Individual Development Accounts: a Vehicle for Low-Income Asset Building and Homeownership

Individual Development Accounts: a Vehicle for Low-Income Asset Building and Homeownership

Highlights

• Asset-building strategies, such as Individual Development Accounts, enable low-income persons to save and invest in long-term assets with return potential.
• Studies show that participants in Individual Development Accounts experience positive outcomes, such as accelerating the move to homeownership, obtaining safe mortgages, succeeding as homeowners, and avoiding foreclosure.

Research suggests that IDA participants are more likely to become homebuyers and tend to be more successful homeowners. Individual Development Accounts(IDAs) emerged in the United States in the 1990s as an asset-building strategy. Although IDA programs vary in design, they all provide matching funds to low-income recipients to promote savings that can be spent later on eligible uses such as higher education, microenterprise, and homeownership. The goal of these programs is to help low-income families save money that they can invest in high-return, long-term assets.

Recent research provides important insight into the success of IDAs as a vehicle for promoting both asset-building and low-income homeownership. As noted in the previous article, homeownership has been a primary means for low-income Americans to build wealth and has been shown to yield positive social outcomes. Evidence suggests that IDAs, when paired with counseling, may promote more sustainable low-income homeownership. This article explores the history of and research underpinning IDAs, both in general and in the homeownership context.

IDAs: History and Implementation

IDAs were first proposed by sociologist Michael Sherraden in his 1991 book Assets and the Poor: A New American Welfare Policy. In his book, Sherraden states, “Unlike traditional welfare programs, IDA accounts would introduce real assets into the lives of many poor people who would otherwise be without them. IDAs would be a different approach to welfare policy, an approach that emphasizes individual development and combines social provision with individual responsibility and individual control. IDAs would enable the poor to bring their own cards to the table and make their own deal.”1 They would also promote longer planning horizons and other positive behaviors.2

The Personal Responsibility and Work Opportunity Reconciliation Act of 1996, which significantly reformed welfare, included IDAs as an eligible use of federal funds. Later, the 1998 Assets for Independence Act authorized the U.S. Department of Health and Human Services (HHS) to provide nonprofit organizations with grants to implement IDA programs in partnership with community development financial institutions, eligible credit unions, and local, state, or tribal governments.3 These projects have yielded critical research about IDA program design, user demographics, and results. HHS funding also spurred a significant expansion in U.S. IDA programs; more than 200 organizations run Assets for Independence projects, with the greatest number in California (22), Pennsylvania (11), Texas (10), Ohio (9), and Florida (9).4 In all, more than 600 IDA programs are active nationally, according to the Corporation for Enterprise Development.5 In addition to Assets for Independence, other major IDA funders include HHS’s Temporary Assistance for Needy Families program and the agency’s Office of Refugee Resettlement, Federal Home Loan Banks, philanthropies, and local corporations and financial institutions.6

Because so many entities sponsor them, IDA programs show significant variation in design characteristics and eligibility requirements. However, most programs share certain key characteristics. Eligibility is typically limited to those whose annual household incomes fall below a certain threshold (often 200 percent of the federal poverty level), and may also require limited net worth and a good credit history.7 Participants are usually enrolled in a program for a period of one to five years, and saved earnings, when withdrawn for eligible uses, are matched at rates that can range from 1:1 (the most common) to as high as 6 matched dollars for every dollar saved. Some programs cap annual and lifetime matches. Matched savings can typically be used to pay for higher education, start a small business, or buy a home, and some programs also allow participants to use matched savings for retirement, home repairs, or work-related car or computer purchases.8 In addition to matching savings, programs also provide general financial counseling, and many offer additional training tailored to the participants’ planned use.

Key Findings on IDAs

Important findings on IDAs have emerged from evaluations of the Assets for Independence Act and the projects it has funded. The formal impact study of the act examined a 3-year longitudinal sample of 485 participants from IDA programs around the country, compared using propensity score matching with a nonparticipant sample constructed from the U.S. Census Bureau’s Survey of Income and Program Participation.9 The average IDA participant in the study saved $935 after 3 years in the program, a net savings rate of 1.2 percent of their earnings. Although participants made more unmatched withdrawals (withdrawals for ineligible uses) than matched ones in the first two years, by the end of the third year, the total average of matched withdrawals exceeded unmatched withdrawals and 31 percent of participants had made at least one matched withdrawal.10 The study revealed several statistically significant effects on the three major forms of asset ownership: IDA participants were 35 percent more likely to be homeowners, 84 percent more likely to own businesses, and 95 percent more likely to pursue postsecondary education than nonparticipants. However, the program did not have a statistically significant effect on monthly earnings, overall financial assets, home equity, or consumer debt — partly because of the short duration of the study.11

Source: "Find an IDA Program Near You," The Corporation for Enterprise Development website (cfed.org/programs/idas/directory_search/). Accessed 19 November 2012.

Another critical source of evidence on the effects of IDAs has been the American Dream Demonstration (ADD). Coordinated and funded by the Corporation for Enterprise Development and its philanthropic partners, ADD was composed of 14 program sites nationwide that offered IDAs, each of which was managed by different entities and employed varying program designs and guidelines.12 ADD targeted the working poor, and as with the programs funded by Assets for Independence, its participants were more likely to be female, African American, and single and never married than the low-income population overall. ADD participants were also significantly more educated and more likely to be employed full- or part-time than the average low-income person.13 The outcomes of this group of 2,378 participants have been studied extensively and continue to inform the evolution of IDA program design.

Research into IDA savings patterns among the ADD participants revealed an important link between income level and savings rates; although average monthly net deposits into IDAs increased with growth in household income, the rate of saving declined substantially, especially for those at higher income levels.14 To gain further insight into this inverse relationship between income and IDA savings rate, Sherraden et al. examined program characteristics that might affect savings. Higher monthly match caps — which can be viewed as savings targets — yielded greater savings and fewer unmatched withdrawals.15 Financial education was also associated with increased IDA contributions, but only for the first 12 hours of education.16 In contrast, savings match rates were not associated with increased savings; though this seems surprising, it is consistent with evidence from 401(k) plans, another form of matched savings.17 The authors also note that the data suggest that IDA program design may affect comparatively poorer participants more strongly.18

Researchers have also analyzed data from ADD to better understand other IDA outcomes. Key findings include the following:

• Prior banking experience is correlated with various positive IDA results; ADD participants who previously used banks made greater average monthly deposits, deposited more frequently, and had much lower odds of dropping out than participants without banking experience.19

• ADD participants who had higher education levels, existing assets, and no debt were less likely to drop out of the program. Program design also mattered; programs with higher match rates and longer programs had lower drop-out rates, and participants who took advantage of direct deposit were also less likely to drop out.20

• Preliminary research into the long-term effects of IDA participation at the Community Action Project of Tulsa County ADD site suggests that, 10 years after the program began, IDA participation had significant positive impacts on education and home maintenance but not on homeownership, microenterprise, or retirement savings.21

But with research on the ADD and other IDA programs, methodology challenges must be considered. Researchers consistently caution that program participants are often recruited by sponsoring organizations and affirmatively choose to participate in an IDA program, which increase the likelihood of consistent differences between program participants and the low-income population as a whole. Although some of this variation in personal characteristics is observable and can be accounted for in statistical analysis, differences in motivation or other unobservable factors may skew results, even in experimental evaluations that randomly assign applicants to either an IDA treatment group or a control group. In addition, many IDA evaluations rely on comparison groups built using data rather than through random sampling; this quasi-experimental design also increases the likelihood of systematic differences between program participants and the low-income population overall.

IDAs and Homeownership

While many articles examining IDAs focus on overall asset-building results, another vein of IDA research has delved into the effects of these programs on specific savings goals, including homeownership. The challenges confronting low-income Americans seeking homeownership opportunities are described throughout this issue of Evidence Matters.

Homeownership is associated with a variety of improved outcomes, but it also carries substantial risks, especially for low-income households. As University of Southern California economists Raphael Bostic and Kwan Ok Lee explain, “The current housing environment…is one in which lower-income households can find themselves quite vulnerable to homeownership failure. Our finding that foreclosure rates are elevated in lower-income communities, holding other factors equal, supports the notion that these elevated risks have come to fruition.”22 In their research, Bostic and Lee discover that low-income households who were able to make higher down payments and build equity early derived greater financial benefits from homeownership. They conclude that “[a] clear dimension to be explored is increasing lower-income households’ savings rates. Higher savings would permit these households to make larger down payments, and we have shown that the risks of homeownership difficulties and foreclosure are significantly reduced if homeowners can acquire equity early in their ownership tenure.”23 Research into the outcomes of IDAs used to purchase homes supports the argument that increased savings can yield more sustainable low-income homeownership.

Buying a home is the most popular eligible use for IDA matching funds.

Homeownership is a key driver of IDA participation and the most popular savings goal in programs where matching funds can be used to buy a home.24 This finding may be related in part to IDA program design, because matching rates are often higher for homeownership or microenterprise goals than for higher education or other uses.25 In addition to the matched savings, IDA programs typically offer prepurchase homeownership counseling and guidance in (and sometimes direct oversight of) mortgage product selection.26 Although buying a home is the most common use of IDAs, Schreiner and Sherraden’s review of ADD finds that this goal is also associated with failure to complete IDA programs: “About one-half of IDA participants in ADD planned to save for home purchase, and they are much more likely to drop out than those planning for other matched uses.”27 The authors attribute this difference to two factors: that among program participants, renters are likely worse savers than those who are homeowners, and that the process for purchasing a home is more difficult and expensive than that for other potential uses making participants more likely to become discouraged and drop out.28

Although purchasing a home remains challenging for low-income households even when they are enrolled in IDAs, studies show that renters participating in IDA programs were likely to become homeowners more quickly than nonparticipants were. The Tulsa ADD program site structured its IDA program as a randomized experiment, and the outcomes at this site have been heavily analyzed. Grinstein-Weiss et al. examined Tulsa renters belonging to the IDA participant group and control group at four periods: at the start of the program; 18 months into the program; at 4 years, when the program concluded; and 10 years after the start of the program. The treatment group received financial education and case management services in addition to matched savings, whereas the control group did not have access to matched savings but could seek out homeownership counseling from other providers in the area.29 The researchers found that, at 4 years, the “odds of being a homeowner were 75 percent higher for the treatment group than for the control group,” controlling for demographic and financial variation.30 In addition, clearing old debts appeared to be a critical step on the road to homeownership; 32 percent of IDA participants who had reported clearing old debts at 18 months were homeowners after 4 years compared with 15 percent of IDA participants who did not clear debts and 9.6 percent of non-IDA participants who did not report clearing their debts.31

However, recent preliminary research by Grinstein-Weiss and others into long-term followup results has revealed considerably weaker effects of IDA participation on homeownership. Ten years after the program began, both treatment and control groups had experienced large growth in homeownership, and among the full group no statistically significant effect of IDA participation was evident. Thus, IDA participation may only have accelerated rather than increased homeownership among program participants. However, “for the subgroup of people with above-sample median annual incomes at baseline (about $15,500 per year), assignment to the treatment group significantly increased the homeownership rate and duration of homeownership.”32 The authors suggest that this finding could support targeting IDA programs with a homeownership component to those on the higher end of income eligibility. The authors also acknowledge that the relative ease of low-income home purchase between 1998 and 2007, the comparatively low housing costs in Tulsa during this time, and the availability of alternative homeownership assistance for the control group may have dulled the long-term effects of IDA participation on homeownership in the experiment.33

Research suggests that IDA participants not only are likely to become homebuyers earlier than other low-income persons but also tend to be more successful homeowners. Rademacher et al.’s 2010 article “Weathering the Storm: Have IDAs Helped Low-Income Homebuyers Avoid Foreclosure?” examines the outcomes of 831 homebuyers from 6 IDA programs between 1999 and 2007, testing various homeownership measures against a comparison group constructed with Home Mortgage Disclosure Act (HMDA) and other mortgage performance data sources.34 The researchers found that minorities and women composed a much higher proportion of homebuyers in the IDA sample than in their comparison sample of low-income homebuyers: “The proportion of African American homebuyers in the IDA sample is more than three times higher than in the HMDA sample, and for Hispanic homebuyers, the proportion is 1.5 times higher. Similarly, 73.5 percent of the IDA homebuyers are female compared with 44.6 percent of the HMDA sample.”35 Women and minorities were much more likely to receive subprime mortgages during the period of this study. IDA homebuyers in the treatment group, however, received government-insured loans and avoided subprime and high-interest loans in much higher proportions than did their non-IDA counterparts, likely because of their access to counseling and ongoing mortgage product monitoring as well as their ability to make higher down payments than most low-income homebuyers could.36

Perhaps Rademacher et al.’s most important finding was that IDA program participants experienced foreclosure in much smaller numbers: “Among our sample of IDA homebuyers…3.1 percent (or 25 out of 803 homes) entered foreclosure by April 2009. This foreclosure rate is less than one-half to one-third of the foreclosure rates for the comparison samples.”37 As of 2009, 93 percent of program participants had successfully retained their homes without demonstrating any difficulty in making their mortgage payments.38 Despite its quasi-experimental design, IDA participants in Rademacher et al.’s report do not represent a random sample of the national low-income population because both the treatment and control groups took the initiative to apply to the program, creating self-selection bias. Further, the study cannot determine precisely which aspects of IDA program design — the increased savings, the counseling, or the help with selecting an appropriate loan product — contributed most to these positive outcomes. Nevertheless, the authors conclude that “[s]tudy findings are consistent with the hypothesis that the services and features of IDA programs…help low-income populations obtain affordable mortgages and experience successful and sustainable homeownership outcomes.”39

HUD’s Family Self-Sufficiency Demonstration

As research continues to inform IDA policy, successes learned from IDA evaluations are being applied to other initiatives targeting low-income individuals when possible. HUD’s Family Self-Sufficiency (FSS) program is just one example of a program that draws from IDA asset-building strategies. The FSS program helps low-income families who receive housing vouchers build the assets needed to achieve economic self-sufficiency.40, 41 FSS participants meet with case managers who provide training in financial literacy and referrals to supportive services such as childcare and education programs. Each participant receives an escrow account managed by the public housing agency. Rent normally increases as the family’s income rises, but in the FSS program, credits based on increased income are deposited into the family’s escrow account; this savings incentive serves as the equivalent of the matched savings in IDA programs. Graduates from the program receive the total amount banked plus interest.

Unlike participants in traditional IDA programs, graduates of the FSS program may use their savings however they wish; however, they are encouraged to put their money toward asset-building activities. Researchers in HUD’s Office of Policy Development and Research recently assessed outcomes for a group of FSS participants from 2005 to 2009 and found that after 4 years, 24 percent had graduated from the program and received their escrow funds, with which 4 families had purchased a home. Participants achieved meaningful savings in a relatively short period, with graduates averaging $5,300 in escrow by the time they completed the program. Graduates also made significant strides in annual household income, increasing from an average of $19,902 during their first year to $33,390 in the year they completed the FSS program.42

More research on the efficacy of the FSS program is in development. HUD has funded two studies of the program to date, but neither one shows how well families would have done in the absence of the program. The third study, the Family Self-Sufficiency Program Demonstration, will evaluate whether the benefits that FSS participants enjoy (such as increased income and savings, educational attainment, and economic improvement) can be attributed with certainty to the program’s design or to inherent characteristics of the participating families.

Next Steps

IDAs remain a comparatively new policy innovation, and research will continue to shape the development of program guidelines and targeting of services. And this research is by no means limited to the United States; case studies on national IDA programs in both developing and developed countries are yielding important findings. One key dimension of this ongoing research is the need to clearly determine which components of IDA interventions most drive improved outcomes for participants. For example, as mentioned previously, research from the ADD shows that financial education correlates with greater savings for the first 12 hours, but because IDA programs consistently require financial education, little evidence exists as to whether IDA programs without a financial education component would succeed at all.43 Such research can help future IDA initiatives create more efficient program designs that tailor to the needs of specific populations.

Even as evidence refines future IDA efforts, certain challenges are likely to persist. The Assets for Independence Act process study notes a number of ongoing issues, and while several are related to program participants — such as attracting sufficient participants and helping participants set reasonable goals — many are tied to funding. These financial challenges include “navigating the regulations of diverse funding sources and requirements; raising nonfederal funds; [and] coping with limited funds for administrative costs.”44

IDAs remain expensive, and not just because of the need to match participant savings. In 2005, IDAs cost about $64 per participant each month to administer, including “recruitment, financial education, monitoring deposits and withdrawals, and providing other high-tech services.”45 And because IDA programs remain relatively small, they often do not achieve economies of scale.46 Ongoing expansions of IDA programs, technical advances, and the efforts of organizations such as the Corporation for Enterprise Development to share best practices have likely helped to lower administrative costs, but obtaining sustainable funding streams that do not rely overly on a single source, such as a federal grant, will remain a critical challenge.

With demonstrated success at bolstering rates of higher education, microenterprise, and successful homeownership, IDAs are poised to become an increasingly important approach to supporting wealth-building among low-income Americans.

1. Michael Sherraden. 1991. Assets and the Poor: A New American Welfare Policy, Armonk, NY: M.E. Sharpe, 231.
2. Donna DeMarco, Gregory Mills, and Michelle Ciurea. 2008. “Assets for Independence Act Evaluation, Process Study: Final Report,” U.S. Department of Health and Human Services, 2–3.
3. “Assets for Independence Program Summary,” U.S. Department of Health and Human Services website (www.acf.hhs.gov/programs/ocs/resource/assets-for-independence-program-summary). Accessed 5 November 2012.
4. “AFI Grantees.” Assets for Independence Resource Center website (idaresources.org/afigrantees). Accessed 5 November 2012.
5. “IDA Directory.” Corporation for Enterprise Development website (cfed.org/programs/idas/directory_search/). Accessed 5 November 2012.
6. DeMarco, Mills, and Ciurea, 3.
7. Corporation for Enterprise Development. 2009. "Individual Development Accounts (IDAs)."
8. Ibid.
9. Gregory Mills, Ken Lam, Donna DeMarco, Christopher Rodger, and Bulbul Kaul. 2008. “Assets for Independence Act Evaluation, Impact Study: Final Report,” U.S. Department of Health and Human Services, iv.
10. Ibid., v.
11. Ibid., vi–vii.
12. Michal Grinstein-Weiss, Michael Sherraden, William M. Rohe, William Gale, Mark Schreiner, and Clinton Key. 2012. “Long-Term Follow-up of Individual Development Accounts: Evidence From the ADD Experiment,” v.
13. Michael Sherraden, Mark Schreiner, and Sondra Beverly. 2003. “Income, Institutions, and Saving Performance in Individual Development Accounts,” Economic Development Quarterly 17:95, 99–100.
14. Ibid., 104.
15. Ibid., 106.
16. Ibid.
17. Ibid., 104, 107.
18. Ibid., 108.
19. Michal Grinstein-Weiss, Yeong H. Yeo, Mathieu R. Despard, Adriance M. Casalotti, and Min Zhan. 2010. “Does Prior Banking Experience Matter? Difference of the Banked and Unbanked in Individual Development Accounts,” Journal of Family and Economic Issues 31:2, 220.
20. Mark Schreiner and Michal Sherraden. 2005. “Drop-Out From Individual Development Accounts: Prediction and Prevention,” Financial Services Review 14:4, 48.
21. Grinstein-Weiss, Sherraden et.al, vi.
22. Raphael Bostic and Kwan Ok Lee. 2009. “Homeownership: America’s Dream?” In Rebecca Blank and Michael Barr, eds., 2009. Insufficient Funds: Savings, Assets, Credit, and Banking Among Low-Income Households, New York: Russell Sage Foundation, 244.
23. Ibid., 246.
24. Ray Boshara. 2005. “Individual Development Accounts: Policies To Build Savings and Assets for the Poor.” Brookings Institution Policy Brief: Welfare Reform and Beyond 32, 6.
25. Roslyn Russell. 2008. “Saver Plus: More Than Saving — A Brief Comparison With International Programs,” Melbourne, Australia: RMIT University, 3.
26. Ida Rademacher, Kasey Wiedrich, Signe-Mary McKernan, Caroline Ratcliffe, and Megan Gallagher. 2010. “Weathering the Storm: Have IDAs Helped Low-Income Homebuyers Avoid Foreclosure?” Washington, DC: Urban Institute, 1.
27. Shreiner and Sherraden, 43.
28. Ibid.
29. Michal Grinstein-Weiss, Jung-Sook Lee, Johanna K. P. Greeson, Chang-Keun Han, Yeong H. Yeo, and Kate Irish. 2008. “Fostering Low-Income Homeownership Through Individual Development Accounts: A Longitudinal, Randomized Experiment,” Housing Policy Debate 19:4, 720–21.
30. Ibid., 726.
31. Ibid., 729.
32. Grinstein-Weiss, Sherraden et al., 46.
33. Ibid., 22–3.
34. Rademacher et al., 2.
35. Ibid., 10.
36. Ibid., 11–2.
37. Ibid., 12.
38. Ibid.
39. Ibid., 14.
40. Christopher E. Herbert and Winnie Tsen. 2005. “The Potential of Downpayment Assistance for Increasing Homeownership Among Minority and Low-Income Households,” U.S. Department of Housing and Urban Development, Office of Policy Development and Research.
41. U.S. Department of Housing and Urban Development, Community Planning and Development. 2008. “Use of Community Development Block Grant (CDBG) Program Funds in Support of Housing,” Notice CPD-07-08.
42. Lalith de Silva and Imesh Wijewardena. 2011. “Evaluation of the Family Self-Sufficiency Program,” U.S. Department of Housing and Urban Development, Office of Policy Development and Research. Also see Evaluation of the Family Self-Sufficiency Program: Retrospective Analysis, 1996 to 2000.
43. Margaret Clancy, Michal Grinstein-Wiess, and Mark Shreiner. 2001. “Financial Education and Savings Outcomes in Individual Development Accounts,” Working Paper HEW 0108001, EconWPA, 1, 7.
44. DeMarco, Mills, and Ciurea, iv.
45. Boshara, 4.
46. Ibid.