Statement of Ray Boshara, Policy Director, Corporation for Enterprise Development
Testimony Before the Subcommittee on Human Resources and
Subcommittee on Select Revenue Measures
of the House Committee on Ways and Means
Hearing on H.R. 7, the "Community Solutions Act of 2001"
June 14, 2001
Mr. Chairmen, Members of the Subcommittees:
Thank you very much for inviting me to appear before you today. I am honored to be here.
The Corporation for Enterprise Development (CFED) is a non-profit, non-partisan firm committed to widely shared and sustainable economic growth, in particular for low-income families and communities. In our 20 years of existence, CFED has pioneered many innovative and promising strategies, including the creation and expansion of the microenterprise and asset development fields. Through our work, we have enabled families nationwide to participate in the mainstream economy, and to realize their dreams of obtaining good jobs, opening small businesses, going to college, owning a home, and bequeathing a better future for their children. Further information on CFED can be found at our website at www.cfed.org.
I am here today to testify in support of Title III of the Community Solutions Act. Title III is the Savings for Working Families Act of 2001, which proposes a national expansion of Individual Development Accounts, or IDAs, through a tax credit to financial institutions that set-up, match and support IDAs. CFED greatly appreciates your consideration of this important legislation.
I’m pleased that the bi-partisan Savings for Working Families Act of 2001 is included in the Community Solutions Act, and to report that, yesterday, Congressmen Joseph Pitts and Charles Stenholm and Senators Joseph Lieberman and Rick Santorum introduced the Savings for Working Families Act of 2001 as a stand-alone bill. I would also like to recognize and commend my former boss, Congressman Tony Hall, for first bringing the concept of IDAs to Congress back in 1991 as the Chairman of the House Select Committee on Hunger. Finally, I thank Congressman J.C. Watts, as well as many members of the Ways and Means Committee, for their long-standing support of asset building for the poor and IDAs.
As many of you may know, IDAs are emerging as one of the most promising tools to enable low-income American families save, build assets and enter the financial mainstream. IDAs reward the monthly savings of working-poor families who are trying to buy their first home, pay for post-secondary education, or start a small business. This reward or incentive is provided through the use of matching funds that typically come from a variety of private and public sources. Similar to 401(k)s, IDAs make it easier for low-income families to build the financial assets that they need to achieve the American Dream. To further help them move into the economic mainstream, accountholders receive financial education and counseling. Further information on IDAs can be found at www.idanetwork.org.
In this testimony, I shall address three questions:
1. What do we know about assets?
2. What do we know about IDAs?
3. Why should Congress support the Savings for Working Families Act of 2001?
I. What do we know about assets?
We know four things about assets — assets matter; low- and moderate-income people have relatively few assets; public policy plays an important role in determining who gets assets and who doesn’t; and asset subsidies are delivered through the tax code via individual asset accounts, which works well for better-off families but not so well to lower-income families.
1. Assets matter. Assets not only provide an economic cushion and enable people to make investments in their futures; assets also provide a psychological orientation — toward the future, about one’s children, about having a stake in America — which income alone cannot provide. Michael Sherraden — author of Assets and the Poor — observes that, "Few people have ever spent their way out of poverty. Those who escape do so through saving and investing for long-term goals." Melvin Oliver and Thomas Shapiro, authors of Black Wealth/White Wealth, state that "Wealth is a particularly important indicator of individual and family access to life chances…It is used to create opportunities, secure a desired stature and standard of living, or pass class status along to one’s children." And this is not just theory: Edward Scanlon and Deborah Page-Adams find that there is increasing evidence that assets do in fact have a range of important positive effects on children, families, and neighborhoods.
2. Assets are distributed more unevenly than income in the U.S. Researchers Robert Haveman and Edward Wolff constructed, for the first time, a measure of "asset poverty" and conclude that asset poverty (using a range of definitions) greatly exceeds income poverty. Even using their most "liberal" definition of asset poverty — net worth needed to get by for three months at the poverty level — the asset poverty rate of 25.5 percent is twice that of the income poverty rate of 12.7 percent. Stacie Carney and Bill Gale, in a thorough review of previous research on saving and wealth accumulation among the poor, conclude: "Although researchers are uncertain as to why low-income and disadvantaged households accumulate low levels of assets and what can be done about it, the basic fact of low accumulation cannot be disputed…The available data present a unified picture: low-income households accumulate almost nothing." Finally, low levels of assets are a particular problem for African-Americans, Latinos, and children. For example, forty percent of all white children, and 73 percent of all African-American children, grow up in households with zero or negative net financial assets. Their prospects for achieving the good life are, in my view, quite slim.
3. Public policy encourages asset accumulation for some and discourages it for others. In my view, public policy can structure opportunity, or lack thereof. This is true both historically and currently. In fact, using the assets framework, one can discern two distinct social policies in America, and these two policies (combined with lack of institutional arrangements to support asset building among lower-income persons) account for most of the gap in wealth between lower- and higher-income households:
- Asset development. For the non-poor, we have an asset development policy. From the Homestead Act and GI Bill of previous years, to $288 billion in annual tax subsidies for individuals to accumulate assets today (especially homes, retirement accounts, small businesses, and higher education) this country has widely – and wisely – supported asset accumulation by households. But these benefits are highly regressive: for example, according to the Joint Committee on Taxation, households with incomes over $50,000 received 91 percent of homeownership tax expenditures and 93 percent of retirement tax expenditures in FY1998.
- Asset denial. Low-income and poor families face three barriers in trying to accumulate assets. One, lacking an income tax liability, they cannot take advantage of tax-based opportunities such as mortgage and pension deductions. Two, low-income families seeking public assistance must — as a matter of law — spend down their assets to receive such assistance, and face severe asset limits once on assistance. And three, low-income persons are much more likely than others to be among the 10 to 20 percent of "unbanked" households, thus preventing them from accessing the mainstream financial services that make asset accumulation possible.
4. Increasingly, domestic policies in the U.S. and worldwide use individual asset accounts to allocate asset subsidies and achieve social and economic policy goals. Michael Sherraden observes that social policy is moving away from large programs and towards individual assets accounts, most of them provided through the income tax system. Examples include IRAs, Roth IRAs, 401(k)s, 403(b)s, Super IRAs, Medical Savings Accounts, Individual Training Accounts, and proposals for Children’s Savings Accounts. Increasingly, this is how government will deliver benefits to its citizens, since such accounts provide control and flexibility for families in a global economy. However, lacking an income tax liability, low-income families cannot access most of these benefits. IDAs let low-income families participate in this social policy transformation, allowing them to earn public and private matches (provided they have worked and saved first) and then enabling them to make the social policy choice — e.g., financing a home, a business, an education — that best suits them.
II. What do we know about IDAs?
To summarize, IDAs work. The Center for Social Development (CSD) at Washington University in St. Louis — the primary evaluators of IDAs —reported in January that "Data from the American Dream Demonstration suggests that the poor can save and accumulate assets in IDAs."
IDAs were first proposed in 1990 by Michael Sherraden, but the first IDA programs didn’t come into being until 1995 in Illinois, Wisconsin, and Indiana. Since then, IDAs have expanded to nearly every state in the country (only North Dakota, Wyoming and Alaska have, to the best of our knowledge, no IDA activity). Most of the IDA programs are supported by private funding, although state and federal support is increasing.
IDA practice, policy, and demonstrations can be summarized as follows:
IDA Practice: About 10,000 low-income families are saving in IDAs today, all of them through the support of about 300 community-based organizations. IDAs reach a diverse range of disadvantaged people and are implemented by a broad range of non-profits, including churches, credit unions, housing agencies, welfare agencies, workforce development programs, United Ways, and community development corporations. Financial education is provided by these organizations, as well as by others.
IDA Policy: At the federal level, IDAs have been included in existing economic development and safety programs (such as TANF and the Community Reinvestment Act), and two small federal IDA programs have been funded, both at the U.S. Department of Health and Human Services (one in the Office of Community Services, the other at the Office of Refuge Resettlement). IDAs have also been embraced by states: 32 have included IDAs in their TANF plans (although only about half of those actually use TANF funds for IDAs) and 31 states have passed IDA legislation. States are also using CSBG and CDBG funds, as well as tax credits, to expand IDAs.
IDA Demonstration: The most significant source of hard data about IDAs is the Downpayments on the American Dream Policy Demonstration, or "ADD." ADD is a 14-site, privately-funded, 4-year demonstration organized by CFED and CSD. As of June 30, 2000, nearly 2,400 people were participating in ADD, saving for an average of about 13 months in the program. In ADD, 88 percent of the participants had incomes below 200 percent of the federal poverty line.
A full copy of the evaluation report published by CSD can be obtained at http://gwbweb.wustl.edu/Users/csd/, and a copy of the Executive Summary has been included as an appendix to my testimony. Some of the highlights from the report are as follows:
Good questions have been raised about how accountholders are saving—what are they giving up to save in an IDA? While we don’t have conclusive evidence yet, we can say that thus far it’s primarily through more efficient consumption that accountholders can save: they’re eating out less, spending less money on alcohol and tobacco, shopping more carefully for food, and working harder. Also, it has been asked if savings in an IDA represent new savings — is this money that would have been saved anyway? Again, we don’t have conclusive evidence yet, but anecdotal evidence is compelling that IDA savings are, in fact, new savings.
When we started doing this work, everyone said "The poor cannot save." Well, our data have shown that they can. We believe this reflects the "institutional" nature of IDAs — that people of any income will save if properly structured and supported. A good example here is the federal Thrift Savings Plan: one could wonder how much federal employees would save if it weren’t set-up for them: the match, the information, the automatic payroll deductions, the education about the plan all make it easy and attractive to save in the TSP. It’s the same for IDAs.
Finally, let me remark that, to our surprise, the "asset effects" of IDAs (that is, the psychological, civic, and social effects of IDAs) appear to be both sooner and stronger than anticipated. While we will, of course, properly evaluate the accountholders in our demonstration (including comparing them to a control group), we have been struck by the stories of the accountholders themselves – how much they say IDAs have changed their lives, their childrens’ lives, their attitudes about their future. Michael Sherraden has always said, "Income may feed peoples’ stomachs, but assets change their heads." Listening to our accountholders, this is turning out to be quite true.
III. Why Should Congress Support the Savings for Working Families Act of 2001?
The Savings for Working Families Act (the "Savings Act") was first introduced in April 1999, then again in early 2000 and, as I mentioned at the beginning of my testimony, was introduced again yesterday as a stand-alone bill. The legislation was passed by the Senate in 1999 as part of the its $792 billion tax bill, but was dropped in conference (the bill was subsequently vetoed by President Clinton). The Savings Act came very close to being enacted in late 2000 as part of the "New Markets Initiative," but was dropped at the very last minute. I am pleased to report that President Bush has included an IDA tax credit in his budget (modeled on the Savings for Working Families Act), following a campaign event he did on IDAs in Dayton, Ohio last year.
Here’s how the Savings Act would work. Financial institutions (in partnership with community based organizations) would set-up and support the accounts for qualified individuals, and provide matching funds and financial education. At the end of the year (or quarter), a tax credit could be claimed by the financial institution for matching funds provided (up to $500 per accountholder per year), as well as a couple of smaller tax credits for maintaining and supporting the accounts. I have attached a detailed summary of the legislation in the appendix.
Specifically:
- The aggregate amount of all dollar-for-dollar matches provided (up to $500 per person per year), plus
- A one-time $100 per account credit for financial education, recruiting, marketing, administration, withdrawals, etc., plus
- An annual $30 per account credit for the administrative cost of maintaining the account.
There are five reasons why, I think, Congress should support the Savings for Working Families Act of 2001:
1. IDAs are the next step for welfare reform and community renewal. Assets are the one piece of the poverty puzzle that has never really been addressed before. Welfare reform has succeeded in moving people from welfare to work, but without savings and assets it will be hard, in my view, for these families to stabilize their lives and make good investments in their children and futures. IDAs also help complete the community renewal process which began last year.
2. Asset building has a long tradition in the U.S., and reinforces basic American values of work, saving, and responsibility. We’re not asking Congress do something for the working poor that’s not already being done for the middle class. Keep in mind that not one federal dollar is spent until low-income people work and save, and some private sector dollars are leveraged. IDAs are not a government hand-out, nor are they a new poverty program, but rather a true public-private-citizen partnership, one that expands our successful asset-building system to people willing to work and save.
3. IDAs have strong, bi-partisan support and have been endorsed by a broad range of organizations. Both President Bush and former President Clinton support IDAs, as do a very wide range of Members of Congress. In addition, the Savings Act has been endorsed by the Financial Services Roundtable, United Way of America, Credit Union National Association, National Conference of State Legislatures, National Association of Home Builders, National Congress for Community Economic Development, National Council of La Raza, National Center for Neighborhood Enterprise, and many others.
4. IDAs lead to stronger retirement savings and complement the recently-passed retirement savings credit. For many if not most low-income families, retirement savings is important, but the larger issue is getting to retirement. While the vast majority of IDAs do not offer retirement as a use, IDAs nonetheless help low-income people prepare for their retirement the same way many of us in this room prepare for our retirement: by investing in a home, an education, or a small business. We have compelling anecdotal evidence that once people begin to save in an IDA, they then open up IRAs, buy life insurance, and think more seriously about their children’s futures. CFED commends this Congress for including a retirement savings credit in the recently-passed tax bill, but we believe that coupled with IDAs this savings credit will be better utilized.
5. IDAs have been tested and shown to work, but they don’t reach enough working poor families. Our Congressional sponsors have said, "Why are we limiting this great idea to demonstration projects?" The main problem with IDAs, they say, is that there aren’t enough of them. While nearly 19 million persons are potentially likely to open an IDA under the Savings Act, only 10,000 are presently using them. We will never overcome wealth and opportunity gaps through demonstration projects, and private sector funding needs to be leveraged by public sector funds in order for it to expand. Keep in mind, too, that the credit would be authorized for only five years: at that point, we hope we have made the case that IDAs work, and that this credit is worthy of expansion.
Conclusion
Without assets, poor families are likely to remain poor. And without asset development policies, only very few poor families will have the opportunity, incentive, and institutional supports necessary to save for and acquire productive assets.
Ever since the New Deal, America’s public and private sectors have spent billions on the poor in the form of income support, safety nets, rental assistance and transitional aid, but these sectors have rarely invested adequately in the poor, empowered them with assets, enabled them to own a piece of their neighborhoods, or encouraged them to build wealth. Thus, while the U.S. has succeeded in preventing the vast majority of poor families from falling through the bottom, it has failed in offering the asset-building tools necessary to let those families move from the bottom to the middle or top.
IDAs represent a new vision for America’s working-poor families: enable them to build assets, not just income; empower them to own, not just owe; view them as savers, producers, and entrepreneurs, not just recipients, borrowers, and trainees. In other words, through opportunities to save and acquire assets, invite low-income working Americans to be participants in the American economy, rather than recipients of its excesses.
In closing, it is important to observe that the entire process of family development, community building, and neighborhood revitalization begins with low-income people themselves -- it is their investments in themselves that trigger all the other investments. Nothing happens if nothing is saved; America will realize no returns on its investment in IDAs if poor people will not or cannot first invest in themselves.
With your favorable consideration of Title III of the Community Solutions Act, IDAs can move from reaching a few thousand hard-working families to millions. Mr. Chairmen, members of the subcommittees, thank you for your time. I am pleased to answer any questions you may have.
Savings and Asset Accumulation
in Individual Development Accounts
Downpayments on the American Dream Policy Demonstration
A National Demonstration of Individual Development Accounts
Mark Schreiner
Michael Sherraden
Margaret Clancy
Lissa Johnson
Jami Curley
Michal Grinstein-Weiss
Min Zhan
Sondra Beverly
February 2001
Center for Social Development
George Warren Brown School of Social Work
Washington University in St. Louis
http://gwbweb.wustl.edu/users/csd/
(314) 935-7433
Executive Summary
Long-term improvement in well-being requires asset accumulation. While saving is not easy for anyone, it is more difficult for the poor because they have few resources relative to subsistence requirements, because they lack access to some public-policy mechanisms that subsidize saving, and because scarce resources and restricted access may push saving out of their world view.
Individual Development Accounts (IDAs) are a new policy proposal designed to address these constraints and to improve access to savings institutions for the poor. Withdrawals of deposits by the poor in IDAs are matched if used for home ownership, post-secondary education, or microenterprise. Participants also receive financial education and support from IDA staff.
Do IDAs work? Data from the American Dream Demonstration (ADD) suggests that the poor can save and accumulate assets in IDAs:
Average monthly net deposits per participant were $25.42.
The average participant saved 67 percent of the monthly savings target.
The average participant made a deposit in 7 of 12 months.
With an average match rate of 2:1, participants accumulated about $900 per year in IDAs.
The American Dream Demonstration
ADD is a demonstration of IDAs in 14 programs across the United States. It is scheduled to run for four years (1997-2001), with two more years of evaluation through 2003.
The Corporation for Enterprise Development (CFED) in Washington, D.C., designed ADD and guides it. The Center for Social Development (CSD) at Washington University in St. Louis designed the evaluation.
The evaluation of ADD is the first major study of IDAs. The Startup Evaluation Report (Sherraden et al., 1999), monitored the start-up period through June 30, 1998. Saving Patterns in IDA Programs (Sherraden et al., 2000) covered programs, participants, and saving patterns through June 30, 1999. This report discusses savings and asset accumulation through June 30, 2000. A final monitoring report will cover ADD through December 31, 2001.
Data come from the Management Information System for Individual Development Accounts (MIS IDA), a software package created and supported by CSD. MIS IDA offers tools for program management and evaluation (Johnson, Hinterlong, and Sherraden, 2000). Data in MIS IDA were collected by program staff and may be the best ever assembled on high-frequency saving by the poor. In particular, records of cash flows in IDAs come from bank statements and are very accurate. The report notes carefully possible effects of weaknesses in the data.
A Theory of "Asset Effects"
IDAs aim to do more than just transfer resources to the poor. Of course, resources are good to have, if only because they can be converted into consumption. IDAs, however, expect that its transfers will be saved rather than consumed. But standard welfare transfers can also be saved. How are IDAs different?This report develops Sherraden’s (1991) proposed answer in terms of institutional theory. IDAs are packaged in an institutional structure that explicitly asks and expects participants to save their transfers in forms (such as homes, human capital, or business assets) unlikely to be quickly consumed. In contrast, standard welfare is designed to support consumption.
The institutional package matters because people are not the rational, omniscient beings assumed in economic theory. People are subject to suggestion, and they respond to patterns of choices worn smooth by public policy because that takes less effort than to imagine choices and then to weigh possible chances of consequences.Institutional theory suggests that the structure of IDAs encourages the poor to see saving as an option with positive consequences:
The existence of IDAs forges a social pattern as it sends the message that the poor can save.
Matches increase the return on savings, increase asset accumulation from given savings, and attract people to the program.
IDAs are linked to financial education that provides knowledge of how to save.
The match cap becomes a goal in the minds of participants.
Monthly statements give feedback and show progress toward goals. Furthermore, program staff and peers provide informal encouragement. The focus on success makes saving easier.
IDA programs ask for monthly deposits. This encourages saving to become a habit.
IDAs give poor people access to a way to commit to save.
Through budgets, goals, and plans, IDAs focus on the future and increase future orientation.
IDAs point out goals (such as home ownership or post-secondary education) that people might not see (or see as worthwhile) on their own.
Informal discouragement of unmatched withdrawals helps to curb dissaving.
Sherraden (1991) introduced the concept of asset effects, defined as the impacts of ownership. Humans are forward-looking, and current well-being depends in part on expected future well-being. People with more assets in the present expect to have more resources in the future. Thus—for purely economic reasons—they expect to be happier. "Asset effects" occur when ownership improves expected future well-being and thus, for psychological reasons, improves current well-being. Not only do owners think differently, but others also treat them differently. The social and political effects of ownership may matter even more than the individual effects.
Participation in ADD
Enrollment. A participant is defined as someone who enrolled in ADD and who had an account statement in MIS IDA. As of June 30, 2000, ADD had 2,378 participants in 14 IDA programs.
Graduation. About 13 percent of participants had taken a matched withdrawal. A fourth of these "graduated" and left the program, and three-fourths are still active.
Exit. About 16 percent of participants had exited without a matched withdrawal. The cumulative risk of exit in the first 12 months was 11 percent, and it was 16 percent for the first 24 months. As of June 30, 2000, 81 percent of participants were active. These and other outcomes will change with time.
Savings Outcomes in ADD
Gross deposits. The average participant had participated for 13.3 months and had gross deposits of $41.43 per month ($552 total).
Unmatched withdrawals. The size and frequency of unmatched withdrawals has been one of the biggest surprises in ADD. About 37 percent of participants made unmatched withdrawals from matchable balances, removing 25 percent of all matchable deposits. For participants who made unmatched withdrawals, the average number was 2.9, and the amount removed was $320. With an average match rate of 2:1, this implies a loss of potential matches for people who make unmatched withdrawals from matchable balances of about $640. The high opportunity cost of unmatched withdrawals, coupled with their size and frequency, highlights the difficulty of asset accumulation for the poor, even in the supportive institutional context of IDAs.
Net deposits. Net deposits are defined as gross deposits minus unmatched withdrawals minus balances in excess of the match cap. Aggregate net deposits in ADD were $838,443. Net deposits per participant were $353 ($420 for non-exits). The average monthly net deposit (AMND)—defined as net deposits divided by months of participation—was $25.42 (for non-exits, $30.30). Median AMND was $17.96 ($23.35 for non-exits). With an average match rate of 2:1, the average participant in ADD had accumulated about $75 per month.
The average match rate per dollar of net deposits was 1.96:1, and the match that corresponded to net deposits was $1,644,508. If all net deposits were used in matched withdrawals, total asset accumulation in IDAs would be $2,482,951. With exits included, this is $1,044 per participant; with exits excluded, it is $1,245 per participant. These figures will change as ADD progresses.
Matched withdrawals. Aggregate matched withdrawals in ADD through June 30, 2000 were $191,601. The average match rate per dollar of matched withdrawals was 1.82:1, and matches disbursed were $348,373. The average participant with a matched withdrawal had 2.0 withdrawals for a total of $603. Their total asset accumulation averaged $1,698.
Matched withdrawals became more common as balances were built through time; 9 percent of participants had a matched withdrawal by their 12th month, and 27 percent had one by their 24th month.
Matched uses. As of June 30, 2000, 13 percent of participants had a matched withdrawal. About 24 percent made a home purchase, 24 percent invested in microenterprise, and 21 percent pursued post-secondary education. The rest used their matched withdrawals for home repair, retirement, or job training.
About 87 percent of participants had no matched withdrawals. Of these, 57 percent intended to buy a home, 18 percent intended to spend on microenterprise, and 15 percent planned for post-secondary education. About 10 percent planned for home repair, retirement, or job training.
Net deposits as a percentage of the pro-rated match cap. On average, participants had net deposits of 67 percent of the monthly savings target (median 49 percent). At this pace, they will use two-thirds of their total match eligibility.
Deposit frequency. On average and at the median, participants made a deposit in 7.0 months per year. Non-exits made a deposit in 7.6 months per year. Some evidence suggests that frequent depositors accumulate more than infrequent depositors.
Savings rate. On average, AMND was 2.2 percent of monthly income (median 1.3 percent). The savings rate decreased as income increased. Perhaps the institutional effects of IDAs are stronger than the economic effects of greater income, and perhaps these institutional effects are somehow stronger for poorer people.
IDAs and EITC. Net deposits increased markedly in tax season. IDA participants save some chunk of tax refunds or payments from the Earned Income Tax Credit.
Costs
Policy choices require data on both outputs and costs. Cost data in MIS IDA are measured with error and are probably overstated for many reasons (for example, due to start-up costs, provision of technical assistance to other IDA programs, and data collection for the evaluation of ADD). Average program expenses (without matches) were $70.38 per participant-month, or $2.77 per $1 of net deposits. A study of the first 14 months of the experimental-design program in ADD also found costs in this range (Schreiner, 2000a). Costs in ADD did decrease with time. Average program expenses per participant-month through June 30, 1999, were $117.58; in the next 12 months, they averaged $43.06.
With a 2:1 match, total outlays in IDAs were thus roughly $6 per $1 of net deposits ($1 savings, $2 match, and $3 program expenses). This is about $2 of total outlay per $1 of asset accumulation.
Are these costs high or low? The answer depends on the as-yet-unmeasured benefits of IDAs. A standard financial benefit-cost analysis is planned for the site of the experimental design (Schreiner, 2000b). Even without precise knowledge of benefits, however, measurement of costs highlights trade-offs and sets a benchmark that encourages efficiency.
Qualitative evidence from the evaluation of ADD suggests that participants believe that intensive service is a key element of program design. A key challenge for IDA programs is then to provide such services in such a way that benefits can exceed costs. The tension between intensive service and cost structures that would allow broad access to IDAs may lead to two tiers of IDA designs, one with fewer services, lower costs, and broader outreach, and another with greater services, higher costs, and narrower targets (Sherraden, 2000).
New Savings versus Shifted Assets
IDA deposits can come from new savings or from assets converted from other forms. Even if the poor (or the non-poor) do not explicitly shift liquid assets, they can implicitly shift illiquid assets if IDAs lead to reduced investment and maintenance in non-IDA assets. High returns on IDAs may also lead savers to borrow or to repay debts slower than otherwise.
Qualitative evidence from the evaluation of ADD (Moore et al., 2001 and 2000) suggests that IDA deposits came in some unknown measure from both new savings and from shifted assets.
Program Characteristics and Savings Outcomes
The association between program (institutional) characteristics and savings outcomes matters because policy can affect program design. The results below are derived from multivariate regressions that control for a wide range of program and participant characteristics.
Match rates. A central feature of IDAs is the match rate. In regressions, higher match rates have large, strong associations with reduced risk of unmatched withdrawals and with reduced risk of exit. Match rates do not, however, have a statistically significant link with AMND.
Qualitative evidence suggests that matches attract people to IDAs; quantitative evidence here suggests that higher match rates keep people in IDAs and encourage them to maintain their balances. But higher match rates do not seem to lead to greater deposits. We believe that these estimated associations result mostly from institutional factors, but economic factors, two-way causation, and censored data also matter to some unknown extent. The data from ADD do not allow a sharp test of the effect of match rates on savings outcomes.
Monthly savings target. The monthly savings target is the amount that, if saved each month and not removed in unmatched withdrawals, would produce net deposits equal to the total match cap. On average in ADD, AMND was 67 percent of the savings target.
Higher savings targets were strongly linked with large reductions in the risk of unmatched withdrawals and the risk of exit. Higher savings targets were also strongly linked with higher AMND.
At least three forces may drive this. First, participants may change match caps into goals, leading to greater savings effort when match caps are higher. Second, AMND is cut-off for participants at the match cap. Third, programs may have assigned higher targets to groups expected to be high savers. These last two factors may induce a spurious positive correlation between the match cap and savings.
Financial education. Required financial education is a central feature of IDAs in ADD. The average participant attended 10.5 hours of general financial education. Each hour up to 12 was linked with large increases in AMND, but hours after that had little effect.
In broad terms, AMND increases with financial education (whether general or asset-specific), but only up to a point, probably somewhere between 6 and 12 hours. The content of classes probably also matters, but we did not measure it.
Participant Characteristics and Savings Outcomes
Participants in ADD are not a random sample of people eligible for IDAs; they are program-selected and self-selected. Programs target certain people, and eligibles in the target group who expect the greatest net benefits are the most likely to enroll. Results in this report pertain only to eligibles who, if they had the choice, would enroll in IDAs.
Compared with the overall U.S. population at or below 200 percent of the poverty line, IDA participants are more disadvantaged in that they are more likely to be female, African-American, or never-married. IDA participants are less disadvantaged, however, in that they are more educated, more likely to be employed, and more likely to have a bank account. These patterns likely reflect the explicit targeting of the "working poor" by programs in ADD and the client base of the host organizations.
Gender. About 80 percent of participants were female. Gender had no link with savings.
Race/ethnicity. About 47 percent of participants in ADD were African-American, 37 percent were Caucasian, 9 percent Hispanic, 3 percent Native American, 2 percent Asian-American, and 3 percent "Other." Although average AMND for all groups was at least $19.50, differences between groups were large. For example, compared with Asian Americans, average AMND was $10.58 less for "Other," $11.62 less for Hispanics, $12.77 less for Caucasians, $20.82 less for African Americans, and $22.30 less for Native Americans.
These differences are not due to race/ethnicity per se but rather to a constellation of socially produced characteristics correlated with both race/ethnicity and savings. In a perfect model that controlled for everything, the estimated link between race/ethnicity and savings would be zero.
IDAs aim to increase inclusion in institutions for saving and asset accumulation. We do not know whether IDAs increase saving or whether they increase saving more for disadvantaged groups. Although IDAs in ADD did narrow relative racial/ethnic gaps, they are not a panacea.
Education and employment. Given their income, participants in ADD were highly educated: 24 percent had a college degree of some sort, and 85 percent completed high school. Education was not linked with the risk of exit. AMND was highest for people with 4-year college degrees.
Participants in ADD also had a high incidence of employment: 78 percent worked full-time or part-time. Employment status was not significantly associated with any savings outcomes.
Receipt of public assistance. About 50 percent of participants in ADD had received some form of public assistance at enrollment or before. Current receipt of public assistance was not associated with any savings outcomes.
Income. Mean income/poverty in ADD was 111 percent (median 100 percent). About 21 percent were under 50 percent of the poverty line, and 12 percent were over 200 percent of the poverty line. The level of income was not associated with the risk of an unmatched withdrawal, the risk of exit, or AMND, but higher income was associated with a lower savings rate. Possible explanations include institutional factors, censored data, and measurement error, but we believe that institutional factors matter most and that they may be strongest for the poorest.
Insurance coverage. About 51 percent of participants in ADD had health insurance, and 31 percent had life insurance. Health insurance did not have a significant association with exit, unmatched withdrawals, or AMND. Life insurance was not associated with AMND, but it was correlated with reduced risk of exit and of unmatched withdrawals.
Asset ownership. Participants who owned assets likely had unobserved characteristics that predisposed them to save more in IDAs. For example, participants with a checking account were much less likely to exit, they were much less likely to take an unmatched withdrawal, and they had much higher average AMND. The same pattern holds for home owners and car owners.
Summary
These mid-way results from ADD will raise questions, spark debate, and inform policy. The goal of this discussion and of future research—in ADD and elsewhere—is to build knowledge about how programs that aim to encourage saving and asset accumulation can be more inclusive and generate greater net benefits.
References
Johnson, E., Hinterlong, J., & Sherraden, M. (2000). Strategies for creating MIS technology to improve social work practice and research. Paper presented at the Fourth Annual Technology Conference for Social Work Education and Practice, Charleston, South Carolina, August, ejohnson@gwbmail.wustl.edu.
Moore, A., Beverly, S., Schreiner, M., Sherraden, M., Lombe, M., Cho, E. Y., Johnson, L., & Vonderlack, R. (2001). Saving, IDA programs, and effects of IDAs: A survey of participants, research report. Center for Social Development, Washington University in St. Louis, http://gwbweb.wustl.edu/users/csd/.
Moore, A., Beverly, S., Sherraden, M., Sherraden, M., Johnson, L., & Schreiner, M. (2000). How do low-income individuals save, deposit, and maintain financial assets? Paper presented at the Inclusion in Asset Building: Research and Policy Symposium, Center for Social Development, Washington University in St. Louis, Sept. 21-23, aam1@gwbmail.wustl.edu.
Schreiner, M. (2000a). Resources used in 1998 and 1999 to produce Individual Development Accounts in the experimental program of the American Dream Demonstration at the Community Action Project of Tulsa County, research report. Center for Social Development, Washington University in St. Louis, http://gwbweb.wustl.edu/users/csd/.
Schreiner, M. (2000b). A framework for financial benefit-cost analysis of Individual Development Accounts at the experimental site of the American Dream Demonstration, research design. Center for Social Development, Washington University in St. Louis, http://gwbweb.wustl.edu/users/csd/.
Sherraden, M. (1991). Assets and the poor: A new American welfare policy. Armonk, NY: M.E. Sharpe, ISBN 0-87332-618-0.
Sherraden, M. (2000). On costs and the future of Individual Development Accounts, comment. Center for Social Development, Washington University in St. Louis, http://gwbweb.wustl.edu/users/csd/.
Sherraden, M., Johnson, L., Clancy, M., Beverly, S., Schreiner, M., Zhan, M., & Curley, J. (2000). Savings patterns in IDA programs—Downpayments on the American Dream Policy Demonstration, a national demonstration of Individual Development Accounts. Center for Social Development, Washington University in St. Louis, http://gwbweb.wustl.edu/users/csd/.
Sherraden, M., Page-Adams, D., Johnson, L., Scanlon, E., Curley, J.,
Zhan, M., Bady, F., & Hinterlong, J. (1999). Downpayments on the
American Dream Policy Demonstration, start-up evaluation report. Center
for Social Development, Washington University in St. Louis, http://gwbweb.wustl.edu/users/csd/.
The Savings for Working Families Act of 2001
Included in H.R. 7 and S. 592
|
Provision |
Language |
|
Eligibility |
|
|
Allowable uses (qualified expenses) |
|
|
Matching funds |
|
|
Federal tax credit |
|
|
Sources and limits on individual deposits into accounts |
|
|
Federal tax treatment of accounts |
|
|
Penalties for withdrawal of individuals’ own savings |
|
|
Role of non-profits and Tribes |
|
|
Role of financial institutions |
|
|
Financial education training |
|
|
How asset purchased |
|
|
Legal structure of the account |
|
|
Program certification |
|
|
Effect on mean-tested federal programs |
|
|
Administering agency |
|
|
Reporting and evaluation |
|
|
Applicable years |
|