An empirical analysis of self-reported, work-limiting disability.

This study considers the correlates of the incidence of self-reported disability. Because the data base used followed individuals for a number of years, simultaneity bias is avoided. A great number of regressors are considered. Holding a hazardous job, being recently divorced or widowed, smoking cigarettes, experiencing unemployment, and frequently working overtime are positively associated with becoming disabled, while education and risk preference are negatively associated. Moreover, the findings suggest that some of the reasons for the strong association between health and schooling in prior studies include the association between schooling and access to a safe job, less chance of unemployment, fewer cigarettes smoked, and risk preferences.