Brazil submitted the first Reduced emissions from deforestation and forest degradationenREDD + reference level to the United Nations Framework Convention on Climate ChangeenUNFCCC -Implications regarding climate effectiveness and cost-efficiency

Brazil was the first country to submit a reference level (RL) to the United Nations Framework Conven-tion on Climate Change (UNFCCC) aiming to claim results-based payments for reduced emissions fromdeforestation in the REDD+ mechanism. This voluntary RL submission has passed an official technicalassessment process by the UNFCCC on the accordance of the data and method applied, which was finalisedin December 2014. After years of international negotiations this has become possible by the REDD+ framework finallyagreed. Although never formally adopted, the rules of procedure of the UNFCCC foresee consensus amongthe 196 parties to the convention. Due to the wide heterogeneity in national circumstances and capaci-ties between these parties, the decided REDD+ framework represents a compromise, and in some pointsspace for interpretation remains due to a lack of clear definitions. E.g., no internationally standardised method exists for the design of RLs under UNFCCC; rather each country has the possibility to apply itsown approach within a given range of requirements. We argue that these nationally designed RLs involve the danger of raising a ‘cherry-picking’ mentality, which would reduce cost-efficiency of the payments. Especially in the light of the ongoing pressure to fight anthropogenic global climate change, the overall objective of REDD+ should be climate effectiveness. Therefore, apparently contradictory objectives - of integrating as many developing countries as possible, by keeping the barriers low, and of effective emission reductions by persisting on strict rules for national carbon accounting - had to be balanced. According to the technical assessment by the UNFCCC, the submitted Brazilian RL fulfils all formal requirements. However, as implication of the negotiated framework, there remain some questions on the climate effectiveness and the cost-efficiency of the submitted approach, which this viewpoint paper aims to critically reflect. Brazil chose for its present RL design the eligible options of a subnational coverage, with consideration of the activity of deforestation only. But only comprehensive accounting of all activities on national scale can ensure that all potential emissions are accounted for. By the present accounting gap, doubts on climate effectiveness arise. Furthermore, Brazil chose a fixed starting year for historic emissions and did not consider national circumstances, which results in reference emissions of about 1100 Mt CO2/a between 2006 and 2010. This potentially profitable RL setting for Brazil raises doubt on the cost-efficiency of the implied REDD+ payment. Whether these problems arising from the four eligible but critical RL options applied by Brazil emerge inother RLs as well, was checked for further five RL proposals submitted up to now. None of these country submissions has passed the technical assessment yet. The submitted approaches for RL design differ among the considered countries but one can see a tendency towards profitable individual RL definitions, similar to the Brazilian RL. From our findings we conclude that individually determined RLs without standardised methods bear the risk of national ‘cherry-picking’ which can compromise climate effectiveness and cost-efficiency of the whole REDD+ mechanism.