The paper evaluates the treatment effects of the wage subsidy program in Hamburg with propensity score matching as well as linear regressions. It is found that the estimated treatment effects differ only slightly between both methods. When analysing heterogeneous treatment effects, however, linear regressions are easier to implement and the results are easier to interpret than with propensity score matching. All in all, the program has significant positive effects, which are especially large for problem groups. Therefore, the Hamburg wage subsidy program has not only overall positive effects but also reaches the target group.