CGD Climate and Forest Paper Series #4

In March 2004, the Brazilian government initiated a range of policies and enforcement actions (under the Action Plan for Preservation and Control of Deforestation in the Legal Amazon) that brought sharp reductions in the rate of deforestation. In 2008, Brazil signed an agreement with Norway to receive payments during a 5-year period for bringing greenhouse gas emissions from deforestation below a 10-year average (1996–2005). Norway pledged up to US$1 billion for this agreement, which stipulated that these funds would be donated to the Amazon Fund (Fundo Amazônia), managed by the Brazilian National Development Bank and invested in actions to prevent deforestation and to promote the conservation and sustainable use of the Amazon biome.

In May 2014, we visited Brazil as part of a three-country study to gain insights into the value of future expansion of performance-based payments in other countries and other sectors. The following notes are based on discussions with government officials, NGO staff, private entrepreneurs, and independent researchers in Brazil about the policies and programs that are associated with reduced deforestation and forest degradation in Brazil, with particular attention to the influence of the Brazil-Norway Agreement and the Amazon Fund.[1] The views expressed here are ours alone.


In 2008, Brazil and Norway completed an agreement (hereafter “the Agreement”[2]) under which Norway would contribute to a Brazilian environmental fund if the annual rate of greenhouse gas emissions from deforestation were brought below the average rate of the 1996–2005 period—a “performance-based payment” agreement.[3] Norway pledged up to US$1 billion for this purpose. Under the Agreement, donations go to the Amazon Fund (Fundo Amazônia), which had been created for this purpose at the Banco Nacional de Desenvolvimento Econômico e Social (BNDES) to finance projects that contribute to reducing deforestation and promoting sustainable development in the Amazon. Up to 20 percent of the funds could also be applied in other biomes, whether in Brazil or other tropical countries.

The Agreement with Brazil was the first of three bilateral performance-based agreements established by Norway with forest-rich countries in the context of broader international negotiations and experimentation with Reducing Emissions from Deforestation and Forest Degradation (REDD+). In December 2013, Brazil and Norway expanded and extended the Agreement to December 2021. While Norway remains the largest donor to the Amazon Fund, Germany and Petrobrás have also contributed.

The remainder of this note presents our views of the successes, challenges, and lessons from this experience based primarily on our interviews with participants and observers in Brazil.


  1. International endorsement enhanced domestic legitimacy. The Norwegian offer of US$1 billion endorsed the Brazilian government’s ongoing and dramatic efforts to reduce deforestation in a way that improved the domestic legitimacy of these policies. The Agreement enhanced the standing of the Environment Ministry in relation to other ministries and raised the level of attention and organizational status of environmental concerns within BNDES. Earmarking the money for environmental projects gave legitimacy and resources to domestic constituencies who are promoting a conservation-oriented agenda in and out of government.

  2. One billion dollars is significant. Though small compared to the size of Brazil’s economy and its government budget, the US$1 billion amount was consistently described as a significant and healthy addition to funding in support of reduced deforestation. Although a pure cash-on-delivery aid model would have no restrictions on the use of payments, earmarking the funds for efforts to reduce deforestation in the Amazon made them a significant increment to this particular type of expenditure rather than a tiny contribution to the national treasury.

  3. Offsets were avoided, and sovereignty was respected. In international climate negotiations, the government of Brazil has opposed allowing rich countries to buy international offsets rather than reduce their own emissions and has opposed approaches that could be seen as compromising sovereignty. Norway’s offer to provide funds that (1) reward performance without generating offsets and (2) are managed by a domestic financial institution made it possible for Brazil to accept the Agreement because it respected both its negotiating position and its sovereignty. An important consequence of the Agreement was to demonstrate an approach to performance-based international cooperation on forests that is compatible with these two principles and that allowed Brazil to support REDD+ in international fora.

  4. Institutional choice promoted domestic accountability. The Brazilian decision to house the Amazon Fund in BNDES, guided by an independent steering committee, is accepted by many people as having protected the Norwegian funds from being politicized or wasted. BNDES was criticized for starting slowly and for supporting projects that were poorly aligned with the Amazon Fund’s intent, but this criticism has abated as the pace of commitment and disbursement increased along the directions laid out in the Agreement. One informant noted that the Amazon Fund has committed and disbursed funds much more quickly than a previous multilateral program (Pilot Program to Conserve the Brazilian Rain Forest).[4] About 75 percent of the original Norwegian pledge has now been transferred to BNDES, which has approved specific projects equivalent to about 50 percent of the original pledge.[5] It is analyzing additional projects equivalent to another 25 percent of the original pledge.[6]

  5. The funder is a helpful outsider. Norwegian engagement through the Agreement is seen positively by BNDES and NGOs. BNDES staff see Norway as a source of technical and political support and consult regularly with Norwegian foreign officers and environmental staff. NGOs see Norway as a useful channel of communication to the Brazilian government and as an honest broker. The Brazilian government does not see itself as an aid recipient, so the framing of the partnership as cooperation is important.

  6. The role of trust was enhanced. The Agreement clearly demonstrates a high degree of trust and partnership between Brazil and Norway. Typically, bilateral funding commitments are made under strict conditions to give the funder assurances that funds will not be misused or applied in ways that violate social and environmental norms. The Agreement relies instead on BNDES and an independent steering committee to address the relevant fiduciary, social, and environmental risks without direct supervision or involvement of Norway in the Amazon Fund’s governance.


  1. Performance outpaced funds available for payments. Brazil’s deforestation rate started to decline rapidly in 2005 and continued to decline in subsequent years because the federal government began actively enforcing rules prohibiting illegal encroachment in public lands (about two-thirds of the Amazon forest is publicly owned), established new protected areas, and undertook other related measures. The declines in 2006 and subsequent years were so great compared to the average of the prior 10 years that the external funds promised by Norway were far from sufficient to cover the total reduction in greenhouse gas emissions (when priced at US$5 a ton as mentioned in the Amazon Fund project document). If the Agreement had based transfers from Norway to the fund on emission reductions alone, the pledged funds would have been transferred much more quickly. Instead, the agreement linked transfers from Norway to the Amazon Fund’s pace of commitments and disbursements (or financial need). Consequently, even though eligibility for Norwegian transfers is achieved when there is progress on reducing emissions from deforestation, the timing and actual amounts transferred appear less like performance-based payments and more like payments for committing and disbursing funds to Amazon Fund projects.

  2. The Agreement is transparent but poorly understood. The Agreement between Brazil and Norway is public, and the amounts transferred are also easily found on the web. Once these funds are transferred to the Amazon Fund, BNDES appears to be publishing the use of this money in a well-documented and timely fashion. Nevertheless, most informants in Brazil had an imperfect understanding of the terms of the Agreement, including (1) what exactly determines Norway’s transfers under the Agreement, (2) what types of activities and grantees are eligible for support, and (3) how decisions are made on individual proposals. BNDES maintains that this latter information is readily available from public documents or by making queries.

  3. There has been a slow start. Despite large declines in deforestation, Norwegian transfers were affected by the pace with which the Amazon Fund could develop a pipeline of projects. It took time for BNDES to develop its system for soliciting and evaluating project proposals in ways that fulfilled the Amazon Fund’s mission while meeting BNDES’s fiduciary and other standards. These issues have now been addressed so that the mechanism is functioning well in terms of committing and disbursing funds. Some informants argued that the fund’s procedures are too difficult for local and indigenous communities to access. BNDES maintains that it is reaching indigenous groups by working with various partners, as it has done in projects with the State of Amazonas and Fundo Brasileiro para a Biodiversidade.

  4. Is this the right strategy? People we spoke with had strong opinions about whether the Amazon Fund is supporting the right kinds of actions. Some view the fund’s initial approach as too focused on projects and not sufficiently strategic; they would prefer to see the fund aim for larger, more systemic impact by engaging in policy-level initiatives. An early example of such interventions is support for large-scale implementation of cadastral registration of private lands under the revised Forest Code, which will in turn support a new market that promotes efficient land use. One informant proposed using the fund to purchase verified emission reduction credits from REDD+ projects, echoing other discussions of the possibility of programming funds on a payment-for-performance basis to subnational jurisdictions or project-level initiatives (e.g., The Zero Deforestation Pact).[7] Another expressed concern about the risks of the fund’s resources being used inappropriately to subsidize inadequate government budgets.

  5. Will progress continue and be sustained? The Agreement and Norwegian funds are not prominently connected in people’s minds with the main domestic political debates about whether and how to maintain low deforestation rates and create incentives for forest protection. Many persons interviewed expressed concern that maintaining progress on reducing deforestation will become more difficult over time. We heard contradictory perspectives about whether the domestic strategy on enforcement is being pursued with sufficient vigor; some informants argued that the command-and-control approach has reached its limits and that the focus should be on a positive agenda of finding and supporting a strategy of forest-friendly sustainable growth. Some informants worried that Brazil is not progressing toward such a sustainable growth model for rural communities (i.e., income generation compatible with low or zero deforestation) for lack of a national strategy and because the federal government has ambitious proposals for new roads and other infrastructure in the Amazon region. Some see a particular disconnect—and the possibility of a future collision—between BNDES’s role in administering the relatively limited Amazon Fund and its role as a lender of significantly larger amounts in support of large traditional infrastructure development in sensitive areas. Ministry of Environment officials maintained that they do have a national strategy and that current policies are establishing innovative linkages with the private sector that permit economic growth and environmental protection to proceed together.

Lessons for Performance Agreements

  1. Domestic ownership drives change, but performance payments can help. Performance-based payment arrangements work best when recipient governments are committed to the goal and have the support and the healthy pressure of active civil society groups. In such cases, the funding is not the core driver of successful change. Rather it serves to facilitate and amplify changes by providing legitimacy, drawing attention, and strengthening domestic constituencies, as has been the case in Brazil. International offers to pay for reductions in deforestation can help by making performance visible in terms of the outcome and in terms of funding. However, it will never guarantee or force domestic action.

  2. Domestic ownership is stronger when funders are hands off. By limiting Norwegian involvement in how funds are used and tracked, the Agreement enabled Brazil to assume full responsibility for managing the money while giving it discretion about how funds would be applied. Norwegian officials are still consulted, but they have not required a formal role in the governance of the Amazon Fund (in contrast to other funders who have insisted on having representation on governing boards of such programs). As a result, BNDES in particular and the Brazilian government more generally have taken strong interest and pride in creating and implementing the Amazon Fund.

  3. Domestic ownership is stronger when funders are modest. The government of Norway has certainly promoted its contributions to tropical forest conservation efforts internationally and to domestic audiences in Norway, but it has not prominently advertised its engagement within Brazil. Therefore, the Brazilian public’s awareness of Norwegian involvement may be limited. This seems to be a good strategy for lowering domestic resistance from nationalist constituencies while still providing support to groups in favor of forest conservation.

  4. Use existing institutions when possible. Both domestic and international legitimacy of the Agreement were enhanced by channeling the Agreement’s funds through a domestic institution (BNDES). In combination with the creation of an independent steering committee, BNDES manages the Amazon Fund with credible mechanisms for addressing fiduciary, environmental, and social risks. Another advantage of using BNDES is that it agreed to charge very low overhead for the Amazon Fund (capped at 3 percent of contributions according to the Agreement). Nevertheless, care must be taken to ensure that an existing institution’s mission and procedures do not constrain innovative or systemic uses of funds.

  5. Look for ways to pass performance-based incentives downstream. In the context of REDD+, it is important to explore how to translate at least some portion of international performance-based payments into domestic results-based incentives at the level of subnational jurisdictions, projects, and/or individual landowners. The legitimate financing needs for readiness activities should not obscure the risk of lapsing into business-as-usual funding of inputs not clearly linked to achievement of the ultimate objective.

CGD is grateful for contributions from the Norwegian Agency for Development Cooperation and the UK Department for International Development in support of this work.

[1] We are grateful to the people who took time to discuss Brazil’s deforestation policies and the Amazon Fund with us, including the ambassador of Norway; officials from the Ministry of External Relations, Ministry of the Environment, and Secretary of Strategic Affairs of the Presidency; and staff from Banco Nacional de Desenvolvimento Econômico e Social (BNDES), the World Bank, the Inter-American Development Bank, the British Embassy, Germany’s KfW development bank, The Nature Conservancy, Instituto de Pesquisa Ambiental da Amazônia, and Biofílica.

[3] Details regarding how the Agreement would be implemented are contained in the “Amazon Fund Project Document,” available at­ export/sites/default/site_en/Galerias/Arquivos ­/Boletins/Amazon_Fund_-_Project_Document_Vs_18-11-2008.pdf.

[4] The Pilot Program to Conserve the Brazilian Rain Forest is a World Bank–administered program of US$428 million that started in 1992. While it is credited with many successes, it took more than a decade to complete.

[5] Based on information at, accessed on May 27, 2014.

[6] BNDES moves projects beyond the screening phase into analysis only when it knows it has funds to support them. BNDES argues that this procedure precludes spending time on applications for projects unless funding is available—a savings to both applicants and BNDES.

[7] “Pacto Pela Valorização da Floresta e Eliminação dos Desmatamentos na Amazônia—‘Pacto Desmatamento Zero’” (2008) can be found at