Examining Racial Inequities in BOND Impacts

This paper reexamines impacts from the Benefit Offset National Demonstration (BOND) to explore previously unexamined racial differences in participant outcomes.  It examines whether the impacts of BOND differed by participants’ race/ethnicity and the extent to which community-level racial inequities in economic conditions are correlated with participant outcomes that were central to BOND’s goals.  It pairs Social Security Administration (SSA) data from the BOND evaluation on a sample of nearly 1 million Social Security Disability Insurance (SSDI) beneficiaries with multiple measures of racial inequalities in employment and economic conditions for beneficiaries’ county of residence.  Exploration of race differences using SSA data has been limited due to inconsistent collection of race/ethnic data.  We overcome this issue by combining historical and contemporary sources of data to assign race/ethnicity to 99 percent of our sample.

The paper found that:

• In the absence of the BOND intervention, the highest levels of earnings and employment are observed among beneficiaries in the control group who are Black or non-Hispanic Asian.  Beneficiaries who are White have the highest amount and months of benefits.
• BOND increased employment and the proportion with earnings above a programmatic threshold called the BOND Yearly Amount (BYA), SSDI payments, and SSDI months for beneficiaries of color.  Increases in employment and earning above BYA were larger both for Black beneficiaries and all beneficiaries of color relative to White beneficiaries.  While absolute values of impacts were small, impacts represent about a 5 percent increase in employment and a 12 to 14 percent increase in the proportion earning above BYA relative to the control group mean.
• Our results reveal mixed associations between area-level inequities and beneficiary outcomes.  Racial parity in earnings and unemployment rates are associated with declines in employment-related outcomes for White beneficiaries and beneficiaries of color, whereas intergenerational mobility is uniformly associated with increases in employment-related outcomes for all beneficiaries.
• Combining multiple sources of SSA administrative data allowed us to identify or refine information on beneficiaries’ race/ethnicity for 99 percent of the BOND sample.

The policy implications of the findings are:

• The impacts and benefits of SSA programs may differ depending on a beneficiary’s racial or ethnic identity.  Moreover, analyses of beneficiary experiences by race and ethnicity capture experiences only after acceptance to SSDI and may not account for differential application experiences or award rates by race.
• Inequalities in local economic and social conditions are often correlated with employment and benefit outcomes that are of interest to policymakers, though not always in a uniform way.  Understanding local conditions may help inform efforts to tailor program implementation in consideration of the context in which they occur.
• Better collection and analysis of information on the race and ethnicity of beneficiaries would expand understanding of differential program experiences.