EGP 1654-4951 1654-6369 Ethics & Global Politics Vol. 3, No. 3, September 2010, pp. 217–238 sici xxx PUBMED Abbreviation ChristopherL.Pallas

*Correspondence to: Christopher L. Pallas, London School of EconomicsLondonUK

London School of Economics

LondonUK

New School for Social Research

New YorkUSA

Ethics & Global Politics Developing a context-based standard of democratic legitimacy for CSOs The literature on transnational civil society encompasses a number of conflicting views regarding civil society organizations’ (CSOs) behavior and impacts and the desirability of civil society involvement in international policymaking. This piece suggests that this lack of consensus arises from the diverse range of contexts in which CSOs operate and the wide variety of activities in which they engage. This article seeks to organize and analyze the disparate data on civil society by developing a context-based standard of democratic legitimacy for CSOs. The article disaggregates democracy into input, throughput, and output components, and shows how CSOs must support or manifest different aspects of democracy in order to be democratically legitimate in a given context. Applying this standard to existing works, the article identifies several problems in current research, including a failure to recognize ways the democratic imperatives of transnational advocacy differ from national advocacy, and the potential for international civil society interventions to undermine local democratic processes. transnational civil society democracy global governance NGOs international policymaking As civil society organizations (CSOs) have become a prominent feature of international relations, a number of problems have arisen regarding their study. Early research and analytical theory on civil society in international relations was dominated by a desire to demonstrate the potential for an alternative to the state-based study of international affairs. Thus, the early literature, particularly from the constructivist school, focused on demonstrating civil society's impact. Now that that impact has been proven, new questions have arisen regarding civil society's behavior and its contributions to democracy. Although civil society is frequently heralded as a key component of more democratic global governance, a noticeable gap has been observed between the hoped-for impacts of civil society as a normative construct and the real-world impacts of the CSOs of which it is composed. A number of researchers have called into question CSOs’ motivations, representivity, and democratic credentials.1 Even among those authors that hold that CSOs can contribute positively toward global governance, one finds a variety of competing and sometimes contradictory prescriptions for judging these organizations’ legitimacy.This article addresses debates over the legitimacy of civil society actors by offering a democratically based standard of civil society legitimacy. It argues that civil society's legitimacy is based not on wholly endogenous factors like transparency or participation, but rather on civil society's effects on the democratic rights of the populations it impacts. The article clarifies the conditions under which CSOs can contribute to creating or enhancing democracy by identifying the democratic needs of different political contexts and highlighting the role CSOs play in contributing to democracy in each one. This article also highlights the dangers of confusing or conflating the role of CSOs in national and international spaces, a problem that appears to occur frequently as CSOs expand their activities in the global realm.This article divides democratic legitimacy into three parts: input (such as grassroots participation), throughput (such as transparency), and output (mainly impacts). As will be shown, the relative importance of each element varies in different political contexts. CSOs’ legitimacy is judged based on their contributions to the aspects of democracy most necessary in their political context.To help highlight the impacts of context and the ways in which civil society can respond to it in a democratically legitimate way, I have used the language of ‘roles.’ Within the national context, a CSO (or a coordinated group of CSOs, like an advocacy network or campaign) can act as a Revolutionary seeking to reform or replace an undemocratic regime and install a democratic one. It can also act as an Advocate representing the interests of particular groups within a democratic system. In either the state or international context, a CSO can act as an Agent working on behalf of the state or international institutions. Finally, in the global context, a CSO may also act as an Authority contributing actively to the creation and enforcement of global norms and policy.This article proceeds in five parts. I begin by reviewing the literature on transnational civil society (TCS) organizations and identify questions in need of clarification. Next, I disaggregate democratic legitimacy and discuss its various elements. Third, I identify the four contexts in which CSOs may act and elaborate the rationale for judging democratic legitimacy differently in each context. Fourth, using insights from this analysis of contextual legitimacy, I highlight two important problems in the current analysis of the legitimacy of transnational non-governmental actors. Finally, I discuss ways that a contextual understanding of legitimacy can contribute to the remaining questions in the literature. Among academic authors, civil society is discussed as both a theoretical object and in empirical terms as the aggregate of some delimited set of CSOs. Studies examining civil society in empirical terms also often narrow their focus to a particular type of CSO or network or to a specific campaign. This article engages with civil society as a real-world phenomenon and proposes legitimacy standards for individual CSOs and for collective groups thereof, including campaigns and networks. Nonetheless, this article identifies and explicates tensions across the spectrum of the civil society literature, from those works dealing with civil society in wholly theoretical terms to those working wholly in empirical terms.Academic writing on civil society involvement in international affairs dates back over 30 years. The literature on international civil society, however, did not reach critical mass until the early 1990s, when a more regular dialogue began on the role of TCS in global affairs. Whereas earlier efforts were largely empirical, these new writings were prompted by a variety of theoretical, normative, and empirical interests. Since then, shifting foci within the field and a multiplicity of approaches have led to a fractured and at times meandering body of literature.Some of the first contemporary references to TCS appear in articles published by Martin Shaw and Ronald Lipschutz in 1992. Lipschutz wrote that ‘global civil society’ was creating a form of transnational demos that would ‘challenge, from below, the nation-state system.’2 He embraced an explicitly normative agenda, calling for academics to ‘undertake the reconstruction of world politics’ to facilitate civil society's growing role.3 Shaw took a slightly more cautious approach. He agreed that the growing power of civil society ‘challenged the principles of sovereignty and non-intervention’ at the heart of the state system.4 However, he did not think the ‘global society perspective’ would ‘become central to world politics in the short- or medium-term.’5 These articles were followed by a host of others, many of them arguing for the potential of civil society to revolutionize global governance. Some argued for civil society's democratizing potential6 whereas others simply emphasized its power and influence.7Most of this writing reflected the effort, often led by constructivists, to break free from a state-based, realist depiction of international relations.8 Constructivists and others contested the dominant materialist perspectives in international relations which depicted state power as the predominant explanation for international events and decisions. The influence of civil society and the existence of transnational networks were important proofs that states were no longer the sole legitimate focus of study. Civil society's power to create norms and influence policy indicated a locus of power outside the state and a means of power other than material dominance. The creation and adoption of international norms also indicated that, contrary to rational choice theory, states’ interests could shift over time.9 Similarly, the idea of globe-spanning citizens’ networks provided an alternative to the vision of international anarchy and isolated states favored by realists and rational choice theory.The focus on theory-building in the mid-1990s sometimes eclipsed empirical research. International relations writings about civil society often seemed to rely on media depictions of current events rather than carefully investigated case studies. At the same time, the case for civil society's democratizing potential was more inferred than proven, sometimes from the precedent of the civil-society driven democratization of South American and Eastern European nations.10 The constructivist focus on norms and impacts left civil society advocates open to the charge of ignoring questions of agency by not specifying clearly the means by which civil society achieved its influence.11 Other critics challenged the relevance of national experiences to the global context.12 A number of authors also suggested that early writing on civil society ignored the ways in which the complexities of global advocacy might inhibit genuinely democratic representation. They noted that effective global advocacy relied on coercive power available only to a minority of non-governmental organizations (NGOs) and that the political bargaining in which powerful NGOs engaged was neither transparent nor accountable to many of the people it affected.13 This new body of more critical writing marked a shift away from the discussion of civil society primarily as a theoretical construct and toward a more empirical analysis of the observed behavior and impacts of CSOs.This wave of critique resulted in a number of strong, interdisciplinary works that combined constructivism with rational choice analysis. They focused on CSOs’ ability to upend the state system by introducing new norms while still relying on some state mechanisms to implement and enforce standards.14 Chief among these was Keck and Sikkink's Activists Beyond Borders which laid out the ‘boomerang theory’ of transnational advocacy, depicting how weak advocates in developing nations might enlist the aid of partners in powerful states who would exercise political leverage on behalf of their developing country partners.15 Fox and Brown's The Struggle for Accountability made an empirical assessment of CSOs’ impacts on the World Bank, using case studies authored primarily by practitioners.16 Florini's The Third Force built on both of these approaches. It stressed civil society's increasing importance in global governance while emphasizing the tendency of CSOs to pursue ‘their [own] conceptions of what constitutes the public good.’17These new works helped address some problems of theory and method but they did little to tackle growing concerns about the legitimacy of civil society. Questions remained about CSOs’ democratic credentials, particularly their accountability and transparency.18 Other authors raised queries about whether CSOs could function democratically outside the boundaries of the state.19 Still other works challenged constructivist assumptions about CSOs’ autonomy from the state or material interests.20One persistent concern was that civil society, particularly at the global level, was fundamentally dominated by elites. Research from both development studies21 and political science22 highlighted CSOs’ use of elite mechanisms. Some argued that civil society replicated and magnified the power imbalances of the old state system rather than remedying them.23 Stone suggested that the complexity of global networks concentrated their benefits among actors with the ‘resources, patronage or expertise’ to participate in transnational discussions.24 These concerns contribute to the continuing debate among academics and practitioners on the appropriate role for CSOs in global governance. Numerous rifts exist within this body of literature. These rifts center on three main issues: the understanding of civil society, the model of global governance, and the definition of democracy. Further complicating factors are the divisions between normative and empirical approaches, and among various schools of academic research. When seeking to address the democratic legitimacy of CSOs, the challenge is two-fold. First, most of the rifts are multilateral; none of the debates around civil society, global governance, or democracy can be neatly partitioned into two sides. Second, most authors writing about civil society manage to cross multiple fault-lines as they write simultaneously about civil society, global governance, and democracy.The most obvious debate centers on CSOs’ behavior and motivation. Some authors maintain that CSOs genuinely ‘rescue the causes of marginalized or excluded groups.’25 Others insist that CSOs pursue their own understanding of the public good.26 Some focus on the diversity of organizations, reminding us that not all CSOs are truly ‘civil’—i.e. non-violent or interested in upholding the common good.27 The skeptics insist that CSOs are highly parochial, with the tendency to promote rich-world policies for Southern or poor populations.28 The most pessimistic of all insist that CSOs are materially driven and self-interested29 or even assist in an imperial agenda.30A second debate is over the shape of global governance. Authors of the cosmopolitan school argue for the eventual dissolution of national governments or predict the rise of a global superstate.31 Others argue for the enduring power and importance of states and institutions, including a role for states or international organizations in implementing CSO agendas.32 Again, a critical minority questions whether democratic global governance is even possible33 and whether CSOs are just tools of the state.34The definition of democracy forms another debate. Bexell, Tallberg, and Uhlin have observed that while normative democratic theory manifests a ‘trichotomy’ of separate representative, participatory, and deliberative models, writers on global governance feel free to sample from and combine these strands.35 Held's vision of cosmopolitan democracy, for instance, mixes elements of all three models.36 Nanz and Steffek take a more purely deliberative approach.37 An emphasis on participation is common among advocates for CSO participation in global governance,38 while others use accountability as a proxy for equal representation.39Over this fractured ground are laid other complicating factors. Political theorists have created elaborate normative models, while empirical researchers have taken issue with the gap between civil society's idealized behavior and the reality of CSO conduct. At the same time, variations in approach among international relations, international political economy, development studies, and non-academic practitioners further complicate the literature. This article addresses these debates in the literature by offering a democratic standard of legitimacy for CSOs that defines the ways in which CSOs can contribute to advancing democracy in different contexts. Before elaborating on this standard, however, it is necessary circumscribe the CSOs to which it applies and to provide an underlying definition of democracy. A CSO, in principle, can refer to any association without an explicit role in government or explicit profit-making purpose. Robert Putnam famously includes bowling leagues and bridge clubs in his discussion of civil society. What most of the organizations and movements whose activities are addressed by this article have in common, however, is an impact on the creation or use of policy. Focusing on policy may exclude consideration of the associational value of some types of organizations, but it does provide a lens for exploring the importance of civil society in international relations and it succeeds in encompassing the vast majority of the literature. Moreover, it is able to incorporate both formal and informal organizations.For the purposes of this piece then, I will define a CSO as any formal or informal association of individuals which is involved in the creation, reform, or implementation of policies and norms, provided that the association is neither primarily a part of government or governance institution nor of a profit-making enterprise. This definition thus includes professional NGOs, social movements, trade unions, and foundations. It makes no distinction between ‘civil’ and ‘uncivil’ groups, since this article examines the democratic legitimacy of all groups. It excludes government departments and corporations, but recognizes that some third-sector organizations will engage and even act on behalf of the political and commercial sectors. The definition includes all non-governmental and non-commercial groups acting to impact or implement policy and norms. It thus includes both advocates and service providers.An increasingly prominent term in the literature is transnational or ‘global’ civil society. TCS refers to those civil society actors who engage in activity beyond the borders of their own states or whose domestic activities are linked to non-governmental and non-commercial actors beyond their own states. This article pays attention to such actors, including international NGOs headquartered in one country yet working in another; local civil society actors receiving international funding; global justice movements; and any CSO which is connected to a regional or international network. In choosing a definition of democracy, my major concerns are applicability and fairness. The definition must apply equally well to states, international institutions, and CSOs. Much of the literature on civil society and democratization is implicitly (or explicitly) a comparison of the relative democratic credentials of these three categories of actors. To fairly judge between them, the definition must apply to all three. Likewise, the definition must engage with the existing literature on civil society, and be applicable to CSOs’ stated objectives, capacities, and limitations.State-based democracy utilizes majority rule, in which some equally distributed measure of voice or authority is used by citizens to exercise control over the government.40 This understanding of democracy is not only embraced by state actors, but also a large number of practitioners, particularly those from the developing world.41 Eliminating the strict emphasis on voting can allow a majoritarian definition of democracy to accommodate the pluralistic or participatory standards typically used to test democracy in non-state settings and creates space for the evolving representative mechanisms of international institutions.Granted, Habermasian depictions of deliberative democracy are non-majoritarian. However, deliberative democracy requires open access in order to function. Only when all ideas can be brought to the debate is it guaranteed that discussion will reveal the ultimate good. There is abundant evidence that international policy discussions are not universally accessible, and that discussions among CSOs may be dominated by elites.42 Were it currently applied to CSOs, it seems unlikely that any organization, network, or campaign would meet the standard of universally accessible deliberations.In addition to being majoritarian, democracy, as defined in this article, is representative. Representative practices are nearly universal in modern democratic states. As Dahl notes, ‘in practice, all democratic systems, with the exception of a few very tiny communities, allow for, indeed depend on, delegation of power and authority; the citizen body delegates some decisions to others.’43 These designated persons are commissioned to represent or act on behalf of a particular population.44 Moreover, as noted, data suggests that much of transnational non-governmental civil society advocacy is elite-driven. Thus, the resulting test is essentially two-fold. On the one hand, it asks whether elite CSOs are representing the people they claim to represent. On the other, it asks whether the influence of different elite actors can be balanced between them, such that their aggregate impact still reflects the will of the majority of stakeholders.Finally, democracy includes the protection of citizen rights.45 Rights may be protected by either norms or laws allowing this definition to engage with constructivist, functionalist, and realist literature. Of course, some of the ‘rights’ protected by civil society groups are quite controversial, so a specific list of the rights encompassed would be useful.46 However, as Held has pointed out, the understanding of liberalism (and its constituent rights) has shifted historically.47 For the purposes of this piece, it is more important to establish that democracy involves both popular sovereignty and the protection of rights than it is to enumerate those rights in great detail.In summary, this article defines democracy as a system of equal citizen authority or value expressed via some representative mechanism and resulting in government or institutional responsiveness to the will of the majority, but under which the government or institution is also constrained to protect the liberal rights of its citizens or stakeholders. The definition does not require that all citizens make use of their voice or actively participate, only that the mechanism of input (e.g. voting or otherwise) be equally accessible and provide for all participating voices to be equally valued. It also requires that the governance organization respond to the expressed will of the majority and protect commonly recognized rights. The democratic legitimacy of CSOs is judged on the basis of their contributions to the democratic well-being of the persons impacted by their actions. That is, the legitimacy of individual CSOs, and of collective endeavors such as campaigns and networks, are judged based on whether they contribute to a system of equal citizen authority, majority rule, and government responsiveness or the protection of basic rights. Applying this, however, requires dividing democratic legitimacy into several parts. In his work on ‘The Democratic Legitimacy of Transnational Actors,’ Uhlin writes: In order to organize the various concepts related to democratic legitimacy, I find it useful to distinguish between input legitimacy (the relationship between the actor and its constituencies or people affected by its activities), throughput legitimacy (the actual procedures for decision-making within the actor), and output legitimacy (the consequences of the actor's decisions and other activities).48The concepts of input and output derive from Scharpf.49 Uhlin adapts these standards to transnational actors and borrows from Dingwerth the useful category of throughput legitimacy.50Each element of democratic legitimacy prompts distinct questions. Input legitimacy focuses on issues of representation and inclusion. It examines whether a state or non-governmental actor is representative of its constituents or stakeholders, whether stakeholders have equal voice in formulating policy positions, and, particularly in the case of advocacy organizations, to what extent they advance the interests of those populations they claim to represent. Throughput legitimacy examines transparency, accountability, participation, and deliberation. It asks how actors promote participation and discussion, whether they are transparent, and how and to whom they are accountable. Output legitimacy focuses on the consequences of actors’ activities. It includes both the impacts of a successfully implemented policy and the ways in which activism can change the political system.51Naturally, some of these elements of input, throughput, and output, and the specific questions they prompt, resonate more strongly with some definitions of democracy than with others. They also vary by context. Uhlin suggests that ‘forms of democratic legitimacy differ … between social, cultural, and political settings.’ However, the focus of his work is on varieties of actors rather than varieties of contexts so he does not elaborate on this point. Furthermore, he elects not to operationalize his framework to make a critical evaluation of the literature. This article builds on his work by taking both of these steps. The implication of Uhlin's argument is that many of the apparent disagreements over the meaning of democracy in the civil society literature are implicitly debates over the type of legitimacy most relevant to CSOs. His argument highlights the extent to which many authors fail to explicitly define democracy in their work and also fail to clearly associate desired behaviors or outcomes, including much-vaunted standards like accountability or participation, with a specific democratic theory. This article addresses these issues by developing a framework for contextual legitimacy that reveals how some seeming tensions in the literature actually result from an effort to apply the same standard of democratic legitimacy in disparate contexts.The literature on civil society and democracy tends to describe CSOs as though they operate in a single, global context. Some authors distinguish between domestic and international civil society, but organizations from these two contexts are frequently treated as equal parts of transnational networks. Although potential disparities between national and international actors are acknowledged,52 their combined efforts are frequently treated as monolithic campaigns.53 Moreover, transnational campaigns to change the national policies or practices of a single country (e.g. by stopping a dam or freeing imprisoned journalists) are treated as equivalent to transnational efforts to create new global policies (e.g. banning landmines or improving financial regulation). Finally, as noted earlier, cases from specific national contexts, like the civil society-driven democratization of countries in Eastern Europe or Latin America, are used as models for the democratization of global governance.I would argue that, with regards to its democratic legitimacy, CSOs actually operate in three different contexts. First, they operate in undemocratic states. In recent history this would include places like Eastern Europe or apartheid South Africa. Second, they operate in democratic states, i.e. states that have some measure of liberal democracy and that are acknowledged as democratic by their peers. Historically this includes the USA and Western Europe and, more recently, much of Latin America, Southeast Asia, and parts of Africa. Finally, CSOs can operate in the international realm, where they are often beyond the control of any one state or institution. This is the newest of civil society contexts but arguably the most powerful. It includes CSO lobbying of the UN, World Bank, or WTO, the development or implementation of aid programs, and transnational campaigns and advocacy networks working to construct new international norms or regimes. Each of these three contexts includes substantial internal variation. For example, the context of democratic states includes states with varying forms (and, some would say, ‘quality’) of democracy. For the sake of parsimony, this article does not model the variation within these three contexts or detail methods for evaluating the contexts themselves (e.g. by specifying the threshold at which a state becomes democratic). I would argue that these are second order concerns compared to the question of how different contexts create different standards for CSOs’ democratic legitimacy, which much of the civil society literature to date has ignored.The different contexts should inform and shape our understanding of legitimacy, with the multiplicity of contexts leading to a diversity of standards for democratic legitimacy. If the democratic legitimacy of civil society is judged by its success in developing or facilitating democracy, the goal of any organization wishing to be democratically legitimate must be to enhance the democratic rights of the entire population it impacts. This requires gauging the interaction between the CSO(s) and other structures (namely governments or institutions) which could or should grant enduring democratic rights and protections. These governments and institutions exist, they have impact, and, in many cases, they have better established democratic credentials than competing civil society actors. If CSOs claim to enhance the democratic well-being of their stakeholders, then the roles and impacts of governments and institutions must be taken into account. Insofar as the legitimacy and authority of these actors vary by context, so too will CSOs’ interactions with them.This piece uses the language of ‘roles’ to describe the requirements of a given context. Each role defines the means by which civil society addresses the democratic needs of a particular setting, as determined by the presence or absence of other democratic structures. Therefore role, as used here, cannot be divorced from context.The needs of each context create specific democratic legitimacy requirements for CSOs operating therein. Meeting these requirements leads to particular types of behavior. Roles encompass both the standards of democratic legitimacy and the resulting behavior. However, it is important to emphasize that these roles are used in an analytical sense, to assess democratic legitimacy, not as abstract descriptions of possible CSO activities. The roles are simply short-hand for the requirements of context. Thus, this piece does not label a CSO as a ‘revolutionary’ simply because it is working to change the system or as an ‘advocate’ because it claims to be representing a certain group or interest. Civil society actors which do not meet the democratic needs of a given context are not said, on the basis of their behavior, to be fulfilling an alternative role. Instead, I would describe them as failing to meet the democratic legitimacy standards of the current context. To reiterate, for CSOs to be democratically legitimate in any given situation, they must interact with other structures in a way that develops the democratic rights of its stakeholders. Each context presents one or two possible behaviors and a set of standards by which such behaviors may be judged. The behaviors and standards are summarized in the four roles. For the sake of clarity, each of the roles is described as it applies to a single CSO. However, the roles are equally applicable to multiple CSOs operating in a given context, including coordinated groups such as campaigns and networks.In an undemocratic state, a CSO must be legitimated by its efforts to reform or replace the existing regime. In this context, it must play the role of revolutionary. It is important to note that efforts to create or enforce rights are commonly described as advocacy activities. In this context, however, such ‘advocacy’ is actually a form of revolution, insofar as it changes the system of government to make it (by the standards of this work) more democratic via its recognition of rights. Non-revolutionary activities, such as educating children or providing healthcare may certainly be judged legitimate by any number of moral or technical standards, but they do not provide democratic legitimacy. Providing services or advocating on behalf of specific interests can have little enduring effect on citizens’ control over government without a wholesale change in the means of governing. Service provision may prop up an undemocratic state, even as it mitigates its impact on its citizens. Advocacy which does not push for political reform may likewise legitimate the state. Even where such CSO advocacy wins concessions, it is reliant on an undemocratic regime to maintain them. Instead, a CSO must promote change in government, including the development of representative mechanisms and the recognition of basic rights.In this context, a CSO's legitimacy should be judged purely on outputs. On the one hand, representation, participation, or transparency are meaningless if national democracy is not established. CSOs themselves might be internally democratic, but internal democracy will not succeed in obtaining democratic rights or protections for the country's citizens. On the other hand, if liberal democracy is established, new legal or constitutional standards will be developed to govern representation and protect rights. Thus, a CSO in this context does not necessarily need to manifest these standards itself. Civil society is not the government; it is the means to establishing (or reforming) the government. Revolutionary CSOs, like Solidarity in Poland, are judged to be democratically legitimate when their efforts succeed in establishing a democratic state. Conversely, revolutionary organizations that act on hierarchical or otherwise undemocratic lines may be looked upon skeptically if they attempt to govern national affairs in such a fashion after national democracy has been established (e.g. ZANU in Zimbabwe).When acting within a democratic state, a CSO may play the role of advocate. Democratic rights are guaranteed by the state, and a CSO can enhance the democratic rights of a state's citizens by monitoring or facilitating state processes. It does this either by seeking to represent marginalized populations, ensuring that they are fully empowered within the political process, or by acting as a watchdog, ensuring that the government continues to function democratically and protect citizen's rights. In this role and context, a CSO is judged on either input or throughput. When acting as a representative, a CSO must be judged on both input and throughput. If a CSO claims to speak on behalf of a given population, then its claims must be verifiable. This requires both representation and a measure of transparency and accountability. Without these things, a CSO risks tipping the scales in favor of special or even imaginary interests or co-opting the causes of marginalized populations to achieve ends other than those desired by those populations. When acting as a watchdog, a CSO must be judged based on its throughput. A CSO can and should support the practices of transparency, accountability, and deliberation which enhance democracy, but in order to legitimately enhance them, it must also model them, creating a standard for the behavior citizens should expect from their government. A CSO operating within the democratic state context is not judged on outputs. Democratic representation is already provided by the state and, in a majoritarian regime, sometimes a CSO should lose, i.e. if it is representing an interest at odds with the will of the majority. The exception, of course, is when a CSO is attempting to enforce and protect the recognized rights of a particular minority. However, even in this case a CSO's legitimacy is not judged by its outputs, because a CSO ultimately has no control over the state. When a good-faith effort (input and throughput legitimate) to protect minority rights fails, it reflects negatively on the democratic credentials of the state, but not on those of the CSO involved.In either the democratic national context or in the international arena, a CSO may act on behalf of a state or institution. When acting on behalf of an established authority, a CSO operates in the role of agent. Historically speaking, the agent role is a result of the neoliberal shift and ‘hollowing out’ of government observed in some Western (i.e. North American and European) states, whereby private actors were delegated responsibilities previously held by the state in the belief that such delegation would increase efficiency or diminish financial risk to the state. The role of a CSO as an implementer of state policy, however, features heavily in some of the more critical literature on civil society,54 and in realist and functionalist perspectives. Separating out this role helps isolate these critiques and understand the relationship between policy implementation and policy or norm formation.Many CSOs combine the agent role with other activities. For instance, religious organizations in the USA may receive government funds to run homeless shelters and yet also act as advocates on behalf of the homeless. Organizations like Oxfam receive bi- and multi-lateral funding for international development, yet are also powerful voices in debates on development policy. It is likely that taking on the role of agent either diminishes or magnifies an organization's capacity for advocacy or revolution, but in the interest of parsimony, the various roles will be treated discretely.When operating as an agent, a CSO must be judged by the democratic credentials of the state or institution on whose behalf it acts. If a CSO acts on behalf of a democratic state, it may be considered democratic; if it acts on behalf of an undemocratic one, it may be considered as undemocratic because of the type of regime it is supporting. It is important to note that this must be examined differently in the national and international realms. In a wholly domestic context, in which a CSO is funded by the government on whose behalf it works, the principal–agent relationship is clear. Internationally, the situation is more complex. A CSO may be funded by a government or a multi-lateral organization, for work in another polity. In this case, the will of the people in the polity in which the work is done must be considered, insofar as they will reap the benefits or suffer the consequences of the CSO's activities. Even in those instances in which a CSO's intervention is approved by a local democratic government, one must also consider whether the local government truly desires the CSO's services or whether those services have been forced upon it by more powerful states or organizations.55 Thus, a CSO acting as an agent may be considered democratically legitimate if it works under contract to a legitimate representative of the people impacted by its work, or if a majority of these people themselves approve that work. The choice of principal (including the alignment of interests between an external principal and the local will) may be considered a form of input. Therefore a CSO acting as an agent is judged based on input legitimacy.The first three roles occupied by civil society—revolutionary, advocacy, and agent—have been thoroughly examined in the literature. It is tempting to assume that the observations made about CSOs acting in these well-recognized roles and contexts transfers to civil society involvement in international policymaking. In reality, however, CSO involvement in international policymaking requires recognition of a new role.When a CSO engages in global policymaking (either in a de jure way through formal participation in international decision-making or in a de facto way through the propagation of international norms), it is acting as an authority. The reach of both individual CSOs and networks of organizations frequently spans national boundaries, and transnational activism often results in the creation of international networks. Activists make broad claims of popular support. At the same time, the rise of global problems like terrorism and climate change has necessitated international collaboration to a degree unprecedented in political history. Technology has further facilitated multi-lateral collaboration, and international institutions like the UN, World Bank, and WTO have laid the framework for global governance. Populist claims of non-state actors, international communication, transnational problems, and global governance have all combined to challenge states’ claims to act as the sole voice of their citizens in international fora.At the same time, the research during past decade has demonstrated that CSOs have the power to change the international behavior of nation-states and institutions and to create new norms and regimes. Individual CSOs and networks of concerned actors have been credited with playing a significant role in expanding human rights standards and environmental norms, and even in nuclear disarmament.56 CSOs were the driving force behind debt forgiveness and the Ottawa Convention banning landmines. CSOs, either independently or through campaigns and networks, have greater agency or reach in the international realm than many states or institutions possess. CSOs are neither merely acting against them (in a revolutionary role) nor is acting within them (as advocate). A CSO may claim to occupy these roles and indeed CSOs often conduct themselves as though they are revolutionaries or advocates. In truth, however, civil society has established itself as a new mechanism of citizen influence. CSOs are a part of contemporary global governance arrangements. Thus, CSOs are best described as acting in the role of authority.A CSO's democratic legitimacy when acting in this new role is determined by the international context. The global arena lacks clearly defined democratic protections for its citizens. The possibility of structured representation has been proposed,57 and some authors contend that the European Union (EU) is a successful test case for the possibility of a cosmopolitan global government.58 Currently, however, no enforceable democratic rights exist for global politics. Thus, as a participant in global governance, a CSO must be judged on the same criteria by which other international actors, i.e. states and institutions, have been judged: whether they provide for equal representation, respond to citizen control, and protect fundamental rights. These are essential questions of input and output. Thus, in this context a CSO's legitimacy is judged on both input and output. When we compare works which examine CSOs within equivalent contexts, we see the contextual standard of democratic legitimacy vindicated by the similarities among context-equivalent theories. At the same time, revisiting the literature and applying this new standard highlights ways in which writing on TCS has ignored the unique democratic requirements of the international context.Despite the seeming jumble of civil society literature and the numerous fault-lines identified, most theories of civil society are more complementary (or deviations between them take place on clearly identifiable theoretical grounds) when they are viewed through the lens of role and context. The literature on civil society and national democratizations clearly reflects the standards of the revolutionary role.59 Such literature describes undemocratic regimes and judges civil society, usually positively, for its role in contesting them. According to the standards presented here, organizations involved in national democratizations should be judged based on their output legitimacy. Studies of such organizations can be judged on the extent to which they recognize and theorize the importance of output legitimacy in determining CSOs’ democratic credentials.Writing on advocacy and interest groups and their roles in the democratic process should reflect the standards of the advocacy role.60 Again, much of it does. The context and rationale of the advocacy role explains why this literature focuses on the behavior of organizations or coalitions, their mechanisms of influence, and ways such behaviors and influence model or create democratic throughput.Studies of NGOs as implementers of state policy come under the agent role. Here we find much of the development studies literature. The input legitimacy criteria of this category are reflected in the emphasis of much of this literature on principal–agent relationships and the impacts of foreign intervention on local representation and autonomy.In short, this pattern in the literature supports the use of context and the disaggregated components of legitimacy in judging CSOs democratic credentials. At the same time, however, it highlights the dangers of transferring a model from one context to another without sufficient study or adaptation. In particular, the four-role parsing allows us to identify two common flaws in the literature on TCS actors. One is the failure to recognize the significance of the international context and thus treating global policymaking (i.e. the authority role) as though it is taking place within an established state. The other is treating local policymaking (i.e. intervention by TCS actors in local political affairs) as though it is taking place within the global context. Misidentifying a CSO's role can lead to two errors when examining CSO involvement in international policy or norm formation. The first is to treat civil society actors as though they are revolutionaries. This attitude is particularly common among practitioners, who are prone to interpreting the unwillingness of some global institutions to accede to a CSO's demands as ‘evidence’ for a democratic deficit at those institutions.61 Adherents to this view frequently push for the elimination of international governance mechanisms or for them to be reformed in a way that gives greater voice and authority to CSOs. A CSO is legitimated by its opposition to the perceived injustices of the current international system. The emphasis is thus primarily on outputs, i.e. how much change a CSO can force on the current order. Representative inputs are largely assumed.The alternative error is to judge a CSO as though it is acting as an advocate, occupying that role as it does within established democratic states. This perspective is more common among academics.62 This attitude presupposes that civil society activism is legitimate as long as it is supporting someone or something. The emphasis thus is primarily on inputs, on a CSO's ties to its claimed constituents, clients, or ideas. Throughput is sometimes suggested as an additional measure of legitimacy, usually in the form of transparency and accountability to clients or constituents. However, it is not clear that such throughput mechanisms always make CSOs more responsive to the people they impact.Both of these approaches ignore the consequences of the absence of a democratic, global superstate. In the absence of a global state (and without any realistic, near-term prospect of creating such a state), revolutionary or advocacy behaviors at the transnational level do little to enhance the democratic rights of citizens. As Bowden writes, citing Hegel, when civil society exists without the state ‘the interest of individuals as such becomes the ultimate end of their association.’63 CSOs can act as effective interest advocates, but there is no state government which can subject individuals to the concerns of others who do not share their interests or needs. Similarly, there is neither a mechanism capable of enforcing democratic representation nor any superior authority capable of protecting the rights of those stakeholders without a powerful interest group of their own. Walzer writes regarding the synergies between the state and civil society: [A]cross the entire range of association, individual men and women need to be protected against the power of officials, employers, experts, party bosses, factory foremen, priests, parents, and patrons; and small and weak groups need to be protected against large and powerful ones. For civil society, left to itself, generates radically unequal power relationships.64This is indeed the problem with CSO activity in the global context: it is civil society largely left to itself. Whereas CSOs may be regulated and their influence counterbalanced within the confines of any given polity, TCS activities transcend the authority of any state or supranational institution. Because of its power in this context, a CSO is a de facto authority and its democratic legitimacy must be judged based on its fulfillment of this role. (Alternatively, it can subject itself to a democratically legitimate authority and act as an agent.) Removing or reforming existing institutions (often in a way that gives the CSO or its allies more power within them), only exacerbates the problems of the stateless context.Acting as an advocate makes use of the situation without mitigating it, exploiting the absence of a superstate to advance the CSO's own agenda. A CSO functioning as an advocate may be accountable to those whom it claims to represent, but unless the CSOs involved in creating or implementing a given policy are in some fashion accountable to everyone impacted by their work, the situation can easily facilitate tyranny and the abuse of power. The result is a situation in which those who ‘shout the loudest’ win.65 Neither revolutionary nor advocacy behavior can be democratically legitimate in this context. Judging a CSO as though it is occupying a revolutionary or advocacy role only serves to rationalize democratically illegitimate behavior.For a CSO to be democratically legitimate in the transnational context, it must rise to the standards to which states and institutions are held. Cosmopolitan theorists seem to have gone furthest in recognizing this problem. Held's acknowledgment that a true cosmopolitan democracy will require representative political structures is informative.66 Insofar as CSOs themselves are part of global governance, CSOs in the aggregate must seek to represent all stakeholders in any given policy, not just those to which the organizations are most closely tied, and to achieve outcomes that reflect the will of the majority while protecting liberal rights. Only in those cases in which CSOs, campaigns, or networks seek to determine and enforce majority rule and the protection of acknowledged rights can CSOs’ involvement in transnational policymaking be said to be democratically legitimate. Any analysis of the democratic legitimacy of civil society activity in the transnational context that does not recognize and grapple with the fundamental problem of statelessness is critically flawed. One must also be cautious when writing about the interventions of TCS actors in the domestic policies of a democratic nation. Examples of such intervention abound. International NGOs or movements may apply direct pressure to a national government (via publicity campaigns, boycotts, lobbying, etc.). International actors may also apply pressure indirectly, for example by pushing donors to make aid funding conditional on specific policy change. Foundations or NGOs may initiate ‘grassroots’ campaigns, establish and staff local offices, or fund existing indigenous movements. Such domestic interventions are a key means by which TCS actors or networks have impact.Unfortunately, much of the writing on transnational advocacy has failed to delineate between such domestic intervention and CSOs’ involvement in international policymaking. There are several reasons the two have been conflated. First, the majority of TCS campaigns over the last 30 years have focused on problems and policies in the developing world. In developing nations, national decisions frequently involve some international component because many domestic policies or programs rely on international funding. Thus, a decision by Brazil to build a rail line or an undertaking in Niger to prioritize primary-school education can easily be depicted as an externally driven World Bank project (or an EU program or an IMF policy), rather than as a national decision. Second, the process of global norm formation can take place on both the national and international level. Transnational activists may promote an international norm of condemning torture or protecting children's rights, and then seek to have that norm applied to individual states. Its adoption by successive states, in turn, helps establish it as a global norm. Third, during the initial development of modern international CSOs and transnational networks in the 1980s and early 1990s, the majority of the world's population did not live in democratic states. Ignoring the role of the local state was easy because many states were perceived as illegitimate. Undemocratic states and undemocratic international institutions were easily tarred with the same brush, dismissed as mere obstacles in the pursuit of ‘good’ or ‘democratic’ policy.The most significant change during the past 10 years with regards to this trend is that the majority of the world's citizens now live in recognized democracies. This evolution has permitted more governments to function as the legitimate representatives of their citizens and has created an environment in which local CSOs can fulfill the advocacy role with its functions of watchdog and representative. As two civil society leaders from the global South have written: The new political context that has emerged, marked by democracy and citizen participation, has increasingly led to more collaborative modes of relating [to government]. It is civil society's participation in political life, in that realm of public life in which societal decisions are made and carried out, that provides the conditions for sustainable development.67In such a context, it is imperative that academics and practitioners draw careful lines between national and international policies. For instance, it is helpful to distinguish between ‘World Bank’ projects like the Narmada Dam that are actually planned by national governments before the intervention of the World Bank, and policies like structural adjustment that are largely international creations. Likewise, one must distinguish between when a CSO (or campaign or network) is developing an international norm that is widely accepted within national democracies, and when a CSO is seeking to use international politics to impose the will of a well-resourced minority on weak states. An example of the former would be developing an international norm against torture and imposing it on Indonesia.68 An example of the latter would be when US-based environmentalists seek to direct development funds for Brazil toward decreasing energy consumption rather than building more power plants.69 The latter ‘norm’ is not democratic because it seeks to impose via fiduciary fiat a policy that would ordinarily be the subject of public debate in a developed, democratic nation.When a CSO is interacting with an internationally created project or policy, it occupies the authority role described earlier and is subject to the legitimacy tests described in the preceding section. However, when a CSO is working transnationally to influence a domestic policy, it must be judged by those standards that are applied to national civil society. Essentially it can act in either the revolutionary or advocate role and its legitimacy must be assessed accordingly.It is tempting to treat transnational networks or campaigns as though they are above such considerations or as though their large international followings are an automatic source of legitimacy. Yet the non-local members of such campaigns are neither subject to the local polity nor part of the demos it governs. Allowing such international voices to overwhelm local democratic procedures is to give the members of international organizations or movements power on par with that of local citizens despite the fact that these global citizens have no allegiance to the country they are impacting and are unlikely to bear the immediate consequences of the plan they impose. International organizations or movements should not be presumed to be democratically legitimate simply because they are large.Instead, as per the revolutionary and advocacy roles discussed earlier, the legitimacy of TCS intervention in local affairs must be judged by the degree to which it contributes to sustainable, national democracy. A transnational actor, by definition, is not tied to a single country. When the international network leaves a dam cancellation fight in India to protest a dam in Pakistan, it leaves the local citizens behind. It is unable to offer them long-term democratic protections. If it does not contribute to the democratic functioning of the national state, then it has had no long-term impact on the democratic well-being of local citizens. It may have helped some of them win a particular battle, but it has done nothing to win the proverbial war. If the international intervention has promoted special interests at the expense of majority rule, the situation is even worse. In such a case, global activists may have actually undermined the functioning or legitimacy of the national regime responsible for ensuring most day-to-day democratic rights of local citizens. Granted, an exception may be made if an international campaign intervenes to protect minority rights against a tyrannical majority. In this sense it is supporting the ‘liberal’ portion of the definition of liberal democracy outlined earlier, provided that the rights supported rise to the level of internationally recognized liberal norms. However, if CSOs’ involvement merely swaps an overweening local majority for an overweening global minority, it has done little to support long-term, sustainable democracy. To be democratically legitimate, transnational CSOs operating must either work within (and by the rules of) any existing local democratic system, or seek to replace an undemocratic system with a democratic one. The input–throughput–output vision of democratic legitimacy used in this article establishes a series of roles in which CSOs must operate in order to achieve democratic legitimacy. Identifying the link between context and legitimacy helps explain similarities and tensions among different sub-sections of the civil society literature. Thus, in addition to creating a systematic approach to determining CSOs’ democratic legitimacy, these contextual criteria may have the effect of resolving certain debates within the literature. Even where it does not, this more nuanced approach can be used to establish lines of argumentation and help develop the sort of clear-cut intellectual debates necessary to move research and theory in this area forward.This article's use of democracy and context to evaluate the existing literature reveals two areas of special concern regarding the study of the democratic legitimacy of CSOs’ involvement in transnational policymaking or norm formation. The first is the way in which global models borrow too heavily from the national context, ignoring the fact that CSOs, campaigns, and networks operating outside state boundaries can themselves be tyrannical. As cosmopolitans have already begun to do, future researchers should define the components of global democracy. Once those theoretical elements are defined, however, they must be joined to an empirical assessment of the ways in which CSOs currently contribute to or inhibit the development of democratic global governance. A related area of concern is the failure to delineate between TCS involvement in international policy and transnational involvement in local or national policy. Too often in such instances, researchers judge transnational CSOs’ democratic credentials based on their international input or throughput processes, without fully considering the organizations’ impacts on local democracy. This can lead to normative prescriptions that negatively impact the democratic rights of many stakeholders. Future research on ‘The Democratic Legitimacy of Transnational Actors’ must remedy these shortcomings. The tools and approaches presented here should facilitate that process. The author wishes to thank Anders Uhlin, Eva Erman, Jonas Tallberg, and Enrique Peruzotti for their comments on early drafts of this article. 1. Brett Bowden, ‘Civil Society, the State, and the Limits to Global Civil Society’, Global Society 20 (2006): 155–78; Alexander Cooley and James Ron, ‘The NGO Scramble: Organizational Insecurity and the Political Economy of Transnational Action’, International Security 27 (2003): 5–39; Michael Foley and Bob Edwards, ‘The Paradox of Civil Society’, Journal of Democracy 7 (1996): 38–52; Paul Nelson, ‘Deliberation, Leverage, or Coercion? The World Bank, NGOs, and Global Environmental Politics’, Journal of Peace Research 34 (1997): 467–70. 2. Ronald Lipschutz, ‘Reconstructing World Politics: The Emergence of Global Civil Society’, Millennium: Journal of International Studies 21 (1992): 391. 3. Ibid., 420. 4. Martin Shaw, ‘Global Society and Global Responsibility: The Theoretical, Historical, and Political Limits of International Society’, Millennium: Journal of International Studies 21 (1992): 432. 5. Ibid., 434. 6. David Held, Democracy and the Global Order: From the Modern State to Cosmopolitan Governance (Cambridge: Polity, 1995); Peter Spiro, ‘New Global Communities: Nongovernmental Organizations in International Decision-Making Institutions’, The Washington Quarterly 18 (1995): 45–56; Rodger Payne, ‘Deliberating Environmental Politics’, Journal of Peace Research 33 (1996): 129–36. 7. Peter Willetts, ed., The Conscience of the World: The Influence of Nongovernmental Organisations in the UN System (London: Hurst, 1996); Carrie Meyer, ‘The Political Economy of NGOs and Information Sharing’, World Development 25 (1997): 1127–40; Nelson, ‘Deliberation, Leverage, or Coercion?’ 8. Cf. Alexander Wendt, ‘Constructing International Politics’, International Security 20 (1995): 71–81. 9. E.g. Matthew Evangelista, Unarmed Forces: The Transnational Movement to End the Cold War (London: Cornell University Press, 1999); Ann Marie Clark, Diplomacy of Conscience: Amnesty International and Changing Human Rights Norms (Princeton, NJ: Princeton University Press, 2001). 10. Bowden, ‘Civil Society, the State, and the Limits to Global Civil Society’; cf. Lipschutz, ‘Reconstructing World Politics’; Foley and Edwards, ‘The Paradox of Civil Society’. 11. Richard Price, ‘Transnational Civil Society and Advocacy in World Politics’, World Politics 55 (2003): 579–606. 12. Bowden, ‘Civil Society, the State, and the Limits to Global Civil Society’. 13. Foley and Edwards, ‘The Paradox of Civil Society’; Nelson, ‘Deliberation, Leverage, or Coercion?’; cf. Terje Tvedt, ‘Development NGOs: Actors in a Global Civil Society or in a New International Social System’, Voluntas: International Journal of Voluntary and Nonprofit Organizations 13 (2002): 363–75. 14. Price, ‘Transnational Civil Society and Advocacy in World Politics’. 15. Margaret Keck and Kathryn Sikkink, Activists Beyond Borders: Advocacy Networks in International Politics (Ithaca, NY: Cornell University Press, 1998). 16. Jonathan Fox and L. David Brown, eds., The Struggle for Accountability: The World Bank, NGOs, and Grassroots Movements (Cambridge: MIT, 1998). 17. Ann Florini and P.J. Simmons, ‘What the World Needs Now?’ in The Third Force, ed. Ann Florini (Washington, DC: Carnegie), 1–16. 18. Paul Nelson, ‘Heroism and Ambiguity: NGO Advocacy in International Policy’, Development in Practice 10 (2000): 478–90; Jan Scholte, ‘Civil Society and Democratically Accountable Global Governance’, Government and Opposition 9 (2004): 211–33. 19. Bowden, ‘Civil Society, the State, and the Limits to Global Civil Society’. 20. Tvedt, ‘Development NGOs’; Cooley and Ron, ‘The NGO Scramble’; Kal Raustiala, ‘States, NGOs, and International Environmental Institutions’, International Studies Quarterly 41 (1997): 719–40. 21. Nelson, ‘Heroism and Ambiguity’; Jonathon Murphy, ‘The World Bank, INGOs, and Civil Society: Converging Agendas? The Case of Universal Basic Education in the Niger’, Voluntas 16 (2005): 353–74. 22. Kenneth Anderson, ‘The Ottawa Convention Banning Landmines, the Role of International Non-governmental Organizations and the Idea of International Civil Society’, European Journal of International Law 11 (2000): 91–120. 23. Ngaire Woods, ‘The Challenge of Good Governance for the IMF and World Bank Themselves’, World Development 28 (2000): 823–84; Firoze Manji and Carl O'Coill, ‘The Missionary Position: NGOs and Development in Africa’, International Affairs 7 (2008): 567–83. 24. Diane Stone, ‘Knowledge Networks and Global Policy’, in Global Knowledge Networks and International Development, ed. Diane Stone and Simon Maxwell (New York: Routledge, 2005), 89. 25. Candido Grzybowski, ‘We NGOs: A Controversial Way of Being and Acting’, Development in Practice 10 (2000): 442. 26. Florini and Simmons, ‘What the World Needs Now?’; Nelson, ‘Deliberation, Leverage, or Coercion?’ 27. Foley and Edwards, ‘The Paradox of Civil Society’. 28. Ngaire Woods, ‘Making the IMF and the World Bank More Accountable’, in Reforming the Governance of the IMF and the World Bank, ed. Ariel Buira (London: Anthem Press, 2005), 149–70; Nelson, ‘Heroism and Ambiguity’. 29. Cooley and Ron, ‘The NGO Scramble’. 30. Manji and O'Coill, ‘The Missionary Position’. 31. David Korten, Globalizing Civil Society: Reclaiming Our Right to Power (New York: Seven Stories Press, 1998); David Held, ‘Democratic Accountability and Political Effectiveness from a Cosmopolitan Perspective’, Government and Opposition 9 (2004): 364–91. 32. Keck and Sikkink, Activists Beyond Borders; Thomas Risse, ‘The Power of Norms versus the Norms of Power: Transnational Civil Society and Human Rights’, in The Third Force, ed. Ann Florini (Washington, DC: Carnegie, 2000): 177–210. 33. Robert Dahl, ‘Can International Institutions Be Democratic?’, in Democracy's Edges, ed. Ian Shapiro and Casiano Hacker-Cordon (Cambridge: Cambridge University Press, 1999), 19–36. 34. Raustiala, ‘States, NGOs, and International Environmental Institutions’. 35. Magdalena Bexell, Jonas Tallberg, and Anders Uhlin, ‘Democracy and Global Governance: The Promises and Pitfalls of Transnational Actors’ (paper presented at the annual conference of Millennium: Journal of International Studies, London, October 25–26, 2008). 36. David Held, Models of Democracy, 3rd ed. (Stanford, CA: Stanford University Press, 2006). 37. Patrizia Nanz and Jens Steffek, ‘Global Governance, Participation, and the Public Sphere’, Government and Opposition 9 (2004): 314–35. 38. E.g. Payne, ‘Deliberating Environmental Politics’; Gryzbowski, ‘We NGOs’. 39. Scholte, ‘Civil Society and Democratically Accountable Global Governance’; Nelson, ‘Deliberation, Leverage, or Coercion’. 40. Dahl, ‘Can International Institutions Be Democratic?’ 41. Kumi Naidoo and Rajesh Tandon, ‘The Promise of Civil Society’, in Civil Society at the Millennium, ed. Civicus (Hartford, CT: Kumarian, 1999); see also, Ezra Mborgori and Hope Chigudu, ‘Civil Society and Government: A Continuum of Possibilities’ in the same volume. 42. Simmons, ‘Learning to Live with NGOs’, Foreign Policy 112 (1998): 82–96; Nelson, ‘Deliberation, Leverage, or Coercion’; Anderson, ‘The Ottawa Convention Banning Landmines’. 43. Dahl, ‘Can International Institutions Be Democratic?’ 21. 44. Ibid. 45. Ibid.; also Andrew Moravcsik, ‘Is there a “Democratic Deficit” in World Politics? A Framework for Analysis’, Government and Opposition 39 (2005): 336–63. 46. Dahl, ‘Can International Institutions Be Democratic?’ 20; see also, Moravcsik, ‘Is There a “Democratic Deficit” in World Politics?’ 339. 47. Held, Models of Democracy, 59. 48. Anders Uhlin, ‘The Democratic Legitimacy of Transnational Actors: A Framwork for Analysis’, in Legitimacy Beyond the State? Re-examining the Democratic Credentials of Transnational Actors, ed. Anders Uhlin and Eva Erman. (London: Palgrave, 2010). 49. Fritz Scharpf, Governing in Europe: Effective and Democratic? (Oxford: Oxford University Press, 1999), 6–11. 50. Klaus Dingwerth, The New Transnationalism: Transnational Governance and Democratic Legitimacy (New York: Palgrave MacMillan, 2007). 51. Uhlin, ‘The Democratic Legitimacy of Transnational Actors’. 52. Keck and Sikkink, Activists Beyond Borders. 53. See, for example, the cases in Fox and Brown eds., The Struggle for Accountability; Dana Clark, Jonathan Fox, and Kay Treakle, eds. Demanding Accountability: Civil Society Claims and the World Bank Inspection Panel (Oxford: Rowman and Littlefield, 2003). 54. E.g. Cooley and Ron, ‘The NGO Scramble’; Manji and O'Coill, ‘The Missionary Position’. 55. For a detailed discussion, see Tvedt, ‘Development NGOs’. 56. Clark, Diplomacy of Conscience; Susan Burgerman, Moral Victories: How Activitsts Provoke Multilateral Action (London: Cornell University Press, 2001); Evangelista, Unarmed Forces. 57. E.g. Held, Models of Demcoracy, 304–7. 58. Jurgen Habermas, ‘Toward a Cosmopolitan Europe’, Journal of Democracy 14 (2003): 86–100; Moracvsik, ‘Is There a “Democratic Deficit” in World Politics?’ 59. E.g. Petr Kopecky and Cas Mudde, ‘Rethinking Civil Society’, Democratization 10 (2004): 1–14; Lorenzo Fioramonti, ‘Civil Societies and Democratization: Assumptions, Dilemmas, and the South African Experience’, Theoria 52 (2005): 65–88. 60. See, for example, Paul Sabatier and Hank Jenkins-Smith, eds. Policy Change and Learning: An Advocacy Coalition Approach (Oxford: Westview, 1993). 61. See, for example, Bruce Rich, Mortgaging the Earth: The World Bank, Environmental Improvishment, and the Crisis of Development (Boston, MA: Beacon Press, 1994); Lori Udall, ‘The World Bank and Public Accountability: Has Anything Changed?’, in The Struggle for Accountability, ed. Fox and Brown (Cambridge: MIT, 1998), 391–438. 62. E.g. Risse, ‘The Power of Norms versus the Norms of Power’; Margaret Keck, ‘Governance Regimes and the Politics of Discursive Representation’, in Transnational Activism in Asia, ed. Nicola Piper and Anders Uhlin (London: Routledge, 2004), 43–60. 63. Bowden, ‘Civil Society, the State, and the Limits to Global Civil Society’, 163. 64. Michael Walzer, ‘The Concept of Civil Society, in Toward a Global Civil Society, ed. Michael Walzer (Providence, RI: Berghahn Books, 1995), 23. 65. Held, Models of Democracy, 307. 66. Ibid., 303–11. 67. Naidoo and Tandon, ‘The Promise of Civil Society’, 9. 68. Risse, ‘The Power of Norms versus the Norms of Power’. 69. Rich, Mortgaging the Earth.