Introduction
A metric that scholars often use to categorize an international hierarchy is the extent to which the leading state uses coercion or consent to generate the political outcomes it favors. Although this distinction has intuitive appeal, it is misleading. Do so-called ‘benevolent’ hegemons always use carrots and never sticks? Current theoretical frameworks on hierarchy suggest the answer is ‘yes’, but evaluating the balance between coercion and consent is often highly subjective. Determining levels of consent is problematic because hierarchies often operate in the shadow of a skewed distribution of power. Still, one should not make the opposite error. International orders cannot rely on coercion alone. A hierarchy predicated on the use or threat of violence lacks the self-sustaining attributes normally associated with ‘order’. It is possible to conceive coercion and consent as two ends of a continuous measure, yet this conceptualization hardly solves the problem at hand. The amount of coercion a state would need to exercise before it crosses the critical threshold between ‘benevolence’ and ‘coerciveness’ is unclear. More critically, all international orders rely on some mix of coercion and consent.
We need to go beyond such conceptualizations of international hierarchy. To evaluate different forms of hierarchy, I draw on republican theory to emphasize that domination characterizes some orders and non-domination characterizes other orders. Domination, to borrow Philip Pettit’s terminology, refers to a situation in which an agent’s freedom of action is wholly contingent on the preferences of another agent. The external, domineering agent possesses certain prerogatives that can affect the subordinate agent’s welfare independently of the latter’s own interests. To use a heuristic example, a slave might seem free and able to pursue its own preferred set of activities. Nevertheless, that very freedom could be a function of the benevolence of a master who still might assert her prerogatives at any given time. Thus, domination does not automatically entail the severe duress that accounts of coercion in international relations theory typically imply. More importantly, under domination, the subordinate has no option to peacefully exit the hierarchy if it so wants. Indeed, existing conceptions of hierarchy neglect to ask whether the subordinate possesses the ability to exercise some outside option or negotiate a sunset clause to terminate the relationship. Situations of non-domination differ because the subordinate agent is not beholden to the interests of the superordinate agent. Rather, the use of coercion under non-domination reflects an earlier decision by the subordinate agent to delegate the ‘power of attorney’ to the superordinate agent so as to uphold the contract between them.
These insights have significance for key debates in international relationships scholarship. Too much analytical confusion arises when scholars attempt to weigh the net balance of coercion and consent underlying a hierarchy. Take, for instance, a debate between Randall Schweller and Robert Jervis on John Ikenberry’s work on international order. Ikenberry contends that, after the Second World War, the United States fashioned a consent-based liberal order in the West. Schweller dismisses Ikenberry’s characterization of the US-led order because it understates the role of coercion that the United States used in advancing its own interests. Jervis alleges that Schweller’s criticism ignores the transformative nature of the institutions undergirding the US-led hierarchy. Specifically, these institutions, though created by the United States, shaped the long-term development of interests and values of the United States so as to narrow the scope of possible coercion (Ikenberry Reference Ikenberry2001; Schweller Reference Schweller2001 and Jervis, Nau and Schweller Reference Jervis, Nau and Schweller2002). This debate, though illuminating, misses the fundamental property of the US-led hierarchy that an emphasis on consent or coercion overlooks: that of non-domination. As I argue below, what distinguished the Western alliance from the Soviet bloc is not what mixture of consent and coercion that characterized either relationship. Instead, in the Western alliance, the relationship took on contractual features that were predicated on a respect for the autonomy and political interests of weaker states. By contrast, the Soviet bloc constituted domination because the autonomy of weaker states (if they had any) was at the mercy of the Soviet Union.
My proposed framework contains an important critique of recent conceptualizations of international hierarchy. I argue that some hierarchies exhibit different properties than those described by theories that these asymmetrical arrangements as contracts (Lake, Reference Lake2009). According to this view, agents consent to a Pareto-efficient but unequal division of rights. I recast this form of hierarchy as non-domination because the superordinate agent enjoys the ‘power of attorney’ by helping to manage the interests of the subordinate agent. Nevertheless, hierarchies that feature domination are not contractual for several reasons. First, under domination, contractual opportunities or clauses for terminating the hierarchy are not available to the subordinate. Second, the leaders that emerge from these societies have oftentimes no alternative but to agree to an arrangement offered to them. Even if a leader were to be ideologically predisposed to the rule of the domineering agent, the lack of available (however ideologically unattractive) alternatives deprives their agreement of being done in good faith. Third, the welfare of the dominated agent is ultimately subject to the predilections of the domineering agent. Such predilections or preferences may not have been revealed at the time when the contract was initially made. Indeed, so-called ‘contracting’ parties face a highly asymmetric capacity to renege on the original contract.
Thinking in terms of coercion and consent obscures important similarities and differences between hierarchies of domination and non-domination. Republican theory enriches our understanding by clarifying important attributes of hierarchy. Fixing our attention on whether states experience domination raises new issues for theory and understanding the world. Do dominated states respond to their superordinates differently than non-dominated states? How do states that have experienced domination in the past choose alignments in the future? If we expect slaves to behave or perceive the world differently from free persons, then we should expect such differences to appear when we look to their international analogues. Adopting the republican framework advanced here recasts questions important to international relations scholars in a way that focuses our analysis on how different hierarchies should influence state behavior.
This paper proceeds as follows. I first highlight, and critically discuss, how scholars refer to levels of consent and coercion to distinguish hierarchies. I then develop the central concepts of domination and non-domination. I demonstrate the superiority of these distinctions by examining Soviet and American relations with Central-Eastern and Western Europe, respectively, during the Cold War. I conclude with some thoughts on implications for theory and research.
Describing hierarchies
Hierarchies are:
‘… ordered relations between units where power and authority is centralized and the units in the system are functionally differentiated. In a hierarchical international order, states are integrated vertically within well-defined superordinate and subordinate positions. In anarchical systems, order is the equilibrium that results from the balancing of a decentralized array of competing states. In hierarchical systems, order is established or imposed by a leading state wielding concentrated power and authority. Hierarchical orders are characterized by stratified relations between leading and secondary states’ (Ikenberry Reference Ikenberry2011, 11).
Discussions of hierarchy tend to center on the different forms of rule practiced by the superordinate state. Though this scholarship has produced invaluable insights regarding international hierarchy, I demonstrate in this section that it has largely thought in dichotomous terms (consent vs. coercion) when describing different frameworks for understanding hierarchies. I then discuss how using such distinctions neglects the subtle but important ways by which consent and coercion often overlap.
Benevolent and coercive leadership models
Duncan Snidal distinguishes between what he calls ‘benevolent leadership’ and ‘coercive leadership’ to denote two contrasting forms of hierarchical rule (Snidal Reference Snidal1985). ‘Benevolent leadership’ describes a situation in which one state with preponderant resources provides all other states in the international system with public goods. These goods are non-rival and non-excludable, meaning that the consumption of a good by one agent does not affect the ability of another agent to consume that good. No restrictions exist as to who can consume these goods. The concept of public goods is central to a theory of ‘benevolent leadership’ called hegemonic stability theory. According to this perspective, the provision by a leading state of public goods (e.g., medium of exchange, sufficient liquidity, and a set of basic property rights that facilitate free trade) is a necessary condition of international stability (Kindleberger Reference Kindleberger1973). Otherwise, states would not be able to overcome collective action problems in producing these goods.
Ikenberry builds on these arguments (Ikenberry Reference Ikenberry2001). Forsaking short-term opportunities for exploitation, the United States opted instead to craft a constitutional international order following the Second World War that grants voice opportunities to weaker states while also credibly committing self-restraint. This rules-based order is ‘benign’: the leading state not only takes into account the interests of weaker states in its own grand strategic calculations, but also because the leading state actively pursues strategies of self-restraint through the use of binding international institutions. This liberal order also enables an open economic system that produces net gains and increasing returns for states that participate in these arrangements.
David Lake also premises his description of international hierarchy on consent. For him, hierarchies involve contractual bargaining in which an agent allocates some amount of sovereign rights to another (superordinate) agent (Lake Reference Lake1996, 7). Hierarchy is, therefore, an exchange of goods or services between the superordinate state and the subordinate state. To gain the weaker state’s compliance the leading state offers such goods as security, order, and economic stability. In part because it supplies these goods, the leading state acquires legitimate authority over the subordinate. The hierarchy is ‘benign’ because neither state would enter into the contractual arrangement unless it is Pareto-improving.
‘Coercive leadership’ describes a situation in which the preponderant state compels subordinates to contribute to the operation of the international order and its general infrastructure (Snidal Reference Snidal1985, 588). Robert Gilpin writes that though it can provide public goods on its own, the leading state nevertheless possesses the ability to extract contributions toward the provision of those goods from the weaker states (Gilpin Reference Gilpin1981). Accordingly, the leading state becomes a central authority that can tax weaker states to help pay for those goods. The distribution of benefits, however, is tilted to its favor.
Alternative conceptualizations of coercive hierarchies exist. Ikenberry writes that imperial arrangements are those in which superordinate states coerce subordinate states into compliance (Ikenberry Reference Ikenberry2011). That is, the imperial state compels adherence from the weak state. According to another perspective, the leader fashions an international order in which subordinate states are functionally differentiated so as to have them render goods or services that cater to its own needs. This iniquitous hierarchy privileges a ‘core’ group of industrialized states at the expense of underdeveloped states located in the ‘periphery’.Footnote 1
In sum, scholars offer diverse conceptualizations of international hierarchy. Many formulations emphasize the extent to which coercion or consent is used. Yet these models of hierarchy exhibit two complementary problems: ‘benevolent leader’ models systematically overstate consent and understate coercion, whereas ‘coercive leadership’ models overstate coercion and understate consent.
‘Benevolent leadership’: overstating consent and understating coercion
Scholars have criticized the ‘benevolent leadership’ model of hegemony for overstating consent. Arthur Stein argues, in contrast to Kindleberger, that the presence of a leading state does not automatically produce an open trading system (Stein Reference Stein1984). The leading state might even have to coerce other states to adopt free trade policies (Gallagher and Robinson Reference Gallagher and Robinson1953). Indeed, states often contain powerful interest groups that would otherwise prefer protectionism.Footnote 2 This form of paternalism risks constituting domination if the weaker state did not freely determine what best advances its own welfare or has not delegated the ‘power of attorney’ to the ‘benevolent leader’. Furthermore, Joanne Gowa challenges the notion that the leading state would invariably prefer free trade (Gowa Reference Gowa1989). As large countries are able to influence the world price of a good, they can impose tariffs that generate more revenue than the amount of consumer surplus that is lost as deadweight.
In a different line of criticism, Schweller alleges that all great powers, including those characterized by liberal democracy, would use coercion in order to advance their interests (Schweller Reference Schweller2001). The United States occasionally displayed a lack of self-restraint during the Cold War, particularly when it was seeking to achieve foreign policy objectives in Southeast Asia and Latin America. Ultimately, the liberal order that the United States allegedly helped establish and support following the Second World War rests on a heavily skewed balance of power. Even if a liberal superordinate state appears to gain the consent of a weaker state, it remains unclear as to whether the weaker state’s decision was genuinely consensual given these power disparities. The weaker state might have had no choice but to acquiesce to the superordinate (Gruber Reference Gruber2000). Thus, the coercive aspects of ‘benevolent leadership’ risk being understated.
‘Coercive leadership’: overstating coercion and understating consent
It is equally difficult to determine whether ‘coercive’ hierarchical arrangements in the manner described above are really so coercive. They accordingly risk understating consent. Terminologically, the phrase ‘coercive leadership’ is oxymoronic: a degree of voluntarism should underlie any relationship in which one agent assumes the role of ‘leader’ and another takes the place of ‘follower’. Assent cannot be compulsory if we wish to properly differentiate the act of following from the act of submission. Yet the use of sticks need not be completely absent in the relations between the ‘leader’ and the ‘follower’. Rather, the use or threat of negative inducements should not be what motivates the ‘lesser agent’ into either participating in that relationship or making a specific action that only the leading state desires. ‘Coercive leadership’ implies less of the voluntarism that leadership should connote and more of the compulsion that a relationship of bondage would entail.
It is unclear why superordinate states should compel subordinate states into contributing toward the provision of a public good. After all, this arrangement might still be Pareto-superior for all the states involved in such arrangements. Further, if the purpose of public goods is to legitimate the leading state, then how can the order still be construed as ‘coercive’ (Gilpin Reference Gilpin1981, 34)? To be sure, the scope of reasons for why a state would choose to obey a rule or commitment is considerably wider than some have previously assumed. Consider the three generic reasons for why states might choose to follow a rule or a commitment. These reasons are fear of punishment, self-interest, and the belief that the rule is legitimate (the rule ought to be obeyed) (Hurd Reference Hurd1999). Gilpin’s account of coerced resource extraction and legitimacy formation fits awkwardly with this rendering of legitimacy.Footnote 3 It is more plausible that in Gilpin’s account states would agree to pay taxes to the leading state for one of the first two reasons for state compliance. The state might be offering its resources without much resistance for fear of retribution. More consistently with Gilpin’s account, the state behaves self-interestedly so as to capture the gains made from the transaction. However, if public goods benefit those subordinates on net, then weaker states should not resist making contributions to it. If the leading state’s aim were to legitimate its own supremacy, then having the costs of extraction outweigh the benefits of the public good would be counterproductive.
This observation highlights another problem of the ‘coercive leadership’ model. Resource extraction intended to produce international public goods implies that the leading state assumes the role of some ‘quasi-government’. The leading state thus provides a (partial) source of central authority and order within the international system. To describe this arrangement as ‘coercive leadership’ is odd because the very same basic arrangements are commonplace in the domestic composition of most states. Democratic and autocratic regimes alike rely on domestic resource extraction in order to render their institutions more effective, finance their militaries, and provide public goods (Huntington Reference Huntington1968, Mastanduno et al. Reference Mastanduno, Lake and John Ikenberry1989, Tilly Reference Tilly1990). Thus, existing conceptualizations of ‘coercive leadership’ appear to draw on a salient feature of domestic political systems and transpose it pejoratively to the international arena.
States vary by the extent to which they use or threaten violence against members of their own societies. However, an important attribute of statehood concerns the ability of the central government to exercise a monopoly on the legitimate use of force. For my purposes, whether the state is democratic or autocratic matters insofar each has different expectations regarding the uses of violence to generate cooperation between members of society and the regime. Yet political scientists resort to metrics (other than coercion) to differential domestic political regimes, including openness to competition, selectorate size, or even prevailing ideology (See, e.g. Bueno de Mesquieta et al. (Reference Bueno de Mesquita, Smith, Siverson and Morrow2003), Dahl (Reference Dahl1971), Geddes (Reference Geddes2003), and Huntington (Reference Huntington1968)). Using coercion/consent to describe international hierarchies becomes less compelling if comparative political scientists themselves rely on other metrics to differentiate domestic hierarchies.Footnote 4
Accordingly, for Gilpin, an international hierarchy can only be coercive if its domestic analogue is that of the autocratic state. Two implications, which Gilpin himself does not explicitly specify, follow from this principle. First, decisions regarding matters of international governance (e.g., public goods provision) are made only by the leading state with little to no input from subordinates. Second, international institutions disproportionately advance the interests of the leading state to the extent that they are detrimental to the welfare of weaker states. However, autocratic regimes that perhaps embody these characteristics do not rely solely on violence. Governing on the basis of violence alone is too costly to sustain political rule beyond the short-run.Footnote 5
Lake’s work reveals another reason underlying the systematic overstatement of coercion. As noted, a leading state acquires authority on the basis of the Pareto-efficient bargain it strikes with the weaker state. Lake thus conceives authority as a relationship between two different actors, one of which possesses rights to issue and enforce orders where as another has a correlative obligation to satisfy its demands (Lake Reference Lake2009). Leading states apply coercion, according to Lake, mainly to sustain political orders that subordinate states collectively regard as legitimate. Consequently, irrespective of the interests of the target, American interventions may be palatable for other subordinate states within a region because they accept American leadership and the rules that it enforces. Coercion is a means to enforce contracts; in other words, the superordinate state uses coercion to discipline wayward states. Coercion must be understood as operating within the wider context of regional cooperation. Nevertheless, Lake’s argument assumes that states entered into these hierarchies with the United States voluntarily and that their leaders do in fact regard American leadership as legitimate.
Summarizing the problem
Analytical challenges appear when using alleged patterns of coercion and consent to distinguish international orders. Theoretical descriptions of ‘benevolent leadership’ neglect how coercion can characterize the formation and maintenance of international order. Conversely, accounts of ‘coercive leadership’ unsatisfactorily deal with the levels of consent often generated to sustain it. Some theories of hegemony explicitly seek to find a balance between the two. Antonio Gramsci applies the concept of hegemony to explain why socialist revolution had failed to take root in southern European societies in the early 20th century. He argues that elites exercised power over the majority of domestic populations by relying not only on coercion and violence, but also on ideological or cultural consensus (Gramsci Reference Gramsci1971). Institutions (such as ecclesiastical authority) help ensure that the subordinate classes positively identify themselves with ‘ruling class’ values (Joll Reference Joll1977). Hegemony, simply put, entails a mixture of coercion and consent.Footnote 6
Yet these conceptualizations face the same analytical problems. Many hierarchies depend on some mixture of coercion and consent. Moreover, ambiguity exists over whether the mechanisms adopted to elicit consent ultimately depend on coercion. If they do depend on coercion, then subscribing to the prevailing ideology may be inauthentic and epiphenomenal if elites remain sufficiently powerful. If not, the presence of consensus-building institutions still remains the product of an agent that was once capable enough to determine their design. In the next section, I suggest borrowing some analytical concepts from republican political theory to overcome the larger difficulties associated with using dimensions of coercion and consent to distinguish between international hierarchies.
Republican political theory and domination
To go beyond such distinctions, I propose framing international hierarchies in terms of domination and non-domination. This approach relies on republican political theory.Footnote 7 In this section, I outline the central concerns of republican political theory and discuss its key concepts before contrasting its view of domination against other conceptions of power found in international relations theory.
Before proceeding, I should clarify what I mean by ‘interest’. My use of this term in the following discussion mirrors the realist assumption that preferences of sovereign units (or agents) are exogenous, fixed, and uniformly conflictual. That underlying preferences are conflictual, however, does not preclude non-conflictual outcomes. As Jeffrey Legro and Andrew Moravcsik write, ‘observed political conflict may be deterred or dissuaded by domination, bribery, threats, or balancing (Emphasis mine, Legro and Moravcsik Reference Legro and Moravcsik1999, 16)’. Accordingly, I seek to describe hierarchies as relations between those societies that, prior to such arrangements, would have bargained with each other over the allocation of scarce resources. Under hierarchy, such bargaining dynamics might lose their salience, but even subordinate political units are at minimum striving to ensure their own survival (Waltz Reference Waltz1979, 118). Thus, in my application of republican theory, I treat superordinates and subordinates as both ‘black-boxes’ for the purposes of analytical clarity and precision. This theoretical move is consistent with how many existing treatments of hierarchy (at least) implicitly embrace the state-as-unitary-actor assumption as well.Footnote 8
Defining domination
According to Pettit, liberal theorists typically conceive liberty as freedom from constraints (Pettit Reference Pettit1997). This understanding, however, neglects the paradoxical possibility that circumstances exist (such as having a ‘benign’ master) in which a slave can be construed as free. To address this concern republican theorists point out that, although the slave may be free from external constraints, its freedoms are ultimately contingent on the preferences or disposition of his master. To use Pettit’s language, the slave is vulnerable to having its freedom of action revoked or subject to arbitrary interference. For Pettit, ‘interference’ refers to the ‘intentional or quasi-intentional worsening of someone’s choice situation … [that] may reduce the range of options available, or … alter the expected payoffs assigned to those options (Pettit Reference Pettit1997, 272)’. With the modifier ‘arbitrary’, Pettit specifically refers to the use of coercion that is independent of the interests of the subordinate actor but correlative of those held by the superordinate actor.Footnote 9 Domination thereby denotes a situation in which an agent’s set of choices may be subject to the arbitrary interference by another agent.
Pettit refines this theory by outlining several axioms so as to further clarify the concept of domination (Pettit Reference Pettit2008, 104–110). First, he affirms the base-line reality of personal choice in which individuals can select some option from a choice menu that they have. Second, a possibility of control by an external, domineering agent exists such that some social relationships can be described as domination.Footnote 10 This external actor exercises arbitrary power as defined above. Third, domination is a relationship or positioning amongst agents in which the subordinate actor cannot countervail the control held by the superordinate actor. There cannot be equality between agents. Moreover, no outside recourse is available to the subordinate agent. This agent cannot appeal to a third-party for relief from domination. Applying these concepts to world politics further clarifies them. Pettit recognizes that some strategies used in interstate relations such as ‘intentional obstruction, coercion, deception, and manipulation’ constitute examples of interference. However, they do not ipso facto constitute forms of arbitrary interference that can occur under domination.Footnote 11
Interference can occur independently of domination. This situation is possible when the interfering agent conditions its actions on the interests of the target agent. As Pettit writes, ‘[t]he person envisaged relates to me, not as a master, but more in the fashion of an agent who enjoys a power of attorney in my affairs (Pettit Reference Pettit1997, 23)’. The agent that experiences non-dominated interference permits it, and can inhibit it, at any point.Footnote 12 When the superordinate agent wields the ‘power of attorney’, the situation approximates the contractual authority that Lake describes whereby the subordinate agent grants the superordinate agent residual rights and other responsibilities.Footnote 13 The superordinate agent then uses these rights to help manage the interests of the subordinate agent and uphold the original contract. Similarly, Ikenberry’s description of constitutional order recognizes the elimination of arbitrary power as a key value (Ikenberry Reference Ikenberry2001, 29–49). In this particular contractual arrangement, the superordinate agent credibly exercises self-restraint and grants voice opportunities precisely in order to ensure that subordinate states’ interests are taken into account. Contractual hierarchy may thus be recast as a hierarchy of non-domination.Footnote 14 Finally, Pettit’s formulation of the term ‘interference’ parallels how international relations scholars understand coercion. In international relations scholarship, coercion refers to a situation in which agent tries to ‘affect the behavior of an opponent by manipulating costs and benefits (Pape Reference Pape1996, 12)’. Agents use coercion to strike a bargain such that the efficacy of coercion relies on some convergence of interests (Schelling Reference Schelling1966). Pettit’s definition of interference is consistent with coercion insofar as the interests of the affected agent are still taken into account. When this sort of tracking is absent, interference becomes arbitrary.
Domination and non-domination are not semantic substitutes for coercion and consent, respectively. The level of invigilation varies as a function of the superordinate agent’s own preferences or the subordinate agent’s ability to ‘hide’ or escape sanctioning. Moreover, consent and coercion can at once characterize the subordinate’s relationship with the superordinate under domination. A benevolent master may permit the slave some freedoms, thereby increasing the possibility of gaining the slave’s acceptance of its position. At certain junctures, however, the master might activate its prerogatives to generate compliance in a manner that is independent of the interests of the subordinate agent. Compliance by the subordinate state takes place in the shadow of this form of power. Conversely, a hierarchy marked by non-domination can encompass many of the features of hierarchy as described by Lake. In such a contractual arrangement, coercion (or ‘interference’) lacks arbitrariness because its purpose is to uphold the contract. That is, under non-domination, compliance follows from non-arbitrary contractual obligations.Footnote 15 Moreover, as Lake indicates, a contract-based hierarchy can still confer upon the superordinate agent political authority that allows it to issue commands that the subordinate state should obey. That coercion can take place with or without domination highlights the need to avoid describing international hierarchies in terms of coercion and consent.
Another distinguishing feature of domination is the absence of any agreed-upon rules or understandings regarding the conditions under which the subordinate can withdraw from the hierarchy. Under domination, as Pettit argues, the subordinate can use its cunning to escape undue interference. Still, whether the superordinate agent is unable to exercise its prerogatives or exhibits benevolence, the subordinate remains at least nominally tethered to it. To break out of this relationship entirely, the subordinate agent has to resort to measures (e.g., violence) that are otherwise circumscribed by the superordinate agent. By contrast, under non-domination, some understanding exists that the subordinate can either abrogate the relationship or set an ‘expiry date’ on it under peaceful terms. The superordinate agent possesses the ‘power of attorney’ within the limits set by its subordinate.Footnote 16
Is it possible to experience domination in one issue area and not another? The answer is a qualified yes. An individual could suffer domination at home vis-à-vis his or her spouse while retaining autonomy at the workplace. This observation can be extended to international politics as hierarchies may differ in their extent over military and economic issues. Theoretically, it is conceivable that states experience domination in the economic sphere and not the military sphere, or vice versa. Yet this point should not be overstated. In international politics, as in domestic politics, issues are often linked and packaged in order to be consistent with particular ideologies or policy preferences. Issue areas are often interdependent and complementary: a state’s interests regarding economic issues might correlate with its interests regarding military issues.Footnote 17
Domination as a non-contractual form of hierarchy
Domination does not comprise contractual arrangements that agents use to obtain Pareto-efficient though unequal outcomes. First, the relationship does not permit the subordinate to unilaterally and peacefully terminate the hierarchy (either through withdrawal or expiry). In contractual settings, opportunities for peaceful withdrawal exist – even if they are costly, the subordinate accepts the costs and ensuing liabilities. Even if subordinates were predisposed to making these arrangements, their agreement was not made in good faith because they were deprived of having an array of competing options from which to choose. Consequently, domination does not accord with aspects of agency theory, which also emphasizes contracting between agents. In principal–agent relationships, problems arise when the agent fails to live up to the group goals specified in a pre-existing agreement.Footnote 18 The critical difference between a principal–agent framework and the republican understanding of domination lies in the premise. To continue the analogy, the master exerts existential control over a slave rather than an indentured servant who forgoes wages and some rights so as to be in the service of an employer for a fixed period of time. Thus, in terms of agency theory, interference is not arbitrary because it is consistent with the contract on which the relationship is based. Second, leaders that emerge in these societies have oftentimes no alternative but to agree to the arrangement bestowed upon them. Describing the surrender of a wallet under gunpoint as consensual on the basis that the victim could have chosen instead to be shot stretches the meaning of consent. Third, the subordinate agent is at the mercy of its superordinate. Theoretically, the superordinate is not bound by any contractual limits in its behavior toward the subordinate.
These latter two observations bear implications for whether legitimate rule can characterize domination. As noted, Lake argues that reliably providing certain goods such as order, stability, and prosperity confers upon the superordinate agent ‘legitimate authority’. This agent could use its ‘legitimate authority’ to obtain further compliance from the subordinate, even applying coercion in order to preserve its own rule. Yet, in a situation marked by non-domination, Pettit argues that an important condition of legitimacy is whether the government is ‘subject to effective, popular influence’. This influence, he adds, should reflect the values and associations of the citizenry (Pettit Reference Pettit2010a, 144–149). The absence of these constraints constitutes domination. A domineering agent might supply order and stability, but still lack legitimacy if its rule fails to satisfy the conditions that Pettit describes.
Analogical reasoning helps clarify the distinction between domination and Lake’s conception of authority. Consider a course taught by a professor and with any number of students. The professor begins the course by offering a syllabus. This document summarizes the material to be covered in the course, outlines the classroom rules and grading policy, and explains the tasks students are expected to complete for course credit. The syllabus constitutes a contract between the professor and her students. It defines each of their respective roles. The students are subordinate to the professor, and the professor even enjoys legitimate authority as her qualifications confer upon her a status that elevates her above the students. Exercising these rights, the professor can use various positive (e.g., praise and encouragement, bonus marks) and negative inducements (e.g., grading penalties) to universally motivate students to perform classroom tasks in a timely and satisfactory manner. On her part, the professor cannot propose radical changes to the syllabus and apply a different (or arbitrary) grading policy without first gaining the informed consent from her students. This situation is consistent with how Lake describes hierarchy: a contractual understanding between superordinate and subordinate agents in which the former possesses a legitimate right to rule.
Consider next a situation in which an armed robber takes several individuals hostage during an attempted heist. These individuals simply happened to be at the wrong time and place, and are now unable to leave the supervision of their captor. This hostage situation is another form of hierarchy insofar as the robber attempts to constrain the movements of her captives by fiat. Should a captive fail to submit to her demands, the robber might have to punish him in order to reassert her command. Yet hostage situations also tend to exhibit several peculiarities. First, captors cannot fully deprive their hostages certain material resources, particularly if the hostage-taking lasts a while. They are sometimes willing to concede certain items such as nourishment in order to ensure their captives’ quiescence. Second, and more importantly, hostages can develop emotional attachments to their captors. Known as the ‘Stockholm Syndrome’, this phenomenon occurs when hostages self-identify with their captors and develop sympathies with their captors. Thus, the balance of coercion and consent becomes ambiguous even in such a forceful act as a hostage-taking. Domination characterizes this situation because the safety of the hostages depends on the whims of their captors. A contractual description of hierarchy does not properly account for this dynamic.
Students might bristle at these analogies, yet it is apparent that significant differences exist between an academic course and a hostage situation. Students are not at the mercy of their professor’s whims and diktats. They voluntarily undertake an academic program and enroll in its courses. They are free to choose how much effort they wish to exert toward fulfilling course requirements. Conversely, professors uphold certain responsibilities in accordance with the expectations of their students.Footnote 19 By contrast, individuals are held hostage due to circumstances beyond their control. They may obtain concessions and develop an emotional bond with their captor, but they are constrained in determining the level of effort in submitting to the captor’s demands.Footnote 20 Moreover, the captor could suddenly withdraw these concessions and jeopardize the safety of her captives so as to advance her own interests. Domination is, therefore, qualitatively different from contractual authority.
Domination, as constituted by a hostage-taking scenario described above, is in fact a feature of international politics. The colonialism practiced by European powers during the late 19th and early 20th centuries in much of Africa represents a historical example. European states claimed rights over African territories, monitoring these societies to ensure compliance. Domination initially was not coercive. Some observers of European colonialism remark that the nature of European colonial relations on the continent was of ‘informal empire’, which relied more on the willing collaboration of local indigenous groups. Nevertheless, ‘informal empire’ changed into ‘formal empire’ – a situation increasingly marked by assertions of military power and formalized rule. This change was independent of the interests of (pre-colonial) African sovereign entities because it was a function of the imperial competition amongst European powers (Smith Reference Smith1981, 47).
One plausible counterargument is that imperial states exercised their rule through local intermediaries and collaborators. Accordingly, colonial governance still featured contractual characteristics. This view is flawed. First, in advancing their interests, imperial authorities often reorganized and restructured local societies in order to manipulate local preference formation. Indeed, imperial institutions fostered discursive practices and structures intended to socialize elites and non-elites into accepting colonialism (Ikenberry and Kupchan Reference Ikenberry and Kupchan1990). Thus, Western colonialism is not a story of autonomous agents identifying bargains that reflect overlapping win-sets. Second, as David Abernethy writes, accommodation to imperial rule was often an ambiguous response rooted in neither approval nor disapproval (Abernethy Reference Abernethy2000, 302–303). Accommodationist attitudes were a function of diverse motives, including the desire to develop a stronger capacity to contest imperial rule. Third, the costs of exit were potentially existential.Footnote 21 Whatever their motives, local elites and collaborators were held captive. It was impossible to switch, for example, from Belgian rule to British rule. Nor was peaceful unilateral withdrawal possible until emerging anti-imperial norms and empire fatigue encouraged decolonization after the Second World War.Footnote 22 Domination characterized these relationships.
We can now delineate two ideal-type forms of hierarchical orders. Hierarchies of non-domination are arrangements in which the superordinate agent might use coercive instruments against weaker agents in order to produce mutually desirable policy outcomes. The use of these instruments is such that the political existence of the weaker state does not depend on the superordinate. The welfare of the weaker state is not contingent on the disposition of the more powerful state. Hierarchies of domination are arrangements in which weaker agents might face a situation whereby they have no choice but to submit to whatever is demanded of them. This situation might not occur if the stronger state grants the weaker state with some freedoms or if the weaker state’s leadership is cunning enough to evade reprisals for insubordination. Nevertheless, any provision of liberty here depends on the behavior of the strong state.
Domination and other conceptions of power
Applying the republican conceptual framework to interstate relations raises new insights for conceiving power in world politics. States can exercise interference on the basis of material power projection, as per realist theory. However, interference may not necessarily be based on material power. Interference might even include more indirect forms of coercion such as social practices or conventions. Laura Valentini distinguishes between what she calls ‘interactional coercion’ and ‘systemic coercion’. ‘Interactional coercion’ takes place when ‘an agent A coerces another agent B if A foreseeably and avoidably places nontrivial constraints on B’s freedom, compared to B’s freedom in the absence of A’s intervention’. This definition resembles Pettit’s definition of interference. By contrast, ‘systematic coercion’ is described as ‘a system of rules S … if it foreseeably and avoidably places nontrivial constraints on some agents’ freedom, compared to their freedom in the absence of that system’. A system of rules includes informal or formal institutions that an individual (or group) agent installs to ensure other agents conform to desirable patterns of behavior (Valentini Reference Valentini2011, 210–212). Domination can feature this latter form of coercion. In delineating the duties, obligations, and even privileges imposed on the subordinated agent, the system of rules constitutes domination if the subordinated agent’s interests are not taken into account. Indeed, the archetypal example of the master and slave requires a certain normative structure that defines the identities those two agents have of themselves and their relationship with each other.
Republican conceptions of power fit uneasily in existing schematics for understanding power in world politics. One oft-cited typology describes four conceptions of power along two dimensions (Barnett and Duvall Reference Barnett and Duvall2005). The first dimension concerns the relational specificity that power is projected, distinguishing between direct and diffuse forms of power. The second focuses on the channels through which power produces its effects. These channels include the specific interactions of agents or how their social relations are even constituted. The standard definition of power in the political science literature serves to capture the ability of an agent A to get agent B to do something that they would not otherwise do. This view of power is a more compulsive form of power that operates directly through the interactions of agents.
Because it is consistent with standard definitions of coercion, non-arbitrary interference is in line with the foregoing conception of power. Yet it is unclear where to properly situate arbitrary interference in this taxonomy. On the one hand, with domination its superior status confers upon the superordinate state privileges vis-à-vis the subjugated agent. One may argue that structural power characterizes this relationship due to ‘the very social capacities of structural, or subject, positions in direct relation to one another, and the associated interests, that underlie and dispose action (Barnett and Duvall Reference Barnett and Duvall2005, 52–53)’. An agent can only be considered to be a ‘slave’ if and only if another agent can be identified as a ‘master’ – they are mutually constituted agents. On the other hand, the problem with understanding domination as simply ‘structural’ is that the actual behaviors of the agents embedded in these relations are not necessarily pre-determined. A slave need not behave in a manner consistent with our expectations of that role if he were sufficiently cunning or enjoys certain latitude that the master permits him. Either agent might have the ability to choose their actions according to their interests; indeed, the slave can still act like a free individual under special circumstances despite experiencing domination. The slave might not behave slavishly. Only when the master threatens or practices interference without heeding the interests of the slave does domination feature compulsion. Put differently, the structure of the relationship defines the scope of control available to the domineering agent, who may or may not be able to enforce some level of direct compulsion on the part of the subordinate agent.
The meaning of domination differs from Gramscian notions of hegemony. In Gramscian political thought, hegemony constitutes both a process and outcome by which elites maintain power by foisting upon the population their preferred ideology, distributing resources, and structuring daily life (Flic Reference Flic2009, 121–122). Hegemony is a social phenomenon in which ruling classes generate the consent from subordinate classes in sustaining the prevailing political, social, or economic system (Gramsci Reference Gramsci1971, 461 and 258–259). However, domination does not rely on an account that emphasizes the deliberate production of consent. More critically, the use of coercion under Gramscian hegemony is not ‘arbitrary’. Any use of coercion must be consistent with the ideological imperatives that society members have already recognized as legitimate. Indeed, ideological consistency is not a necessary feature of ‘arbitrary interference’. The preferences of the superordinate agent can change or be revealed in a way that the execution of ‘arbitrary interference’ does not align with an established system of beliefs. Finally, republican theory is more complete because it still offers an account of hierarchy with non-domination. Gramscian theory does not offer a coherent view of a hierarchy in which neither ideological hegemony nor an exclusive reliance on violence exist.
Domination and non-domination in cold war Europe
I have already alluded to European colonialism in Africa and the transformation of ‘informal empire’ to ‘formal empire’. But what are other historical examples of domination and non-domination in international politics? To further illustrate the utility of these concepts I juxtapose two more recent examples, that of Soviet relations with Central-Eastern Europe and American relations with Western Europe during the Cold War. The Soviet bloc constituted domination, whereas the United States had the ‘power of attorney’ over Western Europe. Non-domination characterized the latter set of relations.Footnote 23 To argue that the Soviet bloc featured domination, I must show that the satellite states were subject to ‘arbitrary interference’. I also need to demonstrate the inadequacies of alternative conceptualizations of hierarchy. Regarding the Western alliance system, I affirm the applicability of contractual (and constitutional) theories of hierarchy while noting that this arrangement embodied non-domination.
Soviet domination
Nazi defeat in Central-Eastern Europe and the weakness of local societies that had already endured 6 years of warfare and Nazi occupation ensured the Soviet Union an extremely favorable geopolitical position in the region. Soviet leader Joseph Stalin recognized the opportunity to gain considerable influence, noting that ‘[t]his war is not as in the past; whoever occupies a territory also imposes on it his own social system. Everyone imposes his own system as far as his army can reach’.Footnote 24
The Soviet Union imposed communist regimes across the region during the immediate post-war period. Once marginalized communists filled the power vacuum in East Germany. In Poland, the government-in-exile could not assert its authority while local communists (concentrated in the Lublin Committee) were acquired control over Polish territory with Soviet assistance. The Soviet military exercised de facto control over Romania after 1944, ensuring the ascendancy of a communist party to power. Hungary similarly became a Soviet satellite. In Czechoslovakia, however, an unpopular communist party assumed control of the country through the overthrow of a coalition government with Soviet backing. A Bulgarian communist resistance party took over shortly after the Soviet Union declared war on Bulgaria. All of these states remained communist until 1989.
Coercion and consent in the Soviet bloc
Existing conceptualizations of hierarchy have made little sense of the ambiguous nature of Soviet power in Central-Eastern Europe. Consider how the Soviets exercised brutally repressive policies in establishing subordinate regimes. According to Lake, Soviet leaders adopted these arrangements because they feared high levels of opportunism from these regimes (Lake Reference Lake1996, 26). The Soviet Union’s relative size posed a regional threat to its smaller neighbors. Furthermore, orientation toward the Soviet Union rather than the West was distasteful because many of these states saw themselves as ‘European’, ‘Western’, or non-Orthodox Christian. For both realist and ideational reasons, therefore, these subordinate states would have otherwise balanced against the Soviet Union. Concerns over alliance withdrawal were warranted. After all, members of the societies governed by these regimes mounted occasional protests (e.g., the 1953 demonstrations in East Germany, the Poznan 1956 protests in Poland, non-violent protests following the Warsaw Pact invasion of Czechoslovakia in 1968) and uprisings (e.g., the Hungarian Revolution in 1956). More frequently, individuals practiced subtler forms of resistance and contestation under communist rule in Central-Eastern Europe (McDermott and Stibbe Reference McDermott and Stibbe2006).
Nevertheless, active collaboration took place at both the elite and mass levels. Either out of ideological affinity or political opportunism, some individuals sought to not only lead their national communist parties but also to work with the Soviet Union. For example, Boleslaw Bierut, was a Polish communist and had even been a member of Stalin’s police agency. When the Second World War broke out, he helped organize the new Polish Worker’s Party, enabling him to be the Polish President following the Soviet takeover of Poland. Bierut was not unique in his genuine affinity for communism and the Soviet Union. Czechoslovakia’s Gustáv Husák, East Germany’s Erich Honecker, Hungary’s János Kádár, and Romania’s Gheorghe Gheorghiu-Dej all engaged in communist party activities before the Second World War. When these leaders did engage in opportunism, it was to exploit Soviet anxieties in order to extract greater concessions rather than outright defection.Footnote 25 However, it is unclear whether the risks of opportunism stayed constant or declined with the ascendancy and consolidation of more ideologically attuned leaders. At the mass level, communist authorities often relied on positive inducements to generate political support. In Czechoslovakia, 1.7 million members of a total population of about 10 million were members of the Communist Party in 1966. However, it is unclear whether the risks of opportunism stayed constant or declined with the ascendancy and consolidation of more ideologically attuned leaders. At the mass level, communist authorities often relied on positive inducements to generate political support. In Czechoslovakia, 1.7 million members of a total population of about 10 million were members of the Communist Party in 1966 (Wightman and Brown Reference Wightman and Brown1975, 410). Many joined the party so as to receive such benefits as education and improved social mobility. Members of these societies often found positive elements in the transformative policies that communism brought. Even Czesław Miłosz, author of the anti-totalitarian The Captive Mind, remarked that ‘I was delighted to see the semi-feudal structure of Poland finally smashed, the universities opened to young workers and peasants, agrarian reform undertaken, and the country finally set on the road to industrialization (Michnik Reference Michnik2011, 102)’. In short, periodic instances of outward resistance should not lead us to overlook the benefits that communist regimes used to appease members of the societies they ruled.
Classifying Soviet-led arrangements as either coercion-based or consent-based is problematic. Though Lake does not depend on a description of the Soviet bloc as more ‘coercive’ than ‘consensual’, he refers to coercion as a means upon which the Soviet Union relied to control the region. Lake’s evaluation of the Soviet bloc is conflicted. He writes that ‘[a]ll relationships, whether entered into voluntarily or as a result of coercion, can be considered as based upon some “contract”’ but then later notes that ‘coercion is a substitute for contracting (Lake Reference Lake1996, 7 and 19)’. Thus, Soviet arrangements in Central-Eastern Europe were ‘relatively hierarchic’ and constituted an ‘informal empire’ that contrasts with the ‘relatively anarchic’ character of American relations with Western Europe (Lake Reference Lake1996, 23). However, in a later refinement of his theory, Lake notes that the Soviet Union exercised ‘domination’ over Central-Eastern Europe but does not offer analytic reasons for using this very term. He adds that the Soviet bloc featured both high levels of both security and economic hierarchy (Lake Reference Lake2009, 58–59). Nevertheless, the difference between the Soviet bloc and the US-led alliance system is a matter of the quantity of hierarchy and not the quality of hierarchy.
Republican theory and the Soviet bloc
Republican theory sheds light on the hierarchical nature of the Soviet bloc. Recall that domination characterizes a hierarchy when the superordinate state can assert its primacy without regard for the interests of its subordinates. Domination also exists when the relationship does not permit ‘exit clauses’ or other unilateral but peaceful opportunities for withdrawal. In applying republican theory, several questions are worth raising. Did the Soviet Union assert its preferences independently of its satellite states? Were satellite states able to willingly terminate their relationship without their security directly threatened by the superordinate state?
On the first question, it is evident that pursuing wayward policies elicited the Soviet use of force. The secret trial and execution of Hungarian leader Imre Nagy following the Hungarian Revolution of 1956 demonstrated the Soviet Union’s readiness to use force against those leaders who pursued policies that were inconsistent with its interests. This lesson was reinforced later with the ouster of Czechoslovak leader Alexander Dubček during the Warsaw Pact invasion of Czechoslovakia in 1968. However, even more indicative of a satellite leader’s precarious situation was the alleged treatment of Bolesław Bierut. A staunch Stalinist and supporter of the Soviet Union, Bierut died in mysterious circumstances shortly after Nikita Khrushchev denounced Stalin’s leadership in his ‘Secret Speech’ at the 20th Congress of the Communist Party of the Soviet Union (Kemp-Welch Reference Kemp-Welch2008, 72–73).
The implication of Bierut’s death for other satellite leaders was that firm ideological support and sycophancy did not guarantee political or even personal survival. The Kremlin’s preferences independently of those of Bierut, demonstrating that Bierut’s freedom of action and leadership had always been at the mercy of his Soviet superiors. A contractual theory of hierarchy would make little sense of Bierut’s experience because it would not be able to account for the killing of seemingly docile subordinates. What were the terms of the contract that Bierut signed and was later mortally punished for violating? It appears instead that Bierut experienced domination and was subject to an extreme form of arbitrary interference.
Domination in the Soviet bloc became official with the articulation of the Brezhnev Doctrine. Initially serving to justify the suppression of the Prague Spring, Soviet leader Leonid Brezhnev asserted that ‘[e]ach Communist party is free to apply the principles of Marxism, Leninism and of socialism in its own country, but it is not free to deviate from these principles if it is to remain a Communist party… The weakening of any of the links in the world system of socialism directly affects all the socialist countries, and they cannot look indifferently upon this’. Put differently, all members of the Soviet bloc shared a common fate, and any deviation from the preferences of the Soviet Union was strictly forbidden. Indeed, this foreign policy statement embodies a form of invigilation that is consistent with Pettit’s account of domination. Satellite states were now ‘on watch’. There was no escape from the hierarchy imposed by the Soviet Union.
Some might contend that actions invoking the Brezhnev Doctrine would not qualify as ‘arbitrary interference’ for several reasons. One possible argument is that the doctrine is explicit and transparent, implying that any enforcement on this basis loses its arbitrariness. Yet Soviet interpretations of Marxist-Leninist principles were hardly stable throughout the Cold War. Soviet leaders oftentimes debated over the ‘correct’ meaning of these principles. Official dogma evinced change with each new party chairman. Thus, the Brezhnev Doctrine provided no ideological guidance but to confirm that leaders of satellite states were beholden to the Kremlin. A second argument might claim that political elites in Central-Eastern Europe approved of the Brezhnev Doctrine. After all, many Warsaw Pact members contributed to the invasion of Czechoslovakia in 1968. Unfortunately, the sincerity of any such approbation is difficult to determine. Those elites had incentives for showing support for Soviet actions, not least because they wanted to bolster their pro-Soviet credentials and avoid a similar fate that befell the Prague Spring.
Regarding the second question posed earlier, only one satellite state succeeded in sustaining a foreign policy that diverged from the Soviet bloc. During the Prague Spring, Romanian President Nicolae Ceauşescu condemned the invasion of Czechoslovakia, aligned with communist China, and openly antagonized the Soviet Union. Nevertheless, partly as a consequence of this decision, he later believed that his survival was at stake (Deletant Reference Deletant2007, 499–500). Ceauşescu still felt that the Soviet Union invigilated over his actions in a way that is consistent with a relationship marked by domination (Deletant Reference Deletant2007, 88–89). Recall that a slave of sufficient cunningness might still act as if it were free. If the most defiant satellite leader felt this way, it is conceivable that even more docile leaders had similar fears and anxieties.
US non-domination in Western Europe
The alliance system in Western Europe originated less with American pretensions for domination and more with the United States acting as an ‘attorney’. That is, the United States helped manage the economic and military affairs of Western European societies in order to strengthen those societies and make them one day independently capable of deterring Soviet aggression. After all, Western Europe also experienced significant devastation and social disruption during the war and its aftermath. For American decision makers, the mistakes of post-conflict macroeconomic management from the interwar period were too vivid to ignore. To reconstitute Western European economies, and thus the global economy, the United States offered an extensive package of economic grants embodied in the Marshall Plan. Despite the ambitious scope of this program, American decision makers did not believe initially that their involvement in European affairs would be indefinite. Rather, they believed that the economic restoration of Europe would allow them to leave to politically and militarily disengage from those societies (Judt Reference Judt2006, 109). Nevertheless, the Berlin Airlift and other Soviet maneuvers in Central-Eastern Europe as well as the Korean War convinced American decision makers that a robust presence on the continent was necessary to check Soviet aggression (Gaddis Reference Gaddis1982, 87–124). Barely 5 years after the Second World War the United States established the North Atlantic Treaty Organization (NATO). This action ensured an American presence on the continent and provided a powerful bulwark against the Soviet Union.
How to characterize American primacy in Western Europe during the Cold War has provoked much debate in international relations scholarship. Lake, as noted, describes American alliance relations with Western European countries as ‘relatively anarchic (Lake Reference Lake1996, 1)’. There was little fear that these states would defect to the Soviet Union, especially as they relied on the American economy for their recovery. To assert greater control over them and establish a tighter form of dependency in the manner of the Soviet Union would have entailed exorbitant costs. In later work, however, Lake describes these same relations as hierarchical, at least during the first half of the Cold War (Lake Reference Lake2009). Western European ‘contracting’ featured the combination of NATO, trade dependence, and currency regimes tied to the US dollar. Similarly, Ikenberry confers upon the United States a superordinate role in Western Europe that decisively shaped the pattern of intra-regional relations within a constitutional order (Ikenberry Reference Ikenberry2001). The American-led system was Pareto-improving, generated increasing returns, and allowed allies opportunities to voice their concerns. Geir Lundestad describes the United States presence in Western Europe as an ‘empire by invitation’, arguing that ‘the Europeans even more strongly than most others attempted to influence the Americans in the direction of taking greater, not lesser, interest in their affairs (Lundestad Reference Lundestad1986, 268)’. These works highlight that American-led alliance system is that of a hierarchy that exhibited major differences with the Soviet bloc.
Still, the Western Alliance experienced conflict and United States coercion, particularly during the 1950s and 1960s. The United States used its financial clout to pressure Great Britain as well as France into withdrawing from the Suez in 1956 (Kunz Reference Kunz1991). Further, due to the expenses involved with its extensive military presence on the continent, successive US governments practiced coercion to encourage its allies (e.g., West Germany) to ‘offset’ those costs and refrain from economic behavior that risked undermining the Bretton Woods system (See Gavin Reference Gavin2004). The United States contributed to alliance disarray in its efforts to discourage Great Britain from maintaining its independent nuclear deterrent (Costigliola Reference Costigliola1984). The West German government also resisted American pressure to sign onto the Nuclear Non-Proliferation Treaty (NPT) before finally assenting when Willy Brandt became Chancellor. These frictions sometimes provoked anti-Americanism at the mass level, yet detractors of the United States hardly articulated anything more substantial than rhetoric.Footnote 26
Republican theory and the Western alliance
The Western alliance exemplified non-domination. The United States practiced coercion in order to advance agreed-upon group objectives and prevent free riding. Moreover, states that sought to reduce the subordinate status vis-à-vis the United States remained assured of their security. To begin, the foregoing efforts at coercion, nevertheless, lacked the significance of what had occurred in the Soviet bloc. The United States did not articulate an analogue to the Brezhnev Doctrine to its Western European allies.Footnote 27 In fact, the United States predicated these efforts in coercion on maintaining a unified Western alliance that would be strong enough to independently deter Soviet aggression. This particular goal had motivated institutional development (e.g., the Bretton Woods system) and formed the basis of alliance relations (Trachtenberg Reference Trachtenberg1999). For example, US insistence that allies make offset payments was intended to preserve both the Bretton Woods system and the US military presence on the continent – two things that Western European allies saw as beneficial (Gavin Reference Gavin2004).
With this observation in mind, calling the US presence in Europe as an ‘empire by invitation’ is similar to Pettit’s description of the ‘attorney ’ (Pettit Reference Pettit1997, 23). Western Europeans desired the Marshall Plan, despite its conditions regarding trade liberalization. Several states were especially keen in participating in American-led defense arrangements and receiving security guarantees. Yet they sometimes failed to meet US expectations for their contributions to the US-led international order. In such instances, the United States practiced coercion so as to ensure that mutually desired policy outcomes were attainable.
Finally, the Western alliance was never meant to exist in perpetuity: its founding purpose was to reconstruct and strengthen European societies vis-à-vis the communist threat. As the alliance matured, opportunities for withdrawal and dampening the hierarchy became available to the weaker states. Analogous to Romania’s readjustment of its relationship with the Warsaw Pact, France withdrew from NATO’s military command structure. This action, however, did not trigger the same apprehensions regarding superpower punishment. Moreover, American leadership in this region was largely a function of its relative economic power. When the United States no longer could meet its obligations to the Bretton Woods system, its role in Western Europe changed substantially, particularly with the emergence of the European Economic Community.Footnote 28 To some extent this development had been a key objective of American foreign policy. The United States assumed the ‘power of attorney’ so long as its presence was necessary for reconstituting Western Europe and deterring Soviet aggression. That the United States performed this role is consistent with Lake’s description of contractual hierarchy. By contrast, the Soviet Union maintained its hold on Central-Eastern Europe until economic stagnation, rising nationalism, and an expensive military intervention in Afghanistan led to its political disintegration. The articulation of the Brezhnev Doctrine took place at a time when both Western European and Central-Eastern European states had recovered economically from the Second World War. Although the Soviet Union became revanchist toward its subordinates, the United States began to curtail its involvement on the continent.
Conclusion
Descriptions of international hierarchy are often premised on a particular balance of consent and coercion. I show that placing emphasis on these terms is conceptually problematic. ‘Benevolent’ models of leadership neglect how superordinate states can exploit power asymmetries. The consent of weak states is also difficult to disentangle when a skewed distribution of power exists. ‘Coercive’ models of leadership ignore how some consent is needed to make any political order durable. Empirically, observers often highlight that allegedly benign and coercive regimes still feature coercion and collaboration, respectively. Taken together, these concerns cast doubt on the utility of using levels of consent and coercion to distinguish between international orders.
Drawing insights from republican political theory, I argue that we can better distinguish hierarchies on the basis of whether they feature domination. Under domination, a subordinate state is at the mercy of the leading state’s interests. The subordinate state might even behave as if it were autonomous, electing to choose policies at its discretion. But such freedom is illusory – it is contingent on the predilections of the superordinate state, which can assert its supremacy whenever and (possibly), however, it may please. Moreover, the subordinate state does not have the legal ability to terminate the relationship, either through withdrawal or expiry. By contrast, hierarchies of non-domination are relationships in which the leading state assumes the ‘power of attorney’, by which it has advanced permission to coerce in order to advance an agreed-upon goal. This arrangement does not grant the superordinate state status or prerogatives over the subordinate state. The subordinate state enters the hierarchical relationship with the understanding of the contractual terms governing its termination. These categories improve upon existing theories of hierarchy by taking into account how coercion, as understood by international relations scholars, can occur in both domination and non-domination. In only one of these situations, that of domination, coercion is practiced without regard for the interests of its target. I contrast Soviet and American relations with their subordinates in Central-Eastern and Western Europe, respectively, to give examples of hierarchies that feature domination and non-domination, respectively.
My conceptual analysis opens up new theoretical avenues and empirical questions on issues relevant to international relations scholars. To begin, we should expect observable differences in the behavior of dominated and non-dominated subordinates. For example, Lake describes how conventional military deployments signify a subordinate state’s security subordination to its patron (Lake Reference Lake2009, 68–69). In particular, the patron’s military presence could constrain and protect the subordinate while also reassuring it so as to allay fears of abandonment. Accordingly, unilateral reductions in the military presence might reflect the patron’s desire to renege on the alliance contract it had signed with the subordinate, thereby provoking fears of abandonment. If non-dominated subordinates were to fear abandonment, then they might undertake nuclear proliferation, diplomatic confrontation, and other actions intended to displease the patron. Under domination, however, unilateral military redeployments do not signify ‘contract-breaking’ because no contract exists. Consequently, dominated subordinates have less reason to develop fears of abandonment. This observation suggests that interstate communication operates differently under domination than under non-domination. What might be a costly signal of reassurance and commitment in one context might not have the same meaning in another.
Experiencing domination should also influence the form by which subordinates undertake collective action. Lake claims that under (contractual) hierarchy states might practice symbolic obeisance, defined as ‘costly acts that do not involve direct compliance with commands but are nonetheless public, often collective displays of submission that acknowledge and affirm the authority of the ruler’ (Lake Reference Lake2009, 165). Because they hold the ruler as authoritative and legitimate, subordinates in these relationships are less likely to practice civil disobedience – the ‘inverse’ of symbolic obeisance. Therefore, states are more likely to joining a military coalition led by their patron (Lake Reference Lake2009, 173). Recent research suggests, however, that states do not ‘genuflect toward authority’ by coalition participation without getting anything in return (Henke Reference Henke2012). Just as contractual hierarchy reflects a bargain, so does participation in US-led coalitions. Yet it might be the case that under domination genuflection does occur because weak subordinates are cowed into following the ruler’s lead. Nevertheless, if their ruler were to weaken considerably, then they might collectively engage in civil disobedience so as to flaunt the ruler’s authority. Put together, patterns of domination or non-domination should affect costly signaling and shape collective action. This conceptual framework recasts these issues familiar to international relations theory in a way that was not possible under existing approaches for understanding hierarchy
Acknowledgments
I thank Michael Hunzeker, Ian Hurd, G. John Ikenberry, Richard Jordan, Chris A. Kendall, Anna Lanoszka, Leonard Winogora, Sherry Zaks, and the co-editors of International Theory for their generous support and feedback. I greatly appreciate the excellent criticisms made by three anonymous reviewers. An earlier draft of this paper was presented at the 2011 CPSA Annual Conference in Waterloo, ON. All errors and other infelicities are my own.
