LAUSR

Abstract Environmental effects of development programs are the subject of an ongoing debate. We contribute to this debate by studying effects of a randomly allocated, nationwide development program in The Gambia on deforestation, focusing on parts of the country with meaningful baseline forest cover. Our main finding is that the program caused significant increases in annual forest loss. Conservative benchmark estimates imply that 5.6% of all forest loss occurring within 1 km of treatment villages during the eight post-program years resulted from the program. Accounting for spillovers, we estimate that the program explains about one quarter of the forest loss around all villages. Looking at possible channels, we find moderate treatment effects of the development program on household wealth and livestock holdings. Further, villages with limited access to markets drive the effect of the program on deforestation.