Friday, 14 January 2005 - 2:00 PM

This presentation is part of: Employment and Income

Racial Differences in Performances in a Matched Savings Program

Purpose: In order to study and understand racial inequality in America, wealth should be taken into account. Traditionally, the major indicator of well being used by social scientific researchers was income. Recently, researchers have recognized the importance of measuring household wealth independently from income. And indeed while research using income and earnings data suggests that African Americans earn significantly less than Caucasians, these differences are greater when using wealth measures. Explanations for these differences have deep historical roots and are complex. To deal with this racial wealth gap, a variety of public policy proposals have been developed in recent years. However, Wolff, (2001a) argues that despite the existence of these proposals, there is not enough evidence on their potential success to reduce the racial wealth gap (Wolff, 2001). Thus, the purpose of this study is to examine the performances of African Americans in Individual Development Accounts (IDAs), one of these policy proposals.

Methods: This study utilizes data from the final data collection of the American Dream Policy Demonstration (ADD), which is the first large-scale test of IDAs that followed participants (N=2,364) from 1997-2001. Descriptive statistics were first generated for the two groups. Next, an Ordinary Least Squares (OLS) regression analysis controlling for a wide range of factors was used. Then, with the aim of exploring unique predictors of AMND for African Americans and Caucasians in IDAs, two separate OLS regression analyses were conducted for each sub sample: African American (n=1,100) and Caucasian (n=884). In order to examine if the regression slopes in these two separate analyses are statistically different from each other, the Welch-Satterthwaite t test was used. To verify these results, an additional regression was executed to test interaction effects between Blacks and Whites with the significant independent variables from the two separate regression analyses.

Results: The results of the multiple regression indicate that while low income African Americans save in IDA programs, they save smaller amounts than Caucasians. Results from the two individual regressions indicate that hours of financial education and ownership of a checking account are associated with AMND for both the Black and White groups. Match rate is associated with AMND among African American participants. For Caucasians, several additional variables are associated with AMND; these include marital status, household size, number of dependents, level of education, homeownership, and car ownership. The Welch-Satterthwaite t test results indicate that the regression slopes of financial education, marital status, household size, the number of dependents, and home ownership are statistically different between the Black and White groups.

Implications: These results indicate that institutional structures associated with IDA programs may have more positive effects on saving performance among African Americans than among Whites. This suggests that IDA programs might be able to target Blacks differently from Whites in IDA program design and implementation. Possibilities include higher match rates, increased financial education, and assisting African Americans to navigate the banking system.

Reference

Wolff, E. (2001). Racial wealth disparities: Is the gap closing? (Public Policy Brief). Blithewood, NY: The Levy Economics Institute.

See more of Employment and Income
See more of Oral and Poster

See more of Celebrating a Decade of SSWR (January 13 - 16, 2005)