The role of multi-stakeholder forums in subnational jurisdictions Framing literature review for in-depth field research

This Framing Literature Review for In-depth Field Research draws on the knowledge produced from 30+ years of experience in participatory processes. It informs the Center for International Forestry Research’s (CIFOR) research of multi-stakeholder forums (MSFs) set up to address land use and land-use change at the subnational level in Brazil, Ethiopia, Indonesia and Peru. This literature review should be read in tandem with the project’s Methods Training Manual for In-depth Field Research.MSFs are set up as purposely organized interactive processes that bring together a range of stakeholders to participate in dialogue, decision making and/or implementation regarding actions to address a common problem or to achieve a goal for their common benefit. The growth of MSFs related to land use/land-use change reflects the awareness that environmental problems cannot be addressed without the effective engagement of the actors that determine land-use practices on the ground; nor can such problems be resolved within a conservation community when the drivers are located in other sectors. MSFs may produce more effective and sustainable outcomes by getting actors with contradictory development priorities to coordinate and align goals through discussion, negotiation and planning. In contrast, MSFs may also help to implement top-down approaches and create the illusion of participation. Scholars and activists note that ‘MSF’ may reify these approaches, and take the ‘participation’ of local stakeholders for granted in box-ticking exercises to please donors.This research is timely, because MSFs have received renewed attention from policy makers and development and conservation practitioners in light of the growing urgency to address climate change and transform development trajectories. The comparative project aims to contribute empirically to the study of MSFs and similar participatory processes.

Multi-stakeholder mechanisms have been hailed as "the collaboration paradigm of the 21st century" (Austin 2000:44).Also known in the scholarly and gray literature as multi-stakeholder forums, platforms, processes, partnerships, and networks, these initiatives are purposely organized interactive processes that bring together a range of stakeholders to participate in dialogue, decision making, and/or implementation to address a common problem or achieve a goal for their common benefit. 1 These initiatives are based on the idea that bringing a wider variety of relevant stakeholders together to make decisions, negotiate, and/or share knowledge will lead to more equitable and efficient outcomes than those arrived at by business-as-usual models of decision making (see, 1 As expected, there are many different ways of defining these processes in the specialized literature that bring to the fore different aspects of the method behind this approach to decision making.Multi-stakeholder networks, for instance, are "voluntary cooperative arrangements between actors from the public, business and civil society that display minimal degree of institutionalization, have common nonhierarchical decision-making structures and address public policy issues" (Streets 2004: 5).Multi-stakeholder platforms are "roundtable [s] where people with very different perspectives are gathered" (Warner 2006: 17).Multi-stakeholder initiatives "involve a variety of actors of diverse character and power involved in a variety of interrelated practices that take place in a variety of sites" (Kohne 2014: 471).e.g.Campbell 1994;Pretty 1995; Buchy and Hoverman 2000;Beierle 2002;Reed 2008).Business-as-usual models are commonly top down, unisectoral, and/or expert driven.
In recent decades, a series of different institutions have been devised and put into practice in order to increase and improve the quality and impact of popular participation and empowerment in decision making and planning (see, e.g.Fung and Wright 2003).Due to these developments, the participation of stakeholders in different strategies for decision making around environmental issues, including multi-stakeholder forums (MSFs), is now widely expected globally.Addressing the shortcomings of business-as-usual approaches, these forums base their claim for legitimacy on their capacity to include the participation (and voices) of a wider range of stakeholders than the usual actors involved in decision making (see, e.g.Berkes et al. 1989;Botchway 2001;Hemmati 2002).The benefits of MSFs, as a method of practice, range from the upholding of rights and participatory democracy, which accepts the key role that local peoples play in the sustainability of policies and projects, to the coordination among different sectors and/or levels of governance (see, e.g., Backstrand 2006;Chatre 2008;Pruitt and Thomas 2007;Reed 2008;Reed et al. 2009;Gambert 2010).

Introduction
A radical alternative or the new 'new tyranny'?
"Any claims that participation can challenge the problems of 'uneven development' must be grounded in evidence and theoretically informed argument rather than in opposition to previously dominant models of development".(Hickey and Mohan 2004: 4) "Communities serve less as decision makers than those consulted, less as regulators than rule-followers, less as licensing authorities than as licenses and less as enforcers than as reporters of offenses to still-dominant Government actors".(Wily 2004: 3) "If you don't sit at the table, you end up on the menu".(Roberto Borrero, International Indian Treaty Council) "To learn, you must participate".(Francisca Arara, Organization of Indigenous Teachers of Acre) These initiatives have received much interdisciplinary attention, leading scholars and practitioners to propose MSFs as a transformative solution for more sustainable decision making in forestry, land use, and climate change interventions.Given this interest in participatory processes, it is important to draw on the lessons from prior experiences and scholarship.The growing mainstreaming of multi-stakeholder processes draws attention to the variety of actors that determine land-use practices on the ground, and two key issues stemming from this.The first is the lack of coordination between these actors.For example, environmental problems or deforestation cannot solely be addressed by the environment sector and instead require effective engagement of agriculture, mining, infrastructure and the other sectors that commonly drive deforestation and degradation.The idea, even urgency, of bringing together multiple actors and sectors -"in order to build a common vision, for a country or for a landscape" (Rodriguez-Ward et al. 2018: 103) -is also central to initiatives such as Reducing Emissions from Deforestation and forest Degradation (REDD+; see Thompson et al. 2011;Fujisaki et al. 2016) and integrated landscape approaches (Denier et al. 2015;Kusters 2015;Minang et al. 2015).Thompson et al. (2011:105), for example, call for "the alignment of the viewpoints and needs of many different actors toward a shared goal of limiting climate change and its human impacts".In similar vein, Reed et al. (2016Reed et al. ( : 2548) ) assert that, "[a] landscape approach must attempt to not only understand the basic needs of local stakeholders but to foster empowerment of community members.By providing local stakeholders an active voice in the design and management of the landscape, it can be determined what people want and expect, rather than what they are prepared to accept".
Then, there is a recognition that a wider scope of stakeholders and sectors must be included in decision-making mechanisms in order to achieve better outcomes than those produced by businessas-usual approaches.
The second key issue is the recognition of power inequalities between the stakeholders and sectors involved.Generally, participatory processes are seen as a way to address power inequalities among stakeholders, to understand the perspectives of those most affected by land-use policy and decisions, and to try to bring on board those with the power to affect the implementation and sustainability of proposed initiatives (see Dougill et al. 2006;Tippett 2007;Reed 2008;Reed et al. 2008).In relation to this recognition of power inequalities, the renewed emphasis on local participation is also a reflection of calls from academia and grassroots organizations for a rights-based approach to development and the concomitant recognition of the link between climate change and human rights that has taken place at the global scale,2 rooted in calls for greater community participation in both conservation and development since the 1980s (e.g.Chambers 1983;Chambers et al. 1989).These critical positions acknowledge how the negative impacts of climate change are more likely to affect local populations that live in precarious environments and who are not historically large contributors to greenhouse emissions.This interest is best illustrated in the Paris Agreement itself, as it calls on its signatory Parties to "respect, promote and consider their respective obligations on human rights"3 when taking action to address climate change.The closer attention to the link between human rights and development has also had an impact on the United Nations Sustainable Development Goals.
However, the optimism brought about by the transformative potential of multi-stakeholder mechanisms has also been met with criticism in the scholarly and gray literature and by grassroots representatives of local communities.In the case of our research, the latter are commonly rights-holders to the spaces that will be impacted by the decisions made by the forums we are interested in.In general, these positions follow the broader critique that describes the participatory development paradigm as a technology of governance that confirms the power inequalities among stakeholders, or between project implementers and beneficiaries (see the contributors to Cooke and Kothari 2001;Cornwall 2003).These critiques also argue that, in practice, these initiatives trivialize local participation, limit opportunities for meaningful public debate, lead to outcomes that confirm unequal development practices, and fail to challenge the status quo (see, among many others, Williams 2004;Hickey and Mohan 2005;Perret and Wilson 2010).
These two main positions on the transformative possibilities of these initiatives reflect the arguments around the wider discussion of the participatory paradigm in international development.These discussions agree in their recognition of different kinds and scales of power inequalities between stakeholders.In the literature (see, e.g.Cornwall 2001Cornwall , 2003;;Chambers 2006;White et al. 2015), these inequalities are based on a series of different characteristics, including political (e.g.power to have the last say), resources (e.g.financial power to get things done), technical (e.g.power to know how to get things done), epistemological (e.g.power to decide on the 'acceptable' knowledge to be implemented in a project) and gender (e.g.power to exclude women from processes).This attention to power and the ability of an MSF's process to address inequalities is especially important when it comes to multi-stakeholder processes designed to address land use and landuse change in many parts of Africa, Asia, and Latin America.These are often set within contexts marked by histories of (post)colonialism, conflict, and deep inequalities.We recognize these fields (analytically and in practice) as multilevel and multisectoral (Pierre and Peters 2000;Termeer et al. 2010), where trade-offs are inherent, and the actors involved are significantly diverse (Robbins 2012;Barnes and Child 2014). 4Yet, these positions diverge in their assessments of whether participatory mechanisms can address these inequalities by leveling the playing field and changing the status quo.To summarize these positions, the discussion becomes one about whether participatory mechanisms empower or disempower local actors, stakeholders, or project beneficiaries (in practice, the term used depends on how an initiative and those that coordinate it imagine 'locals' and present them discursively).
Considering this evidence, and the growing resources for and expectations placed on MSFs, the Center for International Forestry Research (CIFOR) has addressed the pressing need for on-the-ground, comparative research on the process and outcomestwo distinct aspects -of such forums through a global comparative study of 13 subnational MSFs in Brazil (3 case studies), Ethiopia (2), Indonesia (4), and Peru (4).Through this study, we aim to analyze the equity and effectiveness of MSFs, and how their processes and outcomes are shaped by the beliefs and actions of those who design these processes and are involved in them.
We examine, among other things, the processes used to select participants, and how they are included (e.g. in discussions, decision making, implementation, monitoring); how decisions are made on meaningful matters (which key issues are and are not included in the processes); and the degree to which decisions are binding or enforceable.
We also analyze the relations among participants and their perceptions of the process and its outcomes, both in terms of equity and land-use practices.We have explicitly chosen MSFs at the subnational level because they are closer to the geographical spaces of land-use planning and management, and due to the current interest in jurisdictional approaches to tackle climate change and deforestation (see, e.g., Fishman et al. 2017).At the same time, this scale will allow us to explore how strategies of global environmental governance are pursued and reshaped through the 'friction' of local encounters (Tsing 2005).
Our working definition for MSFs views them as organized interactive processes that bring together a range of stakeholders to participate in dialogue, decision making, and/or implementation to address a common problem or achieve a goal for their common benefit.Faysse and colleagues suggest that MSFs tend to have a generic objective defined as: "To enable the empowered and active participation of stakeholders in the search for solutions to a common problem" (Faysse et al. 2006: 220).For them, there are usually two main expectations behind a decision to set up one of these processes.
The first, and perhaps the most common, is that MSFs are expected to lead to outcomes or decisions that are more widely acceptable, for stakeholders in general, in comparison to decisions resulting from business-as-usual processes with no or little stakeholder participation.The second is that they lead to better and more acceptable decisions than those arising from one-to-one negotiations.In general, Faysse et al.'s outlook is positive, arguing that MSFs should be analyzed as an on-going negotiation process, one that is always imperfect, but where positive outcomes may nevertheless outweigh negative ones. 5he "empowered and active participation of stakeholders" that Faysse et al. (2006: 220) attribute to multi-stakeholder initiatives refers to Arnstein's (1969) ladder of participation (see Figure 1 below).Arnstein's is the most cited representation of the different degrees of stakeholder participation to an issue.Power and power relations are central to his ladder, as power is at the center of citizens' ability to participate in decision-making spaces: "the ladder juxtaposes powerless citizens with the powerful in order to highlight the fundamental divisions between them" (1969: 217).Each step on the ladder represents increased power for disempowered citizens in terms of their decision-making ability.
The optimism on the possibilities for transformation brought by MSFs permeates most of this literature, as well as the guides for multistakeholder processes that have been developed by scholarly and non-governmental bodies (e.g.The MSP Guide 6 ).The majority of these positions move on from the perspectives set out in the critique of participatory development to tease out the transformative potential that MSFs could have in the right situations.For example, Brouwer and Woodhill (2015)  And why are we interested in them?
The role of multi-stakeholder forums in subnational jurisdictions Although these ideal characteristics are useful to understand the possibilities brought about by MSFs, it is as important, or perhaps more so for our purposes, to understand the factors and dynamics that interfere with or even prevent such ideal functioning.For example, should the definition of the problem or opportunity the MSF engages with not be part of the process rather than its starting point?We could argue that it is often when stakeholders do not hold a shared vision that a process with equitable power distribution or special attention to procedural justice is necessary.
Setting out the definition of the problem or opportunity as a starting point might also lead organizers to include stakeholders that already hold similar points of view.

Why is multisectoral and multiactor coordination so desirable (and difficult)?
The variety of stakeholders invested in land-use outcomes suggests that the transition to more sustainable land use proposed by MSFs requires a strategy of coordination across both sectors and scales (see Larson et al. 2018

Figure 1. Arnstein's Ladder of Participation
Source: www.citizenshandbook.org 8 Citizen Control.Have-nots handle the entire job of planning, policy making and managing a programme e.g.neighbourhood corporation with no intermediaries between it and the source of funds.
7 Delegation.Citizens holding a clear majority of seats on committees with delegated powers to make decisions.Public now has the power to assure accountability of the programme to them.
6 Partnership.Power is in fact redistributed through negotiation between citizens and power holders.Planning and decision-making responsibilities are shared e.g. through joint committees.5 Placation.For example, co-option of hand-picked 'worthies' onto committees.It allows citizens to advise or plan ad infinitum but retains for power holders the right to judge the legitimacy or feasibility of the advice.
Scholarly discussions on the participatory paradigm hinge on whether or not participation has the potential to transform mainstream approaches (Chambers 1983;Chambers et al. 1989).These approaches, often referred to as 'business-as-usual', are commonly top-down, unisectoral, and expert driven.Analysts on both sides of the discussion acknowledge the problematic nature of power inequalities in business-as-usual approaches, but diverge on whether participatory processes, such as MSFs, can transform them.One position highlights the potential for more horizontal decision-making processes, with more equitable and effective outcomes for local populations (Sayer et al. 2013;Estrada-Carmona et al. 2014;and Bastos Lima et al. 2017  These positions link stakeholder participation to positive outcomes ranging from the normative (as an ideology) -including the upholding of rights, justice, and participatory democracy-to the pragmatic (as a method), such as the proposition that stakeholder participation leads to more sustainable and cost-efficient initiatives with more local ownership (Buchy and Hoverman 2000;Hemmati 2002;Reed 2008).
The other strand of this literature concentrates on the challenges to coordination processes presented by the kinds of conflicts that take place among land use and land-use change stakeholders.From this perspective, one of the most important challenges faced by coordination processes is in the different (and at times incommensurable) development priorities held by different actors across sectors and levels.Generally, studies show a complex struggle around REDD+ policy making, which we could extend to land-use governance more generally, where the most economically powerful actors tend to win (see, e.g.Korkonen-Kurki et al. 2015;Ravikumar et al. 2015).We propose that this recognition of power inequalities and the political dynamics underlying land-use decision making and driving businessas-usual practices are key to understanding how coordination processes can be made to lead to more equitable and sustainable outcomes.A clear example of these inequalities in multisectoral coordination is in the power disparities held by different ministries or subnational offices, as those that oversee land or development schemes (e.g.finance, agriculture, mining) tend to have more resources and decision-making power than those that seek to address deforestation and the rights of indigenous/local peoples (e.g.environment, culture; see Larson et al. 2018 andRavikumar et al. 2018 on the shortcomings of, and potential solutions for, mainstream collaboration approaches).Recognizing how power may allow certain actors to reify business-as-usual development practices is key to addressing the shortcomings faced by collaboration approaches.
Equally important to our current research is to inform how coordination processes may be set up to have more equitable and sustainable outcomes.
Our Realist Synthesis Review of the literature on MSFs (Sarmiento Barletti et al. n.d.) suggests that recognizing the power inequalities between the different sectors and stakeholders in a coordination process is a necessary first step to improving such processes.Our review proposed three key findings to inform actors setting up or participating in MSF-like processes.Firstly, that processes should be designed to be adaptive to the contexts that give rise to the issues the MSF addresses and that affect the stakeholders involved.Secondly, that MSFs are more likely to transform business-as-usual decision making if they recognize power differences between different stakeholders by setting up a procedurally just approach.Thirdly, that MSFs are more likely to be transformative if they have a working understanding of equity as a combination of access to both material resources and rights.
The need for an attention to these findings is also illustrated in the results of a study of MSFs in the Peruvian Amazon that was carried out prior to the in-depth research that was informed by this document.Sarmiento Barletti and Larson (2019) found a link between the ineffectiveness of collaborative processes and inequity in the context where the process sits.The study reveals that in San Martin and Madre de Dios regions, the only MSFs that reached an agreement were those that did not counter the development priorities of the most  (Ravikumar et al. 2018).
We transition to the next section, on the wider benefits and challenges of participation, by proposing that one of the key challenges of coordination processes is that they tend to take participation itself for granted.There are two key issues to consider in this regard, which we address below.The first is that we need to understand who is coordinating and to what end.For example, Ravikumar et al. (2018) found that the actors driving deforestation and unsustainable land use (including agriculture and mining offices, private companies, and local elites) coordinate effectively and are able to strategically push coalitions for conservation and rights out of the picture.The second challenge is that coordination as a solution, at least in theory, must be set on the recognition that not all participants are equal.There are important differences to be considered between actors, such as who convenes the process, who funds it, what kind of participation in decision making is available to different participants, and who is not taking part in the process (and why).
We argue that treating all participants as if they had the same access to informing the outcome of a collaboration process may lead to the reification or exacerbation of the existing power inequalities underlying the structures upon which inequality and injustice are constructed.Thus, we recognize that the problem is not merely 'more coordination', but also 'better' coordination, and are setting out to find what this entails in different contexts.

What are the benefits and challenges of participatory processes?
Table 1 summarizes the benefits and challenges of participatory processes, as synthesized from different positions in the literature.
We continue this discussion in the section below, but will look more explicitly at how power and the disparate power relations between participants to an MSF may affect its outcome and process in Section 3.

What are stakeholders, and how are their interests managed by an MSF?
While the necessity for inclusive decision making is widely accepted, agreement over who or what exactly stakeholders are is not as straightforward.
The different positions stem from Freeman's (1984) distinction between those who affect or are affected by a decision or action, a distinction that Grimble and Wellard (1997) later described as 'active' and 'passive' stakeholders.Within this distinction, it is also possible to distinguish between those stakeholders who are directly or indirectly affected by an issue (see Coase 1960).For the purpose of our research, we follow Hemmati's (2002: 2) definition of stakeholders as "those who have an interest in a particular decision, either as individuals or representatives of a group.This includes people who influence a decision, or can influence it, as well as those affected by it".The literature also presents an interesting discussion on what constitutes a legitimate stake on an issue, as at times there is an implicit assumption about the legitimacy of stakeholders (and thus the legitimacy they bring to a multi-stakeholder process) without thinking critically through their different experiences of the issues that an MSF may address (see Friedman and Miles 2002 for a critique). 8In this regard, the most important distinction that we explore is that between the rights that different stakeholders have over the matter at hand.Generally, in the context of our work the term rights-holders commonly applies to indigenous peoples, who have had certain rights 8 Participation/inclusiveness is one of Dingwerth's (2007) three sources of democratic legitimacy (democratic control and discursive quality are the other two).Yet, Dingwerth argues, inclusiveness depends on how stakeholders are defined.He writes that although this depends on the 'legitimacy' of the actors themselves (see also Ottaway 2001), they are "usually limited to those actors deemed necessary for the success of each initiative" (2007: 194).At the same time, linking democracy and participation "at this local or grassroots level raises fundamental and normative questions about the nature of democracy and about the skills and strategies for achieving it" (Gaventa 2004: 162).In an important sense, this can be related to how people perceive the state in the positive (transparent, accountable, supportive) or negative (corrupt, unaccountable, oppressive).

Table 1. Benefits and challenges of participatory processes Benefits Challenges
• Channel for direct participation by different stakeholders • An alternative to state-driven processes for input, conflict and collaboration • Bring people together who might otherwise not have collaborated or provided input • Create opportunities for different groups to learn about each other, communicate, build relationships and trust • Can create a more level playing field for disadvantaged groups • Can shift power to local or previously marginalized groups • Do not assume win-win outcomes and are more explicit about winners and losers • More realistic about time required to bring people together and to reach agreement • Bring more diverse viewpoints and skills that produce synergies and enhance capacities to innovate and cope with complex environments • Allow networking between underrepresented groups and more powerful allies • Allow access to spaces of discussion with more powerful actors that underrepresented actors can use to raise issues out of the MSF's mandate • Foster multisectoral collaboration • Difficult to know people's interests • Much depends on the nature of the convener and facilitator • Rarely have a sustainable institutional base • Create an artificial context that may not persist after they end • Representatives of interest groups may not be accountable to a constituency • Not necessarily legitimate or accepted by authorities • Lack the checks and balances and accountability measures of public decision-making processes • Have many aspects that cannot all be handled at once • Not all stakeholders participate • Transaction costs can be high • Where large numbers participate, in-depth discussion and debate of complex ideas may be difficult • Can give the impression ideas are only legitimate when approved by all stakeholders • Difficulty in getting and retaining the input of stakeholders who really matter At the same time, we must be aware of the 'local' analytical trap, related to the longer discussion in international development about the pointlessness of understanding a 'community' (or even a 'household') as having a joint perspective on any single issue.In general, we note that while all local people take an active part in place-making and pursuing their well-being, not all of them are committed to the same agenda (or to the sustainability of the same place).Power relations are key here as they are the (sometimes explicit, sometimes implicit) proof that places are not homogeneous or harmonious entities.For example, feminist critiques of development noted early on the male bias in participatory processes, as women were either not included in participatory spaces, or were deemed to be 'free labor' (as social capital) to be applied in the work that came with the implementation of their outcomes (what Ribot (1996) called participatory corvée).In general, there was little effort placed toward understanding the intersectionality of gender, as women were understood as a unitary category (see critiques by Cornwall 2001Cornwall , 2003;;Wilson 2015).
This kind of gender-based exclusion has also been noted by students of wider participatory processes.Miller (2001), for example, engages with the idea of the empowerment of women through political practice in democratic contexts.Writing on women's movements, she notes how the reified social structures that inform (and are informed by) how societies construct gender can be "the densest and most elusive obstacle" to (women's) participation.Teasing out the impact of these exclusionary practices, the review of MSFs carried out by Sarmiento Barletti et al. (n.d.) revealed that a common contextual factor that prevented an MSF from reaching its desired outcome was that those forums did not pay specific attention to the experiences of women and the kind of socio-political and/or legal institutions that often prevent them from having a greater say in participatory processes.
So, while bringing different stakeholders and sectors together over an issue is definitely laudable, we must consider the different stakes they hold over the issue at hand, as well as the power inequalities between them (see Section 3).Although the literature does not say much about the potentially 'negative' aspects of broader stakeholder participation, interviewees in our preliminary research with MSF conveners, mention the time-consuming and more expensive aspects of these modes of decision making and/or coordination.Indigenous respondents, although this claim could be extended to other underrepresented groups, highlighted the danger that can come with participation, as their presence may be used to validate the MSF's outcome by more powerful participants with clashing agendas.In these cases, pressure from other stakeholders stemming from power inequalities, the use of majority votes to decide on outcomes, or lack of capacities to participate in, for example, technical forums, coerce underrepresented groups to accept outcomes that may not benefit them.In an important sense, this makes us aware of the possibility of MSFs as a method of decision making that can turn rights-holders into stakeholders.
Bringing everyone together at the table and granting them the same say suggests they all have the same rights to the issue at hand, potentially ignoring the special rights indigenous peoples have had recognized by international agreements, over private actors with different claims to land and land-use practices.However, indigenous respondents also noted that participation in an MSF that has government or private sector presence can empower them to make demands on topics that are unrelated to the MSF but that they have no other spaces available to discuss.As might be expected, some forum organizers interviewed during our preliminary research noted that this was a negative aspect of broadening participation to those they considered to be 'non-experts'.Table 2 outlines some of the key benefits and some related challenges of stakeholder participation, as set out in the literature.
Then, the benefits of the wider scope of stakeholders brought together by participatory processes like MSFs, in comparison to businessas-usual decision-making processes, are generally agreed upon in the literature.The same is true of the recognition of the imbalances in power relations between participant stakeholders.
Recognizing the potential difficulties of the variety of interests that come together in an MSF, Wollenberg et al. (2005: 47-48) outline five relevant features to understand how multistakeholder processes accommodate, coordinate and manage these interests (see Table 3).
We now move on to discuss the underlying role of power and power inequalities in the contexts that give rise to MSFs and in which they operate and produce outcomes.

Improved quality of outcomes
More perspectives lead to a more complete overview of the issue at hand and thus of solutions or outcomes of an improved quality (Woodhill and Roling 1998;Berkes 1999;Olsson et al. 2004).Generally, the literature considers local inputs as leading to better results and sustainability (cf.Arheimer et al. 2004).It is worth noting that some positions warn that more perspectives may not always lead to better outcomes because of the interaction of competing interests in a forum (Brody 2003;Connelly et al. 2006).

Insight into values that cannot be gained through technical approaches
The participation of non-technical 'experts' provides insight into knowledge that goes further than 'science' (Middendorf and Busch 1997).This is not uncontested, however, as some scholars question the usefulness of local knowledge in contemporary discussions (Krupnik and Jolly 2002;Doolittle 2003;Briggs and Sharp 2004).This goes back to the older discussion of the role of indigenous/traditional knowledge in development, such as Sillitoe's (1998) seminal work on the topic.

Legitimacy and democratic ideals
As top-down approaches to decision making go against democratic ideals, proponents of participatory processes highlight their potential to uphold such ideals by granting people more control over the initiatives that affect their lives (Colfer et al. 1995;Colfer 2005).In doing so, local participation also grants legitimacy to decision-making processes.

Achieving political goals
Political goals such as the empowerment of underrepresented groups may be used to justify participation.Participation is also applied to issues that need consensually agreed targets (Arheimer et al. 2004) or when the government needs access to relevant information held by specific groups (Geurts and Mayer 1996).This does not mean that the participatory process will be fair or follow some kind of procedural justice, or that it will not be used by the government as proof of consultation to justify policies.Conversely, these spaces could also be used by local representatives to make unrelated claims that fit with their agendas and priorities.

Social learning
By encouraging stakeholders to work together, MSFs can foster social learning.This can transform relationships, change people's perceptions of each other's positions and demands, and thus enable them to identify new ways of working together and/or reaching a more satisfactory outcome (McDougall et al. 2008;Akpo et al. 2014).

Characteristics of the stakeholders, conveners, and facilitators
• How were the participants to the MSF selected?
• Who is the forum accountable to?
• How legitimate is their participation?
• What kind of influence do they have over decision making both inside and outside the MSF?
• Do they all have the same stakes over the issues discussed at the MSF?

Context that frames the process
• What is the history behind the different demands held by forum participants?
• What conflicts have defined or keep defining their interactions?
• What factors beyond the boundary of the forest or beyond the control of the stakeholders affect the process?

Shared principles and strategies guiding decisions
• What are the principles for making decisions in the MSF?
• What shared/clashing visions or interests do they hold?
• How do stakeholders agree on what roles each of them will hold?
• Are all stakeholders equally committed to the MSF?

Cycles of conflict and cooperation
• Did the MSF have a built-in conflict management mechanism?
• How were differences negotiated or bargained?
• Was the MSF's process consciously formulated toward reaching cooperation and agreement?

Mutual adjustments
• Did the process allow stakeholders to adjust their positions as the MSF ran?
• Was the MSF set up to be adaptive, or did it strictly follow its original focus in the way set out by its organizers?
Source: Wollenberg et al. ( 2005) The generally positive outlook on the potential of MSFs that we explored above must be understood within the context from which multistakeholder processes arose.In an important sense, MSFs follow in the footsteps, and attempt to address the shortcomings, of the participatory paradigm to development that was introduced as a transformative decision-making alternative in the 1980s.The proposition of the benefits of participatory planning and decision making is not new and has been part of lengthy debates in multi-disciplinary scholarly and practitioner discussions in international development about the participatory paradigm.We have summarized these positions above, but now move on to a more in-depth engagement with that literature.Following the critique of development as an imposition of foreign paradigms (e.g.modernity, progress, improvement, and, more recently, wellbeing) on local communities, a rich corpus of texts has proposed participatory research methods and decision-making spaces as a way to include 'project beneficiaries' in the process through which projects are designed and implemented (e.g.Chambers et al. 1989).We understand MSFs as the new iteration of this participatory paradigm.
Yet, this assessment of the potential of MSFs to make decision-making processes more democratic and horizontal may prove to be too optimistic.This is especially so in light of the critique that followed the introduction of participatory practices in development.
Perhaps most prominently proposed in Cooke and Kothari's Participation: The New Tyranny?(2001), analysts have engaged with participatory approaches as a white-washed (or even greenwashed in environment-related cases) version of former practices.Analysts following this line of criticism have often described participatory approaches as neo-colonial impositions on local populations by previous analysts (see, e.g., Ferguson 1990;Escobar 1995).For Cooke and Kothari, and the contributors to their edited volume, the participatory paradigm signified a new representation of control over local peoples in their new guise as project beneficiaries.From this perspective, assemblages such as MSFs are a waste of resources that only serve to confirm mainstream governance and business practices, and the discourses that discriminated against local knowledge (or traditional/indigenous knowledge in the scholarly debates) in favor of scientific/ technical knowledge (see Sillitoe 1998).
Contemporary proponents of participatory processes claim that this critique only addresses the mainstream form of the paradigm, which has been applied to support rather than subvert business-as-usual development. 9Still, due to this common application of the participatory paradigm, the critique of the participatory 9 For example, organizations that commonly set up top-down practices seldom have formal training in participatory practices (Campbell 1994;Stenseke 2009; Westberg et al. 2010).

Power and empowerment in participatory decision making
"[Power is often] associated with […] authoritarianism, bossing, control, discipline, domination -and that only reaches 'd' in an alphabetical listing.In these negative usages, power is abused and exploited.All power corrupts.All power deceives.Bad people are power hungry […] and use it for their own ends."(Chambers 2006: 100) "[Power is not] static or monolithic.There is a continuous process of resistance and challenge by the less powerful and marginalized sections of society, resulting in various degrees of change in the structure of power.When these challenges become strong and extensive enough, they can result in the total transformation of a power structure."(ASPBAE 1993).
process in development can be summarized into four key points: 1.From its setup, it carries an insufficiently sophisticated understanding of how power is constituted and operates, and thus of how empowerment may occur (Mosse 1994;Kothari 2001).2. It is obsessed with the 'local' as opposed to wider structures of injustice and oppression (Mohan and Stokke 2000).

It is formed around an inadequate
understanding of the role of structure and agency in social change (Cleaver 1999).4.There is a tendency for certain agents of participatory development to treat participation as a technical method of project work rather than as a political methodology of empowerment (Rahman 1995;Carmen 1996;Cleaver 1999).
This demonstrates that "participatory development has often failed to engage with issues of power and politics and has become a technical approach to development that, in various ways, depoliticizes what should be an explicitly political process" (Hickey and Mohan 2004: 4).
Although we will not expand further on this here, it is worth noting that the link between power and knowledge has been explored by students of international development based on the work of Michel Foucault (e.g. 1980).This has perhaps most notably been done by anthropologists Arturo Escobar (1995) Namely, that the making of discussions at MSFs into 'technical' ones as a way of making them unbiased, moves them away from the contextual issues that may create inequalities in the first place, and de-politicizes issues that may have profound structural (and thus political) origins.In fact, when taken to extremes, we could argue that multi-stakeholder participation itself is a technology of governance that is imposed on local peoples, limiting their ability to take social collective action, and the government's role in regulating the private sector or simply enforcing its own laws.For example, Cooke and Kothari and the contributors to their seminal edited volume (2001) propose that the imposition of participatory mechanisms becomes a sort of 'tyranny', an externally driven discourse that does not necessarily represent local experiences or knowledge, and that never extends far enough to address the structures on which power inequalities sit.In similar vein, Agrawal (2005) extends Foucault's 'governmentality' as 'environmentality' as an analytical tool to think through how the actors leading conservation and development initiatives may use participation in order to justify initiatives following their own goals and interests.The awareness of these different possibilities, which may happen in parallel or may have been replaced by an MSF, should make us mindful of the need to be conscious of the different forms of local political action that occur in the contexts we engage with in our research, and how these are related to wider concepts and practices of participation.
Thinking of MSFs as an imposition resonates with Carothers' (1999) assertion that democratic and governance initiatives sponsored by Euro-American donors or NGOs tend to carry a highly romanticized discursive take on democracy.With this comes little understanding of the complex local dynamics of power set out in the national (and even less so in the local) contexts in which they are implemented.This is reminiscent of Massey (1995: 286; see also Massey 1994) who points out that "[p]olitical subjects are indeed constituted in political practice, but they are not constructed out of nothing".The specificity of socio-political life in different contexts (which is not to say that there might not be similarities between contexts) are the result of the historical construction (and thus shaped by prior socio-political processes) of place-based identities that may themselves be shaped by forces that originate out of that space.Then, research must recognize the historical and structural qualities of these identities, and thus also the source of the kind of power relations through which different actors engage with each other, as a potential pathway to transform the power relations that support exclusion and inequality.Following Partzsch (2016: 193), we agree that "[s] erious reform is inextricably tied up with questions The role of multi-stakeholder forums in subnational jurisdictions of power, understood here as conscious, structural mechanisms that shape decisions affecting a collective environment".
One of the ways through which power inequalities may become more obvious in the context of MSFs is in the difference between some stakeholders' ability to participate physically and verbally yet nominally (as mere headcount/token presence), and the ability of others to participate meaningfully with powersharing in decision making.Although not explicitly about multi-stakeholder mechanisms, this focus on justice is related to what students of wider participatory governance have noted as the possibilities it carries for the empowerment of local peoples. 11 Ribot's (1999: 49) assertion that "[e]mpowering indigenous authorities does not automatically resolve issues of equity, representations and accountability, nor does it constitute community participation" is important here.For Ribot, participation must be tied to significant devolution of power; that is, powersharing in decision making (see also Donnelly-Roark 2015).From this perspective, participation cannot be empowering or transformational if people are unable "to shape the policies that govern them" (Ribot 2013: 93).This is also related to what Fraser (2005) calls 'participatory parity', in which all citizens have equal opportunities to participate in democratic politics and are able to shape "the fate of the polity in which they are involved" (Isin and Turner 2002: 4).

Participation and power
The positions that we summarized above show that 'participation' occupies a complex continuum of characteristics.We are well aware that meaningful participation is conditioned by power contests and relations that are framed by a series of socio-cultural, political, economic, legal, and historical variables that determine the composition, interactions, procedure, outcome, and impact of an MSF.In this section, we set out a discussion of the different strands of the scholarly literature on power to help us think through these issues.Following Partzsch (2016), we recognize the multidimensional nature of power as central to any understanding of the ability (or not) of MSFs to produce transformational change.In so doing, our analysis not only differentiates between sources of power, but also between the different mechanisms through which power is exercised.We begin with a deep awareness of the necessity of understanding how power is exercised, and thus the subtleties and dynamism of power relations.
Conceptual discussions of power in the social sciences are broad and follow a wide range of analytical positions.Yet, within that difference, if we simplify their arguments, we can find two main trends in how social scientists have approached the study of power, one that cognizes it as domination, and another that regards it in terms of empowerment and agency.Partzsch (2016), bringing together previous discussions on the topic, distinguishes these as three types: power over, power to, and power with.From her perspective, power over is about coercion and manipulation that leads to a result where there are winners and losers; power to deals with processes of resistance and empowerment that allow actors to 'get things done' to reach alternatives to business-as-usual practices; and power with involves learning and cooperation that lead to win-win situations.These distinctions work well as analytical ideal types through which to understand the different mechanisms of the exercise of power through a prism rather than a single lens.These types, to which we will return below, are helpful as they allow for a better understanding of the relationship between mechanisms of power and the results of power, but also the key relationship between actors and structures in these processes.It is worth remembering that, as with all analytical types, more than one may be at play in a single MSF, and thus we must also consider the relationship between them.These ideal types allow us to think of power as more than coercion, and to look at how collaborative action and individual agency can transform business-as-usual decision making around land use and land-use change.
Partzsch argues that, when it comes to analyses of environmental policies, scholars have commonly understood power relations in terms of power over, that is, the ability to coerce and manipulate in order to influence decision-making processes (see also Allen 1998;Haugaard 2011).For Partzsch, there are four different dimensions of power over, which we summarize in Table 4.In general, these dimensions tend to be influenced by Max Weber's (1947) seminal definition of power as "the ability of an individual or group to achieve their own goals or aims when others are trying to prevent them from realizing them".Thus, this type understands power in zero-sum terms, as an actor can only have power if it can prevent others from having it, and thus the powerful can always have their way (see Rowlands 1997;Chambers 2006;White et al. 2015).Power over, then, is about control and it is relative: "one actor's increase in power is another's decrease" (Fuchs et al. 2016: 4).From this perspective, analysis is about finding who has power, where and how they exert it, upon whom, and for what end.The following distinction by Sadan (1997) of how power is exerted can help us think through these issues.The first is the most straightforward and overt, in the outright exclusion of some actors from a participatory process.The second is about the covert face of power, of the ability of actors to control a participatory process by setting agendas and thus preventing the discussion of some issues that do not fit (or may counter) their own priorities.The third and final one is the most hidden one; the one through which actors are able to set their interests behind the whole participatory process and the wider context it addresses.Now, this does not mean there is only one kind of power that is exerted in different ways.Escobar's (2010) distinction between four dimensions of power is helpful here: physical (force to coerce), political-economic (authority and control of resources), discursive and symbolic (the creation, naturalization, and application of discourses and symbols), and ontological (who is behind the governance system that structures us).
The analytical problem with the power over type is that it tends to be equated with agency: those who have power have agency, those who do not have power do not have agency.However, power is not always deployed in a unilinear manner, or imagined as being so negative (or asocial).Partzsch also argues that, either explicitly or implicitly, another analytical position in the field of environmental politics is based on an understanding of power with, in a positive sense that "implies learning processes that allow actors to question self-perceptions and to actively build up new awareness of individuals or groups" (Partzsch 2016: 195; see also Eyben et al. 2006;Gaard 2010).This is a rehashing of power that is not static or centered on specific actors, but highlights its relationality and contested nature.Importantly, it recognizes its positive potential, especially in terms of collective action and agency (see VeneKlasen and Miller 2007;White et al. 2015; and the wider literature on social movements).From this perspective, Partzsch argues, actors can create transformational change if they work together.Partzsch (2016: 196)

expands, "when
The role of multi-stakeholder forums in subnational jurisdictions | 15 scholars share this perspective of power […] they do not think in antagonisms such as top-bottom and winner-loser [but] understand power […] as serving the common good (i.e.environmental protection, planetary stewardship).[…] Structures are irrelevant and can be overcome, if people take coaction".Thus, proponents of this analytical stance hold the perspective that gridlock in environmental politics is incomprehensible.Taking this stance to an extreme, Partzsch (2016: 196) summarizes that from this viewpoint, "[a]s there are no subordinates from this power perspective, no imperative to resist follows".If we approach this stance critically, however, we could argue that it would miss the strategies chosen by groups that refuse to participate in an MSF because of their own recognition that participation may not serve their own strategic purposes (see Edmunds and Wollenberg 2001;Faysee et al. 2006).That is, this viewpoint often overlooks the real place of unequal power relations between MSF participants, and how power may structure how stakeholders engage with each other outside the MSF.
Partzsch also guides us to a third analytical outlook on power, that of power to.As Partzsch explains, this perspective carries an outlook of productive agency that seeks to understand not the conflict at hand in the contestation of power, but the alternative (and potentially transformational) ideas and values that may change the state of affairs.This relates to Parsons' (1963) definition of power as the ability to get things done despite structural constraints.Comparing power with and power to, Partzsch (2016: 199) notes that while the former "implies a strong notion of collective accountability for nonaction, power to emphasizes the ability of individuals and groups to get thing done on their own".For Partzsch, this perspective involves an accountability for non-action, and making sense of gridlock in environmental conflicts as structural constraints or a lack of creative thinking.The confrontation between Greenpeace and Shell during the former's Brent Spar campaign in 1995 is an example of this perspective, as it underlines the agency of the environmental movement over powerful agents, such as one of the world's largest multinational extractive companies.
In summary, power over is asymmetric and seeks winners and losers in environmental politics; power with reveals the pioneers seeking consensual solutions by working together; and power to represents the selfempowering agents of transformation.For the first, the second and third perspectives seem overly 'naïve'; for the second and third, the first's conception of power is always too negative and does not take into account the agency of the less overtly powerful (see also Scott's seminal work on the topic, e.g.1985).Importantly, power over proposes a state of affairs in which change can only take place if it satisfies the powerful or if there is major external intervention.For power with and power to, power can come from creative thinking and alliances.See Table 4 for further comparison between these perspectives.

Definition
Who is responsible for change?
Power with Power is understood as the ability to act in concert, based on learning and cooperating with one other.
The collective, e.g.nation state and international community

Power to
Power is the ability to act in an indirect relational way.Individual actors and groups, e.g.NGOs, businesses, consumers

First dimension
Power is the potential of powerful actors to directly determine the actions of others.

Second dimension
Power manifests itself through some issues that did not make it on to the political agenda or were discarded before (observable) negotiations start.

Third dimension
Power is exercised by means of influencing, forming, and constituting ideas and intentions.

Fourth dimension
Power is inherent (inscribed) in social constructions of subjectivity and individuality that are described in historical terms.
'New thinking' that does not reproduce system and positions Source: Partzsch (2016)

Participation and empowerment
If we combine Partzsch's and Arnstein's takes on participation and power, we see that empowerment is a contested category.The key issue to consider for research purposes is that, while most participants in an MSF will expect to have some kind of influence over the forum's outcome, most government actors (and potentially powerful actors in general; e.g. the private sector) tend to be reluctant to release control over the decisions that affect them (see Arnstein 1969;Pretty 1995;Buchy and Hoverman 2000).In the case of MSFs related to land use and land-use change, this is usually about the development agendas and priorities of the powerful (see Sarmiento Barletti and Larson 2019).Yet, this may play out as a process that leads neither to joint gains nor to a positive effect for its participants, which goes against most scholarly definitions of what makes a process participatory (Pretty 1995;Warburton 1997;Beierle 2002).To illustrate this further, Ribot (2004; see also Ribot et al. 2006 and his texts cited in the previous section) has noted that decentralization may be commonly set around discourses of devolution and empowerment, but these discourses are rarely matched with institutional reform.Instead, the discursive broadening of participatory inclusiveness has advanced in a way that has allowed centralized government agencies to keep different degrees of control over the matter at hand, and especially so in regards to natural resources (see, e.g., Ribot et al. 2006;Mustalahti and Lund 2010).
The devolution of decision-making powers to local institutions, guised as decentralization, is superficial and does not carry enough support for the local level (see Ribot 2004).
This critique is related to Ribot and Peluso's (2003: 153) definition of access as "the ability to benefit from things".For these analysts, it is not about rights per se, but about a person's ability to access them.Access, they explain, is directly related to power relationships and can be applied to both procedural and distributive justice.From Ribot and Peluso's perspective, 'access theory' becomes a way to understand the different bundles of resource-based power (both material and non-material resources) that they deploy in their negotiations with the different actors and institutions through which they can gain, maintain, and control access to resources.Thus, they reveal a direct relationship between access and power that must be kept in mind by those interested in studying MSFs.For Ribot and Peluso (2003: 154), this perspective allows analysts to understand the wide "range of social relationships that can constrain or enable people to benefit from resources without focusing on property relations alone".This is something to consider as analysts navigate the key role(s) of power in the MSFs that they study.
In engaging with the participatory paradigm, we are engaging with a mechanism for decision making that is controversial, as it sits on discourses of democratic ideals (one of the central discourses in Western civilization) but has been criticized from different positions for falling short of its intended mission.MSFs are set up as purposely organized interactive processes that bring together a range of stakeholders to participate in dialogue, decision making and/or implementation regarding actions to address a common problem or to achieve a goal for their common benefit.The growth of MSFs related to land use/land-use change reflects the awareness that environmental problems cannot be addressed without the effective engagement of the actors that determine land-use practices on the ground; nor can such problems be resolved within a conservation community when the drivers are located in other sectors.MSFs may produce more effective and sustainable outcomes by getting actors with contradictory development priorities to coordinate and align goals through discussion, negotiation and planning.In contrast, MSFs may also help to implement top-down approaches and create the illusion of participation.Scholars and activists note that 'MSF' may reify these approaches, and take the 'participation' of local stakeholders for granted in box-ticking exercises to please donors.
This research is timely, because MSFs have received renewed attention from policy makers and development and conservation practitioners in light of the growing urgency to address climate change and transform development trajectories.The comparative project aims to contribute empirically to the study of MSFs and similar participatory processes.
CIFOR Occasional Papers contain research results that are significant to tropical forest issues.This content has been peer reviewed internally and externally.

Center for International Forestry Research (CIFOR)
CIFOR advances human well-being, equity and environmental integrity by conducting innovative research, developing partners' capacity, and actively engaging in dialogue with all stakeholders to inform policies and practices that affect forests and people.CIFOR is a CGIAR Research Center, and leads the CGIAR Research Program on Forests, Trees and Agroforestry (FTA).
Our headquarters are in Bogor, Indonesia, with offices in Nairobi, Kenya; Yaounde, Cameroon; Lima, Peru and Bonn, Germany.
The CGIAR Research Program on Policies, Institutions, and Markets (PIM) leads action-oriented research to equip decisionmakers with the evidence required to develop food and agricultural policies that better serve the interests of poor producers and consumers, both men and women.PIM combines the resources of CGIAR centers and numerous international, regional, and national partners.The program is led by the International Food Policy Research Institute (IFPRI).www.pim.cgiar.org

2 What is an MSF?
We set out this piece to inform research that seeks to engage analytically with MSFs.Although this review has considered a great deal of literature, it is by no means exhaustive.For those interested in looking further into the topic, we would recommend Moeko Saito-Jensen's (2015) Theories and Methods for the Study of Multilevel Environmental Governance, which was written as part of CIFOR's research into multilevel governance.We also recommend that this review is read in tandem with our project's Methods Training Manual for In-depth Field Research.researchwas carried out by CIFOR as part of the CGIAR Research Program on Forests, Trees and Agroforestry (FTA).FTA is the world's largest research for development program to enhance the role of forests, trees and agroforestry in sustainable development and food security and to address climate change.CIFOR leads FTA in partnership with Bioversity International, CATIE, CIRAD, INBAR, ICRAF and TBI.FTA's work is supported by the CGIAR Trust Fund: cgiar.org/funders/This Framing Literature Review for In-depth Field Research draws on the knowledge produced from 30+ years of experience in participatory processes.It informs the Center for International Forestry Research's (CIFOR) research of multi-stakeholder forums (MSFs) set up to address land use and land-use change at the subnational level in Brazil, Ethiopia, Indonesia and Peru.This literature review should be read in tandem with the project's Methods Training Manual for In-depth Field Research.