Compensating wage differentials for workplace accidents: Evidence for union and non-union workers in Tunisia

This paper estimates the compensating wage differentials for job risks for union and nonunion workers. these estimated compensating wage differentials are used to measure the statistical value of life. to avoid a potential problem of a selectivity biais arising if richer people choose safer jobs,we consider risk as an endogenous variable. the endogeneity of job risk implies that ordinary least squares estimates of the wage equation may be biased and this should be corrected. accordingly,we use instrumental variables techniques. using original data from “la caisse nationale de la sécurité sociale”,we found evidence of wage differentials for hazardous work. in addition,organizing workers in union generates a value of statistical life (vsl) at least two times higher than for non-union (344,595.2 dinars for non-union and 985,459.7 dinars for union workers). however,these values are much lower than those estimated in developed countries. this study could provide useful recommendations for policymakers to reduce the risk of death in tunisia. © 2017,statistical economic and social research. all rights reserved.