Barriers to equity in REDD+: Deficiencies in national interpretation processes constrain adaptation to context

A national interpretation process involving diverse actors and interests is required to transform global environmental initiatives into policies appropriate to the national or subnational context. These processes of localising norms are critical spaces to formulate equitable pathways to environmental conservation, yet have received limited attention from policy makers and researchers. We explored national policy processes for Reducing Emissions from Deforestation and Degradation (REDD+) in Uganda and Nepal from the perspectives of ‘intermediaries’, state and civil society actors at subnational and national scale who promote the interests of various stakeholder groups. Through think-tank meetings and semi-structured interviews with a range of intermediaries, we uncovered that REDD+ implementation processes in both countries are dominated by international actors, applying a demanding administrative agenda and restricting space for deliberation. Consequently, social aspects of policy were compartmentalised, reduced to technical exercises and local equity concerns inadequately addressed in national REDD+ policies. For example, social safeguards approaches were perceived to lack substantive guidelines to promote equity. Limited national political space to criticise government policy and lack of attention to relevant evidence further restricted ability to address entrenched injustices such as status inequalities faced by marginalised groups. Although civil society organisations choose to maintain official involvement with REDD+, many expressed a possibility they would oppose REDD+ in future, or serious doubts about its design and expected outcomes. Concerns centred on lack of recognition of indigenous peoples’ and local communities’ values, identities, practices and institutions such as customary tenure systems, alongside possible detrimental impacts to decentralised forest governance regimes, well established in Nepal and emerging in Uganda. We suggest features to be enshrined in REDD+ policy for adapting national interpretation processes to become more effective spaces for empowering diverse intermediaries to negotiate and influence localisation of international norms, ultimately to promote more equitable pathways to reduced deforestation and degradation.


Introduction
Transforming internationally-conceived environment and development goals into mechanisms and policies that are perceived to be fair, or equitable, by local people is a major challenge. The Sustainable Development Goals, the Paris Agreement and Aichi Targets of the Convention on Biological Diversity are prominent global policy initiatives which explicitly target social equity as an important goal alongside environmental conservation. Social objectives of environmental policy are essential for moral reasons, with the minimum goal being to avoid imposing risks and harms. The pursuit of equity (comprising three dimensions: distribution of costs and benefits, decision-making procedures and recognition of diverse identities and values) is also increasingly acknowledged as crucial to gain the widespread support and compliance required to attain ecological goals, and in many cases to address drivers of environmental degradation (McDermott et al., 2012b;Myers et al., 2018;Schlosberg 2013). To move towards implementation, global environmental initiatives commonly involve a process for in-country actors to transform the internationally-determined elements into policies acceptable to diverse stakeholders in the national or subnational context. We term these national interpretation processes (after Okereke, 2007;Peskett and Brockhaus, 2009) and pay attention to the debate and prioritisation of norms within them, or how people feel things should be and the way things should be done (Walker, 2012). A key role in national interpretation processes is played by 'intermediaries', comprising a range of subnational and national, state, civil society and private sector actors with diverse objectives, who influence whether and how international norms fit with local norms and practices (Acharya, 2004;Bratman, 2014;Finnemore and Sikkink, 1998;Keck and Sikkink, 1999;Lewis and Mosse, 2006;Stovel and Shaw, 2012).
Although a crucial arena and scale of policy negotiation, national interpretation processes related to global environmental or climate governance have been the subject of limited scholarly attention. We use processes for debating and formulating approaches to Reducing Emissions from Deforestation and Degradation (REDD+) in Nepal and Uganda as examples, and REDD+ Working Groups as the institutional loci for the actions of state and CSO intermediaries. While numerous studies have addressed which stakeholders participate in REDD+ discussions at national level, few have examined the nature of their participation and potential to influence the prioritisation and interpretation of equity-related norms (Bernstein and Cashore, 2012;Okereke and Dooley, 2010). Nepal and Uganda are both advanced in administrative processes and entering the REDD+ implementation phase, but represent different political environments or opportunity structures for intermediaries to function within. We explore the perspectives of actors involved in REDD+ working groups in Nepal and Uganda to highlight some of the factors enabling or restricting attention to and localisation of equity-related norms and the pursuit of diverse social interests.

REDD+ national interpretation processes: overlooked forums for achieving equitable REDD+
Approximately 70 countries are formulating approaches to REDD+, with many having begun or approaching implementation (Bayrak and Marafa, 2016). Since the inception of REDD in 2006, details of this emerging mechanism have been subject to ongoing negotiation in annual United Nations Framework Convention for Climate Change (UNFCCC) meetings, with notable progress reached through the 2013 Warsaw Framework and the 2010 Cancun safeguards (Arhin, 2014). Due to the potentially profound impacts of changes in forest governance not only on forests but also on people living in or near forests, and the importance of the support of such people in determining environmental outcomes (Dawson et al., 2017;Pascual et al., 2017), equity concerns have increasingly formed part of REDD+ negotiations. Items such as benefit sharing criteria and social safeguards, including full and effective participation, transparency and respect for local knowledge and rights have therefore entered international and national policy debates (Krause and Nielsen, 2014;Okereke and Dooley, 2010;Turnhout et al., 2016). The 'full and effective' participation (both continual and with ability to influence decision-making) of non-state actors, including indigenous peoples, local communities, various social groups and the civil society groups representing them, in REDD+ readiness and implementation stages is considered essential for the legitimacy and effective design of REDD+, and is included as a 'safeguard' principle under both UN-REDD and the World Bank Forest Carbon Partnership Facility (FCPF) (Chhatre et al., 2012). Despite safeguards, the practical means to avoid harm or ensure positive outcomes for indigenous people, local communities and vulnerable people is still a major area of contention (Evans et al., 2014;McDermott et al., 2012a;Suiseeya, 2017). Indeed, early impacts of REDD+ type projects suggest that negative impacts are likely to be experienced by some local inhabitants, through limited participation, disruption of livelihoods, institutions and social systems, impacts on food security and land tenure, with powerful actors capturing most of the benefits (Bayrak and Marafa, 2016). REDD+ policies must be debated at national and subnational level to support formulation of policies compatible with existing regulatory frameworks, suitable for the promotion of effective emission reductions and to uphold relevant principles of equity at the relevant scale of implementation . The importance of national interpretation processes is recognised by the UN and other international agencies: "The UN-REDD Programme supports nationally led REDD+ processes and promotes the informed and meaningful involvement of all stakeholders, including indigenous peoples and other forest-dependent communities." (http://www.un-redd.org/, accessed 21/11/16). Yet, beyond considerations of which stakeholders are represented on REDD + committees and cursory attention to the relative power of civil society organisations, the dynamics of REDD+ national processes and their implications for local outcomes have received limited critical scrutiny by researchers or policymakers (Bastakoti and Davidsen, 2017;Corbera and Schroeder, 2011;Ravikumar et al., 2015). There are several reasons to consider whether national interpretation processes may address local equity concerns. Firstly, although commonly assumed nationally-led processes, they include extensive involvement of international organisations (such as the World Bank, UN and donor agencies ). Second, the processes are often guided by extensive, externally-designed administrative procedures requiring detailed, resource-intensive consideration of technical issues, notably monitoring, reporting and verification (Ravikumar et al., 2015;Romijn et al., 2015). As a result, instruments for addressing equity, such as social safeguards, are commonly reduced to administrative monitoring and reporting exercises rather than meaningful debates about the politics of justice within implementing countries (Krause and Nielsen, 2014;Myers et al., 2018;Schroeder and McDermott, 2014). Through this technocratic lens, negotiations can become detached from wider national issues regarding land and forests, leading to weak integration with other sectors and relevant drivers of change (Corbera and Schroeder, 2011). Finally, the political space within implementing countries may be dominated by the state or other political actors such that the ability of civil society organisations (CSOs) to be critical or raise questions of rights and justice for marginalised groups is often constrained (Chhatre et al., 2012;Mason, 2010;Thompson et al., 2011). Host government repercussions against CSOs for voicing dissent can be severe, and such expression of state power can shape both the composition and functioning of civil society (Scholte, 2011).

Case studies
Nepal and Uganda form illustrative case studies for understanding processes of national interpretation. Both countries have participated in REDD+ readiness activities since 2009 and are progressing towards implementation. Both have high levels of rural poverty and livelihood dependence on natural resources, which are particularly acute among minority social and ethnic groups. Dalits and indigenous people together make up approximately half of Nepal's population and their participation in, and impacts of policy upon them, are key issues in land and forest governance (Paudel and Vedeld, 2015). Nepal's forest sector implemented community-based forest management in the late 1970 s with 20% of forests now under such governance (Larson et al., 2010). In contrast, in Uganda forest governance comes under state control to a large extent but with extensive customary tenure prevailing, leading to sometimes overlapping management regimes (Naughton-Treves et al., 2011). There are several marginalised social or ethnic groups inhabiting forests or land adjacent to them, including the Batwa and Benet, who have suffered severe impacts of forest conservation in Uganda and who may be considered to justify specific attention within the REDD+ process (Cavanagh, 2012). Civil society is active in both countries but freedom of expression is constrained or unequal between social groups (Bhandari, 2014;Robinson and Friedman, 2007). Nepal has faced longterm political instability, including a ten-year civil war (1996)(1997)(1998)(1999)(2000)(2001)(2002)(2003)(2004)(2005)(2006) and frequent changes of government since becoming a multi-party democracy in 1990, whereas Uganda has remained under the leadership of Yoweri Museveni since 1986.
In Nepal the national REDD+ Implementation Centre was established as the co-ordinating state agency under the Ministry of Forestry and Soil Conservation, and a REDD Working Group comprising a variety of state and non-state actors was formed, in January 2009. Amendments to the Forest Act 1993 were proposed in January 2010. The Readiness-Preparation Proposal (RPP) was submitted to the World Bank's Forest Carbon Partnership Facility (FCPF) in September 2010 and Nepal's REDD+ Package was endorsed by FCPF in Sep 2016. CSOs from conservation and social development sectors have substantial involvement in the process (Paudel and Vedeld, 2015). The Implementation Centre aims to fit REDD+ into existing community forest governance, although, in addition to providing opportunities for community forestry, REDD+ may also be considered a threat to community forestry arrangements through recentralization of control (Bushley, 2014). The REDD+ Implementation Centre has undertaken several pilot projects, and drafted an Emissions Reduction Program Document to guide development of an initial phase of projects entrusted to WWF Nepal, focusing on 12 districts in the Terai Arc Landscape.
In Uganda the national REDD+ working group was also set up in 2009 with broad civil society and state representation, initially under guidance of the National Forest Authority. Due to governmental reorganisation, the Ministry of Water and Environment took over and a new working group, the National Climate Change Advisory Committee, was established under its Forestry Sector Support Department. In 2011 Uganda submitted its RPP to the FCPF. Although the RPP was assessed and recommendations made to improve it, the process then stalled until 2014 due to staff changes within the World Bank and the need to amalgamate FCPF and UN-REDD approaches, particularly regarding safeguards. UN-REDD then began to lead the process with support from the Austrian Government. Uganda's REDD+ process involves a wide variety of civil society organisations (Peskett et al., 2011). The few REDD+ type projects which have been piloted have involved diverse designs, managed by conservation NGOs including Ecotrust, the Wildlife Conservation Society and the International Union for the Conservation of Nature.

Conceptual framework
We view processes of policy formation and implementation (in this case REDD+ national interpretation processes) as being complex, multi-scalar and dynamic, comprising numerous actors and organisations with multiple objectives and notions of justice (Bierschenk, 1988;Blaikie and Muldavin, 2014;Mosse, 2008). Through this lens, policy debates are shaped by ideas about the way things 'should' be, or norms. Norms are influenced by people who both generate and receive them, and those that act as intermediaries between them. Building on Acharya's (2004) constructivist argument about the major role of "insider proponents", we identify intermediaries, in similar terms, as those subnational and national actors attempting to reinterpret and reconstitute externally received norms to find congruence with local norms and practices. Conceptually, this approach means that outside norms, such as those included in international policies, are not simply diffused down to national and local scales (Finnemore and Sikkink, 1998;Keck and Sikkink, 1999); rather, sub-national and national actors, often mobilized in specific environmental and social struggles, have the potential to influence global environmental norms and debates from below (Blaikie and Muldavin, 2014;Upton, 2014;Wolford, 2003). However, the extent to which they may realise policy influence is enabled or constrained by the institutional context, including power relations and interactions with other intermediaries. Nor should it be assumed that intermediaries act selflessly and consistently in pursuit of social justice for local actors. Rather, they may pursue multiple economic, environmental or social causes for different stakeholders at various scales, as well as prioritising their individual or organisational status, network-building or tactical positioning in different forums (Finger and Princen, 2013;Routledge et al., 2006).

Study methods
We convened participatory workshops, or think-tank meetings, over two days in Kampala and Kathmandu in August 2016. In Nepal, 23 participants represented 17 organisations (approximately half of the 34 policy actors in REDD+ processes ): 12 diverse CSOs; the state agency responsible for REDD+; two academic institutions; one private consultancy and; a national newspaper. In Uganda 17 participants represented 14 organisations (approximately 42% of organisations listed as part of Uganda's REDD+ Working Group (Namirembe, 2011): 10 diverse CSOs, the Uganda Wildlife Authority, National Forest Authority and two relevant academic institutions. The meetings were facilitated by the authors as open discussions, drawing on participant expertise to deliberate over equity issues relevant to REDD+ and means of addressing them.
We then conducted a semi-structured interview with each of seven individuals in both Nepal and Uganda who represented intermediary organisations in REDD+ processes. Participants were selected to include state and non-state organisations representing a range of equity issues, including gender equality, indigenous people and local communities, human rights, land tenure, commercial timber production, climate change, forest conservation and community forestry, as well as those operating at different scales, from grass-roots activism and project implementation to national-scale advocacy. In Nepal representatives of six CSOs and one state official co-ordinating RIC were selected from the list of wider think-tank participants. In Uganda one state official involved in REDD+ implementation was selected, alongside five representatives of CSOs and one private sector representative. In addition to exploring their perceptions of REDD+ and related policy processes, interview questions addressed the different objectives interviewees and their organisations prioritised in their work, the various forums they operated in and types of actions and interactions they undertook at different scales to advance those causes as well as examples of successes and difficulties in influencing policies and local outcomes. Qualitative data from think-tanks and interviews were then analysed thematically to draw out different intermediaries' objectives and the factors perceived to enable or constrain the localisation of equity-related norms in REDD+ processes, with an iterative approach taken to identify commonalities and differences.

Intermediaries in REDD+ policy processes and their conceptions of equity
National REDD+ processes in Nepal and Uganda grouped together diverse intermediaries operating at multiple scales, from local to global. In both countries, participants expressed a range of priorities, from recognition of minority values, customary land tenure systems and inequalities of social status to biodiversity conservation, climate change mitigation and human resource capacity within the forest sector, suggesting broad participation within national REDD+ policy processes. Furthermore, intermediaries were not solely focused on one priority but instead described multiple, concurrently held social and ecological objectives. All participants professed to be proponents of both forest conservation and equity for rural inhabitants, and their principles were therefore well aligned on a basic level suggesting potential for REDD+ working groups to work as deliberative platforms for negotiating localisation of global norms to address equity issues. According to one interviewee from Nepal, "we all have a common interest in keeping trees. The question is how? And for that we need to recognise the rights of local communities and indigenous peoples." Despite broad agreement about the importance of pursuing social equity alongside forest conservation, contrasting understandings of equity and how to pursue it through REDD+ polarised participants. Think-tank discussions returned repeatedly to issues of recognition, human rights and linked procedural issues. During each two-day thinktank discussion, participants for whom recognition was a central pillar of equity recurrently articulated several equity concerns in both Nepal and Uganda: the disproportionate costs likely to be borne by groups such as Dalits, women, indigenous peoples and cultural minorities through reduced access, place detachment and overriding of traditional knowledge and institutions; elite capture of any entitled benefits, and; social and political barriers to their effective participation. In Uganda, participants highlighted the nationwide issue of overlapping formal and customary tenure systems for forest resource use and grazing, recounting numerous experiences whereby externally-driven projects undermined the rights of local people, particularly the poorest, in favour of more powerful, often distant and politically connected holders of property rights. As one Ugandan think-tank participant explains, "we should consider an inventory of multiple rights under different arrangements. Other countries have taken that path under equity or fairness. Tenure systems are different in Mbale (Eastern Uganda) compared to even here (Kampala)." In Nepal, participants stressed that inclusion criteria in forest policies, to ensure marginalised groups participate and receive benefits, were meaningless because their quality of participation and influence is negligible and benefits commonly diverted or inappropriate to offset the substantial detrimental impacts on their values and way of life. According to one Nepalese think-tank participant, "when your identity gets lost, you have no dignity. When we are not allowed to enter the forest, then our knowledge, skills and practices are refrained from getting handed over to the next generation and knowledge is lost forever. These are some of the key issues and agendas we have been raising because conservation and protection is for every human being, so let's respect and recognize human rights." In contrast to the emotional pleas to acknowledge entrenched injustices faced by marginalised groups, others prioritised the protection of forests or expressed a view that the processes of consent and distribution of monetary benefits to those groups were adequate mechanisms to deliver equitable outcomes and that the major barrier to implementation (beyond establishing reference carbon levels and monitoring methods) was simply to raise awareness among local communities about REDD+ and build capacity among relevant implementing organisations to establish participation and benefit sharing. As explained by a Nepalese interviewee, "the main guiding principle would be livelihood diversification to reduce dependency on the forests….. Second is the social safeguards for marginalised groups. We have to think about that as well so a percentage (of monetary benefit) will go for their welfare, also technical stuff for training them too." In summary, those representing marginalised groups and issues related to recognition of their values and practices were more sceptical of the potential of REDD+ processes to address equity issues, despite the presence of a diverse range of stakeholders on decision-making bodies like the REDD+ Working Groups.

International factors constraining effective norm localisation in REDD + national interpretation processes
Although actors involved in REDD+ processes appeared polarised regarding the extent to which recognition should be central to the definition and pursuit of equity in REDD+, some consensus was evident on procedural limitations attributed to international factors. As shown in Table 1, international factors were perceived to constrain the ability of national and subnational REDD+ policies in Uganda and Nepal to deliver equitable procedures and outcomes for local stakeholders.

International actors control 'national' interpretation processes
REDD+ processes, specifically meetings and activities of the REDD + Working Group in each country, were widely perceived to be unsuitable forums for debate through which global policy norms could be adapted to fit local realities. Indeed, international donors (including the international REDD+ authorities, UN-REDD and World Bank FCPF who fund the national REDD+ mechanisms and the work of many of the intermediary organisations involved, were perceived by 13 of the 14 interviewees, including government officials, to control the processes and agenda with minimal influence from CSOs and state actors. National intermediaries considered the terms, content, order and timeline of national policy processes to have been conceived in international spheres, with little opportunity to reflect on the norms guiding and contained within policy, their appropriateness and need for adaptation, thus compromising the influence of national actors on national REDD+ approaches and their local outcomes (Table 1).

Technocratic and prescriptive approaches to REDD+ foreclose negotiation of equity
The internationally-dictated agenda for REDD+ processes resulted in debate among REDD+ Working Groups being centred on how to meet predetermined milestones as part of a technically complex and burdensome administrative process, segmented into predefined topic areas. Participants in both countries, particularly those focused on recognition for minority social and cultural groups, voiced concern that key equity issues, including social safeguards, would only be addressed in detail much later in the process relative to items such as methods and capacity for measuring, reporting and verification of forest cover and carbon stocks (Table 1). Segmentation into topics such as participation and benefit sharing served to simplify complexities of social and political inequalities, rather than addressing or even debating the issues of recognition and political marginalisation which may underpin those inequalities.

Tick-box approaches to equity issues
When addressing individual components of REDD+ governance, many participants, including some of those holding relatively simple conceptions of equity, perceived the approach to have been reduced to a tick-box reporting exercise at a broad scale, designed to meet the requirements of funders and program managers rather than to deliver equitable outcomes for affected rural inhabitants. Upward accountability to donors was superseding downward accountability to affected communities. The proposed REDD+ safeguard reporting systems were considered ungrounded, abstract and inadequate to promote equity at the local level by many intermediaries, who suggested a need for more substantive, responsive support to detrimental impacts or conflicts arising. Think-tank participants urged assessment and integration of subnational or local institutions to help implement safeguards in practice. This view was in part guided by the profundity of entrenched injustices faced by local communities and indigenous people within their societies. Numerous think-tank participants, and 10 of the 14 interviewed expressed views that pervasive social injustices surrounding tenure, ethnicity, caste and class were unlikely to be addressed through a REDD+ project due to the reduction of complex social and political issues to national scale requirements for intermittent evaluation and reporting (Table 1). They also highlighted that extensive in-country expertise and evidence exists from decades of development and conservation projects for identifying and working towards social objectives in different regions, such as the establishment of national park buffer zones in Nepal, but was being overlooked in REDD+ processes.

Centralisation of governance and knowledge
Participants in think-tank meetings highlighted a risk that imposition of complex and technocratic REDD+ governance was resulting in  Table 1 International factors perceived to constrain the ability of national and subnational REDD+ policies in Uganda and Nepal to deliver equitable procedures and outcomes for local stakeholders. Quotes are taken from think-tank discussions and semi-structured interviews, though not attributed to specific individuals or organisations for ethical reasons. The final column presents potential actions to overcome barriers put forward by the authors.
Barriers to achievement of equity Supporting quotes from semi-structured interviews Potential actions for enhancement of processes International actors control 'national' interpretation processes "The involvement of World Bank is critical worldwide, not only here. I have put my thoughts many times about the involvement of World Bank. I refer to it as climate colonization, and that is happening here in Nepal. So how can we Nepali work to democratize the carbon business here?" "The stagnation (of REDD+ debates) is due to the international level, what they bring and how they put all these burdens on the lower national level people. It becomes absolutely impenetrable at this high level and never gets going.….. People are not challenging that hierarchy, the knowledge barriers they impose. Other people who would do something get lost in that process…… Allow space to develop common understandings of problems and to debate dominant norms, free from administrative burdens, predetermined issues and timings; facilitate multidirectional learning, between sectors and scales.
Technocratic and prescriptive approach to REDD + forecloses negotiation of equitable procedures and outcomes "The implementing parties are trying to place certain ways of acting from REDD. They involve many required activities…. International interest in protecting carbon can overpower the voice of local communities." It's not that you pick specific issues for Uganda, you simply have to work through the components in order…..The FCPF packages have it all covered. There are still national barriers of course, they are high. But the international ones really drive it." "Justice and equity issues haven't been covered yet in the REDD work, so those issues haven't dropped yet. We will have to negotiate in enough detail to get consent (of communities) but that hasn't come up yet. That will be part of the second phase." Enable national intermediaries to co-design categories of issues and the methods and timings to address them in REDD + policy process. Facilitate open discussion of optimal governance and approaches to equity; promote equity concerns ahead of technical development Tick-box approaches to equity issues; lack of substantive procedures to ensure mitigation of risks or responsiveness to impacts "It is not just about a UNREDD or World Bank standard being in place, ticking boxes, getting the necessary signatures. It's a big risk, that if you take that tick-tick approach the people doing the consulting don't even actually believe what they are doing is enough, the locals don't believe, the trust isn't developed. So when the officials agree to the project with communities then you go back after 10 or 15 years and see the forest has been burnt down." "Their methods, like the safeguards don't have any effect at the field level." Integrate project-level social monitoring with local level guidelines and substantive procedures for participation and distribution Centralisation of governance and knowledge; cooption of critical organisations "We are trying to make the point that REDD+ can paralyse the community forestry process in Nepal, firstly through narrowing the decision making space.…. that reduces the strength of the voice which is resisting and debating REDD+ …… So it works in subtle ways to the detriment of local people." "They (other civil society group) are being mobilised to make REDD+ work, talking of the benefits and awareness rather than being critical. So that reduces the strength of the voice which is resisting and debating REDD+….. (other civil society group) are being co-opted to implement REDD without empowering or even bringing in local people. They're not really representing their interests." Establish subnational working groups matching nested governance, allowing multi-directional learning; identify practices that have succeeded in similar initiatives or related fields, while minimising bureaucracy.
Lack of common understanding of REDD+ "It is clear that no one has a common understanding of REDD." "Most people report benefits but say very little about the challenges and possible costs based on lessons from pilot studies." "Pilots have been like low hanging fruit. They have been set up where it is easier to do and have been selective in the rules they implement." "There have been some pilot projects but they have all been CSO driven. That evidence is not really being used by NFA (the National Forest Authority) so when they (NFA, come to) implement projects they will be far removed from the pilots." Draw evidence from wider set of environment and development projects in relevant political context; ensure pilots incorporate realistic design and methodology to maximise policy-relevant learning (continued on next page) N.M. Dawson et al. Environmental Science and Policy 88 (2018) 1-9 increased centralisation of forest decision making. The heightened bureaucracy associated with REDD+ was seen to privilege knowledge at national level, to direct funding to higher level processes and capacity building, with limited benefits or control granted to communities and less to specific poor and marginalised groups. Although REDD+ had endured for more than five years in each country and was accelerating towards implementation due to administrative milestones having been achieved, all interviewees felt there had been limited opportunity to challenge the dominance or prioritisation of certain norms. For example, the focus on national forest monitoring capacity as a key priority and benefit of REDD+ and financial distribution as a mechanism for delivering equity were seen to preclude attention to the relevance of local institutions for implementing safeguards and delivering tenure security. Furthermore, specific sections of the REDD+ administrative timeline in both countries were contracted to various CSOs and research organisations, commonly those involved in the national working group, for consultancy work. This led to accusations between CSOs in Uganda and Nepal of having been co-opted in the process and failing to fulfil their role to provide a critical voice for protecting the interests of or advancing recognition for local communities (Table 1).

Lack of common understanding of REDD+ and undetermined CSO response
The perceived lack of deliberation during REDD+ processes was illustrated by divergent understandings between CSOs and state actors about design and likely outcomes of REDD+ projects, particularly in Nepal. Lack of common understanding was considered in part a result of a lack of pilot projects, particularly involving the responsible agencies and pervading political context, and limited evidence drawn from other carbon forestry initiatives or relevant conservation and development projects (Table 1).
Despite the considerable shortcomings listed above, intermediaries maintained involvement not due to committed support for REDD+ but due to the need to overcome the complexity of the REDD+ architecture and lingering uncertainty about potential gains or harms to their causes. Indeed, support for REDD+ in its current form was far from guaranteed, particularly for those actors prioritising recognition-based justice issues for local communities in Uganda and Nepal (Table 1).

National factors constraining effective norm localisation in REDD+ national interpretation processes
As shown in Table 2, national political deficiencies were also viewed as a substantial barrier to realising equitable REDD + policies and projects.

Lack of integration and weak implementation
Poor capacity, limited resources, frequent turnover of government staff and lack of vertical and horizontal integration within the state often compromised decision-making processes and outcomes for Ugandan and Nepalese people. The REDD+ policy process is overseen by one small implementation centre in each country and the enormity of the task to integrate with other agencies with sometimes contradictory policies and objectives, particularly between economic utilisation of forests (for timber, agriculture or energy projects) and their conservation, was highlighted by all interviewees (Table 2).

Decision-making based on short-term political objectives
However, all fourteen interviewees emphasized that positive local outcomes had often failed to materialise from conservation and development projects, not only due to deficiencies in implementation, but due to political decision making based on short-term objectives to capture benefits or to secure support and political power (Table 2). A common example was the tenure assurances given by politicians to rural populations living alongside Uganda's forests prior to elections, which were rarely upheld if elections had passed or rescinded if support was not provided. In order to overcome these national political influences and constraints on forest governance, intermediaries perceived that the REDD+ Working Group should collate evidence from and reflect on past failures as well as successes in the environment and development sectors and establish procedures to ensure accountability such as clarifying responsibilities and establishing clear dispute resolution mechanisms (Table 2).

Limited space for critical debate
Despite doubts about REDD+, many CSOs chose not to resist REDD + or actively undermine it because of the political context in which they operate. The Ugandan government exerts intermittent authoritarian control over CSOs, and in Nepal the frequent overturn of governments and ensuing power struggle means political affiliation had become an important aspect of coalition building and policy influence. Civil society in both countries actively protested various policies related to forestry, but perceived a need to do so sparingly, based on concrete evidence of severe local impacts rather than speculatively. Only when faced with a more immediate social or environmental threat do they garner support and mobilise against it. For instance, in Uganda, CSOs coordinated a movement to overturn a government plan to grant a company license to clear Mabira forest for sugarcane (Child, 2009). These results were unsurprisingly expressed more commonly in individual interviews than think-tanks. Ten of the 14 interviewed stated the importance of caution in voicing objections to government policies and eleven stated that restricted political space made it difficult to raise issues of social injustice faced by specific groups within society (Table 2). Although CSO actors used organised campaigns through newspapers and radio as well as public protests to influence policy, they expressed a need for a strong coalition to do so, rather than reactive, singular actions. Many saw open protest against REDD+ as detrimental to their organisation's relations with the state, particularly as it represented an unclear effect rather than imminent and dramatic impacts on those they sought to represent.
The politicisation of certain struggles over others can even mean some organisations are excluded from parts of the REDD+ process. Barriers to achievement of equity Supporting quotes from semi-structured interviews Potential actions for enhancement of processes Undetermined civil society response to REDD+ "We see many potential conflicts around REDD+." "When we started looking at it (REDD+) we were not pro or anti. We looked very carefully and it has been an opportunity to argue for our rights, we can try to advance our rights through market based conservation institutions. But there are threats: It's a business, it goes through MRV like a commodity market but many of the people to be affected are unaware of it. So they might be deprived. The remaining rights and livelihoods IPs (indigenous people) have, they are already deprived, so there is chance of their titles being sold, whole areas can be restricted to people." N.M. Dawson et al. Environmental Science and Policy 88 (2018) 1-9 Intermediaries of all types are not free from political influence. Many CSOs have objectives of maintaining or increasing their power and influence, which, particularly in the case of Nepal has led them or individuals within them to be politically affiliated. This can result in civil society actions being politically diluted. For example, if representation of ethnic or social groups such as Dalits or indigenous groups who lack political representation in Nepal is considered unpalatable to political parties, they may not gain widespread support among a coalition of CSOs. Four of the fourteen interviewees perceived they had been excluded at some stage due to their demands being politically unwelcomed by both state and other non-state actors. Regarding procedural aspects of REDD+ debates at national and subnational level, CSOs, particularly those representing minority cultural groups, felt their presence at meetings was designed to give the process participatory legitimacy without any power or political will to effect real change (Table 2).

Discussion and conclusion
Our interpretive study emphasizes national interpretation processes as critical space for the negotiation of global environmental norms and adaptation of policy to local social and environmental contexts. A broad range of civil society organisations participate in national REDD+ working groups in Nepal and Uganda, as in most countries involved in rapidly formulating REDD+ approaches. Those working groups represent a key structure and forum for debating national and subnational REDD+ mechanisms. However, our exploration of intermediaries' perceptions of REDD+ processes at the national level in Uganda and Nepal reveal considerable procedural deficiencies which act as barriers to adaptation of international norms to national and local context and to realisation of equitable REDD+ for local interest groups. Indeed, after more than a decade of REDD+ at the international stage and eight years in both Nepal and Uganda, one of its key limitations, as demonstrated in this paper, remains the lack of inclusiveness and integration of different stakeholder perceptions, suggesting persistently low quality of stakeholder participation. REDD+ proponents aspire to transform social and environmental outcomes, yet achievement of these high ambitions will require a much greater transformation in governance quality than currently evident in national approaches.
Our analysis of national level processes uncovers an agenda predetermined by international agencies favouring a technically-driven, narrowly conceived, administratively burdensome approach which deprioritises equity concerns and forecloses spaces for negotiation such that national processes simply mirror international equity norms espoused in UNFCCC policy without effectively localising (or even nationalising) them. These deficiencies are exacerbated by heavily Table 2 National factors perceived to constrain the ability of national and subnational REDD+ policies in Uganda and Nepal to deliver equitable procedures and outcomes for local stakeholders. Quotes are taken from think-tank discussions and semi-structured interviews, though not attributed to specific individuals or organisations for ethical reasons. The final column presents potential actions to overcome barriers put forward by the authors.

Barriers to achievement of equity
Supporting quotes from semi-structured interviews Potential actions for enhancement of processes Lack of integration and weak implementation.
"Integrating those other fields is not yet happening. People think REDD is separate. Like human rights for example. We should have been integrating those ideas into environment issues 20 years ago. It's a big risk that we may have stakeholders who should be talking about REDD but aren't involved." "Even when policy is well developed, the central level coordination is so weak!…. It (decision making) only ever involves one ministry….. The Ministry of Environment does an EIA when the forest is to be cut for a road development and the Ministry of Forestry don't even know about it!" Ensure involvement of diverse state actors in national interpretation processes; Recognise past failings and requirements for integration and trust-building; direct sufficient resources to cover key equity issues Decision-making based on short-term political objectives "The politicians want the big schemes to go ahead. They are in business. There is an element of that personal gain. Corruption, transparency are lacking….those big cases of licenses being granted get streamlined, without scrutiny." "Projects are run very haphazardly, and they are just doing it for their personal interest. You don't know what the government is planning. Local people distrust the government at all levels." "People are so preoccupied with power. And that makes it very hard for people to focus on doing things. The changeover of power is so frequent in government. We still have no guidelines for our (Nepal's) change to federal, provincial government." "We are also losing forest cover rapidly. The government institutions themselves are behind that, through the decision they choose to make over forests. Where people support the government they are unlikely to be evicted from forests, whereas in other areas they will readily evict people living inside or using forests." "Looking at the syndicate in Nepal's forestry sector, I am not so enthusiastic about REDD." Ensure safeguards are substantively underpinned by laws and regulations at regional or national level; establish adequate grievance mechanisms and spread awareness of rights to hold decision-makers to account; promote transparent decisionmaking in line with agreements made between local communities and institutions implementing REDD + or forest governance programs.
Limited space for critical debate; perceived exclusion of some groups from REDD+ processes.
"If we try and raise our political voice, to be powerful and claim power in the number of (people in social group we represent), then they won't invite me again (to the REDD+ Working Group), so we have to adapt the arguments we use and not overstep." "We have support from some big politicians. But when we have one success we inevitably suffer a backlash. The barriers to gaining success are party political. The main leaders are nonindigenous, and they have an entrenched mindset against IPs." "They are for political parties, the other CSOs, we are not. We don't like to be politicised. The individuals in those NGOs are affiliated to parties, not the institution itself. People meet with others closest to their political beliefs, sometimes even if they are not in the same sector." Encourage openness; use alternative forums with independent chairs or facilitation; develop group position statements; create strength from difference; Seek to highlight existing forms of inequity and social and political barriers to making advances; ensure procedures do not mirror pervasive social inequalities but provide potential for transformation N.M. Dawson et al. Environmental Science and Policy 88 (2018) 1-9 constrained political space within each country. Such constrained space for national interpretation may forego consideration of a broader range of possible governance regimes or lessons to be drawn for effective and equitable implementation (Gebara and Agrawal, 2017;Gupta et al., 2014;Turnhout et al., 2016). These inadequacies are far from trivial, given that forest governance in Nepal and Uganda affects the tenure security and human rights of some of the poorest and most vulnerable people on the planet. Our findings highlight divergent conceptions of equity between REDD+ actors, with those intermediaries who seek recognition of the rights, identities, practices and inferior participatory power of minority groups finding little space for normative influence or in some cases even suffering subsequent exclusion from REDD+ processes. This contrasts with the findings of authors using approaches based on institutional rationality with limited analysis of power dynamics, which suggest for example that widespread participation of civil society in Nepal can indicate a high degree of national ownership and progressive distribution of power and co-operation . Our more critical findings appear to be corroborated, because although both Nepal and Uganda are viewed by REDD+ proponents as performing strongly in administrative preparations and are (as of early 2018) accelerating towards implementation of a wave of projects with international funding, numerous CSO actors participating in REDD+ processes detailed barriers to effective deliberation of norms, expressed very limited buy-in and voiced potential for them to mobilise against any emerging injustices and actively oppose REDD + implementation in future.
Perceived procedural deficiencies followed surprisingly similar themes for Nepal and Uganda, despite contrasting political contexts and political barriers affecting the work of intermediaries, varying forest and land policies and specific minority groups affected. Key differences noted through our analysis included Nepal's longstanding community forest management initiatives, meaning a key source of contention over REDD+ relates to levels of centralisation in forest governance, which will be affected by ongoing processes of constitutional amendment and federalisation. In Uganda, a key concern related to local communities' mistrust of externally-driven forest and land governance due to the tendency to override customary tenure systems, which often exist in parallel with legally recognised property rights and must therefore be considered central to REDD+ safeguards. Respective REDD+ working groups must create space to address these potential weaknesses as priorities. Advantageously, both decentralised forest governance in Nepal and customary tenure in Uganda cases are already well supported in legislation, if not in implementation.
The consistency of the procedural shortcomings related to REDD+ suggests similar solutions could be applied internationally. Effective negotiation of diverse interests in the face of political constraints necessitates attention to the quality of participation and power relations which influence its functioning (Cornwall, 2008;Enns et al., 2014;Turnhout et al., 2010). To enable national processes to function more effectively and realise more equitable procedures and outcomes at local level, attention is required not just to the presence of diverse intermediaries on committees, but to provision of suitable space and forums for open questioning of and opportunity to collaboratively influence the national REDD+ approach by adapting global policy templates through recognition of past failures and drawing on evidence from a range of relevant sectors. Such features of national interpretation processes which may empower diverse intermediaries to deliberate and influence localisation of international norms must be enshrined in transparent, locally grounded policies, supported by regulations or laws (with adequate mechanisms to ensure accountability) and actively promoted, disseminated and upheld by relevant international or implementing agencies such that equity concerns are given equal prioritisation to ecological, administrative and capacity building aims (Tables 1 and 2). The worthy principles expressed in REDD+ safeguards must be underpinned at multiple scales by relevant, functioning regulations and systems that allow even the most vulnerable affected people to ensure those governance standards are upheld.
Our results also shed light on the role played by intermediaries in multi-stakeholder policy negotiations. Although their role, particularly of CSOs, is undeniably crucial in advancing the interests of local communities, intermediaries act strategically to achieve moments of influence and do not always act consistently or selflessly as 'norm entrepreneurs' (Finger and Princen, 2013;Routledge et al., 2006). In both Uganda and Nepal, CSOs who normally provide the most critical voices benefit financially from consultancy contracts to implement donor agendas, and are criticised by potential allies for acting in self-interest to the detriment of local communities. Division between CSOs is more likely to suppress coalition-building, criticism and influence. In those circumstances intermediaries who in principle and in their normal advocacy work would argue for the progressive recognition of rights of marginalised groups or past injustices, may adapt to focus on more simple equity issues like participatory inclusion or monetary distribution, justice norms commonly espoused and granted higher priority in global environmental policy (Okereke, 2008;Suiseeya, 2014). Yet, in support of strategic restraint on the part of Ugandan and Nepalese CSOs, persistent uncertainties regarding local impacts of REDD+ and even what a project will look like on the ground may justify the 'wait and see' strategy adopted by some intermediaries so many years on. Ultimately, without due attention to recognition-based concerns, REDD + risks reproducing injustices for local communities and marginalised groups, which can in turn undermine project support at local and intermediary level, and impair efforts to reduce deforestation and forest degradation.