Does Short-Term School Tutoring have Medium-Term Effects? Experimental Evidence from Chile

This paper explores how short-term tutoring affects educational outcomes in the short- and medium-term. We implemented a randomized experiment of a three-month small group tutoring program in Chile that aimed at improving reading outcomes among fourth graders using college student volunteers. We find small short-term effects on reading outcomes. Using administrative data covering up to eight years after the program ended, we find significant decreases in the probability of dropping out, increases in the probability of timely school progression, and increases in attendance, school grades and test scores. These effects are stronger for students who were ex-ante more likely to drop-out from school. The program effects are stronger for students who established stronger personal connections with the tutors. Our results suggest that tutoring programs may have relevant medium-term effects that go beyond short-term impacts on specific subjects with stronger effects on more at-risk children.