1 Introduction

The Earth System Governance Project emphasizes the need for further research on questions of accountability and legitimacy of international institutions, at the theoretical level, and on the practical mechanisms for addressing existing shortcomings (Biermann et al. 2009, 53; Biermann et al. 2010). One facet of this burgeoning field of research relates to the legitimacy of intergovernmental organizations and rule-making processes; another one to the legitimacy of the myriad of private actors who have become increasingly important participants in global governance over recent years (Held and Koenig-Archibugi 2004). This article aims to bring together some of the questions arising from the debate on the legitimacy of these two types of actors in the area of global climate governance. More specifically, the article seeks to link a discussion of the demands by non-governmental organizations (NGOs) for more equitable forms of representation and participation in the United Nations Framework Convention on Climate Change to an analysis of the representation and participation practices within the NGO community itself.

The dual focus of this article is justified since the contribution of NGOs to global policy-making is often presented as a double-edged sword: on the one hand, their participation is seen as a potential remedy to some of the existing legitimacy deficits in international institutions; while on the other hand, the accountability and legitimacy of these private actors themselves is also increasingly subject to critical scrutiny (Biermann et al. 2009, 53; Scholte 2004). This tension is further compounded by the fact that NGOs play a crucial role in highlighting existing legitimacy shortcomings and calling for higher standards of equitable representation and participation in global institutions. This article therefore addresses two separate but related issues. It shows firstly how NGOs have responded to the perceived participation and representation deficits in the climate convention by analysing their proposals for more equitable forms of representation and participation in climate convention-related climate policy-making. It then turns its focus to the NGO community to consider the extent to which these groups themselves are able to put into practice comparable standards in their own internal decision-making processes. The analysis is situated in a broader theoretical discussion of the problem of democratic deficits in global governance and the role of civil society actors in this context.

The empirical findings presented here are mainly based on an analysis of documents produced by NGOs, published between the meeting in Bali in December 2007 and prior to the fifteenth conference of the parties in Copenhagen in December 2009. These include the written statements by environment and development NGOs submitted to the climate convention from late 2007 to mid-2009 and all available editions of the civil society newsletter ECO published between December 2007 and August 2009 (Bali 2007, Bangkok 2008, Bonn 2008, Accra 2008, Poznan 2008, Bonn I 2009, Bonn II 2009, Bonn III 2009). Besides the NGO submissions to the climate convention, I have also reviewed selected NGO-published materials on climate-related issues produced for a wider or policy audience. The discussion in sections four on representation practices within NGOs uses existing secondary sources on these organizations, together with primary documents relating to their governance structure and internal decision-making procedures and policies. The analysis of policy documents undertaken for this article has been further supplemented with a number of interviews with NGO representatives. The focus of my analysis is primarily on the written statements put forward by CAN due to its prominent position within the climate convention NGO community; however, I have also included the separate submissions by individual environmental and development NGOs (such as the Worldwide Fund for Nature, Greenpeace International, FoE International) and other NGO networks working on climate change (e.g. Ecosystems Climate Alliance) that are available on the climate convention website.

In this article, I use the term ‘international NGO’ for groups not associated with business or governmental interests, which have either constituent members (national organizations) or conduct operations in several countries (usually both). The NGOs I am focussing on are also engaging with intergovernmental organizations as part of their advocacy (and service delivery) work and probably constitute the “category of civil society organizations (CSOs) with most presence in UN system policy forums” (McKeon with Kalafatic 2009, x). In the case of the climate convention, this category of NGOs is predominantly made up of environmental and development NGOs, with many of the most influential and active groups based in Northern countries. The concept of ‘affected communities’ is deliberately left unspecified—any definition of ‘affectedness’ is likely to be contested,Footnote 1 which may be why NGO practitioners and policy-makers also tend to employ this concept very loosely. In this article, the concept of ‘affected communities’ is linked to concerns about their potential and actual marginalization from decision-making processes that impact their lives.

The next section will give a short overview of the theoretical and policy debates that provide the background to the questions addressed in this article. Section 3 distinguishes between two types of ‘external’ NGO responses to the perceived representation and participation deficits in the climate convention: responses to address representation inequities at the intergovernmental level, and proposals designed to strengthen participation by societal stakeholders at all levels of climate policy-making. Section 4 uses the case of the Climate Action Network (CAN) to illustrate how forms of representation by the NGOs themselves could conceivably present a third (‘internal’) form of response to the perceived representation deficits at the international level.Footnote 2 Section 5 offers a critical discussion of the potential problems associated with the mechanisms outlined in the previous sections and highlights various issues for further research.

2 Non-governmental organizations to the rescue: representation and participation deficits in global governance

The democratic legitimacy of intergovernmental rule making relies predominantly on the assumption that citizens are represented through their governmental delegates. In practice, however, this model of democratic representation encounters a number of serious challenges and citizens and local communities may be underrepresented in international organizations for a range of reasons (Dingwerth 2007; Held 2004). One obvious problem is the lack of democratic institutions at the local and national level—governmental delegates are often neither democratically authorized nor accountable to their citizens. However, even in functioning democracies, citizens may only have limited access to (and interest in) opportunities for holding governments to account for their actions in international negotiations. Crucial reasons for this are a lack of transparency and available information on international issues. A particular challenge to democratic representation in international organizations arises in the case of minorities within different states, such as indigenous peoples, that are consistently underrepresented domestically and therefore unlikely to be effectively represented by governmental delegates (on indigenous people, see Schroeder 2010). Moreover, the design and features of the international organizations themselves leave much to be desired from the perspective of democratic legitimacy. Many of the world’s poorest countries are formally and/or informally underrepresented, which means that the citizens of these countries have even less scope for influencing policies that affect them. The international financial institutions in particular are criticized for their donor-driven systems of governance, which fail to give sufficient voice to those countries most affected by their policies (Woods 2001; Stiglitz 2003). Some observers point to the fact that more and more power is vested in informal groupings of states with only limited number of members, such as the Group of 8 or the Group of 20 (Forman and Segaar 2006). These formal inequities are compounded by capacity and resource imbalances that sometimes also undermine the effective participation by poorer countries in international negotiations.

The problem of representation failures is further compounded by the fact that those communities who risk being marginalized from intergovernmental rule making for the reasons discussed above are often the same communities who are particularly vulnerable to the global problems that these institutions are set up to address (such as climate change) and also disproportionately affected by development, environment or trade policy-related decisions taken at the global level. As a result, the interests of the most vulnerable are often unrepresented or passed over in the making of those decisions that are likely to affect them the most (Mason 2005, 11). This discrepancy between democratic control and actual or potential affectedness risks undermining “the congruence between the ‘people’ that is being governed, and the ‘people’ that is supposed to govern” (Scharpf 1998). In the context of environmental politics (but also in other issue areas), this means that international institutions are unlikely to pass the democratic litmus test that “all those potentially affected by risks should have some meaningful opportunity to participate or otherwise be represented in the making of the policies or decisions which generate such risks” (Eckersley 2000, 118).

A number of reforms have been implemented within international organizations over recent years, many in response to growing concerns about democratic shortcomings. These include, among others, initiatives to increase the transparency of intergovernmental organizations, changes to the formal governance structures to give more voting rights to Southern governments and a marked increase in the level of engagement with civil society actors (Conca 1996; Woods 2010). These institutional changes are designed to contribute to a more balanced representation of interests and perspectives and hence intended to strengthen the perceived legitimacy of the organizations. My concern here is primarily with the idea of greater civil society participation as one possible remedy to the legitimacy shortcomings of international organizations (Scholte 2004). In practice, efforts designed to enhance the participation of ‘global civil society’ are frequently synonymous with granting NGOs greater scope for involvement with intergovernmental organizations (McKeon with Kalafatic 2009, 17, Sands 1998, quoted in Mason 2005, 30). The Report of the Commission on Global Governance, for instance, states that “Global Civil Society is best expressed in the global non-governmental movement” (Commission on Global Governance 1995, 254).

The normative case for involving civil society organizations such as NGOs in intergovernmental policy deliberations rests largely on the assumption that through the participation of these private actors, the voice of a wider affected public may be brought to bear on the global decision-making processes. This implies that the participating private actors derive some of their own legitimacy vis-à-vis global institutions from their relationship with particularly affected groups on the ground. Hudson (2001, 342), for example, argues that an NGO’s attempt to justify its advocacy to the World Bank is often based on the quality of its relationship with its Southern partners. Princen points out that environmental NGOs are uniquely positioned as transnational actors to bridge the gap between top-down and bottom-up approaches to sustainable development by “linking the local to the international levels of politics” (Princen 1994, 33). By acting as a link between the two levels, environmental NGOs make it possible for local communities to ‘upstream’ their grievances to international institutions, in the hope that changes at the global level will protect their interests more effectively than local politics alone (Princen 1994, 40). Keck and Sikkink’s ‘boomerang pattern’ of transnational politics captures a similar dynamic (Keck and Sikkink 1998). This model implies that local and national NGOs, whose domestic channels for representation are blocked, follow a strategy of reaching out to international allies, such as international NGOs and transnational advocacy networks, to build pressure on their governments from the outside (Keck and Sikkink 1998, 12).

There are therefore two broad (although far from uncontested) ideas in the literature that have also managed to gain credence with many policy-makers: the notion that the participation of civil society actors such as NGOs contributes to the democratic legitimacy of international organizations; and the idea of international NGOs as ‘links’ between affected and potentially marginalized local communities and international institutions. These two ideas are often brought together, for example, in the assertion that NGOs (the authors use the term ‘organized civil society’) have “the potential to function as a ‘transmission belt’ between a global citizenry and the institutions of global governance” (Steffek, Kissling and Nanz 2008, 3). Similarly, Payne and Samhat maintain that their “participation lends voice to excluded constituencies in global politics” (Payne and Samhat 2004, 27).

While this article accepts the idea that NGOs have a potentially valuable role to play in strengthening the participation and representation of affected communities in international organizations, it suggests that there is a need for more research on the practical mechanisms through which this can be achieved by the NGOs. The article attempts to provide further empirical insights on this issue by looking for the potential mechanisms through which NGOs may be said to be pursuing this objective in the context of the climate convention. It asks, firstly, to what extent the NGOs seek to directly address institutional deficits that potentially preclude fair and balanced representation and participation in the climate convention and provide an overview of the NGO proposals for tackling these problems. The analysis then turns to the NGOs themselves and examines their potential to act as ‘representatives’ (this notion will be discussed in more detail in Sect. 4 below). It suggests that, despite not being democratically elected in the same way as domestic governments, NGOs may still be engaged in representing particular constituencies, on particular issues and in a particular context. Provided that the relationship between the NGO and the constituency in question exhibits features of democratic responsiveness (linked to ‘functional equivalents’ of authorization and accountability (Castiglione and Warren 2006, 15)), the NGO could argue that it has some ‘mandate’ to represent this constituency in particular circumstances. Moreover, it is then worth asking whether the constituencies thus represented by NGOs are, in fact, the same communities who are potentially marginalized or underrepresented in global policy-making processes. A positive response to this question would indicate that NGOs are indeed able to act as ‘transmission belts’ between particularly affected but marginalized communities and international institutions or as democratically legitimate links between the local and the global (Steffek et al. 2008, Princen 1994).

3 Visions for more democratic climate governance: non-governmental organization responses to representation and participation deficits in the climate convention

The purpose of this section is to set out the NGO demands and suggestions for correcting perceived representation and participation deficits in the climate convention. The empirical analysis of relevant NGO documents shows that it is possible to distinguish two broad types of proposals in the NGO positions: those designed to remedy particular representation inequities at the intergovernmental level (that is, among the parties to the convention) and those intended to strengthen societal access to the relevant levels of climate policy-making.

3.1 Non-governmental organization responses to representation inequities at the intergovernmental level

Firstly, NGOs are responding to concerns around potential representation inequities at the intergovernmental level by supporting demands for ‘fair’ representation. The focus here is on the question of weight (Koenig-Archibugi 2006, 14)—how fairly are representation rights distributed among the member states? Linked to these procedural concerns is the question of content—what issues are included in the agenda and in the final policies? It is interesting to note that many of the NGO documents surveyed here include demands for ‘fair’ representation by governments; the NGOs are also concerned with drawing attention to issues that are of particular relevance to potentially underrepresented countries in the negotiations. In practice, procedural demands for equity tend to go hand in hand with these substantive demands although this analysis tries to focus on the former. Overall, demands for equitable representation rights among governments feature less frequently in the NGO documents than demands for greater participation by societal stakeholders (discussed in the next section).

As a UN treaty, the climate convention system of representation corresponds to the one state—one vote arrangement. As such, the formal allocation of votes among countries proves less problematic from the viewpoint of democratic legitimacy than alternative arrangements based on financial contributions or economic power that are used, for example, in the Bretton Woods institutions. Of course, formal equality among the parties does not preclude considerable power imbalances due to the different distribution of capacity and resources among member states when it comes to the reality of political bargaining. Moreover, there are still many uncertainties regarding governance arrangements for possible new or reformed institutions, particularly with respect to climate finance (Mueller 2009). The submissions to the climate convention by environmental and development NGOs contain a number of references to the need to ensure ‘fair’ representation for developing and vulnerable countries in these institutional arrangements. They also advocate the principles of country ownership and subsidiarity, which are intended to strengthen the influence over climate-related policy-making of those countries most affected by and vulnerable to the impacts of climate change.

The detailed draft design for a future Copenhagen treaty submitted by a group of environmental NGOs illustrates this point.Footnote 3 The proposal stresses that the governance structure of the proposed Copenhagen Climate Facility “should reflect a democratic decision-making structure with an equitable and balanced regional representation, ensuring significant representation from developing countries”.Footnote 4 Moreover, “Securing the representation of the most vulnerable countries should be a priority, as they will be most impacted by unchecked climate change”.Footnote 5 The Worldwide Fund for Nature makes a similar point in its proposal for Technology Action Programmes, which envisages an Executive Board of Technology “with balanced representation of developing and developed countries”.Footnote 6 There is some—perhaps deliberate—ambiguity regarding the precise meaning of ‘balanced’, ‘equitable’ or ‘significant’ in these submissions. While some NGOs shun away from specifying a clear ratio in their proposals, others are explicitly demanding majority representation for developing countries. CAN, for instance, calls for ‘equitable representation’ and ‘representative governance’ in a future financial mechanism and argues that it should be modelled on the structure of the Adaptation Fund Board.Footnote 7 This means that a majority of developing countries should make up the board of the proposed financial mechanism, together with ‘specific and significant representation’ for the ‘most vulnerable developing countries’.Footnote 8 According to CAN, “a developing country majority is more equitable than an equal distribution between AI and non-AI countries” since “there are 41 Annex-I countries and 151 non-Annex-I countries”.Footnote 9 A group of development NGOs is also more explicit in their proposal for a UN Climate Fund, whose Executive Board should operate on the basis of the “one country one vote rule and a majority representation for non-Annex 1 countries”.Footnote 10

The governance structure of the Adaptation Fund Board of the Kyoto Protocol is repeatedly referred to in the NGO documents as exemplary: it “scores a first in representative governance”, mainly because of the fact that it has majority representation from developing countries.Footnote 11 As such, it serves as a template for the design of other climate funds.Footnote 12 The demands for equitable representation and participation in climate policy-making are justified on both normative and efficiency-based grounds. The normative argument presents more representation for vulnerable countries (and communities) as a precondition for achieving global justice as it is these countries that will be suffering most from the negative impacts of climate change without having contributed substantially to the current levels of greenhouse gases in the atmosphere. NGOs also employ the rationale of effectiveness, arguing that institutions that are accepted as legitimate by developing countries are more likely to succeed in promoting a shift towards low-carbon development. These normative and practical considerations also seem to explain NGO support for any future financial mechanism to be based within the UN structure (under the authority of the conference of the parties) rather than in other existing or new international institutions.

The twin principles of country ownership and subsidiarity in policy-making and implementation are also crucial for enhancing opportunities for participation by developing countries and for maximizing the effectiveness of climate action and therefore enjoy wide support across the NGO community. They are intended to ensure that the most affected countries—as well as particular stakeholder groups within these countries—have the greatest possible influence over determining national priorities for adaptation and mitigation action and the allocation of financial resources. One way of operationalising these principles is the creation of a country-level coordinating body, which would be linked not only to the global climate convention process but also to various domestic stakeholder groups. The NGO proposal for so-called in-country coordinating mechanisms, for example, is modelled on similar mechanisms employed by the Global Fund to Fight Tuberculosis, AIDS and Malaria.Footnote 13 In the context of the climate convention, these are presented as a way of facilitating “a country-driven process, representing all relevant stakeholders, particularly the most vulnerable communities, ensuring a bottom-up approach to identify adaptation needs on local, sub-national and national levels”.Footnote 14 Similar to this is the proposal for the creation of ‘Multi-Stakeholder National Groups’ as national implementing partners for the climate convention.Footnote 15 These proposals bring together two modes of responses to the alleged representation and participation deficits in global climate governance outlined above. They are intended to address representation and participation failings at the intergovernmental level (by giving developing countries a greater say in how funding, particularly in the context of adaptation, is to be allocated and used nationally), but also firmly endorse the principle of stakeholder participation. This second aspect NGO support for the participation of particularly affected and especially vulnerable stakeholder groups at all levels of climate policy-making—will be discussed further below (Sect. 3.2).

Interestingly, despite expressing their support for the principle of subsidiarity in the climate convention context, the NGOs submissions analysed here contain hardly any explicit references to the role of sub-national entities or cities in the global climate regime. Local governments and municipal authorities present a separate constituency group and in their own submissions demand a more substantial formal role for cities and local authorities in the climate convention. Besides their broad calls for the participation of all relevant stakeholders at all levels of decision-making, the environmental and development NGOs do not appear to attach much priority to supporting these demands. This is noteworthy, given that “by 2030, two-thirds of humanity will live in urban centres where more than 73% of all energy is consumed today”.Footnote 16 Besides the fact that actions taken by cities and other sub-national actors already present an important contribution to global mitigation and adaptation efforts, these sub-national levels of governance present potentially valuable opportunities for facilitating the participation by locally affected communities in the politics of climate change.

As mentioned before, it is often difficult to distinguish between procedural and substantive equity and fairness demands in the NGO positions. The demands for more equitable representation by governments outlined here tend to go hand in hand with NGO support for issues that are considered especially important by developing countries, both as underlying moral principles and as desirable policy outcomes. For example, many NGO submissions to the climate convention emphasize the fact that the moral responsibility for climate change lies with industrialized countries and that these countries need to take the lead in cutting emissions. NGOs also support demands for the provision of financial resources and technology transfer to strengthen adaptation measures in those regions of the developing world that are particularly vulnerable to the consequences of climate change. Finally, as pointed out by analysts of NGO influence in climate negotiations, a number of NGOs have offered expertise and information services as direct support for some developing country delegations, as in the case of the close cooperation between the Foundation for International Environmental Law and Development and the Alliance of Small Island States during the negotiations for the Kyoto Protocol (Newell 2000, 143). In this way, NGOs are, to some extent, addressing the problem that capacity and resource shortages can be important reasons for less effective participation by smaller developing countries’ delegations in global negotiations.

3.2 Non-governmental organization responses to lack of societal access to climate policy-making

NGO support for the greater participation by affected stakeholders at the different levels of policymaking constitutes the second category of NGO responses identified here. In this sense, NGOs are pushing for broader societal access to decision-making (Koenig-Archibugi 2006, 14, Bäckstrand and Lövbrand 2006, 55), beyond, and in co-existence with, the channels for representation through national governments. In the case of climate change, NGOs highlight the impacts of climate change and specific climate change mitigation and adaptation policies on local communities in developing countries and espouse the general principle that those affected should also be given opportunities to participate in policy-making and implementation. They are particularly concerned with protecting the rights and enabling the participation of marginalized and/or especially vulnerable groups, such as indigenous peoples, women and youth. In expressing their support for the direct participation of these groups, NGOs are tapping into the ‘stakeholder participation’ discourse, which advocates, among other measures, the establishment of multi-stakeholder forums, consultation processes, stakeholder advisory boards and the creation of independent complaints mechanisms. It is possible to distinguish between two types of NGO proposals designed to support the participation of particularly affected and/or vulnerable groups: proposals for climate convention-linked domestic-level mechanisms that allow for participation by stakeholders in the local and national context, and proposals for institutionalizing participation at the international level directly.

National policy-making bodies such as the ‘in-country coordinating mechanisms’ discussed above are intended to address potential representation inequities at the international intergovernmental level by giving governments, especially those of recipient countries, more control over their national climate policy priorities. However, these mechanisms are also intended to enable the participation of a diverse range of societal stakeholders in the policy-making process. NGOs are particularly careful to point out that the participation of those who are especially affected or vulnerable, notably marginalized communities, indigenous peoples, women and youth should be prioritized. CAN, for example, argues that “National Adaptation Plans or Strategies should be prepared with the full involvement of civil society, vulnerable communities, and the private sector”Footnote 17 and that: “It is imperative that the most vulnerable people, who have contributed least to climate change but are most affected by it, are at the heart of decision-making about adaptation and risk management”.Footnote 18

A number of references indicate a preference by some NGOs for the development of mandatory standards for stakeholder participation, applicable in particular to indigenous peoples and local communities.Footnote 19 Such standards would be developed and agreed internationally (and in cooperation with the affected stakeholders) and implemented in the domestic context. The proposed in-country coordinating mechanisms, for instance, are supposed to “follow guidelines for adequate, active and meaningful stakeholder participation”, to be developed by the Executive Committee of the proposed Copenhagen Climate Facility.Footnote 20 The Guidelines developed by the World Commission on Dams (WCD) are repeatedly cited as an example of best practice with regard to stakeholder participation, which could serve as a template for similar guidelines in the context of climate policy.Footnote 21 Several of the NGO submissions express support for some level of monitoring and enforcement of such standards, especially in their positions on the REDD (Reducing Emissions from Deforestation and Forest Degradation) mechanism. This would include third party monitoring or the creation of an independent ombudsman to guarantee that countries follow social and governance-related standards, including those relating to the participation of indigenous peoples (Schroeder 2010) and local communities, in their REDD policies. Country performance may even be linked to the allocation of REDD funds.Footnote 22

While making the adoption of certain principles on stakeholder participation, a precondition for funding eligibility may, at first glance, appear to be an effective way of securing public participation in national climate policy-making, such proposals also risk alienating those governments who consider the imposition of social standards an interference in domestic political processes. This problem became apparent during the debate on formulating mandatory standards for participatory impact assessments and independent appeals mechanisms in CDM projects, whereby “Developing country governments opposed the imposition of all such protections, arguing that as sovereign states they alone would design CDM national processes” (Fogel 2004, 113). Another problem with relying on standardized participation mechanisms in the domestic context lies in the fact that some of the most affected communities are often marginalized from domestic political processes for a host of wider economic and socio-cultural reasons. Capacity building and the recognition of cultural, political and economic rights are hence important elements of democratic participation—a challenge also widely acknowledged by the NGOs.

Besides expressing support for multi-stakeholder processes in the national context, the NGOs also emphasize the need for participation by particularly affected groups, especially indigenous peoples, at the international level, namely in the climate convention directly. In order to achieve this, they are calling on parties to “create means by which indigenous peoples are directly represented in the climate negotiation process”.Footnote 23 The participation practices employed by the Convention on Biological Diversity and the UN Convention to Combat Desertification are repeatedly referred to as instances of good practice that the climate convention should also adopt. More specifically, indigenous peoples should have the “right to speak directly to texts under negotiation and to participate in contact groups and friends of the chair meetings where matters (like forests and related issues) may affect them”.Footnote 24 Parties should also be encouraged to include representatives of indigenous peoples and local communities in their official delegations.Footnote 25 The importance of allocating funding to support the participation of indigenous peoples is also acknowledged.Footnote 26

While most of the NGO submissions consistently link ‘indigenous peoples and local communities’, the identity of these non-indigenous communities remains unspecified. Moreover, the frequent references to the UN Permanent Forum on Indigenous Issues and the UN Declaration on the Rights of Indigenous Peoples as appropriate guidelines for developing participation norms and mechanisms within REDD also indicate that the primary concern lies with protecting the rights of indigenous peoples specifically (Schroeder 2010). It is, however, not unreasonable to suggest that REDD will also affect other local communities who do not identify themselves as ‘indigenous’ and whose interests may in some cases even conflict with those of indigenous peoples in the area. For example, one group, which does not appear to be engaged in the climate convention processes at all, are small-scale farmers, although climate change issues and agriculture are intimately related. The emphasis in the NGO submissions on the need to ensure the participation of indigenous peoples as opposed to other potentially affected communities suggests that NGOs are likely to be particularly responsive to the demands of stakeholder groups who have already succeeded in achieving a level of political mobilization and visibility.

4 Whose voice? Non-governmental organizations as representatives in the case of the climate convention

This article suggests that in addition to analysing the ‘external’ procedural demands made by NGOs vis-à-vis international organizations (in this case, the climate convention and associated instruments), it is instructive to turn to the NGOs themselves and to ask whether and how they can act as representatives of potentially marginalized communities in global politics. This section therefore adopts a somewhat different perspective to the previous one: it suggests that a third type of NGO responses to participation and representation deficits in international organizations may be internal to the NGOs and found in the representative and participatory practices employed by these private actors themselves (Biermann et al. 2009, 53).

According to Pitkin (1967), the activity of representation in a democratic context is characterized by an inescapable tension between ‘mandate’ and ‘independence’, or trusteeship (see also Grant and Keohane 2005). A representative has to be more than just a passive mouthpiece for the represented—he or she needs to be able to employ independent judgement. At the same time, however, a degree of responsiveness to the represented is a fundamental requirement of democratic representation—this determines the representative’s mandate to speak for his constituents. Seen from a democratic perspective, representative claims therefore need to be linked to some form of authorization and accountability to the represented; they are more than just ‘objective’ judgements of what outcomes are in the represented constituency’s best interest. Adopting this perspective of NGOs as ‘representatives’ requires us to shift our focus away from the level of interaction between NGOs and international institutions to the level of interaction between the NGOs and those constituencies they claim to represent. The challenge then becomes to identify ‘functional equivalents’ (Castiglione and Warren 2006, 15) to authorization and accountability in the relationship between the NGO in question and the represented constituency. Suggested examples of such forms of authorization by and accountability to those they claim to represent include membership democracy, network democracy, consultation processes with those represented, some types of self-regulation and ombudsman processes.

One key player among the NGOs engaged with the climate convention process is the Climate Action Network (CAN). CAN is “a coalition of more than 450 environmental and development non-governmental organizations worldwide, committed to limiting human-induced climate change to ecologically sustainable levels”.Footnote 27 CAN also acts as the designated ‘constituency focal point’ for the environmental and development NGOs with observer status at the climate convention. Its responsibilities in this role include acting as a conduit for the exchange of official information between the constituency members and the secretariat and coordinating observer interactions at sessions. Its prominent position among the ever-growing number of environmental NGOs with observer status and the long-term and persistent nature of its engagement with the convention make CAN a suitable case for analysis in the climate convention context. It is important to bear in mind, however, that many environmental and development NGOs engage with the climate convention independently of CAN and that the designated environmental NGO constituency is only one of several categories of civil society organizations working with the climate convention.Footnote 28

One of the most obvious—and generally least contested—constituencies represented by NGOs and NGO networks are their members (organizations and individuals) (Van Rooy 2004). In this case, the ‘mandate’ of the NGO to speak for its members may be based on the assumption that joining the organizations and/or providing financial contributions constitutes a form of authorization, or it may be based on more specific mechanisms for member control and influence inherent in the organization’s governance structure. In fact, besides the strategic benefits gained from sharing knowledge and expertise and coordinating campaign work, the legitimacy derived from their claim to represent a broad constituency of like-minded members probably constitutes one of the most important raisons d’être of transnational NGO coalitions and networks. This is certainly true for the Climate Action Network: “CAN enjoys a collective global membership of 20 million people” (ECO, issue 6, August 1994). This forms the basis of CAN’s claim to speak for a constituency beyond its own organization alone, to represent the public interest to a greater degree than other non-state actors (Newell 2000, 128).

In fact, CAN’s membership has expanded rapidly over recent years: in 2002, it consisted of 333 member organizations, while the website now refers to over 500 member organizations.Footnote 29 This considerable growth in membership can itself be seen as a form of authorization (Castiglione and Warren 2006), strengthening not only the legitimacy of the network but also its influence with governmental delegates (Newell 2000).

CAN relies largely on the structure of its network to support its claim to represent its members. CAN members are autonomous organizations, who, when they decide to join CAN, are asked to also become member of their national and regional nodes (in fact, they apply for membership to their regional node). ‘Nodes’ are associations of CAN members in a particular country or group of countries. National and regional nodes are expected to have their own governing bodies, “based on principles of accountability to its members and transparency”.Footnote 30 Strategic or policy decisions at the international level are taken by consensus, which has to include a large majority of national and regional nodes as well as most of the representatives of the international member NGOs with offices in more than 20 countries.Footnote 31 Of course, as is the case for most horizontal networks similar to CAN, resource, language and geographical differences among members are likely to constitute practical challenges to equal participation by members (Duwe 2001; Doherty 2006). Power differences are also evident in the fact that CAN Europe and US-CAN maintain permanent offices with full-time staff in Brussels and Washington, respectively, whereas most of the other regional coordinators are dividing their time between their own NGOs and the work for CAN (Duwe 2001, 179). Nonetheless, the network appears to be aware of these potential problems. The secretariat is charged with ensuring “sufficient voice from the developing world” when organizing global meetings and making the funding arrangementsFootnote 32 and has appointed a designated ‘Southern Capacity Programme Coordinator’.

While more in-depth research is needed, especially on the practical challenges to achieving equal participation in network decision-making (and the potential costs thereof), this rough sketch of CAN’s network structure allows us to make a number of preliminary points about how CAN’s claims to represent a large constituency of members could be rooted in forms of authorization and accountability. Firstly, the rapid increase in the number of organizations joining CAN over recent years can be seen as a form of authorization (member entry). This growth has enhanced CAN’s geographical spread and led to a greater number of national and regional nodes (especially in the South)—and as such has lent greater weight to the notion of the network speaking with a global voice. The relatively horizontal decision-making process used by the network means that decisions (relating to strategy, policy and the content of submissions made at the international level) enjoy the broad support of the individual member groups. Member organizations are able to comment and contribute to the process of policy formulation within CAN through the internal email exchange and in person during or just before the conferences of the parties to the climate convention (Duwe 2001, 180). Although CAN’s charter sets out very detailed processes to address and try to resolve conflicts between members, individual groups who find themselves consistently in conflict with the network’s decisions have the ultimate option to leave. On the other hand, the emphasis on reaching consensus constitutes a relatively strong form of authorization—any submissions made by CAN at the international level are likely to enjoy a strong backing by its members and are not just endorsed by a narrow majority.

The possibility of NGO representation of non-members is also worth considering, although such representative claims are more likely to be publicly contested. However, if the relationship between these groups and the NGO contains genuine elements of authorization and accountability, it may be possible to justify the claim that the NGO has a mandate to represent these groups on particular issues. While many NGOs in the environmental and development sectors pursue participatory practices in their operations on the ground, they are, however, less likely to include non-members in the formulation of their global policies and strategies. For network organizations such as CAN, the distinction between members and non-members and the difference between internal and external (Keohane 2003) lines of accountability may in practice be less of a problem than sometimes assumed in the literature, provided that membership is open to politically mobilized community groups.

5 Concluding remarks and further questions

The analysis presented above allows us to draw some preliminary conclusions, point to a number of potential problems and highlight areas for further research on the potential contributions of civil society actors to strengthening democratic legitimacy in earth system governance. The article investigated two types of proposals by environmental NGO intended to contribute to more representative and participatory decision-making processes in the climate convention context, together with a third potential mechanism, internal to the NGOs: the notion that NGOs themselves are representing local constituencies in global politics.

The first type of ‘external’ NGO interventions discussed relates specifically to the inter-governmental nature of the decision-making processes. By highlighting the need for formal representation structures that give sufficient voice to particularly affected and vulnerable countries, by drawing attention to issues that are of particular concern to these countries and by providing practical support and expertise to certain delegations, the NGOs are to some extent seeking to address obstacles to the effective representation by governments of their citizens in international forums. Intergovernmental channels for representation are also strengthened through the principles of country ownership and subsidiarity. Of course, the solutions put forward by NGOs in this respect do not address potential democratic shortcomings or “blockages” (Keck and Sikkink 1998) at the domestic level, which may work against an effective upward transmission of citizens’ (as ultimate principals) preferences to the international level. Strengthening the voice of undemocratic governments in international organizations does not necessarily pave the way for better representation of their citizens.

Why are environmental NGOs concerned with these procedural issues at the intergovernmental level? At least part of the explanation lies in their assumption that these procedural elements are also likely to determine the effectiveness of a global climate regime. As pointed out above, NGOs argue that strengthening the perceived legitimacy of global climate policies will make it more likely that these are implemented effectively at the national level. Furthermore, it may be that the NGOs are hoping to use these procedural demands to indirectly raise the level of ambition in the climate negotiations. By repeatedly calling for better representation for the most vulnerable countries, they are effectively supporting those parties with the greatest stake in an ambitious agreement.

Another important factor is the growth in number and influence of Southern groups among the environmental and development NGO working at the global level. Southern environmental groups have always been concerned with drawing more explicit connections between environmental issues and broader economic and political factors (Chatterjee and Finger 1994). This is well illustrated in the case of CAN, whose growing constituency of Southern member groups has pushed for more emphasis on the equity aspects of climate politics in CAN’s policy statements (Duwe 2001). While demands for compensating developing countries for the harmful impacts of climate change or the recognition of historical emissions are different from demands for ‘fair’ representation by countries, the latter are important for establishing legitimate channels through which specific rights and obligations may be claimed or contested. The fact that CAN has put greater emphasis on equity issues over recent years is seen critically by some members who argue that the focus on financial transfers to developing countries diverts attention from the pressing need to reduce global emissions as rapidly as possible.Footnote 33

Calls for the participation of affected stakeholders at all relevant levels of decision-making constitute another important element of the NGO submissions to the climate convention. Many of the demands for introducing requirements for ‘stakeholder participation’ apply to the national and local level. At this point, of course, a global policy will already have been developed and participation is essentially limited to the implementation and follow-up phases. Of more interest, therefore, are the demands for direct participation by affected communities in international institutions, for example, through advisory bodies containing community representatives. Of course, it can be argued that NGO is using demands for more and better participation by civil society representatives in international decision-making processes to extend and consolidate their own opportunities for influence. This outcome, however, is not guaranteed. Once intergovernmental organizations are starting to reach out directly to people’s organizations, community organizations or ‘major groups’, NGOs become just one constituency group among many and risk losing privileged access to these organizations. While this is an unlikely development in the case of the climate convention, McKeon’s analysis of the FAO’s efforts to strengthen the participation of grassroots organizations such as Via Campesina shows that many NGOs were left feeling marginalized and alienated (McKeon 2009, 62).

This analysis finds that there is a tendency towards convergence around support for particular procedural norms relating to participation and representation among the NGOs engaging with the climate convention. While some of the NGOs put more emphasis on these procedural demands and mention them more frequently in their submissions than other groups, there is a broad consensus within the NGO community in support of equitable governance structures within international organizations, better representation for particularly vulnerable countries, participatory decision-making processes ranging from the local to the global level, and mechanisms designed to encourage the participation of marginalized groups in particular. This finding would support the assumption that NGOs engaging with international organizations are indeed contributing to the democratization of these institutions and to making them more accountable to the communities affected by their decisions, though not by acting as ‘transmission belts’. There are, however, at least two important qualifications to this argument: firstly, while NGOs are including these procedural demands in their lobbying activities directed at policy-makers, the extent to which they are also able to bring about change of this type in international organizations is questionable. As shown by a number of studies on NGO influence in the climate negotiations, their impact on concrete policy outcomes remains limited (Arts 1998; Newell 2000; Betsill 2008). If anything, it could be that they are contributing to a broader process of delegitimization of unrepresentative and non-participatory governance mechanisms. Secondly, a broad consensus around particular procedural demands does not preclude the possibility of significant divergence when it comes to substantive proposals. An international NGO may support the right of a local community to have a say in the policy-making processes that affect this community but may still fundamentally disagree with the substantive demands put forward by the community. Similarly, different NGOs may espouse the same procedural demands but vary fundamentally in their broader assessment of what constitutes an appropriate response to the climate crisis. It is noteworthy, for example, that the very similar procedural standards on participation and representation are supported by both the Worldwide Fund for Nature and Friends of the Earth International, although both groups differ markedly in their ideological outlook and acceptance of market mechanisms (which the Worldwide Fund for Nature tends to endorse and Friends of the Earth International opposes).

The NGOs themselves remain a ‘black box’ in many ways, and the extent to which they rely internally on functional equivalents to authorization and accountability (Castiglione and Warren 2006, 15) to members or to other communities on the ground needs further analysis. The example of CAN was used in this article to explore these questions. In this case, the network may be justified in claiming to have a ‘mandate’ to represent its members, although this leaves open the question of whether they are also able to justify potential claims to speak for other, non-member constituencies. One of the specific research questions on accountability set out in the Earth System Governance Project relates to the sources of accountability and legitimacy in earth system governance. While this article has dealt with equitable representation and participation as important sources of legitimacy, these are not the only ones. Relating to NGOs, for instance, it would be wrong to apply the same standards of representativeness and participation across the board—many organizations derive at least part of their legitimacy from non-democratic but equally important qualities, such as the ability to supply technical expertise and knowledge. Even in the case of those organizations that act more explicitly as representatives for particular constituencies, we are likely to come across important differences in how these representative claims are justified. The issue of variation among NGOs and the implications of different practices for the potential of NGOs to assume representative functions in global politics warrant further empirical work.

An interesting challenge to the idea of NGOs acting as a link between the local and the global lies in a potential conflict between the two levels. What if the ‘local’ is actively resisting the ‘global’ (Fogel 2004)? The case of Friends of the Earth International is an interesting example of a network organization that embarked on a conscious strategy of basing all its work (including its positions on global issues) on its commitment to the grassroots (Doherty 2006). One interesting consequence of this internal process of prioritising its relationship to the grassroots has been a shift of focus away from international institutions. In becoming more participatory and committed to local voices, Friends of the Earth International has also become more radical in its opposition to dominant institutions and seems to be distancing itself from pursuing direct engagement with state elites, market players and many international organizations. The website states, “Our participation in international fora will thus become secondary to our resistance work”.Footnote 34 This stance contrasts with earlier depictions of Friends of the Earth International that locate the organization’s activities at the “intersection between national and international processes” (Wapner 1996). The case of Friends of the Earth International therefore raises many new fascinating questions. Rather than acting as the frequently portrayed link between the local and the global, are NGOs in fact forced to choose between the two levels? Does the concept of NGO as ‘intermediaries’ between local communities and international organizations only work if the interests, preferences and values of these local communities do not fundamentally challenge the dominant paradigms which shape the policy solution that are under consideration in the international organizations?

The Science and Implementation Plan of the Earth System Governance Project also points to the importance of understanding potential trade-offs between requirements of effectiveness and high standards of accountability and legitimacy in earth system governance (Biermann et al. 2009, 56). A few preliminary observations based on this study strongly support the call for further research on this problem. As NGOs and NGO networks try to adopt more participatory decision-making processes, they face the very practical challenges of balancing inclusiveness with the ability to respond quickly and effectively.Footnote 35 A more profound challenge, however, lies in a possible trade-off between ‘mandate’ and ‘trustee’ based forms of representation. Attempts to become more representative of living constituencies do not necessarily make NGOs more suited to represent future generations, a role which has also been used to justify demands for more civil society engagement in international organizations. To the extent that the development of a mandate depends on authorization by and accountability to living human beings (‘capable of independent action and judgement’, see Pitkin 1967, 209), no NGO can enjoy a democratic mandate to represent future generations. In fact, if NGOs are to represent this ‘constituency’, it is perhaps necessary to accept their role of ‘trustees’, which proves problematic from the viewpoint of democratic legitimacy. A controversial argument may be that a possible trend towards greater accountability by NGOs to living stakeholders could also compromise their ability to act as ‘trustees’ for future generations (a role which could involve supporting actions that are not favoured by living stakeholders).

The problem of potential trade-offs between inclusiveness and effectiveness is not limited to the NGOs but may be similarly applicable to the global rule-making processes they are targeting. A number of commentators have suggested that the UN’s principles of unanimity and inclusiveness contributed to the failure of Copenhagen.Footnote 36 This raises larger questions about possible contradictions between the input and output legitimacy (Scharpf 1998, 1999) of a given rule-making process in earth system governance. It also poses more immediate challenges for the NGOs and their commitment to the UN process. It remains to be seen to what extent the experience of Copenhagen will result in a strategic and normative repositioning within the NGO community.