Payment for ecosystem services programs in the Brazilian Atlantic Forest: Effective but not enough

[Display omitted] •The first PES impact evaluation in South America with a counterfactual approach, separately evaluating forest regeneration and deforestation.•PES programs in the Atlantic Forest incentivized forest regeneration, but its overall conservation impact was low.•Achieving robust forest restoration gains with PES programs in the Atlantic Forest would require approximately 180 years. Payment for ecosystem services (PES) are economic incentives intended to generate conservation benefits, principally on private properties. The global portfolio of PES programs is estimated to represent an annual investment of more than USD $36 billion. Despite this substantial investment, the continued lack of systematic and rigorous impact evaluation of PES has contributed to uncertainty regarding the effectiveness of this tool for conservation. Here, we assessed the ability of two watershed PES programs to promote native forest conservation in the Brazilian Atlantic Forest. Those two focal programs are among the most well-established PES programs in Brazil, and form part of a larger network of PES programs intended to protect the drinking water supply of over 19 million people in the São Paulo metropolitan region. Using a counterfactual approach, we examined if enrollment in a PES program contributed to the conservation of on-farm native forest cover. With propensity score matching, we identified a set of neighboring, non-enrolled ‘control’ properties, with similar size, altitude, soil type, demographic density, presence of water sources and forest cover. We then estimated forest cover on enrolled and control properties before and after PES implementation with a differences-in-differences method, modeled as the probability of an observed change in on-farm forest area as a consequence of PES enrollment. We found that PES has a positive effect on forest cover, with PES enrollment over a five-year period associated with an additional 2.8–5.6% of farm area coverage in native Atlantic Forest, through forest regeneration. PES enrollment was associated with a non-significant trend toward decreased loss of vegetation. We discuss the implications of these results for understanding the contribution of PES to additionality in forest conservation. While positive, the relatively slow impact of PES on forest regeneration suggests that environmental managers should not count exclusively on PES mechanism to achieve conservation goals.