Replication data for: “But Everybody’s Doing It!” The Power of Peer Effects in High Peer Density Areas

Mexico’s conditional cash transfer program (PROGRESA) incentivized school en- rollment for poor families in selected villages beginning in 1997. Poor children who enrolled in school as a result of the incentive influenced their wealthier peers to also enroll. We use a two-stage least squares instrumental variables model to identify these peer effects in PROGRESA eligible villages. This paper adds to the literature on peer effects as well as previous evaluations of PROGRESA by demonstrating that peer ef- fects have the greatest impact when at least 67% of families in treatment villages are receiving the incentive. This result has the policy implication that peer effects may only operate if the density of the peer group exceeds a threshold of approximately 60%.