Union Threat Effects and Nonunion Industry Wage Differentials

We investigate the impact of union strength on changes in nonunion wages and employment. The prevailing model in this area is the threat model, which predicts that increases in union strength cause increases in nonunion wages and decreases in nonunion employment. In testing the threat model, we are also testing two alternatives, the crowding and complements models. In contrast to the prediction of the threat model, decreases in the percent organized (reflecting a declining union threat) are associated with increases in the nonunion wage. Furthermore, increases in union wages appear to decrease, rather than to increase, nonunion wages. Evidence on the determinants of intra-industry variation in nonunion wage premia is somewhat more consistent with the crowding model and is strikingly consistent with the complements model of union and nonunion wage determination. Further evidence on the determinants of intra-industry variation in nonunion employment is consistent with the complements model and the threat model; movements in nonunion industry employment are negatively related to changes in proxies for union strength. Thus, the combined evidence supports the complements model, but neither the threat model nor the crowding model.