Facilitating multistakeholder dialogue to manage natural resource competition: a synthesis of lessons from Uganda, Zambia, and Cambodia

Rural development or natural resource management program planning and implementation frequently confront challenges of environmental resource competition and conflict, particularly where common pool resources are a major component of rural livelihoods. This paper reports on an approach to multistakeholder dialogue, supported by participatory action research, to address the roots of such competition and conflict. The approach, called “Collaborating for Resilience,” includes principles and guidance on building a shared understanding of risks and opportunities, weighing alternative actions, developing action plans, and evaluating and learning from the outcomes. Working in partnership with government, community and civil society actors, the approach was developed and refined through applications in large lake systems in Uganda, Zambia, and Cambodia. This paper presents a synthesis of lessons addressing practitioners in government, non-governmental development organizations, and international development agencies. These lessons include guidance on the context of multistakeholder dialogue processes, addressing gender equity, building stakeholder relationships and accountability across scales, and encouraging learning and innovation over time.


Introduction
Some degree of competition or conflict is intrinsic to natural resource management. In many places, community-based institutions have developed to manage local competition over the environmental commons; however, these institutions are typically inadequate to address more complex challenges involving diverse actors across multiple sectors and scales (Gruber 2010). Also, governance sys-tems at subnational, national, and international levels often lack appropriate mechanisms to ensure access to justice and public participation in decision-making about environmental resources (Ribot and Larson 2005;Layzer 2008). In the absence of such mechanisms, resource conflict can aggravate other social or economic divisions, contributing to broader social conflict. It can also undermine and reverse development gains in other areas, such as health, education, and nutrition.
While tools and approaches have been developed to assess the linkages between environmental resources and conflict, as well as to identify opportunities for peacebuilding through collaborative resource management, these primarily target use by external agencies in an expert assessment mode (Matthew et al. 2004;UNEP 2009). Much remains to be learned about undertaking collaborative assessments with local stakeholders and building on the insights gained to support institutional innovation and learning, including approaches that draw on and enhance existing, traditional conflict management processes (Sanginga et al. 2007;Ratner et al. 2013).
Taking these challenges and observations as a starting point, the Strengthening Aquatic Resource Governance (STARGO) project developed guidance and tools to support multistakeholder dialogue (Ratner and Smith 2014;Rüttinger et al. 2014) and adapted these through application in three freshwater ecoregions -Lake Victoria (Uganda), Lake Kariba (Zambia), and the Tonle Sap Lake (Cambodia) during 2012-2014. The STARGO project used the dialogue approach to develop institutional innovations aiming to build resilient livelihoods among poor, rural producers who depend on wetland and freshwater resources; generate gains in nutrition, income, welfare, and human security; and reduce the likelihood of broader social conflict.
Partners in all three regions used a common approach to stakeholder engagement and action research that we call CORE (derived from "Collaborating for Resilience"). In each ecoregion, partners assisted local stakeholders in developing a shared understanding of risks and opportunities, weighing alternative actions, developing action plans, and evaluating and learning from the outcomes. Whereas these outcomes and the implications for natural resource governance and development policy are detailed separately (see Ratner et al., under review), the present paper focuses instead on the processes of multistakeholder engagement and action research, synthesizing lessons from the three country cases (Burnley et al. 2014;Madzudzo et al. 2014;Oeur et al. 2014).
This paper aims to situate CORE within a spectrum of conflict management approaches, introduce its key attributes, and detail a synthesis of lessons drawn from its application in the three regions. While suitable as well for private companies and community-based organizations, these lessons primarily address practitioners in government, non-governmental development organizations, and international development agencies. The lessons concern: (a) investing in understanding prior patterns of conflict and collaboration by systematically comparing multiple perspectives; (b) building gender equity in the dialogue process, including working outside formal convenings to ensure all voices are heard; (c) linking actors across scales and strengthening accountability, including responsiveness of government authorities; and (d) learning to adapt, tap new support, and scale through structured reflection and joint learning.
The paper is organized as follows. In the next section, we present an overview of conflict management approaches, to specify the conditions under which the CORE approach is appropriate, when it is not, and the principles it shares with other approaches to negotiation and consensus building. Next, we detail the principles of the approach and summarize how it was applied in the three countries. The following section details lessons for practice in four areas, organized with reference to different elements of the 'action arena,' as specified by Ostrom (2005) and adapted by Ratner et al. (2013). In the concluding section, we highlight the distinction between participatory consultation and collaborative decision-making, and the importance of practitioners engaged in this work to exercise an appreciation of existing institutions and relationships, including questions of equitable representation and access to decision-making fora.

A spectrum of conflict management approaches
Natural resource management initiatives inherently address issues of competition over scarce resources. Initiatives that envision an important role for communitybased institutions alongside those of the state and private sector must necessarily be prepared to address organizational capacity in conflict management (Gruber 2010).
We use the term "conflict management" instead of "conflict resolution" because providing ultimate resolution to a conflict is often beyond the power of the actors involved, or not feasible given continuing changes that need to be addressed over time, such as changing global trade policies. Working toward fundamental change is also important but may require longer time periods and engaging actors at other scales. The broader concept of "conflict transformation" emphasizes transforming the relationships that support violence and conflict, along with the systems in which these relationships are embedded.
Effective engagement in resource conflict processes can be a foundation for improved collaboration, strengthened tenure security, and collective action (Dhiaulhaq et al. 2015). Conflict management, therefore, can be part of conflict transformation. Conflict management tries to maximize the positive and minimize the negative effects of a conflict. It is "the practice of identifying and handling conflicts in a sensible, fair and efficient manner that prevents them from escalating out of control and becoming violent" (Means et al. 2002). Compared with conflict resolution, which concentrates on solving an already existing dispute, conflict management also aims at prevention (Swatuk et al. 2008). This can be summarized in terms of three goals (Engel and Korf 2008): 1. Identify latent conflict potential and prevent it from turning into conflict. 2. Prevent existing conflict from escalating. 3. Manage conflict in a way that promotes positive social change.
Alternative conflict management approaches can be used as a complement to the formal legal system or traditional conflict management mechanisms such as those exercised by customary chiefs or local religious leaders, which often reproduce existing inequities (Ratner et al. 2013). They can also be used independently; for example, as part of a process initiated by a non-governmental organization (NGO). These alternative approaches are based on shared decision-making and try to include all affected stakeholders and groups. As such, they can empower local communities and build capacities for sustainable natural resource management. The goal is to reach a mutually acceptable agreement that creates long-term gains for all stakeholders.
Below we provide an overview of elements of alternative conflict management, presented as a spectrum from more intense to less intense levels of dispute among stakeholders. Although in practice the categories cannot always be separated as clearly as described here, distinguishing them theoretically highlights when and how to use these approaches, as well as how to identify the need for external support by third parties, including conflict management specialists (Engel and Korf 2005).

Conciliation.
Sometimes a conflict has already reached a state in which the parties are not willing to enter a conflict management process. This can be especially problematic in the case of alternative conflict management, which needs considerable goodwill by all stakeholders to be successful. In this case, conciliation approaches can be applied: A third party communicates separately with each party to reduce tensions, build confidence and create an acceptable process for conflict management (Means et al. 2002). Conciliation can also help to identify negotiation incentives to bring stakeholders into the process. One very powerful incentive can be external actors that have influence to persuade stakeholders to participate. This role can be played by the facilitator or mediator or by an external actor such as an important national figure or celebrity.

Mediation.
Mediation is the preferred approach if stakeholders are willing to discuss their positions, interests, and needs, but need support to engage in working on a mutually beneficial solution. For example, not all stakeholders may feel confident if there are substantive authority and power differences. Accepting or seeking third-party intervention is also easier and more likely if it is sanctioned by society through formal laws or informal traditions and if interventions in the past have been seen as successful (Wall et al. 2001). Mediators, like facilitators, ensure that the stakeholders agree to the process and logistics. Unlike facilitators, they can have considerable influence in bringing conflicting parties to the table and actively put forward their own ideas and views. In some cases, members of one of the parties in conflict may have sufficient capacity and legitimacy to play a mediation role (Dhiaulhaq et al. 2015).
Negotiation and consensus building. Negotiation is the most common form of conflict management. In a negotiation process, all stakeholders voluntarily search for a solution that is both mutually acceptable and leads to reduced conflict potential (Warner 2000;Means et al. 2002). In alternative conflict management the goal is not just to reach an agreement, but to find a solution that benefits all stakeholders, creating an interest in sustained collaboration. Negotiation can take place with or without a facilitator, but facilitation is part of most participatory approaches. Facilitators focus on supporting the process and logistics of bringing the different participants together. If they act as moderators, they focus on improving communication between the stakeholders, focusing the discussion and ensuring an equitable exchange of views.
One key role of a facilitator is to build trust between the stakeholders. This can be done by clarifying interests and assumptions, establishing a mutually defined system of accountability, and making trust an explicit discussion topic; for example, by assessing together the consequences of breaking trust or discussing how trust can be built as part of the negotiation process. This often takes time. Sometimes local organizations such as NGOs have already built up trust and can act as neutral facilitators, if they are not too involved in the conflict. This is where the CORE approach fits -an example of a facilitated approach to multistakeholder dialogue and action. It is not meant as an approach to intervene in active, violent conflicts, nor to mediate between opposing groups who are unwilling to meet in dialogue and explore options for the future. In such circumstances, other approaches such as conciliation or mediation are needed (Rüttinger et al. 2014). The essential precondition for CORE is a willingness on the part of key groups to meet in dialogue.

Process of stakeholder engagement and action research using the CORE approach
Building on several decades of prior learning in the fields of conflict management and participatory development (Ratner et al. 2014b;Rüttinger et al. 2014), the STARGO project set out to develop and apply a common approach to stakeholder engagement and action research. This section introduces the principles of the CORE approach, compares its application in the three country cases for participatory dialogue and action planning, and then describes the processes of monitoring, evaluation, and learning aimed at building further capacity to manage and transform local resource conflict.

Principles of CORE
Collaborating for Resilience, or CORE, provides a framework for understanding stakeholder interactions and organizing for social and institutional change. This framework is distinguished by its emphasis on whole systems, by an open search for solutions and by its explicit treatment of power. These characteristics make the approach especially well suited to catalyzing collective action to address shared challenges of natural resource management.
The principles outlined here draw on three decades of experimentation and practice in a diverse range of settings. The CORE approach is built around a more general framework for understanding the relationship between purpose, power, and organization. This framework is known as AIC, which stands for "appreciation, influence and control" (Smith 2009). The underlying concepts from AIC as applied in the CORE approach are the following: Multiple levels of purpose. Conflict is usually focused on the immediate goals and interests of competing groups. In their highly-influential guide to "principled negotiation, " Fisher et al. (2011) distinguish between positions and interests. But how are interests derived? In extending the frame of dialogue beyond the scope of a particular dispute and opening up avenues for collaboration, the CORE approach focuses on identifying multiple levels of purpose. In addition to immediate goals, these are expressed through values -how we believe we should relate to one another; and through ideals -how we ought to live or an image of a positive future. Identifying new opportunities for collaboration and collective action requires stepping back to explore commonalities and differences at these higher levels of purpose.

Multiple dimensions of power.
Power is often understood to mean control or decision-making authority -"power over" others. Yet the ability to influence others and engage in joint efforts toward a common purpose is also a form of power -"power with" others. As Gaventa (2006, 24) notes, "power 'with' refers to the synergy which can emerge through partnerships and collaboration with others, or through processes of collective action and alliance building." A third level of power is the power to understand or appreciate the context in which we live, the perspectives of others, and new possibilities -a "power of awareness." This is what Gaventa (2006, 24) describes as a "precondition for action." Multiple levels of organization. Addressing any resource management challenge involving competition among diverse groups requires a whole-systems approach -an understanding of the perspectives of individual groups, their interactions, and the broader setting of institutions, governance arrangements and other factors that influence their choices. Bringing all key stakeholders into the process ensures that multiple perspectives are represented. The CORE approach integrates insights from systems thinking in organizational development (Senge 1990) and socialecological resilience (Folke 2006).
The approach aims to transform stakeholder relationships in ways that promote collaboration, learning, and resilience. In a nutshell, the principles (Figure 1) can be understood in terms of purpose, people, and process: 1. The CORE approach is purpose-driven. Collaborating for resilience requires transforming social relationships. The most fundamental condition for transformation is clarity of purpose. There is a tension in finding a purpose that is broad enough to bring all the key players to the table, yet specific enough to address real needs and motivate action. 2. People make the CORE approach work. In preparing for an initiative, organizers actively seek out the participation of key people from a wide range of stakeholder groups. In conditions of natural resource competition, this means going beyond a particular sector to address the root causes of the problem, potentially bridging several geographic and institutional scales. 3. The CORE process aims at continuous development of institutional capacity to address the roots of resource competition and build resilience. While the principles of the approach can be used in small planning meetings or large, multiday dialogue events, the premise is that complex challenges require multifaceted responses over time. This means that action, reflection, and learning from experience are embedded in the process.
The CORE approach provides a set of orienting concepts, principles, and practices that different groups -including civil society organizations, development agencies, and governments -can adapt to the socio-cultural context and particular challenges at hand. [For a detailed treatment of the CORE approach, see Ratner and Smith (2014)]. For the STARGO project, the process included several months of scoping in preparation for a sequence of multistakeholder workshops. These workshops, while adopting different tools, followed a common format broken into three phases (Figure 2), roughly equal in time: 1. Building a shared awareness of the issues, the possibilities for the future, and the constraints and opportunities of the current situation (the listening phase). 2. Debate over different possible courses of action to pursue a common purpose, including an assessment of the groups that may support and oppose such actions (the dialogue phase). 3. Deciding on an action plan comprised of commitments by individuals and multistakeholder teams, including a reflection on the degree to which these actions will achieve the common purpose (the choice phase).
The CORE approach shares attributes with collaborative learning (Daniels and Walker 2001), designed to address situations of complexity and controversy in natural resource and environmental decision-making. The approach "encourages people to learn actively, to think systematically, and to learn from one another about a particular problem situation" (Walker et al. 2006, 195). The collaborative learning workshop approach follows three stages: developing common understanding; probing relationships among actors and their differing concerns and perspectives, along with potential improvements; and, lastly, debating desirable improvements and setting action plans (Walker et al. 2006, 195). Also in alignment with the CORE approach, collaborative learning "values emerging consensus, but is not consensus-driven (Walker et al. 2006, 196). Guidance on the CORE approach (Ratner and Smith 2014), as well as a suite of tools for use in assessment, planning, monitoring, and evaluation (Rüttinger et al. 2014), were developed in advance of initiating the multistakeholder dialogue processes in each case study site, then adapted on the basis of learning from these cases. The following summaries give an overview of how the process was adapted to each local setting.  Ratner and Smith (2014)].

Applications in each ecoregion
Stakeholders in Lake Victoria were familiar with participatory mobilization meetings, but were not acquainted with multilevel dialogue processes, especially around fisheries governance. Therefore, the STARGO team spent time explaining the broad concepts of dialogue, ownership, and agency behind the CORE approach. To address power imbalances among stakeholders, the team organized a preparatory workshop to give community participants the opportunity to make their voices heard and to enhance their capacity to engage other stakeholders. This was followed by a multistakeholder workshop bringing together government representatives from various levels alongside representatives of three lakeshore and island communities, and later, smaller meetings to review progress.
In the Lake Kariba region, reviews during scoping showed that previous initiatives were limited in large part because of biases toward sectoral interests. STARGO's first workshop mobilized all the key stakeholders involved in the use and management of the lake to envision a desired future against the current realities. This was followed by a smaller workshop that focused on actions that artisanal fishers, on the lowest rungs of the social hierarchy, and other stakeholders like the Department of Fisheries could pursue to promote dialogue. The organizing team opted for a learning-by-doing strategy to foster a locally owned and locally driven approach to developing the capacity for co-management, including linkages with private investors.
In the Tonle Sap region, civil society groups have long contested the fairness of commercial fishing lots that skewed access to the fisheries in favor of a few powerful groups. Recently, state reforms suspended and then permanently cancelled commercial lots and required agencies to plan and implement changes to increase community-based management. However, civil society networks and a range of relevant agencies were poorly prepared to coordinate their efforts in response to the changing policy context. STARGO supported a lake basin-wide dialogue workshop, followed by local and provincial-level workshops focused on facilitating institutional innovations among communities in Kampong Thom Province.

Monitoring and evaluation
The monitoring and evaluation approach aimed not only to report on outcomes but foremost to foster learning among local stakeholders (see Figure 3). Therefore, the monitoring and evaluation systems were designed in a participatory manner, taking into account that most of the participating actors had little or no experience in the use of such tools. This involved documenting the theory of change underpinning each of the local initiatives selected for detailed monitoring, then developing associated output, outcome, and impact indicators to assess change.
It was important that these change indicators be defined in terms meaningful to local actors, including communities, non-governmental organizations, and government institutions. Indicators addressed change in individual attitudes toward members of other groups, as well as change in relations among groups. In addition to providing meaningful measures of progress to support local learning and adaptation of strategies to address conflict and cooperation, the monitoring and evaluation data helped researchers assess the effectiveness of dialogue processes and identify related challenges.
Monitoring and evaluation activities included structured approaches such as questionnaires, focus group discussions and individual interviews, and narrative descriptions of personal experience such as participant diaries. Research team members convened local stakeholders periodically to discuss and review findings as a means of validation and collective learning.

Lessons on building effective multistakeholder dialogue for equitable resource management
The CORE guidance on multistakeholder dialogue helped foster collaboration under difficult circumstances in a range of socio-political and ecological settings, demonstrating the value of the underlying principles. As a result of dialogue and action research processes, the project helped launch a range of locally-driven institututional innovations addressing community-based fisheries co-management, commercial aquaculture investment, resource protection, and public health. Significant outcomes, detailed separately in Ratner et al. (under review) included: improved attitudes toward collaboration and heightened dialogue among community groups, non-governmental organizations and government; new and successful engagement with private investors delivering negotiated agreements to protect local resource access rights and livelihoods; influence on government priorities in addressing the needs of local communities; and new sources of support secured from national and international agencies to scale out innovations.
In addition to assessing and validating outcomes from the individual cases, the project also convened researchers and practitioners from the three case study regions to meet in Uganda for a joint exercise aimed at systematically comparing and synthesizing lessons of broader relevance for policy and practice. The following lessons are oriented toward field-level practitioners in government and civil society working with diverse stakeholders to build collaboration in order to better manage resource competition and increase local livelihood resilience. The lessons emphasize the importance of adapting the general approach to specific local conditions, taking into account new obstacles and opportunities as they emerge.
The lessons are organized below with reference to different dimensions of the 'action arena' as presented in the conceptual framework on resource conflict, collective action, and social-ecological resilience proposed by Ratner et al. (2013), building on the institutional analysis and design (IAD) model (Ostrom 2005;di Gregorio et al. 2008). These lessons concern: (a) understanding the prior patterns of conflict and collaboration as an input to the design of multistakeholder dialogue processes; (b) recognizing the distinct characteristics of different actors and their unequal action resources, with a focus on gender equity in the dialogue process; (c) linking actors across multiple levels and strengthening mechanisms of accountability; and (d) incorporating learning on the outcomes of dialogue processes to adapt, find new sources of support, and pursue broader changes at scale.

Understanding prior patterns of conflict and collaboration
Understanding stakeholders' prior experiences with conflict and collaboration can provide important insights that help shape the approach to dialogue. In Lake Victoria, the team responsible for designing the dialogue process recognized that multiple pre-existing conflicts had led to strongly negative attitudes on the part of community members toward government officials, including Department of Fisheries officers. Expecting that community members might be hesitant to participate openly and confidently, the team organized a separate preparation workshop with community members prior to the main workshop. This preparation helped them become some of the most active participants during the larger dialogue workshop. In Lake Kariba, the research team noticed that participants repeatedly praised the effectiveness of this dialogue forum and subsequent action planning compared to other platforms for presenting their interests. Since commu-nity judgments of effectiveness are constantly reassessed and can change quickly based on perceptions of progress, the team focused on early achievements to build confidence.
Intentionally comparing multiple perspectives can help to maximize learning. At times this required skillful facilitation, paying attention to differences in views, and exploring where these differences came from. In Lake Victoria, personal interviews were arranged around the working schedules of fishers (mostly men) and fish processors and sellers (mostly women) in order to access all possible perspectives. In Tonle Sap, researchers found that local residents not directly engaged in the innovations sometimes had the most valuable insights as relatively impartial observers. They also found that, because a number of related activities were ongoing with support from different outside groups, it was important to take the time to clearly distinguish what actions people were evaluating, and what were the sources of changes they described.
A quality dialogue process requires flexibility to build stakeholder engagement. One measure of a quality process is the authentic ownership that participants develop in their action planning and implementation. Simple decisions such as conducting the workshops in the local language helped encourage active participation, even if it meant outsiders had to adapt. In Lake Victoria, when community representatives returned to their villages after the main workshop, two of the three sites ended up changing their plans. Bringing together three communities encouraged people to reflect, to compare their experiences, and to rethink their priorities. In an unusual show of local commitment, villagers in Kachanga raised money from within the community for building materials, got district council approval for building the latrine and biogas facility, and secured a commitment from the leader of the district government to provide trucks to transport the building materials. The sense of shared purpose brought in additional supporters.

Building gender equity in the dialogue process
Recognizing the distinct characteristics of different actors and their unequal action resources is essential to creating a platform where each has an opportunity to have their voices heard and influence collective decisions. The STARGO project demonstrated how observing gender inequities and other power imbalances can lead to creative adaptations to include all voices. In Uganda, a system of quotas is in place to make sure that less powerful stakeholders, such as women, boat crew, and other fish workers, are included in decision-making bodies such as beach management units. However, during initial community consultations it took several tries and some creative childcare arrangements to find a small number of women to participate in the stakeholder workshop. During the workshops, women and boat crew members rarely spoke or suggested actions unless they were specifically asked. By contrast, male boat owners were very outspoken. Women were also in the minority of those nominated to participate in capacity-building actions. Recognizing the gap between an official policy of inclusion and typical processes of decision-making biased toward men and economically privileged groups propelled the research team to seek out ways to address these imbalances.
A range of informal consultations can help reveal unspoken concerns. Having observed the gender dynamics in the workshop setting, the research team in Lake Victoria held additional side meetings where the more reserved participants could express their concerns. These concerns were subsequently validated in the full dialogue, shifting the focus of planning toward community sanitation. Likewise, in Lake Kariba, the team found that even when women were less vocal in the workshop, facilitators were able to actively seek out their concerns, making sure they were heard by all participants. In the Tonle Sap region, where workshop organizers lacked long prior experience in the selected communities, team members undertook several days of informal consultations, including meetings with small groups of women and men separately, before convening a more structured dialogue event.
Supporting individual change agents can also lead to more lasting institutional change. In Lake Victoria, the team was able to identify individual women representatives from the beach management units and local councils who were particularly active in the early stages of the initiative. The team then found ways to encourage them in the role of change leaders in supporting the community-led activities and in getting other community members involved. On the Tonle Sap Lake, the team found that the participation of a former commercial fishing concession operator proved pivotal when, after suspension of the commercial lots, she committed to helping the community explore different management regimes.

Linking actors across scales and strengthening accountability
A quality dialogue process not only links actors across various social groups but also across scales of governance. Addressing local disputes often requires support from higher levels of administration. The village management committees in Zambia, beach management units in Uganda, and community fishery organizations in Cambodia each faced similar challenges in accessing higher-level support to help resolve local disputes. Many prior efforts at building community-based management institutions focused on local-level organizational capacity in relative isolation, presuming higher-level administrative structures would pass down resources and lend assistance as required. In Lake Kariba, the research team found that involving the Department of Fisheries and Environmental Management Agency at each stage in the process lent legitimacy to local actions. This involvement also helped build linkages so that local change agents could have a voice in longer-term policy, institutional and legal reform. Special efforts were also required to bridge communication gaps, including recruiting a trainer on environmental impact assessment procedures who was originally from the Kariba region to explain key concepts in the local language.
Successful examples of collaboration can help strengthen mechanisms of accountability over time. As a result of local actions in the Tonle Sap floating villages of Peam Bang and Phat Sanday, the commune councils became supportive of joint patrolling. This strengthened relationships that are helping community fishery committees seek support for the more difficult task of piloting the community-based commercial production model. In Uganda, local actions to improve community sanitation attracted interest from government actors at different levels. The district council's public commitment to assist in maintaining the facility provides community members with a point of reference to hold the council accountable in the future. Noting the strong local leadership and commitment, transparency in decision-making and fund management, and timely implementation, the Masaka district head noted that the sanitation improvement project "set new standards of quality… the district would emulate for future projects." In applying the CORE approach, the research team found it necessary to question assumptions about stakeholder roles. In Lake Kariba, the team assumed that the government's role in fisheries management would be focused on surveillance and enforcement. The team was surprised, therefore, to find that community members felt the Department of Fisheries needed to be present in addressing other issues, such as discussions with the traditional chief about the approach to engaging investors. Indeed, though the chief was proximate, community members felt the fisheries officers could play a critical brokering role and lend legitimacy to the process. In Cambodia, the research team initially tried to hold to the principle of equal roles among partners in planning the initiative, then recognized that having different leaders for different activities was appropriate. Therefore, in the concluding policy dialogue forum, the Fisheries Administration played the convening role, while research partners facilitated the event.

Learning to adapt, tap new support, and scale
Effective dialogue can settle disputes before they escalate. In Lake Kariba, investors in cage aquaculture and lakeshore tourism development proved much more willing to cooperate than community members and organizers expected. Local villagers realized that competition among investors meant they were eager to show good will to communities, resulting in spoken agreements to ensure routes of travel on water and land, and to safeguard local employment. This showed it was possible for local communities to engage with investors and build some measure of accountability without resorting to adversarial legal processes. Investors also commented that they would be more likely to request this sort of dialogue in the future as a way of avoiding deteriorating relationships.
Critically assessing past experience at the start of an initiative can help partners avoid repeating past mistakes. As part of the appreciation stage of the CORE process, establishing a shared understanding of prior and ongoing efforts can help prompt reflections on past experiences. For example, when different communities in Lake Victoria jointly reviewed past development efforts, they identified misuse of funds as a recurrent problem. Alert to this risk, the organizing committee in Kachanga at one point halted payments to the contractors hired to build the sanitation facility and opened the accounts to public review. After investigation, the allegations of corruption proved unfounded, yet the action sent a critical message by demonstrating the commitment of local leaders to transparency and accountability in the use of funds.
Structured reflection during implementation is critical. Reflection activities need to be focused, yet flexible. In all three ecoregions, research teams found it challenging to organize community members to record detailed information about activities, such as the number of meetings held with various government groups, or in the case of Tonle Sap, the number of joint patrolling trips undertaken. Few community members found this information helpful in evaluating progress. In Cambodia, the team therefore shifted to focus on broader questions: "What changes have you seen since the last period? What do you see as the obstacles remaining?" These yielded very rich stories, and helped launch discussion about ways to adapt that would help achieve local goals. Similar reflections in Lake Kariba helped community members and local leaders learn what approaches worked in engaging investors.
Engaging local actors and government planners in joint learning is essential to scaling out innovations. In all three ecoregions, there was evidence that community groups and partners had adopted practices from the dialogue approach and were applying these to access new channels of support or scale out local innovations. In each of the ecoregions, authorities were well aware of the international support behind the dialogue and action planning processes, and researchers agreed this brought additional attention to the outcomes. To sustain such reflective learning over time, particularly in the absence of international support, communication channels that directly link community actors and government officers at higher levels are especially important. While decentralization policies may promote local authority and initiative in principle, these cases show that there is often a need to overcome barriers to effective communication. This includes creating a safe space within dialogue processes for perspectives that are critical of government performance, as well as cultivating a readiness on the part of government agencies to learn from local initiative in revising policy goals or implementation strategies.

Conclusion: from conflict management to collaborative governance
As demands on common pool natural resources increase in response to both local livelihood needs as well as distant markets, the practice of effective, multistakeholder dialogue can play an essential role in both reducing conflict risk and strengthening equitable governance institutions. Yet, too often, even where there is a willingness on the part of government, investors, and communities to engage in managing resource competition, the rush to reach negotiated settlements sacrifices the broader benefits of an open dialogue process (Silva-Castañeda 2015). These additional benefits can include reshaping stakeholder relationships in ways that extend beyond the immediate issues of dispute, mutual learning, and reduction in future conflict risk (Calton and Payne 2003;Poncelet 2004).
While it is too early to assess longer-term effects, outcome evaluation of the cases reviewed in this paper provide initial evidence of conflict transformation at the local level, as defined through criteria of changes in attitudes, behavior, and conflict intensity (Augsberger 1992). Yet, as analysis of conflict management efforts in the forest sector has also found, addressing the more persistent roots of conflict "requires long term engagement and trust building with the government as policy makers and NGOs at sub-national and national levels" (Dhiaulhaq et al. 2015, 140).
A more structural perspective on conflict transformation emphasizes such underlying policies, institutions, and actors beyond the conflict site (Lederach 1997). The intent of the CORE approach is that, by increasing a shared appreciation among local actors of the possibilities as well as the barriers to change, dialogue can help identify avenues for further joint action at broader scales. This can include engaging new domestic or international actors, and working through new routes of formal or informal policy influence ). In the terms of Gaventa (2006), this can mean moving beyond the "invited" spaces of power to those that are "claimed" or "created," or working to build transparency or challenge barriers to participation in closed decision-making fora. A structured multistakeholder dialogue process like CORE cannot achieve these changes, as by definition it is applicable to cases in which stakeholders are willing to convene, but it can help reveal the possibilities for additional collective action.
"Small wins" through collaborative dialogue processes such as those described here can also build relationships that improve trust and mutual understanding as a basis for more institutionalized patterns of "collaborative governance" over time (Ansell and Gash 2007). Well-structured dialogue can create new patterns of interaction that, if institutionalized, shift the 'rules-in-use' structuring future patterns of conflict and collaboration. Even in the absence of policy shifts, these changes in stakeholder relationships and institutional capacities can measurably improve governance arrangements. Indeed, this possibility of identifying and contributing to changes that no single actor could achieve alone -power 'with' others, or what Innes (2004) terms "network power" -is a key reason why groups often stay engaged in the face of daunting obstacles. "Creation of this power can be one of the most potent incentives for participants to stay at the table and continue to work together even after the immediate project is completed. Network power is the glue for collaboration over time" (Innes 2004, 13).
As the cases summarized in this paper illustrate, proactive efforts to convene dialogue to address the roots of resource competition can help generate new forms of collaboration among civil society, private sector, and government stakeholders at multiple levels. Too often, "participation" in the design of development projects or resource management policies means little more than consultation with intended beneficiaries on problems and needs, as opposed to shared decisions on priorities and action plans (Haider 2009).
In promoting collaborative decision-making, the CORE approach puts the burden on those organizing multistakeholder interactions to develop an appreciation of existing institutions and relationships, including the inequities in representation and power that must be countered in longer-term efforts at governance reform and conflict transformation.