Helping Low-Income Families Save: Lessons from Tax-Time Savings Accounts

| Posted: April 20th, 2012

With the tax season now over, it’s a good time to point out that tax-time matched savings programs are among the more promising approaches to boost low-income savings. In these programs, tax filers are encouraged to deposit some of their tax refund—primarily from the earned income tax credit (EITC)—into a savings account. Those deposits are then supplemented with match funds. One successful example is New York City’s $aveNYC Program, run through participating Volunteer Income Tax Assistance (VITA) sites. The program pays a 50 percent match on amounts up to $1,000 held in a $aveNYC account for a full year. In 2008-2009, the initial years of the program, 79 percent of participants received match funds. Of these accountholders, 71 percent re-upped the following year.

A just-released report by UNC-Chapel Hill’s Center for Community Capital provides insights on low-income savings behavior based on in-depth interviews with 48 $aveNYC participants—mostly African-American and Hispanic women between 25 and 50 years old. Consistent with other research, the study highlights the fact that children provide the strongest motivation for low-income workers to save because of the obligation to meet kids’ basic living needs, the instinct to serve as a role model for them, and the desire to provide them a better environment for growing up. To be realized, these impulses to save must be combined with confidence in one’s ability to save and a sense of trust in the financial institutions that hold one’s savings. And low-income households often lack that confidence and trust.

The basic message from this recent research—and the emerging body of work in behavioral economics—is that savings interventions need to nurture the intention to save and also make the act of saving easier. Whether someone is able to save (given that they can afford to) depends on a complex tug-of-war between their current impulses and their future plans. Savings tools such as precommitment, default-in/opt-out decision framing, and envelope budgeting are ways to enable our better selves to prevail.

Community-based savings collaboratives, such as savings circles where people pool their savings, have had some success helping the very poor save in developing countries. Is there a scalable way to do something similar in urban America? Can the motivational influence of one’s children be more effectively harnessed?

One possible untested approach is setting up a virtual savers’ club comprised of parents whose children share the same birthday. Parents could create a savings account on their child’s behalf and set a savings goal to be reached by the child’s next birthday. Members would receive periodic savings reminders and could compare their progress with the progress of other club members. This kind of self-induced competition has worked in energy conservation initiatives.

As with $aveNYC, the financial incentive to save could be strengthened by match funds—in this case, funds directed to a child’s account through donations from other family members or friends (or even a noncustodial parent). If this virtual savers’ club were widely marketed, other people who share the same birthday might be interested in making charitable donations of match funds to a child’s account.

In each of these cases, the necessary focus for savings interventions is clear: nurture the intention to save and enable the act of saving.

Share this:

To Save or Spend: Meeting the Short-Term and Long-Term Consumption Needs of U.S. Households

| Posted: March 16th, 2012

Two just-published reports provide useful insights on patterns of household saving in the United States. The analyses highlight the policy dilemma that comes during and after recessions. As American consumers, we tend to save more than we should during recessionary periods, when more spending would stimulate economic expansion. We then tend to fall back into myopic spending habits, when more saving would promote the economy’s long-run growth potential and help provide for our own long-term needs.

Amidst the statistical avalanche of the 2012 Economic Report of the President are the most recent numbers on personal saving (as a percent of disposable personal income) before, during, and after the Great Recession. From a pre-recession level of 2.4 percent in 2007, the saving rate more than doubled to above 5 percent in 2008–2010, exceeding 6 percent in some quarters during the recession. This reflects the collective reaction to the enormous drop in household wealth that was a result of plunging stock market values and housing prices. The total wealth decline was the equivalent of 1.8 years of income for the average household, the steepest drop since such data were first collected in the early 1950s.

During 2011 the saving rate then fell to below 4 percent as pent-up demand for consumer durables (especially cars) buoyed consumption. This has been welcome news for the economic recovery, but it suggests a return to a historical path of not saving, a matter of concern for our long-term economic health.

Barry Bosworth of the Brookings Institution focuses on this long-run horizon in his book The Decline in Saving. As Bosworth points out, our personal saving rate has declined over the past three decades. From an average of 7.7 percent in the 1970s, it dropped to 7.2 percent in the 1980s, 4.7 percent in the 1990s, and then 2.4 percent during 2000–2007. As explanations for this decline, Bosworth points to easier credit availability and financial innovations that enabled households to extract equity from homes and other assets, fueling a more consumption-oriented economy. He notes that Canada is an instructive comparison, observing that “the mortgage market innovations that led to the growth of the subprime mortgage market in the United States were largely absent from Canada.”

Whether American households under-save or over-save is a complex question from a macroeconomic perspective; it hinges importantly on the level of saving or dis-saving in the corporate and public sectors.  The evidence suggesting a return to long-term trends of under-saving in the household sector is discomforting, with discernible risks apparent as one moves down the income distribution. Do households have sufficient assets to weather financial emergencies, to withstand national economic downturns, and to meet their retirement needs? If upcoming data show continued low rates of personal saving, in a world where low interest rates do little to encourage thrift, we should heighten our focus on savings initiatives, especially targeted to low- and middle-income consumers, to break the pattern of saving too little.

Share this:

“Liquid Asset Poverty” and Prolonged Joblessness: The Recession Ripples On

| Posted: February 29th, 2012

On January 31, the Corporation for Enterprise Development released its annual Assets and Opportunity Scorecard. For the first time, the report included estimates of “liquid asset poverty”—the share of American households with insufficient liquid assets (e.g., bank accounts, stocks, mutual funds, and retirement accounts) to subsist at the poverty level for three months. In 2009 an estimated 43 percent of U.S. households did not have enough liquid assets to protect themselves from a major income loss or emergency expense. This represented a slight uptick from the pre-recession level of 41 percent in 2006. Among states, liquid asset poverty rates in 2009 varied by a factor of more than two, with Hawaii, New Hampshire, and Vermont below 25 percent, while Alabama, Georgia, Mississippi, and West Virginia each topped 55 percent.

At a time when the government’s official measure of income poverty has undergone serious review, we should also apply the same critical eye to asset poverty statistics. In light of the prolonged unemployment spells recently experienced in the American labor force, the three-month measure seems too limited. In 2009 more than half (51 percent) of the nation’s jobless were without work for 15 weeks or more.  This figure jumped to 59 percent in 2010, and these estimates do not count discouraged workers no longer looking for jobs.

In which states was the combination of pre-recession liquid asset poverty and subsequent prolonged joblessness particularly acute? The map below shows [with horizontal gridlines] the 21 states whose 2006 liquid asset poverty rate exceeded the 41 percent national average.

States With High Rates of Both Liquid Asset Poverty and Prolonged Unemployment Are Concentrated in the South

Source: Corporation for Enterprise Development and U.S. Bureau of Labor Statistics

The map also identifies [with vertical gridlines] the 19 states that experienced disproportionately high rates of longer-term joblessness, as evidenced by a 2009 median unemployment duration of 13 or more weeks (i.e., three months or longer) coupled with an unemployment rate at or above  that year’s 9.3 percent national average. Of the 10 states (in cross-hatch) where both liquid asset poverty and prolonged joblessness were prevalent, 8 were in the South. Workers in the South have clearly faced a dual challenge, with a risk of hardship greater than that confronted in other regions.

The focus on liquid asset poverty is long overdue. As more and more data become available on this important dimension of household economic security, we should be attentive to issues of measurement and geographic concentration.

Share this:

Family, Friends, and Communities: Support Networks Reduce Hardship

| Posted: January 25th, 2012

As our economy recovers fitfully from the Great Recession, households throughout the United States strive to avoid hardship while dealing with lost income and depleted assets. Middle-class families find themselves in a situation they have rarely encountered: staying one step ahead of deprivation. For households across the income distribution, effective coping requires accessing the economic and emotional support of nearby family members, friends and neighbors, and community organizations.

My colleague Sisi Zhang and I explored recent patterns of social support and material hardship using data from the Survey of Income and Program Participation. We have analyzed the responses of households interviewed in the summer of 2010, about one year after the declared end of the recession.

Expectations Vary in Level of Help From Family, Friends, and Others

Source: Urban Institute analysis of data from Survey of Income and Program Participation

Nationally 42 percent of households indicate that, in a situation of need—such as becoming sick or having to move—they would expect to receive all the help they need from family living nearby. Lesser percentages would rely primarily on friends (33 percent) or other people in the community (19 percent). Interestingly, nearly one-third (31 percent) expect to get no help from such community sources as social agencies or churches. This group is divided equally between those who expect to rely fully on family or friends—and thus needn’t seek other support—and those for whom the lack of established connections with community-based organizations might put them at risk of hardship in a situation of need.

How important is an established support network for fending off hardship? The evidence suggests that it’s very important, based on our multivariate analysis of the SIPP data. Let’s focus on the findings for four indicators of hardship—a missed rent or mortgage payment, a missed utility payment, a forgone doctor visit, and a forgone dentist visit. For each measure, between 9 and 12 percent of households experienced such a circumstance within one year of their interview. (Some of these households experienced multiple forms of hardship.)

For each indicator, those with an established support network—as evidenced by their expectation of receiving all help needed from family, friends, or others—experienced a significantly lower rate of hardship. These effects, relative to those expecting no help from each source, are estimated while controlling for such characteristics as the householder’s age, gender, race, ethnicity, and education, plus the household’s size, income, assets, and debt.

Strong Support Networks Reduce Likelihood of Hardship

Source: Urban Institute analysis of data from Survey of Income and Program Participation

Having a strong support network appears to provide considerable protection, lowering by about one-third the incidence of hardship—specifically, by 2 to 5 percentage points for each indicator, relative to its average of 9 to 12 percent.

The support available through family, friends, and community is clearly a key aspect of maintaining a household’s economic well-being. The role played by one’s support network seems especially important in a weakened economy, when these sources of help constitute a critical layer of protection beyond that provided by one’s own resources and publicly provided benefits and services.

Share this:

Getting American Households to Save: Lessons from Abroad

| Posted: December 28th, 2011

American households save far less than their counterparts in other OECD countries.  In his insightful new book, Beyond Our Means: Why America Spends While the World Saves, Princeton historian Sheldon Garon puts our low net savings rate in perspective.  Yes, it trended upward during the 2000s, but it’s still less than half of the double-digit rates in such economically robust European countries as Austria, Belgium, France, Germany, Sweden, and Switzerland.

Net Household Saving Rates For Selected OECD Countries, 2000-2009 (Percent Of Disposable Household Income)

Source: Sheldon Garon, Beyond Our Means: Why America Spends So Much While The World Saves

As for why, Garon points to institutions that emerged in Britain, continental Europe, and the Far East during the 19^th and early 20^th centuries to promote small-dollar savings among younger generations and the working class.  Thrift was bred through savings banks in schools, post offices, and other community institutions.  We saw their like here in the Northeast and Upper Midwest, reinforced by patriotic savings campaigns during World Wars I and II.  Yet, our government has never nurtured saving behavior the way many others have.  As Professor Garon observes, instead of trying to help democratize saving, our policy has increasingly deregulated credit, especially during the 1980s and 1990s when credit card borrowing, home equity lines of credit, and subprime mortgage lending all exploded, sweeping  unwitting low-income borrowers into the feeding frenzy.

Now that we’re striving to spend our way through an economic recovery, let’s not allow the low-income population to get caught in the groundswell again.  Working families need savings incentives even stronger than those our tax system gives to middle- and upper-income families.  At a minimum, we shouldn’t allow deficit reduction to undermine the few savings programs designed to help the poor and near-poor.  It’s consoling that in the pending FY12 omnibus appropriation Congress didn’t gut individual development accounts (IDAs) under the Assets for Independence Act, which has provided matching funds for the savings accounts of more than 78,000 low-income households since 1998.  (The conference committee reduced this program by 16 percent, far less than a 63 percent cut earlier proposed in the House.)

As the economic recovery moves along, the short-term challenge in boosting personal saving will be to help households weather income shocks and meet emergency expenses as they regain liquidity.  Looking farther ahead, we need easier ways for would-be small-dollar savers to realize their good intentions.  School savings banks and postal savings banks seem antiquated now, but we should focus our energy on devising modern-day, higher-tech equivalents.

Share this:

Reaching The Unbanked: What Would Willie Sutton Do?

| Posted: November 28th, 2011

Now that Bank of America has pulled back its fees for debit-card holders, let’s turn to those who are really on the outside looking in at mainstream financial services: American consumers with no bank account at all. As measured by the FDIC’s January 2009 survey, 7.7 percent of U.S. households are unbanked. That’s 17 million adults.  Another 17.9 percent of households—some 43 million adults—are underbanked. They have bank accounts, but still make some use of payday loans, pawnshops, and other alternative financial products.  

The unbanked population is predominantly urban. Fully 81 percent of unbanked households reside in Metropolitan Statistical Areas (MSAs). If we look at the unbanked rates of very large MSAs that are also represented in the FDIC survey by at least 100 sample households, we find that 25 of these 69 metros have unbanked rates above the 7.7 percent national average.

All Regions Contain Large MSAs With High Unbanked Rates

Among these high-need areas, Memphis ranks highest at 17.3 percent.  Indeed, three-fourths of its Census tracts have unbanked rates exceeding the 7.7 percent benchmark, according to recent estimates by Corporation for Enterprise Development (CFED)—which has just released a powerful data tool for researching local patterns of bank use.

In the past five years CFED has also supported the creation of BankOn initiatives. These public-private partnerships promote access to mainstream retail financial services by negotiating with banks and credit unions to get them to offer starter or “second-chance” accounts. Among the 56 such programs up and running now, most focus on a core city or metro area while others are state or county entities.

How well do BankOn programs cover the neediest metro areas?  Quite simply, much has been accomplished and much remains to be done.  Of the 25 high-need MSAs, only 12 have a BankOn program.

BankOn Initiatives Are Widespread, But Do Not Reach Many High-Need MSAs

Among the 13 others are some with unbanked rates as high as Tulsa’s 12.6 percent.

Because most financial services providers work in a particular state or region, we need to develop strategies to better serve the unbanked residents of the high-need communities they serve. That means concentrating on the interior southwest (Riverside, Tucson, and Albuquerque), the central plains (Wichita, Tulsa, Oklahoma City, and Des Moines), the Great Lakes (Milwaukee, Buffalo, and Rochester), and the interior southeast (Birmingham, Atlanta, and Charlotte).  BTW: Charlotte is Bank of America’s corporate headquarters.

Willie Sutton robbed banks because “that’s where the money is.”  Banks, credit unions, and other mainstream institutions need to reach the underserved segments of their retail markets because that’s where the need and opportunity are.

Share this:

Alpha Testing and Randomized Control Trials: Improving on the Gold Standard

| Posted: October 19th, 2011

In the world of policy research, experimental evaluations—or randomized control trials (RCTs)—are the gold standard.  Indeed, they are the most rigorous way to estimate a program’s effects on participants. That’s because the results for the participants—the “treatment group”—are measured against a randomly picked “control group” that doesn’t enter the program.

Increasingly, RCTs are being used to find out if programs providing financial services to low­-income people work. One recent large-scale example—evaluated by the Urban Institute—is a Treasury Department study of the effects of offering (through the Green Dot Corporation) prepaid debit cards to low-income tax filers so they can direct-deposit their federal tax refunds. In the 2011 tax filing season, 950,000 filers nationwide were randomly assigned to either a control group or one of eight other groups that received differing card offers.

A key element of any experimental study is the take-up rate of whatever is being offered to the treatment group. The closer the treatment group’s take-up rate to program expectations for an eventual operational roll-out, the more reliable the study’s assessment of program impacts.

Behavioral economics has taught us that our financial decision-making is deeply influenced by subtle contextual factors that frame our choices. Given the importance of this “choice architecture” and the pivotal role of the treatment group’s take-up, it’s surprising that so few program offers are pre-tested before they are evaluated. Ironically, government routinely requires pre-testing of the program evaluation questionnaires even though the program’s offer is rarely vetted in advance.

Social science should take a lesson from the computer industry. It uses “alpha tests”—small-scale short-term acceptance testing in an operational setting. Such testing goes beyond customer focus groups (which is what the Treasury study used), and a few funders are trying it. The StabilityFirst pilot test, conducted in 2010 by Harvard’s “ideas42” center on applied behavioral economics, enrolled 20 students at Central New Mexico Community College in Albuquerque into a prepaid debit card program. The students were interviewed at length both before and after to gauge their reactions to the program. A range of issues surfaced, including difficulty resolving customer service matters. Participants were reluctant to make calls to the customer service line, not wanting to commit scarce cellphone minutes for a possibly lengthy call with time spent being transferred or on hold.

Alpha tests like the one in Albuquerque can help researchers identify design features that inhibit take-up. And once these “blocking factors” are known, they can be corrected before an experimental evaluation is launched, making randomized trials far more useful and the outcomes sought more likely to come about.

Share this:

New Census Data Suggest Who's Likely to Stay Poor and Why

| Posted: September 26th, 2011

We all learned last week from the Census Bureau that the poverty rate rose to 15.1 percent in 2010.  Amid the blizzard of statistics was this factoid:  20 percent of U.S. households live on less than $20,000 a year (cash income in 2010 dollars). Underlying this one-year snapshot is a complicated dynamic. Often, household incomes fluctuate week by week and month by month, especially among those in the lower income ranges.

Given these gyrations, what’s likely to happen to the incomes of the 24+ million American households in this lowest quintile in the coming years?  For starters, recent published Census data for the 2004 panel of the Survey of Income and Program Participation--a national sample tracked from 2004 through 2007-- justifies some optimism. In those pre-recession years, the economy was reasonably stable, with unemployment below 6 percent (compared to today’s 9+ percent) and the poverty rate was below 13 percent (compared to the just-announced 15+ percent). But even ignoring that we’re in harder times now, these numbers raise concerns.

Bottom-Quintile Households in 2004 Experienced Wide-Ranging Income Changes by 2007

If we look at what happened to households in the bottom fifth during the three years tracked, it’s a mixed bag. More than two-thirds (69 percent) stay  in the bottom quintile, divided roughly equally among those whose annual incomes drop by more than 10 percent, stay about the same, or rise by more than 10 percent. The good news—again, based on 2004-2007—is that 19 percent move up to the second income quintile, and—remarkably—12 percent move up by two or more quintiles.

Based on this mid-decade experience, which low-income households enjoyed the greatest upward economic mobility? The educated, the married, the young, and non-Hispanic whites. The chances of advancing two or more quintiles within three years were appreciably higher for those with a bachelor’s degree: five times higher than for those who didn’t finish high school. Married people were twice as likely as the never married to jump two quintiles. And those 15 to 24 were twice as likely as those 45 to 64 to make the leap. Meanwhile, the odds for non-Hispanic whites were two-thirds higher than for blacks.

Here’s the concern.  Over the past decade, the demographic profile of those in the bottom fifth has changed in ways that dampen upward mobility.  Relatively more householders in this group have never been married. More are 45 to 64. And more are black. Combined with the risk that the economy performs fitfully over the next several years, income growth isn’t likely to follow the normal pattern of recent economic recoveries.

Remember that earlier factoid?  By this time next year, we may learn that households at the 20^th percentile of the income distribution have incomes of less than $20,000.   We’re climbing out of a hole that’s deeper and steeper than before, and making progress against poverty is all the more difficult.

Share this:

Credit Access for the Unbanked and Underbanked—Economic Recovery’s Missing Ingredient

| Posted: August 17th, 2011

Low interest rates are good for those seeking credit, right? In principle, the Fed’s resolve to stay its course in promoting economic recovery over the next several years should help the unbanked and underbanked population.  In fact, interest rates in mainstream financial markets mean little to those who access financial services—for credit, savings, and payments—through nonbank institutions such as payday lenders, pawnshops, rent-to-own stores, refund anticipation lenders, auto-title lenders, check cashers, and the like.  These alternative providers market their services to low- and –moderate income (LMI) consumers, who can ill-afford the high associated fees and charges.

Just how prevalent is the use of non-bank credit products? Looking especially at state differences in product use, we tabulated data from the Current Population Survey (CPS) January 2009 supplement on the unbanked and underbanked. This FDIC-sponsored and Census-administered questionnaire was completed by about 54,000 households nationally, roughly one year after the Great Recession began.   The survey asked “whether you or anyone in your household have . . .”:

• used payday loan or payday advance services (ever);
• sold items at a pawnshop (ever);
• rented or leased anything from a rent-to-own store (ever); or
• taken out a tax refund anticipation loan (in the past five years).

Those who said yes to one or more of these credit sources we classified as alternative credit users.

The findings? Among households nationally, the average rate of alternative credit use was 11.6 percent, with multi-fold differences among states reflecting variations in supply and demand factors.  By Census region, the highest rate of use was in the South (13.6 percent), followed by the Midwest (12.3 percent), West (10.9 percent), and Northeast (7.4 percent).  The rate of use exceeded 12 percent in eight states, five of them in the South.  Arkansas topped the list (19.8 percent) followed by Kentucky, Oklahoma, Texas, Alaska, Montana and South Carolina (ranging between 17.8 and 16.3 percent).  At the other extreme, usage rates were under 8 percent in seven states, all but one in the Northeast.  Delaware came in at only 4.3 percent, followed by Massachusetts, New Jersey, New York, New Hampshire, Vermont, and Rhode Island (ranging between 5.9 and 8.0 percent).

Rate of Alternative Credit Use, by Region and State (% of Households)

You’d hope that historically low interest rates would open the door somewhat wider to mainstream  credit products for the LMI market segment (those below 200 percent of the poverty level).  To the contrary, lenders are now requiring higher down payments and are imposing tighter underwriting standards  (e.g., credit scores and debt-to-income ratios), even  for their more credit-worthy customers.  This will continue to force the LMI borrowers to rely on alternative high-cost loan products, which may become even more costly as regulatory actions (such as the curtailment of refund anticipation loans) limit the products available to the already underserved market.

To promote a broad-based consumer-led recovery, policies aimed at making low-interest credit available will need to improve credit access across income levels and geographic regions.  One indicator of success will be diminished reliance of the unbanked and underbanked on high-cost alternative credit sources.  This will bear watching as the economy struggles to gain traction over the coming years.

Share this:

On Again, Off Again: "Churn" Among Food Stamp (SNAP) Participants

| Posted: July 13th, 2011

The Great Recession and the Not-So-Great Recovery have triggered enormous growth in the Supplemental Nutrition Assistance Program (SNAP), once called Food Stamps.  Unquestionably, the program has buffered the nation’s recessionary jolt and has averted food insecurity and material hardship.  As of April 2011, more than 45 million persons—one in seven—are receiving SNAP benefits.  With the recession officially ending in June 2009, some states—New Jersey, New Mexico, and Maryland—have seen subsequent growth rates in SNAP enrollment of more than 20 percent (for April 2010-April 2011).

Large cities are experiencing even higher SNAP growth, with year-over-year caseload increases topping 30 percent (for May 2009-May 2010) in such major metros as Houston, Jacksonville, Las Vegas, San Diego, and Wichita.   For some of these cities, the growth reflects catch-up from previously below-average SNAP participation rates (recipients as a share of program-eligible individuals).   As measured in 2008, participation rates were below the national average of 67 percent in San Diego (40 percent), Houston (60 percent), and Las Vegas (64 percent).  This pattern of catch-up growth in major cities is in part responsible for the desired increases in the national SNAP participation rate over the past decade, from a low of 54 percent recorded in 2002 following the nation’s prior recession.

So is SNAP doing its job?  Mostly yes, but a troubling problem is “churn”: recipients who leave the program and then re-enter within four months.  To some degree, churn means that the program is responding, as intended, to fluctuating incomes and shifting household arrangements.   But many churners qualify for benefits when they are cut off for procedural reasons, perhaps missing an appointment or omitting a form.  These households, often with children, aren’t getting the steady stream of help for which they’re otherwise eligible.

Many program practitioners think  that churn among benefit-eligible households  is becoming more prevalent.  If so, churn makes it even harder to achieve further progress in raising the national participation rate so that more of those who need help from SNAP will receive it.  The prevailing policy concern focuses not only on benefit loss among needy clients but also on added administrative costs, as state and local workers spend countless hours to reopen cases when these clients re-apply a few months later.  (The federal government covers only 50 percent of the program’s administrative costs, versus 100 percent of the benefits.)

Many states are streamlining administrative procedures to reduce the time, expense, and hassle of households’ becoming initially certified and then periodically recertified for benefits—e.g., through on-line applications, simplified rules for reporting changes in earnings, and not requiring face-to-face interviews at recertification.  More can—and should—be done to help low-income families get and keep the nutritional support they need and are entitled to receive.

Share this: