The Effect of Teacher Bonuses on Learning Outcomes and the Distribution of Teacher Skill: Evidence from Rural Schools in Peru

Teachers tend to avoid working in places with poor basic services, where transport costs are high and students show low performance. As a result, less advantaged students living in rural areas usually get paired with less qualified teachers. In many developing countries, teachers are offered monetary incentives to work in rural or remote schools. The literature, however, offers very little evidence about their effect on teacher qualifications. Moreover, this is the first study to produce causal evidence about the effect on these incentives on learning outcomes. This paper analyses the effect of unconditional monetary incentives on learning outcomes and the distribution of teacher skill in public rural schools in Peru. Teachers working in a rural school receive, on average, an additional 430 soles each month (around US$ 130 and approximately 30% of the starting salary). Schools are classified as rural based on the population of their community and their distance to the nearest province capital. We use a regression discontinuity design that exploits the exogenous shift in the amount of the bonus that occurs around the population and distance thresholds used to classify a school as rural. We find that the average bonus had a positive effect of around 0.16 standard deviations on reading comprehension and mathematics test scores obtained by second grade students in the 2014 and 2015 national evaluations. One of the mechanisms by which teacher bonuses can have a positive effect on learning is by making rural schools more attractive for talented teachers. We find evidence in favor of this channel. In fact, the bonus caused a shift of 0.38 standard deviations in the average score obtained by rural school teachers in the 2015 recruitment evaluation.