Governance Values in the Climate Change Regime : Stakeholder Perceptions of REDD + Legitimacy at the National Level

This paper presents the results of two national-level studies of REDD+ governance values in Nepal and Papua New Guinea (PNG), using a hierarchical framework of principles, criteria, and indicators (PC&I), with evaluation at the indicator level. The research was conducted by means of an online survey to determine general perspectives on the governance quality of REDD+, as well as stakeholder workshops, in which participants were asked to rank indicators on the basis of perceived national significance. In the online survey, respondents in both countries identified inclusiveness and resources as the highest and lowest scoring governance values, while inclusiveness, resources, accountability, and transparency, were given priority, although their relative importance differed between countries given national circumstances. The reasons for the commonalities and differences of perceptions between these countries are discussed. The findings suggest that while a generic set of governance values may be usefully applied for determining the institutional legitimacy of REDD+, their relative importance is different. This leads to the conclusion that it may not be appropriate to use a simplified approach to REDD+ governance, focusing for example on safeguards, given different national priorities and contexts.


Introduction: Governance Challenges Confronting REDD+
The United Nations Framework Convention on Climate Change (UNFCCC) mechanism 'reducing emissions from deforestation and forest degradation and the role of conservation, sustainable management of forests and enhancement of forest carbon stocks in developing countries', referred to as REDD+, has been the subject of intense academic scrutiny in recent years, notably concerning the initiative's governance at the national level.A search on Proquest using the term "reducing emissions from deforestation and forest degradation" yielded 663 peer-reviewed scholarly articles between 2008 and May 2016.Adding the term "national" generated 537 results, while inclusion of the term "governance" produced 307.This article focuses on the types of values that might be used to determine the institutional legitimacy of REDD+, and what governance priorities national-level stakeholders place on the mechanism.'Governance' refers to the structures and processes used to steer and coordinate interactions [1].It is a more useful term than 'government' for understanding the political and social relations that occur in the intergovernmental realm, and the related contexts of state, society, and the market, encompassing both state and non-state actors' relationships within contemporary systems of global policy-making, and the structures and processes which underpin them [2].
Previous experiences with REDD+ have been largely based on 'nested' (i.e., sub-national) demonstration or pilot projects, with involvement of NGOs and other non-state actors.The trend now appears to be one of increasing centralisation and government control, with the danger that any financial benefits arising from emissions reduction payments will not flow to local actors, thereby reducing compliance.These issues, combined with inconsistent institutional approaches, as well as a high level of policy uncertainty in the wake of the Paris Agreement, are all contributing to a lack of clarity around what governance arrangements will ultimately be used to implement REDD+ on the ground [8].
At the intergovernmental policy level, there has been some recognition of the challenges confronted by national REDD+ initiatives.The Cancún Agreements of the Conference of Parties (COP) 16 in 2010 refer specifically to the need for "guidance and safeguards" including "transparent and effective national forest governance" and the "full and effective participation of relevant stakeholders, in particular indigenous peoples and local communities", but they do not stipulate how these should be implemented, referring only to "national legislation and sovereignty" [9] (p.26).In Article 72 there is an acknowledgement of the importance of "land tenure issues, forest governance issues, [and] gender considerations", but again there is no further elaboration [9] (p.13).Nevertheless, the recognition of the UN Declaration on the Rights of Indigenous Peoples (UNDRIP) in the Cancun Agreements means that there is an implied requirement for the free, prior, and informed consent (FPIC) of Indigenous people in REDD+ activities [10] (p.2407).Under the Cancún Agreements, tropical forest countries involved in the various stages of REDD+ (readiness, pilot studies, and implementation) are expected to ensure that national activities do not impact negatively on the environment or people, whilst also contributing positively to governance and generating environmental and social benefits.Each country is required to demonstrate compliance by developing its own national safeguard system.Compliance is informed by domestic policies, laws, and regulations, and reporting occurs via what is referred to as a safeguard information system.UNFCCC itself does not stipulate what form safeguard systems should take, only that safeguards should be addressed [11].
The effectiveness of REDD+ has been questioned especially as national practices and related power dynamics have had a significant impact on the efficacy of REDD+ governance on the ground, notably regarding the mechanism's inability to protect biodiversity [12].Indeed, some NGO studies have Forests 2016, 7, 212 3 of 17 gone so far as to argue that deforestation and forest degradation are "still being compounded by poor governance, including corruption, conflicts between national and local authorities, and insufficient resources and institutional capacity" [13] (p. 2).Concerns that REDD+ might increase deforestation led to calls from the NGO community in the lead-up to COP 16 for the initiative to make sure that the safeguards it adopted were 'strong' [14].
In the wake of the Cancún Agreements, the REDD+ policy community responded by adopting a range of measures, notably around benefit sharing arrangements.The World Bank's Forest Carbon Partnership Facility (WB, FCPF) has played a significant role, which requires that allocations from its Carbon Fund occur in the context of a national benefit-sharing plan, but exact arrangements are not specified [15].In an effort to create a broader approach to managing safeguards, the FCPF released its own Common Approach to Environmental and Social Safeguards for Multiple Delivery Partners in 2011 including guidelines as to how partners were to prepare Strategic Environmental and Social Assessment (SESA), a WB requirement for funding [16,17].The WB revisited its policies in July 2015, and in the case of FPIC, the revisions required that consent must be demonstrated and where it could not, those project aspects relevant to Indigenous peoples could not proceed.In the case of stakeholder management, the revision required increased and ongoing levels of engagement, as well as providing new standards for international financial institutions (IFIs), leading the Bank to claim it was at the 'forefront' of safeguard policies [18].
A number of voluntary standards have arisen around benefit sharing arrangements and other safeguards, including the Climate Community and Biodiversity (CCB standards), a largely NGO-driven initiative, which requires FPIC, and the REDD+ Social and Environmental Standards, which contain provisions for the transparent, participatory, inclusive, effective, and equitable distribution of benefits [15] (pp.271-272).It should be noted here that there are institutional linkages between the two, with the Climate, Community and Biodiversity Alliance (CARE International, Conservation International, The Nature Conservancy, Rainforest Alliance, Wildlife Conservation Society) and Care International also being involved in REDD SES.The CCB covers 'site based projects', while REDD SES covers 'government led programmes ' [19].Other organisations, including the United Nations' own Collaborative Programme on Reducing Emissions from Deforestation and Forest Degradation in Developing Countries (UN-REDD), and research organisations, such as the International Union of Forestry Research Organisations (IUFRO) and the Centre for International Tropical Forestry Research (CIFOR) and the International Union for Conservation of Nature (IUCN) have also developed their own "guiding principles", but in the absence of formal rules, countries ultimately determine their arrangements [15] (p.271).
In the light of these issues, this article explores the country-level expressions of REDD+ governance in two countries, Nepal and Papua New Guinea (PNG).REDD+ activities started in Nepal and PNG in 2008.In Nepal they focused largely on capacity building around carbon accounting and arrangements for sharing the benefits arising from carbon payments.Various pilot projects were conducted by a range of international NGOs and local partners in Nepal's well-developed system of community forests, which are managed by community forest user groups (CFUGs), while remaining under state control [20].The experience of REDD+ on the ground to date has been both positive and negative.On the one hand the financial and social capacity of the CFUG network has improved, but this has been at the expense of autonomous decision-making and customary rights to forest access [21] (p.39).Despite Nepal's history of inclusive forest management, the technical orientation of REDD+, it has been suggested, has contributed to inhibiting local participation [8] (p.66).In PNG, the implementation of pilot projects has been somewhat sporadic, due in part to an unstable political climate and forest governance issues.Although land tenure resides with customary landowners, projects have been largely directed by government agencies with only a secondary role for NGOs [22].Local communities have had no formally specified role in either REDD+ design, or monitoring reporting and verification (MRV) [8].As a result, REDD+ projects have been largely unable to secure social license for their activities in PNG (particularly in relation to FPIC), not the least because of the difficulties of registering landowner groups-and thereby demonstrating formal land tenure-which is a precondition for participation in REDD+ projects [23] (p.152).
Scholarly analysis suggests that while safeguards have the potential to balance the over-simplification of social-ecological systems that such globally driven and technically oriented initiatives as REDD+ can lead to, this is dependent on the design of standards under which safeguards operate.The standards that have arisen in the context of REDD+ safeguards recognise FPIC, but interpretation can vary, understanding 'consent' as 'consultation' and leaving the ultimate determination in the hands of auditors, and can be largely procedural, merely ensuring that duties and responsibilities are fulfilled when rights are ignored [24] (pp.3350-3354).This can be due to political sensitivities in developing countries around issues of democracy and human rights, resulting in the application of national approaches, rather than international norms [25] (p.2406).This has led to contestation around issues of legitimacy, notably regarding which approach to environmental governance should be adopted.As a result, non-state actors have begun to develop standards themselves, based on their own conceptual frameworks [24] (p.3355).Consequently, in the world of REDD+ (as in other market-based instruments) there are inconsistencies between standards and the values that inform them [26] (pp.[3][4].
In order to determine REDD+ legitimacy and effectiveness in these countries, and draw some broader conclusions, the article uses a comprehensive framework of governance values, which bear directly on the question of how the initiative is steered and its collective action coordinated.The findings suggest that a generic set of governance values may be usefully applied for determining the institutional legitimacy of REDD+, but the emphasis placed on values is different across the two case study countries, with implications for national REDD+ initiatives more broadly.

Materials and Methods: Identifying and Assessing REDD+ Governance Values
Before outlining the methods adopted in this study, a few preparatory comments regarding the approach towards governance and legitimacy is required.It is beyond the scope of an article of this length to go into any great detail, as both subjects have been submitted to exhaustive exploration in recent years, especially amongst scholars of international relations and public policy.There was something of a turn in regime theory in the 1990s and early 2000s, which emphasized a shift away from the notion of 'government' to 'governance', as the latter more accurately described the nature of the structures and processes underlying social-political interactions between state and non-state actors within institutions of global governance [2,[27][28][29].
A parallel debate centered upon the notion of institutional legitimacy.The orthodox view was that the state was the sole repository of political power [30] (p.187) [31] (pp.[3][4].But the view emerged that legitimacy resided within institutions, state or non-state.Two distinct schools of thought emerged: one emphasizing the importance of democratic processes and rules of procedure, referred to as input legitimacy (the means justify the ends); the other focusing on results, referred to as output legitimacy (the ends justify the means) [32] (pp.152-155), [33] (p.12), [34] (p.45).This approach to understanding legitimacy has been subsequently expanded to encompass 'throughput legitimacy', which focuses on the institution's internal organizational arrangements-both in terms of who is represented, and how deliberation occurs [35] (pp.2-5).
Given the institutional emphasis in this discussion, governance can therefore be understood as a value-neutral term, referring to the way in which an institution engages with its stakeholders, makes decisions, and accounts for itself to its constituents and the public at large [36].'Governance arrangements' refer to the structures and processes an institution uses to facilitate steering and coordination [37] (footnote 13, p. 24).Governance values in turn, as explained below, delineate the specific attributes necessary to inform a global public standard for the normative legitimacy of global governance institutions [38] (p.405).
In order to critically interrogate the governance underpinning REDD+, it is necessary to make some methodological decisions about what types of values will or will not be included in the analysis.The method includes consideration and analysis of eleven different governance values (at the level of 'indicators', which can be grouped into the larger categories of 'criteria' and 'principles'): accountability; transparency; inclusiveness; resources; equality; democracy; agreement; dispute settlement; behavioural change; problem solving; and durability.While these governance values exist in many explorations of governance and legitimacy in the literature [39] (pp.12-15), the use of this specific list nevertheless requires justification in two directions.On the one hand, it must be shown why the research method limited itself to these, rather than including further potentially attractive values.On the other hand, it must be shown why the use of all these eleven values was required, rather than just considering a narrower focus on, for example, transparency and accountability.First, these are of course not the only values that could conceivably bear on an institution's legitimacy.Buchanan and Keohane's analysis of the legitimacy of global governance institutions also includes substantive values, such as respect for human rights [38] (p.419).A similar list by Sampford also includes human rights, and adds the substantive values of fraternity and environmental value [40].The term 'substantive values' connotes specific goals the institution should seek to further, or moral constraints it should recognize.As such, rather than evaluating who was involved, and how the process occurred, substantive values make up considerations that should inform the institution's executive decision-makers.While these substantive values are no doubt crucial to an all-things-considered appraisal of an institution's legitimacy, the research method here takes a more focused approach by considering only the values that bear on the nature of the structures and processes of governance themselves.Leo Huberts marks this distinction by separating the ethics of governance from the results of its process [41].In other words, because 'governance' refers to the structures and processes used to steer and coordinate interactions, 'governance values' should refer to the qualities and characteristics of these structures and processes-rather than more substantive moral goals and purposes an institution might pursue.
This research project employs this exclusive focus on qualities of governance (rather than wider substantive values) for several reasons.First, as seen earlier in Section 1, a consistent charge laid against REDD+ has been on the basis of its quality of governance, in particular in the form of local stakeholder input.This particular instance follows a larger pattern of local communities challenging the legitimacy of global governance [42].Without prejudicing the validity of other moral concerns, this specific charge about governance failures warrants a dedicated treatment, which this research aims to provide-at least as it applies to the two case-study areas analyzed.Second, REDD+ already has its substantive values laid down by the UNFCCC, and as expressed institutionally via UN-REDD, FCPF, national level agencies, and so forth (reducing emissions, conservation and sustainable management, and enhancement carbon stocks, and so forth).Achieving these will be captured by the aforementioned governance values of behavioural change, problem solving, and durability.To include in the analysis additional substantive values outside the program's mandated goals would court controversy that a strict focus on structures and processes may avoid.Third, and perhaps most important, a strong performance on the governance values presented here will tend to facilitate the types of substantive goals lauded by Buchanan, Keohane, and Sampford.For example, improved inclusiveness and equality, as well as deliberation and democracy will necessarily further the human rights requirement of giving subjects (including local stakeholders) involvement in decision-making, and input into the decisions of the political authorities that affect them, as well as helping ensure the respect of their other human rights (such as their property or cultural rights pertaining to local environments).Improving these governance values therefore advances the human right to take part in government enshrined in Article 21 of the Universal Declaration of Human Rights [43], UNDRIP [10], and the social safeguards laid down in Cancun [9], and helps facilitate the specific rights of property and culture pursuant to Articles 17, 22, and 27 of the Universal Declaration.
What then of the second concern, which in coming from an opposite direction queries why so many governance values are employed, rather than just cleaving to a narrow focus on usual suspects like transparency and accountability?Again, there are several reasons for this methodological choice.Centrally, it is plausible to think that transparency and accountability are not valuable in themselves: they are valuable insofar as they contribute positively to an institution's performance and legitimacy [44].If an institution has transparency, but stakeholders possess no way to impact upon its working or decision-making, then the transparency will fail to improve the institution's quality or its legitimacy.Similarly, accountability itself does not guarantee good outcomes-it just ensures that someone can be held responsible for rule-breaches that lead to bad outcomes, which is a much narrower quality [45].These two points hint at the deeper reason why it is necessary to employ a comprehensive array of governance values, namely, governance values inter-link with each other.Each governance value's full worth remains dependent on the presence and quality of the others.For example, the benefits of having inclusive practices of allowing many stakeholders a seat at the decision-making table are lost if there is no equality in power relations (or at least efforts made to address imbalances), or access to resources, so that poor, local stakeholders can afford to take the seats allocated to them.A deliberately comprehensive array of governance values can therefore identify an institution's weaknesses that prevent it from achieving its goals and responding to stakeholders.Finally, considering a wide array of governance values allows the project to test which methods the stakeholders prioritize themselves.This helps the institution to respond to the concerns of specific cultural groups, and helps ensure that the research remains sensitive to the priorities of local and national stakeholders.As Section 4 shows, the cultures in the two case-study areas did indeed prioritize different qualities of the governance system differently, suggesting the merit of the methodological approach employed.
In order to evaluate REDD+ institutional expression at the national level, a conceptual continuum of values is presented here.The tripartite typology of governance value sets presented here moves from 'thin' to 'thickish' to 'thick'.Each set differs from the others in terms of the richness of the governance values it includes, with 'thick' governance values incorporating all of the qualities earmarked under the 'thickish' category-and both of these including the sparse requirements laid down by 'thin' governance values.
Thick governance values capture the full gamut of social and moral qualities that can be demanded of an institution's mechanisms of steering and coordinating.If an institution possesses all of these qualities, then (while this may not cover everything that morality might demand of the institution) the institution will possess legitimacy in the ways it goes about making, authorizing, implementing, and appraising its decisions.Thick governance values thus constitute a plausible answer to the question of what sorts of internal organizational arrangements would legitimize institutions and institutional complexes [38].
Institutions and networks gain in legitimacy the more that their actions are transparent and accountable, the more they employ constructive deliberation, the more they provide affected stakeholders with the capacity to provide input into their workings, and the more they produce effective results [39] (pp.[15][16][17][18].Legitimacy in an institutional context is therefore closely linked to governance quality and also applies to the structures and processes that steer an institution, in the theoretical terms of input-, throughput-and output legitimacy, as discussed above.Figure 1 below is an integrated model that captures these three previously discrete theories [32,33,35].
In order to evaluate the case studies selected, Table 1 below contains a comprehensive set of governance values using a hierarchical framework of principles, criteria, and indicators (PC&I), derived from a review of contemporary governance literature [12] (pp.[12][13][14][15][16][17][18]. PC&I have become a common method of assessment for sustainability, and sustainable forest management (SFM) in particular, popularised as a consequence of the 1992 'Rio' Earth Summit, and reflected in its foundational document Agenda 21 [46] (p.61).Frameworks such as these have been developed to ensure the consistency of evaluation, by placing each element in its appropriate location, from principle to criterion and thence to the relevant indicator, to avoid duplication or overlap.A principle represents a fundamental rule or value to be determined, and which is categorised into criteria, which in turn are broken down into indicators, or parameters, for assessment purposes.
Assessment itself occurs at the indicator level, as both principles and criteria are ideational in nature, and not directly measurable [47] (pp. .In order to evaluate the case studies selected, Table 1 below contains a comprehensive set of governance values using a hierarchical framework of principles, criteria, and indicators (PC&I), derived from a review of contemporary governance literature [12] (pp.12-18).PC&I have become a common method of assessment for sustainability, and sustainable forest management (SFM) in particular, popularised as a consequence of the 1992 'Rio' Earth Summit, and reflected in its foundational document Agenda 21 [46] (p.61).Frameworks such as these have been developed to ensure the consistency of evaluation, by placing each element in its appropriate location, from principle to criterion and thence to the relevant indicator, to avoid duplication or overlap.A principle represents a fundamental rule or value to be determined, and which is categorised into criteria, which in turn are broken down into indicators, or parameters, for  In terms of governance, 'values' refer to specific aspects of the structures and processes that determine how things are done within an institution.'Thin' values tend to be the focus of staple governance mechanisms around ensuring organisational responsibility to other participants and the public at large, such as accountability and transparency.'Thickish' values add more elements beyond the 'usual suspects' of accountability and transparency, incorporating qualities around interest representation, notably the degree of inclusiveness of stakeholders (access), and the extent to which an institution resources those who would otherwise not be able to participate meaningfully.They also contain some recognition of the need for democratic practices for reaching agreement (such as consensus or voting).However, viewed from the perspective of assessing legitimacy, thickish values remain incomplete.By contrast, 'thick' governance encompasses the existing suite of values, but is concerned with ensuring equality of relations between stakeholder interests as well as 'deeper' procedural values around how disputes arising from decisions are settled, and how decisions are implemented.Implementation in this thick context looks beyond outputs, such as rules of procedure, or standards, etc., to see if these products result in substantive outcomes that all stakeholders can claim ownership of and support, and which result in lasting behavioural change and solutions to the problem that the institution was established to address (in the case of REDD+, greenhouse gasses arising as a consequence of deforestation and forest degradation) [48].
The two studies were conducted in Nepal over the period from July-December 2011, and in PNG from April-May 2015, following the methodology developed by López-Casero et al. [49] (pp.12-13).The research design was intended to foster collaboration with REDD+ participants in country, using action research methods, whereby an atmosphere is created that allows stakeholders to develop their own solutions to the problems they are tackling, and solve them through their own efforts [50].
At the commencement of the research, participants were invited to express their opinions regarding the structures and processes of governance relevant to forest management and REDD+, and were enlisted from the economic, social, and environmental sectors.A purposive sampling method was used for the selection of participants for the survey because only a subset of the national population was familiar with REDD+ [51][52][53].Their contact details were sourced from publically available online documents using the search terms 'REDD+', 'participants' list', 'Nepal' and 'Papua New Guinea', as appropriate.In the case of PNG, these contacts were supplemented by a further database of stakeholders provided by the research project's national contacts.They included non-state actors active in the forest sector, i.e., non-governmental organisations/civil society organisation (NGOs/CSOs), as well as state interests, i.e., governmental agencies.
Given the specific national context of each country, it should be noted that the relevant stakeholder sectors included in the survey varied slightly: in Nepal, Indigenous people play a role in REDD+ as a specific interest grouping; in PNG, the entire national citizenry sees itself as indigenous, making such a designation less relevant.Care was therefore taken in ensuring the relevant stakeholder sectors were included in each country survey.
The views of stakeholders were collected by use of on online survey tool.The survey was in English, and translated into one of each country's national language (Nepali and Tok Pisin).Participants were contacted by email, while the survey itself was conducted anonymously to encourage participation.Respondents were asked to identify their sector, as well as country status (national citizens or international) and to provide a rating for their perceptions regarding the governance of REDD+ using a Likert scale of 1-5 ('very low' to 'very high') using the indicators of Table 1 above (resulting in a possible minimum and maximum total score for the 11 indicators of 11 and 55, respectively).Opportunity for comment was provided under each of the indicators.A selection of relevant comments is included in the analysis below.
Some caveats to the research must be made here.The surveys were deployed in Nepal in July 2011 and in PNG in April 2015, and represent perceptions regarding REDD+ prior to any direct interaction with the researchers.While this reduced the impacts of researcher-subject influence, the surveys were conducted in the two countries over different time periods.Comparison is therefore not entirely like-for-like, and respondents' perceptions were probably influenced by developments at the point in time in which the surveys were conducted.Furthermore, the mechanism is also still undergoing development, and it is likely that perceptions will continue to change as REDD+ processes further evolve in the two countries.
Table 2 below breaks down the survey respondents by country, stakeholder sectors, nationality, and number of surveys commenced and completed.The survey was followed by a two-day national workshop in each country; in Nepal forty-three multi-stakeholders were present, in PNG thirty-three.Emphasis was again placed on ensuring a diverse representation of interests, targeting those groups least represented in the survey (e.g., women, minority groups).This led to a mixed group of participants in both workshops: some were recruited from the online survey, while others were identified over the course of the research.During the course of a two day workshop.Participants were provided with further information on the indicators and survey on the basis of the results in their respective countries.This provided them with more detailed knowledge to assist in the ranking of the indicators.At the conclusion of the workshop, participants were asked to rank all indicators on a 1-10 scale on the basis of their importance in their respective countries (1 being the least important and 10 being the most important).In the case of the workshops it should be noted that the prioritisation exercise happened after interaction with the researchers, and after the online survey (December 2011 in Nepal, and May 2015 in PNG).While the survey may have influenced some (but not all) of the workshop participants, the objective was to identify national-level priorities amongst the workshop participants in countries.Unlike the survey, the prioritisation exercise reflected future aspirations (i.e., what workshop participants thought was significant for REDD+ moving forward).In both countries, the total number of participants in the workshops was over 30 and the respondents were completely different from each other in each workshop.On that basis and in order to determine whether the mean perceptions of Nepalese and PNG stakeholders for the 11 indicators were significantly different statistically, the independent t-test (or student-t test) was applied [54] (p.158).Given the variations in the size of each sample, the limitations of such a test should be recognised, but it should also be acknowledged that the overall sample size is large enough to warrant such an approach [55].

Online Survey
The results of the online survey are reproduced in Table 3 below.The overall survey scores for REDD+ governance in both countries were relatively similar (34.85 out of 55 in Nepal cf.32.73 in PNG) (see Table 3 above).These equate to what the researchers refer to as a 'legitimacy rating' of 63% and 60%, respectively (with rounding up)-satisfactory performances, but not overwhelming.Secondly, the highest and lowest performing indicators at the national level were inclusiveness (3.83 out of 5 for Nepal cf.3.49 for PNG), and resources (2.3 and 2.42).In the case of resources, the ratings provided by respondents were below the threshold 'pass' of 2.5 out of 5.In both Nepal and PNG, accountability and transparency were the second and fourth lowest indicators in both countries (in Nepal 3.01 and 3.08 respectively, while democracy was the third lowest indicator at 3.02; in PNG 2.71 and 2.78, with dispute settlement as the third lowest indicator at 2.64).
Despite the differences in sample size, the similarity of the results in both countries led the researchers to conclude that they could be interrogated by means of an independent t-test to determine if there were any statistically significant differences between the two sets of results at the indicator level.The analysis of the online surveys revealed that the perceptions of Nepalese and PNG stakeholders for nine of the 11 indicators were not significantly different statistically (i.e., that they rated them equally, at least statistically).In the case of the two statistically significant indicators (p-value ≤ 0.05), the mean value of perceptions for respondents from PNG was lower for agreement (2.84 in PNG and 3.31 in Nepal) and dispute settlement (2.64 in PNG and 3.17 in PNG) than for those from Nepal.
It should be noted here that adding or subtracting ordinal values is not common when using Likert scales to determine respondent perceptions.However, it is not uncommon in political science for indicator-based assessments to convert verbal values (low, medium, high) to numerical scores, although ordinal scales are usually applied to measure the extent of variability rather than quantifiable levels of difference [56] (p.373; footnote 7).Certification bodies routinely apply numerical values to verbal assessments.For example, in the case of Forest Stewardship Council-accredited certifier, Smartwood, compliance is evaluated at the criterion level using a five-point score (extremely weak, weak, satisfactory, favourable, outstanding), based on the averaging of related indicators, which in turn are assessed using a high-to-low scale [57] (p.33).The point being made here is that the quantification of indicators is not a precise science, and a range of methods for aggregating verbal assessments and determining pass/fail thresholds are used in the field of SFM [58].

Prioritisation Execrcise
The results of the prioritisation exercise are presented in Figure 2 below.range of methods for aggregating verbal assessments and determining pass/fail thresholds are used in the field of SFM [58].

Prioritisation Execrcise
The results of the prioritisation exercise are presented in Figure 2 below.The four most significant governance values at the indicator level for both countries were inclusiveness, resources, accountability, and transparency, but in different orders of priority.In Nepal, stakeholders ranked them in the order of transparency (8.6 out of 10), inclusiveness (8.4), accountability (7.8), and resources (7) (TIAR), and in PNG, inclusiveness (7.4), accountability (7.3), resources (7.2) and transparency (7.1) (IART).The lowest prioritised indicators were also the same for both countries-agreement, dispute settlement, and problem solving, but again in different orders; in Nepal, problem solving (3), dispute settlement (3.3), and agreement (3.6), while in PNG they were dispute settlement (3.3), problem solving (4.6), and agreement (5).
The analysis of ranking by Nepalese and PNG stakeholders reveals that for eight of the 11 indicators there was no significant difference statistically.In the case of the three statistically significant indicators (at a 95% confidence level), the mean value of rankings for respondents from PNG was higher for agreement (5 in PNG and 3.6 in Nepal) and problem solving (4.6 in PNG and 3 in Nepal) than for those from Nepal, but the mean value of ranking for respondents from PNG for transparency was lower than that of Nepal (7.1 in PNG and 8.6 in Nepal).

Discussion and Comments
The results of the two governance surveys reveal more than just which indicators were the best and worst performers.While it is commendable that both countries' participating stakeholders thought REDD+ was inclusive, the low score for resources is a cause for some concern, as organisations lacking the capacity (technical, financial, institutional) to represent their interests cannot participate meaningfully in making and implementing decisions.A review of the comments of the survey participants reveals why resources may have been rated so low.In the Nepal survey one respondent ('other' who identified as a PhD student) commented that: "The resources available for the participation of different interest groups is currently limited to certain elite groups who are misleading the representation of that particular group".Another respondent (also 'other', with The four most significant governance values at the indicator level for both countries were inclusiveness, resources, accountability, and transparency, but in different orders of priority.In Nepal, stakeholders ranked them in the order of transparency (8.6 out of 10), inclusiveness (8.4), accountability (7.8), and resources (7) (TIAR), and in PNG, inclusiveness (7.4), accountability (7.3), resources (7.2) and transparency (7.1) (IART).The lowest prioritised indicators were also the same for both countries-agreement, dispute settlement, and problem solving, but again in different orders; in Nepal, problem solving (3), dispute settlement (3.3), and agreement (3.6), while in PNG they were dispute settlement (3.3), problem solving (4.6), and agreement (5).
The analysis of ranking by Nepalese and PNG stakeholders reveals that for eight of the 11 indicators there was no significant difference statistically.In the case of the three statistically significant indicators (at a 95% confidence level), the mean value of rankings for respondents from PNG was higher for agreement (5 in PNG and 3.6 in Nepal) and problem solving (4.6 in PNG and 3 in Nepal) than for those from Nepal, but the mean value of ranking for respondents from PNG for transparency was lower than that of Nepal (7.1 in PNG and 8.6 in Nepal).

Discussion and Comments
The results of the two governance surveys reveal more than just which indicators were the best and worst performers.While it is commendable that both countries' participating stakeholders thought REDD+ was inclusive, the low score for resources is a cause for some concern, as organisations lacking the capacity (technical, financial, institutional) to represent their interests cannot participate meaningfully in making and implementing decisions.A review of the comments of the survey participants reveals why resources may have been rated so low.In the Nepal survey one respondent ('other' who identified as a PhD student) commented that: "The resources available for the participation of different interest groups is currently limited to certain elite groups who are misleading the representation of that particular group".Another respondent (also 'other', with experience in a donor agency) observed that resources tended to stay within the REDD+ working group, while "outside access is rare".An 'NGO' respondent made the case that "resources should be transparent and shared among the stakeholders".In PNG, one local NGO was of the view that it was "absolutely important that communities in our areas are assisted to be proactive", while another international NGO (INGO) indicated that they had been able to "meet local communities, local landowner group representatives and district as well as provincial authorities" in pursuit of their own REDD+ activities, but stressed that: "more resources should be out into building their capacity on what roles they play in reducing emissions, in whatever development activities taking place in their areas".A further INGO noted that while the PNG authorities had been "doing a lot in the REDD domain", "more of what they have done" needed to be "in the public domain".Seeing a relationship between resources and transparency, another government respondent also agreed, that "prior information on REDD+ resource provision [would] be appreciated".One community forest user saw a "need to consult more land owners and take their options" regarding resources, while a third INGO was concerned that "information provided" was not compatible "to the Resource owners' understanding" of REDD+.
Comments relating to inclusiveness were also revealing.In both countries, respondents were keen to ensure that in addition to government, donors, and implementing agencies, that local communities were included in REDD+ project activities.In Nepal, emphasis was given to Indigenous people, and-in the words of one student, who had previously worked with an international NGO-"different interest groups based on gender, caste, class, ethnicity and geographic origin".In PNG, it was not surprising to see several respondents stress the importance of including local landowner groups, given their significance in relation to forest management.In the words of one government respondent it was important because: The people own the land, where REDD+ Projects would most likely be implemented (not government).Therefore we have to put more focus on the landowners and how the practical benefits (not in monetary terms) will help the local people on the ground.
Respondents also made linkages between inclusiveness and transparency.In PNG, one community based organisation representative who lived in a village within a REDD+ project area was of the view that "we should be updated on the progress of this project".Another national respondent who identified as being with a "global NGO/CSO" made the point that "implementing agencies need more transparency, and to involve all stakeholders within the scope of REDD+".As one 'research-based' NGO from Nepal explained, inclusiveness could only happen: By engaging the diverse stakeholders in REDD decision forums; existing forums are dominated by government officials and very few, elite, members from indigenous communities and community networks.
Respondents made a range of pertinent observations regarding accountability.In Nepal, one respondent who identified as 'NGO' felt that "the powerful can influence and the powerless may lose".They saw a connection between accountability and implementation, and wanted to see measures that would lead to "accountable service providers".In PNG, one 'government' respondent believed that there were "no transparency or governance mechanisms in place to account for such projects".They thought that "until such measures" were in place that it could not be said that REDD+ projects were "acting in the accountable manner".Another PNG national who identified as 'INGO' pointed to "the irresponsible nature of the process as it is happening in PNG", putting this down to "the regulatory system, complex landowners issues, and the inability of those responsible to be held accountable".This meant that the "REDD+ process is bound to be entangled in all this".This led one Women's organisation respondent to conclude that: "some systems need to be put in place in order to clearly see that it is accountable".
Problem solving and agreement were rated comparatively lower in PNG than Nepal.In relation problem solving, one PNG academic/government respondent was of the view that "OCCD [Office of Climate Change Development; now the Climate Change Development Authority] lacks technical leadership".Another local NGO thought that: "we need political will and commitment from the government of the day to address governance issues and corruption in the country's forestry sector".Concerning agreement, one INGO respondent noted that landowners were not really involved, observing that: "most time it is only the policy makers, and the very owners of the forest are not included."The academic/government respondent commented further that: The Agreements in REDD+ are not that effective because nationals with technical expertise are not often involved while most technical expertise [is] from outside (overseas) to cover up for those at OCCD.There should be a core team of technical expertise including foreigners to be involved in any agreements related to REDD+ in the country.
Both agreement and dispute settlement were perceived not to be such significant issues in Nepal as in PNG.This may be because REDD+ in Nepal has focused exclusively on the country's well-established community forestry programme, which has its own policies, rules, and legislation [59], even if social and environmental tensions exist with REDD+ [60].In PNG, REDD+ is occurring in a larger and more complex forest governance environment as discussed above.Notably, there are likely to be interactions between the reaching of agreements and resolution of conflicts (or failure to do so).As the local NGO respondent who previously commented on agreement noted, "it is often difficult to resolve conflicts over natural resources in PNG".Another INGO respondent explained that it was hard for local stakeholders to meet with those responsible for managing REDD+, as they were based in the urban centres, which meant that: People don't have access to them in discussions to address any issues related to REDD+.It would be better to have offices dealing with these kinds of things placed in the Districts or Provinces so people have access to them, and discuss and settle issues.Sometimes people just discuss amongst themselves knowing that issues will never get resolved because no one is listening and taking their concerns in, so they are just wasting their time.
Clearly given the low score for resources, it is not especially surprising that in both workshops, it was a priority for stakeholders.In the case of inclusiveness, it is possible that despite the high score in the survey, stakeholders were concerned to make REDD+ more inclusive, or to make sure it remained so.In the case of transparency, and in the light of comments from the survey respondents, workshop participants may also have made a link between resources and their allocation in an open and visible manner.In this case, transparency was perceived as a bigger issue in Nepal than that of PNG.This may be on account of the fact that at the time of the ranking exercise, Nepal was fully engaged in a series of REDD+ pilot projects, and research participants may have been concerned about fund details, and the allocation of project resources to the different levels of government (national, sub-national, and local).

Conclusions
In view of the complex relations between the different governance values discussed above, the researchers observe that the legitimacy of governance systems employing only thin values may be limited.The variations between the two cases are also noteworthy, such as the different perceptions of governance quality regarding dispute settlement.The relationship between different governance values is also interesting-for example, the linkages identified by respondents between accountability and implementation.An analysis measuring only transparency and accountability for example would have awarded REDD+ a 'pass', remarking only on the mild difference between the two countries.However, in drilling down to a more granular level, and by including a broader suite of governance values, it is possible to identify areas of shared satisfaction and dissatisfaction.In this regard, the researchers especially note the scores for inclusiveness and resources, as these national level results accord with a range of global surveys of REDD+ governance they have undertaken over the past five years [26,[61][62][63][64][65].It might be easy to dismiss the low score for resources as being a reflection that there is never enough money, or capacity building, to satisfy stakeholders' needs.However, the comments made by respondents, and the linkages they made to issues around accountability and transparency of resource allocation, lead to a further conclusion that this may be a systemic problem throughout the REDD+ programme.This may be further reflected by the prioritisation of inclusiveness, resources, accountability, and transparency amongst research participants at the national level in Nepal and PNG.This might demonstrate that these governance values are important to REDD+ stakeholders in developing countries regardless of their location and socio-political circumstances.In measuring the impacts and effectiveness of REDD+ policies, an emphasis on broader governance values may be merited.This leads to the conclusion that a 'safeguards' based approach, which stresses only a limited set of values, may not be sufficient to carry the whole burden of responsibility for ensuring the 'good' governance of REDD+ at the national level.

Figure 2 .
Figure 2. Prioritisation at the indicator level of 11 different governance values in Nepal and PNG.

Figure 2 .
Figure 2. Prioritisation at the indicator level of 11 different governance values in Nepal and PNG.

Table 2 .
Summary of REDD+ survey participants in Nepal and PNG.