Eisenbarth, S., Graham, L, Rigterink A. S. (2021) Proceedings of the National Academy of Sciences, Volume 118(29), e2015172118

Significance: To halt deforestation, communities are increasingly being given the authority to manage their own forests. Although standard economic theory predicts that community management leads to overexploitation, field studies have reported that communities can sustainably manage their forests if specific conditions are present. One condition that is correlated with successful common pool forest management is community-led monitoring of the forest. However, whether such monitoring causes improvements in forest conditions is unclear. Using a randomized controlled trial, we provide causal evidence about the impact of community-led monitoring on forest use. Unlike prior studies, we estimate the effects of monitoring on both monitored and unmonitored forests. The results suggest that monitoring may simply displace forest loss to unmonitored forests, rather than reduce it.

Abstract: Rapid deforestation is a major driver of greenhouse-gas emissions (1). One proposed policy tool to halt deforestation is community forest management. Even though communities manage an increasing proportion of the world’s forests, we lack good evidence of successful approaches to community forest management. Prior studies suggest that successful approaches require a number of “design conditions” to be met. However, causal evidence on the effectiveness of individual design conditions is scarce. This study isolates one design condition, community-led monitoring of the forest, and provides causal evidence on its potential to reduce forest use. The study employs a randomized controlled trial to investigate the impact of community monitoring on forest use in 110 villages in Uganda. We explore the impact of community monitoring in both monitored and unmonitored areas of the forest, using exceptionally detailed data from on-the-ground measurements and satellite imagery. Estimates indicate that community monitoring does not affect our main outcome of interest, a forest-use index. However, treatment villages see a relative increase in forest loss outside of monitored forest areas compared to control villages. This increase is seen both in nonmonitored areas adjacent to treatment villages and in nonmonitored areas adjacent to neighboring villages not included in the study. We tentatively conclude that at least part of the increase in forest loss in nonmonitored areas is due to displacement of forest use by members of treatment villages due to fear of sanctions. Interventions to reduce deforestation should take this potentially substantial effect into consideration.

View Paper