Organizing for transformation? How and why organizers plan their multi-stakeholder forums

their plans and expectations compare to previous lessons on invited spaces. Findings reveal that the organizers conceived of power inequalities as obstacles that could be overcome by including historically disempowered peoples in the MSFs, but generally failed to consider specific measures to address inequalities; nor did they develop clear strategies to engage with unsustainable local development and political priorities. enfrentar las prioridades políticas y de desarrollo locales que no apoyan la conservación.


INTRODUCTION
Over the past decades, multi-stakeholder forums (MSFs) have been organized for all kinds of collaboration and coordination under different initiatives, such as community forest management (Mohanty 2014, Nayak andBernes 2008), participatory budgeting (Shan 2007, Wampler 2010, and resource management (Chímère et al. 2009, Søreide andTruex 2011). MSFs have received renewed attention worldwide due to the urgency to transform development trajectories to address the climate crisis. This attention builds on a tendency -particularly among NGOs, government actors and donors -to see MSFs as a transformational solution to the challenges posed by land and forest degradation (Gonsalves et al. 2005, Bastakoti and Davidsen 2015. The debate over the transformational potential of MSFs is framed in part by a discussion regarding whether they can address the power inequalities inherent to the interactions among their participants (see Sarmiento Barletti et al. 2020 for a review). Power inequalities are a central challenge to MSFs that address unsustainable land and resource use, as they are often set within contexts marked by histories of conflict and deep inequalities, and where trade-offs are inherent and the actors involved are significantly diverse (Robbins 2012, Barnes and Child 2014). This is important to consider, as recent experimental research on reciprocity and collaboration concluded that inequality undermines cooperation (Hauser et al. 2019).
Many MSF proponents argue that bringing stakeholders together to discuss a common challenge or opportunity confronts power inequalities among participants; leads to solutions that are more likely to reflect the priorities of historically underrepresented actors (as opposed to top-down decision making or bilateral negotiations); allows more powerful participants to understand the perspectives of vulnerable groups; and includes actors that can affect the implementation and effectiveness of consensus outcomes (Buchy and Hoverman 2000, Hemmati 2002, Dougill et al. 2006, Faysee 2006, Tippett et al. 2007, Reed 2008. Given the centrality of collaboration to current discourses about the transformative change needed to address the climate crisis, these claims merit investigation. This paper is based on such research. It considers the unexplored role of MSF organizers by considering why they set up MSFs and whether they had transformation in mind when they did so. The paper comparatively analyzes interviews with forty-five organizers of thirteen MSFs established to address unsustainable land and resource use in twelve subnational jurisdictions in Brazil, Ethiopia, Indonesia and Peru. 1 Interviews examined organizers' perceptions of how they thought they would achieve their objectives; how power relations and other contextual factors may have impacted on their proposed pathway to change; and the role that their MSF would play in such pathway. Research shows that the MSFs in this sample are "invited spaces" -"those into which people (as users, citizens or beneficiaries) are invited to participate by various kinds of authorities, be they government, supranational agencies or non-governmental organizations" (Cornwall 2002: 24; see Cornwall and Gaventa 2000). Past research found that invited spaces tend to be government-led and respond to top-down decision making (Cornwall 2008), which is strikingly different from how proponents discuss their potential. The paper reveals that although most organizers perceived the MSF method as a way of doing things 'differently', they did not perceive them to be transformative on their own. This seeming contradiction between the optimism for MSFs and their real potential can be explained through two overlapping conceptions of these forums -as an event and as a method of practice. Most organizers proposed MSFs as events that balanced inequalities between stakeholders by bringing them together. They highlighted the potential of MSFs for capacity development, raising awareness and collaboratively designing, or just collaboratively implementing, solutions. This positive perception of power, regarding its potential for collaboration, may be influenced by the organizers' privileged positions as government or NGO actors. Yet, MSFs, as a method, were not seen as enough to curb the political interests and development priorities that drove unsuitable land and resource use in their jurisdictions. As organizers aimed to counter mainstream land and resource use interests and priorities that were shared by the more powerful actors in each jurisdiction, it is unsurprising that their MSFs were underfunded and led to outcomes that had no real impact. Additionally, their MSFs sought to increase, and thus include more of the same type of actors, rather than widen participation to include historically underrepresented actors. The two overlapping perspectives -event and method -raise the question of whether MSFs are organized as platforms for people to plan and enact change together, or as platforms primarily aimed at including relevant actors in the implementation of the ideas proposed by their organizers. And are either of these transformational?
The following section considers the relation between power, participation and MSFs' potential for transformational change. This is followed by a discussion of research methods, and the presentation of results and analysis. The paper concludes by discussing the relevance of these findings for wider debates on MSFs.

MSFs AS INVITED SPACES
The current optimism associated with MSFs comes largely from the notion that enabling citizens to participate more directly in governance, and particularly in communication with government actors, may lead to enhanced democracy by creating better understanding between citizens and public servants and more equitable and effective governance (Cohen andSabel 1997, Avritzer 2002). These participatory processes are said to present an "entirely different kind of interface with policy processes than other avenues through which citizens can articulate their demands -such as protest, petitioning, lobbying and direct action" (Cornwall and Coehlo 2007: 21, see also Gaventa 2001, Goetz and. MSFs are commonly organized as invited spaces as part of the wider effort for more sustainable land and resource use. Although there is no agreed upon definition for what transformational change means with regard to forest and land use, the term has become a rallying cry to address the climate crisis. Atmadja et al. (n.d.), based on a synthesis of common features of how the concept is addressed in scholarly and practitioner publications, outlined five agreed-upon critical elements for moving towards transformational change. Two of those are 'collective learning and reflection' and 'consultation and participation', both of which are reflected in the current interest in MSFs, revealing their central role in how transformational change is being discussed and designed in environmental circles.
Regardless, there is doubt that this transformative potential will be reflected in practice (see Ravikumar et al. 2018).
Participatory processes do not guarantee equality, as the interactions within them and in the wider contexts where they are enacted are shaped by power relations that define what kinds of actions are possible (Gaventa 2006, Grönholm 2009, Perrault 2015. Power inequalities manifest in different ways in the jurisdictions examined in this paper. As the MSFs all engage with land and resource use, the most relevant inequalities are stakeholders' access to political power (e.g. influence over decision makers), economic and natural resources (e.g. financial power to get things done, access to natural resources), technical knowledge (e.g. knowing how to get things done), epistemology (e.g. deciding on the acceptable knowledge to be implemented in a project), and gender (e.g. excluding women) (Chambers 2006;White et al 2015).
In this paper, the exercise of power is more than an actor's ability to control others ('power over'); power is also in actors' agency ('power to') and in the possibility to collaborate ('power with'; see Dewulf andElbers 2018, Partzch 2017), which permeates the optimism set on MSFs. The possibilities within this optimism range between what Avelino and Wittmayer's (2016) consider as 'innovative' powerwhich enables the development of ways to change power relations -and 'transformative' power -which leads to the development of new institutions to challenge power relations. MSFs, by design, are supposed to be a shift away from 'reinforcive' power, which maintains the status quo.
As noted earlier, the MSFs in this sample are what have been previously defined as 'invited spaces' (Cornwall 2002). These spaces have been associated with decentralization and democratization, especially in their potential for enhanced accountability, to empower marginalized actors, and to attend to citizens' demands (Robins and Von Liers 2004, Aiyar 2010). They have also been found to have unintentional outcomes, including the creation of political subjects (Mohanty 2004, Cornwall andCoehlo 2007); the organization of civil society and the expansion and strengthening of alliances (Taylor 1998, Barnes 2007; and non-participation as a form of protest (Cortez Ruiz 2004). More critically, these spaces have been challenged by various factors including: the need to develop the capacities of their members to participate more effectively (Harrison 2003, ANGOC 2006; the power asymmetries between participants (Mahmud 2004, Gaventa 2006; the tendency for top-down decision making (Cornwall 2002); the participation of non-representative actors (Baud and Nainan 2008, Acharya et al. 2004, Coelho 2004, Cornwall 2004; the lack of a clear supportive legal framework for participation (ANGOC 2006); and the finding that they rarely produce binding outcomes (Manor 2004, Williams 2004. Research on invited spaces has affirmed the importance of addressing power inequalities among their participants and of understanding how these processes are affected by their contexts. Participatory spaces, as Mohanty notes, never occur in a vacuum but rather "react upon already existing spaces, on spaces which are simultaneous and overlapping and on the wider social-economic-cultural setting in which they are embedded " (2004: 26). Critics have argued that invited spaces are frequently theatres for participation so that organizers, commonly government actors, can implement rather than A semi-structured, open-ended questionnaire was designed to understand how and why interviewees decided to establish their MSFs. Research materials were translated into Amharic, Indonesian, Portuguese and Spanish in order to interview participants in their national languages. Responses were transcribed, translated and organized in an excel file and manually grouped and coded for similarities (see below). This paper focuses on questions intended to understand if and how organizers took into account the power relations and political and development priorities in the context where they implemented their forums (see Sarmiento Barletti and Larson 2019: 25-27 for the full questionnaire). These are: what organizers aimed for by organizing their MSFs and what conditions were needed to bring about change 3 ; whether they took into account local politics and development priorities when planning their MSFs 4 ; and whether they took power relations between participants into account, including their own 5 . In the results section, the answers to these questions are presented comparatively by MSF, also noting any contrasting views by multiple organizers of the same MSF.

What did MSF organizers aim for?
Given that all case studies were MSFs addressing land-use sustainability, organizers were asked what they thought would be necessary to change land-use behavior among key local actors in their jurisdiction and what role their MSF would play in achieving it. Interview results were synthesized into eight strategies or activities aimed at changing land-use behavior (see Table 3). No interviewee across thirteen case studies explicitly mentioned MSFs as a way to change landuse behavior. The closest mention was in four MSFs, where at least one organizer mentioned coordination between different actors as a necessary catalyst for more sustainable land use. At least one organizer in each of eight MSFs mentioned improving conditions for sustainable production and providing economic benefits from conservation. Interestingly, in eight MSFs -organized by different types of actors -at least one organizer mentioned better enforcement of existing regulations. Other responses included raising environmental awareness (at least one organizer in seven MSFs), capacity development for sustainable development (at least one organizer in seven MSFs), closer communication with and participation of indigenous and local community (IPLC) representatives in environmental planning and initiatives (at least one organizer in five MSFs), and tenure formalization for IPLC (at least one organizer in three MSFs).
negotiate their agendas (Cornwall 2002), or legitimize their decisions and disarm any potential local opposition (Alonso and Costa 2004). Similarly, Baud and Nainan (2008) noted that participants may experience these spaces in different ways. For vulnerable groups these spaces serve to negotiate their rights; for more powerful civil society groups these spaces allow them to exercise their power to negotiate and hold their governments accountable.
In sum, previous research on invited spaces underlines the importance of recognising and understanding the manifestations of power as a first step towards assuring equal and equitable participation and, given the current interest in MSFs, this paper moves from there to assess their transformational potential. In what follows, the paper explores these issues by analyzing the perspectives of and actions taken by MSF organizers.

METHODS
This paper presents the results of interviews implemented as part of a comparative study of subnational MSFs that aimed to achieve more sustainable land or resource use (Sarmiento Barletti and Larson 2019). In total, forty-five organizers of thirteen different MSFs in twelve jurisdictions of Brazil, Ethiopia, Indonesia and Peru were interviewed (see Table 1 for short descriptions of each case study and Table 2 for a breakdown of interviewees). 2 Interviewees were selected because they were recognized by scoping research as the main organizers of each forum. In most cases, the MSFs had more than one organizer.
For the purpose of this research, we defined MSFs as "purposefully organized interactive processes that bring together stakeholders to participate in dialogue, decisionmaking and/or implementation regarding actions seeking to address a problem they hold in common or to achieve a goal for their common benefit" (Sarmiento Barletti et al. 2020:2). The MSFs in our sample fulfil this definition and were selected after scoping research because they also fulfilled five other criteria: they included a forum for in-person interactions; they included different types of actors -at least one government and one non-government local actor; were organized at the subnational level; sought to address unsustainable land or resource challenges; and had been meeting for at least a year. In the data below, interviewees are categorized by the region where they organized their MSF and their actor type -national and subnational government, NGOs, local or indigenous community representatives, university/research institutions, and donor organization. These different types are revealing of the variety of stakeholders in the case studies. Mato Grosso (MG)

Social-Economic and Ecological
Zoning Commission Conduct a participatory and inclusive Social-Economic and Ecological Zoning process.

Ethiopia Oromia (JU) Jamma-Urji Farmer Managed
Natural Regeneration Increase biological diversity, forest carbon stock and natural forest regeneration; transform local people's livelihoods by channeling carbon revenues and access to environmental benefits.

Oromia (SH) SHARE Bale Eco-Region
Contribute to sustainable land use through an inclusive design and process that brings together stakeholders from different sectors to discuss common problems and find solutions that benefit them all.

INOBU-UNILEVER Palm Oil
Initiative Solve overlapping land-tenure issues, promote sustainable agriculture practices and help small-scale farmers obtain ISPO/RSPO certification to access better markets for their palm oil products.
East Kalimantan (EK) Regional Council on Climate Change Coordinate East Kalimantan's climate change-related programs.

Jambi (JAM) Regional Peatland Restoration
Team Raise awareness about the negative impacts of conversion of the community's forest and loss of its natural resources; formulate a solution for customary forest management and conservation together with the local community.
West Java (WJ) Integrated Citarum Water Resources Management Investment Program Established as part of an agroforestry project that aimed to build connections between farmers and the subnational government at provincial and district levels, and to obtain political and financial support to ensure the sustainability of interventions. Answers to what role their MSF would play in changing land-use behaviour were synthesized into six themes (see Table 4). Despite the attention on coordination and joint decision making in the literature, the most common answer (at least one organizer in all cases except Mato Grosso) highlighted the capacity development role of MSFs in changing land-use behaviour. Only in three MSFs (Acre, Pará and Ucayali, all organized or co-organized by a subnational government actor) did at least one organizer note the MSF would play a role in changing land-use behaviour by opening a space for dialogue that included and respected different opinions. For example, an organizer in Ucayali said that the MSF was designed to bring together different interests and organize them towards the common good, understood as decreased deforestation and degradation through sustainable forest management.
At least one organizer each in San Martin, Ucayali, East Kalimantan and West Java -all organized by different actor Economic benefits from conservation Raising environmental awareness Capacity building Communication and participation of IPLC Land tenure formalization x x x 3 types -referred to capacity development, focused on sustainable economic activities for IPLC. In Jambi, it was directed at local communities and their customary institutions to allow them to comply with national environmental regulations. In Pará, the focus was on landowners undergoing land registration, as well as on government officers to better monitor deforestation. In Loreto it was also focused on government actors, who were trained on laws and regulations regarding isolated indigenous peoples. Continuing with this perspective of MSFs as didactic spaces, at least one organizer in six MSFs described them as raising awareness on environmental issues. In San Martin it was directed at raising awareness about conservation among civil society organizations. At least one organizer in nine MSFs (Acre, Mato Grosso, Pará, Jamma-Urji, SHARE-Bale, East Kalimantan, Loreto, Madre de Dios and San Martín) described their MSFs as spaces for coordination and joint decision making for more sustainable land and resource use, planning and management. For example, in Acre, Mato Grosso, Para and Loreto -all led by a subnational government actor -, at least one organizer said their MSF focused on coordination for territorial planning. The MSFs in Acre and Mato Grosso were set up as collaborative spaces to develop zoning maps and promote sustainable development. In Loreto, an organizer noted that the MSF was designed for coordination between actors with a stake in an area with overlapping land-use regimes where reserves for isolated indigenous peoples (who themselves could not participate) would be introduced. As for land and resource management, at least one organizer in each of Jamma-Urji, SHARE-Bale, East Kalimantan, Madre de Dios and San Martín -all organized by different types of actorsnoted that their MSF served as a platform for different actors to coordinate the sustainable management of an area. The MSFs in Madre de Dios and San Martín were legally mandated spaces to support and inform the management of protected areas. In Jamma-Urji, an organizer explained that the MSF was organized to bring together representatives of participatory forest management cooperatives to discuss their concerns over land and resource management and come up with consensual solutions. Organizers in SHARE-Bale and East Kalimantan described their MSFs as spaces for different actors to discuss and coordinate their sustainable development interventions.
In summary, organizers did not pose that their MSFs, on their own, would resolve the issue they sought to tackle (see Table 4). MSFs would support such processes through mainly two roles: capacity development (in twelve MSFs) and coordination (in nine MSFs).
What did organizers believe were the necessary conditions to ensure sustainable changes?
Organizers were asked about the conditions they thought would ensure the sustainability of the land-use changes their MSF supported. Five factors were synthesised from interview results (see Table 5). For the first, at least one organizer in each of nine MSFs (Loreto, Madre de Dios, Ucayali, Acre, Pará, Central Kalimantan, Jambi, Jamma-Urji and SHARE-Bale) proposed that land-use behavior would change if stakeholders had financial support, technical advice, capacity development and materials for sustainable production, or were compensated for losses in income from shifting to more sustainable land use. The second factor, mentioned by at least one organizer in seven MSFs (Loreto, Madre de Dios, San Martín, Ucayali, Pará, Jambi and Jamma-Urji), was the need to raise the environmental awareness of relevant stakeholders and educate them on the environmental impact of their land-use practices. The third one, mentioned by at least one organizer in seven MSFs (Loreto, Mato Grosso, Pará, Central Kalimantan, Jambi, West Java and SHARE-Bale), involved setting clear regulations and enforcement of existing laws, agreements made at MSFs, or pledges made by relevant stakeholders.
The fourth factor, mentioned by at least one organizer in six MSFs (Madre de Dios, San Martín, Acre, Jamma-Urji, SHARE-Bale and East Kalimantan), was the funding available for the MSF. Organizers noted that budget limitations affected the sustainability of their MSFs and how inclusive they were in terms of participation and outreach activities. However, few organizers had a funding strategy for long-term work. 6 Those MSF organizers that did (e.g. Pará) implemented strategies through which they shared the economic and human resources costs of the process with NGOs. The lack of a funding strategy is telling of how inclusive these processes actually were (as opposed to who was invited to participate): most MSFs in the sample met in the jurisdictional capital but could not afford to bring in actors that lived elsewhere. An exception is Acre where the MSF held decentralized meetings in municipalities. Finally, at least one organizer in five MSFs (Madre de Dios, San Martín, Ucayali, Pará and SHARE-Bale) mentioned the need for communication and coordination in an inclusive space, both in terms of the participation of historically underrepresented groups and assuring that nontechnical language was used.

Did organizers take into account local political interests and development priorities when planning their MSF?
Organizers were asked if they had considered the impact that local political interests and development priorities could have on their MSFs. As in most tropical countries, the primary political interests and development trajectories in most of the jurisdictions hosting the MSFs in the sample followed wider historical trends of natural resource extraction. Their MSFs, in seeking more sustainable land use, went against such trends. In seven cases (Loreto, Madre de Dios, San Martín, Ucayali, Jamma-Urji, SHARE-Bale, East Kalimantan) at least one organizer responded that their MSFs could not curb mainstream political interests or development priorities aligned with unsustainable land use. At least one organizer in six MSFs (Madre de Dios, San Martín, Pará, Central Kalimantan, SHARE-Bale, East Kalimantan) acknowledged clashes between the development priorities held by some government agencies and the private sector in their jurisdiction and the environmental agendas held some MSF stakeholders, including conservation NGOs, government environmental agencies, and IPLC organizations. For example, one organizer in Madre de Dios mentioned informal road construction and gold and timber extraction as a challenge. These were driven by private interests and by the Madre de Dios government's own development agenda at the time of research. As such, the sustainability agenda promoted by the MSF may have been in tune with the national government's protected area policies but against that of the regional government and elites who favored the expansion of alluvial gold mining. This perception of local priorities as a challenge is also illustrated in how at least one MSF organizer each in Loreto, Ucayali, Madre de Dios and San Martín (all in Peru) noted a lack of interest in the topic by their subnational government and posed that they organized their MSFs knowing that they would be challenged by natural resource extraction. Both Loreto and Ucayali were co-organized by environmental or indigenous affairs offices in the subnational government; Madre de Dios and San Martin were co-organized by a national agency.
Other issues related to government actors that were noted by at least one organizer were: the fast rate of staff turnover in the jurisdiction (East Kalimantan, San Martín); overlapping forest governance mandates between different government agencies (Jamma-Urji, Loreto); and lack of law enforcement to stop deforestation (Madre de Dios, SHARE-Bale). Only in Acre were development priorities not seen as an obstacle by interviewees, as the jurisdiction promoted sustainable land use. Organizers noted that they had designed the MSF as a process to raise awareness of relevant legislation and for collaboration towards continued sustainable land use.

Organizers' strategies to address political challenges
Although local political and mainstream development agendas tended to be understood as a challenge because they supported unsustainable activities, some interviewees had strategies to address them. Those mentioned by at least one organizer included: planning to engage the government for the long term so that staff turnover did not affect agreements (Central Kalimantan, West Java); including different government agencies in the process (Pará); framing the MSF's work within existing government regulations (Mato Grosso) or customary institutions (Jambi); and adjusting the MSF's activities and expectations to different government agencies and levels (Loreto, Mato Grosso and Pará -all led by subnational government actors). For example, organizers in Pará tailored their promotion of sustainable land use to the different local priorities and challenges in the state's municipalities. They sought to understand the development priorities and political dynamics of each municipality to include them effectively in the MSF. In similar vein, at least one organizer in Loreto noted that the MSF had originally been organized to reconcile overlapping land-use regimes in areas delimited for reserves for isolated indigenous peoples. However, organizers shifted their objective from collaborative problem solving to capacity development after finding great knowledge gaps among participants on isolated indigenous peoples and the laws that protect them. The NGO sector organizer noted that although participants had been learning about isolated indigenous peoples and their rights, the MSF was far from shifting subnational development trajectories that infringed on those rights. The government sector organizer noted a challenge in that the MSF sought to implement national law without considering local development priorities. Loreto is illustrative of cases where national and subnational government priorities were mismatched; the Ministry of Culture sought to promote areas for isolated indigenous peoples, but Loreto promoted extractive activity and infrastructure expansion in the same areas.

Did organizers take unequal power relations among MSF participants into account?
Organizers were asked if they had considered how power inequalities between participants could have affected their MSF and if they had planned to address them. None of the organizers of three MSFs (Jamma-Urji, Central Kalimantan and East Kalimantan) acknowledged that there were power inequalities among the participants in their MSF. In five cases -all with government organizers -(Loreto, Madre de Dios, San Martín, Ucayali, Jambi), at least one organizer considered that their MSF would balance out power inequalities between stakeholders because they were open spaces for different people to come together, participate and be heard. An organizer in Loreto said that their MSF brought people with different degrees of power "together in open dialogue and shared decision making". The interviewee considered that it was necessary to bring together stakeholders with less power (e.g. IPLC organizations) with those actors with decision-making power (e.g. the Ministry of Culture and Loreto's regional government). This position emphasizes the view that bringing people together in an MSF is sufficient to balance power inequalities. Similarly, the government sector organizer in Madre de Dios considered that the fact that an IPLC organization led the MSF, that it is an open space, and that economically powerful actors do not dominate conversations proved that the process was unaffected by power inequalities. At least one organizer in ten cases (Loreto, Madre de Dios, San Martín, Ucayali, SHARE-Bale, Jambi, West Java, Acre, Mato Grosso, Pará) acknowledged power inequalities among the participants to their MSF, understood mainly in terms of their socio-cultural, economic, political and technical capacities to influence the MSF and its outcome. Not all of those respondents described any special measures to address this challenge. In Jambi, for example, at least one organizer referred to gender-based inequalities in the daily interactions in the community where the MSF was focused but did not report any actions to address that challenge. However, at least one organizer in three MSFs (Acre, Mato Grosso and SHARE-Bale) described strategies to balance power inequalities. In Acre, non-technical language was purposefully used to level out differences in technical expertise. Organizers also ensured that there were spaces organized separately for IPLC to discuss MSF-related topics and that decisions at the MSF were made by consensus. In SHARE-Bale, organizers reported conducting a power analysis during the MSF planning phase, as they sought a forum where all participants would be treated as equals and local community representatives would have a direct voice. Although they did not set quotas for women, they did place special attention on gender when planning projects and involving women in the MSF's activities. In Mato Grosso, an interviewee noted that as the first phase of the MSF had mostly included government actors, she broadened the types of participating actors in its second phase.

How did organizers address the absence of key stakeholders?
Power inequalities can also be illustrated by how organizers discussed the absence of key stakeholders from their forums. Although proponents describe MSFs as more inclusive than mainstream participatory processes, wider research in the same project found that some key stakeholders did not participate in the forums studied for reasons such as their lack of interest or resources to participate, having conflicting interests to those of the MSF, or being excluded by an organizer. At least one interviewee in all but two cases (Acre and Pará) noted absences, yet few assumed any responsibility over those absences. For example, at least one organizer in five cases (Madre de Dios, San Martín, Central Kalimantan, West Java, and Jamma-Urji) responded that conflicting interests kept some stakeholders from participating in their MSFs. In San Martín, some local communities demanded collective titles to the land they occupied within the protected area supported by the MSF. One organizer noted that the local communities decided not to participate in the MSF, while another said they had not been invited because they refused to leave the protected area. In Madre de Dios there were local communities involved in alluvial mining close to the protected area supported by the MSF. One organizer said that miners refused to participate in the MSF, but another one stated that they were not allowed to participate because mining was against the MSF's objective. In Central Kalimantan, an organizer said that oil palm associations and many large oil palm companies self-excluded from the MSF because it questioned the sustainability of their businesses. An organizer in Jamma-Urji mentioned that private sector actors were invited to participate in the MSF but chose not to, as they were afraid of the conflicts that could arise from open debate with local communities. In San Martín and Mato Grosso organizers pointed out that the absence of indigenous representatives from the MSF was not their doing as they had been invited to participate. An organizer in Madre de Dios had the same position over the lack of women in the MSF, which was open to "any relevant organization".
At least one organizer in four MSFs (Loreto, Madre de Dios and Jamma-Urji and SHARE-Bale) recognized some responsibility regarding these absences. In SHARE-Bale, an organizer noted problems inviting the participants early on as organizers did not clearly convey their message and most participants joined the platform when it was already running. At least one organizer in SHARE-Bale, Loreto and Madre de Dios mentioned that their own budget constraints prevented the MSF from being more inclusive, since travel expenses could not be covered. Organizers in Acre and Pará did not report missing stakeholders. In Pará, organizers noted that no one had declined the MSF's invitation and that other actors had requested their inclusion. However, interviews with nonparticipants and the review of documents published by the MSF (carried out as part of the wider research project) revealed that IPLC organizations were not invited to participate in the MSF. Acre organizers commented that there were no excluded stakeholders, although IPLC organizations had not participated in the past as much as other actors because they could not afford the costs incurred in travelling to meetings. Decentralized meetings were introduced to address this -we know of no other forum that took such a step.

DISCUSSION
Based on the data presented above, MSFs can be understood in two non-mutually exclusive forms: as events and as a method. As events, MSFs are seen as able to balance inequalities between stakeholders by bringing them together periodically. Because organizers aimed to counter different aspects of unsustainable land and resource use in their jurisdictions, their events challenged the status quo and thus were potential opportunities to collectively think outside the box together and enact change. Within the framing as an event, invited participants collaborated as equals towards their common goals. Yet, those events were short-lived, excluded some stakeholders, and did not always lead to tangible outcomes.
Most interviews, however, conceived of MSFs as a method. This was framed by the political interests and development priorities that drove unsuitable land and resource use in their jurisdictions and that limited the MSF's actions. This led to forums that brought actors together for implementation -and sometimes to co-design the implementation -of their organizers' ideas. To avoid conflicts, MSFs were designedor shifted -to deal with the effects of unsustainable land and resource use, rather than the structural issues framing those contexts. They tended to deal with issues as technical problems, raising awareness and developing capacities rather than following their political ramifications. Also, because organizers aimed to counter different aspects of unsustainable land and resource use, they were not supported by some of the most powerful actors in their jurisdiction as the latter had vested interests in maintaining the status quo. This is probably why their MSFs were underfunded, led to outcomes that still had to be approved by decision makers, and increased the participation of actors of the same type (e.g. government and NGOs) rather than widening participation to underrepresented actors (e.g. IPLC). We discuss each of these types further below.

MSFs as an event
Given the political interests and development priorities in most of the jurisdictions in which the MSFs in this sample were organized, their pursuit of more sustainable land and resource use could be read as transformational. MSF organizers tend to understand their forums, as events, in terms of what Avelino and Wittmayer (2016) described as 'innovative' power, which enables the creation or discovery of new resources to change power relations. Although three case studies had organizers that described deploying measures to address power inequalities inside their MSF 7 , the rest saw the act of inviting people around the same table as enough to undo them. Organizers perceived a constructive potential for power, as participants would collaborate to achieve an outcome for their common benefit.
This positive perception of power may have to do with the fact that organizers were among the most powerful actors in the MSF, were not going to be affected by its outcome, and had organized the MSF with their own solution in mind. Two actor types dominated among organizers: subnational government actors (six MSFs: Acre, Mato Grosso, Pará, Jambi, Loreto and Ucayali) and subnational and national NGOs (in five: Jamma-Urji, SHARE-Bale, East Kalimantan, Jambi, Loreto). Six MSFs were organized by more than one actor type combinations: subnational government and NGOs (Jambi, Loreto), indigenous people and government (Madre de Dios), indigenous peoples and subnational government (Loreto, San Martín), donor and academia (West Java), and subnational government and donor (Ucayali). Subnational government and/or NGO-led MSFs tended to have outcomes to inform policy (e.g. Acre, Mato Grosso, Pará) or the management of a project or coordination between projects (e.g. East Kalimantan, Jambi, Jamma-Urji, SHARE-Bale). The the organizers already had in mind. Most MSFs had at least one government organizer or were organized by NGOs that worked with government institutions, and so their solutions tended to fall within the limits of what was acceptable for mainstream political and development interests.
To support these solutions, MSFs raised the awareness of relevant actors, developed their technical capacities, facilitated their conversations over time, and led to preliminary outcomes that had to be confirmed by decision makers outside the MSF. In some cases, differing opinions were controlled by excluding stakeholders with different goals (e.g. Madre de Dios, San Martín, Pará). In other cases, organizers passed the blame of non-participation to stakeholders themselves, claiming that they were not interested in the topic (or, arguably, in the organizer's approach), or unable to afford the travel expenses to the urban areas where the MSFs met (e.g. Loreto, Ucayali). That organizers in only three MSFs made an effort to address absences raises questions as to whether the others understood equitable participation as supportive to their goals. This is important because selfexclusions may represent the performance of an actor's agency, but they also reflect power inequalities. While powerful stakeholders self-excluded to avoid debates over their land-use practices (e.g. Central Kalimantan, Jamma-Urji), interviews revealed that underrepresented groups did so because they doubted that an MSF would lead to real change for their communities (e.g. Ucayali) or because they could not afford abandoning their livelihood activities to attend the MSF (e.g. Jamma-Urji).
The process through which actors were included and excluded in some MSFs, and the fact that most had set agendas by the time participants joined in -failing to allow underrepresented actors to 'increase the visibility and legitimacy of their issues, voice and demands' (Gaventa 2006: 29) -gives them the characteristics of 'closed' spaces (Gaventa 2006). The MSFs were framed by the 'hidden' power through which organizers 'maintain their influence by controlling who gets to the decision-making table and what gets on the agenda' (Gaventa 2006: 29). These approaches to organizing MSFs reproduce the notion that underrepresented peoples are able and willing to participate when granted a space, without considering their access to funds and technical capacities. This is not to say that organizers acted in bad faith, but it reveals the consequences of being limited by powerful interests in their jurisdictions. The limited resources available for their MSFs are another consequence of the impact of those interests. Limited budgets may have widened inequalities, as organizers lacked funds to more equitably include participants or to set up activities to develop their capacities to participate more effectively. Thus, setting up an MSF without sufficient funds is not only challenging but also may worsen existing inequalities by blocking the possibility of equitable and effective participation. Also, limited funding may result in short-term initiatives, when transformational change requires a long-term perspective. organizers to some of these MSFs did not invite IPLCs to participate in the forum (e.g. Central Kalimantan, East Kalimantan, Mato Grosso), and when IPLCs organized MSFs (Loreto, Madre de Dios and San Martín), they did so together with a government or NGO actor. These three latter MSFs addressed issues that were directly related to their communities or organizations, such as adjacent natural protected areas or the representation of isolated indigenous peoples. The two MSFs that were co-organized with a national government actor were legally mandated participatory spaces for protected areas (Madre de Dios, San Martin) and the third followed a jurisdictional regulation (Loreto). The first two constituted one of the entities through which management plans for protected areas were approved, but interviews carried out as part of the wider research project revealed that the MSF had no real input on how the plan was designed or implemented. The third MSF dealt with isolated indigenous peoples, represented politically by the indigenous organizers who held technical expertise on the topic. The MSF started as a roundtable to support the recognition of five areas for isolated indigenous peoples but became a forum to raise awareness about their rights. At the time of research, it was yet to impact the national and regional development policies that placed isolated peoples in Loreto at risk (e.g. hydrocarbon and forestry concessions and motorways).

MSFs as a method
When MSFs were perceived as a method to jointly develop solutions to unsustainable land and resource use, it was in jurisdictions with a development trajectory that supported the MSF (e.g. Acre) or at a small scale and with the direct participation of a community (e.g. Jambi). As the wider research showed, those MSFs had a specific outcome (e.g. completing a zoning map or achieving the recognition of a communal forest), included the active participation of underrepresented actors, took at least some of the MSF meetings to these participants (e.g. holding meetings closer or in local communities), and facilitated the process (e.g. made it non-technical) to allow for their effective participation.
Elsewhere, the method was informed by the limits placed on MSFs by local politics and development priorities driving unsustainable land and resource use in their jurisdictions. This impact can be explained in at least two ways: the influence those priorities had on decision making -including cases where the actors driving deforestation were represented in legislative assemblies (e.g. Mato Grosso, Pará) -and the lack of power or political influence of relevant national government agencies to enforce regulations (e.g. Peru's ministries of Culture and Environment). This resounds with findings from research in Peru where MSFs that sought outcomes that countered local development priorities tended to fail (Sarmiento Barletti and Larson 2020). As the interests driving unsustainable land and resource use in their jurisdictions were too powerful to curb, MSFs followed a method to implement and sometimes co-design the implementation of solutions that CONCLUSION Assessing the transformational potential of participatory processes is not a new endeavor. However, this paper has inspected MSFs after their re-introduction as a 'new' method of practice in light of climate change, with hopes placed on broad participation and collaboration to design and implement new landscape-scale solutions. Recognizing this, the paper comparatively considered the perspectives of the organizers of thirteen MSFs in four different countries. This was done to understand how and why they organized their MSF and, in order to assess their transformational potential or desire for transformation, whether they had recognized power inequalities between their participants and whether they had seen it as possible or necessary to address them.
Why are MSFs still being considered as a transformative solution despite evidence of their limitations? There is no simple answer. It may be a matter of discourse -ideals of deliberative democracy that are central to Euroamerican political ideals. MSFs as events are portrayed as a step in bridging the representation gap for historically underrepresented groups who come together as equals with more powerful actors with whom they may have historical grievances. However, in leaving power relations unaddressed, MSFs may limit the effectiveness of the participation of underrepresented groups and expose these groups to 'common good' outcomes that may not be in their best interest. Also, although it is conceivable that an event can be catalytic or transformative, we did not see any evidence of this in our cases.
Most interviewees organized their MSFs knowing that their goals would be challenged by the development and political trajectories driving unsustainable land use in their jurisdiction. Viewing MSFs as a method of practice, it may be that organizers sought to share their plausible solutionsplausible because they were within the means of what was acceptable to political and development interests -with key actors and develop their awareness and capacities over time to participate in the implementation of the solution the organizers proposed. This largely concurs with assessments of 'invited' spaces. Furthermore, although this paper did not concentrate on the effectiveness of these MSFs, some of them may have indeed been effective. However, while broadening participation was understood as part of any MSF's work, there was less emphasis on who it was broadened to, why, and whether it could support its effectiveness. This shows that organizing a space for discussion and coordination does not ensure that its process or outcome will overcome the historical structures that frame how land and resources are used and how environmental issues are experienced by its participants. The current rate of social protests around the world, and the authors' experience of indigenous protests in Peru, is testament to the current role of social action as a potentially more effective yet costly pathway for historically underrepresented peoples to be heard and have their rights recognised.
Exploring how organizers plan their MSF matters because research keeps showing that despite much optimism for participation and democratic practice, simply sitting people around a table is not enough, as MSFs are superimposed upon existing patterns of relationships, institutions and, importantly, power relations. Notably, when asked about solutions for unsustainable land and resource use apart from their MSFs, organizers offered options that can be synthesised into four elements: financial support, technical advice and capacity development for communities and smallholders for sustainable land management; clear environmental regulations and enforcement; raising awareness of the environmental costs and benefits of sustainable land use and associated regulations; and effective multi-sector and multi-level communication and coordination. These options are of interest for at least two reasons. Firstly, there is no mention of power inequalities and the deeper structural issues that drive deforestation and degradation in their jurisdictions. Secondly, at least part of them -raising awareness, capacity development, communication and coordination -were already central to the efforts across all case studies.
The analysis of organizers' perspectives on their MSFs reiterates that for MSFs to reach their transformational potential, the first step is an old one -recognising that power differentials cannot be addressed simply by bringing people together, whether as event or as method as currently conceived in our cases. This research revealed two issues that need special attention (see also Coelho 2004 andSarmiento Barletti et al. 2020). The first is the need to design and implement strategies to address power inequalities between participants -for example, by using non-technical language in meetings, and assuring the effective participation of historically marginalized actors through capacity development and covering their expenses related to participation. These strategies must make all participants feel that they are being listened to and are getting something out of the MSF. The second is that MSFs must have funding strategies that permit organizers and participants to think beyond the short-termto bring people together, develop their capacities to participate and engage with policy and decision makers to implement their decisions. Further analysis on how participants perceive the strategies applied by MSFs organizers will help identify those that are more successful.