How (not) to pay - Field experimental evidence on the design of REDD + payments

- We design a visual game depicting the decision situation of a cattle rancher in Brazil.
- We apply it in a randomized economic experiment to assess alternative PES designs.
- Payments conditional on forest stock change reduce deforestation most.
- Payments conditional on forest stock promote sustainable intensification.
- Relative cost-effectiveness of the two conditionality types depends on leakage.

Payments for Ecosystem Services (PES) can use many design features. We investigate the impact of payment conditionality types, i.e. different specifications when to pay and when not to pay, for PES on deforestation and agricultural production in a lab-in-the-field experiment. Our experiment also tests variations in contract period and payment volatility. We designed a highly visual simulation game to characterize the decision situation of a cattle rancher in Brazil. The player can expand extensive pasture by deforestation or intensify existing pasture. The model includes both a land and a capital constraint. We applied the game in an economic, framed lab-in-the-field experiment in Tocantins, Brazil. Payments conditional on forest carbon stock lead to slow, but steady deforestation, while payments conditional on forest carbon stock-change suppressed deforestation more strongly. But payments conditional on stock increase cattle production while payments conditional on stock-change have no effect on production. Thus, depending on the level of leakage, either type of conditionality can be more cost-effective in reducing global carbon emissions. Contracts with limited periods lead to strong deforestation after the end of the payment period. Payment volatility had no significant effect.