The effectiveness of remedial courses in Italy: a fuzzy regression discontinuity design

Skills are important for many social and economic outcomes. Whereas interventions during childhood are considered crucial to increase the skills of disadvantaged individuals, in this paper, we try to understand whether interventions which take place later in life can also be effective. With this aim, we evaluate the effects on student achievement of a number of remedial courses provided by an Italian university. We use a fuzzy regression discontinuity design to identify the causal effect of remedial courses relying on the fact that students were assigned to the treatment if their performance in a placement test was below a certain cutoff point. We deal with partial compliance by using the assignment rule as an instrumental variable for the effective attendance to remedial courses. From our analysis, it emerges that students attending the remedial courses whose results were just below the cutoff point acquire a higher number of credits than students whose results were just above the cutoff. We also find that remedial courses reduce the probability of dropping out from an academic career.