Evidence on the effectiveness of participatory development approaches in low-income countries is mixed. We use an experiment to study elite capture in a large sample of Ethiopian forest user groups, varying governance modalities and exploring implications for livelihoods of group members. External monitoring by forestry officials increases average consumption and income levels, and decreases within-group inequality. In contrast, bottom-up monitoring by group members or the provision of incentives to group leaders for inclusive management have no impact on livelihoods for the average group member. Additional analysis suggests that bottom-up monitoring fails because many group members do not care enough about flows of forest benefits to invest time in monitoring. In the sub-sample of groups where forest benefits are an important component of rural livelihoods, internal monitoring had a positive effect on economic outcomes.