ABSTRACT. We use land use data from Guatemala to estimate the impact of a community forest concession policy on avoided deforestation when accounting for group heterogeneity. Our analysis includes communities with significant forestry experience, along with communities comprising recent settlers with primarily agricultural backgrounds. Employing a matched difference-in-differences approach, we compare deforestation in community concessions relative to a matched control area. We find the concession policy reduced deforestation throughout the study region, with the most significant effects in the recently settled concessions. While there is also evidence of leakage in this group, overall, there