Understanding the Local-Level Predictors of Disability Program Flows: New Adult Awards and Beneficiary Work Activity

Abstract

This paper examines factors that are associated with area-level benefit awards for Social Security Disability Insurance (DI) and Supplemental Security Income (SSI) as well as the work activity of DI and SSI beneficiaries. Although the Social Security Administration (SSA) cannot directly affect state policies or local economic conditions, there is value in understanding the extent to which these policies and conditions might correlate with application rates, benefit receipt, and beneficiary return-to-work rates.

We conducted our analysis at the level of Public Use Microdata Areas (PUMAs), which are geographic units created by the U.S. Census for statistical purposes. PUMAs are within-state geographies that have a population of at least 100,000 people and are large enough to produce statistics on low-occurrence events such as beneficiary suspensions and terminations for work. We aggregated data from the Social Security Administration’s Disability Analysis File, the American Community Survey, and other national sources.

We assess the variation across PUMAs in the rate of new benefit awards and beneficiary work outcomes in 2017. We also consider the association between area-level demographic, economic, health, and health services availability and those beneficiary outcomes from 2005 through 2017. We find that:

• Award rates in both DI and SSI were highest in 2017 in Appalachia (particularly where Kentucky, Virginia, and West Virginia meet) and southern states such as Mississippi and Alabama. They were also relatively high in western states such as New Mexico and Washington. We found that these patterns were relatively consistent across the years of our analysis.

• In a multivariate framework, new benefit awards from 2005 through 2017 were higher in areas with higher shares of the population that were female, did not have a college degree, had a disability, received Supplemental Nutrition Assistance Program (SNAP), or were in poverty and in areas with a higher cost of living (as proxied by wages, rent, and housing values). Most other factors we considered were only weakly associated with new benefit awards.
• Among both DI and SSI beneficiaries, the shares with positive earnings and with cash benefits forgone because of substantial work activity in 2017 were highest in the Great Plains region—with especially high shares of beneficiaries who work in the eastern parts of North Dakota and South Dakota, the southern and western parts of Minnesota, and the northern part of Iowa. This general pattern held in earlier years of our analysis as well.
• In a multivariate framework, the shares of beneficiaries with positive earnings from 2005 through 2017 were higher in areas with higher concentrations of the population over age 65 and in manual labor or service sector jobs, with a higher employment rates of people with disabilities, and with higher obesity rates. The share with positive earnings in those years was lower in areas where a larger share of the population was female, childless, had a disability, received SNAP, or in poverty and in areas that had higher cost of living. The associations between these factors and beneficiaries whose DI and SSI cash benefits were forgone for substantial work activity generally were in the same direction. Other factors were less strongly associated with beneficiary work outcomes.

 
The policy implications of the findings are:

• SSA does not directly control policy levers to affect the demographic, economic, health, and health services factors we considered. Nonetheless, knowing which area factors are correlated with better or worse work outcomes among beneficiaries may help targeting mailings for the Ticket to Work program or devoting resources to programs like the Work Incentives Planning and Assistance program to certain areas. Our cross-sectional and longitudinal analyses painted a consistent picture that areas with lower levels of economic opportunity—and areas with reductions in “affluence” over time—might be advantageous areas to target. Using trend data may help predict areas of opportunity before waiting to observe new awards or beneficiary outcomes.

• It is important to note that our analysis is descriptive and we cannot and do not ascribe a causal relationship to observed factors. Ultimately, the intent of our analysis was to shed light on factors that SSA may want to use to target supports or outreach to potential applicants or current beneficiaries. For that purpose, associations between factors may be sufficient.