Payments for Ecosystem Services (PES) are a conservation policy in which landowners receive financial compensation conditional on actual environmental service delivery. PES has proven to be effective in inducing conservation of privately owned forests, but it is an open question whether this policy will also be effective in (de facto) communal forests. If land is managed by communities, offering payments to the community in return for conservation may give rise to strong free-riding incentives. We implement a Randomized Controlled Trial in arid Burkina Faso, aimed at inducing reforestation in community-managed forest areas, to test whether threshold group payments give rise to improved conservation outcomes compared to the standard PES policy scheme, in which (group) payments increase linearly with conservation outcomes. Threshold payments change the nature of the game from a social dilemma into a coordination game, and hence the former are predicted to be more conducive to public goods provision than the latter. Contrary to this prediction we find that linear payments actually give rise to higher survival rates in our reforestation sites than threshold group payments, and we use both survey evidence as well as lab-experimental results to uncover the underlying mechanism giving rise to this surprising result.