A Note on the "Union Effect" in VSL Studies

PHILIP E. GRAVES, University of Colorado at Boulder

Abstract

Viscusi and Aldy (2003) observe that “most studies of the U.S. labor market find that union affiliation is positively correlated with a greater wage-risk tradeoff while international evidence is much more mixed.” They provide several arguments as to why the risk premium might be higher for union members (marginal versus average worker preference, the quasi-public good nature of workplace safety, and better safety information for the unionized). An alternative explanation–concentration of union membership in undesirable locations–can account for both the apparent higher risk premium in union jobs in the United States and the failure to find that gap in the international setting. Moreover, the explanation advanced here can account for the anomalous finding in several papers that non-union workers appear to have negative compensating differentials for risk.