Wages and hazardous working conditions

Publisher Website
Google Scholar
Abstract
The purpose of this paper is to provide a new empirical test for the existence of wage premiums paid to workers who are exposed to dangerous or unhealthy working conditions. Prior empirical studies have encountered problems with omitted ability data, omitted working conditions data and measurement error. The empirical model presented in this paper uses a difference specification to eliminate bias from omitted ability data. By assuming that all hazardous working conditions are measured by a single endogenous unobserved variable, this empirical model also eliminates bias due to omitted working conditions and measurement error. The empirical model is estimated using the 1973–7 version of the Quality of Employment Survey. The unique data in this survey facilitate estimation of a model in which working conditions are assumed to be unobservable. Using a two-stage technique for consistent and efficient estimation, the empirical results reported in this paper show that a compensating difference for hazardous work does exist.