The strong wage effects related to mismatches between a worker’s education and that required in the job are usually attributed to assignment theory. This theory asserts that productivity and wages depend on the education-job match, which determines the utilization of skills. However, recent research shows that educational mismatches are only weakly related to skill utilization, which in any case fails to account for the bulk of the wage effects. Two alternative theories have been put forward to explain the observed wage effects. One points to wage setting institutions that cause wages to be based on job characteristics regardless of individual performance, the other to the heterogeneity of skills within a given educational level. Both theories explain existing results, but have never been tested directly. In this paper we show that the former theory explains observed wage effects in the public sector, and the latter theory those in the private sector.