Research That Matters (January 17 - 20, 2008)

Objective: This study is the first quantitative examination of the effects of participation in Individual Development Accounts (IDAs) on accumulation of assets other than savings in IDAs. Previous research has tested both the individual and institutional effects on savings in IDAs and demonstrated that IDAs effectively help low-income persons save for approved assets. However, research has not established whether IDA participation influences participants' larger approach to saving and building assets beyond the structured environment of the IDA program. Therefore, a primary and critical issue of IDA policy is whether participation in these programs leads to a significant growth in assets beyond the saving in the IDA accounts.

Methods: Using randomized experimental data, this study examines whether IDA participants accumulate more assets than a control group, and measured using five asset categories: liquid assets (e.g., saving in checking account and saving account); other financial assets (Bonds, stocks, educational accounts, etc); total financial assets; real value assets (Value of business, car, property, and home); and total assets. This study uses three waves of data from the American Dream Demonstration (ADD), which was the first large-scale test of IDAs. Specifically, we examine data from the ADD program operated in Tulsa, Oklahoma with a sample of 1,103 participants: 537 participants assigned to the treatment group, and 566 participants assigned to the control group. Over the course of the 4-year ADD study period, sample attrition and non-participation reduced the control group to 428 participants, and reduced the treatment group to 369 participants. Using a series of multivariate ordinary least square (OLS) regression models, we investigate IDA program effects on asset accumulation beyond savings in IDAs among low-income families.

Results: The multivariate regression models show that IDA participants do not differ from control participants on measures of liquid assets, other financial assets, and total financial assets. However, IDA participants have more real value assets and total assets than members of the control group. In addition, there are no significant differences in liquid assets and total financial assets between the treatment and the control group. Participants in IDA programs were expected to transfer any liquid assets into IDA accounts; therefore, we hypothesized the treatment group would have smaller values in the two types of assets than the control group. However, we show that IDA participants have greater value in the two types of assets than the control group, although the differences were not statistically significant.

Implications: Significant increases in real value assets can be explained by the institutional structure of IDA programs where the accumulated savings are intended for investment in long-term asset accumulation through, for example, homeownership or starting a small business. The increase in real value assets may lead to increased total assets among participants in the treatment group. Thus, this study has a strong implication for public policy: inclusion should be a priority of asset-based policy. Increasing the numbers, varieties, and access to inclusive asset-based programs, such as IDAs, should be developed or expanded to serve low-income households.