Introduction

Anarchy is a special concept for scholars of international relations. Some view it as the core organizing category (Gilpin, 1981, p. 7; Herz, 1959, p. 231; Lake, 2009, p. 2; Oye, 1985, p. 1; Waltz, 1979, p. 88; cf. Donnelly, 2015; Hobson, 2014) of an autonomous discipline of International Relations (IR; Schmidt, 1998) which sets it apart from cognate disciplines such as Political Philosophy and Political Science. Within social theory more broadly, this category has been used in two generic senses (Baldwin, 1993, p. 14; Milner, 1991, pp. 69–70). In the first generic sense, it designates a condition of interaction between multiple individuals without a common superior. Typically, this superior is an institution namely government, so “anarchy” means the absence of common government. This default meaning allows for a host of variations. It may refer, in a narrower sense, to the absence of a state (centralized government with a territorial base and population), and in a broader sense, to the absence of common authority (office of rule) or the absence of a common ruler (the meaning of the Greek term anarkhia). In its second generic sense, anarchy designates chaos and disorder. No conceptual (logically necessary) connection ties the two senses of the term: The absence of government need not entail disordered interaction or vice versa. That order is achievable in an anarchical environment is a premise adopted by certain theories of international anarchy (the original expression is attributed to Dickinson, 1937) or anarchy as it pertains to the relations of states, which is the theme of the present discussion.

International anarchy, in effect, can refer to (a) the absence of world government, (b) international disorder, or (c) international order. The first and the third senses of “anarchy” are central to IR. The last option presents it as the ordering principle of international relations defined as relations between sovereign states. In this third sense, compatible with the first basic sense, (international) anarchy is conceptually (logically) linked to sovereignty. If a state is sovereign internally (is the highest authority in a realm), then necessarily its external relations to other states are relations of sovereign equality or anarchy (in the third sense).

Anarchy so construed is a horizontal arrangement between formal equals to be distinguished from hierarchy, a vertical arrangement between subordinate and superordinate units. It is conventionally accepted that the modern epoch of international relations, traceable to the Treaty of Westphalia (1648), emerged as an anarchical political order between sovereign states. It superseded the hierarchical order of medieval Christendom where a constellation of diverse units—principalities, independent cities, bishoprics—were subordinate to the higher authority of the Holy Roman Emperor in temporal matters and to the Pope, in spiritual matters. Like any other concept, “Westphalian sovereignty” is an abstraction, superimposed on a complex historical actuality (for a genealogy of sovereignty, see Bartelson, 1995; and for the “myth” of Westphalia, de Carvalho et al., 2011).

The idea of anarchy as the absence of central government or state derives from the concept of a “state of nature,” developed by the 17th century philosopher Thomas Hobbes in The Elements of Law (1650/1969) On the Citizen (1642/1647/1998), and refined in Leviathan (1651/1968). The state of nature is a hypothetical condition of statelessness. It can be read in two ways: either retrospectively, as a condition of what human life would be like if the existing civil state would be dismantled; or prospectively, as a condition in which the civil state would be created for the first time. The prospective view of anarchy is the standard one within IR. Among the rest of the classical contractarians such as John Locke and Jean-Jacques Rousseau, Hobbes alone envisages a parallel between the domestic state of nature (among human beings) and the international state of nature (among sovereign states). Hobbes’s international state of nature is the closest analogue of international anarchy in contemporary terms (Hollis & Smith, 1991, p. 102; Waltz, 1959/2001, p. 163).

In providing an analytical review of the scholarly literature on international anarchy (for a historical overview, see Armitage, 2013, pp. 59–74; Schmidt, 1998), the present account assumes that concepts acquire meaning within a semantic field or theory. The polysemic concept of “international anarchy” will be analyzed within the context of three major families of IR theory: (a) realism and neorealism; (b) the international society approach, sometimes termed the “English School” of international relations; and (c) Kant’s republican peace. Understandably, these theories attach different analytical and normative significance to international anarchy. But they all build upon a common puzzle about the state of nature, which Hobbes formulated thus: What happens when a multitude of agents are locked into a common interaction domain in the absence of a common superior? Hobbes’s state of nature will be examined first as a basic model before considering the three models of international anarchy inspired by it—the realist, the international society, and the Kantian one. These models represent analytical maps, not ideal types (Weber, 1904/1949). The concluding pages summarize challenges to the discourse of international anarchy posed by the methodology of economics, and economics-based theories of international politics as well as by approaches that favor global hierarchy and global governance.

Hobbes’s State of Nature

In Chapter XIII of Leviathan, Hobbes describes the state of nature as a “warre . . . of every man against every man” (Hobbes, 1651/1968, pp. 185, 188, 189, 196). Its consequences are dire: misery, anxiety, and often physical destruction of the individuals involved. To see why conflict of this kind arises under the condition of “mere nature,” it is pertinent to attend to Hobbes’s five core premises specifying (1) the nature of the individuals, (2) their goals, (3) their means, (4) their motives, and (5) the domain of interaction. Hobbes’s first premise states that individuals are equal (Hobbes, 1651/1968, pp. 115, 183). All possess reason (the linguistic capacity for computing the names of effects from the names of causes, Hobbes, 1651/1968, pp. 96, 101–103, 115), and all are vulnerable to bodily harm and death (Hobbes, 1651/1968, p. 183). Yet, individuals differ in certain aspects. Hobbes portrays human beings as agents attracted by objects of desire and repulsed by objects of aversion (Chapter VI of Leviathan). It is the intensity of these passions that differs from person to person. Hobbes’s second premise states that individuals set goals for themselves in the external world. Different persons pursue different goals (goal pluralism), and one and the same person might reevaluate the desirability of one’s goals over time. The good for Hobbes is not objective but subjective, and it is merely “apparent good” or what appears to be good to the subject (Hobbes, 1651/1968, pp. 120, 129). This represents a doctrine of ethical subjectivism, not egoism (cf. Elements of Law, Hobbes, 1650/1969; McNeilly, 1968, pp. 126–127). Its implication is that no agent in the Hobbesian state of nature can know with certainty what other agents desire (Tuck, 1989, pp. 52–58). The sole desire known (since it is experienced by everyone) is a negative one: death aversion. Hobbes’s third premise concerns the means required for goal satisfaction—or “power.” Power is a person’s “present means to obtain some future apparent Good” (Hobbes, 1651/1968, p. 150).

Two principal accounts explain the unpalatable character of Hobbes’s state of nature. The first is a motivational account. Hobbes lists three conflict-generating motives (premise 4): “glory” (joy arising from imagining one’s own power or ability), “diffidence” (mistrust), and “competition” (greed) (Hobbes, 1651/1968, p. 185). The primary root of discord is mistrust (Hobbes, 1651/1968, p. 184)—an interpersonal relation that catalyzes the dormant fear of death. Imagine that two strangers encounter each other in a dark alley. Each might fear the worst, to suffer deadly injury; and, given mutual mistrust, it may be prudent for each to attack the other preemptively. This preemptive logic underwrites Hobbes’s “right of nature.” The right of nature is not a right proper (to qualify as proper, a right must be tied to a correlative duty; see Hohfeld, 1919) but a license to act as one pleases, including harming or killing which everybody holds against everybody else in Hobbes’s state of nature.

Mistrust is manifest also in time-asymmetric contracts called covenants (Hobbes, 1651/1968, p. 193). If Alan performs his part of a covenant today and Bob has to perform his tomorrow, Bob might renege on the deal when his turn comes. Hobbes’s more general point pertains to the uncertainty of transactions: There is no guarantee that people in the state of nature will honor their agreements. One standard reading of Hobbes is that a sovereign (coercive public authority) steps in and makes the contracts between private parties enforceable. In game theoretic terms, the sovereign transforms a prisoner’s dilemma game into an assurance game (Hollis & Smith, 1991, p. 129; Moehler, 2009).

A second explanation of why Hobbes’s state of nature is a condition of misery is non-motivational. It factors out agents’ motives and focuses on their external goals. Agents are portrayed as actively seeking out various goals, and therefore also means for attaining the goals (or “power after power”; Hobbes, 1651/1968, p. 161). Key here is Hobbes’s premise 5, which stipulates the state of nature to be a finite, densely populated interaction domain (Forsyth, 1988, p. 136). Inside it, individuals behave like atoms locked in a chamber—their collision is inevitable in the long run. What explains their collision (given goal directness) is the structure of interaction: the proximity of the units in an enclosed space. This particular account of Hobbes’s state of nature may be termed structural. It entails chaos (anarchy in the second sense) generated by overlapping trajectories. The chief inconvenience in this case is not the prospect of physical injury or death but mutual frustration (lack of “felicity”; see McNeilly, 1968, pp. 129–135). Mutual frustration may be the outcome of scarcity, or of time-overlapping preferences for non-scarce goods. There might be enough phone booths in the neighborhood but Cian and Darrel might end up in a fight if both seek access to the same booth simultaneously. The solution is to introduce rules that space out their interactions.

Scholars of international relations are bound to inquire about the role of power in Hobbes’s analysis of anarchy. Power is the universal means for goal satisfaction. But is it always rational to attempt to amass more power than competitors? The answer hinges on Hobbes’s premise of equality. Imagine that agents were vastly unequal. If a David knows his opponent to be a Goliath, then competing in the face of overwhelming power asymmetry would be an irrational (because it is a certain-to-fail) strategy. Inversely, if individuals are roughly equal, it is not irrational for any of them to hope to gain a competitive edge over others. The perceived equality of agents is what fuels competition under anarchy, in Hobbes’s view.

It follows that the greater the known power differentials between agents, the higher the degree of certainty about their situation—a premise pivotal to realism in IR (Guzzini, 1998, p. 145). This is an epistemological claim about what agents know about their environment and about each other. For Hobbes, the problem of knowledge is manifest on two planes: language and action (power). From the standpoint of language, Hobbes’s state of nature constitutes a linguistic chaos (in IR, see M. C. Williams, 1996) and the task of the sovereign is to arrest chaos by generating shared knowledge—a public language that the subjects can reliably use to communicate. From the standpoint of power, the sovereign commands obedience by virtue of “irresistible power” (Hobbes, 1651/1968, p. 397) but only if the subjects have shared knowledge that such power differentials exist. This prompts the question about uncertain knowledge involved in the extrapolation from past to future events (the problem of induction). The art of drawing inductive inferences warily is what Hobbes calls prudence (Hobbes, 1651/1968, pp. 97–98). Prudence, in Hobbes’s terminology, is not an ethical category but a decision making category describing forward-looking agents in a situation of uncertainly.

Indeed, the fundamental predicament of Hobbes’s state of nature at the plane of action is uncertainty understood in this epistemological, forward-looking sense. As Hobbes clarifies, the “war of every one against every one” does not refer to an actual battle but to the “known disposition thereto” (Hobbes, 1651/1968, p. 186). This disposition arises from an expectation of a future threat and is paradigmatic in international relations. In the international state of nature, one state cannot be certain about the intentions of other potentially hostile states, as Hobbes notes in a well-known passage in Leviathan where he compares states to gladiators “having their weapons pointing, and their eyes fixed on one another” (Hobbes, 1651/1968, p. 187). In the IR literature, this situation of mutual hostility deriving from uncertain knowledge about each other’s intentions is familiar as a security dilemma (Booth & Wheeler, 2008; Herz, 1950, 1959, pp. 231–235; Jervis, 1978). More generally, for Hobbes the international state of nature illuminates the predicament of the domestic state of nature. Rob Walker (1993) has suggested that Hobbes’s argument symbolizes the beginnings of modern politics, as a political space where the “inside” of the sovereign state by default has an “outside”: the international sphere populated by other dangerous states.

Two routes are available for exiting Hobbes’s anarchy: by instituting coercive sovereign authority (juridical state), or by acquiescing to a domination by superior power (Chapters XVII and XX of Leviathan). This turns on a basic distinction between political authority (right to rule; Raz, 1990) and power (ability to compel someone through a threat or use of force; Wolff, 1970, p. 4). In the first case, Hobbesian individuals transfer their rights of nature, via a social contract, to a representative body (the state) authorizing it to be sovereign authority.

Authorization in a general (nonpolitical) sense implies that an “author” transfers certain basic rights of action to another person, where this “right of doing any action is called authority” (Hobbes, 1651/1968, p. 218). Hobbesian political authority is juridical in essence: It makes, coercively enforces, and interprets the law of the realm. In the second case, a superior power acquires control over a population in a territory without authorization from below. Either way, the elimination of anarchy implies singularity—or sovereignty. One single authority or one single power must be setting the rules of the game in a realm. If more than one ruler claims supremacy (think of the Pope and the Emperor in the Middle Ages), the subjects would no longer know whom to obey: The problem of uncertain knowledge (the plight of anarchy) would reappear.

Hobbes did not prioritize either law or power as a solution to the problem of uncertainty in his early works. But law has precedence in his mature treatise, Leviathan, where the new doctrine of authorization is introduced (Chapter XVI). Law, when authorized from below, produces legal authority. This is a stable (uncertainty mitigating) solution to uncertainty since the act of authorization presupposes an obligation (commitment) to keep laws one has authorized oneself. Conversely, superior power remains an inherently unstable solution because the dominated would not cease hoping to reverse this superiority one day.

In sum, Hobbes’s juridical thesis is that law and order exist inside the realm of the state, whereas outside it in the international realm, chaos, violence, and lawlessness prevail. Anarchy in the first generic sense (a lack of supreme juridical authority) for Hobbes serves to explain why there is anarchy in the second generic sense (chaos).

Realism and International Anarchy

-Realists have reworked and enriched the puzzle of international anarchy inherited from Hobbes. Other classical figures of political thought, Thucydides and Machiavelli, have been a profound inspiration for the classical realists (political realists) Hans Morgenthau, John Herz, E. H. Carr, George Kennan, Henry Kissinger, Arnold Wolfers, and Raymond Aron. A distinct, scientific paradigm of realism—neorealism or structural realism—was inaugurated by Kenneth Waltz in his influential Theory of International Politics (1979). The neorealist paradigm has its variations. While it is notoriously difficult to specify what unifies realism as a school of thought (Guzzini, 1998, pp. 226–235), realists accept four core assumptions: (a) the central units in international relations are unitary actors, states; (b) states seek power; (c) the sphere of domestic politics differs in principle from the sphere of international relations; and (d) the relations between states take place under anarchy. By adding different auxiliary premises to this core, classical realists and neorealists reach different conclusions about the character of international politics and about the role of international anarchy.

Realists, regardless of denomination, reject the “domestic analogy.” This is the idea that a domestic society composed of human beings is, in principle, similar in its workings to a society of states (Bull, 1966a, p. 35; Suganami, 1989; Walzer, 1977/1992, pp. 58–59; cf. Beitz, 1979/1999, pp. 49–50, 154–155). This leaves open the further question of whether a domestic anarchy among human beings is analogous to an international anarchy among states. While realists reject the first argument (about society), they embrace the second (about anarchy). They think that the idea of a society of states is a contradiction, and that states behave like individuals in a Hobbesian state of nature. On this basis, they assert an asymmetry between the domestic and the international sphere (premise 3).

Classical realists express this asymmetry in normative language. Domestic political theory, as Martin Wight (1966, p. 33) remarked, is a theory of “the good life,” whereas international theory is a theory of “survival.” The idea goes back to Machiavelli who distinguishes private persons from public officials or princes, and allows the latter to commit crimes to protect the welfare of the state (Machiavelli, 1532/2003). According to this Machiavellian tradition of reason of state (raison d’état), protecting the interest of the domestic state in international relations means that violence among states is permitted and constitutes the normal state of affairs.

Typically, the classical realists use the term international anarchy to refer to disorder (chaos) and violence rather than to the absence of a common superior. Given the premise of international disorder, key explanatory factors of state conduct are the ethical priority of the domestic over the international realm, or human nature. Hans Morgenthau, for example, identifies a fundamental drive of human nature, the “desire to dominate” (1948, pp. 17–18), which trails from the arena of domestic politics into that of international politics.

International politics is “power politics” in classical realist analysis (Herz, 1942; Morgenthau, 1948; Wight, 1946/1978). The category of power tends to overshadow anarchy. Some realists from the classical school regard hierarchy—empire or “hegemony”—as the central category for analyzing international relations. Hegemony implies that one agency dominates others through influence, and empire refers to an institutionalized structure of rule where a metropolitan center exercises imperial control over subordinate units in the periphery. Adam Watson (1992, pp. 13–18) has proposed a spectrum (or pendulum) model of various international systems located between the poles of anarchy and hierarchy (empire). These considerations show that the anarchy–hierarchy distinction, even though acknowledged, is not theoretically basic in classical realism.

In sharp contrast, neorealist theory treats hierarchy and anarchy as basic, mutually exclusive categories. The pioneer of neorealism, Waltz, employs anarchy as a demarcation criterion to differentiate the domain of international politics from domestic politics. Domestic politics consists of relations of political rule that are arranged hierarchically (vertically) under domestic government; international politics is ordered anarchically (horizontally) in the absence of world government (Waltz, 1979, p. 88).

Waltz uses the term anarchy in a second sense, to denote a basic structural property of the international system (Waltz, 1979, p. 89, 1959/2001, pp. 6–7, 182–183, 231–232). To explicate the notion of system, Waltz contrasts “theory of foreign policy” with “systemic theory of international politics” (Waltz, 1979, pp. 18–19, 71–78). Theory of foreign policy, exemplified by classical realism, views each state as an atom: The actions of other states may inadvertently obstruct its action but they do not systematically constrain it. Conversely, systemic theory (such as neorealism) portrays each state as a “part” whose actions are constrained (structured) within the framework of a larger “whole,” the international system.

This may appear to be intended as a contrast between social atomism and social holism, but commentators have raised doubts whether Waltz’s account is genuinely holist (Hollis & Smith, 1991, pp. 115–118; cf. Guzzini, 1998, pp. 128–129; Wendt, 1991). In Waltz’s analysis, system is a concept used in an interactionist (non-holist) sense. It is borrowed from neoclassical economics (microeconomics) and it refers to the unintended consequences of interaction (which constrain individual action). This Waltzian notion of system is analogous to a market system (Waltz, 1979, pp. 71–72, 89–91). Analytically, any such system consists of interacting units and structure (Waltz, 1979, pp. 79–80). The structure has three tiers: an ordering principle (unit arrangement), the functional differentiation of the units, and distribution of capabilities (Waltz, 1979, pp. 82, 88).

Waltz relies on two parallel concepts of anarchy that he does not always distinguish consistently. First, anarchy, as a demarcation criterion, refers to the absence of world government. The anarchical international system is a system of “self-help” (Waltz, 1979, pp. 107, 111) where states cannot appeal to a final arbiter to resolve their disputes. Like a solitary individual in Hobbes’s state of nature, each state must procure for its survival alone (for it would be dangerous to entrust this job to other states), generating security dilemmas and war. Second, anarchy, as a horizontal order of nominally equal units is the ordering principle of the international system. It comprises the first layer of Waltzian structure. Its second layer is supposed to capture the functional differentiation of the units. However, no functional differentiation (division of labor) characterizes international politics: States are “like units” that perform an identical range of tasks (Waltz, 1979, pp. 96–97, 107). Because of this, the second tier of structure is a fixed parameter with respect to state interaction.

Anarchy, the first tier of Waltzian structure, has the character of “deep structure,” which is unobservable except through its effects (Ruggie, 1983, p. 266). All three tiers represent structural properties of the international system as a whole and cannot be reduced to the unit-level properties of its constituent units, states. The key unit-level property of each state, considered singly, is power defined as “capabilities” (internal capabilities, military might, and organizational resources; or external capabilities, military alliances). In an anarchical international system, states differ in capabilities but perform an identical task—security provision. Hence their basic motive is to secure their own survival and, if possible, to accumulate power (Waltz, 1979, pp. 105, 118, 126, 1959/2001, p. 203). Consequently, the international system has polarity: the system-wide distribution of capabilities (power) across all the units (Waltz, 1979, p. 98). This is the third layer of Waltzian structure (closest to the “surface”) that can be measured by counting the number of great powers in the system.

Waltz’s argument of how states interact under the structural constraints of anarchy replicates Hobbes’s structural account of conflict in the state of nature. For Waltz, states are “billiard balls” colliding inside an enclosed space. The outcome of their collision is a system of balance of power (distribution of capabilities) that emerges unintentionally. Waltz elucidates it by means of a laissez-faire analogy (Waltz, 1979, pp. 91, 118): States are like firms, the international system is like a market, and security is like a pricing mechanism. Firms that sell over the market price will go out of business, and, similarly, states that “bandwagon” with, instead of “balance” against the preponderant state(s) will perish (Waltz, 1979, pp. 126–127). Once formed, a system-level balance of power exerts structural constraints on subsequent state action and tends to restore systemic equilibrium.

Barry Buzan has refined Waltz’s systemic theory of international anarchy. Apart from the international system, he claims, its individual units, states, have deep structure. Unlike a “weak” state, the internal structure of a “strong” state is characterized by a high degree of political cohesion and legitimacy (Buzan, 1983/1991, pp. 97, 103). Such a state does not face internal threats to its national security and its external relations with other states are more stable. A critical mass of strong states promotes the emergence of mature anarchy in international relations (Buzan, 1983/1991, p. 175). Mature anarchy corresponds to a Hobbesian state of nature modulated by common norms and rules, and immature anarchy is an unmodulated Hobbesian state of nature as in Waltz’s scheme. One important difference is that on Waltz’s interactionist market model, the limits on state action are not set by the intentional pursuit of common rules and norms but, rather, by fortuitously generated power differentials.

The neorealist view of international anarchy has been criticized for its neglect of change and historicity by constructivists and poststructuralists. Poststructuralist Richard Ashley has branded Waltz’s anarchical structure a “frozen abstraction” (Ashley, 1984, p. 255). John Ruggie (often classified as a constructivist) has argued that Waltz’s neorealism cannot account for the transformation of the international system, from medieval to modern, since for Waltz the second tier of the anarchical structure (functional differentiation) is a fixed parameter. If functional differentiation is reintroduced as a variable, it would reveal a systemic change—from functionally heteronomous units in the medieval system to homogeneous (sovereign) units in the modern one (Ruggie, 1983, pp. 273–275, 285).

Constructivists have pointed out that neorealists are too swift to associate international anarchy with mutual hostility, security dilemmas, and worst-case scenarios. Alexander Wendt observes in an often-quoted article, “Anarchy is What States Make of It” (1992) that the identities and interests of agents (states) are not given prior to interaction, hence anarchy need not generate a system of self-help. Criticizing Wendt’s emphasis on individual agents, Jonathan Mercer has suggested that intergroup competition matters and that self-help “is a consequence of intergroup relations in anarchy” (Mercer, 1995, p. 251). In Social Theory of International Politics (1999), Wendt challenges the neorealist view of anarchy as an immutable, deep structure. Instead, anarchy is constituted by culture (macro-level normative structure), which may be a Hobbesian culture of enmity, a Lockean culture of rivalry, or a Kantian culture of friendship (Wendt, 1999, pp. 246–312).

International Society and International Anarchy

Hobbes’s austere picture of anarchy appears to leave no room for social rules or society. But international society, aptly christened an “anarchical society” by Hedley Bull (1977/2002), claims to unite these seemingly opposed ideas. To say that states form their own society, an international society, is to assume that they are like socially competent human beings who follow common rules and norms together. This idea is traceable to the writings of early modern jurists and political thinkers Hugo Grotius (1631/2005), Samuel Pufendorf (1688/1964), and Emmerich de Vattel (1758/2008), and Bull himself refers to his international society approach as “Grotian” (Bull, 1966b, 1977/2002, pp. 25–30). Nonetheless, unlike domestic society, international society is organized anarchically because its member states maintain sovereign equality in their external relations. The puzzle of the domestic analogy is focal here—to what extent does international society resemble domestic society?

The contemporary international society approach has two distinct strands. The first, the English School of international relations was set up in the 1950s as the “British Committee on the Theory of International Politics” under the lead of Hedley Bull, Martin Wight, and Herbert Butterfield joined by Adam Watson and John Vincent, and influenced by Charles Manning (for its history, see Dunne, 1998; Linklater & Suganami, 2006). Since the 1990s, English School writers have included Barry Buzan, Richard Little, Nicholas Wheeler, Tim Dunne, Andrew Hurrell, James Mayall, and Michael Donelan. The second strand of international society theory is a branch of international political theory dedicated to questions of international morality. Its prominent representatives are Terry Nardin (1983), Michael Walzer (1977/1992), and Robert Jackson (2000). The difference between these two strands can be measured as an analytical distance from classical realism (Brown, 2001). Bull and Wight maintain an ambivalent but close connection to classical realism, Walzer grants that realist premises may be a limiting case of international morality, and Nardin dissociates international morality from realism, treating them as ideal types. Such differences should remind us that the expressions “English School” and “international society approach” are not, strictly speaking, synonymous.

A hybrid type of an anarchical international society that bridges English School thinking with Waltz’s neorealism has been championed by Buzan, Little, and Charles Jones (Buzan et al., 1993). They reserve the label structural realism for their historicized analysis of international systems. It integrates the components of anarchical structure and functional differentiation, and introduces a new variable called “interaction capacity” (Buzan et al., 1993, pp. 66–80).

After this capacity reaches a threshold, an international system evolves into an international society (Buzan et al., 1993, p. 71; Guzzini, 1998, p. 221) that is best elucidated by an interpretive, hermeneutic epistemology as opposed to Waltz’s Popperian philosophy of science (Little, 2000). Interpretation matters since the rules of international society are assumed to have meanings. States inside international society are akin to intelligent human agents who can understand meanings—a point that reflects Bull’s distinction between system (mechanical) and society (rule-based and meaning-based Bull, 1977/2002, pp. 10, 13).

Bull’s Anarchical Society (1977/2002) remains the classic statement of why anarchy is the organizing principle of the society of states, despite impressive growth in the English School literature (see Buzan, 2014). By “anarchy” Bull means the absence of government, as a juridical concept (Bull, 1966a, p. 38; 1977/2002, pp. 55–56). Yet at its base, Bull’s theory is sociological. It is a theory of international order where order is defined in sociological terms (Bull, 1977/2002, p. 51) as a “pattern of behaviour that sustains the elementary or primary goals of social life” such as prohibitions against violence, the stability of agreements, and property. Having distinguished government from rules (external constraints on conduct) as order-maintaining devices, Bull argues that states inside international society recognize common rules without recognizing a common government (Bull, 1977/2002, pp. 51–73). This complex of rules—rules of balance of power, rules of war, rules of great powers, rules of international law, and rules of diplomacy (Bull, 1977/2002, pp. 26–37)—constitute the “primary institutions” (Buzan, 2004, pp. 167–176) of the society of states. International society does not include all nominally existing states (around 200). It is a “club” where membership is conditional on keeping the rules of the game, a proviso that binds the great powers as well.

The primary institutions of international society impose normative limits on what is permissible to do for any state that claims to be a member in good standing. One important issue concerns the primary institution of international law. What sets the global law of a global state (a hierarchical order) apart from the international law of international society (an anarchical order) is that inside the latter, member states make the law together, as formal equals. As long as no single state or group of powerful states “prescribe the law to the others” in Vattel’s phrase (de Vattel, 2008, Book III, §47, p. 469), international society is compatible with international anarchy.

The compatibility of international society and international anarchy implicates the domestic analogy: It invites us to examine the degree of similarity between a society composed of states and domestic society. Despite disagreements in other respects, various international society theorists all endorse a qualified domestic analogy. They agree that states, like competent social beings, act under the constraint of common institutions but insist (and this is the qualification) that the primary institutions of international society are sui generis.

International society is a normative structure in its own right—it is not just an epiphenomenon of domestic society. Three factors elucidate the sui generis character of international society, as Bull contends in “Society and Anarchy in International Relations” (1966a) and “Hobbes and the International Anarchy” (1981). The first and cardinal one is that unlike human beings, states are very hard to kill. So in international relations, Hobbes’s postulate of fear of death can be discounted and, with it, the ubiquitous quest for relative security. Second, states are not compelled to invest the bulk of their resources into the pursuit of security the way Hobbesian individuals in the domestic state of nature are, so a state can invest considerable resources into internal economic development and international trade. Third, a state (especially a great power) is highly self-sufficient economically and its self-sufficiency is not easily destroyed during interstate war. Unlike domestic anarchy, therefore, international anarchy is a palatable condition that allows for international cooperation among states.

Theorists of international society such as Nardin, Jackson, and Walzer accept Bull’s idea of international society as sui generis but support it with a different argument. Breaking with classical realism, they advocate the principle of international morality or the “morality of states” (Beitz, 1979/1999, pp. 65–66). One way to ground this morality is to appeal to the fundamental right of self-preservation, as Grotius (1631/2005) did in The Rights of War and Peace (Tuck, 1625/2005). Grotius’s logic is prudential (in the ethical sense). Given that in the international state of nature all states have an identical aim, self-preservation, and if no state should be blamed for pursuing this aim, then the best course of action for all states collectively is to grant each other a reciprocal right to exist (Tuck, 1989, pp. 20–21, 51).

A moral (non-prudential) ground for the morality of states is offered by Walzer. Walzer (1977/1992) depicts each state as a legal shell that protects the rights of individuals as members of a self-determining political community. Since these human rights (to life, liberty, and community) cannot be enjoyed if the community is obliterated, it must be granted that the community’s guardian, the state, has a right to exist that is recognized by other states (Walzer, 1977/1992, p. 59). The result is a basic principle of peaceful coexistence among states, which implies the recognition of specific state rights to territorial integrity and nonintervention. The international sphere, on this view, is not a jungle but a civilized society of states. Only in extremis, a political community threatened with annihilation might resort to preemptive strikes, the usual arsenal of realism (Walzer, 1977/1992, pp. 74–85)—but this represents an exception rather than the normal course of international conduct.

The morality of states, then, is a morality of peaceful coexistence. It presupposes an anarchical society because anarchy respects the fact of international plurality where various political communities, each with its distinctive conception of the good, coexist. Walzer contrasts the “thin,” common morality of international conduct with the “thick” local codes of particular communities (Walzer, 1994, pp. 17–19). Nardin calls the former a “morality of accommodation” concerned with “procedures rather than purposes, means rather than ends, and compromise rather than perfection” (Nardin, 1983, p. 232). It is a minimalist morality: It imposes an obligation on states not to injure one another but it does not require them to practice mutual aid. In the international society literature, this position goes under the label of pluralism. Pluralists hold that a group of agents accept common rules or norms as procedural limits on conduct without accepting common substantive purposes such as moral perfection or economic redistribution. Bull (1966b) is a pluralist who restricts the defining norms of international society to a “thin” set of procedural norms: nonaggression, nonintervention, keeping of treaties, and diplomatic immunity. A contrasting view, which presents international society as a “thick” normative order committed to shared goals and purposes such as humanitarian intervention is solidarism (Wheeler & Dunne, 1996).

This brings up the question about the values and norms that underwrite the anarchical society of states. One value implicit in the discussion so far is the freedom of the individual (be it a state or a human person). From a neo-Hegelian perspective, Mervyn Frost has developed constitutive theory (Frost, 1996) that links the value of freedom to two anarchical practices (Frost, 2009): the society of sovereign states and global civil society (a global rights practice). Both are anarchical in form, and exclude the hierarchical social forms found in slave-owning societies and in empires. According to Frost, the most distinctive feature of our present historical epoch is that nearly everyone, everywhere simultaneously participates in both anarchical practices. This argument invokes Hegel’s notion of freedom (Hegel, 1821/1981) understood as a valued identity of being a free individual, constituted through mutual recognition between self-conscious actors. In modernity, freedom is realized as rights holding. What the notion of anarchy contributes to this Hegelian framework is that a horizontal (anarchical) social form assures its participants of their equal standing as free agents. Thus, in the global rights practice, participant individuals constitute one another as holders of equal sets of basic rights (freedoms) and in the global practice of states, participant states constitute one another as holders of equal sets of sovereign freedoms (Frost, 2009, pp. 55–58, 78–115). But freedom has its dark side. For example, men and women who exercise their freedom of contract as economic agents in global civil society are in permanent competition with one another, generating huge power differentials globally. Although such differentials reveal a strain between the society of sovereign states and global civil society, these two practices are not incongruent—taken together, they represent “double anarchy” grounded in the value of freedom.

Apart from Hegel’s and Hobbes’s view of freedom (the right of nature), there is another prominent notion of freedom that has relevance for international anarchy—it was expounded by the 18th century German philosopher Immanuel Kant.

International Anarchy: Kant’s Republican Peace

Kant presents us with an argument for a “league of free states,” which is a special kind of an anarchical society structured around the ideal of freedom. Kant’s view is worth examining since Kant became a proponent of anarchy in international relations after some hesitation. In his pre-1793 writings he espoused a blueprint for world government (“world republic”) but later, in Perpetual Peace (1795/1991) and The Metaphysics of Morals (1797/1996) opted for a “negative substitute” (Kant, 1795/1991, p. 105)—foedus pacificum. (Kantian scholars continue to debate which of the two institutional forms was Kant’s true preference; Bohman & Lutz-Bachmann, 1997; Cavallar, 1994; Hurrell, 1990; Kleingeld, 2004). The English translation of foedus pacificum is “federation” but Kant means “confederation” by it—a loosely organized “congress” or “league” of states.

Kant’s position merits attention also because it fleshes out the normative connection between freedom and international anarchy. In order to appreciate the significance of this connection, it is requisite to briefly consider Kant’s theory of the state. Following Hobbes, Kant holds that unless a centralized legal order, a state, 237 is created, a condition of lawlessness and disorder (anarchy in the second sense) will prevail. This is the reason Kant embraced the idea of world government initially. Such an alignment of Kant and Hobbes might seem awkward at first blush. But although Kant’s deontological (duty-based) theory of morality does not sit well with Hobbes’s prudential morality of self-preservation, their theories of law and state are strikingly similar in permitting one of two options: law and order (inside the realm of the state) or, conversely, disorder and violence (outside the state). Thus, the IR trichotomy between “realists” (Hobbesians or Machiavellians), “rationalists” (Grotians), and “revolutionists” (Kantians) popularized by Wight (1991) and Bull (1977/2002, pp. 23–26) must be treated with caution (see Hoffmann, 1986, p. 186).

It may be recalled that Hobbes defined the state as a centralized juridical institution for making, interpreting, and enforcing the law. While Kant operates within this Hobbesian tradition, he has a special task in mind: to spell out the features of a properly constituted state: a republic. In the first definitive article of Perpetual Peace, Kant urges that the constitution of each state must be republican. This invocation of the domestic make-up of the state has promoted the view of Kant as a liberal. Here liberalism refers to a doctrine in contemporary political science that construes of the international sphere as an extension of domestic law and politics.

In this vein, political scientists have hypothesized an empirical law of “democratic peace” or “liberal peace” stating that liberal democracies do not fight each other (Doyle, 1986). Despite ample references to Kant in the burgeoning democratic peace literature (a key collection is Brown et al., 1996), Kant’s argument departs from this hypothesis in two ways. First, Kant envisages a universal principle of international right that would abolish war among states for ever—hence Kant’s talk of “perpetual peace.” Neither universality nor perpetuity can be reduced to an empirical, inductively supported law. Second, Kant wants to preclude the possibility that a well-organized society domestically might resort to war internationally without proper ground (e.g., unilaterally)—a weakness that Kant attributes to just war theory (Williams, 2012). In this aspect, Kant differs also from contemporary liberal political philosophers like John Rawls (1999) who think that “well-ordered societies” (liberal democracies) would automatically behave responsibly on the international scene.

Kant is making two claims about the normative structure of the international realm. The first is that states must recognize a common principle of international justice—or “international right.” International right regulates the relations between states; civil right regulates the relations between citizens inside the state; and cosmopolitan right, those between citizens regardless of location. Kant’s second claim is that actualizing international right demands an international authority, a society of free states. The challenge is to understand what Kant means by free states, and how he connects freedom to anarchy. Of relevance here is not only Kant’s idea of freedom as autonomy familiar from his moral theory in The Groundwork (Kant, 1785/1997) but rightful freedom expounded in The Doctrine of Right (Part I of The Metaphysics of Morals, Kant, 1797/1996) which is a work on legal theory.

In The Doctrine of Right, Kant argues that in the domestic state of nature, individuals hold an innate right to freedom (Kant, 1797/1996, VI: 237). This is a freedom to choose one’s ends in the external world (external freedom) where the subject interacts with other subjects. It embodies a full-fledged right to act, not a naked license to act (Hobbes’s right of nature). Still, like Hobbes’s state of nature, Kant’s state of nature represents a condition of uncertainty since individuals might interfere with each other’s freedom of choice. The hazard of interference arises because they are physical entities confined to a finite space (Flikschuh, 2010; Ripstein, 2009, p. 12). To ban interference once and for all, a principle of right (justice) must be introduced so that the freedom of one “can coexist with the freedom of everyone in accordance with a universal law” (Kant, 1797/1996, VI: 230). This principle must have teeth: It must be coercively enforceable. The state as a public realm of law, where law regulates the external freedom of agents by coercive means, creates the principle of right. However, public coercion must be rightful. It must be universal in that it makes everybody, not just some, secure in their relations, including their normative relations or rights; and it must ensure non-domination by preventing the will of one person from arbitrarily determining the choice of another person (Ripstein, 2009, pp. 30–56).

Hobbes and Kant both define the state as a public realm of law: a res publica. (Conversely, in a despotism, the state is a private property of the despot.) A republic comprises a public system of rules (laws) created by authorized officials. It distributes and determines the juridical rights of the citizens; it enforces these rights coercively (as a matter of right); and it treats citizens as equal under the law. But formal equality is not enough. In a properly constituted Kantian state—a republic or Rechtsstaat (a rule-of-law state)—the citizens ought to be equally free. This is a freedom of co-legislation: The law ought to embody the “general united will” of the citizenry (Kant, 1797/1996, VI: 313–316; Kersting, 1992). This requires the law-making prerogative of the legislators to be limited by a civil constitution. Such limits ensure that the law is just: that it is universal and non-arbitrary or that it applies to everybody, including the legislators themselves.

In this light, Perpetual Peace can be read as a proposal for establishing an international rule of law. It envisages an international public authority (a league of states) to uphold an ideal international constitution whose aim is to permanently outlaw war. Perpetual Peace lays down the articles of this international constitution. Its “definitive articles” (the three principles of right) apply to relations among states inside the pacific league, whereas its “preliminary articles” (principles of nonaggression and trust), to those between league insiders and outsiders (Laberge, 1998, p. 91). The definitive articles demand a republican constitution for each state (domestically), a league of free states (internationally), and a cosmopolitan principle of universal hospitality (globally). Once ratified, these articles become binding on member states, but entry into the pacific league is voluntary or non-coercive (Cavallar, 1994, p. 468). The impermissibility of coercing any state into membership makes the foedus pacificum an anarchical (freedom-respecting) international society.

Besides freedom (whose meaning will be clarified in a moment), “anarchy” designates a state of nature for Kant: a domestic and an international one. Kant contends that human individuals have a duty to exit the domestic state of nature and to enter a rightful condition (civil state) since only the civil state can secure their rights to freedom (Kant, 1797/1996, VI: 312). This entitles one individual to force another into entering the civil state. After its creation, the civil state becomes an artificial individual in the international state of nature. Kant emphasizes that it would be wrong to coerce a state to exit the international state of nature in order to enter the league of states (Kant, 1795/1991, p. 104). Some scholars think that Kant’s position reflects anti-paternalism or respect for a people’s right to political self-determination within a state (Kleingeld, 2004, p. 309). Others argue that since Kant views the state as the supreme enforcer of right, the application of right against a state cannot be coercive (Flikschuh, 2010, p. 471). Both perspectives assume that the state must remain a free agent, but refer to two different aspects of freedom: autonomy of the moral person (Kleingeld, 2004, p. 309) versus autonomy of the state, as a person who is subject to moral constraint but whose nature is artificial or juridical (Flikschuh, 2010, p. 480). The latter implies the idea of freedom as independence under conditions in which interaction with others cannot be avoided. Autonomy (freedom as self-legislation or internal freedom) belongs to the domain of Kantian ethics (where the individual, considered in isolation from all other individuals, legislates in one’s inner court of consciousness), whereas independence (freedom of choice or external freedom) must be secured in the external relations between multiple individuals—its purview is Kantian law. To count as free, the agent’s choice to act in the external world must remain independent from the choice of other agents (Ripstein, 2009, pp. 15–16, 34–35). The moral ground which justifies the use of legal coercion is freedom (Kant, 1797/1996, VI: 340): Law is there to enable agents to interact with one another as equally free agents.

In sum, Kant’s league of free states is an anarchical society whose members, mutually independent juridical persons, are free to bind themselves to the authority of common rules. It is a “pluralist” anarchical society limited to the prevention of war. It differs from its English School counterpart in two aspects. First, for Kant, international law is the fundamental institution of international society, whereas for Bull and Wight the balance of power and great powers are equally important. Moreover, while Bull allows international law to be violated for the purposes of maintaining the balance of power, Kant regards international law as the institutional embodiment of international right which necessarily excludes considerations of power. Although states outside the Kantian pacific league inhabit a domain of disorder and lawlessness, the principle of international right prohibits the waging of war against an outsider, a pariah state, as a prudential strategy of preventing more wars in the long run (Cavallar, 1994). Nor is it permissible to export republican values abroad by waging wars of democracy promotion (Williams, 2012, pp. 113–140). Only wars of self-defense are permitted under Kant’s principle of international right.

Finally, Kant’s skepticism about a global state in Perpetual Peace appeals to the principle of cosmopolitan right. It is a principle of universal hospitality: a right of refuge granted by a state to foreign citizens abused by their domestic authorities. Kant fears that if a global state turns, it would leave no route for escape for persecuted citizens and minorities: It would degenerate into “soulless despotism” (Kant, 1795/1991, p. 113). This reminds us of the normative appeal of international anarchy as a pluralist framework that enables political communities with different conceptions of the good to coexist.

Curiously, the premier neorealist Waltz acknowledges the force of Kant’s argument in his lesser-known article, “Kant, Liberalism and War” (Waltz, 1962). On Waltz’s reading, Kant is a liberal thinker—not an exponent of harmony of interests—but a liberal who recognizes the war-prone character of international relations. As Waltz concludes , Kant believed that “states may improve enough and learn enough from the suffering and devastation of war to make possible a rule of law among them that is not backed by power but is voluntarily observed” (Waltz, 1962, p. 337). This shows that the three models discussed so far—the realist, the international society one, and the Kantian—notwithstanding key differences, map out a common discourse of international anarchy.

Challenges to International Anarchy

The discourse of international anarchy has its limitations sketched in this brief concluding section. One well-known limitation is conceptual—it has to do with the tight conceptual link between “sovereignty” and “anarchy.” As mentioned previously, a system composed of internally sovereign units, states, necessarily entails an anarchical (horizontal) order among equals externally. But this presupposes Bodin’s (Bodin, 1606/1962), Hobbes’s, and Kant’s notion of sovereignty as indivisible, where a single institution, the state, holds sovereignty (the supreme right to rule). It does not apply to contemporary federal systems of government where sovereignty is shared between the government and the people or, further, between different branches of government and the civil constitution. Kant’s league of free states is a confederation (rather than federation)—in this type of institutional structure, member states do not surrender sovereignty to a supranational authority.

The most robust challenge to international anarchy stems from the discipline of economics, as reflected in the methodological principles of fungibility and efficiency that are central to economic theory. According to the dominant paradigm of neoclassical economics, economic behavior is explained by the theory of rational choice, which assumes that the rational economic agent chooses an action based on an order of preferences over alternative outcomes under (information and other) constraints (Elster, 1986; Heap & Varoufakis, 1995, pp. 4–31). Two central assumptions here are (a) rationality defined as formal consistency, which implies preference transitivity (Sen, 1986, pp. 64–67); and (b) money, a standard of measuring economic value that is “fungible” or translatable across different sectors (on fungibility in IR, see Baldwin, 1993, pp. 20–21). Fungibility is a methodological principle of value assessment. It regards values as mutually translatable, and measurable on a uniform scale (the measurement units can be “utils” or money). This means that qualitatively distinct values are treated as if they were economic “goods” whose difference is purely quantitative (goods differ only in prices). IR theorists for whom economics constitutes the foundation of IR theory follow this fungibility logic. They accept that different types of goods can be exchanged without limit other than the economic agents’ endowments and willingness to transact. On this view, a sovereign state may trade its status of a sovereign entity in exchange for benefits. Recently, David Lake (2009) has argued that weaker states are prepared to relinquish aspects of their sovereignty to a club of powerful states in return for obtaining international security. Even though some modicum of legitimacy must be ensured, the envisaged system is essentially a global hierarchy dominated by the great powers.

It is notable that theories buttressed by economic methodology such as “hegemonic stability” (Gilpin, 1981; Kindleberger, 1973), “complex interdependence” (Keohane & Nye, 1977), and globalization represent world politics as a global political economy (Gilpin, 2001) or global hierarchy (Lake, 2009). In general, students of economics tend to prefer hierarchy as an organizing principle of world politics, whereas students of law, ethics, and security favor anarchy. In his defense of hierarchy, Lake approaches state sovereignty and international security as economic “goods” to be traded against each other, and even legitimacy is a strategy for reducing the cost of great power management globally (Lake, 2009, pp. 9, 99–103). In contrast, scholars of international law and ethics hold that certain domains of value (e.g., sovereignty, rules of peaceful coexistence) are protected and not subject to trade-offs since they constitute the identity of states as players on the international scene (Frost, 2009; Nardin, 1992, p. 23). So far this does not yet explain why economic theory should prioritize hierarchical systems over anarchical ones. The explanation is that in economics the assumed purpose of agents is efficiency (or value maximization). Moreover, the domain of economic value is all-inclusive—translated in territorial terms, it produces the image of a single, borderless global economy. Relations of hierarchy may emerge within the global economic system simply because they best fulfill the task of value maximization—any imposition of limits would frustrate the task. This image of global hierarchy is radically distinct from international anarchy—a world of territorial and normative limits, of territorially bounded, juridically independent units: sovereign states.

Earlier IR debates exploring the tension between economics and security have implications for international anarchy. The discussion was spurred when critics of Waltz noticed that he combines economic and security reasoning within his neorealist theory. On the one hand, states seek relative security vis-à-vis other states; on the other, the anarchic international system is modeled as an oligopolistic market. But the maximization of profit by firms has no analogue in the realm of international security, pace Waltz, since it makes no sense to talk of the maximization of security (Baldwin, 1997, p. 22). In the specialized IR literature, this has stimulated a related debate over relative versus absolute gains (Powell, 1991). Neorealist Joseph Grieco (1988) joined Waltz (1979, p. 105, 1993, p. 60) in claiming that the structure of anarchy compels states to worry primarily about security and therefore about relative power (relative gains): How much more does A get relative to B? Neoliberal institutionalists such as Robert Keohane (1984, p. 27) objected that absolute gains matter—or how much each party, considered alone, would gain in absolute terms from an exchange against a baseline of no exchange. These logics pull in different directions: Relative gains reflect the security logic of an arms race, and absolute gains, the economic logic of comparative advantage and, ultimately, value maximization.

Neoliberal institutionalism has played a central role in repositioning the assumption of international anarchy within the framework of economic models of cooperation, and away from models of pure conflict favored by neorealism. Contra neorealists, neoliberal institutionalists contend that international anarchy does not prevent cooperation among self-interested states. Rather, the prospect of international cooperation depends on the ability of states to create institutions. In accordance with game theory and rational choice, states are equated to rational egoists who pursue utility maximization under uncertainty (Axelrod, 1984; Axelrod & Keohane, 1985). This form of rational choice (“rationalist”) institutionalism treats uncertainty as a problem of incomplete information, and emphasizes mix-motive games where players hold simultaneously cooperative and conflictual preferences (for a criticism, see Alker, 1996). In this context, international institutions constitute procedures for solving monitoring and compliance problems in certain “issue areas” (international trade, nuclear proliferation), and operate by reducing transaction costs (Keohane, 1984). One notable species of international institutions are international regimes: “sets of implicit or explicit principles, norms, rules, and decision-making procedures around which actors’ expectations converge in a given area of international relations” (Krasner, 1983b, p. 2; see the 1982 issue of International Organization on international regimes reprinted in Krasner, 1983a). Stephen Krasner’s diagnosis of international regimes as “intervening variables” (Krasner, 1983b) indicates that international institutions mitigate but do not abolish the condition of international anarchy. For theorists like Krasner (who is closer to the realist camp), international institutions complement the agency of hegemonic states, whereas for neoliberals like Keohane (1984) and Gilford Ikenberry (2001), once created (by a hegemonic state), institutions begin to exert independent effects on international outcomes.

Since the 1990s, the inability of states and international institutions to manage global problems such as environmental pollution, financial crises, crime, and underdevelopment has attracted the attention of theorists of globalization. They argue that the Westphalian model of international anarchy, which prioritizes security over economic issues, and which treats sovereign states as primary agents, is no longer an adequate representation of an increasingly interdependent, globalized world (Held, 1995; Held & McGrew, 2007; Strange, 1999). These claims are difficult to assess because “globalization” is a label for a diverse range of developments pertaining to the shirking of space and time, deterritorialization, instantaneous transactions (Scholte, 2005, pp. 101–117), and virtual reality (Der Derian, 1999). Global processes are often defined in terms of trade, investment, and communication flows across state borders (Hirst & Thompson, 2002, p. 247). More abstractly, globalization has been linked to the increased permeability of boundaries in social space (Hofferberth, 2015), thus promoting changes in social divisions based on class, age, or gender. It has also been linked to changed attitudes towards risk, uncertainty, and threats posed by science and technology in post-industrial societies in late modernity (Beck, 2000). These arguments imply that the problem of anarchy and the idea of international relations (relations between states) it entails seem to have become obsolete, and that we should focus instead on the alternative idea of global relations which transcend state borders.

The earlier literatures on international regimes and globalization have recently been reinvigorated under the rubric of global governance (Buchanan & Keohane, 2006; Finkelstein, 1995; Lake, 2010; Weiss & Wilkinson, 2018, 2019). It has been suggested that private agencies above and below the state—banks, firms, professional associations—produce and manage regulative networks across state borders in a decentralized manner. The concept of “governance” refers to regulative procedures that differ in principle from the procedures of state “government” (Rosenau & Czempiel, 1992). The state is a public, coercive, and supreme juridical authority with a jurisdiction over a territory whose purpose is to regulate two sets of relations—between its officials and citizens, and among its citizens (see Weinrib, 2019, pp. 23, 31–37). In contrast, governance includes regulatory techniques typical of business enterprises such as standard setting, monitoring performance, and self-regulation. As a rule, techniques of this kind lack a territorial or juridical component, operate in a diffuse or polycentric manner, and imply no order of precedence among the actors involved.

Like the idiom of globalization, global governance comprises a loosely connected cluster of arguments. Paradigmatically, the term global governance is a shorthand for institutions for managing the post-WWII global order exemplified by the United Nations, the International Monetary Fund, and the World Bank. In a broader sense, the term can designate the emergence of regulatory hierarchies in world politics which include various norms, regimes, and quasi-institutionalized structures for tackling global issues that either circumvent or encompass the authority of the sovereign state (Weiss & Wilkinson, 2019, pp. 17–18). In this latter sense, the distinction between globalization and global governance becomes blurred. The emphasis on institutions therefore seems to provide a clearer analytic focus in thinking about global governance.

Two broad hypotheses concerning post-WWII global governance institutions have emerged. The first holds that these institutions expose (and can correct) weaknesses in government exercised by the state, or by a multitude of states in the international realm. For example, the International Criminal Court (set up in 2002) is an institution with a global jurisdiction allowing it to prosecute perpetrators of war crimes and crimes against humanity when their home state is unable or unwilling to do so. Conversely, the second hypothesis suggests that global governance institutions are weak and in danger of unraveling (Ruggie, 2014, p. 5; Zürn, 2018). In this connection it is necessary to differentiate between empirical, conceptual, and normative standpoints (Hofferberth, 2015). One conceptual puzzle concerning the area of climate governance is whether it is possible to make sense of the myriad of overlapping, partly incompatible regulatory regimes, norms, and standards in that area which coexist at the domestic, international, and global level (Keohane & Victor, 2011, p. 8). When questions instead are raised about the (perceived) lack of legitimacy, democratic accountability, or procedural transparency of global governance institutions (Tallberg & Zürn, 2019; Zürn, 2018), the problem of global governance is approached from a normative standpoint.

Rosenau and Czempiel’s motto of “governance without government” (1992) brings home the idea that international anarchy presupposes the institution of government and, indeed, the sovereign state as a public and centralized form of government. Even though the plurality of global governance and globalization discourses cannot be compressed into a single theme, they all seek to replace the model of government associated with the sovereign state. The rationale behind this suggestion is the state’s inability to effectively manage global problems thus undermining its authority. However, it is important to acknowledge that the state’s ability to rule effectively or its power—exercised under conditions of globalized economy or globalized security threats—is not the same as its right to rule or its authority (see Morris, 2010, pp. 1323–1324). The discourse of international anarchy is premised on the state construed as a locus of authority; it does not per se purport to address the problem of state power. This is one more reason why the conversation between proponents of the international anarchy and its critics is bound to continue.