Disempowering democracy: local representation in community and carbon forestry in Africa

All major agencies intervening in community-based and carbon forestry – such as international development agencies, conservation institutions, and national governments – state that their interventions must engage local participation in decision making. All say they aim to represent local people in the design and implementation of their interventions. In practice, decision-making processes are rarely ‘free’, barely ‘prior’ poorly ‘informative’ and seldom seek any form of democratic ‘consent’ or even ‘consultation’. Through case studies of representation processes in forestry programs in the Congo Basin region, Ghana, Kenya, Nigeria, Senegal and Uganda, this special issue shows how forestry interventions weaken local democracy. We show that participatory and ‘free, prior and informed consent’ processes rarely reflect local needs and aspirations, they are rarely democratic and they do not permit participants to make significant decisions – such as whether or how the project will take place. The intervening agents’ choices of local partners are based on expedience, naïve notions of who can speak for local people, anti-government and pro-market ideologies informed by a comfort with expert rule. Although elected local governments are present in all cases in this special issue, they are systematically circumvented. Instead, project committees, non-governmental organizations, customary authorities, and local forestry department offices are recognized as 'representatives' while technical project objectives are favored over democratic representation.


INTRODUCTION
Public management of natural resources around the world is guided by discourses of local participation. Yet, the modes of participation that are based on substantive and empowered forms of engagement and, support and strengthen local democracy, remain elusive in forestry interventions. Recent to consultation, implying that consent is no longer required, thus enabling intervening agencies to choose to override local people's preferences (FCPF 2009). 1 It looks like a green light for project implementers to consult local people and then do as they please.
For the for-profit corporations, the World Bank, and other bilateral and multilateral donor organizations, FPIC draws from the concept of Informed Consent and Participation (ICP) (Baker 2012), which is generally required for what these institutions classify as high-risk projects. ICP involves an in-depth and good faith exchange of views between project proponents and the populations that projects choose to define as 'affected communities' in a way that should influence how a project is carried out. Clearly, consent should be understood as the right of democratically represented communities to approve or reject proposed actions or projects that affect them (Tamang 2005;Lewis et al. 2008;Ribot 2004Ribot ,2013. The issue of consent is one of the most critical challenges facing the development process. Government and project practitioners certainly have the urge to secure consent in order to avoid the risk of project refusal or resistance. Yet, they merely engage in activities that appear to provide local inputs and consent. Both FPIC and participation are problematic means for providing substantive social protection. There is no 'freedom' of choice when people lack the right to refuse predetermined options they are offered (Ece this issue). There is no freedom of choice when 'prior' provides insufficient time to reflect and organise. There are no real choices when in lieu of 'informing' people of their rights and project implications, intervening programs 'educate' them on why they should agree to interventions (Mbeche this issue). There is also no 'consent', certainly not with the toothless form 'consultation', when the community is not engaged as a whole citizenry through local institutions or processes that democratically represent them (Lewis et al, 2008;McGee 2009;Colchester 2010;Baker 2012;Ribot 2015:44-box 16,137-138;Nuesiri this issue, Baruah this issue).
The contributions in this volume show that, in practice, neither 'community participation' in community-based forestry, nor consent or consultation in carbon forestry, are based on substantive exchange of views (Colchester 2010;Ribot 2016:32,44;Marfo 2015; Mbeche this issue; Nuesiri this issue, Ece this issue). They remain mere theatrics. In international, national and local fora, where key decisions regarding access, use and ownership of public forests are taken, consent, consultation and participation are reduced to a mere presence of certain categories of 'relevant' people -defined to be relevant by those managing the process and usually called 'stakeholders'. Whether these 'relevant' people democratically 2 represent local people's interests is, at best, of secondary concern. Project implementers often choose to work with non-democratic local actors, such as experts, customary chiefs, local committees or NGOs. In this respect, intervening agents or agencies, including national governments, forestry administrations, donors, international development agencies or environmental groups -that is, any supra-local institutions 3 that impose laws, programs or projects -fail to understand or to care that who is authorized to speak for or represent local people' matters deeply if participation in decision making is to be substantively representative and legitimate. Our position is that representation is not about talking with interested parties that projects identify as 'stakeholders'; it is not about having customary authorities and other influential elites at the table, nor should it be about consulting some vaguely defined 'population' on pre-determined goals and choices. Democratic representation required for substantive democratic engagement or democratic 'participation' is about broad-based citizen-driven decision-making. It requires democratically representative local institutions.
In this special issue, we use basic principles of local democracy -concerning the degree to which leaders represent and are accountable to local citizens -to evaluate how people are included in decision-making (Manin et al. 1999;Ribot 2013Ribot ,2004. Our definition is substantive rather than procedural. Hence, we acknowledge that electoral systems are not democratic when the leaders lack powers to respond to local needs or when the elections and other accountable means do not establish accountability -as is often the case. We acknowledge that systems without elections can be democratic, and a few are. Nevertheless, we take the position that electoral systems are preferable and represent the most durable and generalisable form of democratic representation (Manin et al. 1999). Looking at responsiveness and accountability, we find in practice that the democratic ideals behind representation, consent, participation or social safeguards are rarely, if ever, evident in international forestry projects. Based on twenty-seven case studies conducted in thirteen countries (Burkina Faso, Cameroon, Democratic Republic of Congo, Ghana, Kenya, Mozambique, Nigeria, Senegal, South Sudan, Tanzania, Uganda, Nepal and Peru) from 2011 to 2015, the Responsive Forest Governance Initiative (RFGI) research program, 4 from which this special issue is derived, found systematic disjuncture between the donor and government claims of participation, representation and inclusiveness in policy-making, and project development and implementation. The articles in this issue describe and begin to explain that disjuncture.
Political representation is responsiveness of government and leaders to people's needs and aspirations. Representation is democratic when that responsiveness is driven by the accountability of decision makers to the people (Manin et al. 1999). Democratic representation requires political processes by which program and project decision makers can (i.e. are empowered to) respond and are also accountable to (meaning can be sanctioned by) citizens (Manin et al. 1999;Ribot 1999,2012). In the country cases in this issue, governments and donors hold up representation of local people in forestry decisions as an ideal. Policies and project documents for all cases proclaim the importance of representation, but do not define it in enough detail to know what it would entail in practice. Likewise, when asked, most development and forestry practitioners cannot define democracy in sufficient detail to be able to systematically establish or support it -they cannot provide a substantive definition. They do not seem to know that it involves significant decisions in which the deciders are systematically accountable to the people of the jurisdiction. This is not to say that they do not believe in democracy. Indeed, they feel committed to promoting it. But, as important as it appears in discourse, policy documents, and in laws, accountable representation remains secondary to project forest management goals. The result is a slippage between stated commitments and practice. The idea of representation is draped over or performed next to forest management goals -the processes set up to engage local populations do not make the projects respond to local needs and aspirations. These processes are an enactment by intervening agencies of their participation requirements. The acts are used to facilitate local acceptance -through convening and consulting with select sub-groups or elites rather than broad-based public representation. Democratic representation is lost in the translation from discourse to documents or laws and then to practice.
Where they exist, and despite their shortcomings (Manin et al. 1999;Ribot 2013; Chomba this issue), elected local authorities with the legal mandate to politically represent the citizens in their jurisdiction could serve as the institutional and legitimate basis for democratic participatory decision-making processes. Yet, as contributors in this special issue show, elected local governments in all of our case studies are circumvented. The sidelining, and thus delegitimation, of these elected local actors weakened these existing structures (Baruah, Chomba, Faye, Mbeche, Nuesiri this issue). 5 Representatives (democratic or not), however, need to be recognized in order to have authority (Ribot 2006(Ribot ,2013Ribot et al. 2008;Sikor and Lund 2009;Thomassen 2011). 6 Bypassing them undermines their authority -not empowering (or disempowering) them and thereby robbing them of relevance, and therefore legitimacy, vis-à-vis citizens. Forestry projects tend to choose to create, empower, and legitimise, alternative local institutions -local line ministry offices (Faye this issue), committees (Baruah, Chomba, Ece, Faye, Nuesiri this issue; also see Manor 2005), and customary chiefs (Baruah, Nuesiri this issue; ; also see Ribot 1999;Ntsebeza 2005). These institutions are rarely democratically representative. Further, many of them, especially project-initiated committees are ephemeral -non-sustainable -and disappear with the comings and goings of project interventions (Murombedzi 2001;Manor 2005). The outcome is a set of institutional arrangements that are neither democratically representative nor sustainable and that disempower and delegitimize duly elected representative local government -in a process that Faye (this issue and 2015) has called 'derecognition'. This volume characterises how and explores why forestry projects and programs systematically circumvent elected local governments by creating their own alternative arrangements for local 'representation' or 'participation'. The studies and broader literature suggest that the preference of the international agencies and donors to work with institutions other than elected local government has practical and ideological roots (Ribot 2004;Manor 2004;Ribot, Chhatre and Lankina 2008). The first practical justification is the need to achieve the forestry objectives of a program (in most of our cases, carbon forestry). Democracy is slow and intervening agents choose to work with local actors most likely to "efficiently" implement their programs and projects. The second, in service of the first, is control -where local forestry brigades, customary chiefs or user committees are easier targets of manipulation by central agencies or donors than are elected local governments. Third are ideological positions that are anti-government (in a Regan-Thatcher sense), pro private sector (part of the Regan-Thatcher neoliberal economistic stance), or pro customary authority (emerging from indigenous people's movements and from a romantic ideology that everything indigenous is good -even if the leaders are colonially created, hereditary and/or despotic). These beliefs often steer projects toward NGOs, project committees, private user groups, companies or individuals, or customary authorities. These ideological overlays may merely serve as a support for implementing the instrumental objectives of the project, i.e. implementation of project goals -that is, they may be an excuse to work with the quickest and most-efficient institutions, rather than with messy and slow democratic processes. In many cases, the existence of customary authorities parallel to and often in competition with other local authority structures, provides intervening agents with a convenient motive to circumvent democratic processes and implement projects through these authorities. Or, these beliefs may be drivers of local institutional choices in and of themselves. Whatever the motive, intervening agents systematically avoid local democracy while favouring NGOs, user committees, chiefs and other private bodies.
In the 1980s, corruption and lack of efficiency of state bureaucracies became an important focus of the new international development policy in Africa (World Bank 1989, van de Walle 2001. The debates on the 'failure' of African states (Bayard 1989) were translated into the policy field as the 'crisis of governance' (Wunch and Olowu 1990). This provided a strong justification for pushing for decentralisation, democratisation and economic liberalisation under the idiom of 'good governance' (Olowu 2003;Törnquist 2007;Kersting et al. 2009). 7 These reforms also legitimated an expansion of national governing arrangements to include NGOs and private-sector institutions (see Frahm and Martin 2009). In this context, supporting NGOs became increasingly equated to contributing to the deepening of democracy, through the creation of a vibrant civil society (Bangura 1992). 8 Indeed, NGOs are conflated with civil society and often considered as 'de facto agents of democracy' (Kamat 2004;see Saward 2008 as an example of this conflation). In this 'inherently democratic and democratising sphere wherein private actors and institutions can flourish' (Mercer 2003), the NGO is also seen to provide an enabling environment for private enterprise. 9 Their proponents claim that NGOs are of and for the people; their representational claims -based on their location within the sphere of civil society -remain largely unquestioned and are thus strengthened (Oyono 2015).
'Forest user groups' or 'environmental/forest committees' are often involved in programs or projects as 'representatives of civil society' at the local scale (see Rutt and Lund 2014). These may be existing entities that forestry projects give new responsibilities and powers, or that which forestry projects create. These user groups and committees are hard to distinguish from NGOs -both are usually private bodies (self-defined as having public interest) that are in most cases paid or supported by governments and donors. Most NGOs, despite the 'non' prefix, are effectively paid agents of higher-level intervening agencies, often national or international governments (Fisher 1997;Shivji 2007;Banks, Hulme and Edwards 2015). Grassroots and civic organizations are often recognized by, and then co-opted into being paid agents of, outside organisations (Fisher 1997;Hearn 2007;Houtzager and Lavalle 2009;Oyono 2015). This special issue shows that forestry projects that claim to be 'community based' -including the UN and World Bank 'Reduced Emissions from Deforestation and Forest Degradation with co-benefits' (REDD+) projects -rely mostly on institutions other than elected local governments to represent local people and for project implementation (Chomba this issue; Nuesiri this issue; Mbeche this issue; also see Anderson 2012;Rutt 2012;Marfo 2015). 10 Traditional authorities are yet another important locus of recognition by forestry projects discussed in this issue. In some countries like Nigeria (Nuesiri this issue), Ghana (Baruah this issue), chiefs continue to play an important role, especially at local scale in forestry programs. In other countries (e.g. Senegal -Faye this issue; Ece this issue), however, traditional authorities are not formally included in forest governance decision-making. This can be partly explained through their incorporation into the state administration after Independence. In countries where traditional authorities have been recognised as part of the post-colonial state governance, they are also incorporated into the forestry projects as 'stakeholders.' Yet, this incorporation is only partial, and is formally subsumed under 'state' authority (Nuesiri this issue, Chomba this issue, Baruah this issue). In such cases, we observe a re-enforcement of the reification of cultural identities, akin to what had been sought by colonial administrations through native authorities. In other words, international forestry programs continue to re-create colonial patterns -a kind of modern reproduction of indirect rule (Mamdani 1996;Ribot 1999Ribot ,1996. 11 At the same time, traditional authorities are not simply 're-invented' through the reification of cultural identity but they are legitimated and reified when recognised (Povinelli 2002). Similar to the NGOs, whether or not they stand for local people's substantive interests is overlooked. While their role as mediators with ancestors and custodians over traditional land continue to make them locally legitimate and relevant (Nuesiri 2012), in some places they are even at times despised for the way they treat local people and for their history of collaborating with oppressive governments (Mamdani 1996;Ntsebeza 2005).
Clearly the transfer of public powers to private bodies can also undermine democracy. Democratic space is closed down, or enclosed, by the privatisation of public forest resources. The public domain -the powers that are vested in democratically representative authorities -is constituted by control over resources and decisions relevant to local people. As those resources and decisions are privatised in the name of local management of forests, democracy is diminished. What were once decisions of elected rural authorities become private decisions of private collectives or individuals, despite the fact that these resources are privatised in the name of local democratic or participatory management. For example, in Ghana, the privatisation of tree tenure (Baruah this issue) and in Senegal the privatisation of use and commercial rights of public forests and the privatisation of charcoal production and trade (Ece this issue; Faye this issue) are important priorities of community forestry projects. In Kenya, private companies are creating their own private governance mechanisms through REDD+ projects (Chomba this issue). This placing of public resources in the hands of private bodies diminishes local democratic space -it encloses the public domain (see Ribot 2004Ribot ,2013Rutt and Lund 2014; Ece this issue).
In short, the local institutions -NGOs, chiefs or private bodies -chosen as partners by international forestry interventions shape local representation, as well as 'participation' or FPIC. These choices are not tangential to forest governance -they define forest governance in those sites. Discourses and practices deployed in the name of participation -under the umbrella of participation, stakeholder consultations or FPIC -seem to provide a justification to support the establishment of necessary conditions for implementation of specific donor and government objectives. The authors in this special issue scrutinise the claims of representation put forth by different actors in international carbon and conservation projects. These programs include community-based conservation and development projects as well as projects oriented towards carbon markets. They analyse how 'democratic participation' is framed in public policies targeting the forests at national and international scales, and how these policies are put in practice in forestry projects.

CHOICE AND RECOGNITION -FRAMING THE RESEARCH
The studies presented in this special issue used what we call the 'institutional choice and recognition' research framework (Ribot 2006;Ribot et al. 2008). Each study examined the dynamics of decision-making processes involved in forest use, ownership and access; and the governance relations inscribed by community-based forest use and carbon forestry projects. We use the term 'institutional choice' to refer to the process by which development agencies, large international organizations, domestic organizations and national governments choose the local institutions or actors to partner with in their forestry interventions. We use the term choice in distinction from its use in neo-classical economics, institutional choice and public choice economics. In our definition, it is about the will of authorities (intervening agents and their institutions) who make decisions, rather than assuming that governing decisions are an aggregate of individual choices articulated through governing actors (Ribot 2013). In this sense, the term choice attributes agency to these actors so that we can trace outcomes and thus responsibility back to their decisions. This is not to say that people's consciousness and agency are not themselves somewhat structured (a la Bourdieu 1977;Foucault 2010). 12 The idea that individuals (or institutions composed of people who set policies and rules) make decisions is relevant for understanding political-economic relations -in this case the choices by higher-level institutions, such as development agencies choosing the local institutions they work with. Through these choices, they are shaping local governance, and in particular, they are shaping rights, recourse, representation, and recognition. Through the discursive justifications of their choices, they bypass local institutions (Rutt and Lund 2014). 13 We emphasize decisions in order to attribute agency to higher-level intervening agents. These intervening agents do decide, and thus should be responsible for, which institutions are being empowered and which are not; they recognise some, fail to recognise others and derecognise -remove powers from -yet other local institutions (Faye this issue). Institutions do not just emerge organically out of Ostrom's (2009) polycentric bodies floating harmoniously in the luminiferous aether of Pareto optimality. They emerge from strategic decisions within structural relations. They emerge in the context of power struggles in a material world.
Institutions are generated by social processes in which some individuals and organisations become recognised as relevant actors. People in need of services -whether it is the support for property claims or a demand for social protections -seek the institutions (or forums) in which their demands have standing and are likely to find favourable response (von Benda Beckman 1981). They shop for, and thus reinforce, effective institutions and authorities. While there is forum shopping from below -that is shopping by the people in need of services and adjudication -that shapes institutions (Lund 2002;Sikor and Lund 2008), we observe important forum shopping from above (Ribot 2006;Ribot et al. 2008). In top-down forum shopping, higher-level forums, such as international development agencies or governments, are shopping for local forums, such as local institutions and authorities to work with. The local institutions they choose have consequences for local rights, recourse and representation -the local institutional landscape. Thus, the word 'choice' (e.g. the choice of local-level by higher-level forums) helps us to explore the logic of how, that is through which institutions, intervening agents engage with local people -based on interests and ideologies within a set of structural and discursive relations and constraints. These choices are consummated when intervening agencies empower or work with, that is when they recognize, local institutions (see Taylor 1994;Fraser 2000;Povinelli 2002;Ribot 2006Ribot ,2007Ribot et al. 2008).
The recognised, and thus authorised, institutions often stand as representatives of certain groups (e.g. forest users, or indigenous peoples). Yet, they may or may not be considered as representatives in the eyes of these people. Although they are empowered and authorised by donors, they may not be considered legitimate, particularly when their roles and actions contradict or run against the needs and aspirations of those they claim to represent. 14 Many theorists (Manor 1999;Agrawal and Ribot 1999;Manin et al. 1999;Ribot 2004) underline that democratic representation and the legitimacy of democratically elected authorities and institutions is measured by their responsiveness to the needs and aspirations of their constituents and their accountability towards them. Pitkin (1967) and Manin et al. (1999) call this 'substantive representation': where representatives 'stand for' and act in the best interests of the represented, and are responsive to, because they are accountable to, the latter -to the degree to which that is possible (on elite capture and the vicissitudes of local democracy) (Bardhan 1997;Manor 1999;Persha and Andersson 2014;Lund and Saito-Jensen 2013).
The articles in this special issue explore the degree to which the institutions being chosen by intervening agents are substantively representative of, that is empowered to respond to and held accountable by, the full range of local citizens. Their research also explored the reasons that intervening agents choose to support the kinds of local authorities they engage with -the degree to which this engagement is about creating substantive democracy or the degree to which it is driven by other concerns -expedience, control, privileging the efficiency of project goals over democracy objectives or ideological concerns. These case studies are in many senses exploratory rather than decisive. Nevertheless, they illustrate the effects of choices being made by intervening agencies on the local institutional landscape and its ability of institutions to democratically represent the citizens of their jurisdictions.

CONTRIBUTIONS TO THIS SPECIAL ISSUE
The contributions to the volume are described below, organised from higher-scale to local interventions. Walters and Ece examine international REDD+ project negotiations in the Congo Basin region; Mbeche and Nuesiri explore national REDD+ Preparedness Programs in Uganda and Nigeria. Ece examines mid-level elected sub-national government's role in a forestry program in Senegal. Faye, Chomba and Baruah present village-level forestry cases in Senegal, Kenya and Ghana.
Gretchen Walters and Melis Ece (this issue) explore power struggles within a World Bank-led REDD+ project that brings together Congo-Basin countries under the Central African Forests Commission (COMIFAC). Over four years of project development and changing REDD+ engagement, they show how the power of the World Bank to allocate resources comes into tension with the sovereignty of each nation, as they sit at the project negotiation table. What appears as an arena of equals for regional cooperation on REDD+ is an ambiguous space in which agendas are fought out. Cameroon continued to engage under these circumstances while Gabon withdrew from REDD+ as a whole and only partially engaged with the project. The authors observe that "countries are expected to adapt their existing forestry policies to REDD+ process needs, often according to guidelines handed down from international climate change negotiations." These impositions occurred despite that REDD+ social protections require conformity with national laws (UN-REDD no date). Under these conditions, those countries with the ability to withdraw and forego REDD+ opportunities were able to set their own agendas. In this case, the donors, as in all multi-and bi-lateral forestry programs, are mandated to work with authorities that represent their nation states on REDD+ -usually officials from forestry ministries. In the remaining articles in this issue, however, we will see that while these ministries are taken to represent 'nations' in the international arena, their domestic practices (in REDD+ and in community forestry) fall grossly short of even supporting democratic principles in local decision-making. While this article does not directly address choice and recognition of local institutions, it provides background for the kind of international context in which international forestry programs and their national components unfold.
Mbeche (this issue) and Nuesiri (this issue) examine representation in REDD+ Preparedness Programs (R-PP) that support REDD+ policy design in Uganda and Nigeria (also see Marfo 2015). They argue that despite the participatory claims of the national R-PPs, national-scale 'stakeholder' consultations are limited to a biased set of non-representative actors. The R-PP policy consultations invited few representatives of indigenous forest-dependent peoples or elected local governments. NGOs and government officials constitute the majority of invitees. Mbeche and Nuesiri found that REDD+ implementers reasoned that local actors lacked the ability to understand the technical details involved in REDD+ and therefore they needed experts to mediate their interests (see also Faye 2015). In practice, they only incorporated local actors who validated decisions that had been made by experts. Further, they circumvented elected local governments whom they felt would 'slow down' or 'politicise' the R-PP (also see Ribot et al. 2008). These two articles show that there is little democratic or deliberative debate in the 'consultations' leading to the preparation of REDD+ strategy documents.
To receive REDD+ finance, developing countries must prepare an R-PP document through a process with input from indigenous and forest-dependent peoples -a claim towards their substantive representation. Both UN-REDD and the World Bank's FCPF require interested countries and project developers to follow FPIC principles in the development of national REDD+ legal and policy frameworks (FCPF and UN-REDD, 2012 ; also see Anderson 2011:15). FPIC is also required from affected populations before the implementation of REDD+ projects (World Bank 2012;FCPF and UN-REDD 2012). The World Bank's social and environmental safeguards also apply to the R-PP, requiring FPIC before REDD+ financing is provided (World Bank 2012). In addition, the R-PP preparation guidelines call for each country "to give forest dependent populations, in particular, indigenous communities a right to negotiate the terms of REDD design and implementation" and requires "inclusion of indigenous and forest dependent communities through their own appointed representatives" (Mbeche this issue; also see FCPF and UN-REDD 2012).
Robert Mbeche (this issue) describes how stakeholder consultation processes for R-PP were carried out by the World Bank and the Norwegian Embassy in Uganda. Here, the World Bank recognized the National Forest Authority (NFA), as REDD+ focal point. The NFA created a secretariat composed of private international consultants, a steering committee of ministers, and working groups populated mostly by NGOs and consultants. The R-PP working groups had no representation from local groups. Sixty-one percent of its members were from NGOs, 23% from central government ministries, 9% from private companies, and 6% from research institutes. Despite there being 2,372 elected rural councillors and five levels of democratically elected local government in Uganda, the Uganda R-PP was centralised and had no direct input from the elected local authorities. In addition, a parallel 'highly participatory' World Bank consultation process with 154 participants included only seven elected rural councillors in Uganda. Mbeche (this issue) also points out that in consultations carried out by the World Bank the government officials were preferred as 'representatives'. In subsequent consultations by the Norwegian Embassy, more importance was given to forms of representation based on ethnic and autochthonous identity claims.
All of these processes in Uganda managed to avoid a complex and time-consuming democratic process, to perform a consultation in order to meet the requirements of REDD+, UN and World Bank participation and FPIC requirements, and to facilitate the quick implementation of the program by 'educating' local people in the technical details of REDD+ implementation. In short, Mbeche (this issue) found that the REDD+ consultations serve the instrumental outcomes of REDD+ by: i) 'educating' the participants in REDD+, particularly on its 'technical' aspects; ii) legitimising the REDD+ strategy in the eyes of the donors by performing 'representation' of local people; and iii) promoting the quick achievement of the REDD+ objectives.
Emmanuel Nuesiri (this issue), working in Cross River State, Nigeria, found that elected local governments were replaced in the National REDD+ consultative process by NGOs and chiefs who were called on to speak on behalf of their members and their subordinate populations. Citizens of local jurisdictions were not represented through a democratically accountable process. The REDD+ preparation process developed a discourse of representation and even of democracy, yet they did not include substantive representation as part of the REDD+ FPIC process. In the Nigeria consultation process, the REDD implementing agents had their own idea of democracy -rather than leaders being accountable to the people, it is a cacophony of voices handpicked by the Nigeria-REDD secretariat to legitimate the UN-REDD process. This arrangement allows the UN-REDD secretariat and organisation (international and national) to choose the actions they see fit while claiming that they have the consent and backing of local people. The farce of representation and pretence of democracy is a necessary step they must go through -what Nuesiri calls 'symbolic' representation (see Pitkin 1967;Törnquist et al. 2009) -in implementing their programs. These are programs that will be implemented whether or not local people need, want or benefit from them.
As in Uganda, UN-REDD in Nigeria produces a highly skewed notion of representation. At the national scale, the NGOs are included in REDD+ policy consultations as 'stakeholders' and as the representatives of 'local people.' Representation in the drafting of Nigeria's REDD-readiness document (supported by the UN-REDD) is also skewed. The majority of the 'representatives' are from the Forestry Commission, followed by the NGOs, the chiefs and 'community forestry groups'. The elected local authorities are not included in the consultations. None of the chosen 'representatives' are accountable and responsive to forest-dependent or indigenous peoples. The choice of NGOs and chiefs reflects the donors' understanding of participation in FPIC while producing mere 'symbolic' forms of representation -it is simply not intended to be substantive. Nuesiri found that the donor's 'stakeholder' approach that included 'all affected' brought in a variety of interested parties, facilitating elite capture rather than citizen-driven decision-making.
Both Nuesiri and Mbeche found that "efficiency in achieving the program goals" was an important criterion reflecting the instrumental rationality in donors' institutional choices at national R-PP consultations. Mbeche notes "efficiency" was constructed not only in relation to forestry and carbon-related technical expertise, but also in relation to the program's budgetary and time constraints. Nuesiri, however, adds that the ability to wield political power and control over resources and people to achieve the program goals can be equally important in shaping donors' rationality of choice, reinforcing the accumulation of power in the hands of government officials, NGOs and local elites, who are neither elected nor downwardly accountable.
Melis Ece (this issue) shows how democratic spaces are closed down through the conversion of elected regional councils into instruments for privatisation of commercial and use rights to public forests in Senegal. The choice to work with regional councils in forestry projects is partly conditioned by neoliberal decentralisation reforms. When the 1996 reforms re-instituted the regional councils as 'local authorities', they also gave them the role of intermediaries of development. Ece argues that this role helped transform the regional council into a locus for business like contractual agreements involving elected local governments and donor-funded "community-based" forestry projects. This shift helped to displace the decision-making on public forests from rural councils to regional councils. However, this arrangement did not prevent the Forest Department's overriding of both rural and regional councils' decisions. It also enabled donor-funded forestry projects to push further forest-based commodity production and privatisation of access over public forests via the regional council.
Ece also discusses the consequences of undemocratic practices and institutional choices of 'community-based' forestry and conservation projects at the rural community-scale. In the region of Tambacounda, the community forests are created for the commercial production of forest products, such as charcoal for urban use. Ece notes that the donor-funded forestry projects, which initiate the creation of reserves, bypass base-level rural councils in key decisions about land allocation and in the drafting of management plans. Further, despite their competing territorial claims, different projects share a similar neoliberal market rationality, oriented towards increasing commodification and privatization of rights over public forests. Ece argues that this market-oriented rationality is also important in shaping the projects' local institutional choices. They put in place and recognise forest management committees as local institutional partners. The project-based local committees are treated as commercial organisations, with exclusive private access to trade of forest commodities and responsibility for forest management.
The article shows that through these arrangements, the legitimacy of the elected local governments (rural councils) is undermined in two ways. First, the rural councils have the legal rights to manage forests and commercial forest activity, yet they are not included in forestry decisions by the Forestry Department and international donors, who prefer to work through 'committees.' Second, the elected rural councils have been subordinated to intermediate-level elected regional councils, positioned as a hub for donor-funded community forestry projects. The decision-making forum of the regional council is hijacked to subordinate the rural councils, to promote commodification and privatisation of rights over public forests and turned into an arena for performance of participatory democracy.
Also in Senegal, Papa Faye (this issue) details how forestry decisions, legally under the jurisdiction of elected rural councils, are taken from these representative bodies and given to non-representative forest management committees created by a World Bank sustainable development project in Tambacounda. He describes how foresters and projects corralled village-based charcoal producers into the invented category of 'local producers', and then delimited their rights as distinct from and less than those of urban-based merchants. After setting these committees up in the name of participation and representation, Faye shows how forestry laws, practices and discourses systematically limited their rights by loading them with the burdens of forest management and depriving committee members of direct access to forestry markets -they had to sell most of their charcoal to urban-based forestry merchants. In parallel to the committees, foresters allowed urban merchants to buy from committees and to also hire migrant labourers to cut wood and make charcoal. Merchants, unlike local producers, were not saddled with the odium of management. This forestry system hemmed forest villagers into subsistence labour while allocating lucrative trade opportunities to the urban elite. Forest villagers wind up poor, living at subsistence on the brink of disaster. This 'sustainable development' project has recently claimed it will support climate adaptation by alleviating rural poverty through revenue generation from the charcoal business, but instead was enriching forest merchants and impoverishing forest villagers while teaching them to cut and carbonize their forests for low wages: this does not seem to be an effective way of reducing climate-related (or any other) risks. It is also not strengthening the long-term democratic representation of forest villages in forestry or in any other decisions as it is sidelining and delegitimising the elected rural councils.
Faye's contribution illustrates also how the project's institutional choices and the supporting technical justifications are subject to contestation by elected local governments. During the first phase of the World Bank project, the forest service rationalised their choice of committees with specious technical claims arguing that the committees had the capacity to implement technically required management. Their actions showed they valued implementation over the procedural concerns of democracy (Ribot 2008). The forest service and projects used a logic of technical necessity to override local concerns and to circumvent the elected local governments (community councils). Faye also shows, during the course of these management impositions, that: The stakes involved in the control of access rights and the institutions that enforce them had become very clear to ELGs [Elected Local Governments]. Indeed, PCRs [presidents of the elected rural council] understood that in order to respond effectively to local needs and to secure political visibility for themselves, they needed access to resources and the ability to exercise authority. Therefore, they began resisting the technical claims as much as they could, mobilising political arguments that are deeply rooted in decentralisation laws, and working to control the revenues flowing from forest-related activities.
After the first phase of the project, the elected councils challenged the domination by taking decision-making back from forest service and the project created committees -a demand from elected local governments for observance of national law. Here elected councils chose themselves. In the project's next phase, however, the project returned to re-establish its non-democratic committee-based implementation units. They then made these units into 'associations', which are private organisations that elected local governments will not be able to dissolve at the end of the next project phase. The elected local governments believe that the government's logic for creating associations was to override elected local governments (Faye 2015a(Faye , 2015b. In her article, Manali Baruah (this issue) focuses on a Community Resource Management Area (CREMA) that encourages farmers to plant timber trees on cocoa farms in South-Western Ghana. The CREMA approach has been promoted as a collaborative form of natural resource management that would help generate conservation and financial benefits for communities living around the protected areas. However, CREMAs like the one analysed by Baruah, are also implemented in off-reserve degraded forested areas to increase timber production and introduce tree tenure privatisation. In Ghana, the elected District Assemblies' powers had been further limited after the reversal of decentralisations in 1996. Together with the chiefs, who grant access rights to land, forests and trees in their traditional 'wassa' areas, the state forestry administration retains the key decision-making powers over the commercial exploitation of off-reserve forests and trees. In this context, the ownership of the trees, access to tools, seedlings and benefits (including carbon benefits) need to be negotiated, a task undertaken by a local NGO positioned as the CREMA project implementer.
Baruah shows that despite claiming to strengthen democratic resource management through CREMAs, donors reinforced the existing centralised power relations by circumventing the elected district assemblies, recognising traditional elites and chiefs as de facto representatives in CREMA management and, choosing private entrepreneurs as the project implementing "local NGOs". Similar to other cases presented in this volume, the project established natural resource management committees to manage the CREMA. Here, however, chiefs occupied a prominent role both at regional and village-scales. Members of the committees were selected and nominated by the chiefs, often in the presence of the district forestry officers.
In the process of establishment of the CREMA by the Forestry Commission and in the drafting of its bylaws by international donors, the District Assemblies were not even 'consulted'. In this sense, foresters further weakened and delegitimised elected local authorities by diminishing their role in public decision-making. While the choice to work with chiefs and NGOs is justified on grounds that these bodies somehow represent the 'public', the fact that they are not elected or accountable to the people makes them effectively private bodies. Simultaneously, to the degree that they are accountable to donors and the forest service who empower them to implement forestry activities, they are effectively administrative branches of these two extra-local agencies. So, through both this effective privatisation (the degree to which these non-state bodies have new discretion) of public resources and through external administrative control, the space of local public decision-making discretion is reduced; substantively weakening local democratic representation.
Of course, if in Ghana participation means allowing private groups such as local elite-led NGOs and 'community-based initiatives' to engage in and implement resource use decisions on behalf of the local people, then this non-public form of inclusion opposes democracy. Indeed, more democracy here would mean less privatisation since democracy needs public resources and public decisions (a sphere of public decision making) in order to play a democratic role. Forests in Ghana could, if managed by local representative authorities, provide a collective local public domain. The contradictory talk of participation, representation and accountability while promoting private decision-making and benefit, needs to be evaluated. Perhaps private forestry decisions and use can increase efficiency or even make forests more lucrative, but at what cost to fledgling rural democracies? Democracy is not being chosen by these intervening agencies -even if it is one of their stated values. The CREMAs are the proposed basis of REDD+ Readiness in Ghana, yet they are not supporting community representation. If participation means implanting the administrative orders of external agencies it is also not democracy. In this sense we have two enclosures of the discretion of public democratic authorities; even the limited powers devolved to elected authorities are taken away and given to private bodies and others are retained by central agencies as required activities of the intervening agencies.
Membership in CREMAs is exclusive despite that CREMA and donor guidelines insist on elections and inclusiveness. The implementing agents excluded anyone whose interests or activities were not aligned with the objectives of the CREMA. They were also excluded from benefits derived from its activities and from receiving material from the NGO. Although implementing NGOs, like IUCN along with the Ghanaian forestry and wildlife line offices were aware that the process by which the committee members are selected was undemocratic, they chose not to interfere. "Their position is justified by the necessity to adhere to project 'timelines and objectives'." The rationale given by implementing agencies for choosing to work through local NGOs and Chiefs included the merits of community-based initiatives, civil society engagement, donor mandated 'participatory' processes, regard for customary systems and perceived lack of capacity of the local governments. However, the private interests behind these institutional choices were often economic and political. The donors and foresters rationalise their lack of engagement with representation and accountability via institutional mandates, technical and managerial goals.
Susan Chomba (this volume) observes a key dilemma of representation faced by a prominent private-sector REDD+ project in Kenya. Do forestry programs work with unaccountable and ineffective local governments or should they circumvent them by creating their own more-effective committees? Kenya is transitioning from a centrally managed form of local administration towards empowered elected local governments. Prior to the 2013 general election, local administration was made of weak elected local councils without financial powers running in parallel to strong local chiefs appointed by the provincial administration. The 2010 constitution, however, mandated the transformation of the councils into democratically elected 'decentralised' county governments with political and fiscal powers and responsibilities. The 2010 Kenyan constitution also provided for the centralised system to be phased out. Given the representative roles that elected leaders are designed to fulfill under the constitution, they would have likely been the first choice had the project been aiming to support and work through democratically elected local governments. But, the local elected leaders, elected before 2013, were circumvented. Project staff argued that this was to avoid political interference or because they viewed them as corrupt. After the new constitution, the project would not work with the newly elected local leaders because the necessary resources and infrastructure were not yet in place. On the other hand, the centrally appointed local officials made of chiefs were directly engaged in the project in the beginning, but later only as ex-officio members because they were similarly perceived as unaccountable, corrupt and inefficient. So, the REDD+ implementing agents largely skirted newly elected institutions. Instead they created their own single-purpose 'elected' Location Carbon Committees (LCCs), that worked with Community Based Organizations (CBOs) to implement REDD+ at the local level.
The project, being part of REDD+, had a fraught choice in meeting the universal principles of participation, representation and democracy under the United Nations conventions and declarations, as well as those mandated for REDD+. Their choice to circumvent the 'representative' local government institutions appeared to contradict the REDD+ safeguard principle of ensuring full and effective participation of all relevant stakeholders. Indeed, while the project could claim that their carbon committees appeared representative on account of their being elected, the committee elections were hardly democratic. In a community meeting, voters had to bow their heads so they could (wink wink) not see others and raise their hands to be counted by the administrative chief. In this case, voters had to trust the chiefs to count correctly and feared that others would tilt their heads to see who they voted for. Of course, voters might also not feel free to express their positions in front of powerful chiefs. In short, this system was nothing like a secret ballot that would constitute a fair election. In this sense, the REDD+ decisions and processes reflected performed participation and representation while elected local government was left on the sideline.
The studies in this special issue provide democracy lessons for REDD+, carbon forestry, and any participatory forms of natural resource management. Each case illustrates conundrums that projects face when required to represent, negotiate, and protect local interests. In community-based forestry in Senegal, as well as in REDD+ carbon forestry in Nigeria, Kenya, and Ghana, the intervening agents require representation of local people in decision-making in order to protect local people's interests. The Ghanaian and Kenyan cases show how representation safeguards are compromised by an inappropriate circumvention of local government in favour of chiefs, NGOs, CBOs and local committees. Yet, despite their failures, these modes of participation are held up as the model for how representation will be conducted when REDD+ is implemented nationally. The cases all show that representation is given second tier to implementing of other forestry project goals. Representation is required, but merely performed as a theatrical 'symbolic' enactment (Nuesiri this issue). Substantive representation, even when there are democratically elected local authorities, is nowhere to be seen. While we would have liked more analysis of the rationality of institutional choice -the politics of choice and recognition --these papers just begin that part of the analysis.

CONCLUSION
Elected local governments would seem to be good institutions to represent local people in local decisions. But they are not given the opportunity to do so. They are ignored and avoided. This is not because they are weak or even culturally inappropriate. It is because international development agencies -certainly those working on natural resource management -choose not to work with or through them. Local democratic institutions exist in most places. Yet, environmental projects and programs choose to work with institutions that operate in parallel to elected local governments. This choice, unfortunately, perpetuates the view that existing local democratic institutions are incapable of responding to local needs -it makes them appear irrelevant or incapable, without giving them a chance to prove otherwise. Local governments sit powerless on the sideline. Interventions create and work with alternative institutions at a moment when governments across the developing world have legislated into existence new elected local democratic institutions (Crook & Manor 1998;World Bank 2000;Ndegwa 2002). Indeed, these choices support local institutions that masquerade as representative but remain accountable to donors, private organisations, line ministries or to an identity-or interest-based sub-section of the population. They foster committees, NGOs, chiefs, and other private bodies while generating a specious image of representation. The studies in this special issue documented how projects and government agencies chose to create and work through parallel institutions in forestry in Ghana, Kenya, Nigeria, Senegal and Uganda.
The 16th United Nations Framework Convention on Climate Change (UNFCCC) Conference of the Parties (COP16) in Cancun in 2010 established safeguards for implementing the largest global forestry project, REDD+. The Cancun Agreements called for actions that "complement or are consistent with the objectives of national forest programs and relevant international conventions and agreements" (UN-REDD no date), but, they did not address the fact that most national forestry programs fail to adequately uphold local representation or human rights. They called for "transparent and effective national forest governance structures, taking into account national legislation and sovereignty" (UN-REDD no date), but failed to recognize that transparency is only effective where there is sanction 15 (Fox 2007) and that national legislation often fails to provide for local representation or rights. They called for "respect for the knowledge and rights of indigenous peoples and members of local communities…" (UN-REDD no date), yet failed to acknowledge that many people who have lived in forests for generations are not indigenous and require equal representation and protection; safeguards should not be creating second-class citizens. They demanded "full and effective participation of relevant stakeholders, including, in particular, indigenous peoples and local communities" (UN-REDD no date), without defining stakeholders as citizens rather than merely interested parties. Indeed, why should a non-resident merchant hold a 'stake' in someone else's forest just because they stand to profit from it? They may have an interest, but the claim of rights or a 'stake' in the public resource belongs, by definition, to citizens. Stakeholders are indeed citizens if and when they live in the jurisdiction that presides over the public resource in question. So, let them vote if they want a 'stake' in the public resource; they should not be having a binding decision-making power at the negotiating table. They can and should inform (or misinform) decision-making processes, but they do not deserve a disproportional say in decisions. The decision should be by accountable representatives. Decisions over public forests and other natural resources should be made in a democratic manner, rather than in proportion to the interests of 'stakeholders' recognised by actors within line ministries or projects.
The articles in this special issue show that while all forest carbon and conservation programs studied, aimed or claimed to represent local people's needs and aspirations in their decisions, generated very little representation that might be called democratic. Why is this so? Why do project and program managers choose to circumvent elected local government? Given that the safeguards outlined in the Cancun Agreements do not call for representation, it is not a shock that representation does not result from their application. They call for conformity with the status quo and the favouring of the influential, defining them as 'stakeholders' and giving them stakes in decision proportional to their interests. Several other motives for choosing to avoid elected local governments also emerge from the studies. First, elected local governments are avoided due to ideological favouring of markets and privatisation (Ece this issue). These lead to enclosures that remove public decisions from democratic institutions and shift them to the private domain (ranging from individuals and corporations to NGOs). Second, elected local governments are avoided due to collusion between forestry service or project authorities and wealthy actors involved in lucrative activities where democracy or elected local authorities might undermine wealthy elites (Ece this issue; Faye this issue). As we well know, democratic representation can result in redistribution -and the rich and powerful do not seem to want to give up their wealth and power. Third, elected local governments are avoided because many intervening environment and development agents believe that civil society and stakeholder approaches or customary authorities constitute democracy (Baruah this issue; Faye this issue; Mbeche this issue; Nuesiri this issue). While some of these processes or actors may represent people, they lack systematic accountability to the people as a whole (who we consider to be the residents of the jurisdiction where the public resource resides) and therefore they are not democratic (Manin et al. 1999). Fourth, local government is avoided because democracy is a slow and laborious process that requires time and resources, making it an unlikely choice by the agents under pressure to implement forestry management or carbon programs. Fifth, given that many local people might object to the very programs being implemented, their inclusion and consultation may be inconvenient and threatening to the project personnel trying to make interventions -they being under great pressure to demonstrate success (Baviskar 2004). A sixth reason is that many technical agencies and project experts feel that the decisions are technical and belong in their expert hands. So, technical necessity is often used as an excuse for centralising decisions with line ministries and project staff (Faye this issue, Mbeche this issue, 2015;Mitchell 2002;Easterly 2013).
Many intervening agents also believe that local governments are corrupt or inefficient, so they circumvent them (Chomba this issue). This may be true, but if the same intervening agents tried to circumvent a corrupt local government in the United States or Europe, even to implement a park management project or build a playground, these agents would find themselves quickly incarcerated. It is not acceptable to circumvent government agencies just because (or even if) they are corrupt. Many are corrupt (Bardhan 1997). So are many corporations, NGOs and chieftaincies (Ntsebeza 2005;Temudo 2015). Rather than circumventing corrupt local governments, it is incumbent on anyone wishing to intervene to work to make those governments more accountable and effective. Along these lines, Chomba (this issue), working in Kenya where agencies judged local government to be corrupt, recommends that in the short run, REDD+ projects can work with local committees and community-based organizations, but should place them under the authority of elected governments (see Ribot 2004). Integrating these institutions and decision making into elected local government -but establishing project-imposed checks and balances in the form of committee deliberations, public meetings, audits and other public accountability mechanisms -would get communities involved and help them to learn critical lessons on how to articulate their needs to elected leaders and how to hold their leaders accountable. In the process, it would make these leaders relevant and worth holding to account.
Instituting democratic governance in the long run will require the implementers of REDD+ -and any other laws, programs or projects -to entrust democratically elected local government with resources and discretionary powers. People will then learn to trust them when local governments have been made accountable through the normal politics and multiple accountability relations that ensure democratic practice. The Responsive Forest Governance Initiative 16 that generated the research presented in this special issue also produced a set of guidelines for supporting local democratic processes while implementing natural resource interventions (Ribot 2016(Ribot ,2017Barrow et al. 2015). Despite the fact that there are many structural and political-economic obstacles to supporting elected local government, those guidelines can be boiled down to two words: choose democracy.

NOTES
1. The meaning of consent in FPIC has been debated since FPIC's origins in the drafting of the United Nations Declaration on the Rights of Indigenous Peoples. Uncertainties remain regarding the relationship between consultation and consent. In practice, the interpretation of 'consent' is decidedly narrow. In 2000 the World Bank commissioned an independent report on Extractive Industries to determine whether resource extraction aligns with the Bank's mission of poverty alleviation and sustainable resource use. The evaluation report (IFC 2003) argued for FPIC to empower communities with self-determination. However, the World Bank was concerned that 'consent' in FPIC would promote veto by individuals and groups, and opted instead for 'consultation' geared towards gaining the broad support of affected communities. FPIC differs from mere consultation in several important ways, but most significantly in the way decision-making authority is exercised and legitimated. Consultation requires only an exchange of information between project implementers and communities, but does not involve any transfers of powers. It does not require that information be exchanged or that final project decisions be based on the views and needs of the affected communities. Thus 'C' has effectively been re-interpreted to mean a duty to consult without the obligation to obtain consent. We contend that the objective of consultation must be to obtain FPIC Consent.
2. Representation is the responsiveness of leaders to the people. Representation becomes democratic when that responsiveness is driven by accountability of those leaders to the people (Manin et al. 1999). Achieving responsiveness with accountability requires leaders who have powers to respond -that is executive, legislative and judicial powers backed by resources such as knowledge, finance and bulldozers. It also requires citizens who have means to reward or punish, that is to sanction or hold accountable, their empowered leaders. Sanctions can include elections or many other means -such as embeddedness of leaders in community, third-party monitoring, courts, information diffusion, threats of violence, shaming, etc. (see Ribot 2004:annex A;Agrawal and Ribot 2012). Elections are important despite being systematically inadequate on their own (see Chomba this issue). So, accountability can also be achieved without elections,although we see elections and the standard procedures of democratic government as an important way of legislating and institutionalizing -making sustainable and geographically complete -democratic practices. Non-election-based systems can certainly be democratic -as long as there are means of accountability. But representation is not always democratic even when it is effective, appreciated and viewed as legitimate. There are many systems in which people are well represented without accountability of leaders. These are systems in which the leaders are ideologically committed to being responsive to local needs and aspirations. We call this benign dictatorship. But, benign dictatorships of this nature are still not democracy and are always at risk of deviating from popular desire. Electoral systems are also at great risk of being undemocratic, when they are poorly structured, lack powers to respond and lack multiple accountability relations beyond elections.
3. Here, we focus mainly on the institutional aspects of organizations. 5. This delegitimisation can, of course, weaken the entire democratic institution, or just its functioning in the domain of forestry.
6. Thomassen (2011) analyzes the relationship between representation and recognition from the perspective of identity politics.
7. Although the World Bank was the leading institution that spearheaded the good governance approach, multi-lateral and bi-lateral donors adopted quickly a similar approach in their interventions. Policies and measures promoted under good governance differed from one donor to another. For example, Oluwu et al. 2009 argued that the WB's approach was less focused on 'democracy' and more on economic liberalization.
8. This is not to say that there was no revival and proliferation of different grass-roots organizations and associations during this period of liberalization. Similarly, international development interventions also elicited the demands for participation in decision-making from below. However, here we would like to draw attention to the emergence of a new framing of 'civil society' in development discourse, and to attending problematic assumptions regarding representation.
9. Intervening agents have long viewed NGOs as an alternative to centralised state administrations in implementing development and conservation interventions. This trend emerged during the implementation of structural adjustment policies, and is associated with a broader shift in development policy thinking and practice in Africa and beyond (Harrison 2005;Kamat 2004).
10. For an excellent critical overview of REDD+ see Lund, Sungusia, Mabele andScheba 2017. 11. Ribot 1999 argued that the appeal to all of these non-democratic local authorities, chiefs (whether or not part of the current state administration), NGOs, and even local administrative units of government is a modern reproduction of indirect rule. In all cases it is the management of rural subjects.
12. As in legal theory, we stop process tracing analysis of cause at the first recursive node of human will to identify what we would call an agent (Holmes 2009). Further, in our framing objects do not have agency, thus we trace causality to human will; objects do not think and therefore cannot be responsible, humans do think (Arendt 2003). In this way we avoid an infinitely looping analysis of the relation between agency and structure.
13. Rutt and Lund 2014 describe how projects choose civil society organizations based on their prior partners. They chose partners they know they can work with.
14. One complication to this scenario is that donor-recognised representatives can use the powers and means transferred to them to strengthen their client networks, or to use these means in a way to meet their constituents' needs and aspirations, in order to become locally legitimate.
15. Transparency does not cure the widespread 'transparent corruption', corruption that is already visible to all.