In this paper we focus on the role of caseworkers in the assignment and take-up of welfare-to-work programs. We conduct a field experiment that generates exogenous variation in the assignment of caseworkers to different policy regimes. The experiment allows us to provide evidence on the effectiveness of welfare-to-work programs and to study how caseworkers exploit their discretion in assigning these programs to welfare recipients. We find substantial heterogeneity in how caseworkers assign welfare-to-work programs. Participation in the experiment and learning about the effectiveness of the different programs do not induce caseworkers to focus more on the effective programs. Obtaining knowledge about welfare-to-work programs is thus not enough to improve policy, also effort on implementation is required.