The impact evaluation of short repeated programs: The case of parenting skills programs

In this article, we estimate a series of models to evaluate the effects of a short program in a context where randomization is not feasible but the program follows a cyclical pattern. We will focus empirically on a case study involving parenting skills courses, which have become increasingly popular. In order to evaluate the impact of the program on parents' and children outcomes, we employ two different methods. The first method compares the outcomes of families who have just finished the program with those who are about to start it; the second compares the outcomes of the same families over time. Furthermore, we propose a model to test whether families who enrolled early were systematically different from those who enrolled later. We find beneficial effects of the program on the importance of living in an area that offers opportunities and of having good quality relationships with friends and family; on the level of self-confidence in sharing one's experiences with other parents; and, in general, on the opinion that tablets and cell phones can be useful for learning, can give parents the opportunity to do something and can calm children. Moreover, we show that the families who access the course early are not random in every respect: they consistently assign higher importance to being well-integrated into a community and having access to culture for their well-being, and eventually, they utilize their time with children differently.