Distributional Effects of Wage Leadership: Evidence from Sweden

This paper represents the first analysis of the consequences of a formal wage leadership, the Swedish Industry Agreement. We show that leadership in general has implied a lowered wage level for occupational groups having signed the agreement compared to groups that have not signed it. This is as expected as wage leadership should stabilize wage increases. However, the effects differ widely across occupations and skilled groups that signed the agreement have raised their wage level compared to otherwise similar workers outside the agreement. The agreement seems to have had a less binding effect on skilled workers. A possible explanation is that local wage formation is more common among the skilled groups. The agreement has increased the wage level among high educated compared to low educated and thus raised the education premium. Difference-in-differences models are applied using register data 1990-2005.