Steven Bernstein, Department of Political Science, University of Toronto, 100 St George Street, Toronto, Ontario, M5S 3G3 Canada. Email: firstname.lastname@example.org.
Benjamin Cashore, Program on Forest Policy and Governance, School of Forestry and Environmental Studies, Yale University, 230 Prospect Street, Room 206, New Haven, CT 06511–2104, USA. Email: Benjamin.Cashore@yale.edu
In the absence of effective national and intergovernmental regulation to ameliorate global environmental and social problems, “private” alternatives have proliferated, including self-regulation, corporate social responsibility, and public–private partnerships. Of the alternatives, “non-state market driven” (NSMD) governance systems deserve greater attention because they offer the strongest regulation and potential to socially embed global markets. NSMD systems encourage compliance by recognizing and tracking, along the market’s supply chain, responsibly produced goods and services. They aim to establish “political legitimacy” whereby firms, social actors, and stakeholders are united into a community that accepts “shared rule as appropriate and justified.” Drawing inductively on evidence from a range of NSMD systems, and deductively on theories of institutions and learning, we develop an analytical framework and a preliminary set of causal propositions to explicate whether and how political legitimacy might be achieved. The framework corrects the existing literature’s inattention to the conditioning effects of global social structure, and its tendency to treat actor evaluations of NSMD systems as static and strategic. It identifies a three-phase process through which NSMD systems might gain political legitimacy. It posits that a “logic of consequences” alone cannot explain actor evaluations: the explanation requires greater reference to a “logic of appropriateness” as systems progress through the phases. The framework aims to guide future empirical work to assess the potential of NSMD systems to socially embed global markets.
Despite these advances, theory development is being hampered by the conflation of mechanisms with different characteristics, scope, depth, and prospects for transforming the global marketplace. This is particularly problematic for assessing “non-state market driven” (NSMD) governance systems because, unlike the voluntary nature of most other forms of private authority, these are designed to create binding and enforceable rules (Cashore 2002). NSMD systems are defined here as deliberative and adaptive governance institutions designed to embed social and environmental norms in the global marketplace that derive authority directly from interested audiences, including those they seek to regulate, not from sovereign states. Operationally, they use global supply chains to recognize, track, and label products and services from environmentally and socially responsible businesses. They operate in what John Ruggie (2004, p. 504) labels an emerging global public domain: an “increasingly institutionalized transnational arena of discourse, contestation, and action concerning the production of global public goods, involving private as well as public actors.” Here, members of the public increasingly express their demands to moderate the excesses of global liberalism and to “embed” markets in broader societal goals.
Non-state market driven systems have proliferated to address: global problems such as fisheries depletion; deleterious environmental impacts from forestry, food production, tourism, and mining; rural and community poverty; and inhumane working conditions (see Appendix I). Their potential impact is far from trivial. Current systems alone operate in sectors that represent one-fifth of the products traded globally.1
What makes these non-state governance systems unique? How do they gain governing authority? What is their transformative capacity? To address these questions, we develop an analytical framework designed to explicate how NSMD systems might achieve “political legitimacy,” defined as the acceptance of shared rule by a community as appropriate and justified.2 We focus on legitimacy because if NSMD systems are to achieve their goal of moving beyond static systems in which firms and social actors constantly evaluate and reevaluate whether to withdraw support based on short-term cost–benefit calculations, they must become more deeply engrained as legitimate authorities (Levi & Linton 2003, p. 419).
Our framework addresses three shortcomings in existing scholarship. First, we differentiate NSMD systems from other forms of private authority, drawing on and revising Cashore (2002). Second, we highlight how globally institutionalized norms, or social structure, provide the constitutive and regulative basis of legitimacy for these systems. Third, we challenge the tendency in existing scholarship to assess support for NSMD systems as either strategic or norm driven. We assert that while much of the initial motivation for firms to participate comes from market-based incentives, a full-fledged theory of NSMD system emergence and institutionalization must address how firms, as well as consumers, community stakeholders, and non-governmental organizations (NGOs) interact with norms embodied in, and promoted by, NSMD systems. We posit that actor interactions produce a three-phase process through which NSMD systems may gain political legitimacy: initiation (phase I), building widespread support (phase II), and political legitimacy (phase III). As systems move through the phases, we argue that what March and Olsen (1998) identify as a “logic of appropriateness” becomes increasingly important to explain actors’ evaluations, although strategic action based on a “logic of consequences” continues to play a role.
Our method includes both inductive and deductive elements. Inductively, we draw on data collected over the last seven years, including more than 230 interviews and 21 country-level analyses that focused primarily on NSMD in the global forest sector, as well as interviews and documentary analysis specifically for this paper on the range of NSMD systems in Appendix I. Deductively, we draw on theoretical work in political science, sociology, and management to build propositions on conditions for movement from phase II to phase III. Although no system currently is fully at phase III, our attention to the transformative impact on the global marketplace that might occur is especially justified given the ultimate goal of NSMD systems. Our aim is to build a comprehensive analytical framework, which we then use to initiate theory-building. We do not aim to test our arguments in this paper.
We proceed in the following analytical steps. First, we distinguish NSMD governance from other forms of private authority and justify our attention to political legitimacy. Next, we inductively develop the core of our framework, that is, the conditioning effects of social structure and the three-phase process of achieving legitimacy. The final section proposes preliminary causal arguments on the conditions under which NSMD systems gain full political legitimacy (move to phase III).
NSMD systems and their need for political legitimacy
Features of NSMD systems
Drawing on Cashore (2002) and Cashore et al. (2004), we discern five characteristics of an ideal-type NSMD system. Together, these characteristics distinguish NSMD systems from other forms of private authority.
First, NSMD systems do not derive policy-making ability from states’ sovereign authority. As elaborated in Cashore’s earlier work (2002), this feature does not mean that states are unimportant: some state agencies have provided financial support for particular NSMD systems, and domestic and international regulatory environments potentially affect their activities. However, even in cases where governments supported their formation, NSMD systems do not derive governing authority from states nor are they accountable to them.
Second, NSMD institutions constitute governing arenas in which actors purposely steer themselves toward collective goals and values and in which adaptation, inclusion, and learning occur over time and across a wide range of stakeholders. Dynamic governance differentiates NSMD systems from most traditional ecolabeling initiatives (e.g. Nordic Swan), which generally identify a static measure of environmental quality a firm must adopt to receive a label. NSMD system managers justify this design feature on the grounds that it makes NSMD systems more democratic, open, and transparent than many of the business-dominated public policy networks they seek to bypass, as well as most corporate self-regulation and many social responsibility initiatives.
Third, authority granted to NSMD systems emanates from the market’s supply chain. Producers and consumers from extraction to end-users (in the case of commodities such as forest or agricultural products) or from service providers to consumers (in the case of services such as tourism) make individual choices about whether to require that products or services are certified for compliance to an NSMD system.
Fourth, NSMD systems aim to reconfigure markets. They attempt to ameliorate global problems that, in their absence, firms have little incentive to address. This feature distinguishes NSMD systems from new arenas of private authority designed to standardize business practices, such as accounting, or to improve market coordination. In those cases, economic incentives for profit-maximizing firms inherently exist. Thus, they pose no puzzle for compliance (Porter 2007).
Fifth, NSMD systems possess mechanisms to verify compliance and to create consequences for non-compliance. This feature means that, in effect, they develop mandatory standards for those who sign onto the system. The most common compliance mechanism is a third-party audit in which auditors “certify” firm or producer compliance with the rules or identify improvements required for a successful audit. In contrast, self-regulation and CSR standards are usually voluntary and often discretionary, even for those who sign onto them.3
Arguably, the first full-fledged global NSMD system was the Forest Stewardship Council (FSC) certification program. Transnational environmental and social groups created the FSC in 1993 following governments’ failure to negotiate a binding global forest convention. To avoid business domination, which many view as a key problem with state-centered processes, the FSC includes environmental, social, and economic decision-making chambers, each with equal voting weight. It also excludes governments from formal participation. The FSC created nine international principles and criteria (later expanded to 10) to guide the development of environmentally and socially appropriate standards in local settings around the world. The FSC accredits and requires auditors to certify companies that manage their operations according to FSC rules.
Forest industry and forest owner associations subsequently undertook their own initiatives in national settings including the US, Canada, Indonesia, Finland, Brazil, Malaysia, and Australia. In many cases, their interest in competing as a “legitimate” NSMD system in the global marketplace led them to adapt their systems to transnational market requirements or, as in the case of the Program for Endorsement of Forest Certification (PEFC), to create formal global institutions.4 Because debates continue among industry, forest owners, and environmental groups over which program is preferable – like similar debates in other sectors – our framework draws attention to the dynamic interactions among burgeoning NSMD systems and their competitors as they vie for legitimacy.
Buoyed by widespread support in the forest sector, other systems emerged or evolved to fit the NSMD classification.5 The most prominent systems are reviewed in Appendix I. They include the Fair Trade Labelling Organization (FLO), which coordinates under one system groups that had worked separately on consumer campaigns to improve the conditions of poor and marginalized producers in the developing world. FLO covers internationally traded commodities and specialized goods including coffee, tea, cocoa, sugar, bananas, rice, honey, vanilla, nuts, clothing, sporting goods, flowers, wine, and diamonds. Similarly, Social Accountability International, initiated by the nonprofit Council on Economic Priorities to reduce sweatshop labor practices, developed into a system that monitors companies according to specified social criteria, including child labor and worker safety (Bartley 2003; Courville 2003; O’Rourke 2003). The FSC model explicitly inspired the Marine Stewardship Council (MSC) governing natural fisheries management and the Sustainable Tourism Stewardship Council, among others. The International Social and Environmental Accreditation and Labelling (ISEAL) Alliance is an umbrella organization created to develop agreement on “best practices” for any NSMD system (ISEAL 2006).
NSMD systems, as we elaborate below, are unlikely to govern effectively if they depend solely on firms’ strategic interests for compliance (Meidinger 2006). Like governments, whose coercive capacity is reinforced by their legitimacy, NSMD systems require legitimacy to justify policy development and enforcement measures. Unlike sovereign states, which by definition possess legitimate authority, or international organizations, which derive their legitimacy from sovereign states’ consent, NSMD systems must actively achieve “political legitimacy” (Suchman 1995; Cashore 2002).
Whereas the concept of legitimacy generally refers to viewing the actions of an entity as “desirable, proper, or appropriate” (Suchman 1995, p. 574), our analytical framework focuses specifically on political legitimacy because it concerns the acceptance of a governance relationship, where commands ought to be obeyed (Esty 2006, p. 1511). It reflects “a more general support for a regime [or governance institution], which makes subjects willing to substitute the regime’s decisions for their own evaluation of a situation” (Bodansky 1999, p. 602). Political legitimacy requires institutionalized authority (whether concentrated or diffuse) with power resources to exercise rule as well as shared norms among the community. Norms of legitimacy provide justifications and a shared understanding of what an acceptable or appropriate institution should look like and bounds what it can and should do.
We purposely avoid adding specific content to what legitimacy requires by definitional fiat. Instead, our framework focuses conceptual attention on how the normative environment and interactions of actors within NSMD communities determine and shape the process and content of legitimacy granting. This approach highlights that criteria of legitimacy are contingent on historical understandings at play and the shared norms of the particular community or communities granting authority.
Achieving political legitimacy
The conditioning effects of global social structure
Global norms and institutions structure actors’ choices over whether to support NSMD systems. They serve a constitutive or legitimating function that has enabled the emergence of NSMD governance by defining what appropriate authority is, where it can be located, and on what basis it can be justified. They also serve a regulative function by prescribing and proscribing the boundaries of NSMD governance activities.
Constructivist international relations scholarship has used such a notion of social structure, drawing especially on insights from sociological institutionalism (Finnemore 1996; Meyer et al. 1997; Ruggie 1998, pp. 22–25; Reus-Smit 1999; Barnett & Coleman 2005). Although these writings use various formulations – an “environment” in which organizations operate, “normative structure,”“social structure”– their basic insight is that already institutionalized norms define appropriate and inappropriate courses of action, legitimate institutional forms, and create a context in which cost–benefit analysis occurs, even making certain actions unthinkable.
Structuring can be understood to operate through an idea of “fitness,” where legitimacy is understood as embedded in social systems that provide a basis of appropriateness, or that make the purposes, goals, or rationale of an institution understandable and justifiable to the relevant audience in society (Weber 1994, p. 7; Bernstein 2001). Thus, legitimacy is enhanced when the norms and rules of organizations conform to existing social structure, where they compete for legitimacy (Scott & Meyer 1983, p. 140; Barnett & Coleman 2005, p. 598). In the case of NSMD systems, the relevant social structure includes institutionalized global marketplace norms as well as emerging democratic, social, and environmental norms in the global public domain. These norms can be found not only in specific declarations or principles that might apply to the sector, product, or process in question (e.g. the Statement of Forest Principles or Convention on Biodiversity in the case of forestry or core International Labor Organization [ILO] conventions in the case of labor), but also include broadly accepted norms of global environmental, labor, and human rights governance. These may be embodied in international treaties or “soft” declaratory international law, as well as in action programs, or in statements of leaders. Relevant international “hard” law includes legalized trade rules under the World Trade Organization (WTO), especially the agreement on technical barriers to trade (TBT), which includes coverage of non-governmental standardizing bodies.6
Social structure, however, is not wholly determinative of NSMD success or failure for two reasons. First, NSMD systems may themselves be a source of change in international norms and rules. Second, in line with most social theory, we recognize that agents and organizations almost always possess some autonomy. They may therefore pursue a variety of strategies, including acquiescence, compromise, manipulation, or strategic social construction to succeed within given structures (Barnett & Coleman 2005, pp. 600–602). For these reasons, our overall framework stresses the importance of the interaction of social structure with the dynamics of choice, an increasingly common strategy among scholars interested in the interaction of logics of consequences and appropriateness (Finnemore & Sikkink 1998; Risse et al. 1999; Checkel 2005).
Evidence that global norms and rules matter
We observe that three sets of relevant norms in social structure have facilitated the emergence of NSMD systems. First, sovereignty norms have worked to their advantage. Although NSMD systems are disadvantaged in gaining legitimacy in comparison to intergovernmental institutions because they must build authority from the bottom up, their autonomy from intergovernmental processes allows them to tap into emerging norms more quickly. NSMD systems in forestry, fisheries, workers rights, and agricultural production emerged because certification corrected inattention to existing problems or provided a way forward when international negotiations stalled. They also effectively bypassed thorny debates over sovereignty – frequently a source of intergovernmental stalemate – because they targeted firms, not states.7
Second, the emergence of NSMD systems has corresponded with a general shift in global environmental norms, and in the international political economy more generally, toward sympathy with market mechanisms and economic liberalism more broadly (Bernstein 2001). This shift has provided a supportive normative environment for market-based systems such as the NSMD systems reviewed in this article. For example, Principle 12 of the 1992 Rio Declaration on Environment and Development – the most widely accepted consensus statement on sustainable development norms – posits that environmental protection and open markets are compatible. This position is also found in the WTO preamble and in policy statements across the Bretton Woods and United Nations (UN) system. Recent events and initiatives reinforce this normative understanding, including the 2002 World Summit on Sustainable Development, where public–private partnerships emerged as a dominant mechanism to implement sustainable development in the shadow of disappointing progress in most areas of intergovernmental cooperation. These examples illustrate the increasingly shared understanding that working with markets and the private sector is the appropriate course of action in global governance.
Third, a growing normative consensus on the need to “democratize” global governance reinforces the legitimacy of NSMD systems’ inclusive approach to decision-making. These norms include demands for democratic reform and improved public accountability of international institutions to states and/or broader affected publics (Payne & Samhat 2004; Held & Koenig-Archibugi 2005), as well as “stakeholder democracy” that calls for “collaboration” and truer “deliberation” among states, business, and civil society (Bäckstrand & Saward 2004; Vallejo & Hauselmann 2004). Such normative pressure is especially prevalent in international environmental institutions, treaties, and declaratory law that have been on the forefront of promoting increased public participation and transparency at all levels of governance (Bernstein 2005). Specific examples include codification in principles such as Rio Declaration Principle 10 (which states that environmental issues are best handled with participation from all “concerned citizens at the relevant level”) and the Aarhus Convention on Access to Information, Public Participation in Decision-making and Access to Justice in Environmental Matters, which came into force in 2001. Institutional reforms along these lines include the creation of the UN Environment Program’s civil society forum (institutionalized in 2002), the World Bank Inspection Panel (created in 1993 to improve accountability to local communities), and democratic reforms in the Global Environmental Facility in the mid-1990s to improve meaningful participation from developing countries.
NSMD systems both promote and embrace these democratic norms. For example, they help explain the Forest Stewardship Council’s eventual adoption of its three-chamber decision-making process. Likewise, the emergence of the Workers Rights Consortium can be explained, in part, by the initial failure of the Fair Labor Association (FLA) to develop multi-stakeholder processes in which social groups and institutional purchasers of products felt included. Similarly, the domination of Unilever in the development of the MSC contributed to its limited uptake from environmental and social groups. Following complaints of a “democratic deficit,” the MSC undertook a governance review that resulted in an overhaul designed to better ensure openness, transparency and accountability to all stakeholders (MSC 2001).
The internal attention to democratic norms is increasingly matched by external expectations. States and international organizations, including the WTO, World Bank, ILO, and Food and Agricultural Organization, increasingly demand that the development and implementation of standards be inclusive, be transparent, include participation of stakeholders, and be adaptable to local conditions in order to be recognized as legitimate. As one NSMD system official explained, “it’s a chicken or egg” situation, where democratic expectations created by NSMD systems are feeding back to create expectations for all social and environmental standards, including those set by traditional standard setters such as the International Organization for Standardization (ISO).8
In terms of “hard” law, contemporary international trade rules provide an enabling environment for NSMD systems in three ways. First, the neoliberal normative environment reflected in contemporary trade regimes provides enabling conditions for market-friendly systems such as NSMD governance. Second, although NSMD systems must navigate the trade rules that regulate international standards in order to avoid disputes, these rules target state regulation and leave significant leeway for non-state governance. Indeed, governments and intergovernmental organizations have resisted directly adopting an NSMD system standard in order to avoid disputes. For example, the ILO considered but rejected a proposal to certify countries rather than firms with a “global social label” owing to developing country concerns it would constitute a non-tariff trade barrier and contravene WTO rules (Bartley 2003, p. 450).9 Third, while ambiguities in trade law make it uncertain what the outcome of a trade dispute would be if a state adopted an NSMD system standard (Joshi 2004), NSMD systems are increasingly proactive in seeking to conform to international rules. For example, ISEAL has instructed its members to adhere to TBT Annex 3 (the Code of Good Practice for the Preparation, Adoption and Application of Standards) and Annex 4 (Second Triennial Review) that define conditions for recognition of international standards. Leading NSMD systems within ISEAL have also started to register with the World Standards Services Network and begun to conform to relevant ISO guides.
The evolutionary logic of NSMD governance: Three phases
Against this backdrop of an enabling global social structure, we posit a three-phase interaction process that takes the perspective of actors and audiences (the NSMD system's community) along the supply chain, who must evaluate whether to support the NSMD system in question. In identifying these phases, we focus attention on the legitimacy achieving strategies of NSMD systems and learning through which political legitimacy develops. Uncovering these processes addresses the paradox of why profit-maximizing firms would ever agree to abide by a non-state political authority that increases burdens and shapes their behavior, especially when incentives to avoid, shirk, and exit the system are high (Prakash 2001; Rivera 2002; Sasser 2002; Raines 2003). Similarly, it addresses why environmental and social groups would ever agree to a certification system palatable to firms when such systems face overwhelming pressure to devise “second best” standards that do not put supporting firms at a competitive disadvantage vis-à-vis non-participants.
This conundrum, when acknowledged at all, is frequently cited by rationalist accounts to argue that NSMD systems will never be effective. Such accounts assume that evaluations of legitimacy are utilitarian: firms base evaluations on cost–benefit calculations because they value profit maximization while environmental and social organizations base evaluations on whether the system’s rules might ameliorate the environmental or social problem(s) they were created to address.10 Whereas attention to such calculations may be appropriate for understanding functional legitimacy – the idea that organizations serve functions that constituents value (Barnett & Coleman 2005, pp. 597–598) – it is incomplete for understanding how political legitimacy, which emphasizes a shared sense of community and norm generation, might emerge.11
To overcome the limitations of utilitarian assumptions, we draw from complementary literatures in political science, organizational sociology, and management that identify different logics or drivers of action. Political scientists, for example, have been drawn to March and Olsen’s identification of a “logic of appropriateness.” In contrast to a utilitarian logic of consequences, this institutional logic “pictures political action as driven less by anticipation of its uncertain consequences and preferences for them than by a logic of appropriateness reflected in a structure of rules and conceptions of identities” (1996, p. 250). Such processes are “built upon visions of civic identity and a framework of rule-based action… Embedded in this notion are ideas about the obligations of citizenship and office, the commitment to fulfill an identity without regard to its consequences for personal or group preferences or interests” (1996, p. 254). Similarly, management scholars (Oliver 1991; Greening & Gray 1994; Prakash 2001) have shown the explanatory power of enduring norms and organizational cultures on how firms manage relationships with an increasingly complex array of stakeholders. And, in sociology, Suchman (1995) distinguishes subjective and largely non-engrained “day-to-day” evaluations of whether an organization seeking societal approval fits the evaluator’s own goals and interests (producing “pragmatic” legitimacy) from a situation in which the appropriateness of the entity becomes so engrained and durable that “to do otherwise was unthinkable” (i.e. it possesses “cognitive” legitimacy). Political legitimacy almost always has some elements of both, but fits neither ideal type. It cannot rest simply on pragmatic evaluations because it requires accepting an institution that sometimes does not operate in congruence with particular actors’ immediate interests, but it rarely if ever corresponds to the “cognitive” ideal type. It almost always rests on discursive validation based on implicit or explicit justifications. To be clear, the distinction is not between “good” values like environmental protection versus, say, profit maximization, but between actions motivated by an un-embedded utilitarian logic versus a situation where interests are pursued in congruence with norms of institutional appropriateness as defined by the governance system. What is “good” is often precisely the issue to be worked out within politically legitimate arenas recognized by a wide-ranging and diverse community.
Thus, we argue that logics of consequences and appropriateness are always at play in motivating firms and social actors over support of NSMD systems. However, changes in the norms of appropriateness (from, say, laissez-faire liberalism to a more socially embedded market) that both enable the legitimacy of NSMD governance systems and are actively promoted through its socializing efforts shift understandings of how profit maximization should occur as systems progress through the three stages (Fig. 1).12 This can occur in two ways. First, as NSMD systems gain legitimacy, their rules may change material incentives facing firms, thus changing evaluations of actors even when acting strategically. For example, the institutionalization of NSMD systems in niche markets creates new opportunities for participating firms to pursue profit. A second reason is that reference to a “logic of appropriateness” can explain how norms potentially redefine firms’ identities (e.g. from being exclusively profit maximizers to being socially responsible actors) in accordance with the expectations of the NSMD system, whether or not they agree with particular decisions of these institutions (Suchman 1995; March & Olsen 1996, 1998; Checkel 2005, p. 804). We argue that the conditions conducive to norm-driven behavior and socialization increase in phases II and III.
Phase I: Initiation
This phase captures early support for NSMD systems, before any active efforts on the part of NGOs to change company evaluations through targeted boycotts, public shaming, or other information and publicity campaigns.13 Thus, economic demand for certified products will be limited or nonexistent. Without firm support at this stage, the certification program will necessarily die.
According to the utilitarian logic that dominates at this stage, firms must perceive the costs of NSMD governance to be less than current or potential economic benefits. Thus, everything else being equal, “model” firms already performing at, or close to, the requirements of the NSMD system will join first because the system will grant an outside stamp of approval that differentiates them from their competitors. For other firms, the high cost of changing practices to meet a system’s requirements acts as a deterrent.
A variety of intervening factors influence how firms make these cost–benefit calculations, especially whether to take into account strategic considerations and long-term risk exposure or only short-term profit maximization. First, non-publicly traded firms with a low debt load can more easily adopt a longer time-frame than firms with higher debt loads and publicly traded firms. Likewise, companies with highly recognized logos that render them de facto monopsonists, such as Nike, have greater leeway in how to evaluate support for certification programs. Third, managers may find a “business” case to adopt an NSMD standard for efficiency or as part of a long-term strategic decision. For example, they may be convinced the standards will help them become more efficient and lower costs, improve labor or community relations, attract investment, open new market opportunities, reduce a variety of risks, or anticipate and avoid future regulatory burdens. These arguments may be part of NSMD systems’ attempts to convince firms to join or may be generated internally. They depend in part on corporate culture and/or the values of senior managers (Prakash 2001; Howard-Grenville & Hoffman 2003; Vogel 2005, pp. 19–24).
Finally, firms that market environmentally or socially sustainable products have a clear incentive to join because their core values fit with the system. Indeed, when a firm faces uncertainty over its profit-maximizing choices, it may “fall back” on existing values, which could include environmental stewardship or social responsibility (Delmas & Toffel 2004). Such values vary significantly across firms in the same sector. At this point, a consequentialist logic dominates because firms still aim to maximize core values. However, because those values come in part from non-market norms, a logic of appropriateness plays a supporting role. To the degree such values overlap with values held by social and environmental groups, it can facilitate norm generation within the relatively small community at phase I.
Environmental and social group evaluations
Like firms, environmental and social groups in phase I make choices driven by strategic calculations that fit their own core values and organizational self-interest. They have an incentive to participate in and positively evaluate governance systems that respond to their goal of ameliorating social and environmental problems.
Expected result in the marketplace
The dominance of strategic logic will first create segmented or “niche” markets that signal conformity of a small number of firms’ practices with NSMD system requirements. As a result, phase I creates a well-defined, although small, “political community” of firms and social and environmental stakeholders, all of whom benefit from addressing global problems. Their overlapping values and strategic interests predispose them to develop “trust” ties, shared norms and understandings, as well as feelings that they are “fighting the good fight” amidst a wave of poor practices elsewhere.
However, the small group of participating firms will be unable to make a serious dent in ameliorating global problems for two reasons. First, solving them requires widespread support. Second, because NSMD systems initially attract firms already practicing close to a system’s standards, they make relatively limited impact “on the ground.”
Evidence for phase I
Although a full review is beyond the scope of this paper, there is no question that the empirical evidence on initial firm uptake of NSMD systems is consistent with our arguments. In the forestry case, the vast majority of industrial forest companies worldwide refused to support the FSC in the early days. Instead, support came from a handful of firms, such as Collin’s Pine in Oregon, interested in distinguishing their practices from those of their heavily criticized competitors (MacArthur Foundation 1998). Many of those initial supporters were small private forest owners willing and able to put environmental values before profits (Hayward & Vertinsky 1999). Likewise, the first supporters of Fair Trade coffee tended to be small business cafés that marketed their concern with developing country coffee production to appeal to their educated clientele (Courville 2001). In addition, virtually every empirical study reveals that NGOs make strategic evaluations based on whether NSMD systems achieve core values of environmental or social improvement.
Phase II: Building support
A major conundrum confronts systems vying for greater support: To attract firms that face higher compliance costs to join, systems face pressure to ease behavioral requirements, something strategically motivated environmental and social groups will resist. Phase II is thus marked by sector- or firm-level shaming and/or boycotting activity that targets initially recalcitrant companies in an attempt to raise the costs of non-participation.
If firms and NGOs followed this strategic logic, we would expect to see divergence of standards, fragmentation of systems, and the creation of distinct and polarized communities. Whereas some evidence supports this explanation for marketplace dynamics, this phase is also marked by convergence, or, in some sectors, oscillating divergence and convergence. To explain convergence, we argue that normative pressures from global social structure combined with the emergence of shared norms and learning can lead to a redefinition of disparate interests and the prerequisites for widespread community building. When convergence occurs, discerning strategic versus norm-governed behavior becomes more difficult because interests may be redefined. Below we explain these countervailing dynamics and assess these arguments against the empirical evidence.
Firms’ strategic behavior
Firms’ strategic choices during this phase fall into three categories. The first group comprises firms that initially joined the system because they easily met its standards. Their desire to see the system expand rests on the relative evaluation of the economic benefits they receive from a niche system that differentiates them from competitors and the costs they incur from already practicing at a higher standard. Only if costs outweigh benefits does expansion serve their interests, because it would “raise the bar” of competitors operating in less stringent regulatory environments. The second group comprises firms that initially joined because their organizational culture or values and financial health predisposed them to be proactive. They should generally support expansion as it would promote what they believe to be appropriate business practices.
The third group – comprising the vast majority of firms – are those firms that initially rejected participation because of perceived costs or fears of loss of autonomy to NGOs with limited knowledge of how markets function. This group is the primary target of environmental and social groups who aim to increase participation.
Utilitarian logic dictates that to attract this third group, in the absence of increased market demand and/or price premiums, certification systems must relax behavioral requirements to reduce compliance costs because this group has further to go to meet standards. Moreover, the experience of systems in Appendix I has been that price premiums and demand only develop incrementally because of hurdles in the supply chain, including fragmentation of producers and limited awareness of customers and consumers. Even if, miraculously, supply chain support quickly accounted for, say, 50% of demand, the unregulated 50% would still produce countervailing pressure on standards if compliance costs are high because it would leave a sizable market for certified companies that have an incentive to cheat or exit the system.
Environmental and social groups’ strategic behavior
Environmental and social groups essentially fall into two camps: those supportive of the creation of NSMD systems and those who are either indifferent to or critical of such efforts. The strategic interest of the former group is to maintain or increase standards. They have “learned” that existing standards can be met (they can point to companies that joined at phase I) and are thus disposed toward shaming firms to meet those standards. They may even believe standards can be raised because some firms did not have to significantly change practices to meet them. Sasser (2002), for example, found that most NGOs would not be ready to grant legitimacy to non-state governance until the on-the-ground effects are shown to improve environmental or social integrity. The second camp is freer to criticize perceived deficiencies in existing systems. They point to unpopular firms receiving certification or rules and standards they deem inappropriate. Their interest is to raise standards. However, successful efforts to do so risk driving away firm support.
Evidence of strategic action in the marketplace
The evidence indicates that these strategic dynamics do matter, but tell only part of the story. Supporting this logic, we observe many efforts to fend off certification, justify the status quo or existing government regulation, and discredit supporters. For example, most forest companies in North America and Europe initially balked at the FSC, explaining that public policies were wholly adequate (Cashore 2002). Similarly, companies criticized for using developing country “sweatshops,” such as Nike, initially defended their practices. NSMD systems in their early stages, such as in fisheries and mining, continue to see limited uptake and little support from retailers along the supply chain. Likewise, Starbucks and Peet’s coffee initially attempted to justify their coffee-purchasing practices as responsible.
Environmental and social groups responded strategically to denounce such claims. Tactics have included shaming companies through the media with evidence of their destructive practices, and obtaining support from customers, whose operations are not the target, for improved performance from their suppliers.14 For example, the US apparel industry has been much more responsive to activists’“dirty gold” campaign than their mining industry suppliers.
When “fending off” fails, industry associations often move to a “plan B” strategy of “pacifying” through the creation of new initiatives (Oliver 1991). These alternatives – mostly of the kind we contrasted to NSMD systems above – rarely include meaningful stakeholder governance or require significant behavioral changes. Forestry and tourism, in particular, have been marked by a proliferation of systems. Under such circumstances, the “community” becomes polarized with few shared norms.
An observed, though unintended, consequence of fragmentation has been increased public debate among supporters of competing systems. In such cases, as each side has argued about the appropriateness of its preferred responses, the sides simultaneously learned about causal mechanisms that permeate efforts to institutionalize NSMD governance. Moreover, various non-industry actors gained knowledge previously unavailable to them about how the industry actually works, including details of production and processes (McDermott 2003). Strong evidence for this pattern comes from forestry, the sector with the longest history in our phase II. In these cases, competition moved from binary opposition to more nuanced discussions of possible “win–win” solutions and learning among NGOs that some practices previously deemed detrimental may have positive social or environmental consequences.
The mechanisms under which this can occur may vary, but inductively we can discern two patterns. First, business-initiated competitors frequently move toward incorporating characteristics of NSMD systems – they engage in “mimetic isomorphism” or “mimicking the most prominent or secure entities in the field” (Suchman 1995, p. 589). For example, although the chemical industry maintains control over Responsible Care’s governance, it started to require third-party verification of its standards in 2005. Similarly, the FLA, spawned by the US Apparel Industry Partnership, initially lacked mandatory standards or independent verification of compliance, but introduced mandatory third-party auditing in response to competition with the Workers Rights Consortium (Bartley 2003; Göbel 2004, pp. 51–52). And, whereas the American Forest & Paper Association (AF&PA) attempted to retain as much authority as possible over its Sustainable Forestry Initiative when it created an external advisory committee, this body’s failure to fit the norms of what the marketplace considered appropriate led AF&PA to adapt again and create an independent multi-stakeholder board with control over rule development. This strategy is at once an attempt to buttress legitimacy by conforming to established models or standards and a signal that reinforces the legitimacy of the organizations mimicked, as it recognizes those entities as the accepted standard.
Second, these systems frequently act strategically to enlist “business-friendly” environmental and social groups in an attempt to gain credibility in the wider community, but with the consequence of opening up space for shared norms to emerge. This occurs because even business-friendly NGOs will attempt to influence understandings from the “inside” and will share information with mainstream activists. False claims and rhetoric on all sides become less tenable as greater information is made available to all parties. Together, these two patterns of behavior suggest that the transformation of many business-initiated NSMD competitors toward the NSMD model illustrates an appropriateness logic at play. In perhaps the clearest example, many firms and business associations in Europe and North America have recently begun to justify their support of forest certification as an engrained business practice (Cashore et al. 2004).
Convergence effects can be observed within both NSMD systems and competitors as a result. In the former case, as market support increases incrementally for certification in general, more firms might join the original program in the hope of “working from within” to develop market-friendly standards. A greater sense of community can be expected to develop as systems focus on providing strategic advantages for firms that join, sharing experiences, and directing attention to the technical means of implementation. For example, initially, most leading environmental groups opposed certifying timber from plantations or old growth forests. However, intense discussions led activists to “learn” that fast growing plantations could play a role in reducing demand elsewhere and minimize environmental impacts. Likewise, discussions about how old growth forests function led FSC to accept responsible harvesting in them. Such discussions generated the concept of “high conservation value forestry”– a normative term that now permeates forestry discussions generally.
In the case of NSMD competitors, pressure mounts to “ratchet up” their efforts in response to market demand and the broader normative environment. For instance, many large scale coffee companies are working with the Rainforest Alliance to increase their own behavioral requirements in order to access a coffee certification market dominated by small growers (Fair Trade) and organic coffee (IFOAM) producers.
At this point, strategic versus norm-driven motivations and behavior become hard to disentangle. Such difficulties resonate with broader trends noted in the management literature, which points to shifts in business practices from viewing social and environmental initiatives as issues of regulatory compliance to matters of social responsibility or “strategic engagement” (Howard-Grenville & Hoffman 2003, p. 71). Core values of business may remain largely unchanged, but norms of the appropriate way to do business are shifting.
Accordingly, we observe business-led efforts that appear strategically motivated, but operate within a new understanding of appropriate marketplace behavior. The new market environment creates an incentive for firms to convince purchasers down the supply chain to recognize all NSMD certification programs as appropriate. In response, some entrepreneurial environmental groups have opted for a “third way” of initiating programs that fit the NSMD governance category, but with more relaxed standards than earlier programs. In addition, we see the rise of umbrella organizations such as ISEAL to build consensus on best practices, and multi-stakeholder efforts designed to encourage increased understandings, collaboration, and reduction of uncertainty. Such efforts include formal meetings of supporters of different programs or approaches and new proposals to address the legitimacy and appropriateness of both specific standards and “rules of the game.” In the forest sector, this resulted in the International Forest Industry roundtable publicly adopting virtually all of the FSC principles and criteria. When these efforts failed to address environmental NGO concerns, it actively sought agreements with environmental groups through a global multi-stakeholder process, the Forest Dialogue.
Phase II can lead to at least three scenarios. First, the combination of increased public awareness and competition among systems can put pressure on governments to regulate the problem. However, this seems unlikely at the global level because collective action and political obstacles to international agreement remain on problems that prompted most NSMD systems. Second, institutionalization may continue to be elusive, with pressures for divergence and convergence fluctuating. This might be owing to resistance to learning, mistrust, or deep divisions within the NSMD community, whether between firms and NGOs, among NGOs, or because other relevant actors such as indigenous communities may disagree on what governance norms are appropriate. Here, we would expect to see pockets of success, but also confusion in the marketplace and little progress on the broader global problems driving NSMD governance attempts. A third possibility (our phase III) is the institutionalization of NSMD governance to the point that a critical mass of actors in a sector agree on a common project and to abide by the rules of systems to which they are bound.
Phase III: Political legitimacy
Reflecting our description of political legitimacy, in this phase the full range of stakeholders within a targeted sector recognize their membership in a political community that grants an NSMD system authority to govern. Power struggles do not end – thus strategic action still occurs – but the community recognizes NSMD systems as legitimate arenas in which to mediate disputes and address policy problems. While market transactions along the supply chain still provide the mechanisms through which authority is embedded, and supply chains and product labeling are still critical for recognizing and enforcing compliance, firms no longer evaluate from a strictly strategic perspective whether to grant support to NSMD systems. Instead, they work through them to advance their interests. Similarly, environmental groups move from contingent support based on a specific standard to an acceptance that the NSMD system is a legitimate arena in which to develop appropriate standards. Institutional arrangements facilitate processes in which firms and non-business stakeholders jointly participate to develop standards that best encourage efficient and effective ways to address social and environmental problems.
As no current system operates fully at this phase, we cannot give empirical illustrations, nor can we pre-judge the exact institutional form NSMD governance will take. It could be centered on a single sector-wide system, a coordinated network bound together by shared norms, or a common set of minimal standards and practices. Regardless of its institutional form, once fully institutionalized, NSMD governance standards and behavioral requirements can be increased without putting supporters at a competitive disadvantage as virtually all firms in the legitimate marketplace would be part of the NSMD community.15
Toward theory-building: Conditions for achieving phase III
Up to now, our framework has been largely inductively generated, based on our observations of how NSMD systems have formed and unfolded over time. The evidence presented above suggests that strategic action alone is insufficient to explain the dynamics of what has occurred so far in the development of NSMD systems (i.e. phases I or II), and that a rationalist account, by itself, is even more unlikely to explain further movement toward full political legitimacy (Phase III). It also suggests that the most promising avenue for theory development is to focus on norm generation and community building.16 Developing propositions on the conditions for moving from phase II to phase III cannot proceed inductively, however, as NSMD systems have yet to gain this status, which, as we detail, is far from inevitable. We thus proceed deductively, drawing from two distinct literatures that address how diverse members of a policy community may come to accept as appropriate particular procedures and/or policies even as they maintain distinct core values.
The first literature we draw upon is the “advocacy coalitions framework” (ACF) (Sabatier & Jenkins-Smith 1993). The most relevant finding from this literature for our purposes here is that “policy-oriented learning across coalitions” can lead to congruence, or at least mutual understanding, among groups with distinctly different core values under certain conditions. As phase III concerns legitimacy of systems rather than agreement on outcomes among groups, we would expect that such learning processes are essential; that is, we hypothesize that they constitute necessary conditions for moving to phase III. This literature finds that a key mechanism for policy learning is a forum where expert knowledge from either side can be presented, criticized, and justified. A related set of conditions, identified in a recent study that expanded the ACF to include the psychology of stakeholders, concerns procedural factors. It found the strongest statistical associations were among trust, procedural fairness, and a consensus on the legitimacy of their collaborative policy process (Leach & Sabatier 2005, p. 498). These findings resonate with pressures on NSMD systems toward democratic decision-making, as well as recent empirical findings that suggest a lack of participation and accountability in NSMD rule-making, or lack of resources to enable participation, prevents a sense of “ownership” among participants in the scheme, which can in turn influence perceptions of justice and fairness (Raines 2003; de Azevedo 2004, pp. 88–89).
A second literature, which adapts insights from Jürgen Habermas’s theory of communicative action to international relations, reinforces the importance of shared understandings and fairness (Risse 2000, 2005). This scholarship argues that legitimacy is enhanced under conditions in which actors share enough of a “common lifeworld,” of collective interpretations of the world and themselves, upon which they can draw to make truth claims and interact within a system of norms and rules perceived as legitimate. Actors must also be open to arguing and persuasion and avoid pulling rank, using private information, or coercive tactics.
The most important empirical finding in this literature for our phase III is that argumentation as described above, which facilitates legitimacy, has been found mainly to play a dominant role when norms are already well institutionalized (Risse et al. 1999; Risse 2005, pp. 164–165). Although this creates a circularity problem as the legitimacy of the system is required for participants to be willing to forego strategic behavior, it also leads us to hypothesize that learning processes in the lead-up to phase III are important in designing formally democratic or deliberative procedures, because practices associated with true deliberation are unlikely to take hold until late in a process of socialization to the norms of the system.
These findings suggest two additional propositions about the conditions under which NSMD systems might move to phase III. First, learning processes must be established that include forums for exchanges of expert information, the building of databases of experiences, and the development of best practices. Second, systems must be designed to create a learning environment in which stakeholders can “build community” that taps into shared understandings of legitimacy among participants. These understandings in turn often stem from broader legitimating norms globally. One value, fairness, stands out as important, but depends less on formal procedures than on a sense that weaker actors have a reasonable ability to influence outcomes and that mechanisms are in place to ensure, for all actors, a sense of ownership of decisions that affect them. As the FSC and MSC examples indicate, and best practices developed by organizations such as ISEAL dictate, this is currently a priority, although challenging to achieve in practice. Virtually all systems identified in Appendix I are involved in processes to enable better access, transparency, and a sense of ownership among disempowered players, especially firms and social groups in the South, and small producers generally.
Our framework makes three theoretical contributions. First, it reveals the significant influence of social structure on the emergence and institutionalization of NSMD systems. Second, it demonstrates the limits of static analyses where preferences are treated as given, and demonstrates how the interactions of firms and social and environmental actors can create new identities and shared norms that drive the evolution of NSMD governance systems. As our analysis suggests, and empirical examples illustrate, many NSMD arrangements are engaged in legitimating processes that contain elements of logics of “appropriateness” and “argumentation” in which stakeholders and targeted actors can discuss, argue, and deliberate in increasingly legitimate arenas about NSMD governance and standards.
This last point reveals a contribution to the ongoing dispute in International Relations theory over which “logic of action” takes priority. Our analysis supports the growing recognition that both a “logic of consequences” and a “logic of appropriateness” are almost always at play, but one or the other may appear to take priority in different contexts. In the case of NSMD governance systems, the need to pay attention to the explanatory power of a logic of appropriateness increases as institutionalization progresses. This proposition, built upon our preliminary empirical assessment, is especially significant as the governance in question takes place in the global marketplace, a realm in which utilitarian logic is assumed to dominate.
In addition, our framework suggests a number of implications for the future of NSMD systems. First, their transformative potential to socially embed markets is greater than that of other non-state, hybrid, and voluntary initiatives with which they have been conflated. However, if we are right that political legitimacy ultimately rests on community building, achieving it will be much more challenging for NSMD systems than for other forms of private authority. NSMD systems engage a wider array of stakeholders with multiple identities (producers, consumers, environmentalists), geographic locations, and interests. Thus, actors within an NSMD network are more likely to disagree on which performance criteria produce legitimacy and their relative importance vis-à-vis procedural norms.
Second, our framework suggests that analyses based solely on utilitarian firm-level responses severely underestimate the potential of NSMD systems and misrepresent their wider transformative goal of socially embedding the global marketplace. Such analyses would erroneously conclude that while NSMD systems are currently doing some good in niche markets, the prospects for widespread adoption are minimal. Instead, they will serve a purpose mostly of rewarding companies already engaging in good practices or easing the guilt of consumers purchasing those products.
To correct for this truncated picture, which is not supported by the empirical record, our analytical framework identified processes through which authority is generated. While we are not Pollyannish about the prospects for successful transformation – the necessary conditions present formidable hurdles – preliminary evidence of the evolution of some systems in this direction justified our attempt to better understand how and whether such transformations might occur. The analytical framework and preliminary set of theoretical propositions developed here positions the next step for empirical research: theory testing that includes a systematic assessment of the full range of actor motivations, and the factors that influence them, within and across NSMD communities as these systems evolve.
We thank, for their helpful comments, David Apter, Graeme Auld, John Braithwaite, Tim Bartley, Marion Chertow, Will Coleman, Michael Conroy, Claire Cutler, Radislov Dimitrov, Dan Esty, Thorsten Goebel, Neil Gunningham, Carmen Huckel, Stefan Kaufman, Stephen Kellert, Nat Keohane, Jonathan Koppel, Kelly Levin, Erin Mansur, Constance McDermott, Errol Meidinger, Sheila Olmstead, Gus Ranis, Gus Speth, Volker Rittberger, David Vogel, four anonymous reviewers, and the editors of this journal. We thank Erin Hannah, Radhika Dave, Teresa Kramarz, and Larissa Yocom for valuable research assistance and comments. Financial support for this research came from the Social Sciences and Humanities Research Council of Canada, the USDA’s National Research Initiative, the Ford and Doris Duke Charitable foundations, and the Rockefeller Brothers Fund.
This figure was derived from WTO (2003) by dividing the total trade in sectors represented in Appendix I with total global trade.
Examples include the UN Global Compact, the Organization for Economic Cooperation and Development’s Guidelines for Multinational Enterprises (revised in 2000), the Global Reporting Initiative, and, until 2005, the chemical industry’s Responsible Care program.
European forest owners originally created the PEFC as an “umbrella”, “mutual recognition” program for national initiatives developed to compete with, or preempt, the FSC.
Most of the commercial forest lands in the US, Canada, and Europe are under a third-party certification system and concerted efforts are under way to expand support for certification in developing countries (Cashore et al. 2004).
We do not equate law and legitimacy, but law can be an important source and indicator of legitimacy.
Still, sovereignty has worked against legitimacy at the national level in some developing countries where some NSMD schemes are seen as reflecting Northern interests.
Personal interview, senior executive of the Forest Stewardship Council, 12 January 2006.
The impetus for the proposal came from the Clinton administration as part of its promotion of labor standard certification, which eventually evolved into the NSMD system FLO.
Our depiction of firms’ values and environmental and social groups’ values are stylized in order to capture the core conflict within NSMD communities.
Functional legitimacy corresponds roughly to Suchman’s (1995) concept of “pragmatic” legitimacy, where actors evaluate an organization based on whether it produces benefits for an audience, is responsive to its larger interests, or reflects its values.
Ontologically, our approach embraces the notion that markets are always socialized to some degree (Polanyi 1944). We are concerned with how changes in the logic of appropriateness can modify the context in which actors define their interests and identities.
In practice, some firms may be targeted as programs emerge, but for analytical reasons we address targeting as part of phase II.
A general empirical finding in all sectors represented in Appendix I is that firms that are only required to give “preference to” certified products in their procurement policies, but whose own practices are not the target of NSMD systems, are much more likely to give support than firms required to undergo behavioral changes.
Black markets could still exist, but they are a problem not specific to non-state governance.
We offer these as plausible but not exclusive explanations for the movement from phase II to phase III.
Examples of non-state market driven (NSMD) governance systems
Year FLA established a system in which auditors verified samples of companies for compliance. In 2002, the FLA Board, rather than companies, decided which companies to monitor and assigned monitors to them.
FLO united 15 separate initiatives, the first of which was the 1988 Fair Trade Initiative based in the Netherlands.
Founded in 1972, but gradually evolved into an NSMD system. In 1997 established an arm’s-length body to accredit certifiers.
SAN was created in 1992 to develop standards and pave the way for Rainforest Alliance certification of agricultural products, the first of which occurred in 1993. Its first coffee certification took place in 1996.
1994 was the year that the SFI provided for a third-party verification component.
Effort to unify disparate ecotourism programs operating globally.