Counting the disabled: Using survey self-reports to estimate medical eligibility for social security's disability programs

We develop methodology to identify medical eligibility for Social Security's Disability programs using publicly available data. Using a structural model of Social Security's disability determination process estimated on a sample of applicants, we make out-of-sample predictions of medical eligibility for non-beneficiaries in the general population aged 18–64. This methodology defines an observable, publicly-available indicator of disability consistent with Social Security's definition. This would be useful in a wide number of applications and may be of particular interest to empirical labor economists. We find that 2.9% of this population were medically eligible but not receiving disability benefits early in 1992. Our methodology permits use of publicly available survey data to analyze medical criteria that determine who receives disability benefits. We consider effects of sample selection adjustments, sample restrictions, and several methods of estimating eligibility from continuous probabilities. Our preferred measure outperforms the conventional single variable model based on the "prevented" measure.