SSDI receipt is strongly correlated with education but determining whether this association reflects a causal effect of education on SSDI receipt is challenging. Among individuals ages 60 to 64, more than one in five high school dropouts receives SSDI, as compared to about one in fifteen college graduates (Favreault et al., 2013). Education may affect SSDI receipt through various channels, such as occupation and health. Education can also factor into the SSDI award decision directly through its inclusion in the medical-vocational grid. Yet early-life health or family background could affect both education and SSDI receipt, generating an association between them that is not reflective of a causal effect.
This project will estimate the causal effect of education on SSDI receipt and earnings, exploiting variation arising from the elimination of the Social Security Student Benefit Program (SBP) in 1982. The SBP extended dependent benefits to children of deceased and active Social Security beneficiaries through age 22 if the child was enrolled in college. The program served 12 percent of full-time college students at its peak and led to a large increase in college enrollment among eligible beneficiaries (Dynarski, 2003), yet its long-term effects have not been studied. This project contributes to Area #1, trends in disability, as it sheds light on how changes in educational attainment in the population may affect the SSDI rolls. The project contributes to Area #4, informing long-term projections and models, as it estimates the effect of education on earnings.