Legitimacy in Global Governance: International, Transnational and Private Institutions Compared

Transcription

1 Swiss Political Science Review 18(2): doi: /j x Legitimacy in Global Governance: International, Transnational and Private Institutions Compared INGO TAKE University of Go ttingen Abstract: How to generate legitimate forms of governance beyond the nation state is often considered a central question in contemporary world politics (Moravcsik 2004: 336). To proceed in theory-building, scholars need to systematically assign the theory-driven assumptions on legitimate forms of governance beyond the nation-state with the various, already observable, forms of global governance. This article aims to conduct a comparative appraisal of the legitimatory quality of different patterns of governance by applying a framework of indicators for their assessment. The indicators are selected from the scholarly debate within International Relations on the legitimacy of global governance arrangements and structured by a multi-dimensional concept of legitimacy (input-, throughput- and output-dimension). This framework is then applied to international, transnational and private forms of global governance in three policy fields in order to show how each of them try to produce and maintain legitimacy, which strategies they apply and in how they interact with their stakeholders. KEYWORDS: Global Governance, Legitimacy, Internet Regulation, Environmental Standards, Social Standards Introduction Conceptualizing legitimate forms of governance beyond the nation-state is a central concern of the governance-approach, which has received extensive attention in International Relations (IR) scholarly debate since the mid-1990s. To advance in theory-building, scholars need to systematically assign the theory-driven assumptions on legitimate forms of governance with the various already observable forms of global governance (vgl. Mu gge 2011: 69; Bexell Tallberg Uhlin 2010: 96). The latter implies establishing problem-centred rules of cross-border governance arrangements that possess neither formal authority nor central enforcement power, and feature different actors, forms of organisation and steering patterns. Due to the variety of forms of governance beyond the nation-state, it is important to distinguish between international forms of governance (inter-governmental institutions), transnational forms of governance (networks between international, national and non-state actors) and private forms of governance (networks consisting only of non-state actors). 1 Transnational and private governance arrangements in particular depend on the voluntary cooperation of rule-addressees to be effective and therefore must generate legitimacy from within in order to enforce their rules. Against this background, these new forms and mechanisms of effective and legitimate governance beyond the nation-state become essential issues 1 In practice, these varying categories can and do overlap. Nevertheless, for analytical purposes, it remains justifiable to separate the three forms of governance. One of the main disparities between them lies in the different roles non-state actors play in agenda- and rule-setting. Ó 2012 Swiss Political Science Association

2 Legitimacy in Global Governance 221 to investigate. Several scholars and politicians critically assess the problem-solving capacity and the democratic quality of these new forms of governance and see them as an attempt to erode the sovereignty of states. Others consider them to be an adequate answer to steering deficits of states and international organizations. Thus, only a comparison between traditional forms of international governance with mostly novel forms of transnational and private global governance can provide a reference point for assessing to what extent the respective concerns are justified. To date, no such three-dimensional comparative analysis of different forms of global governance exists. This article, based on an extensive empirical study of nine cases, presents a proposal how to conduct such a comprehensive study. Because the capacity of global governance arrangements to enforce rules and to solve problems in an effective way depends on their acceptance as legitimate actors in the view of their stakeholders (cp. Scholte 2011: 111), the focus of this study is on legitimacy. 2 It examines these new forms of governance by conducting a comparative appraisal of the legitimacy possessed by transnational and private patterns of governance in comparison with international governance arrangements in the same issue areas. The nine case studies encompass international, transnational and private global governance arrangements in the fields of forest certification, child labour and internet regulation. The guiding research question is: By what factors are global governance arrangements able to generate legitimacy and why might that answer vary across different forms of global governance arrangements? In addition, the article asks: Is there a correlation between normative and empirical legitimacy? To answer these questions, suitable indicators measuring the legitimacy of different forms of governance and their respective acceptance levels must be identified. The operationalization of such indicators therefore depends on the conception of legitimacy guiding the research, the manner in which the international environment (in which the actors are operating) is conceptualized and the extent to which certain standards of legitimacy are seen as appropriate to this environment (Wolf 2002). 3 Because of diverse contexts, governance beyond the nation-state cannot refer to forms of decision making and mechanisms of enforcement established at the state level (cp. Bernstein 2011; Cerutti 2011: 123; Clark 2007: 193; Buchanan Keohane 2006). Therefore, the criterion for legitimate global governance must be chosen with regard to the structural characteristics of cross-border policy and cannot be derived from the ideals of representative democracy. Because governance beyond the nation-state is incompatible with the democratic principle of sovereignty of the people, a suitable alternative concentrates on functional and sectoral forms of representation (i.e. the organized participation of interest groups to generate input-legitimacy). Furthermore, one cannot revert to elections on the global level in order to hold decision-makers accountable. Instead, independent mechanisms of monitoring and sanctioning have to be established to ensure the decision-maker s accountability and by this throughput-legitimacy. In order to evaluate the legitimacy of selected governance arrangements, first a checklist of indicators of legitimate governance must incorporate the degree to which the observable forms of organization and the procedures of differing governance arrangements meet the 2 According to Moravcsik, this focus is justified: Is global governance [ ] democratically legitimate, or does it suffer from a democratic deficit? This is emerging as one of the central questions perhaps the central question in contemporary world politics (2004: 336). 3 If what is assumed in political science is correct, namely the rise of governance goes along with a transformation of democracy ( ), then it cannot simply be assumed that the standard model of democracy in the nation state can be applied. A concept which conforms to the normative standards of democratic theory but which at the same time is adjusted to the subject of inquiry is needed (Benz Papadopoulos 2006: 4).

3 222 Ingo Take criteria for normative legitimacy (supplemented by substantive legitimacy). Secondly, interviews must be evaluated to empirically appraise the de facto acceptance by the stake-holders of the rules and decisions established by the different governance arrangements. This structured comparison of different forms of global governance should help to identify adequate institutional designs. Such a comparative review of what until now has been considered quite unrelated hypotheses is a first step to a new and higher standard for a plausibility test. Its results may constitute a pre-stage formulation of a theoretical explanation of corresponding causalities. In addition, to the extent that the chosen indicators reveal a legitimizing effect, they could serve as a standard for future research studies and, as such, contribute to further systematization of studies on global governance. This standard can also provide the basis for criticism of global governance institutions and guide further reform efforts. In this way, the study can make a contribution to provide insights regarding the conditions, prospects and limits of legitimate governance beyond the nation state. This article introduces a conceptualization of legitimacy which supersedes the bounds of the nation state, and proceeds in the following five steps. I first differentiate between normative and empirical legitimacy. Second, I further categorize the data into input-, throughput- and output-legitimacy, supplemented by content-related and substantive legitimacy. Third, the selection criteria of the case studies will be briefly sketched out. Fourth, the indicators introduced in the conceptual part will be applied to the nine case studies. Finally, among many specific conclusions, it will be broadly demonstrated that none of the nine empirical case studies show that international governance in principle can claim a higher degree of legitimacy than transnational or private forms of governance and that a systematic correlation between the acceptability of a governance arrangement and the degree of its acceptance does not seem to exist. Legitimacy of Governance Arrangements The following section introduces a checklist of indicators that does not assume any a priori bases of legitimacy (because legitimacy is an essentially contested concept; cp. Brassett Tsingou 2011: 1; Hurd 2007: 2) but draws on recently debated touchstones of legitimate governance beyond the nation-state. While I do not survey the entirety of this discussion, the most often cited indicators are included. 4 These indicators are then systematically combined according to the well-established distinction between input-, throughputand output-legitimacy, 5 without claiming that the chosen indicators are the only necessary and sufficient conditions for legitimacy. 6 Accordingly, this section does not claim to present a full-fledged model of legitimate global governance. Instead, it introduces a systematically structured proposal for how the legitimacy of different forms of global governance arrangements can be appraised. 7 This preliminary approach delivers a framework of analy- 4 Some of the indicators presented here can be equalized with others not mentioned: for example, consensus-orientation covers deliberation and consultation, monitoring encompasses evaluation and responsiveness and sanctioning comprises correction. 5 The differentiation between the input-, throughput- and output-dimension of legitimacy - as sketched in the following sentences is part of the legitimacy concept applied in this study and does not correspond to different phases of the policy process. 6 This approach is justified insofar as there are no universally shared criteria of legitimacy in global governance (Koppell 2008: 192). 7 By avoiding a fully formulated theoretical framework, the risk of neglecting important indicators of legitimate governance is minimized.

4 Legitimacy in Global Governance 223 sis and serves as a first plausibility test for the explanatory power of the chosen indicators. The input-dimension addresses the question of who is entitled to make decisions and who is to be represented in the decision-making process (Scharpf 1999). Here, the equal and active participation in the processes of decision-making by all stakeholders has to be assured in order to derive a direct mechanism that effectively imparts the demands and preferences of the rule addressees (cp. Benz Papadopoulos 2006: 275). The throughputdimension addresses the procedural level (Zu rn 2000) focusing on to what extent decisionmaking is transparent and to what extent decision-makers will be held accountable for their actions. The catalogue of input- and throughput-indicators measuring the normative legitimacy of governance arrangements is supplemented by content-related and substantive legitimacy as well as empirical data on stakeholders acceptance of a global governance arrangement as an indicator for output-legitimacy. 8 This empirical form of legitimacy has its roots in Weberian social science and is conceptualized as the governance arrangement s capacity to engender acceptance on the part of internal and external stakeholders (Beetham 1991: 9). Thus, normatively founded legitimacy (acceptability of rules and decisions) is supplemented by empirically founded legitimacy (societal acceptance of rules and decisions) (cp. Bernstein 2011; Hurrelmann Schneider Steffek 2007: 3; Buchanan Keohane 2006: 405; Zu rn 2004: 260). At this stage of research, it is assumed that the three dimensions complement one another. 9 Accordingly, it is expected that the more indicators a global governance arrangement satisfies and the higher the degree to which it satisfies them, the more legitimate is the organization. Furthermore, it is hypothesized that the higher the degree of normative legitimacy, the higher the empirical acceptance of a global governance arrangement. In the following, the selected indicators of legitimate governance are presented. Representativeness of the Governance Arrangements To claim input-legitimacy, I assume that sector-specific governance arrangements have to include all relevant stakeholders in the decision-making process (matching principle). This is justified by the assumption that stakeholders preferences are not fixed in advance and unchangeable but are formed during the process of standard-setting and decision-making. Therefore, such processes have to be designed in a way which guarantees the maximum possible participation (Dahl 1989). However, it is obvious that for practical reasons not all rule addressees can be included in the decision-making process. In this respect, governance arrangements have to identify their stakeholder-groups and mobilize them for participation. Therefore, one has to ask which criteria are suitable for the selection of stakeholders. In order to measure the extent to which the matching principle is attained, it must be examined if certain stakeholder groups are excluded, and if the system favours a special group of stakeholders. In this study, a high degree of representativeness is given when all stakeholder-groups are provided with formal participation rights; a middle degree is given when all stakeholder-groups are at least provided with informal participation 8 Legitimacy ( ) lies as much in the values, interests, expectations, and cognitive frames of those who are perceiving or accepting the regime as they do in the regime itself (Black 2008: 145). ( ) what constitutes legitimacy results from an interaction of the community of actors affected by the regulatory institution, i.e. the public who grant legitimacy (Bernstein 2011: 19). 9 This lowers at least to a certain degree the problem, that the different indicators can not separated accurately.

5 224 Ingo Take rights; and a low degree is given when several stakeholder-groups are excluded from the work of a governance arrangement. Level of equality Another central problem of input-legitimate governance is the asymmetrical allocation of power, resources and communicative capabilities of the different stakeholders. Asymmetric power can undermine the effective claim of formally equal participation rights and risks leading to the systematic under-representation of particular stakeholder groups (cp. Schmitter 2002: 65 66). This possible problem is countered by institutional provisions that seek to correct - where appropriate - these imbalances. The degree of political equality in a governance arrangement is measurable, for example, by asking to what extent all actors possess equal status as members, have equal access to information, as well as equal voting rights (cp. Saward 1994). In addition, it must be considered to what extent governance arrangements guarantee these freedoms and rights in practice and make resources available to those groups that cannot afford to participant in relevant meetings (cp. Steffek Nanz 2008). Accordingly, in the context of this study, a high degree of equality is assumed when all stakeholders have the same participation rights and when stakeholder-groups with weak resources receive financial support that enable them to participate in meetings on an international level; a middle degree is given when rights are equal but no resources for poor stakeholder-groups are available; and a low degree of equality is given when different stakeholder groups are endowed with graduated participation rights. Consensus-orientation in the negotiation-process Another possibility to strengthen input-legitimacy is through the orientation of the processes of policy formulation and standard-setting at the principle of the unanimous vote. 10 Processes of consensus-building are based on the exchange of knowledge and arguments, which implies an upgrading of structurally disadvantaged actors with only weak resources. The aim of each actor in these processes is to convince the opposing parties of the superiority of its own argument and to identify overlapping preferences. The stronger the interests and functional logics of the participating actors diverge, the more the governance arrangements need to look for an encompassing consensus to generate acceptance on the side of the rule addressees and to ensure the continuity of the arrangement (cp. Jachtenfuchs Kohler-Koch 2004: 88). Based on this assumption, the more stakeholders that agree to a certain policy in a deliberative process the stronger the perception of legitimacy. Correspondingly, in what follows, I differentiate between a high degree of consensus-orientation characterized by multi-level adherence to the principle of unanimous vote, including even external actors; a middle degree of consensus-orientation characterized by covering only the organizational level; and a low degree of consensus-orientation marked by restricting the principle of the unanimous vote to only certain bodies of the organization. 10 Input-oriented arguments often rely simultaneously on the rhetoric of participation and of consensus (Scharpf 1999: 7).

6 Legitimacy in Global Governance 225 Transparency of the Decision-making Processes In the throughput-dimension, the accountability of decision-makers is the main characteristic of legitimate governance (cp. Scholte 2011: 15; Chan Pattberg 2008). 11 As a precondition for ensuring accountability, the transparency of the decision-making processes must be guaranteed (cp. Scholte 2011: 16; Gupta 2008; Buchanan Keohane 2006: 426). This assumes that only an actor who is informed about all the stages of the decision-making process is able to put forth its concerns, to comprehend its integration in the decision-making process, to exert its monitoring function and, if necessary, its veto rights. Transparency implies the unlimited and timely access of respective stakeholders to all available information at a reasonable cost (cp. Keohane 2011: 102) and the approval of structures and decision-making processes by observers (One World Trust 2007: 11; Buchanan Keohane 2006: 428). The transparency of a governance arrangement improves to the same extent that relevant information is made available to the stakeholders, ensuring debates about rules and decisions are publicly available and decisions are attributable to decision-makers (cp. Mason 2005; Florini 2008). With reference to this conceptualization, a high degree of transparency is clear when structures and processes of a governance arrangement enable internal as well as external actors to identify the persons responsible for specific decisions, when reporting commitments for rule-addressees and a codified information policy, in which observance is enforceable, exists. A middle degree of transparency is evident when structures and procedures are transparent but internal decisions are not, when reporting commitments for ruleaddressees are not encompassing and when there is no codified information policy and no chance to file a suit in cases of withholding of information; and a low degree of transparency can be recognized by structures and procedures transparent only to members of the governance arrangement, the restriction to voluntary reporting commitments for ruleaddressees and the lack of a codified information policy. Monitoring mechanisms The throughput-dimension of legitimate global governance is also improved by mechanisms that ensure the control of an organisation s bodies and the practical implementation of its rules and standards (cp. Scholte 2011: 17; Neyer 2004: 119). Such mechanisms compel these actors to inform, explain and justify their actions and in so doing, add to their accountability (Bovens 2007). On that note, monitoring procedures encourage decision-makers to anticipate the reactions of the stakeholders and through this create a connection between rule-makers and rule-addressees (Benz Papadopoulos 2006: 275). The monitoring mechanisms have to be clearly structured, adopted to the specific circumstances and provided with the necessary resources to adequately fulfil their functions. It is of additional importance that the detection of infringements is carried out by independent authorities and that this process cannot be influenced by affected actors (cp. Abbott et al. 2000: 415). Hence, in the following, it will be differentiated between a high quality of monitoring characterized by external control of the organizations practices and encompassing detailed and independent (third party-)monitoring of rule-addressees including onsite-inspections and adequate resources for their execution; an average qual- 11 The One World Trust defines accountability as the process through which an organisation makes a commitment to respond to and balance the needs of stakeholders in its decision-making processes and activities, and delivers against this commitment (2007:11).

7 226 Ingo Take ity of monitoring featured by internal control of the organizations practices and a detailed monitoring system for rule-addressees organized by the organization itself (second-party monitoring), including announced onsite-inspections but restricted by a lack of resources; and a low quality of monitoring cognizable by internal self-control of the organization as well as its rule-addressees (first-party-monitoring) and no onsite-inspections. Sanctioning Mechanisms Finally, in order to safeguard accountability, global governance arrangements should not only contain mechanisms for authoritative rule interpretation (e.g. independent arbitration courts) and the detection of infringements, but also institutionalised procedures for the handling of rule-breakers (Keohane 2011: 102; Buchanan Keohane 2006: 426; Zu rn 2005). These procedures can comprise for example a loss of reputation, regulation or the exclusion from the organization (for a critical perspective see Chayes Chayes 1995: 86). Therefore adequate complaint procedures and suits are required. In the handling of rule infractions, claimants have to be informed about the proceedings and have to be entitled to file suit. Additionally, sanctioning mechanisms should be initiated by non-contract parties (Keohane Moravcsik Slaughter 2000). When examining sanctioning, examining the number of proceedings initiated as well as their outcomes and any resulting changes in behaviour is instructive. In this study, a high capability for sanctioning is evident when detailed complaint procedures open to all stakeholders exist, an independent dispute settlement body is available and substantial punishments are possible, supplemented by the willingness and capacity to make use of them; 12 I consider an average capacity for sanctioning when complaint procedures only exist for rule-addressees and not the organization itself, when the access to these procedures is not open to external stakeholders, when an independent dispute settlement body is available but does not have the capacity for substantial punishment and when there is only a weak willingness to apply sanctions at all; a low capacity of sanctioning can be recognized by complaint procedures applying only to rule-addressees and available only to members and when the governance arrangement abstains from sanctions. Accord with universal principles (content-related legitimacy) The congruence of the aims of governance arrangements with universal principles should serve as a further indicator of normative legitimacy (cp. Franck 1990; Bernstein Cashore 2004: 34; Kratochwil 2006: 303). The assumption in this context is that the acceptance of global governance arrangements increases to the extent to which its aims, principles and norms are compatible with universal principles (Bernstein 2011). Decisions caused by collective actors meet acceptance only when rule-addressees are convinced that the rule-makers are not aiming only for the realization of their own interests, but feel obliged to promote general welfare and common goals. This is measured by the degree of congruence of the aims and statutes of a governance arrangement with the relevant codified universal principles and norms. With respect to this study, a high degree of congruence is achieved when there is close consistency between the principles of a governance arrangement and universal 12 High capacities on the side of governance arrangements helping its rule-addressees comply with the rules could be an alternative way to handle infringements but, as will be shown in the following, sufficient capacities are mostly non-existent.

8 Legitimacy in Global Governance 227 principles; an average degree of congruence is attained when there is a high correspondence with regional principles; and a low degree of congruence exists when conformity with universal or regional principles is weak and when a governance arrangement is restricted exclusively to functional aims. Expertise (substantive legitimacy) A governance arrangement can generate legitimacy by its capacities for problem-solving, as, for example, Majone stated:,,substantive legitimacy depends on such features as the expertise and problem-solving capacity of the regulators, ( ) (1998: 21). In the context of international negotiations, it has long been observed that actors accredited with a high degree of problem-specific expertise enjoy an accordingly high reputation among all negotiators as well as beyond them (Cutler Haufler Porter 1999: 18). To the degree that standards and political decisions reflect the state-of-the-art of scientific-technical knowledge and the (practical) know-how of internal and external stakeholders, the acceptance of these decisions as adequate and legitimate will rise (Cohen Sabel 1997; Arts Kerwer 2007: 152). The expertise, professionalism and the bias to cooperation of NGOs that grew during recent years (Kissling Steffek 2008; Walker Thompson 2008; Take 2002) often fuel these expectations and potentially increase the societal acceptance notably of transnational and private governance arrangements (Scholte 2011; Steffek Kissling Nanz 2008; Schmidt Take 2000: 174). In this study, high expertise is assumed when a governance arrangement is able to conduct an encompassing information policy and when it can thereby resort to scientific findings, technical know-how as well as external knowledge; an average expertise exists when an organization strives toward informing the public, exhibits know-how and participates in scientific research or cooperates with epistemic communities; a low expertise is characterized by the lack of knowledge on site and a lack of effort to promote and coordinate the diffusion of knowledge and know-how between the ruleaddressees. Material resources (substantive legitimacy) The capacity of problem-solving of global governance arrangements also depends on their ability to acquire material resources. Management-theories understand non-compliance mostly as a problem of involuntary breaches of rules (Chayes Chayes 1995). According to that, states and non-state actors do not lack the will but capacities to comply with the rules (cp. Zu rn 2005: 29 30). To measure the potential contribution of a governance arrangement to problem-solving, the suitability and adequacy of existing resources of the respective institution resulting from its capital and its personnel must be accounted for. The assumption here is that the acceptance of a governance arrangement increases to the extent it is able to generate as many capacities as possible or to collect them from stakeholders and in so doing enhance its problem-solving capacity and that of its rule-addressees (cp. Mitchell 1996; Underdal 1998). Accordingly, a high degree of problem-solving capacity is given when a governance arrangement has adequate personal, financial and operational resources at its disposal to fulfil its aims; an average degree is given when an organization has less than adequate resources at its disposal and instead depends on the resources of external actors to accomplish its aims; a low degree of problem-solving capacity is assumed when an organization neither possesses the adequate resources to realize its aims nor is able to recruit them from external actors.

9 228 Ingo Take Acceptance (empirical legitimacy) Output-legitimacy, as conceptualized in this article, is not the fulfilment of targeted goals, (i.e. the solution of common problems) (cp. Scharpf 1999; see also Lo vbrand Rindefja ll Nordqvist 2009), 13 but instead the degree to which a governance arrangement succeeds in generating acceptance by its internal and external stakeholders and by this motivating them to follow its rules (Meidinger 2008: 287; Neyer 2004: 101; Bernstein Cashore 2004). 14 This concerns situations in which a group of stakeholders is subject to rules and decisions by an authority that claims to have a right to be obeyed, and actors intersubjectively hold the belief that the claim is justified and appropriate (Bernstein 2011: 19; see also Lake 2010; Biersteker Hall 2002). An institution holds authority when the rule-addressees acknowledge that there are content-independent reasons to comply with its rules (Buchanan Keohane 2006). For Thomas Franck, the most basic indicator of a rule s legitimacy [is] whether it is validated by community (Franck 1990: 198; see also Bernstein 2011: 19; Kratochwil 2006: 302; Suchman 1995: 574). 15 When applying this conception of output-legitimacy, two aspects need to be differentiated: First, the acceptance of a rule that manifests itself in statements of the rule addressees (approval, constructive critique, critique, denial); 16 and secondly, the acceptance of a rule that manifests itself in the number of stakeholders which voluntarily subordinate themselves to a governance arrangement (Beetham 1991: 92). In the following, I will therefore differentiate between a high degree of acceptance that manifests itself in the positive statements of a plurality of different internal and external stakeholder-groups, in a fast-growing number of actors which submit to the rules voluntarily and the substantial measures taken by these rule-addressees to implement those rules; an average degree of acceptance is expressed in positive statements by mainly internal and a few external actors, a slow but growing number of actors which submit to the rules voluntarily and a restriction on low-cost opt-outs taken by rule-addressees; and finally a low degree of acceptance is demonstrated by positive statements coming only from internal stakeholders, a constant or even decreasing number of actors which submit to the rules voluntarily and insufficient measures of rule-implementation. Table 1 lists the indicators for the appraisal of legitimate governance beyond the nation-state: 13 One reason is that the causal chain between the performance of a governance arrangement and the state of a policy field is difficult to establish. Another reason is that it is not possible to designate what a solution of a common problem is and when it serves the common good. Thirdly, solutions have to be viewed in relationship to the preferences of stakeholders. In this respect, solutions are effective and fair only when they are accepted by the stakeholders. To know the preferences of the stakeholders we need an appropriate input. Thus, there is no output legitimacy without input legitimacy (Abromeit 2002: 19). 14 It is important to differentiate between indicators of normative legitimacy on the one hand and acceptance as an indicator of empirical legitimacy on the other. The reason is that 1) the consent of rule-addressees alone can be completely detached from the normative legitimacy of a governance arrangement (for example in authoritarian regimes) and 2) normative legitimacy does not automatically generate acceptance on the side of the rule-addressees (for example if the governance arrangement is unable to fulfill its aims). Therefore, acceptance does not equate with legitimacy but only one element of it. 15 In this context, it has to be considered that stakeholders can differ greatly with respect to their cultural imprints as well as their social, political, economical and ecological background (cp. Scholte 2011: 113; Bernstein 2011). 16 This form of acceptance of the governance arrangements were captured on the basis of statements by the rule addressees and expert opinions as expressed in interviews with the author.

10 Legitimacy in Global Governance 229 Table 1: Legitimacy Dimensions of Governance Beyond the Nation-State Normative Legitimacy Empirical Legitimacy Input-Dimension Throughput- Dimension Content-related legitimacy Substantive Legitimacy Output- Dimension Representativeness Level of political equality Consensus-orientation Accountability consisting of transparency, monitoring procedures and sanction mechanisms Conformity of the goals with universal principles Expertise and problem-solving resources Acceptance (measured by statements and voluntary participation) Selection of cases Following the above formulation, the different dimensions of legitimacy and their selected indicators must be applied to already observable forms of global governance. Only by such a procedure can theory-building regarding necessary conditions of legitimate governance beyond the nation state advance. Therefore, the following section briefly sketches the criteria for the selection of cases and names the nine case studies from which the empirical data was gathered. The selection of cases is orientated towards the research question and the chosen research design. A first criterion for the selection of cases is the degree of globalization of the issue in question and the corresponding demand for cross-border regulation. Secondly, the object of regulation has to be characterized by a problematic situation, that is, it has to exhibit problems of co-operation and coordination. Thirdly, international, transnational as well as private forms of institutionalized political steering have to be detected. Finally, the selection of cases should be systematized by the differentiation of policy-types according to Lowi (1972), which separates constitutive (creation of general conditions), regulative (steering of actors performance by bans and rules), and redistributive rules (redistribution of resources). Because redistributive rules are rarely observable on the global level, due to their depth of engagement and the poor sense of community, they were not examined in this study. Instead, constitutive rules, observable in the case of internet governance, and regulative forms of governance, in the fields of social and environmental standards, were taken into consideration. The latter are observable in two forms; standardsetting (e.g. in the area of industrial relations) and certification (e.g. in the area of forest protection). Standards are agreed criteria established by different institutions which form the basis on which the technical or physical characteristics of products (product standards) and the processes of their manufacture and transport (process standards) are assessed according to their social and environmental compatibility. Systems of regulation more orientated towards market mechanisms are certification schemes that edit labels. Certifications are procedures by which the compliance of certain standards by companies, production processes, products or services is monitored by an independent accredited certification organization. The label confirms that the company is operating in accordance with the standards. This comparative study investigates the activities of the International Telecommunication Union (ITU), the transnational World Summit on the Information Society (WSIS) and the private Internet Corporation for Assigned Names and Numbers (ICANN) which are geared toward the improvement of internet governance in the field of the domain names system.

11 230 Ingo Take Table 2: A selection of attributes of the organizations under investigation Issue-area Members Date of Incorporation Rule-Authority Outcome ILO Child-labor States, social 1919 Ratification by Resolutions partners member-states RugMark Child-labor Multi-stakeholder 1995 Binding for members Label SAI Child-labor Private actors 1997 Binding for members Certificate PEFC Forestry State actors 1999 Binding for members Label ISO Forestry Public and private Non-binding Certificate ISO corporations FSC Forestry Private actors 1993 Binding for members Label ITU Internet States Ratification by Standards member-states WSIS Internet Multi-stakeholder 2001 Non-binding Declaration ICANN Internet Private actors 1998 Binding for customers Standards Furthermore, the standards setting activities of the International Labour Organization (ILO), the transnational RugMark-Initiative and the private Social Accountability Initiative (SAI) aiming at the elimination of child labour were analysed comparatively. Finally, the certification of sustainable forestry by the international Programme for the Endorsement of Forest Certification Schemes (PEFC), the transnationally organized International Standards Organization (ISO), with its environmental management standard ISO 14000, and the private certification scheme Forest Stewardship Council (FSC) were also investigated. Table 2 lists some attributes of the organizations under investigation. The empirical data The selected global governance arrangements will now be analyzed through the particular dimensions of legitimacy identified in the second section above. I will examine how closely the governance arrangements match the indicators outlined in a particular dimension of legitimacy. The normative legitimacy of these arrangements is appraised by utilizing legal texts (e.g., constitutions, by-laws) and secondary material. I investigated the empirical legitimacy of governance arrangements through conducting qualitative interviews, inter alia, with experts, representatives of these institutions and their addressees (see Annex). 17 The following analysis and interpretation of data is restricted to illustrative sketches because a comprehensive conceptual, methodological and contextual discussion would go beyond the scope of this article. High informal but selective formal representativeness Membership in the three governance arrangements dealing with the formulation of social standards is based on representing a number of selected societal groups (rule addressees) 17 About 2000 questionnaires were sent to the nine organisations under investigation, their rule-addressees and selected experts (derived from the literature). In addition, about 90 interviews were conducted. Altogether we were able to evaluate 158 questionnaires (for a list of interview-partners see Annex 1).

12 Legitimacy in Global Governance 231 allowing external stakeholders only limited participation rights. At the ILO, with its tripartite structure encompassing delegates of governments, employer s and worker representatives of the member states, NGOs participate only during certain policy phases (ILO- Constitution, Art. 3 and 5; At RugMark, stakeholder groups are allowed to participate but processes are dominated by certain companies and NGOs (cp. BMZ 2003). As a private organization, SAI neglects state actors in particular (Social Accountability International 2008). All three governance arrangements are dominated by several stakeholder groups with formal participation rights, while restricting others to informal ways of participation, whereas actors from developing countries in particular have to fight against marginalization (with the exception of RugMark, which is mainly located in India, Pakistan and Nepal). In a nutshell, organizations in this policy field can be credited only with a medium degree of representativeness. In the policy field of sustainable forestry, the FSC stands out quite clearly with its activities towards a balanced representation of ecological, social and economical interests as well as including actors from the South and the North (FSC 2005: Statutes and By-laws, 2005, Bonn: FSC). Although state actors by nature play a subordinated role in this private governance arrangement, the FSC has to be credited with a comparatively high degree of representativeness. In contrast, the PEFC neglects the representation of ecological and social groups as well as of actors from the South on both the national and the international level (PEFC Statutes; The same holds for ISO 14001, where actors from the South, as well as non-commercial groups, are marginalized (cp. Murphy Yates 2009: 25). In other words, the PEFC and ISO reach only a medium degree of representativeness whereas the FSC reaches a high degree. In the ITU, member states enjoy the full range of formal participation rights while private companies enjoy some of them if they are represented as sector members. Civil society actors are not represented in any way in the ITU ( Work in the organization is dominated by the developed countries and the big IT-companies. At the WSIS, only governments own a formal member status. However, civil society actors and private companies as well as international organizations are credited with special functions and they are allowed to participate in all stages of the process (WSIS: Declaration of Principles, B 48), although the invitation procedure is selective, non-transparent and neglects actors from the South. At ICANN, actors from the South are also underrepresented. Apart from that, panels are open to all stakeholder groups and they are allowed to participate actively in the work of the organization, although mostly only in an advisory function. While the role of governments increased in the course of time, the representation of the individual users decreased in the panels of ICANN ( bylaws.htm; see also: Hofmann 2006). In summary, ICANN and the WSIS do not represent all stakeholder groups in formal way, but informally they try to reach a higher degree of representativeness whereas the ITU systematically excludes civil society actors and thus ranks low in representativeness. Predominant graduated participation rights instead of equal participation The organizations active in the policy field of child labour award the different rule addressees and stakeholders with graduated participation rights. In the ILO, governments have two votes, the social partners one each, and all further stakeholder groups possess only weaker graduated participation rights but no voting rights (ILO-Constitution, Art. 3 and 5). RugMark and SAI also grant graduated participation rights to the different stakeholder

13 232 Ingo Take groups whereupon companies and NGOs enjoy more rights than state-actors do (BMZ 2003; SAI 2008). None of the three governance arrangements show the will or capacity to provide marginalized rule-addressees with adequate resources to enable them to exercise their rights. Altogether, the equal participation of stakeholders is realized in each case to a medium degree. In the area of sustainable forestry, the assessment of the three governance arrangements diverges to a greater degree. The PEFC grants non-state members participation but weights its voting rights on the international and national level in a way that economic stakeholder groups can always outvote representatives of ecological or social interests. Furthermore, actors from the South with only weak resources get no material support (FERN 2001: 25). Within ISO 14001, member groups possess graduated rights which are linked to their individual status, while on the level of the Technical Committees, members can decide their status by themselves. In addition, modest efforts are under way to support marginalized members from the South and civil society stakeholders. Apart from that, the members of the developed countries and economic representatives dominate (cp. Murphy Yates 2009: 45). In contrast, the FSC arranges for the widely equal participation of economic, ecological and social stakeholders from the North and the South in the context of its structures and processes. Because of its scarce resources, it does not succeed in complete equality, especially on the procedural level (Dingwerth 2008). However, it can be stated that the FSC formally best matches the criterion of equal participation of all stakeholders. In the ITU, graduated participation rights apply for member states, the sector members as well as partner organizations. Whereas private business actors can acquire participation rights via their status as sector members, civil society actors are excluded from all forms of formal participation (ITU Constitution, Art. 3, 4). The ITU makes almost no effort to enable representatives from the developing countries, as well as the small and medium sized companies, equal participation. At the WSIS, only states possessed all formal participation rights whereas all other stakeholder groups were awarded only with observer status (WSIS Declaration of Principles). On the informal level, interested organizations were granted the opportunity to bring in their needs in numerous preparatory meetings. In so doing, the asymmetrical distribution of resources between private and civil society actors on the one hand, and between actors from developed countries and those from developing countries on the other, became noticeable. This mismatch could not be balanced by the funds provided by the WSIS (Dany 2008: 66). At ICANN, only the organizations bound by contract possess decision-making authority whereas states, epistemic communities or individual users were granted only an advisory function. Yet, by its functionally differentiated organizational structure, ICANN strives for a geographically balanced and active participation of the individual stakeholder groups in their areas of concern (ICANN By-laws). ICANN offers no financial means for the strengthening of actors with weak resources (ICANN: Annual Report : 11). Whereas the WSIS and ICANN at least strive for an equal participation of all stakeholders, this applies for ITU to a much lesser degree. Consensus-orientation without coercion to unanimity The consensus-orientation is significant for each of the three organizations in the social sector. For example, all organizations employ multi-stage deliberative processes for consensus finding which allows even external stakeholders to participate in different forms on different levels. In official bodies, polls take place as well but the participating actors strive for decision-making by consensus (see the by-laws and procedures of the three organizations). The

14 Legitimacy in Global Governance 233 three governance arrangements in the field of environmental standards also show a high degree of consensus orientation on all levels. A two-thirds majority of the delivered votes are necessary for the adoption of a decision (cp. The by-laws of the FSC ( 15), the PEFC (Guideline ) and ISO (ISO IEC Directives, Part, 1,5). In the field of internetregulation, the three organizations also prefer decisions by consensus, but the way of realizing them differs and is restricted to the organizational level. The ITU aims at a broad approval of decisions in consensus-building processes of expert panels (more than a twothirds majority, cp. ITU-Convention, Art. 32B). The WSIS gained approval of all its member states due to its policy of the lowest common denominator and the postponement of disputed decisions. Simultaneously, civil society took this for a motive to pass an alternative Summit Declaration. ICANN strives - according to its own code of conduct for consensus but does not make it necessarily a basic principle of decision-making (cp. ICANN By-laws (2006): Art. IX,1). Transparency particularly on the side of the rule-addressees Of the governance arrangements active in the field of child labour the ILO alone shows a high degree of transparency. The ILO is the only organization that has a codified information policy available and informs not only on its organizational structure, its statutes and by-laws, but also about its decision-making processes whereby stakeholders are encouraged to identify the responsible persons (cp. ILO Constitution). In addition, its rule-addressees are subject to a reporting commitment concerning their implementation of international standards (cp. IAO-database NATLEX: RugMark und SAI oblige their rule-addressees to have transparency (see the websites of RugMark and SA8000, IV, 9.12), but do not provide information about internal procedures and have no codified information policy at their disposal. Therefore, their degree of transparency is rated comparatively weak. In the field of sustainable forestry, the FSC shows the highest degree of transparency by informing the public about its standard-setting-, accreditation- and certification procedures and granting external stakeholders the possibility of direct observation (cp. Moreover, the FSC obliges its rule-addressees to have transparent chains of custody as well as forestry practices (FSC National Initiatives Manual 1998: xxix). The PEFC likewise offers considerable information about its procedures of standard-setting, accreditation and certification (see but does not provide external stakeholders with the chance of participatory observation and asks its rule-addressees for the disclosure of much less information than the ambitious rules of the FSC (PEFC Technical Document, Common Elements and Requirements of PEFC, clause 4.3). The standardsetting processes inside the ISO are also less transparent than that of the FSC. Only NGOs with liaison-status receive all available information, whereas external stakeholders have to pay dues (ISO IEC Information Center: Moreover, companies that submitted to the ISO standard do not have to inform the public to what extent they conform to the standards (cp. Murphy Yates 2009). Thus, PEFC and ISO show only a medium degree of transparency, while the FSC shows a high degree. In the field of internet regulation, all three governance arrangements use the Internet to inform the public about organizational processes and its results. However, only ICANN possesses a codified information policy that allows for complaint procedures for cases in which documents were repressed without legal cause (By-laws Art. III; Chango 2011). Fur-