What have we learned using merged administrative data from the Social Security Administration and the Rehabilitation Services Administration?

Abstract

BACKGROUND: Policy makers have substantial interest in how the provision of employment services to persons with disabilities affects earnings and receipt of disability benefits.

OBJECTIVE: We examined the extent to which studies using matched Social Security Administration (SSA) and Rehabilitation Services Administration (RSA) data inform how provision of employment services affects employment and benefit outcomes.

METHODS: We summarize each study and consider the extent to which the findings address the effects of service provision on employment and benefit outcomes.

RESULTS: The studies provide rich contextual information about how enrollment for services is related to employment and disability program outcomes but limited evidence regarding impacts. Positive relationships between service and outcomes may confound the impacts of services with effects of other factors, such as the unobserved severity of medical conditions, motivation, or strength of the local labor market. Two studies that attempted to rigorously estimate the impacts of employment services found convincing evidence of impacts on service enrollment but no evidence of impacts on employment and benefit outcomes.

CONCLUSIONS: RSA-SSA data can facilitate estimation of employment service impacts, but to differentiate impacts from effects of confounding factors, researchers must exploit serendipitous or planned opportunities that are external to the data themselves.