What explains state variation in Social Security Disability Insurance (SSDI) application rates?
Social Security Disability Insurance (SSDI) applications and receipts vary greatly by state (McVicar 2006; Bound and Burkhauser 1999; Rupp and Stapleton 1998), which has led to concerns about potential inconsistencies in the application of disability standards. This possibility has prompted numerous Congressional hearings and reports and led the Social Security Advisory Board (2001a; 2001b) to express concern about the Social Security Administration’s (SSA) ability to disentangle the potential causes.
Much of the previous work focuses on the SSDI rolls, allowance rates or award rates; not as much attention has been paid to application rates (also referred to as filing rates) since Rupp and Stapleton (1998) summarized the known factors affecting caseloads.
Work by the Social Security Administration (SSA 1988) finds that economic and demographic differences are significant factors in explaining state SSDI application rates, but their sample lacks information about the underlying health of the population. Strand (2002) advances this work by adding some, albeit limited, health and health insurance information but also covers a very short time period (1997-1999). Duggan and Imberman (2009) explore the relationship between applications and the unemployment rate but do not control for other potentially confounding factors.
While it is debated whether the overall growth in the SSDI rolls could be attributable to increased underlying disability, clinical measures of health exhibit substantial state-level variation. For example, age-adjusted mortality rates are 25 percent higher in Mississippi and 22 percent lower in Hawaii than the U.S. average (National Center for Health Statistics 2010). Selfreported disability varies even more, even after controlling for individual characteristics (Subramanian, Kawachi, and Kennedy 2001), although this variability may in part reflect labor market conditions (Parsons 1982; Haveman, de Jong, and Wolfe 1991; Gruber and Kubik 1997; Currie and Madrian 1999; Bound and Burkhauser 1999). Variations in poverty rates between states also may lead to differences in the incidence of mental and physical impairment (McCoy, Davis, and Hudson 1994).
In addition to investigating the extent to which this geographic variation in SSDI applications reflects differences in demographics, disability, and state economies, this paper also examines the correlation between state policies and application rates. States play a critical role in both the administration of the SSDI program and, indirectly, in determining its attractiveness. State governors appoint the top state-level administrators of the program, which could directly affect program administration (Iyengar and Mastrobuoni 2008). If the program administration makes acceptance more likely, decreases wait-time, or in other ways streamlines the application process, applications could increase. State policy indirectly affects the relative value of the Medicare component of receiving SSDI benefits by affecting the accessibility and affordability of other forms of health insurance through regulation of the private insurance market and the administration of the Medicaid program. States also determine the generosity and duration of unemployment insurance benefits, affecting the valuation of the financial benefit of SSDI.
Finally, this paper examines the total SSDI application rate as well as the application rate broken down into its two components: the SSDI-only applications and the concurrent applications to SSDI and the Supplemental Security Income (SSI) program. Individuals must meet work-history requirements, both recent and over their lifetimes, and fit strict disability criteria to be eligible for the SSDI program. SSI is a means-tested program for individuals who can no longer work, using the same disability criteria as the SSDI program for non-elderly applicants but augmenting the work-related benefit. By separating the two populations we test the hypothesis that the two populations respond differently to specific state policies.