In case of regional discretionary on the implementation of policy measures, central governments may consider the differences in outcomes as an indication on the effectiveness of policy. In turn, to motivate regional policy makers uniform incentives might be developed. However, if the underlying population differs, uniform incentives may have an discouraging effect. This paper considers the differences in outcomes in early school leaving between the two largest Dutch cities. Using an exceptionally rich data set of all students in Amsterdam and Rotterdam, we test by probit and matching models whether uniform incentives are appropriate.