We study separately four educational groups. Net migration and commuting are scaled by relevant population variables. For both type of response variables, we include three incentive variables on the log scale corresponding to a relative real hourly wage variable, a relative difference in unemployment rates and a relative difference in employment rates. The equations are estimated with OLS using dummy variables for the different economic regions. The main attention is on the effects of a marginal change in the real wage variable. For individuals with low and vocational education we find that the relative real wage variable enters significantly, whereas mainly no significant effects are found for the two groups with high education.