Conference Summary

Saving Strategies and Innovations for Low-Income Households

December 10, 2010

Columbus, Ohio


Authored by O. Emre Ergungor, Federal Reserve Bank of Cleveland, and Michal Grinstein-Weiss, Krista Holub, Clinton Key, and Andrea Taylor, Asset Building Research Group at the University of North Carolina
Printer-friendly version (PDF)

Low-income households, especially in today’s turbulent economy, face numerous challenges when it comes to saving and asset building. Nonetheless, substantial research over the past two decades has shown not only that the poor have a strong desire to save and increase their economic stability, but also that they can do so successfully. A conference co-hosted by the Federal Reserve Bank of Cleveland and the University of North Carolina’s Asset Building Research Group, “Savings Strategies and Innovations for Low-Income Households,” explored cutting-edge programs and interventions that facilitate saving.

The conference convened top policymakers, researchers, and community development specialists, including keynote speaker Mark Iwry, senior advisor to the secretary and deputy assistant secretary for retirement and health policy, U.S. Department of the Treasury. The speakers shared new research and policy developments and highlighted opportunities and challenges that will be encountered in future efforts to increase the saving rate and make all Americans financially secure.

Five key messages emerged from the discussions:

  1. Literacy training can combat innumeracy, or mathematical illiteracy, which is a major hindrance to financial security. Such training is more effective, however, if accompanied by coaching to change habits and attitudes toward saving.
  2. Low-income individuals rarely are permanently unbanked. Many have used banks in the past and continue to use them for some services. Key factors associated with the use of mainstream financial-service providers include perceptions about banks, incentives for saving, access (banks’ location and business hours), employers’ choice of pay method, and ease of access to funds in emergencies.
  3. Existing programs that encourage low-income individuals to save through addressing the key factors listed above are associated with positive personal and financial outcomes. For example, people who have assets are more likely to seek education.
  4. The biggest successes in the assets-building field have required no public funds. For instance, LMI households realized big saving benefits as a result of clarifications to laws related to 401(k)s and the option to split tax refunds into several accounts, with no cost to taxpayers.
  5. The next generation of asset-building programs needs to focus on being testable and scalable; that is, programs that can be widened to encompass large regions and populations, not just small, specific communities. They should be designed as low-cost interventions that could be delivered to millions, with clear measures of their success and impact.

Developing financial capability

Financial capability is the ability to earn, spend, save, borrow, protect one’s assets and negotiate one’s financial life/wherewithal. Developing the financial capability of all people is essential for increasing economic security and improving the financial lives of Americans. This theme permeated many conversations throughout the conference. Ben Keys of the Federal Reserve’s Board of Governors, Ruby Mendenhall of the University of Illinois, and Eleni Constantine of the Pew Health Group all described lower-income people using high-cost financial services when equal services with lower direct or opportunity costs were available. Numerous speakers also highlighted the role of innumeracy in bolstering predatory lending practices generally; particularly in mortgage lending, it is one of the factors that contributed to the foreclosure crisis.

Michael Collins, director of the Center for Financial Security at the University of Wisconsin, highlighted the challenges of achieving financial capability and security. His research links deficits in financial capability to innumeracy. However, he argues that financial literacy training and coaching are not the whole answer: Only half of the people who receive coaching stick to their financial goals. They also have to build self-control skills, and their progress has to be monitored.

Collins stressed that there is still a lot to learn about delivering financial capability training. He noted that most existing studies in this area have atypically motivated respondents, and that more rigorous evaluation designs would help to separate effective programs from less successful ones. He also suggested that there is no single solution to improving financial capability; instead, many different needs are present in the population and can be met with a variety of approaches.

Given the current interest in financial capability among policymakers and practitioners, all of the panelists agreed that more efforts were needed to integrate the latest research with the services delivered to lower-income populations. In particular, Collins suggested that the research community needs to work together to develop a consistent, validated set of metrics so that the outcomes produced by various programs could be more accurately compared. This would also help to foster broad agreement about the definition and objectives of financial capability. The service providers in attendance expressed their eagerness to engage with the research and policy communities to extend “evidence-based practice” in order to maximize the benefit to clients and to avoid delivering ineffective services.

Responding to community and individual needs and preferences

When attempting to increase financial capability and bolster the financial position of lower-income populations, it is crucial to understand and respond to the unique context in which people make financial decisions and interact with the financial system. First, it is important to understand the social and market forces that shape how lower-income households relate to the financial system.

Keys and Constantine both noted the fluidity between banked and unbanked households. Reporting on data from the Detroit Area Household Financial Services survey, Keys pointed out that many banked households occasionally rely on alternative financial services, and many who do not have an ongoing relationship with a bank use mainstream financial services from time to time. Constantine, looking at Los Angeles, found that many of the unbanked households in the study had previously been banked. Moreover, she found heavy use of check-cashing services among those who have an ongoing relationship with a bank.

Constantine’s research shows that about one-third of the unbanked live in a purely cash economy. They are paid for work in cash, use cash to pay for goods and services, and, when cash is not a payment option, use alternative financial products, such as money orders purchased using cash. Keys highlights the role of institutional geography and structure in shaping financial practices, particularly the three institutional constraints that keep the unbanked out of the mainstream financial system. First, many of them have had negative experiences with mainstream financial institutions; this reduces trust, damages perceptions, and, in some cases, makes it hard for the unbanked to qualify for opening a new account. Second, there is a dearth of mainstream financial infrastructure in the places where lower-income people live. The hours and locations of mainstream institutions are simply not convenient to the lives of the unbanked. Finally, employers of lower-income workers tend to underuse direct-deposit pay systems, although these systems often provide incentives and encourage interaction with banks (and can substantially reduce employers’ payroll fees). Likewise, Constantine’s research finds that an employer’s choice of payment method is an important predictor of whether their employees are banked. Like all populations, lower-income households respond to the incentives and structures they encounter in making banking decisions.

Another part of the economic system that shapes the unique financial needs of the lower-income population is the Earned Income Tax Credit (EITC). Last year, according to Mendenhall, EITC refunds moved 6.6 million people, including 3.3 million children, out of poverty. Mendenhall and others suggest that in many respects, EITC is replacing cash-transfer welfare programs as a key means of helping poor households, and its expansion has important implications for the financial lives of low-income households. First, because the process of claiming refunds is complicated, about 70 percent of EITC recipients pay tax preparers—often several hundred dollars—to access their refunds. Many recipients are not aware of other, often no-cost options for tax preparation, so paid preparers heavily target lower-income neighborhoods.

Because the money comes in one large lump sum each year, it creates both challenges and opportunities. Many find it tempting to spend the money relatively quickly, often to meet previously delayed needs. Alternatively, an infusion of money, usually greater than the household’s average monthly income, presents an opportunity to build savings and increase financial security.

One of the keys to creating savings with EITC money, according to Mendenhall, is making available quality products that are responsive to people’s needs. In particular, Keys, Medenhall, and Constantine all report a desire for flexible savings products that are secure and keep the money easily available for emergencies. This desire for easy access to savings leads many to save in cash. At the same time, Mendenhall interviewed many who value access in emergencies but also talked about “hiding” money from themselves to make saving easier. Often, these savers dedicate their tax refunds for a particular purpose, like a child’s education.

This resonates with the savings goals found in all of the studies. Collins finds that savings goals are vital—and that many of them are intended to benefit children. The savings goals of lower-income households and the constraints they face present several strategies for moving forward. Keys’ research suggests that a stronger presence of mainstream financial institutions in lower-income areas and a greater use of direct deposit by employers would increase the number of low-income individuals who use traditional banks. Constantine believes that stronger social connections between communities and banks could help overcome people’s historic alienation from banks and, combined with well-designed, low-cost products, could entice many into the mainstream banking system. Such changes would also enable people to save more—and more securely—when they have savings opportunities, like EITC tax refunds. New products offered by the government, such as those described by the Treasury Department’s Iwry, could be the key to meeting the unique financial needs of low-income households. Iwry emphasizes ongoing governmental efforts to increase economic opportunity and access for all Americans and to meet the needs of financial institutions as well as consumers.

Improving institutional access and incentives

The afternoon research panel focused on innovative initiatives that seek to address the challenges and meet the opportunities highlighted in the morning session. Michal Grinstein-Weiss of the University of North Carolina, Ida Rademacher of CFED, and Barb Gosse of Social and Enterprise Development Innovations talked about Individual Development Accounts (IDAs), and Caroline Schultz of MDRC talked about a workplace-based savings initiative called AutoSave. All the programs these panelists discussed seek to leverage small-dollar savings by low-income households to build assets and increase financial security. Moreover, they seek to connect underbanked populations to mainstream financial services that are low-cost and tailored to meet the needs of the consumers.

IDAs are incentivized savings accounts that offer low- and moderate-income households access to a mainstream banking product, financial education, and, if used for approved purposes, a financial match to savings they accrue. IDA accounts were first introduced in the mid-1990s and are growing in prevalence. Gosse presented research on a major Canadian IDA demonstration project called learn$ave and Independent Living Accounts. In learn$ave, participants saved for adult education and used their savings to enroll in school, with the goal of increasing future earnings and economic self-sufficiency. In the Individual Living Accounts program, homeless-shelter residents save and receive matching funds (for security deposits, utility hook-up, etc.) to move into rental housing. Both programs are intended to connect disadvantaged populations to mainstream financial services and to target financial incentives to actual needs in the lives of participants. Gosse argued that the provision of matching funds should be understood as an investment, not an expenditure. Both programs showed evidence of a positive impact on saving. Learn$ave increased school enrollment of participants relative to a control group, and no participant in the Individual Living Account program returned to the homeless shelter during the period of data collection.

Rademacher presented data from across the United States that compared the homeownership of people who used IDA-matched funds for homeownership to that of the general lower-income population. Her research shows that IDA participants had lower foreclosure rates than the general low-income homeowner population. These findings suggest that low-income individuals benefit from targeted, incentivized saving and financial education. Grinstein-Weiss presented data on the only randomized control trial on IDAs in the United States. The experimental portion of the American Dream Demonstration (ADD) began in 1998 in Tulsa, Oklahoma. A group of low-income participants were randomly assigned either to a group that had access to an IDA program or to a control group that did not. The study contacted these participants 10 years after random assignment and six years after the end of the IDA program to examine the program’s effect on long-term outcomes. Grinstein-Weiss found that although IDAs increased homeownership for baseline renters after four years, the effect did not persist. Both treatment- and control-group members experienced substantial growth in homeownership over the decade, but by 2009, the group members owned homes at a similar rate. More clear-cut was the fact that all the IDA research presented captured low-income participants’ enthusiasm for the program and the potential benefits to low-income households from well-targeted programs that increase access and incentives to use mainstream institutions.

Schultz’s discussion of AutoSave highlighted another institutional channel for facilitating saving, making it attractive and easy for low-income people. AutoSave works by intervening in the workplace to reduce barriers to saving. It partners with employers to create a channel for workers to directly deposit a portion of each paycheck into a dedicated, flexible savings account. The program responds to several needs mentioned above but also faces several challenges. By being integrated with normal financial life (the receipt of a paycheck), it meets consumers at a convenient place and removes several steps that often lead to unopened savings accounts. At the same time, a number of potential participants were suspicious of the program and worried about hidden fees or pitfalls that would cost them money. These are lessons for other savings programs that target the underbanked population.

Substantial progress, demonstrated by the programs presented at the conference, has been made in devising interventions to help meet the needs and fulfill the goals of low-income people and to connect them to low-cost, incentivized savings products. These programs, in combination with the discussion of the financial lives of low-income households, highlight the need for further innovations to increase savings for low-income households.

Designing the future of saving

The day-long conference concluded with a lively panel on the future of saving for low-income households, featuring David Newville of the Center for Financial Services Innovation, Ian Galloway of the Federal Reserve Bank of San Francisco, and Ray Boshara of the New America Foundation. The panelists identified key themes from the day and offered their thoughts on how researchers, practitioners, and policymakers could collaborate to improve savings outcomes.

Newville and Boshara both spoke of the need for policy and programs to be even more responsive to the needs and goals of consumers. In particular, Newville noted the wide variety of motives for—and modes of—saving. He pointed out that short-term saving behavior may benefit from different policy and products than long-term behavior. Boshara highlighted the need for short-term emergency saving and its potential to improve the financial stability of low-income households. Galloway and Newville also emphasized the importance of integrating public and private efforts toward increasing saving and offered examples of promising developments in this area, such as BankOn USA and the Treasury Department’s tax-refund, debit-card pilot. Boshara agreed, noting that while incentives had a place, the biggest successes in the assets-building field required no public funds. For instance, he pointed out that lower- and moderate- income households had realized big saving benefits by clarifying laws related to 401(k)s and the option to split tax refunds into several accounts, with no cost to taxpayers.

All three agreed that the next generation of programs needed to focus on being testable and scalable. Being testable, that is, demanding clear measures of success and impact, will allow policymakers and researchers to identify programs that deliver the best services for the investment made. It also allows for identifying programs that do not meet goals and could be rethought and improved. Galloway suggested that such rigorous evaluation could be used in a market-style funding system that directs resources to the most promising programs. He pointed out that the government wasn’t particularly good at “picking winners.” Instead, he argued, funding could support a diverse network of local providers, making continued funding contingent on documented success.

Scalability of future programs is the second major issue. To date, most programs designed to increase savings in low-income households use staff time and resources intensively and therefore are small. The next generation of programs should be efficiently delivered to many people and over time. Boshara refers to this as the “think-small agenda,” identifying relatively low-cost interventions that could be delivered to millions. Newville suggests that emerging technologies could help to find modes of delivery that reduce cost and increase impact. All of the panelists agree that evolution in these directions would create a hopeful future for programs that have already demonstrated an important capacity for improving the economic security and quality of life for the low-income population.