smaller savings suggestions (i.e. suggesting

to save 75% of the refund was more effective than

suggesting to save 25%). Higher suggested

savings amounts were also associated with still

retaining some of the refund as savings six months

later.

Providing different types of messaging about tax refund savings

(Grinstein-Weiss et al., 2017; Roll et al., 2018)

Refund to Savings researchers tested an array of

different messages encouraging saving the refund

in TTFFP. Tax filers were randomly shown one of

several messages promoting saving the refund for

either emergencies, retirement, the future, or

general savings goals. Researchers also tested

passive messages versus more interactive

messages.

In general, messaging around the need to save for

emergencies was more effective than other types

of messaging, though the overall differences were

not large. Interactive messaging—in which tax

filers had to select savings goals—was not more

effective than passive messaging approaches.

Making savings deposits the default decision (Grinstein-Weiss et al., 2017)

Refund to Savings researchers tested the impact of

restructuring refund deposit options to emphasize

savings. Rather than seeing a generic "direct

deposit" option as the first deposit options, the

option to directly deposit their entire tax refund into

a savings account was the first option shown to

TTFFP filers.

Of all the interventions in the R2S Initiative, this

was the most effective. Restructuring refund

deposit options to emphasize depositing the entire

refund into a savings account was associated with

a 58% increase in refund savings deposits (from

8.4% in the control to 13.34% in the group with

the new deposit options).

Asking tax filers to pre-commit to save (Roll et al., 2019)

The Refund to Savings initiative also tested the

degree to which asking TTFFP filers to pre-commit

to saving their refund at the beginning of the tax

filing process impacted their likelihood of

depositing their refund into savings accounts.

Participants pre-committed to save their refund at

rates ranging between 8.6% and 11.7%,

depending on the way the pre-commitment was

framed. Pre-commitment was associated with large

savings deposit increases among early tax filers (a


49

21 percentage point increase in the rate of

depositing the refund into a savings account), but

not among later tax filers.

Promoting bond purchase for tax filers without savings accounts (Roll et al., 2019)

The Refund to Savings initiative also tested the

impact of promoting bond purchases as a savings

alternative for tax filers who did not have savings

accounts. If filers did not have savings accounts,

they were shown a set of refund deposit options

that emphasized bond purchases.

While promoting bond purchases as an alternative

to savings account deposits for those without

savings accounts was successful at increasing bond

purchases, the overall effects were still extremely

modest. Bond takeup in the control group was

0.1% and, in the most successful intervention

group, the bond purchase rate was 0.13%. These

results speak to the general lack of demand for

bonds among LMI tax filers.

Encouraging EITC take-up through a reminder program (Bhargava & Manoli, 2015)

Researchers partnered with the IRS to send letters

to households who did not claim the EITC despite

potentially being eligible for the credit. These

letters were aimed at encouraging EITC take-up

and randomly included four messaging conditions

that either (1) provided a generic reminder (the

control condition); (2) emphasized the ease of

enrolling; (3) provided a high amount of detail on

the EITC; or (4) highlighted the maximum payout

of the EITC.

Providing any mailed reminder at all was associated

with a substantial increase in the rate of EITC take-

up and emphasizing the maximum payout of the

EITC was the most effective way of driving take-up

of the credit. Interestingly, providing detailed and

complex information in the mailing was actually

worse than providing a generic mailing with no

information.

Automatically saving a percentage of the tax refund upon depositing (Common Cents Lab,

2016)

Researchers with the Common Cents Lab partnered

with the Digit financial technology app to test the

impact of pre-commitment on refund savings

intentions. Study participants were asked what

percentage of the refund they would like to save

either at the time they received the refund (the

control group) or prior to receiving the refund (the

experimental pre-commitment group). That

percentage of the refund was then automatically

deposited into a savings account.

This intervention was very effective. Control group

participants who were asked to save their refund

at the time they received it saved 10% of their

refund. Treatment group participants who received

the savings pre-commitment offer before tax filing

saved 15% of their tax refund.


50

Using social pressure to motivate savings (Common Cents Lab, 2017)

The Common Cents Lab partnered with a VITA site

to test the impact of providing filers with different

refund splitting options and of informing filers that

most people pick specific options. The

experimental conditions either recommended

different split rates (e.g. noting that saving 40%

was "highly recommended) or noted that "most

people check" the 25% option.

While providing "recommended" savings options

was more effective than providing a generic set of

options, using social pressure to motivate savings

was the most effective approach of those tested.

Telling people that "most people clicked" the 25%

savings option tripled the rate of people opting to

save 25% of their refund (3% in the control versus

9% in the treatment).

Providing environmental cues to promote refund splitting (Common Cents Lab, 2017)

In the same study as above, the Common Cents

Lab added a number of environmental cues to

promote savings at a VITA site. These cues

included a checklist of required items for a tax filer

to bring to the site including a savings account

routing number and "desk tents" on each tax

preparer's desk reminding filers to save part of

their tax refund.

Providing additional environmental cues to

encourage savings was associated with a marginal

increase in the rate of splitting the refund into

savings. VITA filers who did not receive additional

environmental savings cues split their refund at a

rate of 1.3%, while those who received

environmental savings cues split their refund at a

rate of 1.6%.

Promoting Different Savings Vehicles at Tax Time

Outside of the tax time savings interventions that rely on savings matches or behavioral economics to

encourage savings, there have been a handful of other studies investigating the degree to which tax filing

can be an opportunity to promote different savings vehicles. One study involved providing H&R Block tax

filing professionals with a training session on U.S. Savings Bonds and then offering tax filers the opportunity

to purchase bonds at select H&R Block sites (Tufano, 2011). This study found that offering bonds was

associated with a large increase in the take-up of savings products at tax sites and that savings bonds

disproportionately appealed to the asset poor. A different study investigated the impact of a strong bond

promotion intervention conducted at VITA sites and found that automatically directing some of the tax

refund into a savings bond by default (requiring filers to opt-out of the bond purchase) was not associated

with any increase in savings behaviors, indicating that bond purchases may be highly unattractive for VITA

filers and other LMI tax filers (Bronchetti et al., 2011). Finally, another study found that providing savings

nudges as part of the Refund to Savings Initiative was actually associated with an increase in savings

account opening among those who lacked savings accounts in the months after tax filing (Despard et al.,

2018). This is particularly interesting as these interventions were not focused on account opening and

indicate that providing messaging around savings at tax time may lead to increased engagement with

financial products beyond the tax season.

The key lesson from this research is that bond purchases are a good way for tax filers to save but programs

promoting bond purchase may be relatively ineffective. This is not to discount the benefits of savings bonds


51

generally. The more savings options tax filers have, the better and educating tax filers about U.S. Savings

Bonds and encouraging their purchase is a valuable service tax preparers can offer. At the same time,

interventions targeting bond purchases tend to be ineffective: Even relatively strong interventions like

making bond purchases at tax time the default option had no effect on bond purchasing behavior

(Bronchetti et al., 2011). As such, investing resources into promoting bond purchases as a savings option

may not lead to substantial increases in savings rates.


52

Appendix B: Enhanced Data Collection Project Tax-

Time Survey


53

Appendix C: Comparison of Questions Included in

VITA, R2S and SaverLife Data Sources

COMPARISON OF FOLLOW-UP SURVEY QUESTIONS

Data Source VITA TTFFP SAVERLIFE

Tax Refund-Specific Questions

Did you receive a tax refund this year? ✓ ✓ ✓

How much was your federal tax refund this year? ✓ ✓ ✓

[IF NO REFUND] How much did you owe in federal taxes this

year?

✓ ✓ ✓

Upon receiving your tax refund, did you save any of it? ✓ ✓ ✓

How much of your tax refund do you still have saved? ✓ ✓

How did you choose to receive your tax refund this year? ✓ ✓ ✓

Did you use your tax refund to …

Save for emergencies or other unexpected needs ✓ ✓ ✓

Save for specific future purposes ✓ ✓ ✓

Make a large purchase ✓ ✓ ✓

Pay down debt ✓ ✓ ✓

Purchase household necessities ✓ ✓ ✓

Make essential payments in advance ✓ ✓ ✓

Cover essential healthcare expenses ✓ ✓ ✓

Pay for legal expenses ✓ ✓ ✓

Other (please specify): ✓ ✓ ✓

How important was your refund for you to be able to…

[OPTIONS IDENTICAL TO TAX REFUND USAGE OPTIONS] ✓ ✓ ✓

Did you use your any of your tax refund for your child(ren),

grandchild(ren), or any dependents under the age of 18?

✓ ✓ ✓

Did you use any of your tax refund to support any family

members over the age of 18?

✓ ✓ ✓


54

What was the most important item you used your tax refund

for? (Open-Ended)

✓ ✓ ✓

Which of the following statements reflects your view on

withholding taxes from your paycheck?

✓ ✓ ✓

Other Categories

Hardship Experience ✓ ✓ ✓

Manage $400 Expense ✓ ✓ ✓

Financial Well-Being ✓ ✓ ✓

Income ✓ ✓ ✓

Demographics ✓ ✓ ✓

VITA-specific questions ✓

Savings account transaction data ✓

Self-employment/gig economy ✓

Financial shocks ✓

Assets/debts ✓

Health/insurance ✓


55

Appendix D: VITA Programs included in the VITA

Tax-Time Research Project

VITA PROGRAMS INCLUDED IN THE VITA TAX-TIME RESEARCH PROJECT

Stat

e

VITA Program VITA Sites # of sites

participati

ng

Key

Populations

Served

MO Metro St. Louis

Community Tax

Coalition

MCTC St. Louis College Prep 4 Urban

communities MCTC Central Baptist Church

MCTC Waypoint Church - The Realm

MCTC Lindenwood Tax Clinic

AR Arkansas Asset

Builders

Arkansas Asset Builders 2

Arkansas Asset Builders Mobile

FL Campaign for

Working Families

CFWF @ Bethune Cookman University School of Business 2nd Floor

11 Immigrant

communities,

limited English-

speaking

populations,

urban

communities

CFWF @ Boys and Girls Club of Volusia-

Flagler County

CFWF @ Chisholm Community Center

CFWF @ Daytona State College

CFWF @ Flagler County Library

CFWF @ Flagler NE FL Community Action Agency

CFWF @ Four Townes YMCA

CFWF @ Goodwill Self Sufficiency Center

CFWF @ Grace Episcopal Church

CFWF @ Mid-Florida Community Services

CFWF @ Mt. Calvary Baptist Church of Palm Coast

TX Community

Development

Corporation of

Brownsville

Community Development Corporation of Brownsville

2 Immigrant and

limited English-

speaking

populations,

urban

communities

La Hacienda

MI Accounting Aid

Society

AAS Macomb County Central Action 2 Urban

communities AAS Macomb County South

VA Northern VA

Cash Campaign

Ferlazzo 1 Immigrant and

limited English-

speaking

populations


56


57

savings interventions and behaviors - prosperity now...evaluating tax time savings interventions and...

Documents