Next Article in Journal / Special Issue
Beyond Rewards and Punishments in the Brazilian  Amazon: Practical Implications of the REDD+  Discourse
Previous Article in Journal / Special Issue
Co-Operation or Co-Optation? NGOs’ Roles in Norway’s International Climate and Forest Initiative
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:

Towards a Role-Oriented Governance Approach: Insights from Eight Forest Climate Initiatives

Chair Group of Forest and Environmental Policy, University of Freiburg, Tennenbacher Str. 4, D-79106 Freiburg, Germany
Author to whom correspondence should be addressed.
Forests 2017, 8(3), 65;
Submission received: 31 October 2016 / Accepted: 25 February 2017 / Published: 28 February 2017
(This article belongs to the Special Issue REDD+ Crossroads Post Paris: Politics, Lessons and Interplays)


In forest climate governance processes such as REDD+, non-state actors take on various, more or less formal, but in fact potentially authoritative governance tasks when informing, financing, (co)deciding or implementing forest climate action. Drawing on the concept of social roles, we investigate eight different REDD+ governance processes and how a variety of practical authoritative roles are enacted in administration, finance, decision-making and knowledge production. By systematically revealing the distinct ways of how different roles were filled, we developed a first (potentially still incomplete) typology of role practices and underlying rationales within different governance settings. In this endeavor, the role concept offered a valuable and handy analytical tool for empirically operationalizing governance performance, which is principally compatible with both institutional and social constructivist approaches to legitimacy.

1. Introduction

Anthropogenic climate change is one of the most pressing contemporary societal challenges affecting human life and the environment across the globe [1,2,3] and induces high economic costs on human societies [4]. With a contribution of about 10% to 20% to global emissions of greenhouse gases (GHG) [5,6], the unsustainable use and conversion of forests mark a significant driver of climate change, especially in developing countries [7,8]. Considering forests’ high potential as sinks for GHG emissions, the idea of establishing an international and results-based financial mechanism which addresses deforestation in developing countries was broadly welcomed by many Parties and stakeholders as early as 2005 at the 11th Conference of the Parties (COP) of the United Nations Framework Convention on Climate Change (UNFCCC) in Montreal, Canada.
However, it was not until COP 21 (2015) that negotiations were finalized for REDD+ (“Reducing emissions from deforestation and forest degradation and the role of conservation, sustainable management of forests and enhancement of forest carbon stocks in developing countries”). Formally, REDD+ is an international instrument that is implemented by nation-states, e.g., as part of their future Nationally Determined Contributions (NDCs). However, there exists a decisive national discretion in implementing NDCs; and, as the Paris Agreement highlights, there is indeed an increasingly important role for non-state actors in the design and implementation of concrete policies “to address and respond to climate change” [9]. Even before the actual REDD+ negotiations under UNFCCC were finalized, public and private actors were already actively engaged in activities to globally discuss or locally test REDD+, which de facto shaped the mechanism [10]. In addition, a number of preparatory steps were taken for capacity building for the implementing entities in ministries. Examples include the UN-REDD Program or the World Bank’s Forest Carbon Partnership Facility (FCPF). These preparatory steps proved decisive during the REDD+ negotiation process [11], and rather than seeing their legacy disappear, it is likely that they will form the foundation for the future REDD+ system. This includes the many ties that already exist between public and private engagement.
The involvement of private actors in REDD+ governance offers great chances for experimenting with innovative solutions to collective action problems, e.g., finance [12,13]. For that reason, polycentric policy networks [14] are—in theory—often seen as normatively superior in complex cases such as climate change [15,16]. One prominent reason is that they help to combine different advantageous capacities of markets (efficiency) or communities (context-sensitive political support) [12]. In practice, however, governance across the state–society divide may similarly fail or perform poorly [17,18]. One concern is that polycentric governance involves several actors from business, civil society, government and science simultaneously, which disperses governance authority [19,20,21]. Even where processes appear genuinely state-led, non-state actors may in fact fulfill potentially authoritative governance tasks when informing, financing, (co)deciding or implementing climate action without proper legitimation which has critical repercussions for the provision of collective goods [17,22]. For instance, recent (Foucauldian) governmentality studies showed that even seemingly neutral and technical practices such as notation, accounting and auditing may be highly political and influential [23,24]. Accordingly, it would be too narrow to focus only at the more or less formally legitimized decisions made by official representatives in state-driven processes. Joint working of public and private actors may be misused for scapegoating or absolving from accountability [18,25,26]. In addition, the higher complexity of mixed or hybrid (state/non-state) policy networks at and across local and transnational levels makes it increasingly difficult to identify which actors are excluded in these processes [10,14]. More scrutiny is required regarding supposedly “neutral” actor roles such as administration or expertise in polycentric REDD+ governance [22] and how exactly they interplay with formal decision-making pathways.
When answering questions about the performance of distinctive governance settings, e.g., regarding their legitimacy, it appears essential to be able to identify and differentiate the specific roles that actors hold within polycentric settings and whether they actually carry authority—and thus whether they need to be held accountable. Revealing such authoritative roles, however, is no easy task when applied to the myriad hybrid cases that make up REDD+ reality. There are several cases that are concerned with technical information, capacity building, financing or implementation oriented actions, such as carbon standards in forests or reforestation projects—cases that often are not officially coined as REDD+ but have repercussions for how REDD+ performs. Each case simultaneously engages a number of actors from different societal spheres in a broad range of different roles or functions. However, the actual nature of actors does not automatically determine their respective governance strategy. Business is not per se always trying to maximize monetary profits. Likewise, public welfare is not guiding all actions of civil society representatives. It would be highly presumptive or even naïve to suggest that belonging to a certain actor group automatically disqualifies or preserves an actor from engaging in certain actions. Involving some government representatives will not always increase democratic legitimacy, nor will business involvement principally increase efficiency but undermine democratic accountability.
Drawing on a role-oriented account of governance in networks, we investigate eight distinct cases of polycentric governance in the broader context of inter- and transnational forest climate policies. Specifically, we want to elaborate on the disparate ways in which different actors (jointly) fill potentially authoritative roles. Covering a variety of different cases, we want to reveal the distinct patterns that supposedly similar roles—such as secretariat, expertise or financing—may take in polycentric forest climate governance. Based on the first empirical insights gained from different cases, we discuss how such typology may be usefully employed for evaluative considerations about democratic accountability and legitimacy and thus complement established institutional approaches to democratic legitimacy.

2. The Relevance of Roles for Legitimacy

Typically, and drawing on Scharpf’s seminal work [27], legitimacy may most generally be conceptualized as having two analytical dimensions: input and output legitimacy. Input legitimacy generally refers to the inclusion of stakeholders in decision-making processes [28], and is concerned with the “right” procedures of participation, accountability and transparency [29,30]. Output legitimacy refers to the problem-solving effectiveness of a policy instrument and its result [31].
On the input side, representation is seen to be central. However, while official government delegates risk losing national mandates if their policies are not responsive to the voters’ will, the input accountability of non-state actors in transnational governance is typically not as straightforward or formalized. From an institutionalist view, specific qualities of the process may be highlighted through which non-state actors are included in decisions. It is hoped that institutional arrangements or decision-making procedures, such as voting rules, participatory boards or grievance structures help to achieve these qualities.
Output legitimacy in Scharpf’s understanding means effective problem-solving and, further, “that the policies adopted will generally represent effective solutions to common problems of the governed” [31]. As it is difficult to measure if the policy helped to address the original problem, possible criteria to capture output legitimacy are (1) the extent to which self-set goals were reached and (2) the level of compliance by the most important actors [28]. Moreover, output legitimacy depends, in large part, on expertise, where trusted actors with respective knowledge claims are entitled to define effective policies [32]. Where international organizations (IO) cannot deliver the needed expertise—and hence expert legitimacy—to public authorities in the transnational arena, other non-state actors may provide it [33]. However, the principal applicability of expert legitimacy in cases characterized by high uncertainty and stakes, as in cases such as climate change or natural resource use, may be questioned because expertise often excludes relevant stakeholders and their alternative ways of knowing from governance processes and their respective discourses [34,35]. For unstructured problems, there are no impartial evaluations of legitimate outcomes without considering the perspectives of those affected by decisions. Such “justification of actions” by referring “to those whom they affect according to reasons they can accept” [36] is a central theme in social constructivist notions of legitimacy. What matters in this view are the expectations that affected audiences have about an actor’s involvement in governance and whether these expectations are eventually met [36,37]. Legitimacy necessarily varies as it is dependent on the context, issue area and specifically the different legitimacy understandings of distinct political communities [38] (p. 19). Rather than assuming that some sort of pre-defined mechanism or institution will ensure legitimate decisions, legitimacy is taken as a dynamic stance—as a two-sided process between different actors with reciprocal expectations. However, in research practice this assumption lacks sufficient specificity as to identify relevant processes that require attention because—in principle—any more or less formalized processes with state or non-state actor participation may be decisive. Rather than overly relativizing, we follow Bernstein and suggest that legitimacy depends on “historical understandings and the shared norms of the particular community or communities granting authority” [39]. Hence, in the forest climate governance communities around REDD+, we can assume that expectations are not just randomized, but structured and institutionalized. Such a reading of legitimacy has been used for analyzing both singular hybrid and private settings in environmental governance [40,41].
Analytically, the concept of actor roles offers a way of incorporating the advantages of both social constructivist and institutional schools of thought. Simply put, roles are the characteristic behavioral patterns that form part of social life [42] and describe the relationship between the individual and the social system [43]. On the one side, they are typically grounded and expressed in formal rules or informal norms (e.g., statutes, contracts or working task descriptions, or societal functions). At the same time, they rest in individual and subjective expectations of both the role holders and the affected communities toward the respective social position [42]. With this characteristic, roles not only reveal the congruence between a given norm (formal expectation toward role holder) and a practice, but also between different subjective expectations of the role holder and affected communities, which are tied to this social position. Covering, formal (de jure) and subjective (de facto) expectations, the role concept helps to understand legitimacy as a socially constructed and structurally institutionalized (i.e., patterned and hence empirically more easily detectable) phenomenon. If actors are able to fulfill the formal and subjective expectations tied to their roles, they create acceptance and credibility for their actions. Moreover, the concept is not limited to public or genuinely political actors but can be employed to all sorts of actors in governance settings. Such specification appears especially relevant regarding the mentioned concern that authoritativeness may not only be tied to official decision pathways and institutional settings (parliaments, boards, plenums etc.), but also to seemingly harmless functions related to administration, budget, auditing or expertise.
When investigating the plethora of novel polycentric cases in forest climate governance, it appears critical to draw on general analytical categories that help identifying relevant authoritative roles. Börzel and Risse (2005) offered a first attempt to analytically grasp different governance purposes (rule-setting, implementation, service provision) and functions (co-optation, delegation, co-regulation or self-regulation) that private actors may have in hybrid Public–Private Partnerships [17]. One notion was that a self-regulated setting which sets formal rules completely independent from government was more problematic than private governance having evolved just for the ‘service’ of government (i.e., delegation) [17]. While such typologies are very helpful for distinguishing different ways or degrees of exerting authority, they may still disregard apparently unproblematic functions which have potentially decisive impacts on the provision of collective goods [22]; particularly for such contested issues as climate change. Moreover, a functional typology provides no empirical insight as to what specific types of roles, tied to specific organizational strands or tasks of a setting, are typically relevant for analysis. We suggest that authoritative roles may principally be tied to four different organizational branches of governance that hold significance for the input or output legitimacy of this setting’s operations: (1) administration, (2) funding, (3) internal decision-making and (4) knowledge making processes.

2.1. Administrative Roles

From the time Biermann and Siebenhühner conducted their seminal research about international bureaucracy, it has since become evident that administrative roles can be fairly relevant for questions about governance legitimacy [26]. The classical view in the public administration literature is that administrative roles encompass a “cohesive set of job-related values and attitudes” [44]. More specifically, attitudes that are assumed to ensure responsible behaviour. In practice, administrators evidently have to deal with contradictive role expectations that arise from simultaneously having discretion based on expertise while being obedient to external masters [43]. In principle, administrative roles may turn out to be fairly authoritative in operational decision-making, including agenda setting, book keeping, and communicating with stakeholders. However, rather than suggesting that an independent and strong secretariat is always necessary [45] or let alone legitimate, it is rather more important that such strength responds to valid expectations about the respective administrative role within a certain setting.

2.2. Financial Roles

While administration and finance appear as genuinely linked, because one and the same actor may be responsible for executive tasks, accountability relations from a role perspective may be completely different, not least because funding sources of governance setting are often external. The expectations tied to financial roles are typically linked to the dimension of input legitimacy and relate to concepts of transparency and accountability [29]. Considering that activities such as distributing and spending money always carry the risk of inefficiency or corruption, transparency becomes an important approach for ensuring input legitimacy of financial roles. Accountability refers to the “relationship between an actor and a forum, in which the actor has an obligation to explain and to justify his or her conduct. The forum can pose questions and pass judgment, and the actor may face consequences” [46]. Multilateral or bilateral finance is typically tied to a set of formal rules (conditions) and corresponding expectations about performance within or by the funded governance setting, a donor–recipient relationship well known from development cooperation [47]. Principally, recipients are accountable to providers and, for instance, they have to report regularly on the progress of their work, yet this accountability may be weaker if, for instance, the organization itself generates the necessary income (e.g., through selling services or products) or can draw on philanthropic (unconditional) donations.

2.3. Decision-Making Roles

On top of roles relating to financing and administration/organization [48] we need to consider roles that are more straightforwardly related to consequential decisions. Decisions are consequential if they relate substantially, rather than only procedurally, to a policy process. Such implementation and execution oriented roles determine fairly specific rules or methods for project definition, implementation or evaluation. A conceptual differentiation between more consequential, as compared to rather strategic or procedural decision-making modes is useful, because they may not be exerted by the same set of actors.
By contrast, strategic decision-making covers the overall goal orientation carrying expectations about what defines an effective outcome of the governance operations. These roles often relate to governance bodies such as Boards of Directors covering aspects of both input and output legitimacy. Evidently, the specific performance of such boards is not only defined by more or less formalized codes of conduct (statutes), but also depends on social–psychological processes regarding how actors participate or interact in such group and how exactly they exchange information or discuss controversial issues [49,50]. Regarding the input legitimacy of strategic decision bodies, adequate representativeness of those taking these decisions is important, whereas strategic roles may be filled by various actors, such as CEOs, owners, managers, shareholders or Boards of Directors.

2.4. Expert Roles

As mentioned earlier, expert legitimacy allows different non-state actors to be actively involved in public affairs. Experts might be invited to participate because they better understand the cultural context within a project area can provide lessons learned from pilot projects or bring in technical knowledge about legal or technological terms or scientific methods. The concept of epistemic communities [51] reflects this central role of expertise in (transnational) governance. Epistemic communities encompass “a network of professionals with recognized expertise and competence in a particular domain and an authoritative claim to policy-relevant knowledge within that domain or issue-area” [51]. Typically, policy processes draw on evidence-based science where technical, professional and bureaucratic expertise become decisive ingredients for decision-making [52]. In fact, expert knowledge carries high authoritative power because it influences consequential decisions (output dimension) fairly directly. The political power of knowledge is even more significant in unstructured environmental problems, where “both the nature of the ‘problem’ and the preferred “solution” are strongly contested” [53]. In such cases where ignorance and political opposition is strong, bureaucratic or scientific knowledge are hardly neutral. Rather, such knowledge is selective and partially supporting of certain positions while marginalizing others [54,55,56]. Revealing the different role related to knowledge and expertise and how they are institutionalized, including what expectations are tied to them, may be critical for evaluating the (output) legitimacy of governance in practice.
Given the aforementioned assumptions, seemingly neutral functions and their respective operational roles require analytical scrutiny. With the paper at hand, we attempt to systematically elaborate on the different rationales behind authoritative roles in forest climate governance such as REDD+ and, further, work to identify the different patterns that administrative, financial, decision-making and expert roles may take empirically. By exploring eight distinct cases of forest climate governance in a comparative fashion, we intend to provide a systematic typology of authoritative roles. In this way, we seek to generate valuable insights as to whether and how roles may serve as a helpful conceptual and methodological tool for operationalizing and, hence, evaluating the performance of governance settings, specifically concerning their legitimacy.

3. Methods

Our study of eight distinct global forest climate governance formats follows a comparative qualitative case study approach [44]. In order to carve out the variety of authoritative roles in polycentric governance, we follow a dual strategy. Our cases selected cover different hybrid governance settings that—first—range from more to less private actor intensive compositions and that—second—cover different fields of climate action (information, regulation, finance, project implementation). This way, we are able to provide an extensive informative basis that covers varied (potentially hidden) authoritative actor roles in governance across the whole policy cycle from problem-definition and agenda-setting to policy implementation and evaluation [45]. Broadly following the empirical variation in transnational policy processes that are more or less directly tied to the formal UN based REDD+ mechanism, this comparative study allows a nuanced discussion in different policy stages.

3.1. Case Selection

In an extensive online and literature review, we identified a set of 35 potential cases of transnational governance arrangements with relevance to the issue of forest climate policy and finance. This topical focus assured inter-case consistency and comparability. Finally, eight different settings were selected based on the criteria of fulfilling varied relevant REDD+ purposes (e.g., networking, implementing, standard-setting, evaluation and commitments). Four of these cases have—from a formal point of view—been strongly state-led activities, i.e., either informing global UNFCCC negotiations or national implementation: the REDD+ Partnership and the Standing Committee of Finance (both information exchange forums) as well as the Partnership for Market Readiness and the Association of Southeast Asian Nations (ASEAN) Forest Clearing House Mechanism (CHM) (two capacity building and exchange forums in support of national implementation of carbon markets and forest policy). Each time, however, some kind of private actor involvement was given—yet at different degrees and in different functions. In formal terms, the four other cases involve non-state or private actors in concrete implementation oriented activities more proactively, be it in the coordination and implementation of forest restoration projects in several countries, as in the case of the Bonn Challenge and the Green Belt Movement, or in developing and certifying global standards of sustainable land uses, as with the Roundtable on Sustainable Palm Oil and the Gold Standard.
Among the eight selected cases that engage in varied REDD+ related activities, four settings were rather closely tied to the UNFCCC process (UNFCCC Standing Committee on Finance, REDD+ Partnership, ASEAN Clearing House Mechanism and Partnership for Market Readiness), whereas the other four were relatively independent (Bonn Challenge, Green Belt Movement, Roundtable on Sustainable Palm Oil and the Gold Standard)—but still of relevance to the process. To gain in-depth insights into different performative and more or less authoritative roles in distinct governance settings, we look at different social roles of actors in internal operations. The idea is to not only list relevant actors in each setting, but to elaborate on their constellation and involvement in different decision-making procedures and contexts covering different operational branches, from administration and finance to strategic, procedural and consequential decisions as well as knowledge creation. Specifically, we asked which concrete actors are engaged in a certain role and with what expectation. By highlighting the major formal (institutionalized) and subjective (socially constructed) expectations, we are able to identify and reflect upon potential tensions and pitfalls tied to the specific roles within the respective setting, which would matter for considerations of the overall legitimacy of the cases.

3.2. Data Collection and Analysis

The eight case studies draw on extensive document analysis, participatory observation (where available) and expert interviews. Analyzed documents encompass, for instance, primary research about the setting, but also strategic documents and reports of the respective institution such as working programs or annual/evaluation reports. In addition, more popular public relations oriented documents (website, press releases, mission statement, self-description brochures) were used if they delivered insights into the internal procedures and the (expectations about the) roles of different actors. The insights from analyzing this material were iteratively enriched and validated with information gained from (typically) one additional qualitative semi-structured expert interview per case [48]. Most interviews were conducted face-to-face, by telephone or Skype mainly with experts at the strategic or operational level of the organization or with external actors (private or public stakeholders) that were considerably involved in or followed the governance operations at decisive points as to give an insight into the ways of how different actors work jointly together. Considering the diplomatic load of a number of the settings investigated and to encourage detailed and honest answers, interviewees are guaranteed full anonymity.
Table 1 provides an overview of the expert interviews and the year of the interview (besides the ASEAN CHM) all interviews were conducted between December 2015 and May 2016. In the case of the REDD+ Partnership, the Bonn Challenge and the Gold Standard, we build on our own observations collected during the participation in conferences, meetings and workshops. The respective codes are used to indicate the sources of information within the results.
All interviews were tape-recorded, transcribed and interpreted by means of qualitative methods of content analysis [48,49] along a set of predefined analytical governance dimensions (deductively) as well as on the basis of the empirical material itself (inductively). Four broader dimensions guided the analysis and comparison of the cases: administration, finance, decisions and expertise. To contextualize each case, a brief introduction is given about their development and set-up.

4. Introduction to the Cases

In the following, we briefly introduce each case (founding, objectives, general set-up) and its relevance for or relation to REDD+ before elaborating on the varied roles of different (esp. non-state) actors in administration, finance, strategic and consequential decisions as well as in knowledge making in a comparative fashion.

4.1. UNFCCC Standing Committee on Finance (SCF)

The SCF was established in 2010 to improve the coherence and coordination in the delivery of climate change financing by collecting and synthesizing relevant information to inform the ongoing negotiations under the UNFCCC. Accordingly, it is mainly composed of 20 delegates that are elected by members of the Conference of the Parties (COP) for two years. While (official) stakeholders may principally be admitted to attend regular meetings, there is also a forum for the “communication and continued exchange of information among bodies and entities”. It is organized at least once a year and includes different actors from business, science or civil society [57]. REDD+ finance was a major topic at the 3rd forum on Enhancing Coherence and Coordination of Forest Financing which took place in Durban, South Africa, on 8–9 September 2015.

4.2. REDD+ Partnership

Another learning platform to inform the ongoing negotiations on REDD+ was the REDD+ Partnership, which was active between 2010 and 2014. It was composed of official party partners as well as more than 100 “institutional partners”, i.e., stakeholders from science, business or civil society. It provided an informal setting for partners to discuss (but not formally decide on) activities, finance or technical and political issues related to REDD+ implementation. It held organizational meetings and technical workshops on selected topics of interest such as REDD+ safeguards, (forest) reference levels, measuring, reporting and verification (MRV) or REDD+ financing.

4.3. ASEAN Forest Clearing House Mechanism (CHM)

The ASEAN CHM was founded in 2004 and promotes networking and holistic knowledge management between ASEAN stakeholders, processes and institutions in the field of forest law enforcement and governance and REDD+. The knowledge and learning platform linked senior officials and regional experts of various ASEAN networks including state representatives, business actors and scientists. It is managed by the ASEAN Secretariat [58]. The network met twice a year and could rely on a broad digital infrastructure (interactive website, e-learning, webinar tools). The CHM served ASEAN stakeholders to agree on research agendas (e.g., about REDD+) and to learn about assessment tools related to forest law enforcement. A nominated network manager facilitated communications, but members were also expected to contribute actively [57,58].

4.4. Partnership for Market Readiness (PMR)

The PMR was initiated by the World Bank in 2010 with the objective to support middle income countries to prepare and implement financial climate change mitigation policies such as carbon pricing or carbon taxes. It provides capacity building through knowledge sharing as well as funding for technical assistance. The PMR consists of 13 contributing countries and 17 implementing countries, which together form the Partnership Assembly (PA). Decisions within the Partnership Assembly are made by consensus and may be informed by additional technical partners to provide expertise. The meetings are led by co-chairs, with one member from a contributing (high income) country and one from an implementing (middle income) country. The World Bank Group acts as secretariat and partner for the PMR. It organizes meetings, brings in experts, coaches the countries and monitors their progress based on implementation status reports. The technical assistance for the implementation of the proposals is done in partnership with the World Bank's operational units in each country [59].

4.5. Bonn Challenge (BC)

The Bonn Challenge was founded in 2011 and is an implementation platform for the three existing international commitments on forest restoration: Aichi Target 15 on Biodiversity, the “+“ in REDD+ (esp. enhancement of carbon stocks), and the “zero net land degradation” goal of Rio+20. The Bonn Challenge is a declaration of intent and provides a platform for information exchange, networking and capacity building on reforestation for the “coalition of the willing”. It promotes implementation processes by giving high level visibility to voluntary pledged commitments. Its aim is “to restore 150 million hectares of the world’s deforested and degraded lands by 2020” and another 200 million by 2030 [60]. So far, it has received pledges from 30 countries, one private company and six sub-national regions. Together, these pledges account for 112 million hectares [60]. The civil society organization World Wildlife Fund (WWF), scientific advisors from International Union for Conservation of Nature and Natural Resources (IUCN) and the World Resource Institute contributed to the strategic development of the platform. The Global Partnership on Forest and Landscape Restoration (GPFLR) which is hosted by IUCN is providing the infrastructure of the Bonn Challenge (website, meetings, publications and capacity buildings). Repeatedly, financial actors have been invited to Bonn Challenge meetings to attract more private investment.

4.6. Green Belt Movement (GBM)

The Green Belt Movement, founded in 1977, has offices in Kenya, London and Washington. GBM is a Kenyan grassroots organization that encourages and financially supports women in Africa to jointly grow seedlings and plant trees. The GBM contributed to the development of the Kenyan REDD+ process and acted as a partner for the government. Since 1986, the GBM launches a Pan African Green Belt Network across Ethiopia, Tanzania, Uganda, Rwanda and other African countries. In this context, it offers trainings and practical capacity building for grassroots and development groups. Nobel-peace price awardee Wangari Maathai was the founder of the GBM and basically guided the organization until her death in 2011. Since then, Wanjira Maathai, the daughter of Wangari Maathai, is the new head of the organization.

4.7. Gold Standard (GS)

The Gold Standard is a foundation that offers a global standard to certify carbon emission reductions and sustainable development impacts for carbon offset projects both for the compliance market (Clean Development Mechanism under the Kyoto Protocol) and the voluntary market. It was founded in 2003 by WWF and is endorsed by more than 80 NGOs. Their role is to comment on new methodologies and project design documents. Moreover, 25 “market specialists” from consultancies, think-tanks, NGOs and research institutions constitute the GS Technical Advisory Board, which decides on new methods and oversees the whole certification process [61]. To date, the standard has been applied to 1100 projects in 70 countries [61].

4.8. Roundtable on Sustainable Palm Oil (RSPO)

The RSPO, founded in 2004, is a transnational multi-stakeholder roundtable that seeks to transform the palm oil production. It has developed a private standard for sustainable palm oil. The RSPO is a relatively large setting composed of 3100 members with participation rights and 1400 ordinary members with voting rights. The latter are mainly consumer good manufactures, processors and retailers. The remaining ordinary members are palm oil growers and financial actors. Civil society organizations make up only 3.5% [62]. In the annual general assembly, board representatives are elected and resolutions are made based on simple majority vote.
Table 2 provides an overview of the key features of each of the eight governance settings which are relevant for the analysis of the roles outlined before. The table illustrates the fact that non-state actors are essentially engaged in all settings. It also shows that the sources of funding vary, ranging from bi- and multilateral state-based financing (ODA) to (public or private) donations or self-financing through the provision of payable services. Yet, evidently, the simple fact that stakeholders are formally related to a setting as participants or funders does not tell us whether they actually take on decisive roles which require those actors to render accountability. Such evaluation requires empirical substantiation and a more refined look at who fills specific roles in administration, decision-making, finance and expertise. Additionally, understanding their expectations can help answer the aforementioned question for specific cases.
In order to get a more complete picture of the whole spectrum of possible authoritative roles within the settings, we specifically look for commonalities and differences across the cases as to how specific roles are enacted.

5. Results

In the following section, we apply the role concept to the eight different cases. First, we examine administrative and operational decision-making roles that resemble each other regarding the more procedural pathways of influence, e.g., defining how operations are exercised or which actors are accepted to these operations and in what way. Second, we look at funding constellations and financing roles tied to different settings. Third, we analyze strategic and consequential decision-making roles, i.e., roles that are concerned with outcome-related decisions—both broadly and more specifically. Fourth, we aim to identify different expert roles in knowledge production and to reveal the respective rationale of and expectations about “expertise”. For illustration, we provide a figure that covers the different subvariants for each role type. Where identifiable, outer circles describe the (common) rationales and corresponding expectations behind a specific (set of comparable) role(s). Specific expectations about and why a specific actor carries a distinct role (experience, reputation, capacity, professional aspects, etc.) are explicated within each bubble.

5.1. Administrative Roles in Forest Climate Governance

Assuming that administrations (in their bureaucratic roles) may be fairly authoritative [26] for governance performance, we investigated the eight cases regarding what formal requirements as well as inter-personal expectations are tied to this respective role. We identified three different types of administrative roles that seemed to follow different rationales for engaging specific actors. (1) The first type is that of international bureaucracies. In such cases, the secretariat is closely tied to or even embedded in the international UN processes. Two examples are the Standing Committee on Finance, which is administered by the UNFCCC secretariat, and the REDD+ Partnership which is administered by the secretariats of the two major running funding programs on REDD+ (UN REDD PT/FCPF FMT). (2) A second type of secretariat draws on International Organizations to provide the respective services, e.g., in the case of the Partnership for Market Readiness, which is administered by the World Bank, or the Bonn Challenge, which is administered by the ICUN secretary. (3) A third type is the approach of having internal staff that administer—often in private organizations. Examples here are the Gold Standard and the RSPO or NGOs such as the Green Belt Movement. Figure 1 gives an overview of the three different administrative roles.
An important rationale for relying on (typically rather expensive and unwieldy) international bureaucracies (type 1) or international organizations (type 2) is that the respective actors are highly experienced in working professionally in diplomatic settings and possess specialized financial and organizational capacities or relevant networks for facilitating delicate policy processes with state representatives. Noteworthy to say, the cases of SCF and the REDD+ Partnership (type 1) are also more or less formally and directly oriented at and relating to ongoing international REDD+ negotiations. By contrast, such aspects appear to matter less in cases such as the Gold Standard or the Green Belt Movement, which mainly draw on their own (internal) staff. Overall, this administrative role seems to be more autonomous or independent from outside expectations and driven mainly by internal working requirements. The influence of internal staff secretariats (type 3) on relevant IO and UN networks or respective ongoing negotiations is formally much weaker than in cases of type 1 and 2 secretariats (such as the Bonn Challenge, PMR or SCF), where the same people physically take part in different settings. Figure 1 covers the spectrum of different administrative roles as identified in the eight forest climate governance cases. As mentioned earlier, each of the three smaller circles represents one role type (here: International Bureaucracies, International Organizations, Internal Staff) and the outer circles reflect the underlying (common) rationale (e.g., diplomatic vs. autonomous).
Secretariats with strong ties to external UN or IO organizations hold additional accountability relations, in part because of the need for reporting and information exchange with their home institutions (on working tasks, hours and respective funding). Such information however, is often made public, which increases the overall transparency of the setting. In a case such as the REDD+ Partnership, such an accountability relationship may become fairly complex. Its secretariat partly drew on staff from the UN-REDD Program, which itself is composed of three different international bureaucracies, UNDP, FAO and UNEP. Every UN organization has its own disclosure policy [58]. International organizations such as the World Bank, which administers the PMR, are also bound to formally codified transparency and reporting rules which assure accountability—at least internally. The UN is, in many cases, more accepted than the World Bank, especially among emerging economies which perceive the World Bank as an institution that serves primarily the interests of rich countries [63,64]. Even if the World Bank is just administering, it may affect the relationship between different constituencies in a setting and consequently hinder developing countries to participate at eye level (Expert 4, 2016).
Since organizations such as the Gold Standard or the Green Belt Movement are, by contrast, administered by their internal staff, such excessive external accountability relations are not a significant issue and typical employer–employee rules and legislation apply. For external communication, however, introducing the whole staff on their publicly accessible homepages is common sense and understandable if one considers that the administrative roles often include dealing with personal data or funding. By contrast, in type 1, where IO staff are concerned, such public presentation is less common; often to protect these authoritative people from external influence.
Operational decision-making appears to resemble administrative roles in terms of the day-to-day work they engage in, for instance as chairs of meetings that are concerned with the daily procedural decisions about how, when, and where to meet and what topics to set on the agenda. These seem like relatively harmless activities at first sight; but since such decisions, in fact, restrict certain (alternative) topics from being set on the agenda as much as they exclude certain stakeholders from contributing to decisions, this role is potentially highly authoritative. We identified four types of operational roles: (1) delegates, (2) administrative staff, (3) network facilitator and (4) stakeholders; these are visualized in Figure 2. Subtypes represented in the “bubbles” reflect the reasons for why a specific role was employed, in this case especially regarding expectations about professional skills. Moreover, we identified three common or distinct rationales tied to different operational decision subtypes: “closed shop”, “nodal” or “buy in”.
The first subtype of an operational role is that of official delegates (1), i.e., officially entitled state representatives which hold (co)chair positions. This role is typically tied to a whole set of diplomatic rules of the game that were often lent from other official (UN) arenas for defining how to organize meetings etc. The co-chair role in these cases, as in that of the SCF or the REDD+ Partnership, is in fact fairly powerful and occasionally it happened that relevant stakeholders from within and outside the setting questioned the legitimacy of the process because co-chairs used their authority for promoting own political preferences (own observations, 2010–2014). At the forum on forest finance of SCF, stakeholder participants were given the impression that they are not equally welcome and that those present were the usual suspects which could actually not provide the new information hoped for. This is why a number of participants also showed disinterest and left the event earlier (Expert 1, 2015). The driving cause appeared to be the proximity to the politically charged UN process and that some co-chairs simply could not avoid the negotiation modus when setting the agenda and tried to shape the discussions while avoiding alternative views to be taken into consideration (“closed shop”) (Expert 1, 2015).
On the other side, procedural roles of such settings that are close to international arenas are often also bound to make activities public on their website and to specify working procedures respectively, e.g., sharing work plans or progress reports—apart from exceptional cases of confidentiality. An example would again be the Standing Committee on Finance, which is administered by the UNFCCC secretary and co-chaired by two UNFCCC member state representatives. The meetings are videotaped and available online. A second type can be illustrated by the case of the Bonn Challenge, where administrative staff of a small team within the IUCN secretariat makes these day-to-day decisions. The role resembles that of a service-provider. This role also follows a “closed shop” rationale (though with less political charging), where only few people take potentially important operational decisions (see Figure 2). A third variant is when a neutral (3) network facilitator or mediator is employed to ensure a continuous communication process within a given network. This role appears to follow a “nodal” or mediating rationale as observed in the case of the ASEAN Clearing House Mechanism, where the idea was pursued that a neutral moderator with sufficient time resources and networking skills could be decisive for proper knowledge exchange in the rather political network (Expert 3, 2014). A fourth type of procedural role is that where all sorts of stakeholders (4) may actually “buy in” to procedural operations and contribute in administering the governance network if they wish to do so. An example of such a highly open setting that allows all more or less relevant actors to participate (intrude), is RSPO, where members can apply for such administrative roles given they are financed by their home organization, which may in fact favor members from large organizations while smaller ones do not have these capacities (Expert 8, 2016). We see the tendency for independent or self-administering governance settings to have as restrictive transparency rules as most of the international settings have. Dependent on what different audiences (civil society or transnational governance communities) expect and demand via diverse communication channels, they are obliged to justify their operations and make (most of) their decisions openly accessible [65,66]
Among the four types of roles, we find both rather hierarchical top-down roles (type 1 and 2) as well as more horizontally oriented roles of network facilitator or stakeholders settings. This mirrors the debate about the classical trade-off between inclusiveness and efficiency. However, the input legitimacy in settings with network facilitators and stakeholders involved might be significantly higher than in the case of delegates and administrative staff.

5.2. Financial Roles in REDD+ Governance

In the eight cases, we identified five different types of finance coming either from (1) multilateral funds, (2) international organizations, (3) nation states, (4) own funds, or (5) donations, each representing different accountability communities. Figure 3 gives an overview of the spectrum of different financial roles in terms of who provides the funding. The two wider circles illustrate how the funding relationship affects the work of the governance setting and how the role affects the possibilities to spend the money; either conditionally or autonomously. The words in italics below refer to the question of why these roles exist and what kind of relationship and what expectations are bound to them.
Multilateral funds typically expect that money is spent according to official requirements and formal diplomatic standards. This accounts e.g., for the Standing Committee on Finance, which has to consider UN rates, expected professional standards for venues and technical support, as well as documentation requirements if it organizes a forum meeting [67].
If the funder is an international organization, such as the World Bank, and the governance setting is in the role of a service provider, it will often be urged to follow a specific political agenda (Expert 5, 2015). If states finance a setting, accountability constellations typically evolve from development cooperation between donor and recipient. An example may be the German Government in the Bonn Challenge, the German Development Agency GIZ within the ASEAN CHM or the contributing states within PMR. Contrarily, Bonn Challenge funds, that heavily draw on German support, are coined mainly for administrative staff, public relations or the regular ministerial round tables and have no link to actual implementation as pledges are defined and financed by partners. The contributing countries in the PMR had the possibility to effectuate the involvement of a preferred country, which they would like to support with the funds (Expert 4, 2016), showcasing that contributing countries typically expect something back for their money. In external funding arrangements such as this, freedom of decision may be sensitively constrained by the good will of funders. The two wider circles in Figure 3 with the catchwords “conditional” and “autonomous” illustrate this dynamic.
In contrast, self-funded settings such as the RSPO or the Gold Standard are principally more independent from such expectations, particularly if they offer their own products, ideas or services that are bought by customers. While this may principally mean a higher degree of freedom, it also means a higher dependency on market and customer demands (Expert 7, 2015). Besides the private market, international agreements such as the Paris Agreement or the Sustainable Development Goals also have the potential to create a demand for more sustainable and climate friendly products and services. To be responsive to both, the Gold Standard, for instance, has developed a third version of its standard and offers the certification of carbon offset projects that produce additional contributions to several Sustainable Development Goals related to biodiversity and poverty reduction [68]. In this way, they attract funding from multiple sources, such as the Luxemburgish Government, the World Bank or Goldman Sachs [68]. Governance settings that obtain donations, such as the Green Belt Movement, can act relatively independently, as long as they act according to their proclaimed vision, mission or values which typically serve as a source of motivation for donors to support them. Here, GBM is communicating its “love for environmental conservation, community empowerment, volunteerism and honesty” [69].

5.3. Strategic Roles in Forest Climate Governance

Strategic roles often defined the terms under which certain actors may take general, strategic decisions to guide the overall setting. They are also pointing at reasons why actors chose or are chosen to fill these roles. We identified four different strategic roles in decision-making among the eight cases: principals, experts, founders/individuals and stakeholders.
(1) Examples for the principal approach where authority derives from a ministerial mandate are the SCF or the PMR. Experienced state delegates involved in international forest climate negotiations come together in a smaller setting and try to synthesize and agree upon the strategic options (in line with ministerial agendas). Another way to take strategic decisions is to have (2) an expert-based advisory board such as in RSPO or the Gold Standard, where board members are elected or nominated based on their experience and the professional guidance they are able to provide. The Gold Standard’s Board, for instance, does not rest in stakeholder membership but relies on experts from various sectors (civil society, business, science). The Board can consult the NGO supporter community if substantive change is desired; however, their decisions are instructed only by the Technical Advisory Committee [61]. A third type of strategic role rests in (3) the individual founders that initiated and spearhead the initiative. They gain authority from the charismatic leadership of the respective founder, such as Wangari Matthai in the Green Belt Movement or, to a smaller extent, the founding member of the Bonn Challenge, who was a renowned forest climate negotiator for many years. The Bonn Challenge also draws on celebrities, such as Bianca Jagger, to inspire further engagement. Particularly for the founding phases, individual leadership and charisma may be central in these cases to convince a critical amount of people about their ideas (Expert 5, 2015, Expert 6, 2016). Inspired by their commitment and visionary goals they may attract new members to follow. This “inspiring” spirit may also matter for decision-making processes beyond founding and enhance the overall acceptance of strategic top-down decisions (Expert 6, 2016). However, fulfilling this role requires consistency between decisions and commonly shared values about forest climate governance. A possible backside of such strategic role is that the setting depends on the permanent presence of the individual person. If they leave, instability of the setting might arise (Expert 6, 2016). Another possibility is that (4) relevant stakeholders exert strategic roles; possibly in conjunction with government representatives, as in the case of the REDD+ Partnership. Deriving legitimacy from the democratic inclusion of all stakeholders is also a concern for RSPO, which assures each of its seven stakeholder groups a proportional number of representatives within its Advisory Board. These Board members are democratically elected through the general assembly of the RSPO [70]. By contrast, the strategic role of stakeholders in the REDD+ Partnership was not fully achieved, mainly because they were excluded from the respective initial meetings where heated discussions among country partners about whether or not to include non-state actors at all or for specific sensitive topics (e.g., finance) were still ongoing [11,62]. Figure 4 visualizes the four role types including the main reason why a role is carried by an actor.
While the role of principals (type 1) is based on ministerial mandates, e.g., high-level public servant, experts (type 2) are elected or nominated because of the professional guidance they are expected to provide, e.g., in their role as technical advisory board members. Founders or individuals (type 3) carry their strategically important role due to their charismatic leadership. Lastly, stakeholders (type 4) themselves may take strategic decisions, e.g., in plenary discussion, mostly for the sake of ensuring that these decisions that concern everybody in the setting are democratic.

5.4. Consequential Roles in Forest Climate Governance

Decisions are consequential if they relate substantially, rather than only procedurally to a policy output, i.e., if they concern specific rules or methods for policy definition, implementation or evaluation. Such decisions may be a result of (1) a participatory process with all members of an organization (e.g., a General Assembly), such as in the cases of the REDD+ Partnership or RSPO, meant to increase the democratic legitimacy of these processes; (2) by scientific or professional expert committees such as in the case of the Technical Advisory Board of the Gold Standard (Expert 7, 2015). Moreover, such consequential decision-making roles may be filled (3) by individuals (esp. founders) such as in the case of the Green Belt Movement which is based on charismatic leadership. A fourth type is 4) decentralized decision-making that draws on a flexible, self-determined and hence potentially highly inclusive bottom-up approach such as in the case of the Bonn Challenge. Here, countries or regions autonomously decide on the ways of designing, implementing and evaluating their reforestation projects (own observations, 2015–2017). Figure 5 provides an overview of the four role types described above.
When assessing the legitimacy of policies, assigning decision-oriented roles to a range of plural, potentially concerned stakeholders—as in type 1—can be critical for increasing the democratic legitimacy of decisions. Although decisions were of a rather information-based nature in the exchange platform REDD+ Partnership, meaningful stakeholder inclusion in meetings was a rule rather than the exception in various events and meetings. This was especially true in the last years of its existence and after the role of stakeholders has been formally accepted by all partners (own observations 2010–2014 [11,71]).
Also within the RSPO, consequential decisions are decided upon in a General Assembly (type 1). However, in terms of representation, economic actors significantly outnumber NGO representatives that make up only three per cent of the ordinary members [70]. Overall, the plurality of views present in the General Assembly (the central decision-making body) paralyzes rather than eases efficient decision-making. In the context of defining sustainability criteria, although principally supported and taken more seriously, the process become languid as discussions were shifted to smaller and potentially more efficient working groups. However, these have not come up with concrete solutions either, even after years (Expert 8, 2016). Because no other voting rules have ever been defined, consensus-based decisions are also the rule in the PMR, where all members jointly take operational decisions (type 1). In fact, in the PMR, a fairly homogenous group of public servants from different ministries meets in a comparably small setting moderated by World Bank staff. Lively discussions typically arise about how to distribute the money; such as whether to spend it on new members or to support existing members a second time (Expert 4, 2016). Despite the consensus rule, donor countries in fact hold a major share in authority, given that they can always bail out if they are not happy with the result (ibid.).

5.5. Expert Roles in Forest Climate Governance

As outlined above, expert roles matter for the output dimension of legitimacy. Despite the predominant model of neutral scientific and technical knowledge, expertise is hardly non-political and ways of selecting, framing or presenting knowledge is shaped by the beliefs and world views of so called experts. In our eight cases, we identified four types of roles concerned with knowledge delivery: (1) Delegates, (2) International Organizations and Bureaucracies, (3) Multiple Stakeholders and (4) Knowledge Partners which will be elaborated on in the following paragraph:
In the first sub-variant, (1) official delegates provide administrative and technical knowledge. Cases such as the REDD+ Partnership or the Standing Committee on Finance demonstrate that expectations about expertise in these settings can mean that only experienced diplomatic experts are eligible to assure political acceptance. Stakeholders within the REDD+ Partnership, for example, need an official accreditation to the adjacent official UNFCCC process as a prerequisite for participating in the discussions. That states’ provision of official data about their territories and natural resources is seen as critical for ensuring political credibility (own observations 2010–2014, [11]). In other cases, expertise carried a more bureaucratic notion. (2) International organizations and bureaucracies which are experienced in diplomacy, such as experts from GPFLR/IUCN, may also be seen as trusted providers of knowledge, e.g., of spatial data for reforestation options or for organizing capacity building workshops as in the case of the Bonn Challenge (Expert 5, 2015) or technical assistance by the World Bank's operational country units in the PMR) [53].
Another possibility is to involve (3) multiple stakeholders purposefully for the sake that they would actually provide a valuable and diverse basis of knowledge which may allow for synergies and knowledge exchange in communication processes (Expert 3, 2014). Such a more practice oriented approach to expertise (“those who do, know best”) that draws on different perspectives and experiences with a specific policy problem is the overall rationale, for instance, in the RSPO, the Gold Standard, the ASEAN CHM or the roundtable meetings of the Bonn Challenge. A fourth variant is more oriented toward pioneers where singular (4) knowledge partners acting as role models by providing technical advice on their leading example and respective experiences. The idea is to stipulate learning from actors that are seen as—presumably—very successfully working on relevant policies or actions. The Forest Stewardship Council is such a partner organization for the Gold Standard in the context of sustainable forestry projects that mitigate carbon emissions (Expert 7, 2015). The state of Quebec and the state of California play such a role within PMR as regions that are early movers in implementing carbon pricing mechanisms on a sub-national level [72]. Expectations behind such an exclusive role target policy diffusion and emulation of specific policy options rather than new (own) approaches or learning from a range of alternative approaches.
Figure 6 displays the broader spectrum of different types of expert roles, visualizing (in bold) who is considered an expert and the respective qualitative features tied to these roles (outer circles). Accordingly, actors may be considered as experts based on their political experience (in italics, type 1), administrative (type 2) technical knowledge (type 2 and 4) or based on the different perspectives they bring (type 3).
Overall, type 1 and 2 are closely related to expectations of receiving “official” and diplomatically valid input which may not be a surprise since, in our cases, the respective organizations also operated on a high political level where political experience or administrative, technical knowledge increased their credibility. By contrast, type 3 and 4 are meant to deliver practical personal or technical experience which is suggested to be rather related to implementation “on the ground” (e.g., from pilot projects or a leading region). Apart from bringing up best practices, this approach, in principle, may also allow critical voices to be heard more directly (e.g., discrimination of social groups or legal obstacles) as compared to type 1 or 2 where such concerns would have to attract the official expert’s attention or “good” will.

6. Discussion and Conclusions

Our paper departed from the observation that polycentrism in the global governance of problems—such as mitigating climate change in forests, where multiple actors take on varied governance functions—poses a severe challenge for evaluating the performance of the respective settings, due to increasingly blurred accountability relations in hybrid actor constellations [10,14,18,73,74,75]. Drawing on dominant institutional and social-constructivist accounts in the legitimacy literature that either stress the role of specific institutional arrangements for achieving input or output legitimacy (e.g., participatory structures) or interpersonal processes of expectations about what activity is acceptable to concerned communities, we proposed the sociological concept of social roles as an analytical vehicle for analyzing governance performance. We systematically investigated eight distinct forest climate governance settings regarding how different potentially authoritative (and hence accountable) roles related to governance purposes such as informing, financing, (co)deciding or helping implementing forest climate action, which appear harmless or neutral at first glance, but may have decisive repercussions on how things are done on the ground [17,22].
We identified several distinct variants within six broader types of relevant actor roles in governance—bureaucratic, operational, financial, strategic and consequential decision-making and expert roles—which have potential implications on the input or output dimension of legitimacy, or both. In our analysis, we compared which different actors filled a specific role and looked at related expectations regarding why exactly specific actors were preferred in a particular role or in the specific ways of engagement (how). Overall, distinct role patterns became visible when distinguishing different sub-variants of actors in the eight cases.
Most prominently, we found several roles across all dimensions that had a fairly formalized notion and respective diplomatic expectations as to what the right actor was and regarding the right way of engaging actors in a setting. This rationale was typical for settings with a high proximity to the politically charged UN process (such as SCF, REDD+ Partnership, PMR, and ASEAN CHM, not always across all dimensions). In the distinct set of official roles (e.g., international bureaucracies or delegates), concerns about diplomatic or technical/scientific correctness were central for the choice of specific actors and modes of engagement. Either internationally experienced actors, such as IOs, were preferred as they are experienced with diplomacy and deemed politically neutral, or officially legitimized actors such as country delegates filled the role, because they represent a country politically—as compared to some randomly assigned stakeholder. The respective approach to specific governance roles in administration, decision-making or knowledge making appeared to often result in an exclusively closed shop; inclusive only for the assorted set of official actors, as in the case of the SCF Forum.
In contrast, alternative rationales for the selection of specific actors or ways of their engagement were guiding the role approaches seen in cases that were not as close to or rather independent from the international climate negotiations under UNFCCC (such as GBM, Gold Standard, RSPO or Bonn Challenge). The respective “casting” rationales behind different roles were often related to ideas about facilitating the process in technical terms, allowing for more efficiency, autonomy, or effective implementation as well as lending a certain democratic quality to the overall operations. Here, one prominent concern was, for instance, to achieve a high level of inclusivity, e.g., by allowing stakeholders to “buy in” to operational tasks (if they could afford it) or by drawing on multiple stakeholders as experts or decision takers (e.g., RSPO, exceptionally also REDD+ Partnership). Giving relevant actors principally the right and opportunities to take on roles in governance settings can help to foster a sense of ownership for the respective decisions tied to that role and hence contribute to a higher actor’s identification with the new setting. Autonomy, which means political independence and potentially allows experimenting with new approaches was another prominent motive for defining and filling roles, for instance, when drawing on rather unconditional schemes of funding (self-funding, donation) and internal administrative staff, or when engaging neutral mediators in the operations of the organization. At least in financial terms, accountability concerns are of minor importance in mainly self-financed settings such as the RSPO [70] or Gold Standard [68], which can be a blessing (for the organization) or curse (for those that wish to hold these actors accountable). In a few instances, these alternative cases drew on fairly exclusive leadership approaches when, for instance, charismatic founders of the organization made major strategic or consequential decisions (as in GBM or the Bonn Challenge) or when representatives from pioneering regions informed the knowledge creation process in the PMR—which from a legitimacy perspective, may require scrutiny.
An interesting deviant for legitimacy considerations is also the spatial distance and dispersion of authority in the case of the Bonn Challenge where implementation-related decisions were actually made outside of the setting and on the ground in the country context, either by national or sub-national entities or private actors. Authoritative discretion and with that respective legitimacy concerns appear deliberately externalized in this setting. The case also illustrates how the mix of different approaches to roles (charismatic leader, professional bureaucratic staff, decentralized decisions) across different dimensions can lead to less frustration and resistance, at least about what is done transnationally between the partners. Nonetheless, such approaches also require a certain level of trust in the implementation commitment of contributing partners.
Moreover, while knowledge production and expertise are theoretically often related to output oriented aspects of legitimacy, expertise rationales in our eight cases carried both input or output-related notions as well as a mix of both. The output dimension is highlighted whenever experts are pinpointed as providing a credible source of technical knowledge (e.g., IOs, pilot/pioneer knowledge). In contrast, an approach that draws on on the ground knowledge from multiple sources stresses the very ways of how alternative views of (local) stakeholders are represented (input). When involving project managers in the GBM or the Gold Standard, the intention appeared to be to satisfy both rationales.
In our smaller sample of eight cases, we could explore how formal (institutional) and interpersonal aspects (mutual expectations) form two complementary parts that are critical for governance roles and their performance. Specifically, our empirical data suggests that performance is at risk whenever the formal (institutions) and informal (expectations) components of roles become incongruent. For instance, in the case of the REDD+ Partnership where rules about working operations (esp. co-chairs roles) were hardly explicit and informal practices contradicted the ideas of the founding fathers (own observation 2010–2014, [11]).
Generally, the question of who, i.e., which concrete actor group, filled a role appeared relevant in several cases, but was far from deterministic. Although the nature of an actor does not a priori predefine performance, it mattered for relevant others who filled a specific role—at least in some cases. Secretariats that draw on staff from international organizations were expected (and evaluated) to better fit the diplomatic exercises and respective confidentiality expectations. This was a concern especially in settings with a high ratio of state representatives. With this focus on (external) actors that fill roles within a setting, the concept covers essential interpersonal linkages between distinct governance settings and allows—to a certain extent—to reconstruct pathways of mutual influence. These external ties were principally possible across all governance dimensions (administration, decisions, finance, expertise) because actors from other institutional realms could be engaged anywhere. Principally, however, our typology mainly focused on processes within a certain setting as on specific types of roles within particular governance dimensions, for which we identified different practical variants and underlying rationales. While we purposefully blinded out how different roles eventually play together, including whether trade-offs or mutual synergies exist or whether, for instance, it is advantageous or disadvantageous to have the same actor to fulfill different roles (e.g., checks and balances concerns), the role concept principally allows for such questions and may support respective analyses. These aspects, we believe, are central and should inform further research on social roles in forest climate governance and respective legitimacy considerations.
Overall, our analysis offers a more systematic account of the variation in different governance roles. Despite our narrow focus, we believe that different subtypes help to display and reflect different rationales or predominant expectations tied to these positions, while also covering institutional aspects of this position, e.g., predominant normative expectations about specific actor types. What the typology itself does not perfectly reflect are the contradictory expectations that are tied to a specific role, something that, however, is revealed empirically in the process of making use of and applying the typology to concrete cases. The same holds true for processes of social change, e.g., as regards the expectations that are tied to specific roles but may change as the actors learn by doing.
The described variety in types of actor roles—in only eight cases—vividly illustrates the many potentially contradictory expectations that governance settings may be confronted with internally and in relation with external actors. In practice, governance operations hardly ever please all expectations, but they may find better or worse strategies to outbalance and minimise the trade-offs between them. The Gold Standard is a telling example for how a setting tries to, and largely succeeded, in pleasing many sides; including expectations from civil society, specific states, markets and the UNFCCC. This case, together with the GBM, strongly deviated from the more state-led processes on almost all dimensions. Especially financial independence lends a fairly high degree of autonomy to such non-state actor based settings and with that power. The question then is, whether with this autonomy a respective demand for more democratic accountability arises or not [76].
In light of the insights deriving from this fairly exploratory study on roles in forest climate governance, which offers indeed no fully fleshed assessment of legitimacy for each governance setting, we believe that social roles may act as a useful analytical concept to draw on. Conceptually, we see it as highly appealing that roles have both institutional and social constructivist components, which allows linking and thus mediating between two valid claims in the legitimacy literature. Methodologically, the concept offers a handy analytical tool for empirically operationalizing governance performance for all sorts of mixed governance settings that integrate non-state actors to varying degrees and in varying positions. In this context, the role perspective has proven fairly compatible with the input and output dimensions of the legitimacy concept and hence with claims that stress the need for democratic legitimacy based on inclusive decision-making, accountability and deliberative processes [77] or that highlight the importance of problem-solving effectiveness [28,32,74] as tied to specific institutional settings and social interactions (expertise) within the setting.
Obviously, a role approach, as such, may not suffice for assessing legitimacy entirely, but the typology may still offer a useful analytical framework that any such legitimacy assessments could draw on and employ for deconstructing a variety of rather complex polycentric governance arrangements in a way that it allows such assessment to be more easily employed. A more nuanced typology, as developed in this paper, does not only illustrate the grand spectrum of possible roles that are more or less authoritative in transnational governance settings, but in research practice may in fact help identifying legitimacy-related issues, e.g., where roles appear harmless at first sight but are revealed as deserving further analytical attention. For instance, it can help to question the influence and the agendas of home organizations regarding ties to specific principals behind actors in administration, decisions or finance that might not be obvious at first. Also, in practical terms, our study results offer some reasonably useful insights. It illustrates that if new governance settings are founded, involved actors often bring in their previously developed concepts about roles (incl. working culture, values and administrative routines) which may not necessarily correspond with what the founders envisioned with the foundation in the first place. To avoid tensions at later phases, and for more clarity, mutual expectations and common conceptualizations of roles should ideally be iteratively discussed and communicated, especially during but also beyond the founding phase of the setting.


This study builds on the work in the research project ‘‘Integration of REDD? into a post-2020 climate agreement and linkages to the CBD’’. This project is financially supported by the German Federal Agency for Nature Conservation (BfN), with funds from the German Federal Ministry for the Environment, Nature Conservation and Nuclear Safety (BMUB). Moreover, Mareike Blum is financially supported by a scholarship of the German Federal Environmental Foundation (DBU). Special gratitude is dedicated to the three anonymous reviewers for very constructive and helpful feedback. The article processing charge was funded by the German Research Foundation (DFG) and the University of Freiburg in the funding programme Open Access Publishing.

Author Contributions

This paper is based on team work in which both authors contributed with fairly equal responsibilities esp. in the theoretical framework, research design and conclusions. Mareike Blum has a major share in drafting the case descriptions and conclusions and in collecting, analyzing and interpreting the empirical data. As corresponding author, she fulfilled major editing tasks. Dr. Sabine Reinecke was pursuing the overall research idea and was responsible for the introduction, respectively. During revision, she took a leading role for aligning theory, empirical insights and drawn conclusions.

Conflicts of Interest

The authors have declared no conflict of interest.


  1. Adger, W.N.; Lorenzoni, I.; O’Brien, K.L. Adapting to Climate Change: Thresholds, Values, Governance; Cambridge University Press: Cambridge, UK, 2009. [Google Scholar]
  2. IPCC. Climate Change 2007. Synthesis Report. Contribution of Working Groups I, II and III to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change; IPCC: Geneva, Switzerland, 2007. [Google Scholar]
  3. UNEP. The Emissions Gap Report 2013; United Nations Environment Programme (UNEP): Nairobi, Kenya, 2013. [Google Scholar]
  4. Stern, N. The Economics of Climate Change. Am. Econ. Rev. 2008, 98, 1–37. [Google Scholar] [CrossRef]
  5. Pan, Y.; Birdsey, R.A.; Fang, J.; Houghton, R.; Kauppi, P.E.; Kurz, W.A.; Ciais, P. A Large and Persistent Carbon Sink in the World’s Forests. Science 2011, 333, 988–993. [Google Scholar] [CrossRef] [PubMed]
  6. IPCC. Climate Change 2014: Synthesis Report. Contribution of Working Groups I, II and III to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change; IPCC: Geneva, Switzerland, 2014. [Google Scholar]
  7. IPCC. Climate Change 2007: The Physical Science Basis. Summary for Policymakers. Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change; IPCC: Geneva, Switzerland, 2007. [Google Scholar]
  8. IPCC. Working Group III Contribution to the Intergovernmental Panel on Climate Change Fourth Assessment Report. Climate Change 2007: Mitigation of Climate Change. Summary for Policymakers; IPCC: Geneva, Switzerland, 2007. [Google Scholar]
  9. PARIS AGREEMENT, 2015. Available online: (accessed on 15 July 2016).
  10. Corbera, E.; Schroeder, H. Governing and implementing REDD+. Environ. Sci. Policy 2011, 14, 89–99. [Google Scholar] [CrossRef]
  11. Reinecke, S.; Pistorius, T.; Pregernig, M. UNFCCC and the REDD+ Partnership from a networked governance perspective. Environ. Sci. Policy 2014, 35, 30–39. [Google Scholar] [CrossRef]
  12. Lemos, C.M.; Agrawal, A. Environmental Governance. Annu. Rev. Environ. Resour. 2006, 31, 297–325. [Google Scholar] [CrossRef]
  13. Metlay, D.; Sarewitz, D. Decision strategies for Addressing Complex, “Messy” Problems. Bridge J. Nat. Acad. Eng. 2012, 42, 6–16. [Google Scholar]
  14. Kjaer, A.M. Governance: Polity; Cambridge University Press: Cambridge, UK, 2004. [Google Scholar]
  15. Marks, G.; Hooghe, L. Contrasting Visions of Multi-Level Governance in Multi-Level Governance; Bache, I., Flinders, M.V., Eds.; Oxford University Press: Oxford, UK, 2004. [Google Scholar]
  16. Ostrom, E. A Multi-Scale Approach to Coping with Climate Change and Other Collective Action Problems. Solutions 2010, 1, 27–36. [Google Scholar]
  17. Börzel, T.; Risse, T. Public-private partnerships: Effective and legitimate tools of international governance? In Complex Sovereignty. Reconstituting Political Authority in the Twenty-First Century; Grande, E., Pauly, L.W., Eds.; University of Toronto Press: Toronto, ON, Canada, 2005. [Google Scholar]
  18. Burris, S.; Kempa, M.; Shearing, C. Changes in Governance: A Cross-Disciplinary Review of Current Scholarship. Akron Law Rev. 2008, 41, 1–66. [Google Scholar]
  19. Bell, S. Do we really need a new “Constructivist Institutionalism” to explain institutional change? Br. J. Political Sci. 2011, 41, 883–906. [Google Scholar] [CrossRef]
  20. Held, D. Regulating Globalization? The Reinvention of Politics. Int. Sociol. 2000, 15, 394–408. [Google Scholar] [CrossRef]
  21. Kooiman, J. Modern Governance: New Government-Society Interactions; Sage: London, UK, 1993. [Google Scholar]
  22. Abbott, K.W. The Transnational Regime Complex for Climate Change. Environ. Plan. C Gov. Policy 2012, 30, 571–590. [Google Scholar] [CrossRef]
  23. Lövbrand, E.; Stripple, J. Making climate change governable: Accounting for carbon as sinks, credits and personal budgets. Crit. Policy Stud. 2011, 5, 187–200. [Google Scholar] [CrossRef]
  24. Miller, P.; Rose, N. Governing the Present; Polity Press: Cambridge, UK, 2008. [Google Scholar]
  25. Stoker, G. Governance as theory: Five propositions. Int. Soc. Sci. J. 1998, 50, 17–28. [Google Scholar] [CrossRef]
  26. Biermann, F.; Siebenhühner, B. Managers of Global Change: The Influence of International Environmental Bureaucracies; MIT Press: London, UK, 2009; pp. 8–9. [Google Scholar]
  27. Scharpf, F.W. Regieren in Europa: Effektiv und demokratisch? Campus Verlag: Frankfurt, NY, USA, 1999. [Google Scholar]
  28. Lederer, M. From CDM to REDD+—What do we know for setting up effective and legitimate carbon governance? Ecol. Econ. 2011, 70, 1900–1907. [Google Scholar] [CrossRef]
  29. Gupta, A.; Mason, M. Transparency in Global Environmental Governance: Critical Perspectives; The MIT Press: Cambridge, MT, USA, 2014. [Google Scholar]
  30. Keohane, R.O. Accountability in World Politics. Scand. Political Stud. 2006, 29, 75–87. [Google Scholar] [CrossRef]
  31. Scharpf, F.W. Problem Solving Effectiveness and Democratic Accountability in the EU; Political Science Series of Max Planck Institute for the Study of Societies: Vienna, Austria, 2006; pp. 1–2. [Google Scholar]
  32. Mayntz, R. Legitimacy and Compliance in Transnational Governance. MPIfG - Max-Planck-Institut für Gesellschaftsforschung. Available online: (accessed on 14 February 2017).
  33. Green, J.F. Rethinking Private Authority: Agents and Entrepreneurs in Global Environmental Governance; Princeton University Press: Princeton, NJ, USA, 2013. [Google Scholar]
  34. Lövbrand, E. Revisiting the politics of expertise in light of the Kyoto negotiations on land use change and forestry. For. Policy Econ. 2009, 11, 404–412. [Google Scholar] [CrossRef]
  35. Steffek, J. Discursive legitimation in environmental governance. For. Policy Econ. 2009, 11, 313–318. [Google Scholar] [CrossRef]
  36. Williams, M. Citizenship as Agency within Communities of Shared Fate. In Unsettled Legitimacy: Political Community, Power, and Authority in a Global Era; Bernstein, S., Coleman, W.D., Eds.; University of British Columbia Press: Vancouver, BC, Canada, 2010; pp. 33–52. [Google Scholar]
  37. Bernstein, S. Legitimacy in Global Environmental Governance. J. Int. L. Int. Relat. 2005, 1, 139–166. [Google Scholar]
  38. Bernstein, S. Legitimacy in intergovernmental and non-state global governance. Rev. Int. Political Econ. 2011, 18, 17–51. [Google Scholar] [CrossRef]
  39. Bernstein, S.; Cashore, B. Can non-state global governance be legitimate? An analytical framework. Regul. Gov. 2007, 1, 347–371. [Google Scholar] [CrossRef]
  40. Bernstein, S.; Cashore, B. Nonstate global governance: Is forest certification a legitimate alternative to a global forest convention? In Hard Choices, Soft Law: Voluntary Standards in Global Trade, Environment and Social Governance; Kirton, J.J., Trebilcock, M.J., Eds.; Routledge: London, UK, 2004; pp. 33–64. [Google Scholar]
  41. Esteban de la Plaza, C.; Visseren-Hamakers, I.J.; de Jong, W. The legitimacy of certification standards in climate change governance. Sustain. Dev. 2014, 22, 420–432. [Google Scholar] [CrossRef] [Green Version]
  42. Biddle, B.J. Recent development in role theory. Ann. Rev. Sociol. 1986, 12, 67–92. [Google Scholar] [CrossRef]
  43. Stout, M. Logics of Legitimacy: Three Traditions of Public Administration Praxis; CRC Press: Boca Raton, FL, USA, 2013. [Google Scholar]
  44. Selden, S.C.; Brewer, G.A.; Brudney, J.L. Reconciling competing values in public administration: Understanding the administrative role concept. Adm. Soc. 1999, 31, 171–204. [Google Scholar] [CrossRef]
  45. Hale, T.N.; Mauzerall, D.L. Thinking globally and acting locally: Can the Johannesburg partnerships coordinate action on sustainable development? J. Environ. Dev. 2004, 13, 220–239. [Google Scholar] [CrossRef]
  46. Bovens, M. Analysing and assessing accountability: A conceptual framework. Eur. Law J. 2007, 13, 447–468. [Google Scholar] [CrossRef]
  47. Degnbol-Martinussen, J.; Engberg-Pedersen, P. Aid: Understanding International Development Cooperation; Zed Books: London, UK, 2003. [Google Scholar]
  48. Goodstein, J.; Boeker, W. Turbulence at the Top: A New Perspective on Governance Structure Changes and Strategic Change. Acad. Manag. J. 1991, 34, 306–330. [Google Scholar] [CrossRef]
  49. Forbes, D.P.; Milliken, F.J. Cognition and Corporate Governance: Understanding Boards of Directors as Strategic Decision-Making Groups. Acad. Manag. Rev. 1999, 24, 489–505. [Google Scholar]
  50. Elbanna, S. Strategic decision-making: Process perspectives. Int. J. Manag. Rev. 2006, 8, 1–20. [Google Scholar] [CrossRef]
  51. Haas, P.M. Introduction: Epistemic communities and international policy coordination. Int. Org. 1992, 46, 1–35. [Google Scholar] [CrossRef]
  52. Moravcsik, A. Is there a “Democratic Deficit” in world politics? A framework for analysis. Gov. Oppos. 2004, 39, 336–363. [Google Scholar] [CrossRef]
  53. Head, B.W. Wicked problems in public policy. Public Policy 2008, 3, 101–118. [Google Scholar]
  54. Cash, D.; Clark, W. From Science to Policy: Assessing the Assessment Process. Faculty Research Working Papers Series (November 2001), John F. Kennedy School of Government, Harvard University. Available online: (accessed on 14 February 2017).
  55. Hunt, J.; Shackley, S. Reconceiving Science and Policy: Academic, Fiducial and Bureaucratic Knowledge. Minerva 1999, 37, 141–164. [Google Scholar] [CrossRef]
  56. Van der Sluijs, J. Uncertainty as a monster in the science-policy interface: Four coping strategies. Water Sci. Technol. 2005, 52, 87–92. [Google Scholar]
  57. Association of Southeast Asian Nations (ASEAN) Clearing House Meachnism. Available online: (accessed on 31 October 2016).
  58. Goehler, D.; Schwaab, J. Managing Knowledge and Regional Policy Advice: ASEAN Forest Clearing House Mechanism. Available online: (accessed on 14 February 2017).
  59. PMR Website. Available online: (accessed on 30 January 2017).
  60. Bonn Challenge. Available online: (accessed on 31 October 2016).
  61. Gold Standard Website. Available online: (accessed on 31 January 2017).
  62. RSPO Website. Available online: (accessed on 31 October 2016).
  63. Karsenty, A. The World Bank’s endeavours to reform the forest concessions’ regime in Central Africa: Lessons from 25 years of efforts. Int. For. Rev. 2016, 18, 1–16. [Google Scholar]
  64. Ongolo, S.; Karsenty, A. The politics of forestland use in a cunning government: Lessons for contemporary forest governance reforms. Int. For. 2015, 17, 195–209. [Google Scholar] [CrossRef]
  65. RSPO Website. Available online: (accessed on 28 January 2017).
  66. Gold Standard Website. Available online: (accessed on 28 January 2017).
  67. United Nations Website. What is Common System? Available online: (accessed on 14 February 2017).
  68. Gold Standard Website 2017. Available online: (accessed on 28 January 2017).
  69. GBM Website 2016. Available online: (accessed on 31 October 2016).
  70. RSPO Website 2017. Available online: (accessed on 29 January 2017).
  71. Pistorius, T.; Reinecke, S. The interim REDD + Partnership: Boost for biodiversity safeguards? For. Policy Econ. 2013, 36, 80–86. [Google Scholar] [CrossRef]
  72. PMR Website 2017. Available online: (accessed on 30 January 2017).
  73. Bäckstrand, K. Multi-stakeholder partnerships for sustainable development: Rethinking legitimacy, accountability and effectiveness. Eur. Environ. 2006, 16, 290–306. [Google Scholar] [CrossRef]
  74. Bäckstrand, K.; Khan, J.; Kronsell, A.; Lövbrand, E. Environmental Politics and Deliberative Democracy: Examining the Promise of New Modes of Governance; Edward Elgar Publishing: Cheltenham, UK, 2010. [Google Scholar]
  75. Hajer, M.A. Policy without polity? Policy analysis and the institutional void. Policy Sci. 2003, 36, 175–195. [Google Scholar] [CrossRef]
  76. Uhlin, A. Democratic Legitimacy of Transnational Actors: Mapping Out the Conceptual Terrain. In Legitimacy Beyond the State? Re-examining the Democratic Credentials of Transnational Actors; Erman, E., Uhlin, A., Eds.; Chippenham and Estbourne: Palgrave, Macmillan, 2010; pp. 16–37. [Google Scholar]
  77. Dingwerth, K. Private Transnational Governance and Its Democratic Legitimacy (Transformations of the State). In The New Transnationalism: Transnational Governance and Democratic Legitimacy; Basingstoke: Palgrave, Macmillan, 2007. [Google Scholar]
Figure 1. Administrative roles.
Figure 1. Administrative roles.
Forests 08 00065 g001
Figure 2. Operational roles.
Figure 2. Operational roles.
Forests 08 00065 g002
Figure 3. Financial roles.
Figure 3. Financial roles.
Forests 08 00065 g003
Figure 4. Strategic decision-making roles.
Figure 4. Strategic decision-making roles.
Forests 08 00065 g004
Figure 5. Consequential roles in decision-making.
Figure 5. Consequential roles in decision-making.
Forests 08 00065 g005
Figure 6. Expert roles.
Figure 6. Expert roles.
Forests 08 00065 g006
Table 1. Interviewed experts.
Table 1. Interviewed experts.
Expert of …YearCode
Standing Committee on Finance (SCF) of the United Nations Framework Convention on Climate Change2015Expert 1
Reducing Emissions from Deforestation and Forest Degradation (REDD+) Partnership2010–2014Own observations
ASEAN Clearing House Mechanism (CHM)2014Expert 3
Partnership for Market Readiness (PMR)2016Expert 4
Bonn Challenge (BC)2015Expert 5
2013–2017Own observations
Green Belt Movement (GBM)2016Expert 6
Gold Standard (GS)2015Expert 7
2015–2016Own observations
Roundtable on Sustainable Palm oil (RSPO)2016Expert 8
Table 2. Features of governance settings.
Table 2. Features of governance settings.
Features →StakeholdersFundingAdministration and Operational Decision-MakingStrategic Decision-Making
Governance Settings ↓Consequential Decision-Making
UNFCCC SCFSCF: States (20 delegates)
SCF Forum: states, civil society, business, finance, science
Conference Of the Parties (COP) members (states)Administration:
UNFCCC Secretariat
State delegates
Operational decision-making:
COP members (state delegates)
REDD+ P.States, civil society, science, businessConference Of the Parties (COP) members (states)Administration: FMT/PT
Facility Management Team (by FCPF, World Bank) and UN-REDD Program Team (by UNDP, UNEP, FAO)
States (ministerial level)
Operational decision-making:
all members
PMRStates, financial institutionWorld BankWorld BankStrategic: World Bank and contributing countries
all members
ASEAN F. CHMStates, business, scienceGerman Agency for Development Cooperation (GIZ)ASEAN SecretaryStrategic
all members
BC(sub)states, business, science, civil society, financial institutionsGermany, IUCNIUCN, GPFLR, WRIStrategic
(implementing) states/sub-states/companies
GBMCivil societyUN funds, donations, partly self-fundedInternal staffStrategic:
GSCivil society, business, scienceTwo-thirds self-funded, one-third national, international and private fundsInternal staffStrategic
Advisory board
Expert Committee
RSPOBusiness, civil society, financial institutionsSelf-funded (member fees)Internal staffStrategic
Advisory board

Share and Cite

MDPI and ACS Style

Blum, M.; Reinecke, S. Towards a Role-Oriented Governance Approach: Insights from Eight Forest Climate Initiatives. Forests 2017, 8, 65.

AMA Style

Blum M, Reinecke S. Towards a Role-Oriented Governance Approach: Insights from Eight Forest Climate Initiatives. Forests. 2017; 8(3):65.

Chicago/Turabian Style

Blum, Mareike, and Sabine Reinecke. 2017. "Towards a Role-Oriented Governance Approach: Insights from Eight Forest Climate Initiatives" Forests 8, no. 3: 65.

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop