Ten-Year Impacts of Individual Development Accounts on Homeownership: Evidence from a Randomized Experiment

With the housing market still recovering from the subprime crisis, a key question for policymakers is how and whether to encourage homeownership and the role of down payments in the initial homeowner purchase.

A new research paper looks specifically at the long-term impact of an experiment with an Individual Development Accounts (IDA) program, which was designed to help teach financial education and provide matching funds for qualified savings withdrawals (including a 2:1 match for housing down payments) for low-income households.

The study found that, 10 years after implementation, the IDA program had no significant effect on homeownership rates, nor on the duration of homeownership over the preceding decade.

Many IDAs are implemented as Assets for Independence (AFI), a grant program within the Department of Health and Human Services' Administration of Children and Families. From 1999 through 2008, more than 50,000 IDAs were opened at 544 project sites, which provided grants to community-based organizations and local governments. The program has had an annual appropriation of approximately $25 million, which is then awarded as grants to community-based organizations and local governments to administer IDA programs.

The study focuses on a particular IDA program that ran from 1998 to 2003 in Tulsa, Oklahoma, and which is the only randomized experiment with IDAs in the U.S. to date.

Authors William Gale of Brookings, Michal Grinstein-Weiss, William M. Rohe and Clinton Key of the University of North Carolina at Chapel Hill, and Michael Sherraden and Mark Schreiner of Washington University in St. Louis find that the Tulsa program had no impact on long-term homeownership rates for the group that received access to the IDA relative to the control group or on the average amount of time people owned homes during the 1998-2009 period.

They found that about 90 percent of treatment group members opened IDA accounts, and contributions averaged about $1,800. Homeownership rates for both treatment and control groups increased substantially throughout the experiment. Prior work shows that from 1998 to 2003, homeownership rates increased more for treatment group members than for controls. But after the experiment ended, control group members caught up rapidly with the treatment group.

"A plausible explanation for the pattern of results found-a positive effect through 2003 but no significant effect after 10 years-is that is that the specific design of the Tulsa IDA experiment created incentives for treatment group members to accelerate home purchases before 2003 and for control group members to delay purchases," they write. "Specifically, treatment group members had incentives to accelerate home purchase given the 2:1 match contribution they could receive for home purchase, which was available only up to 2003. Control group members had incentives to postpone purchases until the experiment ended in 2003, at which point they could take full advantage of the homeownership programs, including financial assistance for down payment and closing costs."

The researchers did find a positive impact on homeownership rates among those with above-sample median income ($15,840) at the time they entered the program, but not on any of 23 other subgroups tested, which included by race, age, sex, education, marital status, other assets, etc. Looking at the positive results for the above-sample median income group, the authors state that it "may indicate that while IDA programs are not effective in promoting homeownership among very-low income households, they can be effective for households with higher, although still modest, levels of income." However, they note that in multiple other subgroups, they were unable to detect any impact of IDAs. As a result, it is not clear if the positive results for the single subgroup are meaningful or are simply a random "false positive." Future work will be required to sort out those competing views, they say.

The authors find that there are several other areas to explore in future research, including other qualified uses of savings on such things as home repair, small business, post-secondary education, or saving for retirement. They also note that policy interventions can have often have longer-term effects through some channels even if the short-term effects through other channels fade out - "for example, small class size may have temporary impacts on test scores but longer-term impacts on non-cognitive aspects of behavior and earnings. It is important to know whether financial education, the encouragement to save, and the opportunity to have accumulated funds during the IDA program could have longer-term effects, even if controls had caught up six years after the program ended," they conclude.