Work, health and social insurance

Concentrating on two types of social insurance programs, Social Security Disability Insurance and Workers' Compensation, the essays which constitute this dissertation seek to contribute to the empirical literature on social insurance in two important areas. First, most of the existing literature identifies the relationship between benefit levels and program utilization without truly exogenous cross-sectional variation in benefits. Two of the essays in this dissertation rely on alternative methods of identification which give potentially better estimates of this relationship. Second, while many studies estimate the characteristics of the SSDI system and the behavioral responses to it, none have systematically examined the welfare implications of these findings for the program. The second essay undertakes such an examination.;The first essay uses trends in self-reported disability to gauge the impact of the growth of disability transfer programs on the labor force attachment of older working-aged men. Tabulations of data from several sources suggest that between 1949 and 1987, a sizable portion of the decline in the labor force participation of men aged 45-64 represented movement of men out of the labor force and onto the rolls of disability transfer programs. Since the expansion of transfer programs represents only one of the forces behind this movement, these figures represent upper bounds on the impact of such programs on work force attachment.;The second essay develops a model capturing salient features of the DI program, most notably imperfect and costly screening, and seeks to determine whether current probabilities of acceptance faced by applicants are in line with probabilities in an optimally designed program. The findings presented here indicate that while the quality of health screening information may be good enough to warrant its use, the optimal acceptance probability is highly sensitive to changes in our assumptions about the accuracy of screening.;The third essay uses what are plausibly exogenous changes in workers' compensation benefits to identify the relationship between benefit levels and the number of claims filed for workplace injuries. The results of this analysis suggest that while some relationship may exist, this and other studies may be unable to adequately measure its magnitude.