A methodology to estimate national REDD+ reference levels using the Zero-Sum-Gains DEA approach

• The ZSG-DEA approach is proposed to estimate the REDD+ reference levels.
• The use of ZSG-DEA model is more beneficial for Latin American and the Caribbean.
• Countries in Africa, Asia and Oceania may see a revenue drop in REDD+ programs.
• The principle of fairness turns out to be more beneficial to Africa, Asia and Oceania.
• The principle of efficiency is more beneficial for Latin America and the Caribbean.

REDD+ reference levels directly impact the benefits which a country may receive. However, the existing “Compensation Reduction” (CR) and “Compensated Successful Efforts” (CSE) are only considered from a unilateral perspective of outputs or inputs. The combination of these two approaches is considered to estimate the REDD+ reference levels through the Zero-Sum-Gains Data Envelopment Analysis in this paper. The agricultural labor force and agricultural land area are used as input variables, and the gross agricultural production and carbon emissions from deforestation are considered as output variables. The REDD+ reference levels of 89 countries are calculated and classified through the Zero-Sum-Gains DEA model. The results demonstrate that the REDD+ reference levels are estimated efficiently through the Zero-Sum-Gains DEA model, and all countries with deforestation are in the Zero-Sum-Gains DEA frontier, indicating the overall Pareto optimality has been achieved. The empirical results also indicate that the use of Zero-Sum-Gains DEA model is more beneficial for Latin American and the Caribbean, while the countries that may see a revenue drop in REDD+ are in Africa, Asia and Oceania. Consequently, the final REDD+ reference levels should take into account both efficiency and fairness by selecting the appropriate fairness-efficiency weighting factor.