Spatial Externalities and Social Multipliers of Schooling Interventions∗

We study the effects of social interactions on school attendance decisions within a large-scale social program in rural Mexico. In order to identify pro-gram externalities across localities, we use exogenous variations in the number of beneficiary localities in the proximity of each village induced by the exper-imental evaluation design. We find evidence of large and positive spillover ef-fects on secondary school enrollment decisions amongst children in the treated group, but no evidence of such effects for children in the control group. This marked heterogeneity sheds light on the underlying mechanism through which social interactions operate in our setting. Notably, individuals do not seem to influence each others through schooling behaviors. It is instead the inter-vention which is likely to have triggered interactions amongst the targeted population, thereby enhancing individuals ’ perceptions of the benefits of edu-cation. Our preferred estimate implies a substantial social multiplier effect of the program: a one standard deviation increase in the number of beneficiary localities in the surroundings of each village increases enrollment rates by 4.1 percentage points