The Social Security Disability Insurance (DI) program imposes strong work restrictions on beneficiaries; however, the causal effect of the work disincentives on labor supply has been difficult to estimate. We take a new look at this question by exploiting the fact that DI benefits are payable only until full retirement age (FRA), at which point they are converted to retired worker benefits, and the program’s implicit high marginal tax rate on earnings is abruptly
relaxed. Using a quasiexperimental research design, we examine whether the DI work disincentives are binding by comparing changes in labor force participation rates before and after the FRA for DI beneficiaries and non-beneficiaries. We find a relative increase in labor force participation at FRA for DI beneficiaries of 10.4 percentage points, and argue that this is likely a lower bound estimate on the labor supply disincentive effects of the DI program.