Multiple Wins, Multiple Organizations—How to Manage Institutional Interaction in Financing Forest Landscape Restoration (FLR)

: By restoring forest ecosystems and fostering resilient and sustainable land use practices, Forest Landscape Restoration (FLR) contributes to climate change mitigation, adaptation and sustainable development as well as the protection of biological diversity and combating desertiﬁcation. This integrative approach provides the opportunity for multiple wins, but it necessitates the management of complex institutional interactions arising from the involvement of multiple international organizations. Focusing on the pivotal aspect of ﬁnancing, this article surveys the landscape of public international institutions supporting FLR and analyzes the effectiveness of existing mechanisms of inter-institutional coordination and harmonization. Methodologically, our research is based on a document analysis, complemented by participant observation of the two Bonn Climate Change Conferences in May and November 2017 as well as the Global Landscapes Forum in December 2017. We ﬁnd that ﬁnancial institutions have established fairly effective rules for the management of positive and negative externalities through the introduction of co-beneﬁts and safeguards. The fact that each institution has their own safeguards provisions, however, leads to signiﬁcant transaction costs for recipient countries. In the discussion, we thus recommend that institutions should refrain from an unnecessary duplication of standards and focus on best practice.


Introduction
Forest Landscape Restoration (FLR) is an action-oriented set of ideas about how to best achieve several forest-related international policy goals in a multiple-win approach. These goals include Sustainable Development Goal (SDG) 15 on the sustainable use of terrestrial ecosystems [1] (p. 24), the Convention on Biological Diversity (CBD) Aichi Target 15 [2] to restore at least 15 percent of degraded ecosystems, the United Nations Framework Convention on Climate Change (UNFCCC) objectives regarding Reducing Emissions from Deforestation and Forest Degradation in Developing Countries (REDD+) [3] and the United Nations Convention to Combat Desertification (UNCCD) Land Degradation Neutrality (LDN) Project [4]. The fundamental aim of the FLR approach is to shift attention "away from simply maximizing tree cover to truly considering forest functions in the overall configuration of landscapes people depend on" [5]. As such, FLR has become a collective action frame for a wide variety of international actors to scale up forest restoration projects in an integrated landscape approach, based on national priorities and voluntary commitment. With FLR, stakeholders' perceptions and activities can be directed towards "the process of regaining ecological functionality and enhancing human well-being across deforested or degraded forest landscapes", considering The result of our analysis is that all institutions have developed fairly efficient measures to manage positive and negative externalities of their activities for other policy fields related to FLR. While there are few variances between the different safeguard systems, each institution however cultivates its own system. This not only binds a lot of energy in the institutions themselves, but also creates high transaction costs for recipient countries, which often deal with several institutions and consequently requirements and procedures in parallel.

Theoretical Framework, Materials and Methods
To analytically grasp the growing institutional density [45] or complexity [46] of international relations, scholars have worked out concepts like regime complexes [23,24], regime complexity [25] or institutional fragmentation [26,27]. All of them seek to better understand the impact of increasing institutional interlinkages "in the absence of hierarchical coordination on the strategic behavior of state actors and on the governance of specific issue areas" [37] (p. 3), (see also [47,48]).
Functional overlap between institutions can be deliberately created by states which are dissatisfied with the regulatory work of a specific organization and want to shift rule-making to another existing organization or even create a new organization [49] (p. 297). Unintentional functional overlap occurs if an organization is created for functional reasons in an existing issue area already covered by other organizations [49] (p. 298). With regard to climate finance, the establishment of the Green Climate Fund (GCF) could be an example as the funding to be managed by the GCF has so far predominantly been covered by the World Bank. Finally, functional overlap can also occur independently from member states' intentions if an international organization pursues a mission creep, meaning that the organization itself expands its regulatory competencies [49] (p. 298), [50].
Complexes of international institutions with functional overlap present states with strategic opportunities for forum-shopping [24,51] and regime-shifting [52]. Some scholars [53][54][55][56] argue that functional overlap and institutional interplay are "potentially problematic" [56] (p. 29) and harm the effectiveness of global governance because states strategically use institutional fragmentation to pursue their parochial interests. Consequently, rules-based cooperation is undermined by the evolution of contradictory rules and the performance of international organizations is restricted. This position is opposed by Gehring and Faude [57] who develop a functional argumentation based on differentiation theory and institutional ecology. They attribute more positive features to the functional overlap among regulative international institutions and consider institutional fragmentation as a "rational response to the increasing complexity of society" [47] (p. 120). Gehring and Faude argue that states as members of multiple overlapping institutions have a general interest in coherent rules and institutional complementarity and take this into account "when determining their behavior within either of these institutions" [57] (p. 471). Especially when the distribution of power between states is balanced, they expect the emergence of sophisticated patterns of co-governance [57] (p. 481), (see also: [58] (pp. 374-375)). "In the absence of inter-institutional negotiations and a written contract" [57] (p. 482), Gehring and Faude hold that a division of labor will spontaneously evolve in a process of mutual adaptation. Faude further argues that international organizations themselves have an interest in resolving functional overlap because this exposes them to inter-organizational competition [49] (p. 300), (see also: [25,59,60]). He expects organizations to react by specializing in specific issues, functions or policy instruments, but also considers the assimilation of rules among institutions as an alternative solution [49] (pp. 301-303).
Holzscheiter [61] (p. 323) adds a constructivist perspective to the debate. She argues that inter-organizational relations and cooperative structures are a result of influential and contested global norms. The norms which she reconstructs-similar to the functional logic of Faude and Gehring-prescribe harmonization of rules and inter-organizational coordination, including division of labor, as a precondition for effective global governance [61] (pp. 335-336). In contrast to Faude and Gehring, who explicitly delimit their argument to regulative international institutions [57] (pp. 493-494), Holzscheiter focuses on organizations in the global health sector which have more operative functions [61] (p. 335). In this context, functional overlap is problematic not only because it gives countries opportunities for forum shopping and might lead to contradictory rules, but also because a lack of coordination can easily lead to a duplication of structures and efforts, which results in high transaction costs [61] (p. 339), [62][63][64]. Holzscheiter's constructivist perspective allows for an analysis of unintended or irrational effects of inter-organizational cooperation. She argues that institutionalization of such a cooperation often leads to hypercollective action [65] which neither increases efficiency nor reduces institutional structures. Instead, the proliferation of cooperative structures can increase transaction costs [61] (pp. 327-328).
FLR as a collective action frame and related institutional initiatives such as the Global Partnership on Forest and Landscape Restoration (GPFLR) and the Bonn Challenge are premium examples of actors' awareness of a functional overlap between the CBD, UNFCCC and UNCCD. They seek to streamline related restoration goals in an integrated way and to establish an institutional arrangement that creates win-win-win solutions in terms of the protection of biodiversity and the global climate while at the same time promoting and supporting sustainable livelihoods [15]. The funding of FLR activities is central to the realization of the conventions' goals. However, as the financing for REDD+ has already shown [37], both functionally overlapping rule-making and diverse allocation principles as well as competition for funds make the implementation burdensome for those seeking to realize and manage projects on the ground.
In the following section, we will map the diverse funding mechanisms in place and analyze their safeguards and co-benefits rules. The aim is to analyze whether there is-or is the potential for-inter-organizational coordination to streamline FLR into international forest financing, resulting in the harmonization of rules among the different issue areas and organizations to better address FLR governance challenges. In addition to coordination and harmonization efforts at the operative level of financial institutions, we also consider relevant decision at the regulative level of international negotiating bodies. We examine whether inter-institutional coordination and harmonization are produced by processes of mutual adjustment or by negotiated agreements, whether they take the form of functional differentiation or assimilation of rules, and whether they result in increased effectiveness, coherent rules and low transaction costs or have negative impacts.
The analysis is methodologically rooted in a qualitative research agenda. Because the objective of this article is to study the different organizations involved in FLR financing and their respective safeguards and co-benefits rules, this analysis is based on a document analysis of UN and World Bank documents as well as secondary literature on FLR related financing and rule-making. This document analysis is further enriched by insights gained from participant observation of the two UNFCCC climate negotiations in May (SBI, SBSTA) and November (COP 23) as well as the Global Landscapes Forum in December 2017, all taking place in Bonn, Germany.

Results
It is consensus among those involved in international efforts to protect and restore forests that dedicated multilateral public finance will only be able to supply a minor amount of the money needed. It is thus essential not only to attract some niche private investments but to green regular flows of private and public money, most importantly in the area of agriculture [66,67]. However, big multilateral funds-especially climate funds-play a central role in setting standards, catalyzing other investments, and fostering the reform of national regulations. We therefore focus on the Global Environment Facility (GEF), the Green Climate Fund (GCF), the UN-REDD Programme and the World Bank climate funds.
The GEF has existed since 1994 and is the major financial mechanism of the CBD and the UNCCD. It is also one of the most important sources of REDD+ finance. Within its integrated strategy of Sustainable Forest Management (SFM), the GEF provides a specific funding window for FLR [68] (p. 163). The GCF was set up in 2010 as an additional UNFCC funding mechanism and assists developing countries in their climate change mitigation and adaptation activities. It has recently announced a pilot program for REDD+ results-based payments [69] and also provides support for REDD+ readiness and implementation activities [70]. While the GCF has no special funding window for FLR, the "landscape approach" is explicitly mentioned as an important instrument for "coordinating agriculture, restoration of degraded lands, sustainable forest management, and conservation in an integrated manner [ . . . ] [which] applies to virtually all eligible REDD-plus activities" [70] (p. 9).
The UN-REDD Programme, a UN collaborative program established in 2008, supports countries in preparing and implementing REDD+ strategies. It is jointly managed by the Food and Agriculture Organization of the United Nations (FAO), the United Nations Development Programme (UNDP) and the United Nations Environment Programme (UNEP). In 2014, UNEP and IUCN announced a collaboration [71] which aims at making FLR knowledge and restoration opportunity assessment tools available to UN-REDD Programme countries. Initial steps included the institution of an FLR helpdesk and pilot projects in several selected countries.
The World Bank not only plays a central role in multilateral development finance, but also administers three important climate funds which support REDD+ activities: The Bio Carbon Fund (BioCF), operational since 2004, the Forest Carbon Partnership Facility (FCPF), operational since 2008, and the Forest Investment Program (FIP), operational since 2009. The BioCF in its current Initiative for Sustainable Forest Landscapes (ISFL) follows an integrated landscape approach, supports large-scale land-use change programs and pilots a carbon accounting methodology which assesses overall emissions reductions from land-use change including restoration activities [72]. In comparison, the FCPF is more narrowly focused on REDD+ [73]. The FIP is intended to pilot or scale-up innovative approaches to forests as a climate change issue. Similar to the BioCF it aims at large-scale projects which result in nationally significant shifts in forest management practices [74] (p. 5). An increasing amount of FIP finance goes to "landscape approaches" [75] (p. 4). In its overall Forest Strategy, published in 2002 [76], and its Forest Action Plan for 2016-2020 [77], the World Bank proclaims its intention to move beyond small-scale forest projects to a programmatic approach, which supports forest-smart national development policies. One important feature of this are integrated finance packages which combine funding from the World Bank Group development agencies with money from the climate change trust funds administered by the World Bank as well as other sources [77] (pp. [53][54][55]. An "integrated landscape approach" [77] (p. 14) is supposed to guide all World Bank forest interventions [77] (p. 28).
In the overall landscape of international institutions contributing to the governance of FLR (see Figure 1), these funds constitute a middle layer between regulative institutions and implementing institutions. The former include the UNFCCC, the CBD, the UNCCD, and the SDGs. The latter comprise the World Bank and other multilateral development banks (MDBs) as well as several UN subsidiary organizations and international non-governmental organizations (NGOs). Thus, the World Bank, several regional MDBs, the FAO, UNDP, UNEP, Conservation International (CI), IUCN and the WWF are accredited among others as implementing agencies to both the GEF and the GCF. The Inter-American Development Bank (IADB) and UNDP also serve alongside the World Bank as delivery partners for the FCPF. The FIP is implemented by the World Bank together with other MDBs. The BioCF is implemented by the World Bank. UNDP, UNEP and FAO together implement the UN-REDD Programme. Notably, the financing landscape for FLR does hardly differ from the one for REDD+. However, for the more integrative approach of FLR, the issue of cooperation and harmonization between these institutions needs to be reconsidered.
The Inter-American Development Bank (IADB) and UNDP also serve alongside the World Bank as delivery partners for the FCPF. The FIP is implemented by the World Bank together with other MDBs. The BioCF is implemented by the World Bank. UNDP, UNEP and FAO together implement the UN-REDD Programme. Notably, the financing landscape for FLR does hardly differ from the one for REDD+. However, for the more integrative approach of FLR, the issue of cooperation and harmonization between these institutions needs to be reconsidered. The need for inter-institutional coordination and harmonization of rules in the area of FLR arises not because the goals of the relevant regulatory organizations inherently contradict each other, but because their operationalization in specific projects can produce-positive or negative-externalities for other areas of action. Throughout the institutional landscape relevant for FLR, two instruments for the harmonization of rules have evolved, which we will analyze in the following: through the specification of co-benefits, institutions intend to provide for positive externalities and through the establishment of safeguard systems, they seek to avoid negative externalities. In reality, the distinction between the two instruments is not as clear-cut as depicted here. Often, safeguard standards also stipulate some areas where co-benefits are to be sought. Their central purpose, however, is to "safeguard" against unintended negative consequences, while the purpose of co-benefits is to generate other benefits in addition to the core objective. The difference between the two instruments as ideal types is illustrated by Figure 2. implementing agencies in green, and the non-governmental implementing agencies in light green (own illustration).
The need for inter-institutional coordination and harmonization of rules in the area of FLR arises not because the goals of the relevant regulatory organizations inherently contradict each other, but because their operationalization in specific projects can produce-positive or negative-externalities for other areas of action. Throughout the institutional landscape relevant for FLR, two instruments for the harmonization of rules have evolved, which we will analyze in the following: through the specification of co-benefits, institutions intend to provide for positive externalities and through the establishment of safeguard systems, they seek to avoid negative externalities. In reality, the distinction between the two instruments is not as clear-cut as depicted here. Often, safeguard standards also stipulate some areas where co-benefits are to be sought. Their central purpose, however, is to "safeguard" against unintended negative consequences, while the purpose of co-benefits is to generate other benefits in addition to the core objective. The difference between the two instruments as ideal types is illustrated by Figure 2. How safeguards and co-benefits address externalities: Safeguards rule out actions in the red area, but permit all actions, which are fully or partly green (for institutions aiming at goal 1) or blue (for institutions aiming at goal 2). A co-benefits approach focuses on the smaller green-blue striped area (own illustration).

Specification of Co-Benefits
At the level of regulative institutions, co-benefits are not specified in a very concrete or binding manner, but institutions seek to integrate goals from other issue areas into their programs. Most vigorously, the SDGs have brought together economic, social and environmental dimensions of global development. In the preamble of the Agenda 2030, the "protect[ion of] the planet from degradation, including through […] sustainably managing its natural resources and taking urgent action on climate change" comes second, right after the determination to "end poverty and hunger, Figure 2. How safeguards and co-benefits address externalities: Safeguards rule out actions in the red area, but permit all actions, which are fully or partly green (for institutions aiming at goal 1) or blue (for institutions aiming at goal 2). A co-benefits approach focuses on the smaller green-blue striped area (own illustration).

Specification of Co-Benefits
At the level of regulative institutions, co-benefits are not specified in a very concrete or binding manner, but institutions seek to integrate goals from other issue areas into their programs. Most vigorously, the SDGs have brought together economic, social and environmental dimensions of global development. In the preamble of the Agenda 2030, the "protect[ion of] the planet from degradation, including through [ . . . ] sustainably managing its natural resources and taking urgent action on climate change" comes second, right after the determination to "end poverty and hunger, in all their forms and dimensions" [1]. SDG 13 addresses climate change mitigation and adaptation and, in SDG 15, countries resolve to "protect, restore and promote sustainable use of terrestrial ecosystems, sustainably manage forests, combat desertification, and halt and reverse land degradation and halt biodiversity loss" [1] (p. 14).
The UNFCCC Paris Agreement "aims to strengthen the global response to the threat of climate change, in the context of sustainable development and efforts to eradicate poverty" [78] ( § 2). In 2010, REDD+ parties affirmed at the 16th Conference of Parties (COP 16) in Cancún that REDD+ actions should be "consistent with the objective of environmental integrity" and "implemented in the context of sustainable development and reducing Poverty" [79] (p. 26).
The CBD Strategic Plan for Biodiversity 2011-2020 emphasizes that "[b]iological diversity underpins ecosystem functioning and the provision of ecosystem services essential for human well-being [ . . . ] and is essential for the achievement of the Millennium Development Goals, including poverty reduction" [2] (p. 6). Strategic goal D is to "enhance the benefits to all from biodiversity and ecosystem services" [2] (p. 9). Aichi target 14 aims to restore ecosystems that provide essential services, taking into account the needs of women, indigenous and local communities as well as the poor and vulnerable by 2020. Aichi target 15 sets the goal to restore 15 percent of degraded ecosystems by 2020 with the rationale of contributing to climate change mitigation and adaptation and to combating desertification [2] (p. 9).
In its ten-year strategic plan 2018-2028, the UNCCD lists the improvement of the living conditions of affected populations as its second strategic objective right after the improvement of the conditions of affected ecosystems. Biodiversity and climate change are addressed under strategic objective number 4: "To generate global environmental benefits through effective implementation of the UNCCD" [80] (pp. [3][4]. Financing institutions adopt these proclamations of integrated aims from the regulative institutions they serve and make them effective through the more concrete specification of co-benefits recipient countries need to include in their funding proposals. If financial institutions are connected to several regulative institutions, this can foster an enhanced integration of goals as exemplified by the GEF, which has since 2000 promoted projects that cut across several of its convention-related focal areas and integrated funding lines like the GEF-6 SFM strategy [81] (p. 7). The objective of the sub-program on FLR is to "reverse the loss of ecosystem services within degraded forest landscapes" [68] (p. 171). Success is measured in terms of the "area of forest resources restored in the landscape" in hectare [68] (p. 171). Poverty reduction and livelihood enhancement are mentioned as goals of the SFM strategy alongside biodiversity conservation, climate change mitigation, and combating desertification [68] (pp. 153-158), but not translated into indicators.
Member countries are not as enthusiastic about integrated funding lines as the GEF Secretariat, because it is unclear how integrated approaches count in member countries' commitments under the Rio Conventions. Furthermore, ministries want to have clear budgets for their respective policy fields, which is not necessarily compatible with the idea of policy integration. Representatives of the Rio Conventions, too, lobby for clear budgets allocated to the implementation of their programs, which indicates an inter-institutional competition for funds [82,83]. GEF Programming Directions are consequently a compromise between the integrative ambitions of the GEF Secretariat and the needs of member countries. The Secretariat's first draft for the GEF-7 Programming Directions, for example, focused on 14 integrated Impact Programs including one on Landscape Restoration [84]. After discussions with stakeholders at the first replenishment meeting in March 2017, the Secretariat Sustainability 2018, 10, 757 8 of 19 revised this draft. As a result, the focus was shifted back to focal area strategies and the number of Impact Programs was reduced to four. FLR became part of a new Impact Program on Food Systems, Land Use and Restoration. In addition, it will continue to be an important instrument within the Impact Program on SFM. In GEF-7, the focus of the SFM Impact Program will be on three selected ecosystems (the Amazon, the Congo Basin, and important drylands) [85].
The core goals of the GCF are climate change mitigation (measured in tons of carbon dioxide equivalent reduced) and adaptation (measured in total number of beneficiaries and number of beneficiaries relevant to population) [86] (p. 7). The fund's Initial Investment Framework contains further criteria for funding decisions. These include the projects' "sustainable development potential" in terms of environmental co-benefits (e.g., in the areas of eco-system services and biodiversity conservation), social co-benefits (e.g., in the areas of health and safety, education, good governance and culture), economic co-benefits (e.g., in the areas of job creation, poverty alleviation, and economic development), and gender-sensitive development impact (with regard to correcting existing inequalities in climate change vulnerability) [87]. In the GCF's pilot project on REDD+ results-based payments, the provision of information on non-carbon benefits is optional. However, non-carbon benefits can improve the overall scoring of funding proposals with regard to project selection [88] (p. 19). This is in accordance with UNFCCC decisions on REDD+, which recognize the importance of non-carbon benefits [89] (p. 26) but leave their consideration at the discretion of countries [90] (p. 15).
The UN-REDD Programme closely follows UNFCCC guidance. Accordingly, it supports countries to enhance non-carbon benefits "where desired" [91] (p. 13). Nonetheless, the declared overall development goal of the Programme is to "reduce forest emissions and enhance carbon stocks from forests while contributing to national sustainable development" [91] (p. 7) and its safeguard standards address positive externalities for poverty reduction and sustainable development, including biodiversity protection and other environmental and natural resource management objectives [92] (p. 6).
The World Bank's Forest Action plan is designed to address the bank's core mission "to end extreme poverty and boost shared prosperity in a sustainable manner" [77] (p. vii) in an integrated way. The livelihood value and economic potential of forests plays the central role in the Forest Action Plan, but climate change adaptation and mitigation are also understood as genuine goals of World Bank activities. The Forest Action Plan is aligned with the World Bank's Climate Change Action Plan and supposed to serve the "global forest agenda" [77] (p. 2), which includes the SDGs, the UNFCCC Paris Agreement and the Bonn Challenge as well as the United Nations Forum on Forests (UNFF) agenda. The Forest Action Plan's second focus area of forest-smart interventions in other sectors is intended to address trade-offs between different land uses and to deliver integrated solutions for complex developmental problems. Emphasis is placed on comprehensive ex-ante analysis of possible trade-offs and synergies [77] (p. 29).
Reflecting the overall World Bank approach, non-carbon benefits are an integral part of the FCPF emissions reductions program as well as BioCF and FIP programs. The FCPF and the BioCF list improved local livelihoods, transparent and effective forest governance structures, securing of land tenure, protection of biodiversity and/or other ecosystem services as desirable co-benefits [93] (p. 25), [94] (p. 7). The FIP focuses on co-benefits for forest-dependent communities, including indigenous peoples, the protection of biodiversity and the adaptive capacity of forest ecosystems and their human inhabitants. Both in the BioCF [95] (p. 8) and the FIP [96] (pp. [5][6] countries are required to report on the number of people having received livelihood benefits.

Establishment of Safeguard Systems
At the level of regulative institutions, the UNFCCC COP 16 in 2010 in Cancún has adopted a set of broad but not very detailed safeguards for REDD+ activities [79] (pp. [26][27]. The CBD COP 9 in Bonn in 2008 took up the issue of potential REDD+ impacts on biodiversity and an Ad Hoc Technical Expert Group on Biodiversity and Climate Change (ADTEG-BDCC) was established, which produced a report on Connecting Biodiversity and Climate Change Mitigation and Adaptation, published in 2009 [97]. However, the report was not officially endorsed by the CBD COP. Even though UNFCCC negotiators were unwilling to consider the report or include very specific biodiversity provisions into REDD+ safeguards, the report nevertheless raised attention for the issue of REDD+ impacts on biodiversity. At the CBD COP 10 in Nagoya in 2010, important REDD+ recipient countries resisted any formal input to the UNFCCC, but the COP mandated the CBD Secretariat to provide advice on the application of REDD+ safeguards in a biodiversity-protecting manner and to convene an expert workshop on the issue together with the UNFCCC. In addition, the CBD organized three regional workshops. [98] (pp. 15-16). The resulting recommendations [99] were adopted as voluntary guidance for countries by the CBD COP 11 in Hyderabad in 2012 [100] (p. 212). Binding biodiversity safeguards for REDD+ were prevented by several recipient countries, which had no interest in strong and specific regulations. In the UNFCCC as well as the CBD, they raised the argument, that the institution had no mandate to deal with issues of the other institution [98] (pp. 15-16), [42] (pp. 1-2). This case illustrates how complicated and challenging inter-institutional cooperation and coordination can be when regulative institutions try to find ways to better integrate policies and related regulations. Legal aspects and member countries' interests and commitment play a decisive role in the pursuit of more integrated approaches.
At the level of financing and implementing institutions, each organization has its own safeguards policy, which all of its projects need to comply with. There are some differences between these policies, but diversity is limited by the fact that all of them are directly or indirectly informed by the World Bank safeguards. These have become a kind of de-facto minimum standard. The World Bank has recently updated its safeguards policy. The new Environmental and Social Framework (ESF) [101], which is 148 pages long, becomes effective in January 2018 and contains 10 Environmental and Social Standards, which are-in contrast to the UNFCCC Cancún safeguards-spelled out in quite some detail.
The GCF is still in the process of developing its own safeguard standards. It has adopted interim safeguards based on the 2012 Performance Standards for Environmental and Social Sustainability of the International Finance Corporation (IFC) [102], the World Bank daughter for engagement with the private sector [86] (p. 3). While geared more to private sector project partners, the 2012 IFC performance standards are otherwise more or less identical with the 2018 World Bank ESF. The GEF's Agency Minimum Standards on Environmental and Social Safeguards adopted in 2012 [103] are modeled very closely on the 2005 World Bank Environmental and Social Safeguard Policies [104], which are still much less systematic in their coverage than the new World Bank ESF.
The UN-REDD Programme's Social and Environmental Principles and Criteria (SEPC) adopted in 2012 build on the UNFCCC Cancún Safeguards as well as the safeguard policies of the three implementing agencies [92] (p. 2). Considering that the FAO, the UNEP and the UNDP are accredited to the GEF and the GCF, which require their implementing agencies to meet standards modeled on the World Bank safeguards, their respective safeguards systems can be assumed to be consistent with (at least the 2005) World Bank standards. Nevertheless, the UN-REDD Programme SEPC is unique in its integration of the UNFCCC Cancún safeguards.
The other financial institutions funding REDD+ activities have found different ways to address the Cancún safeguards (see Table 1). In its Methodological Framework for the Carbon Fund, the FCPF requires Emissions Reductions Proposals to "promote and support" the Cancún safeguards in addition to "meet[ing]" the World Bank safeguards [93] (p. 18). Tasked with clarifying how the two standards relate to each other, the FCPF management team concluded that the "application of the World Bank's safeguards [ . . . ] should be sufficient to ensure that the World Bank's safeguards successfully promote and support the UNFCCC safeguards for REDD+" [105] (p. 1). The World Bank can revert to this position also with regard to the FIP which, too, requires implementing MDBs to ensure projects' consistency with UNFCCC decisions on safeguards [106] (p. 8). The BioCF does not explicitly require application of the Cancún safeguards, but in the ISFL Emissions Reductions Program Requirements, it is nevertheless pointed out that UNFCCC REDD+ safeguards will be satisfied through the application of the World Bank safeguards [94] (p. 5). Contrary to the World Bank's assessment, the GCF concludes that "there are key differences between the Cancun safeguards and the GCF environmental and social safeguards standards" [69] (p. 8). In its pilot program for REDD+ results-based payments, the GCF has split the task of complying with both standards between actors: While it is the responsibility of the GCF accredited entities implementing REDD+ projects to describe how the GCF safeguards have been met, the countries are responsible for demonstrating how the Cancún safeguards have been addressed [69] (p. 9). The GEF, which does not have a specific REDD+ program, does not require projects to meet the Cancún safeguards.
All of the safeguard standards cover a wide array of topics including provisions on good governance, stakeholder engagement, and the protection of human rights and security as well as obligations to avoid negative (and to some extent foster positive) externalities for environmental integrity and sustainable development. With regard to negative externalities for FLR, three clusters of provisions are most important: (1) provisions addressing GHG emissions, which prevent negative externalities for the UNFCCC goal of climate change mitigation; (2) provisions aiming to protect healthy eco-systems and to prevent negative externalities for the CBD aim of biodiversity protection (if they protect habitats), the UNCCCD mission of combating desertification (if they halt degradation and foster sustainable management of natural resources), and the UNFCCC goal of climate change adaptation as well as the SDG project of sustainable development (as they ensure people's continued access to ecosystem services); and (3) provisions guaranteeing access to livelihood resources by protecting indigenous rights and restricting involuntary resettlement to prevent negative externalities for sustainable development as enshrined in the SDGs (see Figure 3).
Surprisingly, provisions against project-related increases in GHG-emissions are not part of the 2005 World Bank safeguards (or those of the GEF). These were integrated into the World Bank safeguard systems in 2012 when IFC performance standards were set up. There and in the 2018 World Bank ESF, climate pollutants are considered within the standard on Resource Efficiency and Pollution Prevention and Management. The stated objective is to "avoid or minimize project-related emissions of short and long-lived climate pollutants" [101] (p. 61). Due to climate change mitigation being the primary goal of REDD+, the Cancún safeguards and the UN-REDD SPEC focus on the more specific risk of reversals of emissions reductions achieved in the forest sector or their displacement to other forest or non-forest carbon stocks [79] (p. 27), [92] (p. 7).
Safeguards for the protection of healthy ecosystems are developed most with regard to biodiversity conservation. Provisions against land degradation are mainly treated as an aspect of habitat protection. Sustainable forest management and eco-system services are not as prominent as biodiversity concerns, but nevertheless receive increased attention.
The World Bank 2005 safeguards contain standards on Natural Habitats and on Forests (the GEF safeguards combine them under the heading of Protection of Natural Habitats). These standards focus on the protection of forests and other natural habitats from degradation or conversion, including conversion to planted forests. Plantations need to be designed to prevent the introduction of invasive species [107] (pp. 1-2) [108] (p. 1), and commercial harvesting operations must follow standards of "responsible forest management" [107] (p. 2). The maintenance of ecological functions is mentioned as an aim in the standard on Natural Habitats [108] (p. 1) and the standard on Forests specifies that forest restoration activities shall maintain or enhance ecosystem functionality [107] (p. 1). GEF standards are more stringent than the 2005 World Bank standards with respect to the introduction of potentially invasive, non-indigenous species [103] (p. 5).
obligations to avoid negative (and to some extent foster positive) externalities for environmental integrity and sustainable development. With regard to negative externalities for FLR, three clusters of provisions are most important: (1) provisions addressing GHG emissions, which prevent negative externalities for the UNFCCC goal of climate change mitigation; (2) provisions aiming to protect healthy eco-systems and to prevent negative externalities for the CBD aim of biodiversity protection (if they protect habitats), the UNCCCD mission of combating desertification (if they halt degradation and foster sustainable management of natural resources), and the UNFCCC goal of climate change adaptation as well as the SDG project of sustainable development (as they ensure people's continued access to ecosystem services); and (3) provisions guaranteeing access to livelihood resources by protecting indigenous rights and restricting involuntary resettlement to prevent negative externalities for sustainable development as enshrined in the SDGs (see Figure 3).  In the 2018 World Bank ESF (as well as the 2012 IFC performance standards), the issues of natural habitats and forests are integrated into a general standard on Biodiversity Conservation and Sustainable Management of Living Natural Resources. Protection is extended to modified habitats with significant biodiversity value and restrictions on the introduction of potentially invasive alien species are strengthened [101] (pp. [102][103]. Commercial production and harvesting of living natural resources now needs to follow standards of "sustainable management" [101] (p. 103). Reflecting its connection to human livelihoods, the issue of eco-system services has moved to the standard on Community Health and Safety. More emphasis is put on avoiding negative impacts on ecosystem services [101] (p. 69).
In the same way but less detailed as the World Bank standards, the Cancún safeguards call for REDD+ activities to be "consistent with the conservation of natural forests and biodiversity, ensuring [ . . . ] [they] are not used for the conversion of natural forests, but are instead used to incentivize the protection and conservation of natural forests and their ecosystem services" [79] (p. 26). While this is not part of the provisions labeled as safeguards, the UNFCCC decisions taken in Cancún 2010 stipulate that REDD+ activities shall "promote sustainable management of forests" [79] (pp. [26][27]. Elaborating on the Cancún safeguards, the UN-REDD SEPC requires REDD+ programs to "protect natural forests from degradation and/or conversion" [92] (p. 6) through REDD+ activities and to avoid negative impacts on biodiversity. REDD+ programs shall not involve the conversion of natural to planted forest, unless as part of forest restoration. They shall "maintain and enhance multiple functions of forests including conservation of biodiversity and provision of ecosystem services" [92] (p. 7) through land-use planning and forest management and also "avoid or minimize adverse impacts on non-forest ecosystem services and biodiversity" [92] (p. 7).
With regard to negative externalities for people's livelihoods, two issues are addressed: involuntary resettlement and indigenous rights. The World Bank 2005 (and the GEF) safeguards contain rules on avoiding, minimizing and mitigating restrictions in access to land and involuntary resettlement. Affected people shall be assisted in enhancing or at least restoring their livelihoods [109]. The most important strengthening in the World Banks 2018 ESF is a prohibition of forced evictions [101] (p. 77). The Cancún safeguards do not address the issue, but the UN-REDD SEPC even goes beyond World Bank standards by stipulating that countries should "ensure there is no involuntary resettlement as a result of REDD+" [92] (p. 5).
The World Bank standards on Indigenous peoples require full respect for indigenous people, their rights and culture. Adverse impacts on them are to be avoided, minimized and mitigated including through provisions on culturally appropriate social and economic benefits. While the 2005 World Bank standards [110] only ask for Free, Prior and Informed Consultation with indigenous people, the UNFCCC REDD+ safeguards (even though only implicitly) [ Even though the differences between the safeguard systems of the different financial institutions are not very large, the coexistence of several standards constitutes an organizational hurdle unnecessarily burdening national institutions tasked with coordinating REDD+ and FLR activities [37] (p. 11). Recognizing this as a problem, the two major financing institutions supporting REDD+ readiness activities, the FCPF RF and the UN-REDD Programme sought to coordinate their approaches from a very early point on. In 2012, they adopted joint Guidelines on Stakeholder Engagement in REDD+ Readiness [111] and a common Readiness Preparation Proposal Template [112]. While financial as well as implementing institutions routinely collaborate with other institutions in individual forest projects, the FCPF RF and the UN-REDD-Programme have so far been the only two institutions integrating their procedures in a permanent way.

Discussion
All of the financial institutions analyzed in this article have established fairly effective rules for safeguards in order to manage positive and negative externalities relevant for FLR. They thus provide a reasonably well financial infrastructure for FLR activities in developing countries. All funds promote the integration of co-benefits and consider them in their funding decisions, but countries retain great leeway in determining which co-benefits they will integrate in their programs. This is reasonable for general institutional safeguard standards, which need to fit a great variety of activities in diverse contexts. It would be desirable, however, for forest-related multilateral public funds to develop FLR-specific funding lines that contain more stringent requirements pertaining to the multiple benefits associated with the FLR concept.
The safeguard systems of all financial institutions are quite elaborate and coherent across organizations. However, areas for further improvement exist. Safeguards regarding GHG emission could become more detailed and standards for the protection of healthy ecosystems could integrate issues beyond biodiversity like land degradation and eco-system services to a greater extent. Biodiversity NGOs advocate for the inclusion of requirements regarding genetic diversity and adaptation potential of seeds used for restoration projects [113,114]. Indigenous peoples demand that their right to Free, Prior and Informed Consent is respected in all situations and that consent is clearly defined as a positive decision by affected communities' own institutions [115], ( [116], p. 51). It would be desirable that all institutions clearly rule out involuntary resettlement and include general provisions requiring countries to ensure that projects do not negatively affect local communities' livelihoods.
Comparing institutions at the operative level of financial (and implementing) organizations with rules at the regulative level of international negotiating bodies, co-benefits as well as safeguards are more developed and inter-institutional coordination is more successful at the operative level. The UNFCCC's Cancún safeguards for REDD+ are much less detailed and encompassing than the safeguard systems at the institutional level of the funds. While the attempt of the CBD to provide input to the UNFCCC negotiations regarding biodiversity safeguard standards for REDD+ failed to produce a binding outcome, inter-institutional coordination at the operative levels works fairly well. It involves negotiated agreements like the common standards of the FCPF and the UN-REDD Programme as well as processes of mutual adjustment. Prevailing inter-institutional dynamics at the operative level can be seen to support a race to the top: Organizations periodically update their safeguard systems to catch up with advances made by other organizations. In addition, is it usually agreed that in projects involving several funding organizations the more stringent of the safeguard provisions apply (see, e.g., ([117], p. 3), ( [118], p. 12)).
Harmonization, however, takes the form of assimilation rather than resulting in a division of labor. The concept of functional differentiation would suggest that institutions take the lead in developing provisions regarding positive and negative externalities for the area of their own expertise. This has been attempted but proven unsuccessful at the regulatory level, when the CBD tried to provide biodiversity safeguards for REDD+. At the operative level, no attempts in this regard have been made. It is the World Bank that de facto sets minimum standards in all safeguard areas, including environmental issues where it has less expertise than other institutions in the field. However, while the safeguard provisions of all funds are oriented on the World Bank's in some way, each organization nonetheless cultivates its own safeguards system. At first sight, this seems to contradict the idea of an integrated approach to FLR, suggesting that each institution has an ax to grind leading to institutional competition rather than to cooperation. Inter-institutional cooperation has proven the opposite, leading to a harmonization of safeguard rules between financial institutions supporting FLR, which do not provide problematic opportunities for forum shopping. The problem is that they have not led to a reduction of transaction costs, particularly for implementing agencies. The continuous mutual adaptation of standards requires a lot of energy from the institutions and the fact that each institution still has their own safeguard system is burdensome for recipient countries, which often need to fulfill the requirements of several institutions at the same time. It would not be desirable to reduce transaction costs by dropping some provisions from safeguard systems and thus lowering their standards. However, organizations should strive to reduce transaction costs by eliminating unnecessary duplications of standards and harmonizing application procedures (see FCPF RF/UN-REDD Programme). Ideally, all organizations would use the same safeguard system. In a functional division of labor, the standards for each issue area would be supplied by the organization with most expertise. Thus, for instance, the CBD would provide provisions for the protection of biodiversity, which would be used by all organizations. However, such a division of labor is probably too far away from prevailing practices to be easily implemented. In addition, as the CBD/UNFCCC example has shown, member countries seem to be rather reluctant to expand existing mandates. A pragmatic way to reduce transaction costs for recipient countries would be the adoption of the World Bank ESF by all organizations. De facto, all other organizations' safeguard systems are modeled on those of the World Bank anyway. The replacement of their safeguard systems with the World Bank ESF would thus not change the substance of requirements, but it would simplify procedures for recipient countries and spare the organizations the work of periodically comparing and updating their safeguard systems. This, in turn, would require an institutionalized way of inter-organizational cooperation to jointly update and adapt safeguards, to support a race to the top and to avoid a sense of patronization. Where organizations or some of their projects really want to address safeguard issues that go beyond those of the World Bank, they could specify additional requirements in their procedures. This would make actual differences more transparent than the co-existence of several largely congruent safeguard-systems.