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



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.


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. Mügge 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 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 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-, throughput- and 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 analysis 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 throughput-dimension addresses the procedural level (Zürn 2000) focusing on to what extent decision-making 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; Zürn 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 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.

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 rule-addressees 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 rule-addressees 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 quality 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; Zürn 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 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 rule-addressees.

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. Zürn 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.

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 Lövbrand/Rindefjäll/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:

Table 1.   Legitimacy Dimensions of Governance Beyond the Nation-State
Normative LegitimacyEmpirical Legitimacy
Input-DimensionThroughput-DimensionContent-related legitimacySubstantive LegitimacyOutput-Dimension/
RepresentativenessAccountability consisting of transparency, monitoring procedures and sanction mechanismsConformity of the goals with universal principlesExpertise and problem-solving resourcesAcceptance (measured by statements and voluntary participation)
Level of political equality

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; standard-setting (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. 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.

Table 2.   A selection of attributes of the organizations under investigation
 Issue-areaMembersDate of IncorporationRule-AuthorityOutcome
ILOChild-laborStates, social partners1919Ratification by member-statesResolutions
RugMarkChild-laborMulti-stakeholder1995Binding for membersLabel
SAIChild-laborPrivate actors1997Binding for membersCertificate
PEFCForestryState actors1999Binding for membersLabel
ISO/ ISO 14000ForestryPublic and private corporations     1947/1996Non-bindingCertificate
FSCForestryPrivate actors1993Binding for membersLabel
ITUInternetStates     1865/1934Ratification by member-statesStandards
ICANNInternetPrivate actors1998Binding for customersStandards

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) 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; http://www.ilo.org/ilolex/english/constq.htm). 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; http://www.pefc.org/resources). 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 140001 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 (http://www.itu.int/net/about/legal.aspx). 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 (http://www.icann.org/en/general/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 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 2005–2006: 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 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 5/2006) and ISO 14001 (ISO/IEC Directives, Part, 1,5). In the field of internet-regulation, 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 two-thirds 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: http://natlex.ilo.org). 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. http://www.fsc.org). 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 http://www.pefc.org) 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 standard-setting 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: http://www.standardsinfo.net/info/index.html). Moreover, companies that submitted to the ISO 14001-standard do not have to inform the public to what extent they conform to the standards (cp. Murphy/Yates 2009). Thus, PEFC and ISO 14001 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). Furthermore, ICANN admits interested stakeholders to all meetings (aside from the ones of the Government Advisory Council, GAC). This is in contrast to the WSIS-process as well as the ITU, where information is more restricted, thus crediting these two organizations with only a medium degree of transparency.

Monitoring procedures lack capacities

All three governance arrangements for the regulation of child labour come with detailed and independent monitoring systems for their rule-addressees, but permit little external control of their own practices.18 The rule-addressees of the ILO must report on its activities which are additionally supplemented by monitoring inspectors supported by the ILO (ILO-Constitution, Art. 19, 22, 35). However, due to limited capacities, the ILO must rely on the control of social partners as well as the self-control of the member states (One World Trust 2006: Global Accountability Report, London). RugMark also relies on its own inspectors when it comes to the monitoring of the rule-addressees. The inspectors conduct on site inspections supplemented by the mutual control of the license-holders as well as self-control (cp. BMZ 2003: 67f.). SAI outsourced its monitoring function to an independent organization (SAAS) that accredits and monitors independent certification agencies which in turn certify and control the rule-addressees on the ground. The rule-addressees of SAI must report on its activities like the ILO (cp. SA8000, IV, 9.13). The monitoring capacities of RugMark in particular are insufficient for a regular and comprehensive control of the rule-addressees (Betz 2000: 12). At the same time, none of the three governance arrangements provides external actors with incentives or resources for supplementary monitoring activities. Given that the governance arrangements show only a marginal willingness to submit themselves to monitoring mechanisms,19 monitoring is rated only average.

Also, the three organizations active in the field of sustainable forestry have to some extent detailed procedures at their disposal for the monitoring of their rule-addressees by independent third parties. These actors in turn are accredited by independent bodies according to well-defined procedures. Only the FSC commits its accredited certification agencies to conduct surveys on site (cp. FSC Accreditation Manual, Section 15). In contrast, according to the rules of ISO 14001, even self-certifications of environmental management systems are admissible, leaving the decision up to the rule-addressees (Haufler 2002: 24). The standard setting by ISO 14001 as well as the PEFC is not monitored by external actors but by the general assembly in cooperation with the elected council (ISO/IEC Guide 61 and 62; PEFC Accreditation and Certification Procedures). In contrast, the FSC stands out by its practice of allowing not only members but also external stakeholders as well as the public to call the diverse bodies of the organization to account if acting against the principles and statutes of the FSC (cp. FSC-Bylaws). Against this background the monitoring mechanisms of the FSC are rated as highly effective whereas those of the PEFC are classified as average and those of the ISO 14001 as only marginally effective.

Concerning the governance arrangements in the field of internet regulation, the effectiveness of their monitoring systems also differs. In the case of the ITU, the Joint Inspection Unit of the UN functions as an independent monitoring and evaluation authority, however with restricted capacities and competences (see Resolutions 57, 107). Apart from that, the member states – as direct rule-addressees – independently ensure that the activities of their national sector-members and partner organizations are in line with the provisions of the ITU (ITU Constitution, Art. 6.2). The WSIS abstained from the establishment of monitoring mechanisms completely due to the declaratory character of its resolutions, but nevertheless allows for UN evaluation of the implementation of its action plans (Tunis Agenda, § 111–120). In contrast, ICANN allows for the control of the Council as well its staff in multiple ways (By-laws, Art. IV; see also Accountability and Transparency Frameworks and Principles).20 There are complaint procedures available for individuals and entities that feel affected in substance by the activities of the organization (By-laws, Art IV, 2). Furthermore, the by-laws contain provisions for the independent monitoring of the activities of the Council by a third-party (Art. IV, 3). Complementarily, there is an ombudsman, a couple of evaluation procedures as well as the external control of ICANNs’ activities by NGOs (Art. V; Chango 2011: 277). The monitoring of the rule-addressees ranks low in the context of the allocation of domain names, because cases in which domain names were illegally acquired are revealed usually by the owners of brand names and dealt with in the context of the Uniform Domain-Name Dispute-Resolution Policy (UDRP). All in all, the effectiveness of the monitoring mechanisms of ICANN is rather high in comparison to the ITU while the WSIS has no such mechanisms available at all.

Sanctioning mechanisms: to help rather than punish

The governance arrangements dealing with the problem of child labour have differentiated and institutionalized sanctioning mechanisms at their disposal. For example, the comparatively detailed complaints procedures and proceedings of the ILO comprise not only corrective actions and their review but also the possibility of appeals against their results. As the last resort, it is possible to go to the International Court of Justice as an independent body, although this has not yet occurred (ILO Constitution, Art. 20–33). Sanctioning mechanisms are open to all member organizations though external stakeholders have only restricted access (Art. 26). Besides, they are launched only rarely and mostly result in public shaming and to a lesser extent in substantial punishments (http://www.ilo.org/ilolex/gbe/ceacr2008.htm).21 RugMark and SAI also rely on corrective actions, shaming and where appropriate on the suspension or the withdrawal of certifications if norms are violated continually. In addition, RugMark and SAI oblige their rule-addressees to maintain complaints procedures by themselves (cp. BMZ 2003; SA8000, IV, 9). At RugMark, sanctioning procedures are obviously applied only selectively and are not codified. This differs from the practice of SAI, which also stands out against RugMark due to its delegation of complaint procedures to an independent organization (SAAS). Furthermore, SAI allows stakeholders to hold not only companies responsible in the case of breaches of rule, but also the certification agencies accredited by SAAS and even SAAS itself (SAAS Global Procedures Guideline 304, 2008). Admittedly, there were only a few cases so far in which sanctions were actually applied (http://www.saasaccreditation.org/complaintlist.htm). In comparison, the ILO and SAI - in contrast to RugMark – can resort to binding and institutionalized complaint procedures. The capacity and willingness to apply sanctions seems to be weak in all three governance arrangements.

The organizations dealing with environmental standards for forestry also differ with regard to the effectiveness of their sanctioning mechanisms. The PEFC can suspend or even exclude its members in cases of rule violations but there is no way to hold the bodies of the PEFC itself responsible by sanctions (cp. Technical Document, Annex 7, 6.3). With regard to ISO 14001, neither the rule-addressees nor member organizations or the organization itself can be held responsible for breaching the rules. In contrast, the FSC has available not only procedures for the sanctioning of rule-addressees and accredited certification agencies but also - at least formally - internal mechanisms for the sanctioning of member organizations and members of the council (FSC By-laws, para 13). All in all, the three governance arrangements have complaint mechanisms available respectively prescribe their members as well as their rule-addressees to have them at their disposal. Nevertheless, only the FSC enables even non-members (under certain circumstances) to initiate sanctioning procedures. Its sanctioning mechanisms are thus the strongest in terms of institutionalization and differentiation whereas the PEFC shows only a medium degree of institutionalization and ISO 14001 has no noteworthy sanctioning mechanisms established at all.

The ITU and the WSIS can also barely be held responsible for their activities. ICANN differs in that it can at least hypothetically be relieved of its duties by the US Department of Commerce if the organization falls short of its contractual requirements (cp. ICANN By-laws, Art. 2.2 and 2.3). Since the Affirmation of Commitments in September 2009, this possibility is even more restricted.22 Differences between the three governance arrangements exist rather with regard to the possibilities of sanctioning their respective rule-addressees. The ITU offers a dispute settlement process that can be used only by members (ITU Constitution, Art. 56; ITU Convention, Art. 41). Anything beyond a denunciation of breaches of the rules by shaming is only possible when members do not pay their contributions. The WSIS has no sanctioning mechanisms because of its declaratory character. The states themselves are responsible for the implementation of action plans which is why failures rarely become public. In contrast, ICANN offers procedures by which offences against by-laws are punished (removal of accreditations or denial of domain name rights) and differences between rule-addressees are settled by dispute settlement procedures that can result in the denial of domain name rights or the refusal of their adjudication (cp. http://www.icann.org/udrp/). The procedures of ICANN are accessible beyond members, and therefore more open than that of the ITU, but they also allow for substantial sanctions.

High congruence of the rules with universal principles

The governance arrangements in the field of sustainable forestry and those striving for the elimination of child labour each demonstrate a high degree of congruence between their aims and universal principles. Whereas the social standards of the ILO, RugMark and SAI refer in particular to UN-conventions and declarations, the environmental standards of PEFC, ISO 14001 and FSC are geared towards different international and regional agreements and conventions. The PEFC refers to European conventions for sustainable development, the ISO 14001 to WTO-principles and the FSC to the UN-Convention for Sustainable Development. In the field of internet regulation, only the WSIS refers explicitly to universal principles (sustainable development, human rights, Millennium Development Goals), whereas the ITU and ICANN limit themselves to functional aims like interoperability and stability as guiding principles.

More know-how than scientific expertise

The comparison of the expertise of the three governance arrangements in the field of child labour suggests that the ILO have the edge over the other two organizations not only because of its decade-long experiences and its personal and financial strength, but also because of its own scientific work and the broad inclusion of external (practical) know-how and (scientific) expertise. In contrast, the informational capacity of RugMark is disposed primarily for roadshows on a local level, a few activities in the field of adult education and national awareness campaigns in cooperation with other actors. SAI also acts with its know-how as an agency of vocational training and even participates in scientific research, although to a lesser degree than ILO.

The governance arrangements in the field of sustainable forestry also perform different informational activities. The FSC strives for a worldwide diffusion of scientific knowledge and practical know-how, the initiation of learning-processes and raising public awareness. The PEFC expands its informational activities to a lesser degree and focuses instead on a process of mutual learning between the members. Both organizations resort to the know-how of forest-owners and companies of the forest-sector, with the FSC incorporating the local know-how of foresters and environmental NGOs. In the context of ISO 14001, the improvement of environmental management systems is in the sole focus. Therefore forest-related expertise is not an object but is already expected from the rule-addressees themselves. The informational activities of ISO 14001 are therefore restricted to enabling mutual learning and the compilation of manuals for environmental management systems.

With regard to the governance arrangements in the field of internet regulation, a high degree of know-how and expertise can be observed. At ICANN, public procedures of regulation on the one hand, and the institutional inclusion of organizations experienced in regulation as well as resorting to the expertise of states (GAC) and users (ALAC) on the other hand, make for a high degree of expertise. The WSIS has generated a diversified expertise in the context of preliminary meetings and by the establishment of the Working Group on Internet Governance (WGIG), whereas the ITU has access to the expertise of states and the business community by virtue of its institutional structure, but abstains from including the knowledge of civil society experts (cp. the websites of the nine organizations).

Resources are almost nonexistent

With regard to the organizations’ dealing with social standards, the ILO possesses far more personal, financial and operational resources than RugMark and SAI (cp. the annual reports of the organizations). For activities to fight child labour alone, the ILO has financial means that exceed the total budget of RugMark and SAI twentyfold. The same applies for personal capacities (IPEC 2006: 22). RugMark runs some facilities for the rehabilitation of child workers as well as a couple of schools. Admittedly, neither RugMark nor SAI are able to support their rule-addressees with personal or financial means in any noteworthy amount. Therefore, the ILO is the organization with the most resources at hand in this sector, but it also has the biggest target group. Concerning the problem-solving resources of the three governance arrangements in the field of sustainable forestry, it has to be stated that none of them dispose of a budget that allows for making significant material contributions to the support of the problem-solving capacities of their rule-addressees. In addition, none of the three organizations in the field of internet regulation has enough resources to give stakeholders access to the Internet or even attenuate the digital divide between the North and the South (declared goal of the ITU and the WSIS). ICANN and the ITU can at least resort to their own revenues and thus decide autonomously about the use of their financial means inside their organization which is not an option for the WSIS because of its external funding.

Acceptance by several but not all stakeholders

The acceptance of the ILO can be demonstrated first by the numerous positive statements of international organizations, concurring governance arrangements, governments, the social partners, the scientific community and NGOs. The ILO demonstrates acceptance secondly by the dynamics of the ratification process, in particular of Convention 182 (cp. http://webfusion.ilo.org/public/db/standards/normes/appl/index.cfm?lang=EN), and thirdly by the measures taken by the member states in the context of IPEC to fight child labour (IPEC 2006: 37/38). Similar statements about RugMark and SAI are less frequent due to their lack of prominence. But these two organizations also can refer to an increase in the number of actors which submit to their rules voluntarily as well as the number of the products certified by them and their importers (cp. http://www.nepalRugMark.org/achievement.php; http://www.homepages.com.pk/RugMark/RugMark/activities.htm; http://www.RugMarkindia.org/about/facts.htm; http://www.saasaccreditation.org/certfacilitieslist.htm). The standard of SAI, in addition, is developing an ever wider geographical distribution.

The three governance arrangements in the field of environmental standards can each point to a high degree of acceptance. In the case of the FSC, most notably the NGOs and the companies at the end of the chain of custody, credit the organization with legitimacy. A few states from different regions of the world and several international development banks embrace the FSC-standards, therefore demonstrating their acceptance, whereas private forest owners and several forestry companies are sceptical toward the organization. On a material level, the growing diffusion rate and the increasing number of certifications of whole chains of custody point to a high degree of acceptance of the FSC (cp. UNECE/FAO Forest Products Annual Market Review 2007–2008). The PEFC is credited with a high degree of acceptance mainly by private forest owners and companies at the start of the chain of custody as well as by several states whereas NGOs are particularly critical. In addition, its diffusion rate is growing as well as its number of certifications of chains of custody (cp. UNECE/FAO Forest Products Annual Market Review 2007–2008). ISO 14001 gains acceptance primarily among companies, notably among firms from the developing countries but also among global players, several states and the WTO. Critique comes mainly from environmental NGOs. Despite this, the number of ISO 14001-certifications is growing as is its diffusion rate (cp. ISO Survey of Certifications). Obviously, all three governance arrangements were successful in generating the acceptance of at least a number of stakeholder groups whereas critique is expressed by single groups only.

The global governance arrangements from the Internet sector are also credited with acceptance by different stakeholder groups, whereas critique comes mostly from individual groups. The private organization ICANN is viewed as legitimate actor for the regulation of the Internet by Western states in particular. Furthermore, numerous companies, parts of the scientific community and scores of NGOs criticize ICANN on several points but at the same time trust in the ability and willingness of ICANN to reform its structures and processes. In contrast, developing countries are bothered by the private character of ICANN and its location in the US. They prefer the ITU as rule-setter for the Internet infrastructure, because here they have more voice. Admittedly, most of the developed countries till now refuse to confer these competences on the ITU (cp. Drake 2004: 132). External stakeholders criticize the insufficient or even absent consideration of small and medium-sized companies, civil society actors and governmental departments not directly responsible for the telecommunication sector as well as the dominance of powerful states and companies in the ITU (cp. Drezner 2004: 496). In contrast, the majority of states as well as the bigger companies acting in the field of internet regulation demonstrate a high degree of acceptance to the ITU. The WSIS was characterized by its member states and the majority of its external observers as a multi-stakeholder-approach which functions as a role model. Critique came above all from civil society actors, which felt largely marginalized by the WSIS-process and criticized the lack of concrete obligations for the states as well as the renunciation of an effective follow-up process (cp. Ó Siochrú 2004). Table 3 gives an overview of the performance of the nine cases regarding the three dimensions of legitimacy.

Table 3.   Benchmark of the three dimensions of legitimacy according to primary sources and secondary literature
Normative LegitimacyEmpirical Legitimacy
  1. Explanation: ++ (high degree), + (medium degree), 0 (low degree), - (absent/not applicable).



The comparative assessment of the legitimacy of different forms of global governance based on primary sources and secondary analysis on the one hand, and interviews and questionnaires on the other hand, indeed created some important findings about the specific legitimacy of individual governance arrangements. Furthermore, the assessment delivered clues about possible correlations between normative acceptability (“Anerkennungswürdigkeit”) and empirical acceptance. Altogether, the correlations are not clear enough to serve as a starting point for a more parsimonious theory. Rather, they show a high degree of complexity of social phenomena and diverse perceptions among relevant actors. This is an important insight by itself, although not a particularly astonishing one. Furthermore, it does not present a major problem for this study because it did not aim to test theories but instead aimed to generate tentative conclusions about the correlation between different forms of global governance and their normative acceptability and the connectivity between acceptability and factual acceptance. In addition, the potential legitimacy of governance arrangements and important indicators of legitimate governance must be identified and some general insights about the preconditions, possibilities and limits of legitimate governance beyond the nation-state should be detected. To this end, the insights gained by this study can definitely make a contribution and therefore will be presented below.

  • 1 As a start, none of the nine empirical case studies of the three policy fields show that international governance in principle can claim a higher degree of legitimacy than transnational or private forms of governance. Such an assumption has numerous adherents, especially in the scientific community and even more in politics, but can be confirmed only in the case of the ILO, whereas private governance arrangements like ICANN and FSC are convincing counter-examples. For instance, the private governance arrangement of FSC, due to its organizational structure and the design of its procedures, exhibits the highest (normative) acceptability of all organizations under investigation. But this does not mean that private governance arrangements are in principle conceptualized to show a higher degree of acceptability than, for example, transnational or international governance arrangements. Thus, the ILO performs better in the overall comparison of acceptability across policy fields than the private governance arrangements of SAI and ICANN. ITU and the WSIS also show a low degree of normative acceptability. According to this, a systematic correlation between the type of a governance arrangement (international, transnational, private) and the degree of its acceptability does not seem to exist.
  • 2 The supposed correlation between (normative) acceptability and factual acceptance appears most clearly in the cases of the ILO and the FSC, and to a lesser degree ICANN. In addition, the findings about RugMark and SAI support the aforementioned interrelation in the same way, namely, a medium degree of acceptability correlates with a medium degree of acceptance. As for the PEFC and ISO 14001, a medium degree of acceptability correlates with a high degree of acceptance. According to the interviews, the high degree of acceptance could be caused by the flexibility of the standards that grant the rule-addressees a comparatively large freedom of action. This applies in particular to the process standard of ISO 14001. In addition, the ITU and the WSIS provide no confirmation of the assumed correlation between normative legitimacy and acceptance. Their acceptability is rather low, whereas their acceptance is high. In the context of the conceptualization of legitimate governance at hand, this is explainable only by the high degree of expertise that both governance arrangements show. Because they both aim at a constitutive function, namely the regulation of a policy field (internet infrastructure), it can be argued that governance arrangements of this kind have to rely more strongly on expertise than on (normative) acceptability. Thus, a systematic correlation between the acceptability of a governance arrangement and the degree of its acceptance does not seem to exist.
  • 3 The indicator-based comparison across policy fields enables the identification of the potential legitimacy of individual governance arrangements. Thus, the analysis of the primary and secondary sources shows that the FSC, with its individual organizational structures and processes, performs best in the overall comparison of acceptability. According to this, the other organizations could raise their input legitimacy by a more formalized inclusion of till now only informally participating stakeholder groups and by an institutional safeguard for the equal treatment of Northern and Southern actors (supported by targeted material help for organizations with little resources). Admittedly, these findings from the case studies indicate that even under such conditions global governance arrangements would not succeed in enabling the actors from developing countries to adequately participate and exert notable influence (cp. also Reinicke/Deng 2000: 77; Dingwerth 2008). This results from, on the one hand, that the governance arrangements lack (mainly material) resources to enable actors from the South a balanced participation, and on the other hand, that most of the transnational and private governance arrangements, as well as many international organizations, possess only a low level of awareness. This complicates the identification of all relevant stakeholders and even more their mobilization. In order for these imbalances to not become a structural hindrance for legitimacy, consensus-orientated procedures of standard-setting and decision-making are needed, which ensure that even marginalized stakeholder groups are taken into consideration with their concerns. The analysis of the findings from primary and secondary sources as well as the interviews emphasizes the importance of consensus-orientated procedures. Only the arrangements concerned with regulation show merely a medium degree of consensus-orientation. According to this study, they could make a contribution to the enhancement of their input-legitimacy by opening up their consensus-building procedures for external stakeholders and by the factual consideration of the results gained by these procedures.
  • 4 In the throughput-dimension, it is striking that the majority of the governance arrangements show no readiness to open up their decision-making processes for the participation of observers and make them transparent for the public. Technically, it would be possible for the Internet, but obviously many organizations are afraid that such openness would be at the expense of a consensus-orientation and effectiveness. In contrast to this, the results of this study demonstrate, at least with regard to the ILO, the FSC and ICANN that a high degree of transparency definitely can come along with a high degree of consensus-orientation and a high acceptance. In contrast to the other governance arrangements, they stand out due to their codified information policy that incorporate even external actors and ensuring that decisions are attributable to the responsible person. The other governance arrangements avoid this by selective or even restrictive information policies. Concerning the safeguarding of accountability by monitoring control, ICANN and the FSC set standards that aim with their procedures not at the rule-addressees alone but they subordinate themselves to these monitoring standards. Altogether, all analysed governance arrangements suffer from a lack of control capacities23 and should therefore think about ways to encourage external stakeholders that are ready to fulfil the task in a supplementary way. However, NGOs alone are only to a limited degree able to execute this function (cp. Scholte 2011: 319). The FSC has the strongest institutionalized sanctioning mechanisms that are not directed solely at the rule-addressees but also against the bodies of the FSC themselves. In addition, external actors can under certain conditions initiate sanctioning procedures. Other governance arrangements could learn from the experiences of the FSC with this practice. An alternative strategy for sanctioning could exist through enabling rule-addressees to comply with the rules. To this end, appropriate resources are necessary on the side of the governance arrangement. Therefore, it is no surprise that the ILO in particular is committed to this practice.24 On balance, there are good reasons for alternative strategies that justify a neglect of monitoring- and sanctioning mechanisms, but considerable transparency of structures and processes of global governance arrangements seems to be an indispensable indicator of legitimate governance. This assumption is confirmed by the interviews as well as the analysis of primary sources and secondary literature.
  • 5 In addition, the indicator-based comparison across policy fields shows that all investigated governance arrangements with a regulative (i.e. steering) function (by standards or certification) set goals that are strongly aligned with universal principles. Only in the field of internet regulation, where governance arrangements should fulfil a constitutive task (i.e. setting of a regulative framework), functional goals dominate. At the same time, all these constitutive governance arrangements show a high degree of expertise that seems to be the most important explanatory factor for their acceptance, whereas accordance with universal principles, as distinguished from regulatory governance arrangements, plays no major role. It is also remarkable that the ILO alone disposes of noteworthy resources, albeit all other governance arrangements have the ability to generate at least a medium degree of acceptance. This indicates that resources for problem-solving are not an indispensable criterion for the generation of acceptance.

Based on this premise, some assumptions about the conditions, possibilities and limits of legitimate governance beyond the nation state can be formulated. Global governance arrangements can generate legitimacy by themselves only if they are ready to open their organizational structures and procedures for external stakeholders. They must also grant them an equal and effective influence by consensus-oriented and transparent processes of standard setting and decision-making. To what extent they succeed also depends on how effectively they can elucidate the general significance and acuteness of their respective issues, and at the same time, to enhance their degree of popularity.25 In this respect, global governance arrangements should also strive to initiate and accompany learning processes by informational activities that reach as many stakeholder groups as possible in preferably all regions of the world. For such informational activities, as well as for the identification of stakeholders, their material support and the periodical control of their rule-addressees, governance arrangements need resources. No organization has a sufficient quantity of resources available for all of these activities. Thus, global governance arrangements are confronted with the dilemma that through informational activities their acceptability, their expertise and as a result their factual acceptance could increase, but the material resources necessary for it must to some extent come from their own coffers (e.g. by membership fees and dues). Generally, they are dependent on the support of external actors like private foundations or bodies governed by public law what weakens their status as autonomous actors in world politics. Thus, governance beyond the nation state can generate only a limited legitimacy from inside. Furthermore, the findings of this study suggest that the indicators of legitimate governance applied in this context and proven as explanatory could be supplemented by additional indicators. Appropriate candidates could be (1) the degree of flexibility a governance arrangement guarantees its rule-addressees in the implementation of standards and rules or (2) the ability of governance arrangements to learn and adapt to varying conditions or changing demands from relevant stakeholder groups. Before these questions can be addressed in a systematic way, the tentative conclusions about adequate forms of legitimate governance beyond the nation state generated by this study should come under advanced scrutiny. To this end, it would make sense to conduct case studies in other policy fields, for example in the field of private security services (Cutler 2010; constitutive governance arrangements) or in the policy field of fair trade (Ullrich 2011; regulative governance arrangements) (cp. Take 2009). Thus, the insights generated in the framework of this study could be substantiated and transferred into a theory of global governance.


  • 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.

  • 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).

  • 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.

  • 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.

  • 10

     “Input-oriented arguments often rely simultaneously on the rhetoric of “participation” and of “consensus”” (Scharpf 1999: 7).

  • 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).

  • 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.

  • 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.

  • 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).

  • 18

     This seems to apply to most contemporary global governance arrangements (cp. Scholte 2011: 24).

  • 19

     SAI and RugMark as members of ISEAL are at least in certain areas (standard setting) subject to peer-control. The ILO lets at least its operational activities be evaluated by independent actors.

  • 20

     “In September 2009 the Joint Project Agreement between ICANN and the US Department of Commerce was replaced by a so-called Affirmation of Commitments that transferred responsibility to monitor ICANN from the US Government to a global review process” (Chango 2011: 271).

  • 21

     The ILO by itself is not allowed to impose substantial fines, but its Council can request the member states to take legal measures against a member that continually violates norms (Art. 33).

  • 22

     Because sanctioning in the form of the cancellation of competences would be contrary to the interests of the United States, it is unlikely (Froomkin 2000: 50).

  • 23

     That governance arrangements mostly invest more money in the ability of rule-addressees to comply with standards and rules than in monitoring activities and that even in the case of uncovered breaches of the rules, they focus more on their correction than on punishment certainly relates to the fact that all governance arrangements are dependent on gaining a preferably high number of actors which subordinate voluntary to their standards and rules in order to establish themselves in the international system at all.

  • 24

     An additional strategy for the sanctioning of rule-addressees by punishment could be to design the standards and rules so that the different rule-addressees have enough scope for their compliance, accounting for their societal prerequisites. This practice can be observed, for example, at the PEFC and ISO 14001.

  • 25

     According to Scholte (2011: 313), “(…) global governance has become considerably more visible to affected publics over the past several decades.“



Interviewed representatives of organizations:

FSC: FSC Germany, FSC Denmark, FSC Finland, FSC Italy, FSC Cote d′Ivoire, FSC Slovakia, FSC Czech Republic, FSC South Africa, FSC International Center (Bonn).

PEFC: PEFC Belgium; PEFC Germany; PEFC Austria; PEFC Luxembourg; PEFC Switzerland; PEFC Brazil; European Network of Forest Entrepreneurs; Conféderation Européenne des propriétaires Forestiers; American Forest Foundation; Deutscher Forstwirtschaftsrat.

ISO: Sari Sahlberg, Finnish Standard Association; M. Abderrahm Taibi, Service de Normalisation Industrielle Marocaine; Zdenka Bursova, Czech Standards Institute; Javier Garcia Diaz, Spanish Association for Standardization and Certification; Ghana Standards Board; Claudia Laabs, DIN.

ICANN: no responds.

ITU: no responds.

WSIS: no responds.

ILO: ILO SEA; ILO EVA; Andre Dale (ILO Director).

RugMark: Tina Gordon, Transfair, RugMark Deutschland, RugMark USA.

SAI: Eileen Kaufmann, SAI

Interviewed experts:

FSC: Klaus Dingwerth, University of Bremen, Germany; Fred Gale, University of Tasmania, New Zealand; Karina Ahorian and Peter May, Amigos de la Terra, Brazil; Matthias Schwörer, Ministry for Agriculture, Forest Division, Germany; Elke Mannigel, Oro Verde; Scott Paul, Greenpeace; Sébastian Risso, EU Forest Policy Officer, Greenpeace EU Unit; Martin Kaiser, Greenpeace Germany; Ruth Nogueron, World Resources Institute; United Nations Forum on Forests Secretariat.

PEFC: Matti Liimatainen, Greenpeace Nordic; PEFC Watch; Ruth Nogueron, World Resources Institute

ISO: Natalie Eckelt, BUND; Germany; Chris Elliot, WWF Canada.

ICANN: Thomas Hoeren (University of Münster), Milton Mueller (Syracuse University), Jaqueline A.Morris (WGIG), Wolfgang Kleinwächter (Department for Media and Information Sciences of the Unversity of Aarhus), James Love (Consumer Protection Technology);

ITU: Don MacLean (International Institute for Sustainable Development), William Drake (Project on Information Revolution and Global Governance), Wolfgang Kleinwächter (Department for Media and Information Sciences of the Unversity of Aarhus);

WSIS: Tarek Cheniti (Oxford University), Marylin Code (AT&T), James Love (Consumer Protection Technology), Wolfgang Kleinwächter (Department for Media and Information Sciences of the University of Aarhus).

ILO: Eva Senghaas-Knobloch, University of Bremen; Andrea Liese, FU Berlin; Fair Labour Association; Campaign for Labor Rights; National Labor Committee; Werner Sengenberger (former ILO-Director); US Department of Labor; ICTFU United Nations Office; Human Rights Watch; NLC USA; James Howard, International Confederation of Free Trade Unions(ICFTU).

RugMark: UNICEF, Germany.

SAI: NLC USA; Jasjit Sodhi, Lead Auditor; Madhuri Lele, Badacsony and Kiraly Ltd HU; Ana Delgado, Conexion Social; Vic Thorpe, just solutions.

Interviewed rule addressees (anonymized):

FSC: N&V Forest Products Netherlands; Precious Woods, Brazil; Woodland Trust, UK; Cikel Brasil Verde Madeiras Ltda, Brazil; Orsa Florestral, Brazil; Finnforest, Estonia; Kinnarps, Sweden; Espen Holz, Germany; Forest Finance, Germany; Koninklijke Wöhrmann, Netherlands; Burgo Benelux, Netherlands; Elof Hansson AB, Sweden; Staatsbosbeheer, Netherlands; Viking Timber Sweden; RICOH Industry, France; FSC Germany

PEFC: Skogsägarna Mellanskog, Sweden; Forestry Division Carinthia; Arcadian Timber, Canada/USA; ECS Entre Prenörscertifiering SB, Sweden; Anthony Forest Products, USA; Sb Skog, Norway; Forestal Mininco, Chile; American Forest and Article Association, USA; Consorzio Fores tale Valli Stura e Orba, Italy.

ISO: no responds

ICANN: Ministry of Foreign Affairs (Argentina), CAFEC, Intellectual Property and Technology Section of the Hawaii State Bar Association/Of Counsel Godbey Griffiths Reiss, LLP, Hongkong Internet Forum, ISOC Puerto Rico, Asociación de Derechos Informatico de Argentina, ISOC Australia, National IT and Telecom Agency (Denmark), Consumer Reports Web Watch, Estonian Informatics Centre, Ministry of Transport and Communications (Finland), Federal Department of Environment, Transport and Communications (Schweiz), Director of Information Society and Information Technologies Directorate (Bulgaria), Ministry of Economic Development (Newzealand), ISOC France, dot Berlin, CENTR, ISOC Germany;

ITU: Ellipso, VeriSign, Askcomreg (Irland), Vodacom, ERO (European Radio Communications Office, Denmark), Institut Belge des Services Posteaux et des Télécommunications (IBPT), Europäische Kommission, Information Society and Media DG, Bockham Tec., National Communications Authority Ghana, Flashmedia Group South Africa, EU-Kommission, Belgacom;

WSIS: Ministry of Foreign Affairs (Belgium), Delegation of Australia, Delegation of Nicaragua, University of Mexiko City, Belgian Senate, Delegation of Indonesia, Ministre de PTT (DR Kongo), International Center for New Media, Telecommunications Comissioner (Thailand), Communications Regulation Authority (Latvia), First Section.

ILO: Bundesministerium für Arbeit und Soziales, Germany; ITUC; US Council for International Business (USCIB); Ceylon Bank Employees Union; Bundesvereinigung der Deutschen Arbeitgeber, Germany; Deutscher Gewerkschaftsbund, Germany.

RugMark: Nordic Home Interiors, USA; Designer House& Carpets; formation carpets; Emma Gardner Design; Zafar Orient Designer.

SAI: Eileen Fischer; It A Carpets; Kastex; APEX Vietnam; Raagan Exports, India; Inter Market Knits, Pakistan.

PD Dr. Ingo Take is currently supplementary professor for International Relations at the University of Göttingen. His research is focused on legitimate governance beyond the nation state, NGOs, Internet regulation, social and environmental standards and world society. Recent publications: Take, Ingo (Hrsg.) 2009: Legitimes Regieren jenseits des Nationalstaats. Unterschiedliche Formen von Global Governance im Vergleich, Baden-Baden: Nomos Verlagsgesellschaft; Take, Ingo with Dirk Jörke 2011: Vom demokratischen zum legitimen Regieren, in: Politische Vierteljahresschrift 2/2011. Address for correspondence: Institute for Political Science, University of Göttingen. OEC, 0.134, Platz der Göttinger Sieben 3, 37073 Göttingen, Germany. Phone: +49 (0)551 39 12409; Email: take@uni-greifswald.de