Explaining Applications to the U.S. Disability System: A Semiparametric Approach

This study investigates the determinants of applications for U.S. disability benefits between 1986 and 1993 using a semiparametric discrete factor procedure separately for men and women. Approximating a dynamic optimization model, the estimation accounts for a variety of potential biases that were unaddressed in prior studies. Our results indicate different responses of men and women to variations in policy measures. Past labor earnings and fringe benefits as well as benefit eligibility and benefit amounts clearly affect application behavior.