Governance without a state: Can it work?


Tanja A. Börzel, Research Center “Governance in Areas of Limited Statehood,” Freie Universität Berlin, Binger Str. 40, 14195 Berlin, Germany. Email:


In this article we explore how much state is necessary to make governance work. We begin by clarifying concepts of governance and the “shadow of hierarchy” and we follow this clarification with a brief overview of empirical findings on governance research in developed countries. We then discuss the dilemmas for governance in areas of limited statehood, where political institutions are too weak to hierarchically adopt and enforce collectively binding rules. While prospects for effective policymaking appear to be rather bleak in these areas, we argue that governance research has consistently overlooked the existence of functional equivalents to the shadow of hierarchy. We assert that governance with(out) government can work even in the absence of a strong shadow of hierarchy, we identify functional equivalents to the shadow of hierarchy, and we discuss to what extent they can help overcome issues of legitimacy and effectiveness in areas of limited statehood.


Since the 1990s, the literature on governance within and beyond the state has focused on non-hierarchical modes of coordination and the involvement of non-state actors in the formulation and implementation of public policies. The participation of non-state actors in public policymaking was supposed to improve both the quality of public policies and the effectiveness of their implementation since rule addressees could bring in their expertise and their interests (e.g. Reinicke 1998; Reinicke et al. 2000). In the 1990s, this argument was introduced into the governance literature by students of international relations and European politics, who have been discussing governance without government (e.g. Rosenau & Czempiel 1992; Peters & Pierre 1998; Cutler et al. 1999; Hall & Bierstecker 2002; Grande & Pauly 2005) and new modes of governance (Rosenau 2000; Héritier 2002; Héritier & Lehmkuhl 2008b) as alternatives to the traditional top-down, command-and-control approach of hierarchical steering by government.

In the meantime, however, empirical research has demonstrated that non-hierarchical coordination and the involvement of non-state actors do not necessarily hold their promise to increase the effectiveness and the legitimacy of public policymaking. Findings with regard to the Organisation for Economic Co-operation and Development (OECD) world including the European Union (EU) and Eastern Europe show that “governance with(out) government” is mostly likely to be effective if a strong state looms in the background which sees to it that non-state actors contribute to the provision of collective goods (Héritier & Lehmkuhl 2008b; Börzel 2009a; Héritier & Rhodes forthcoming in 2010). Such a “shadow of hierarchy” provides a crucial incentive for both governments and non-state actors to engage in non-hierarchical rulemaking and service provision (Mayntz & Scharpf 1995b).

But if fully consolidated statehood is indeed a precondition for effective governance with(out) government, a dilemma results: On the one hand, it is less clear why non-state actors should contribute to governance when strong states can deliver the goods. On the other hand, if the state is too weak, governance with(out) government is all the more necessary for public policymaking, but it is unlikely to be effective under these circumstances. As most countries outside the global North do not enjoy consolidated statehood, governance in those countries would be seriously hampered in providing much-needed collective goods. Is there a way out of this dilemma?

We argue in this article that there are functional equivalents to the shadow of hierarchy cast by a strong state. First, external actors such as international organizations or states might substitute for a lacking shadow of hierarchy in “areas of limited statehood.” Second, social norms of local, national, or international communities often create a strong logic of appropriateness (March & Olsen 1998) so that the reputation of non-state actors to contribute to governance is at stake. We conclude therefore that governance with(out) government can work even in the absence of a strong shadow of hierarchy.

This article is mostly conceptual and uses empirical examples mainly for illustrative purposes. We proceed in the following steps. We start by clarifying concepts of governance and the “shadow of hierarchy,” and we follow this clarification with a brief overview of empirical findings on governance research in developed countries. We then discuss the dilemmas for governance in areas of limited statehood where political institutions are too weak to hierarchically adopt and enforce collectively binding rules. While prospects for effective policymaking appear to be rather bleak in these areas, we argue that governance research has consistently overlooked the existence of functional equivalents to the shadow of hierarchy. In the final part of the article we present such equivalents and discuss to what extent they can help overcome the governance dilemma in areas of limited statehood.

Governance with or without government: Conceptual clarifications

Following the work of Renate Mayntz and others, we define governance as the various institutionalized modes of social coordination to produce and implement collectively binding rules, or to provide collective goods (e.g. Mayntz 2004, 2009; for a general discussion see Benz et al. 2007; Schuppert & Zürn 2008). Thus, governance consists of both structure and process. Governance as structure relates to institutions and actor constellations. The literature usually distinguishes between the state, competition systems, and networks (negotiation systems). As to actor constellations, state governance involves exclusively governments, while both competition and negotiation systems consist of configurations of state and non-state actors (firms, interest groups, non-governmental organizations [NGOs], and so forth).

Governance as process pinpoints the modes of social coordination by which actors engage in rulemaking and implementation and in the provision of collective goods. Hierarchical coordination usually takes the form of authoritative decisions with claims to legitimacy (e.g. laws, administrative ordinances, court decisions; for a discussion regarding hierarchy in international relations, see Lake 2009). Hierarchies are based on institutionalized relationships of domination and subordination, which significantly constrains the autonomy of subordinate actors.

Non-hierarchical coordination, by contrast, is based on voluntary commitment and compliance. Conflicts of interest are solved by negotiation. Voluntary agreement is either achieved by negotiating a compromise and granting mutual concessions (side-payments and issue-linkage) on the basis of fixed preferences (bargaining), or actors engage in processes of non-manipulative persuasion (arguing) through which they develop common interests and change their preferences accordingly (Benz 1994, pp. 118–127; Risse 2000). Actors may differ with regard to their bargaining power, but no actor is subject to the commands of others.1 Non-hierarchical coordination can either be formalized in negotiation systems or it can be organized in informal networks. Public private partnerships (PPP) are a typical example of governance institutions based on non-hierarchical coordination (e.g. Rosenau 2000; Schäferhoff et al. 2009).

For some scholars, markets are not usually regarded as governance because they are spontaneous orders (Hayek 1960, 1973) that leave “no place for ‘conscious, deliberate and purposeful’ efforts to craft formal structures” (Williamson 1996, p. 31). Yet market mechanisms can be institutionalized to coordinate actors’ behavior through competition (Benz 2007). In our article we use the concept of competition systems to describe the institutionalization of market-based modes of political coordination.

To sum up, in our understanding, governance includes hierarchical steering by state actors, but also includes the involvement of non-governmental actors (companies, civil society) in the provision of collective goods through non-hierarchical coordination. This coordination range from consultation and cooptation, delegation, and/or co-regulation/co-production to private self-regulation inside and outside the control of governments. Non-hierarchical coordination can involve governmental actors as long as they refrain from using their coercive powers. Some authors therefore distinguish between governance by, with, and without government (Zürn 2002; see Fig. 1).

Figure 1.

Governance with(out) government: the non-hierarchical involvement of non-governmental actors.
Source: Based on Börzel and Risse (2005).

To avoid conceptual overstretch, certain forms of interest intermediation remain outside our definition. Governance without government does not cover lobbying and mere advocacy activities of non-governmental actors aimed at governments as well as at supranational and international organizations. Non-state actors who are not active participants in negotiating or competition systems pose few challenges to existing concepts and theories in political science and international relations. Also excluded from our definition are those arrangements among non-governmental actors that are based on self-coordination and do not aim for the provision of common goods and services (markets; see above); and those that produce public goods and services as unintended consequences (e.g. private security companies) or produce public “bads” (mafia, drug cartels, transnational terrorism).

The “shadow of hierarchy”

The various institutionalized rule structures (hierarchy, negotiation systems, competition systems) with their dominant modes of coordination are ideals that hardly exist in reality. Rather, we find combinations, both within and beyond the state (Benz 2001, pp. 175–202; Börzel 2010b).

It is frequently overlooked in the literature that non-hierarchical modes of coordination, such as negotiation and competition systems, are often embedded in hierarchical structures. In the modern state of the highly industrialized OECD world, government, business, and civil society almost always negotiate under a “shadow of hierarchy” (Mayntz & Scharpf 1995b; Scharpf 1997). This means that the state threatens – explicitly or implicitly – to impose binding rules or laws on private actors in order to change their cost–benefit calculations in favor of a voluntary agreement closer to the common good rather than to particularistic self-interests. “The Damocles sword of threatened direct state intervention” is necessary to induce collective private self-regulation or public-private co-regulation (Schmitter & Streeck 1985, p. 131). Take the case of telecommunications: Regulatory agencies closely monitor pricing and competition among private telecom firms to make sure they provide public services of sufficient quality.

A shadow of hierarchy is important for governance with(out) government because it generates important incentives for cooperation for non-state actors. “Legislators can threaten to enact adverse legislation unless potentially affected actors alter their behavior to accommodate the legislators’ demands” (Héritier & Lehmkuhl 2008a, p. 2). Companies prefer self-regulation (e.g. in reducing environmentally harmful effects) over legally binding regulation because of the greater flexibility and the influence they can exert on policy outcomes (Boddewyn 1992; Héritier & Eckert 2008). Moreover, the possibility of hierarchical intervention reduces the incentive for actors to renege on their voluntary commitment. Business associations, for instance, seldom have sufficient sanctioning capacities to deter opportunistic behavior of their members in the implementation of voluntary agreements (the free-rider problem). Therefore, we rarely find effective societal self-coordination without the involvement of state actors that have the capacity for taking and enforcing collective decisions (Mayntz & Scharpf 1995a). In countries with corporatist structures of interest intermediation, business and labor are free to negotiate wage agreements that are, however, endorsed and enforced by the government. Finally, opportunistic behavior of non-state actors is rendered less likely if governments review the negotiation outcomes to ensure they correspond to the common good. This is most important if companies, professional associations, pressure groups, or consultancies are involved. Unlike governments and not-for-profit organizations (e.g. NGOs), they are not obliged by formal institutions or social norms to pursue the common good.2 German health insurance companies, for instance, can set their insurance premiums; the federal government can, however, intervene to ensure affordable health care.

If the shadow of hierarchy provides an important incentive for non-state actors to cooperate in the provision of rules and collective goods, the willingness of non-state actors to engage in governance with and without government should increase with the degree to which state actors can resort to hierarchical modes of governance. For governments, it is exactly the reverse: the higher the government's capacity for hierarchical policy-making, the fewer incentives it has to cooperate with non-governmental actors. Governance research (implicitly) assumes that governments are genuinely problem-oriented. This “functionalist fallacy of governance research” (Mayntz 2004, p. 71, our translation) neglects the fact that governments seek to increase or at least to maintain their autonomy as well as their problem-solving capacity in the policy process. Because the cooperation with state and non-state actors entails a significant loss of autonomy, state actors are only inclined to engage with private actors if they gain and regain problem-solving capacity rather than use hierarchical modes of coordination.

To sum up, the shadow of hierarchy provides both governments and non-state actors an important incentive structure for cooperation, albeit in different ways (Fig. 2). The stronger the shadow of hierarchy, the more non-state actors have incentives to cooperate in the provision of collective goods. In contrast, the incentive structure for governments to cooperate with non-state actors is more curvilinear. On the one hand, weak states are unlikely to engage in governance with non-state actors because they might fear a loss of autonomy vis-à-vis society (agency capture). On the other hand, strong states are unlikely to share governance authority with non-state actors. The two curves meet in the middle; that is, a medium shadow of hierarchy is most likely to yield effective and problem-solving governance, once non-hierarchical modes of coordination and non-state actors are involved.

Figure 2.

The shadow of hierarchy and diverse incentives for cooperation for governments and non-state actors.

Empirical research on “new modes of governance” in the highly industrialized OECD countries as well as in Eastern Europe has confirmed this proposition. Mayntz and Scharpf, for instance, showed that important initiatives of private self-regulation collapsed the moment the credible threat of legal intervention by the state ceased to exist (Mayntz & Scharpf 1995a; see also Héritier & Lehmkuhl 2008b; Börzel 2009a; Héritier & Rhodes forthcoming in 2010). Business actors, in particular, engage in public policymaking but only if national governments and EU institutions are powerful enough to cast a credible shadow of hierarchy. This explains why the role of business actors in the provision of collective goods has remained limited at the EU level and in the new member states.

Governance in areas of limited statehood: A dilemma

So far, our discussion has been confined to theoretical and empirical insights from governance research in the developed world of liberal democracies; that is, in countries in which the ability of the state to ultimately enforce collectively binding decisions is not challenged. Let us now move to areas of limited statehood which by definition lack such a credible shadow of hierarchy.

The essence of statehood is the ability of the state to enforce collectively binding decisions, ultimately through coercive means, via its monopoly over the means of violence. Thus, the Western literature draws an implicit link between hierarchy and the state. It follows closely Max Weber's conceptualization of statehood as an institutionalized authority structure with the capability of hierarchically steering (Herrschaftsverband) and legitimately controlling the means of violence (Gewaltmonopol) (Weber 1921/1980). While no state governs hierarchically all the time, states at least possess the ability to authoritatively make, implement, and enforce central decisions for a collectivity. In other words, states command what Stephen Krasner has called “domestic sovereignty;” that is, “the formal organization of political authority within the state and the ability of public authorities to exercise effective control within the borders of their own polity” (Krasner 1999, p. 4). Moreover, states possess “international sovereignty” based on mutual recognition between territorial entities that have formal juridicial independence (Krasner 1999, p. 3).

The insights into the shadow of hierarchy and the propensity of governments and non-state actors to cooperate in the provision of collective goods and services mostly result from countries that enjoy consolidated statehood; the ability to enforce the law if need be. Yet consolidated statehood is the exception rather than the rule in the contemporary international system also seen from a historical perspective (Beisheim et al. 2007; Risse 2010). Outside the world of developed and highly industrialized democratic states, most countries contain what we call “areas of limited statehood.” While areas of limited statehood still belong to internationally recognized states (even the failed state Somalia has international sovereignty), it is their domestic sovereignty which is severely circumscribed. In other words, areas of limited statehood include those parts of a country in which central authorities (governments) lack the ability to implement and enforce rules and decisions or in which the legitimate monopoly over the means of violence is lacking, or both, at least temporarily. The ability to enforce rules or to control the means of violence can be restricted along various dimensions: territorially; sectorally (i.e. with regard to specific policy areas); socially (i.e. with regard to specific parts of the population); and temporarily. As a result, we can distinguish different configurations of limited statehood.

If we conceptualize limited statehood in such a way, it becomes clear that areas of limited statehood both are prevalent in the international system nowadays and have been prevalent in the past. After all, the state monopoly over the means of violence has only been around for a little more than 150 years. Most of the world's current states contain “areas of limited statehood” in the sense that central authorities do not control the entire territory, do not fully possess the monopoly over the means of violence, and/or have limited capacities to enforce and implement decisions, at least in some policy areas or with regard to large parts of the population. This is what Somalia, Brazil, and Indonesia, but also China, have in common. While China, for example, is able to control most parts of its vast territory, its government lacks the capacity to enforce its own laws, particularly with regard to environmental protection (for details, see Thauer 2009). Last but not least, we need to mention the modern protectorates here, such as Afghanistan, Kosovo, and Bosnia-Herzegovina – internationally recognized states that lack “Westphalian” sovereignty in the sense that external actors rule parts of their territory or in some policy areas (Krasner 1999).

“Limited statehood” is not confined to failing and failed states that have all but lost the ability to govern and to control their territory. Failing and failed states comprise only a small percentage of the world's areas of limited statehood. Most of the developing and transition states contain areas of limited statehood insofar as they only partially control the instruments of force and are only partially able to enforce decisions, mainly for reasons of insufficient political and administrative capacities. Thus, we do not talk about “states of limited statehood,” but areas – territorial or functional spaces within otherwise functioning states that have lost their ability to govern. While the state of Pakistan, for example, enjoys a monopoly over the use of force in large parts of its territory, the so-called tribal areas in the country's north-west are beyond the control of the central government. Mexico enjoys mostly consolidated statehood, but the central authorities are too weak to enforce human rights and the rule of law with regard to public security and policing in various quarters of Mexico City. State institutions themselves are often the source of insecurity and other public “bads” (Braig & Stanley 2007; Müller 2009).

The more a country contains areas of limited statehood, however, the weaker is the shadow of hierarchy, by definition.3 If a state cannot enforce the rules in parts of its territory or with regard to some policy sectors, it cannot cast a credible shadow of hierarchy either. According to the research on governance with and without government discussed above, this results in a serious dilemma for areas of limited statehood. On the one hand, the lower the capacity of the state, the greater is the need for governance through non-hierarchical modes involving non-state actors to compensate for government weakness or state failure. On the other hand, limited statehood implies a weak shadow of hierarchy as a result of which such “new” and non-hierarchical modes of governance are unlikely to emerge and be effective. To sum up, the more such “new” modes of governance are necessary, the less likely they are to be sustainable (Börzel 2009b). So, are areas of limited statehood doomed because weak state institutions result in governance failure too? Would governance and the provision of collective goods then be impossible in most of the developing world?

Empirical evidence suggests the answer to those questions is “no.” Weak or limited statehood does not automatically translate into weak governance. “Governance without a state” (Risse & Lehmkuhl 2007; Risse 2010) appears to be an empirical reality in many parts of the world. Take the most extreme examples, namely, failed states and war-torn societies. For the past fifteen years or more, Somalia has not enjoyed functioning statehood. Yet parts of the country, namely, the northern province of Somaliland, have been well governed throughout this period with rather low levels of violence (e.g. Menkhaus 2006/2007; Debiel et al. 2010). As Sven Chonacki and Zeljko Branovic argue, even warlords sometimes provide security as a public rather than a private good under specific circumstances (Chojnacki & Branovic 2010). Likewise, multinational companies police local communities, voluntarily implement environmental protection standards, provide HIV/AIDS-related services, or agree to use sustainable energy (Hönke et al. 2008; Hönke 2010; Müller-Debus et al. 2010). Moreover, PPPs, including Western development agencies, international organizations, firms, and NGOs, are heavily engaged in areas of limited statehood, providing services in the realms of public health, development, and environment (Beisheim et al. 2008; Liese & Beisheim 2010). PPPs such as GAVI (which provides immunization to children) or the Global Fund to Fight Aids, Tuberculosis, and Malaria invest billions of US dollars for these purposes and all but outspend international organizations such as the World Health Organization (WHO) in their area of governance (Schäferhoff in preparation). Multinational mining companies in the Democratic Republic of the Congo, such as the US company Freeport McMoRan, manage social relations in adjacent communities by engaging in development projects together with donor agencies and NGOs, often bypassing state institutions (Hönke 2010).

How can we explain these findings? Why does governance research in Western developed countries show that “new” modes of governance require consolidated statehood and a strong shadow of hierarchy, while “governance without a state” appears to be widespread in areas of limited statehood? We suggest in the following that the link between statehood and the shadow of hierarchy is weaker than is often assumed in the governance literature. More importantly, there are functional equivalents for a state-based shadow of hierarchy that enable effective and sustainable non-hierarchical modes of governance involving non-state actors in areas of limited statehood.

Functional equivalents to the shadow of hierarchy

Consolidated statehood is a prominent way but not the only way to generate a shadow of hierarchy. Possible alternatives can be conceptually distinguished according to the underlying logic of social action; whether they rely on a logic of consequences or a logic of appropriateness (March & Olsen 1989; March & Olsen 1998). According to the logic of consequences, self-interested and utility-maximizing actors are likely to contribute to governance given the right incentives and/or if those actors are embedded in institutional settings constraining them. Below, we discuss two alternatives to the shadow of hierarchy using the logic of consequences, namely, the risk of anarchy on the one hand and the involvement of external actors able to cast a shadow of hierarchy “from afar” on the other.

According to the logic of appropriateness, actors are embedded in normative structures that induce them to “do the right thing” and to follow social rules. Again, we discuss two alternatives to the shadow of hierarchy relying on this logic, namely, how normative structures affect and embed markets on the one hand and how local community norms lead to governance on the other. We focus empirically on a particular type of non-state actor, firms whose self-interested behavior is constitutive and whose primary function is to provide private goods rather than to engage in governance. Thus, firms are a “hard case” for governance in areas of limited statehood (Flohr et al. 2010). Why should they invest in the common good and contribute to governance, particularly under conditions of high political instability and legal uncertainty, including unclear property rights, all of which are all too common in areas of limited statehood?

The risk of anarchy

While the shadow of hierarchy provides a key incentive for non-state actors to engage in governance and the provision of collective goods, the same might hold true for the exact opposite, namely when there is no political order. If the state is not capable of adopting and enforcing collectively binding decisions, companies are not confronted with a situation in which they have to weigh the costs of participation in governance and voluntary commitment against the possibility of suboptimal and hierarchically imposed policies. Rather, they face the danger of entirely absent governance. If the pursuit of their individual profits depends on the provision of certain common goods and collectively binding rules to produce them, and the state is not capable of or is unwilling to provide these goods, companies have a major incentive to step in and provide governance in areas of limited statehood. Take the case of HIV/AIDS in South Africa (Börzel et al. forthcoming). Multinational companies of the automotive industry, such as BMW, Mercedes Benz and General Motors, require skilled labor that suffers from the HIV/AIDS pandemic. In the absence of a functioning public health system, including public education on the risks of the pandemic, companies have been stepping in to provide health and education services not only for their workers but also for their families and the larger community. Mercedes Benz trains doctors and nurses in East London, South Africa regarding the treatment of HIV/AIDS who are then sent to raise the general level of public health services offered in the public hospitals of the particularly poor province of the Eastern Cape in South Africa (Müller-Debus et al. 2009). In this case, the firms’ self-interest in human capital rather than any orientation toward the public good provides strong incentives to engage in health and education governance in the absence of a functioning state (Thauer 2009; Müller-Debus 2010).

The anarchy problem in areas of limited statehood closely resembles the international system in the absence of an enforcer or a hegemon (Mayntz and Scharpf 1995b: Fn. 4). Transnational or global governance has to cope with the problem that there is no world state to ensure compliance with costly rules. Rather, in the case of international regimes, individual states have to voluntarily enforce norms and rules (Krasner 1983; Rittberger 1993). At the same time, the various international institutionalisms (see overviews in Keohane 1989; Hasenclever et al. 1997; and Simmons & Martin 2002) have convincingly demonstrated that “cooperation under anarchy” (Oye 1986) is indeed possible if and when egotistic actors cannot reach their goals unilaterally or if and when their unilateral action produces negative externalities for everybody involved. In other words, governance and the provision of collective goods is possible even in the absence of an enforcer who can cast a credible shadow of hierarchy.

At the same time, the institutionalist literature has also identified scope conditions for the provision of effective governance in world affairs. The literature on “legalization,” for instance, has argued that compliance with costly rules is all the more likely the better the norms and rules are defined (precision), the higher the degree of obligation, and the more adjudication of compliance is referred to independent monitoring or even judicial systems (delegation) (Abbott et al. 2000; Goldstein & Martin 2000; Zangl & Zürn 2004). This argument was evaluated with regard to transnational PPPs providing governance to reach the United Nations Millenium goals in areas of limited statehood: it held up pretty well against the empirical evidence, particularly in the case of service providing partnerships that involved large amounts of financial resources (Beisheim et al. 2008; Liese & Beisheim 2010). The more institutionalized the PPP, the higher the PPP's effectiveness in contributing to governance in areas of limited statehood.

In sum, two scope conditions provide incentive structures to ensure that non-state actors engage in effective and sustainable governance in areas of limited statehood: the risk of anarchy and highly institutionalized settings.

External actors compensating for limited statehood

The state is not the only source to cast a shadow of hierarchy inducing actors to provide governance in its home territory. External actors, such as international organizations and foreign governments, can commit non-state actors to participate in effective governance. First, external actors may directly exercise domestic sovereignty, holding the monopoly over the means of violence, as well as enforce decisions authoritatively. Prominent examples for such areas of limited statehood are modern protectorates from Bosnia to Afghanistan – states that have lost their “Westphalian” sovereignty or in which it is severely circumscribed. The emerging international norm of the “responsibility to protect” can be invoked if a state is neither willing nor able to provide even a minimum degree of governance. As a result, the international community has at least the legal right to intervene and to provide governance if everything else fails. It is yet to be seen whether foreign actors can exercise sustainable and effective governance under these circumstances (for conflicting views, see Fearon & Laitin 2004; Paris 2004; Zisk Marten 2004; and Schneckener 2010). However, if a state's “Westphalian” sovereignty is constrained through direct interference by the international community, external actors are more likely to cast a credible shadow of hierarchy the more they actually control the means of violence and the more effectively they can enforce decisions.

Second, under international law, NGOs, companies, and local actors can be obliged to comply with standards of good governance in areas of limited statehood (Ladwig & Rudolf 2010). The legal principle of agency of necessity (negotiorum gestio; see Bühring & Hüfken 2008) implies that one actor acts in the place of another who is unable to take the necessary action.

Third, actors who provide governance are obligated by the same rules as the state government that is no longer able to meet international standards. In other words, there are standards of international law that hold non-state actors directly accountable to provide governance under conditions of limited statehood. Whether this results in effective and sustainable governance, however, is an altogether different question. The problem is who enforces international law in areas of limited statehood. In the absence of an effective state, direct obligations of non-state actors under international law to provide governance might not matter much. Unless a company, for example, feels committed to international law for other reasons (the logic of appropriateness or reputational concerns, see below), enforcement of international law has to rely on other mechanisms, such as the ones described in this article.

Finally, national governments of (consolidated and democratic) states in which multinational companies and transnational NGOs have their headquarters may also force non-state actors to contribute to governance in areas of limited statehood. In this case, home country laws are in place that require non-state actors such as companies to comply with standards of good governance or other regulations (e.g. environmental laws) irrespective of where they invest or act. While their scope is still limited, home country regulations can help commit companies in particular to provide governance in areas of limited statehood (Prakash & Potoski 2007; Greenhill et al. 2009; Thauer 2009; Deitelhoff & Wolf 2010, p. 212–214; Flohr et al. 2010, p. 52–66; Börzel et al. forthcoming). There are two causal mechanisms linking home country regulations to companies’ contributions to governance in areas of limited statehood. Companies can be legally bound to comply with regulatory norms of their home country when they invest abroad. While the scope and enforcement is still limited, US extra-territorial jurisdiction on human rights, bribery, and national security issues shows the potential for national governments to regulate their companies abroad (Flohr et al. 2010).

The second mechanism is more indirect and works through economies of scale, transaction costs, or market considerations in a globalized economy. Firms producing for markets with a high degree of regulation (e.g. in the realm of environmental standards) and targeting high-end consumer markets have little incentive to use different production standards in areas of limited statehood, thereby, for example, polluting the environment. They tend to transport their regulatory standards abroad if investing in foreign countries (Börzel et al. forthcoming). In this way, production networks in a globalized economy result in firms providing governance, adopting, and even lobbying for environmental and social regulation going beyond legal requirements rather than “racing to the bottom” in areas of limited statehood (Prakash 2000; Haufler 2001; Vogel & Kagan 2004; Greenhill et al. 2009). There are, however, limits to the extent to which home country regulations can be used to induce firms to provide governance in areas of limited statehood (Flohr et al. 2010). These rules are often limited to product regulation (Vogel & Kagan 2004) and rarely extend to supply chains beyond the first or second supplier (Héritier et al. 2009).

The provision of an external authority casting a shadow of hierarchy points to the opposite mechanism as the one identified above as the “risk of anarchy.” While the latter is defined by the absence of a state, a shadow of hierarchy still requires consolidated statehood; for example, in the home country of a multinational corporation. The two alternatives to domestic sovereignty may complement and actually reinforce each other. In this case, the multi-level nature of governance in areas of limited statehood provides for a functional equivalent to consolidated statehood's ability to cast a shadow of hierarchy. Binding firms and other non-state actors to international law as well as to home country regulations does not require formal intrusion into a state's “Westphalian” sovereignty (Krasner 2004). Such intrusion usually meets stiff resistance, especially by weak and fragile states.4 Rather, the international community and/or Western consolidated states see to it that non-state actors such as firms and NGOs engage in governance and provide public services in areas of limited statehood.

Norms and socially embedded markets

The following two functional equivalents to the shadow of hierarchy cast by consolidated statehood rely on the logic of appropriateness and rule-guided behavior rather than on the logic of consequences. To begin with, NGOs and transnational social movements often launch international campaigns naming and shaming companies who fail to contribute to the provision of common goods in areas of limited statehood. As a result, environmental and human rights norms have started to creep into the core business of many companies, particularly multinational corporations with a “brand name” to defend and whose products target markets in (consolidated and democratic) states in which consumers care about the provision of common goods (e.g. Potoski & Prakash 2006; Prakash & Potoski 2007; Héritier et al. 2009; Flohr et al. 2010). What is now being called “corporate social responsibility” requires firms to integrate environmental and human rights norms into their production, management, and general business practices. In many cases, companies had been subjected to NGO campaigns including consumer boycotts. Today, more and more companies have integrated these norms into their business practices, including their risk management, even though their compliance records still vary enormously. A recent study shows that NGO campaigns are particularly effective when they target firms with a strong brand name, whereas they are much less influential when they target firms that lack such an identity (Thauer 2009). Empirical research suggests that the provision of collective goods in areas of limited statehood depends heavily on whether the particular firm in question has a brand name to defend (Hönke et al. 2008; Deitelhoff & Wolf 2010; Flohr et al. 2010; Börzel et al. forthcoming).

Of course, the incorporation of environmental and human rights standards into their practices does not turn firms into charities. Their business is still doing business, as Milton Friedman's famous dictum suggests. However, companies increasingly realize that their markets are socially embedded and that their customers care about the provision of common goods. As footwear companies such as Nike learned the hard way, their American and European customers did not want to buy running shoes produced through child labor in the Philippines. Having learned their lesson, some companies with vested interests in corporate social responsibility standards even regulate and control their supply chain to ensure a clean production (Héritier et al. 2009). Such inspection activities are another example of a functional equivalent to the shadow of hierarchy cast by the threat of state regulation. In addition, more and more pension funds have started investing in companies that conform to certain social standards or are listed in various social indices such as the Dow Jones Sustainability Index. Finally, companies set up rules for environmental and social protection when they fear the loss of idiosyncratic investments: in sophisticated technology or in skilled labor for instance. The automotive manufacturer BMW and its component supplier make use of the environmental production process standard ISO 14000 developed by the International Standards Organisation. While the predominant aim is to increase the transparency of interaction and the reliability of the transaction partner, the two firms also make a governance contribution to environmental protection (Müller-Debus 2010).

To sum up, corporate social responsibility (CSR) is an interesting case in which the logic of appropriateness intersects with the logic of consequences. The more consumers of companies’ products, particularly in high-end markets, care about governance in areas of limited statehood, including human rights and environmental standards (the logic of appropriateness), the more they use market mechanisms to induce firms to comply with these norms (the logic of consequences). This process translates into reputational concerns for companies, particularly those with a brand name to defend (Börzel et al. forthcoming), who care about their reputation even in cases in which there is little evidence for consumer concerns about responsibly produced goods.

Reputational concerns about socially accepted behavior induce firms to take norms more seriously. Norm compliance can then turn into a strategic advantage in competitive markets leading to a “race to the top” with regard to regulatory standards. A study of South African firms has documented these mechanisms (Börzel et al. 2007, forthcoming; see also Prakash & Potoski 2007; Greenhill et al. 2009). Once again, there is nothing altruistic about this situation. In fact, contributing to governance becomes a matter of self-interest, as it pays off if a firm can set the industry standard with regard to environmental or human rights standards. As Annegret Flohr et al. argue, firms can thereby turn into norm entrepreneurs (Flohr et al. 2010). Yet the mechanism only works if consumers – mostly in the home countries of the multinational corporations – care about social norms and good governance in areas of limited statehood and are prepared to pay a certain price for it. To sum up, socially embedded markets can provide a functional equivalent for the shadow of hierarchy.

“Traditional” normative structures

Social norms of appropriate behavior are not only institutionalized at the international level or in areas of consolidated statehood. Areas of limited statehood are often populated by traditional communities with their own social standards, even if they do not always fully conform to global standards of human rights, democracy, and good governance. For example, while African governments often do not care whether Chinese companies comply with environmental standards, local communities often do (Hackenesch 2009). Likewise, companies may be embedded in local communities defined by clan structures sharing certain standards of appropriate behavior that include the provision of governance. Recent studies on mining operations in Africa and Latin America show how demands for benefits from resource extraction in the form of the provision of collective goods are increasingly raised by communities at the local level, who lay blame and grievances at the doorstep of companies’ operations (Szablowski 2007; Hönke 2010). The South African mining industry has been exposed to this process in a particular way. Mining companies had a terrible record with regard to environmental pollution and workers’ rights during Apartheid. In post-Apartheid South Africa, local communities in mining towns and NGOs used naming and shaming practices to induce companies to provide governance with regard to social rights and cleaning up the environment. At the same time, companies turned to CSR for rebuilding relationships with the ANC government, for regaining legitimacy in the wider public, and for avoiding state regulation against them (Hamann 2004; Hönke 2010, pp. 257–285). Mining companies are also subjected to considerable peer pressure, as one rotten apple can spoil the reputation of the entire sector, this decreasing share values on the international stock market (Hönke et al. 2008). Once again, this is a case in which concerns about norms affect the utility calculations of an entire industry.

Local communities do not provide a shadow of hierarchy, but they expect non-state actors – companies and NGOs alike – to comply with local governance norms and to contribute to the provision of collective goods, particularly when the central state institutions are too weak (or too corrupt) to govern. Non-compliance can then quickly become costly for the non-state actors, particularly when the multi-level nature of governance in areas of limited statehood comes into play. In many cases, local communities are well connected to transnational advocacy networks of NGOs and international organizations and thus can link up with the global civil society (for the mechanisms by which this happens, see Keck & Sikkink 1998; and Risse et al. 1999). An exemplary case involves indigenous peoples in the Niger delta, Nigeria, who linked up with the transnational human rights and environmental communities to denounce the behavior of the Shell company, which had both ruined the environment and violated the human rights of local communities (Frynas 2000). As a result of these campaigns, Shell is a different company today and has fully embraced the norms of “corporate social responsibility” even though compliance with these norms in the Niger delta is still disputed (Zimmer 2010). The example shows again how the norms of local communities and the social embeddedness of markets can work together to force non-state actors to contribute to governance in areas of limited statehood.

Governance with and without state: Effective and legitimate?

Governance through non-hierarchical coordination involving non-state actors has emerged as an alternative or functional equivalent to governance by the state. It gained prominence in OCED countries when political and social differentiation became major constraints on the effectiveness of hierarchical coordination by the state (Kooiman 1993; Mayntz & Scharpf 1995a; Mayntz 1997; Leibfried & Zürn 2005). It then travelled to international politics in the 1990s as a possible solution to anarchy; that is, the absence of central government (Rosenau & Czempiel 1992). Finally, it arrived in areas of limited statehood, where the state is seriously weakened or nearly absent (Risse & Lehmkuhl 2007; Risse 2010).5 While there seems to be an increasing demand for governance, with and without government, within and beyond the state, its promise to compensate for the weakness or failure of government appears to rest on a major premise (Börzel 2010a). On the one hand, the state must have sufficient capacity in terms of both resources and autonomy to cast a credible shadow of hierarchy so that non-governmental actors have an incentive to engage in governance and governments are not afraid of being established. On the other hand, this capacity must not be too strong so that it can provide an incentive for governments to seek cooperation with non-state actors. In other words, to make governance without government work an intermediate level of shadow of hierarchy is necessary. It is clear that this cannot be taken for granted, at neither the international nor the domestic level. Yet if areas of limited statehood are not to be doomed, we have to look for functional equivalents.

Functional equivalents to consolidated statehood and its shadow of hierarchy have to provide sufficient incentives for non-state actors to engage in the provision of collective goods. Moreover, they have to ensure that these actors remain committed to the common good and are held accountable when they seek to renege and maximize their individual profit instead. Interestingly enough, it may be the absence of the state that provides the key incentive for this behavior. If states are not capable of adopting and enforcing collectively binding decisions, actors face the danger of entirely absent governance. Such risks of anarchy provide a powerful alternative to the shadow of hierarchy in areas of limited statehood. We observe externally generated shadows of hierarchies by international organizations or other states as well as market pressures or community norms that induce non-state actors to participate in governance and the provision of common goods.

The extent to which these functional equivalents can effectively substitute the shadow of hierarchy cast by consolidated statehood in areas of limited statehood is ultimately an empirical question. An increasing number of studies finds that non-state actors may contribute to governance in areas of limited statehood under specific conditions (Greenhill et al. 2009; Deitelhoff & Wolf 2010; Risse 2010). The functional equivalents to the shadow of hierarchy provided by consolidated statehood rely on different social mechanisms based on the rational choice logic of consequences, the logic of appropriateness of rule-guided behavior, or a combination of both. Governance without government is seldom as effective as government in areas of consolidated statehood. Yet at times it is the sole effective regulation that exists (Vogel 2009). Moreover, when combined, the functional equivalents can compensate for some of their individual weaknesses, such as ensuring compliance (Börzel 2010b; Flohr et al. 2010).

But how legitimate is governance without government under these conditions? The governance literature has promoted “new” modes of governance as a solution to legitimacy problems caused by government failure. Involving non-state actors in the provision of collective goods not only allows tapping into their cognitive and financial resources but also helps to ensure effective implementation. The more actors affected by a policy have a say in decisionmaking, the more likely they are to accept the policy outcome to be implemented, even if their interests may not have been fully accommodated. In other words, the more the various stakeholders are involved in decisionmaking, the more legitimate governance should be. This is the major theoretical argument in favor of directly involving non-state actors in governance (e.g. Reinicke 1998; Kaul et al. 1999; Reinicke et al. 2000). The counterargument posits that governance without government amounts to the privatization of state functions and to the overtaking of governance tasks by illegitimate actors such as companies (for a critical view, see e.g. Altvater & Mahnkopf 2002; Cutler 2003).

Neither the optimistic nor the pessimistic views are particularly helpful. A common standard of legitimacy requires that those being governed should have a say in the process of rulemaking (“input legitimacy”) and that governance should provide collective goods enabling a decent life for the community (“output legitimacy”) (see Scharpf 1999 on this distinction). As to output legitimacy, the effectiveness and the problem-solving capacity of governance are at stake. As argued above, both crucially depend on either the provision of a shadow of hierarchy cast by consolidated statehood or – in the case of limited statehood – the presence of functional equivalents such as the risks of anarchy, the presence of external actors, and the norms of socially embedded markets and of local communities. As to input legitimacy, things become more complicated. First, a functioning shadow of hierarchy not only serves to increase the effectiveness of governance involving non-state actors but also provides a “horizon of legitimacy” (Ladwig et al. 2007)– at least in cases of consolidated and democratic statehood. In other words, input legitimacy of governance without government should not be of major concern, as long as functioning democratic statehood monitors whether or not collective goods are actually provided.

Second, with regard to areas of limited statehood, it goes without saying that the “horizon of legitimacy” provided by the state is equally weak. Moreover, the weak state of developing countries is usually not very democratic and the rule of law is equally deficient. As a result, we need to have a close look at the various governance processes and institutions to ascertain their legitimacy. As a rule of thumb, we could establish that governance in areas of limited statehood is more legitimate the more local communities and citizens are involved in rulemaking and the provision of collective goods – provided that this involvement does not result in violations of fundamental human rights. In other words, whoever governs must be held accountable against international legal standards of human rights, the rule of law, and democracy. We have argued above that there are various mechanisms by which non-state actors such as companies can be held accountable to these standards, be it through the laws of their home countries or through the norms of socially embedded markets. The more local communities are linked to transnational civil society and transnationally operating social movements, the more they can ensure that governance by non-state actors remains legitimate by taking the views of the local communities into account.

This standard of legitimacy also requires that the more external actors interfere with the right of self-determination of local communities by constraining their “Westphalian” sovereignty through the exercise of hierarchical steering, the less intrusive their governance should be (Ladwig et al. 2007; Ladwig & Rudolf 2010). The more outsiders interfere with the right of groups at self-determination, the more they must secure that their governance is not irreversible. In other words, state-building by international organizations and others aimed at creating stable institutions faces serious legitimacy problems, particularly if they use coercion (hierarchical coordination) and if those being ruled have little input in governance.

Finally, the inclusiveness of the governance contributions by non-state actors as well as their external effects on others are important questions, which are highly relevant in terms of legitimacy but have received little attention so far (Börzel 2009c). Does the pluralization of governance providers result in a fragmentation of collective goods provision? If so, what are the consequences for other collectivities? How does the pluralization relate to and affect state governance? For instance, if multinational companies in South Africa provide HIV/AIDS treatment to their employees and their families, by building clinics next to their production sites for example, do other members of the local community receive access to the clinics? And even if companies provide common rather than club goods, do they help build the capacity of the South African state to fight the disease or promote the privatization of the public health sector?

All in all, the provision of collective goods through non-hierarchical coordination and the involvement of non-state actors can be both effective and legitimate. However, this does not mean that statehood belongs in the dustbin of history. After all, international standards are made by states, socially embedded markets need some state regulations, and the campaigns of NGOs target consumers living in areas in which the state provides most of the collective goods. Moreover, state actors – whether local, national, regional, or international – are regularly involved in non-hierarchical modes of governance with non-state actors even in areas of limited statehood. At stake then is the transformation of the state rather than its disappearance (Hurrelmann et al. 2007).


We are grateful to Nicole Bolleyer, Nicole Deitelhoff, Nicole Helmerich, Jana Hönke, Andrea Liese, Fritz Scharpf, and Christian Thauer for suggestions for and criticism on earlier versions of this article. We also thank the editors of Regulation & Governance and three anonymous reviewers for their very useful comments. Research for this article was funded by a grant from the German Research Foundation (Deutsche Forschungsgemeinschaft) in the context of the Collaborative Research Center 700 “Governance in Areas of Limited Statehood” (


  • 1

    Heavy power asymmetries can, however, reduce the choices of actors (by imposing prohibitive costs) in such a way that coordination becomes largely hierarchical.

  • 2

    This is not to say that governments are motivated to pursue the common good and always do so. But unlike non-state actors, governments are held accountable by rule of law and democratic procedures and may be subject to legal and/or social sanctions if they abuse their legal or moral authority to pursue private self-interests.

  • 3

    Interestingly enough, transnational or global governance faces similar problems, whether in the case of inter-state agreements or of transnational public private partnerships (Schäferhoff et al. 2009). In the absence of a hegemon who could cast the equivalent of a shadow of hierarchy (cf. hegemonic stability theory, e.g. Gilpin 1981 and Kindleberger 1981), international and transnational agreements must solve compliance problems in order to be effective. We come back to this point below.

  • 4

    This is another governance paradox. Particularly weak and fragile states usually cling to their “Westphalian” sovereignty; that is, they fight tooth and nail against foreign interference with their domestic sovereignty, even though they do not actually have domestic sovereignty.

  • 5

    Historic studies show, however, that governance without government is not a new phenomenon in areas of limited statehood but emerged well before the institutionalization of international norms (Hönke 2010).