Association for Public Policy Analysis & Management

One of the enduring challenges of decentralized approaches to managing natural resources is that such approaches may enable exclusion and inequality by ethnicity, class, and gender. Despite this problem, the natural resource decentralization is an increasingly popular conservation tool, as evidenced by the fact that rural communities now collectively manage a substantial and growing share of the developing world's forests. I use quasi-experimental methods and a nationally-representative household survey from Nepal to examine participation in the country’s community forestry program, under which millions of households participate in more than 20,000 user groups that govern local forests. Comparing households in villages that received the reform to households in a counterfactual group of similar villages that had not yet received it, I find that households from marginalized ethnic groups were substantially more likely to start participating in the program after implementation compared to households from elite ethnic groups. I argue that the design of the reform explains why the Nepalese case is so exceptional; By mandating that local user groups implement redistributive, pro-minority policies, the central government and external NGOs created incentives that offset the opportunity costs of participation for minority households. I use machine learning methods to show that economic and gender-based disparities in participation remain despite successful efforts at ethnic inclusion. Contrary to many other cases, the Nepalese case shows that marginalized groups participate actively under decentralization when institutions provide tangible, targeted incentives. However, these incentives may not be enough to alleviate disparities by gender and class.