Results-based payments for REDD + Lessons on finance , performance , and non-carbon benefits

• Results-based finance is a cornerstone in the approach to REDD+, as outlined in the Paris Climate Agreement. • Results-based finance will need to consider not only carbon/emissions-related payments but also incentives for intermediate outputs (such as policy performance) in order to effectively reduce deforestation and forest degradation. • A major gap in the current guidance for REDD+ finance is a lack of clear, context-relevant criteria and metrics to help justify and mobilize payments. • Negotiation and agreement on performance outputs and outcomes and their indicators are critical to ensuring national/local ownership and compliance. • Understanding the variation in costs and who is bearing the different costs of REDD+ will be critical in setting payment levels that can incentivize both carbon-effective and equity outcomes.

• Results-based finance will need to consider not only carbon/emissions-related payments but also incentives for intermediate outputs (such as policy performance) in order to effectively reduce deforestation and forest degradation.
• A major gap in the current guidance for REDD+ finance is a lack of clear, context-relevant criteria and metrics to help justify and mobilize payments.• Negotiation and agreement on performance outputs and outcomes and their indicators are critical to ensuring national/local ownership and compliance.• Understanding the variation in costs and who is bearing the different costs of REDD+ will be critical in setting payment levels that can incentivize both carbon-effective and equity outcomes.

A renewed focus on REDD+
Since 2005, an international policy framework has been in development to financially incentivize emissions reductions from deforestation and forest degradation and to conserve and enhance carbon sinks (REDD+).The framework was completed in 2015 and REDD+ became a central part of the Paris Agreement, 4 which emerged from the 21st Conference of Parties (COP 21) in December 2015.A key element of the Agreement (Article 5): Recognizes the importance of adequate and predictable financial resources, including for results-based payments, as appropriate, for the implementation of policy approaches and positive incentives for reducing emissions from deforestation and forest degradation, and the role of conservation, sustainable management of forests and enhancement of forest carbon stocks; as well as alternative policy approaches, such as joint mitigation and adaptation approaches for the integral and sustainable management of forests; while reaffirming the importance of non-carbon This brief does not claim to provide answers to all dimensions of the questions outlined above, the key objective here is to inform REDD+ decision makers and practitioners, as well as researchers, about critical issues and potential trade-offs when designing results-based finance instruments.We also highlight core areas where further guidance and clearer targets are required to manage these trade-offs.
1 How can adequate, predictable and effective REDD+ finance be achieved?
An important consideration for the GCF in the design of a results-based payment mechanism is related to the setting of payment levels and timing for release of payments.For the incentive to be truly effective for each recipient country, one has to understand the type, size and distribution of the costs of emissions reductions.

Variation in the opportunity costs of deforestation
On national policies, our research in Brazil has shown that command-and-control disincentives (forest law enforcement, protected areas, etc.) are the most cost-effective instrument for achieving the national REDD+ target of an 80% reduction in deforestation.However, an overemphasis on command and control transfers the burdens to private landowners who would incur over USD 1 billion annually in opportunity costs (Börner et al. 2015).This may not be politically feasible for Brazil over the long term when the costs are largely borne by a single group.Hence, integrating incentive policies such as payment for ecosystem services (PES) into the policy mix can make REDD+ strategies more equitable and, at least, partially compensate landowners for lost opportunities (Börner et al. 2014(Börner et al. , 2015)).
As such, the level of what would be considered as adequate REDD+ payments will vary with opportunity costs at different locations, and thus will also result in potential trade-offs between cost efficiency and equity/human welfare objectives.
Hence, a key question relating to trade-offs between effectiveness, cost efficiency and equity for a financing body such as GCS: how much variation can be considered when setting payment levels, whether across or within countries?Setting a fixed price for reduced emissions, as the Norwegian Government has done in their performance-based REDD+ contracts with Brazil, Indonesia and Guyana, will likely focus emission reduction efforts in regions within the country where opportunity costs are low.Arguably, in many cases, this will be where forests are least threatened and have little additionality, unless efforts are coupled with effective threat-targeting mechanisms.This also relates to how reference levels are set.
If set at a national level, similar across all locations, a payment mechanism will often go to areas not under threat, and thus have limited or no additionality.The problem of adverse selection bias is common in other voluntary results-based mechanisms such as PES or forest certification, as those with low or zero opportunity costs would likely have complied even without an incentive.

Understanding who is paying for the costs
In a survey of 22 subnational pilot REDD+ projects and programs in six countries (Sunderlin et al. 2014), we find that stakeholder institutions at the subnational level (as opposed to the project level) are highly involved in implementation of REDD+ and that there is a high level of subsidization particularly by subnational government institutions (Luttrell et al., in review).A total of 84% of stakeholder institutions at the subnational level are incurring higher costs than the benefit they receive from the REDD+ initiative.Many of these costs are not financial.
In places where there is unclear tenure, there are barriers to recognition and compensation of costs.In the survey of subnational REDD+ pilot projects and programs (Sunderlin et al. 2014), we find that only 9 out of 21 sites have clear legal tenure and these are associated with land uses that have the highest financial opportunity cost.More predictably, the opportunity costs of the informal sectors and small-scale actors engaged in customary (though sometimes legally ambiguous) land uses are often not considered and thus not compensated for.

Timing and sequencing of payments
One important aspect is understanding the time profile of costs.In contrast to results-based approaches, lessons from our policy performance research (described in next section), on-the-ground Fairtrade and certification projects (Tjajadi et al. 2015) and the experience of PES (Loft et al. 2014) suggest that some level of upfront payment to cover some of the costs is important to incentivize participation, particularly of the poor.However, whether at the national or project level, payments in advance transfer the risk of non-performance to the buyer or donor.Conversely, payments based exclusively on ex post results might lead to high rates of non-participation, particularly for lower-income countries, or poor subnational regions or socioeconomic groups.This may be an issue where the financing institutions needs to weigh effectiveness objectives with those of widespread and equitable participation.
A mix of upfront and non-results-based finance will be required in some cases and this can provide positive signals of financing certainty and predictability (Angelsen 2013), with ex post payments upon delivery being perhaps the more significant portion of the financing (Loft et al. 2014).
Having a payment mix will require clear performance criteria and credible reference levels (as discussed in next section), including clarity on such matters as definitions and assessments or measurements of performance (Brockhaus et al. 2016), as well as clear triggers for releasing the results-based payment.
2 What should qualify as 'performance' in REDD+, and how can it be measured?
REDD+ performance or results are quite different in the different phases of REDD+ implementation.In this section, we examine REDD+ performance in the implementation phase of policy approaches for REDD+ and subnational pilot REDD+ projects and programs.Paying for policy performance requires that performance is clearly defined and their metrics are appropriate and measurable .Experiences from subnational pilot projects provide insights for the implementation of payment for emissions reductions.
Payment for performance in the implementation of policy approaches for REDD+ REDD+ was originally conceived as a PES-like concept for carbon emission reduction between countries, to be financed through global and national carbon markets and funds.It has since evolved into having a broader mandate (i.e.including policy reforms), with funding primarily from development aid budgets (Angelsen 2013).Conditional aid to induce policy change is not a new concept.
However, conditionality of REDD+ policies and measures may call for some new approaches.Savedoff (2016) argues that a cash-on-delivery model 7 for REDD+ could increase cost efficiency, be more effective, and improve governance as recipient governments would realign behavior to achieve results -if the targeted outcomes are well-defined, agreed upon and adequately rewarded.Findings from a comparative policy study in 13 countries showed that the three countries with access to performance-based finance for REDD+ (combined with national ownership over the REDD+ process) tended to advance relatively more quickly in implementing policy reforms than the others (Brockhaus et al. 2016).However, the same study also found that overall progress with REDD+ was very slow, a tedious, sticky process.Hence, to realize intended behavioral changes within governance in the forest and land sector to reduce deforestation will require more than the promise of resultsbased finance.Research has also shown that established institutions with vested interests in deforestation that have benefited from existing institutional path dependencies will tend to resist change (Pierson 2000

Key issues when designing adequate, predictable and effective REDD+ finance:
• The devil is in the detail: understanding variation in costs and deforestation levels, and who is bearing the costs of REDD+, will be required for the design of a results-based payment that is effective, efficient and equitable.• Timing is crucial: In ex ante payments for policy performance, developing a reporting schedule that enables payments based both on the time profile of costs and performance will be critical.
as threatened, especially if the alternative development policies are not clearly mapped out; (iii) changes in crosssectoral coordination entails trial-and-error that can bring unexpected consequences and unintended side-effects, often affecting those who are already marginalized (Ostrom 1999).2014) who analyzed cooperation patterns within domestic REDD+ policy arenas in six countries and found that to induce behavioral change away from BAU, performance throughout the policy process needs to be rewarded.To identify appropriate performance measures, it is necessary to take into account (and incentivize) shifts in power structures in REDD+ policy domains over time.This requires the political process occurring during each of the three phases of REDD+ to be unpacked in terms of a detailed analysis of which issues are on the policy agenda at specific times, and of the influence, interests and relations of key policy actors (Brockhaus and DiGregorio 2014:25).
A challenge with results-based finance is the question of who will bear the responsibility for under-or nonperformance.Achieving reduced emission results is influenced by a complex and varied set of political, socioeconomic and environmental factors (e.g.El Niñoinduced forest fires) including institutional bureaucracies, enabling conditions and opportunity costs (Angelsen 2013;Brockhaus et al. 2016).If countries are to bear the costs of implementing policy reforms, they may be unwilling to participate in REDD+ because of these in part, noncontrollable, risks (Clist and Dercon 2014).However, large upfront investments (or ex ante payments) are a problem for the logic of results-based payment, which relies on the donor's credible leverage to reduce or withdraw payments when performance is not met.Assuming the full risk for non-delivery of reduced carbon emissions would miss out on the conceptual advantages of results-based payment.
Another challenge is in the setting of forest reference (emission) levels (FREL/FRL).The UNFCCC has invited countries to submit their national FREL/FRL, and, as of May 2016, 15 countries have done so8 .These are subject to a technical assessment by UNFCCC.If these are to be used as the basis for result-based payments, strong incentives exist for countries to bias these reference levels upwards by selecting favorable historical reference periods, scopes of activities, forest definitions, inclusion of degradation or not.Currently, no independent review mechanism is in place to address these issues at the national level.
Without a system to manage risks related to a country's non-performance and bias in reference levels, particularly when there is investment in reform, disagreements may lead to what we call the 'politics of numbers', where a stalemate might occur with none of the involved parties moving forward.

Payment for REDD+ results: lessons from pilot subnational initiatives
Since 2007, hundreds of REDD+ pilot projects and programs have emerged across the tropics.Research on these initiatives has highlighted that in the majority of cases, results-based payments to local landholders have barely gained traction (Sills et al. 2014;Simonet et al. 2015), in part due to unstable REDD+ financing.In CIFOR's analysis of 23 initiatives across six countries, only four have sold carbon credits, with the remainder dependent on inconsistent public and philanthropic funding (Sunderlin et al. 2015).
The source of funding, however, does not solely determine the application of conditionality in payments on the ground.For instance, in the Jari/Amapá REDD+ project in Brazil, where carbon credits have been sold through the voluntary market since 2013, no conditional payments are made to local landholders (Cromberg et al. 2014).Furthermore, public funds can be performance-based, such as the KfW REDD+ Early Mover's Program that supports Acre's State System of Incentives for Environmental Services.While the government of Acre has experimented with conditional payments to farmers engaging in more sustainable agricultural practices, this is only one of many strategies for promoting low emissions development across the state (Duchelle et al. 2014).Given the limited funding for REDD+, converting forest to other land uses is still more profitable than conserving it.
While conditionalities (particularly when accompanied by monitoring and penalties for noncompliance) can demonstrate vastly improved outcomes, they do come with costs.Their feasibility depends on the capacity of countries to set and monitor conditions.The experience of conditional cash transfer programs suggests that complex eligibility criteria to address additionality and equity outcomes will often result in a trade-off with cost efficiency (Wong 2014).
3 How can non-carbon benefits be integrated and safeguarded in a resultsbased payment mechanism?

Safeguard indicators and the need for counterfactuals
The UNFCCC COP16 in Cancun articulated seven safeguards for countries implementing REDD+ (Duchelle and Jagger 2014).These relate to governance, rights, participation, consent, environmental and social co-benefits, permanence and leakage.The intention is to ensure that REDD+ does not harm local people and the environment and to verify that new REDD+ institutions are complementary to existing environment and development policies.To be eligible for resultsbased compensation, REDD+ countries must develop national-level Safeguard Information Systems (SIS) to monitor and regularly report on the social and environmental impacts of REDD+ (Duchelle and Jagger 2014).Often, jurisdictions and projects engaged with multi-and bilateral donors and third-party certifiers may need to consider additional standards for demonstrating social and environmental performance, such as those of the World Bank Forest Carbon Partnership Fund, the UN-REDD Programme, the Climate Community and Biodiversity Alliance and the REDD+ Social and Environmental Standards Initiative.The multitude of safeguard requirements can mean that countries will choose the minimum safeguard requirements or have generic indicators in the SIS to lower the costs of monitoring and reporting to access results-based payments (Menton et al. 2014).
Given the complexity of monitoring the social and environmental outcomes of REDD+, a key challenge is developing simple, yet adequate, methods and performance indicators that are appropriate to the scale of analysis (Duchelle et al. 2015).Social performance encompasses respect for knowledge and the rights of indigenous people and local communities; full and effective participation of local stakeholders; and enhancement of other social benefits.Environmental performance focuses on promoting biodiversity conservation and ecosystem services provision.Furthermore, while the notion of the counterfactual is intrinsic to carbon monitoring through reference level setting, there is little use of counterfactual scenarios that could provide stronger understanding of the social or other environmental outcomes related to REDD+ (Caplow et al. 2011).To select and monitor social and environmental performance indicators, countries can draw on existing national socioeconomic monitoring programs and leverage both secondary and primary datasets.The use of mixed methods at multiple scales can help provide a more accurate understanding of the results-based performance of REDD+, which could be misinterpreted through the use of one dataset or method alone (Jagger et al. 2010).As SIS are consolidated, there is clear opportunity for learning among REDD+ countries on the use of appropriate indicators, data collection methods and reporting frameworks for monitoring and evaluating the non-carbon performance of REDD+.
Safeguards also have a role in managing risks to enable REDD+ results.The legitimacy of REDD+ benefit-sharing arrangements, for example, can be compromised by the lack of broad consultation with local actors, including customary authorities, and can result in local conflicts that affect deforestation behavior (Kowler et al. 2014;Myers et al. 2015).While participation is part of the social safeguards, the indicators will need to be grounded in the local contexts to be relevant.
Learning from other experiences for REDD+ safeguard monitoring and verification: harmonizing standards and managing civil society engagement An important part of the verification aspect of results-based payments is the grievance mechanisms within benefit sharing.How safeguards could act as effective reporting on grievance and conflict resolution will depend on how countries develop their criteria and indicators.The GCF and other financing instutions will need to consider its role in relation to the determination or verification of safeguards relative to disbursement of results-based payments, and the potential role of its own grievance mechanisms with regard to the different levels of dispute resolution and the legal capacity in different countries.
Key issues when assessing and measuring performance in REDD+ for results-based finance: • Incentivizing reform: There is still a long way to go to work out the details of result-based payments.Furthermore, results-based payments by themselves may be insufficient to drive needed policy and governance reforms for REDD+.The GCF could look to other approaches for resolution of issues.The Forest Carbon Partnership Facility (FCPF) Carbon Fund is one obvious example, however there are important differences between the FCPF and the GCF.For example, the FCPF purchases emissions reductions in the form of credits, some of which can be sold on a market, and accessing the carbon fund requires adherence to different safeguards, to those required by the UNFCCC.The GCF has adopted the International Finance Corporation's (IFC) Performance Standards (as opposed to safeguards) as an interim solution, which are different again.For example the IFC standard concerning free, prior and informed consent of indigenous peoples is to apply 'in certain circumstances', whereas the REDD+ Safeguard requires the full and effective participation of indigenous peoples and local communities.Further, the IFC interim standards provide no provision concerning permanence or leakage.
There are also lessons from the process of establishing Voluntary Partnership Agreements in the Forest Law Enforcement, Governance and Trade (FLEGT) mechanism, in particular relating to independent and formal civil society monitors and multi-stakeholder processes (Luttrell and Fripp 2015 ).By formalizing access to information and, hence, allowing for a broad review of performance, these processes can provide credibility by enhancing accountability and transparency.In this way, they increase commitment and confidence in the system.

Moving results-based financing forward
This brief has outlined several challenges to consider in the design of results-based financing for achieving effective, efficient and equitable outcomes.Existing research and experiences from related interventions, such as PES and FLEGT, provide valuable lessons.However, current guidance to respond to the wide variation and complexities between national contexts will require substantive fine-tuning of global indicators and criteria into adequate performance and results metrics that are grounded in realities, and the setting and reviewing of credible reference levels for setting and triggering payments.While this brief cannot provide such detailed guidance, we highlighted some core areas that require further action and outlined potential tradeoffs and ways to manage these.The parties involved, namely financing institutions such as the GCF, will have to find a balance between having global standards and cohesiveness and adequately addressing national-local complexity.It can do so by sending strong signals to its accredited institutions and partners to invest in understanding these complexities and building on datadriven solutions.
Key issues for integrating and safeguarding non-carbon benefits within a results-based payment mechanism: • Monitoring local realities: The establishment and reporting of SIS is one of the triggers for resultsbased payments: for safeguards to be effective in monitoring issues of social equity and environmental outcomes, use of mixed methods at multiple scales and indicators that are relevant to local realities will be needed.• Managing risks: Safeguards not only support more equitable outcomes, they can also be structured to manage risks (such as the nonlegitimacy of participatory processes) to delivery of carbon emission reduction results.• Learning from other experiences: Those that monitor and verify safeguards might be also exposed to risks, in particular, local civil society actors.Building on other experiences and indicators can help to identify and manage these risks.