Voluntary action in International Relations

Abstract

What is voluntary action in International Relations (IR)? Literatures on bargaining, empire, foreign policy, grand strategy, hierarchy, hegemony, international law, international order, legitimacy, responsibility, status, and strategy depend on the answer, yet IR has no definition of it. We argue that voluntary action is an expression of genuine will. Yet, because of theoretical commitments, there remains deep disagreement about which actors are capable of genuine will, what process produces it, and what internal and external conditions interfere with it. We lay out where and why disagreements occur and discuss ways of advancing disputes over voluntarism in world politics.

Introduction

What is voluntary action in International Relations (IR)? Legitimacy and responsibility depend on the answer. For instance, the United States has declared that it would accept the “peaceful voluntary reunification of Taiwan with China” (Kastner and Rector, 2008: 63), though how it would render a judgment in the face of gray zone actions remains opaque (Grossman, 2025). Russia has called on other states to legally recognize its annexations of Ukraine, claiming they reflect “the free and voluntary will” of the people (Grant, 2015). In international migration, “forced” migrants, such as refugees fleeing political persecution, are granted more robust legal protections than “voluntary” migrants, who cross borders for economic or familial reasons (Ottonelli and Torresi, 2013: 783–784). In the causes of war literature, states are compared to “sleepwalkers” with far-reaching analytical and normative implications (Clark, 2012).

Academically, literatures on bargaining, empire, hierarchy, hegemony, order, and status are rooted in voluntary action. In security and economic hierarchies, for example, the kind of authority that dominant states exercise over their subordinates is contrasted with the “simple voluntary cooperation” (Lake, 2009: 13) one finds under anarchy. The primary way we detect status hierarchies is through the “voluntary deference” that low-status states grant their high-status counterparts (Larson et al., 2014: 10; Matsuda, 2022). While some hegemons rule through coercion, others build international orders through the provision of public goods to “instill a degree of voluntary compliance with the hegemon’s preferred order” (Clark, 2009: 207–208).

All IR paradigms, or families of theories, rest on a concept of voluntary action, but none have a defensible definition of it. In the otherwise encyclopedic Traditions of International Ethics and the Oxford Handbook of International Political Theory, the word “voluntary” does not appear in either index and is not defined in either text, though both works rely on it repeatedly (Brown and Eckersley, 2018: 120–122, 138, 213, 226, 295, 324, 390, 642; Nardin and Mapel, 1992: 130, 146). From an economic angle, Hirschleifer (2001: 187–191, 301, 313) equates voluntary action with decisions made without force or fraud, but that is an oversimplification, because as we will see, compulsions can come in many forms. In his study of diplomacy, Ramsay (2011: 1012) considers actions voluntary if “countries can choose to play strategies where they do not bargain seriously if they become discouraged by the diplomatic exchange.” But this reduces voluntarism to a rare node on a game-theoretic tree. In his study of voluntary unions among states, Parent (2011: 5, 10–12, 154) vaguely defines voluntary action in terms of equality of power, then incongruously finds that unions between equals are “shotgun weddings.” In their work on international migration, Ottonelli and Torresi (2023: 408) define voluntary choices by what they are not: voluntary choices are not the result of use or threat of force, not prompted by the lack of acceptable alternatives, not impossible or unacceptably burdensome to reverse, and not based on false information. Yet this does not tell us what voluntary action is or how to apply the concept beyond migration.

We argue that voluntary action is an expression of genuine will. That requires an actor who can will, a process that produces genuine will, and internal and external conditions that permit the expression of genuine will. Paradigms vary, however, about who expresses will, what processes make will genuine, and what internal and external forces counterfeit that will. We develop a typology to map these debates and help advance them. Yet while paradigms differ on how much freedom they expect actors to possess under anarchy, all agree that voluntary action is a meaningful category in world politics, and many are strange bedfellows. Some view world politics as a world of chains, in which internal and external compulsions undercut voluntary action in all but extreme conditions. Others view world politics as a world of possibilities, in which true duress is rare. Still others view things somewhere in the middle, pointing either to internal pathologies or external predators as the main obstacles to genuine will.

To develop our case, the first section reviews conventional wisdoms and offers a shareable definition of voluntary action in world politics. The second section expands on the elements inside the definition. The third section presents a typology that organizes theoretical disagreements about voluntary action. The penultimate section applies the analysis to events to demonstrate the potential of our approach, and the final section summarizes our chief points.

Voluntary action in the literature: debates and definitions

A traditional perspective is that the prospects for voluntary action in world politics are bleak. International politics is depicted as a “brutal arena” (Mearsheimer, 1994: 9) or a “house on fire” (Wolfers, 1962: 13), where force is not only “the ultima ratio, but . . . the first and constant one” (Waltz, 1979: 113). On this view, voluntary action is close to an empty set in world politics because compulsions are too common and coercive. It is unsurprising that a hegemon can bully weaker states or that powerful global financial institutions can dictate terms to potential defaulters; this is simply what happens in a world with wide asymmetries in military power and vast disparities in economic wealth.

But this view is vulnerable. Governments, too, are characterized by their coerciveness, but that does not stop lawyers (or scholars) from claiming that individuals can act voluntarily in domestic settings (Geuss, 2001: 153; Williams, 2005: 6). Force or the threat of force does not preclude voluntary action, and sometimes force or the threat of force makes an action more voluntary (Kant, 1991: 103, 125, 231, 232). For instance, European and American military aid affords Ukraine the capacity to fight harder for its sovereignty. Nor has the traditional view stopped political realists from goading states to make different (and, in their view, better) choices (Mearsheimer and Walt, 2007; Morgenthau, 1969).

A more modern perspective is that voluntary action is abundant in world politics. Unless one is being physically forced to do something, such as the interwar French government securing reparations from Germany by occupying the Ruhr, one always has latitude to choose. And because complete physical compulsion and state death are rare, the range of issues over which states exercise sovereignty is vast (Biersteker and Weber, 1996; Yoon, 2024: 211). States may be constrained by external structures, yet still have considerable latitude to exercise agency in world politics (Klotz et al., 2006; Wendt, 1987). From this perspective, weak states can act voluntarily by exploiting a hegemon’s vanity (Jakobsen et al., 2018) or great power impatience (Maris and Manoli, 2022) to improve their leverage.

Yet this view is similarly vulnerable. Most coercion stops short of physically performing actions for others, yet few would describe most coercive encounters as mutually voluntary. It is one thing to assert that weak actors have agency and are not simply playthings of powerful actors, still another to assert that most of what weak—or even strong—actors do is entirely voluntary. It strains credulity to interpret compliance with a wartime demand for unconditional surrender as “voluntary” disarmament, export restrictions between asymmetric economic actors as “voluntary” trade restrictions, or religious conversion on pain of ostracism as “voluntary” beliefs. In short, folk accounts of voluntary action in international politics are widespread but problematic.

We posit that the best starting point for voluntary action is will. Etymologically, “voluntary” derives from the Latin velle or volō, meaning to want, will, or wish. ¹ This fits with common sense and common usage. When we speak of someone who acts voluntarily, they typically do as they please, act willfully, or impose their will on others. When we speak of someone who acts involuntarily, they typically act against their will or bend to another’s will. Voluntary action is frequently juxtaposed with coercion, compulsion, compellence, corruption, constraint, pressure, and suborning. Although these have different connotations, they are all shades of the same phenomenon: getting others to do an action that they would not do without force, threat, deceit, or inducement.

If IR scholars have paid little attention to the concept of voluntary action, they have virtually ignored the concept of will. The JSTOR database records only one IR book or article with “national will,” “will power,” or “willpower” in the title (Lieber, 2012) and only one about the “will of nations” (Arnoldson, 1909), and neither of these works defines the concept of will. Many IR literature works embrace Clausewitz’s (1984: 75) point that international politics is a bargaining process, which revolves around acts designed to “compel our enemy to do our will.” However, nowhere does Clausewitz define or theorize will.

Yet we can deduce what will means through context. Separable from “the national interest,” scholars describe “the national will,” and connect this concept to cognates such as “morale” and “resolve,” often viewing will as a contributor to one’s relative power or battlefield success (Aron, 1981: 49; Castillo, 2014: 18; Morgenthau and Thompson, 1993: 143–154; Wright, 1964: 73). On a grander scale, historians have long debated the role of national will in the rise and fall of empires in the same manner (Barnett, 1972: 20–24; McCloskey, 2006: 215–216; Porch, 2022: 123, 148, 221).

In Kertzer’s (2016: 3) helpful definition, resolve is “akin to the idea of willpower, resolve is not what an actor wants, but rather, the extent to which she maintains this intention despite contrary inclinations or temptations to back down.” This implies that will is both what an actor wants (e.g. bearing “good will,” “ill will,” or a “will to power”) and how much she wants it (e.g. “strength of will,” “force of will,” or an “iron will”). Preferences plus resolve takes us a long way to will, but wills can belong to someone or something else, and acting on such wills is not voluntary, even if energetically pursued. To fit the commonsense meaning of voluntary, will has to be truly or mostly from the actor in question, even if enthusiastically encouraged by another (Elster, 1996: 81–82; Preston, 1984: 959–964). In a nutshell, voluntary action must be an expression of genuine will.

Voluntary action should not be conflated with the agent-structure debate (Carlsnaes, 1992; Dessler, 1989). Broadly speaking, this debate concerns whether structure (defined as social patterns that shape what opportunities are available for agents to choose from) or agency (defined as the ability of individual units to choose for themselves) is more important in explaining outcomes (which might include the shape of structures or agents themselves). Much of the literature concerns how agents and structures cocreate each other and often how agents produce structure through structuration, cyclical interactions between actors and the rules that influence them (Giddens, 1979).

Yet agency is fundamentally different than will. Agency derives from the Latin agentia, meaning “to act, do, drive, impel, or perform,” even if unintentionally or incidentally (Wendt, 1987: 347, 360, 364, 368). Agency is a broader category that explores what basically unitary actors chose, intended or not. But will presumes that actors are not unitary, they sometimes act “against their will” and that only intentional actions count. Will explores what it means for an identity to intend something; agency explores choices that could have been otherwise (Crossley, 2022: 172–173). In short, will is a special subset of agency that is deemed more authentic or intense than ordinary agency. Will thus carries a heavier onus of responsibility than agency does, and it widens the door for creativity. It is hard to act against, or independent from, structure without a theory of voluntarism. Where agency emphasizes choice and structure centers incentives, will is about mentality.

Voluntary action in theory: actors and their antagonists

Will is what an actor wants and how much she wants it. At a minimum, genuine will is a will that reflects more the will of the actor than the wills of others. So far, so good. But difficulties arise when we consider who can will, where will comes from, and what counterfeits genuine expressions of will. We tackle each of these issues in turn.

Willing actors

First, one needs an actor that can will. Of course, there are long-standing arguments that no actor can. In philosophy and neuroscience, leading voices argue that individuals do not consciously will (Carruthers, 2007; Wegner, 2003). In domestic politics, persistent difficulties in aggregating preferences, captured in Condorcet’s voting paradox or Arrow’s impossibility theorem, limit the possibility of identifying a clear and consistent popular will (McKelvey, 1976; Schofield, 1978). IR scholars make similar arguments about collective actors in world politics (Bull, 1995: 44–49). Kelsen, for instance, argues that states can never truly be said to will anything because they are nothing more than “fictions” (quoted in Wheatley, 2023: 166).

Nevertheless, large literatures in the social sciences push back against the notion that individuals and collectivities cannot be said to will or cannot be held responsible for their actions. They grant that individual selves are just as modular and jumbled as corporate actors, and yet wills are discernible in composite bodies (Kahneman, 2013; Schelling, 1984; Sen, 2002; Wilson, 2002: 48). For most people, any actor can be capable of acting voluntarily—people, groups, states, networks, organizations, and so on—provided it is essentially a unitary egoist with minimal managerial functions. We propose a willing actor as a purposive, systemic whole that is different than the sum of its parts. That is, actors need an intensely shared identity, collective decision-making capability, and agreed-upon mechanisms for turning decisions into actions (Erskine, 2014: 119; List and Pettit, 2013: 76–77). Based on these standards, states and families would tend to qualify as actors, while generations and crowds would not.

We must make clear, however, that no unit of analysis is neutral. Focusing on some actors means other actors will be out of focus, and prioritizing some actors over others is an implicit bet on their causal importance and/or moral responsibility. Ultimately, this is a theoretical wager (Jackson, 2011; Jackson and Nexon, 2013) on the location of sovereignty (Hinsley, 1986; Jackson, 2007; Philpott, 2001). Will could be exercised but not expressed because internal forces overpower it, much like “the people” might speak but be overruled by “the elites” or vice versa. As we will see, one of the major differences between theories is where they situate will: in individuals, groups, classes, coalitions, parties, states, or international society. When we speak of voluntary action in world politics, we usually mean the capacity of nation-states to express their wills. But this perspective might exaggerate the importance of states, or more sinisterly, provide cover for a leader to impose their personal will while claiming to be mere executors of the national will. Seeing actions from multiple perspectives helps guard against the blind spots of each.

Internal factors

Once one determines who can will, then one must determine what produces will. At a minimum, this is a two-step process: which faculties contribute to will and the process by which will emerges from the faculties. In the first step, it is useful to distinguish between three semiautonomous sensibilities or faculties: desiring, reasoning, and moralizing (Recht, 2025: 53). Will has been associated with all three, but need not be. In everyday and legal language, we often associate will with the desiring faculty (e.g. “do as you will” or “fire at will”) and distinguish it from reasoning (e.g. “the willing suspension of disbelief” or “willful ignorance”) and moralizing (e.g. “willful misconduct” or “the spirit is willing but the flesh is weak”). For thinkers like Plato, Rousseau, and Kant, this relationship is flipped: the desiring faculty makes an action less voluntary, and the reasoning faculty makes it more voluntary (Allen, 2023: 205; Glaser, 2010: 30). For thinkers like Frankfurt, first-order (desiring) and second-order (reasoning) wills must align to be free (see Ekstrom, 2001: 9–13). Will, in this context, requires a theory of rationality and implies an actor has enough information, intelligence, and calm to reason rightly (Druckman and Lupia, 2006).

Second, this raises the issue of how will emerges. If will is singular, this is not an issue. The will wants what the will wants, then acts on it. This is Hobbes’ (1991: 170 [V.8]) approach. He presumes all actors are unproblematically unitary, and almost any kind of behavior is tantamount to voluntary consent. While logically consistent, this approach leads to such universal voluntarism that most people find it intuitively appalling and intellectually implausible.

Perhaps for this reason, most thinkers believe the will is an alloy of faculties, and that requires a decision rule. Will can emerge in a hierarchical process, like a monarchy, or a nonhierarchical process, like a democracy. Plato explicitly uses regal analogies for the will, and Kant characterizes will as something akin to the rule of law. In medieval political thought, the prince is said to possess voluntas ratione regulate or “will directed by reason” (Kantorowicz, 2016: 106–107), which later grew into “reasons of state.” Democratic theory offers a more egalitarian alternative. Manin (1997: 6) finds that four principles have been “invariably observed in representative regimes”: (1) those who govern are appointed by election at regular intervals, (2) the decision-making of those who govern retains a degree of independence from the wishes of the electorate, (3) the governed may give their political opinions without being subject to control by the governed, and (4) public decisions undergo trial by debate. Rough adherence to this formula elicits a general will.

There are many ways these principles can be applied. For example, we might draw distinctions between states based on their domestic regimes, arguing that some better reflect the national will than others. Wright (1964: 159, 196) adopts a version of this approach when he contends that despots have “a self-discovered capacity to mold the national will” and that “governments will occasionally prostitute national opinion to illegal ambitions.” ² Similarly, there might be certain situations or bureaucratic arrangements that undermine the national will. Anderson (2024: 5–8, 38–39) argues that states often engage in compulsive “inadvertent expansion” because national policy is hijacked by parochial agents along distant frontiers.

In all these cases, an actor’s ability to pursue its genuine will is tied to proper procedures, and one must specify what separates the proper from the improper. Interpretations of what rules wills should play by are probably as diverse as governmental forms. This diversity is less important than the point that one needs a standard for processes to judge whether genuine will is being expressed or it is being perverted internally. The litmus test of voluntarism is sufficient representation, whether an action can be tenably said to represent the preferences and resolve of a political body. To declare an action voluntary, one must take a stand on how genuine will emerges inside an actor.

External factors

Having set the actors and the internal factors, we now turn to the environmental conditions that affect voluntarism. Given how protracted the agent-structure debate is, it should be plain that lines are blurry between internal and external factors, but most observers agree that some features of social reality are more extrinsic to actors than others, and these more extrinsic features can be called external factors (Crossley, 2022: 177–178; cf. Giddens, 1979). We contend that external factors differ chiefly by degree (how much pressure an actor is under), valence (whether pressures are more rewards or punishments), and instruments (whether the tools of pressure are more military or nonmilitary).

With respect to degree, one must draw some line beyond which action becomes involuntary. As seen above, Hobbes has a very low bar for voluntarism. Pettit (2007: 175) counters with a high bar: “There must have been a grave matter . . . full knowledge of the guilt, and full consent of the will.” Yet these are extreme positions. We believe that will is seldom an all-or-nothing affair, and the bulk of the debate happens in the middle. Furthermore, one must be wary of drawing any line in degrees of will; adjudication is not straightforward. All political actors have a latent capacity to impose potential sanctions (used in a neutral sense: positive or negative in consequence, economic or noneconomic in nature) on others (Clausewitz, 1984: 370). But just because actors can impose their wills does not mean that this is always relevant. Compulsion is absent when actors cannot significantly sanction each other or when they independently possess identical preferences. Nor is compulsion equivalent to power asymmetries. Capabilities need to be extracted, converted into tools of influence, and credibly threatened or wielded on actors, who can respond in offsetting ways and in settings that may properly institutionalize power. Because of this, weaker actors can be in a position to quash stronger actors’ desires or attempts to remove otherwise viable options within a given scope or domain. As always, influence is context-sensitive and actor-specific (Baldwin, 2016: 3; Barnett and Duvall, 2005: 42) and may exist in a dichotomy or on a spectrum.

Valence means whether a sanction or potential sanction is positive or negative. Positive sanctions are incentives, benefits, or socialization, while negative sanctions are disincentives, costs, or negative feedback. Note that external factors can constrain as well as enable behavior; they can be either anarchic (markets or multipolar systems) or hierarchic (empires or hegemonies). Opinions differ on whether only negative sanctions can deprive an action of its voluntariness or whether any sort of sanction has the potential to do so (as the journalistic credo has it: “without fear or favor”). After all, “golden handcuffs” are still handcuffs and can be as confining as their nonmetaphorical cousins. But in general, the literature tends to associate negative sanctions with a lack of voluntarism. Perhaps this tendency reflects prospect theory. Humans hate acting in the realm of losses and tend to run larger risks in such circumstances (Kanner, 2004). Perhaps it is a prejudice.

Instruments refer to whether a sanction employs primarily military or nonmilitary means. Military sanctions take the form of using or threatening physical violence; nonmilitary sanctions do not. Although the two interact in complex ways, they are conceptually distinct. Military sanctions are the least controversial form of compulsion. For example, the US government forcibly relocating Native Americans along the Trail of Tears is unanimously seen as involuntary. Nonmilitary instruments can be broken down into economic and diplomatic instruments. Generally, external compulsions involving military instruments appear more clearly coercive and seem more likely to render actions involuntary, but military coercion need not be more common or consequential than its economic or diplomatic counterparts. One study calculated that aerial bombing convinced targets to change their behavior in 36 percent of cases (Horowitz and Reiter, 2001: 160), while another found that economic sanctions had the same effect 34 percent of the time (Hufbauer et al., 2007: 159). Trench warfare cost approximately 2 million German soldiers their lives during the First World War, yet the British-led starvation blockade of the Central Powers contributed to an estimated 1 million excess civilian deaths across Germany and Austria-Hungary (Downes, 2006: 178).

Economic sanctions take the form of monetary, financial, or trade instruments that can be readily converted to market prices. Although they have a more benign reputation than military instruments, they may also nullify voluntarism through threats and bribes. This is obviously the case when a boycott is as effective as a blockade, but may be equally true when actors use carrots instead of sticks, or indifference rather than engagement. Between destitution and prostitution, some choose the former and others the latter, but both have grounds to complain that their choices were involuntary. Consider tougher cases. The Trump Administration recently offered US$20 billion in economic aid (with potentially US$20 billion more in private financing) to Argentina, conditional on Javier Milei’s party winning an upcoming election. That offer gave Milei’s party a substantial boost in the polls and likely fueled its victory. Although many objected, the results were uncontested. If China offered comparable sums to Taiwan conditional on a favorable vote for reunification, and a majority of Taiwanese voters elected the Chinese Unification Promotion Party, would most people perceive the results as voluntary? One person’s illicit bribe is another’s just reward.

Diplomatic sanctions take the form of informational influence attempts and may invalidate wills through persuasion, education, or socialization. If stronger actors have disproportionate influence over the rules of the game and wider stores of relevant private information, they may be able to prevent, filter, or create information flows. Recall the American campaign to convince states that Saddam Hussein had weapons of mass destruction in the run-up to the Iraq War. Shortly after, the United States released shifting lists of states that supported the Iraq War to bolster its purported legitimacy. While the name of this group, the “Coalition of the Willing,” implied that members had joined voluntarily, critics dismissed them as a “Coalition of the Bribed, Bullied, and Blind” (Newnham, 2008: 183), and subsequent research showed that the United States had exploited its diplomatic and economic leverage to pressure states to declare their support (Henke, 2017: 421). It may also be sufficiently harmful to form preferences in low-information environments (Kuran, 1995). Relatedly, the quantity of information may be high, but the quality of information may be low (Snyder and Ballentine, 1996). This could result in endogenous preferences, where actors grow to like what they can get and disdain what they cannot, which may or may not express genuine will (Elster, 2016). Low-quality information could also facilitate false consciousness, where weak actors want what the strong want them to want because the strong conditioned them to (Lukes, 1974).

Whether diplomacy alone can eliminate voluntarism is an open question, and the boundary between persuasion and propaganda can be ambiguous. First, we should be careful not to mistake dramatic change for voluntary choice. If actors possess conservative preferences, then voluntarism is perfectly compatible with stability. Schroeder (2004) describes the Concert of Europe not as an impersonal balance of power, but rather as a cooperative social system in which the great powers colluded to fend off liberal revolutions. Conversely, actors might be pressed into wars, revolutions, or other wild gambits by forces outside their control. Gilpin (1983: 54, 194) provides a dynamic account of hegemonic consolidation, decline, and war, but it is far from clear that his protagonists had much choice in the matter.

Second, we should be careful not to assume that voluntary actions are necessarily moral or rational or that the world would be better off if constraints on voluntary action were lifted. Democratic states may effectively represent collective will, but sometimes democratic publics want bad things (Caverley, 2014; Downes, 2006). Minimizing internal compulsions for some actors may increase the external compulsions imposed on others, and unshackling “good” actors need not lead to “good” outcomes. So, too, with nondemocratic leaders, who hide behind the rhetoric of sovereignty while subverting the will of their populations.

Third, the prevalence of an action is a poor indication of its voluntariness. Imperial practices are obvious examples. Observing self-Romanization in the 4th century, Heather (2023: 93–94) states that “the process was fundamentally not a voluntary one”:

The only way to live long and prosper was to buy a Roman entry ticket, whatever the cost. For all the apparent agency being exercised at the local level . . . pressures to conform were so strong that that most local agents were only ever going to make one choice.

Lawrence (2013: 163–164) notes the sudden drop in African nationalism among African students studying in Paris in the 1950s when they returned home and explains that “imperial authority was indeed important for restricting political opposition and foreclosing active nationalist mobilization in the territories.”

It is hard to identify the point at which will becomes counterfeit because of threats, bribes, and inordinate social pressures, but it is not impossible. In our view, a modified version of Mill’s (1979: 8–10) decided preference criterion provides a foundation. When actors with equal experience of the options tend to select one over others in a tolerant setting, the rebuttable presumption is that the selection is voluntary. That may sound like a prohibitive condition in international politics, but it is observable even in harsh environments. Power exhibits magnetic properties, and power bases are typically plural, so between their fields is a neutral or null point, where voluntariness is most visible. There are always frontiers, peripheries, interstices, and liminal zones in politics, and actors seldom fail to have decided preferences within them. For instance, De Bruin et al. (2025) find that citizens in weakly governed areas of Colombia view armed groups as more legitimate when they adopt local norms, involve local leaders, provide services, and curtail violence. Getachew (2019) finds that, for some time, anticolonial thinkers and politicians in neutral spaces articulated visions of international order that bucked the pressures of nationalism and economic hierarchies.

To summarize, voluntary action is an expression of genuine will. Any conception of voluntary action must take a position on three questions: which actors are capable of willing, where does will come from, and what are the minimum conditions for genuine will? As a preliminary consensus on minimum conditions, we ventured coarse lines that the degree of pressure be low enough that the action expresses more of the actor’s will than another’s, that the valence of pressures be more rewards than punishments, and the instruments of pressure be more nonmilitary than military. Although these claims likely match many people’s intuitions, they will not match some others’. Reasonable people disagree, or at least their wills will.

Voluntary action in action: mapping major disagreements

We combine the internal and external factors of voluntary action to highlight where IR scholars diverge. Table 1 considers four different ideal-typical worlds depending on the extent to which actors, in most cases states, are confronted by strong external or internal compulsions. We place IR paradigms inside the box that corresponds most to their views, which are driven by philosophic-scientific wagers on how the world works (Jackson, 2011; Jackson and Nexon, 2013). Our sense is that the broadest support for voluntarism is found in the bottom-right box and the narrowest support is in the top-left box. Of course, writers within and across these boxes disagree on many issues, not just voluntarism. But we contend that different theoretical stances inform why groups of theorists tend to view voluntarism in world politics differently, though the strange bedfellows in each quadrant are also intriguing.

Table 1.

A typology of interacting compulsions.

	External compulsions
		Strong	Weak
Internal compulsions	Strong	World of chains: Marxism, Classical realism, Critical constructivism	World of dysfunctions: Democratic liberalism, Economic liberalism, Bureaucratic politics
	Weak	World of dangers: Structural realism, Liberal institutionalism	World of possibilities: Liberal constructivism, Contentious politics

A world of chains: the upper-left box

For some theorists, internal and external compulsions are both relatively strong, and the prospects for voluntarism are significantly circumscribed. Given their encompassing conceptions of capitalism, many Marxists fit into this quadrant. In terms of actors, economic classes are the principal unit of analysis, and states are often fronts for upper-class interests. In their ceaseless search for profits, capitalists, and the government officials that represent their interests, act like drug addicts, desiring short-term gains over long-term goods (Cox, 1987; Marx, 1990). This represents the will of the capitalist class, but not the proletariat’s will, which produces exploitation, meaning “unequal exchange of labor” or “taking unfair advantage of weakness” (Ferguson and Zwolinski, 2024: 2–4). Internally, Marx (1972: 595, 500) tends to see capitalism as both coercive (“men make their history, but they do not make it as they please”) and enslaving (“the proletarians have nothing to lose but their chains”). Others add that capital has the power to warp people’s preferences so much that they suffer from false consciousness and want what hegemons want them to want (Cox, 1983; Lukes, 1990; Parmar, 2019). Because the boundaries of classes are unclear and they do not coordinate as a body, it is doubtful that classes qualify as actors that can will. Vanguards, cabals, and parties, however, could qualify. Externally, capitalist interests compete for markets under the guise of imperialism. Eventually, this internecine competition will lead to catastrophic war and revolution (Lenin, 1989). Every form of compulsion appears energetically in Marxist theory.

With their desirous view of human nature, classical realists, as well as their neoclassical counterparts, also inhabit this quadrant (Kirshner, 2022). With respect to actors, these realists tend to be permissive about who can be a willing actor, equating whichever individual or group has de facto authority in a state with a willing actor. Internally, leaders are prone to make various decision-making errors, whether due to misperceptions, cognitive limitations, genetic predispositions, institutional flaws, or personality quirks. The result is a series of self-defeating tragedies. Status-obsessed leaders are lured into wasteful overseas expansion, which culminates in imperial overstretch and national decline. Parochial military bureaucracies become ensnared by “cults of the offensive,” which short-circuit diplomacy and result in ruinous wars (Snyder, 1984). Insecure decision-makers in late-modernizing rising powers become obsessed with various “myths of empire,” which encourage them to launch bids for hegemony (Snyder, 1991). Externally, states tend to be the willing actors, and they reside in a competitive world that ties their hands. There are highly hedged hopes that actors could will prudently within these internal and external bounds. Aron (1985: 280–281) tries to thread this needle,

To be prudent is to act in accordance with the particular situation and concrete data . . . it is to prefer the limitation of violence to the punishment of the presumably guilty party or to a so-called absolute justice; it is to establish concrete accessible objectives conforming to the secular law of international relations and not to limitless and perhaps meaningless objectives.

E.H. Carr, a classical realist and a Marxist, embodies their overlap.

Constructivists working in critical traditions reach similar conclusions about the paucity of voluntarism, albeit for different reasons. The actors they prioritize are individuals or groups who are the ultimate bearers of moral worth. Feminist scholars have long emphasized how gender hierarchies reproduce themselves both within and across political communities. Internally, restrictions on women’s rights and restricting gender roles result in nonrepresentative male-dominated institutions and self-defeating decisions because not all perspectives are considered. Externally, these hierarchies are reinforced by international institutions that overlook gender-based violence or ignore women’s roles in global development initiatives (Shepherd and Kirby, 2024; Sjoberg, 2013). Postcolonial scholars make similar points about the pernicious power of racial hierarchies (Sabaratnam, 2020). Historically, these hierarchies have their origins in one of the most glaring examples of external compulsion: the expansion of European empires and the racial biases they brought, refined, and left in their wake. Today, these racial hierarchies help create nonrepresentativeness through dynamics of domestic exclusion and externally via the legacies of empires and imperial modes of thinking. The prospects for voluntary action seem small given how deeply embedded these pernicious ideas are in political institutions.

Strikingly, none of these perspectives entirely rule out voluntary action. Indeed, all grant it crucial importance. For Marxists, the arrival of communism may be a historical inevitability, but choices meaningfully affect its timing or spread. Marx and Engels argued that the recognition of necessity would allow people to intervene in history to speed it up (Bourdieu, 1989: 4). Lenin stressed that vanguard parties could act as catalysts to spread communism around the world (McAdams, 2017). Classical realists catalog repeated political failures, yet never give up on the hope that prudence can prevail (Jervis, 1976: 409–424; McQueen, 2018: 14). Feminist scholars such as Enloe (2014) emphasize not only the restrictive power of gender hierarchies but also the transformative power of feminist social movements. Colonialism left deep legacies, but scholars such as Acharya (2016) note how postcolonial states can engage in “idea-shifts,” challenging and refashioning ideas about human rights, sovereignty, and development. Critical theorists like Thame (2025) illuminate pathways to emancipatory action against violent realignments of global order. Even in a world of chains, theorists refuse to accept that bonds cannot be broken.

A world of dysfunctions: the upper-right box

For other theorists, internal factors are stronger and more pernicious than external ones. Voluntarism remains commonplace, but internal dysfunctions occasionally disrupt the coherent pursuit of collective wills. Of the main paradigms, various versions of liberalism populate this quadrant and privilege the reasoning faculties. For them, coalitions are the main willing actors in international politics, preferably popular coalitions, which come to power through free and fair elections. Internally, for democratic liberals, the biggest danger is that the norms or procedures that underpin democratic institutions will erode and that leaders will become less accountable and more domineering (Greene and Robertson, 2019). Externally, there are some who argue that compulsion is required to spread democracy, open markets, and enforce the dictates of international institutions (Doyle and Recchia, 2011: 1438; Hurd, 2022: 274–275). The hope is that fair processes and clear rules will make such acts legitimate, though there remains something paradoxical about “forcing others to be free” (Applbaum, 2007; Cherniss, 2021: 7).

For commercial liberals, the core concern internally is that uncompetitive sectors or entrenched interest groups will demand and receive protectionist policies, thus depriving society of the benefits of free trade. Haas (1964: 8–10, 143–145, 195–206, 212–221) views bottom-up action in pluralistic societies as the future of politics and voluntary groups as especially legitimate. Martin (1992: 26, 39, 43) posits that international institutions can facilitate cooperation by insuring against “irrational or involuntary defection,” suggesting that sometimes states are better off if they accept external compulsions to guard against internal ones. For Keohane (1984: 72–73), voluntarism has to do with the ratio of opportunity costs: the more one party needs an agreement, the less voluntary it is. This suggests that actors should make efforts to distribute the benefits (and costs) of cooperation equitably, to guard against asymmetric dependence. Externally, commercial liberals tend to see the international environment as more permissive than either realists or Marxists because politics tends to be local, there is no sharp distinction between high and low politics, multiple channels of influence dilute the potential for compulsion, and “military force is largely irrelevant” (Keohane and Nye, 1989: 255).

Scholars of bureaucratic politics and principal-agent problems come out in roughly the same place as liberals, which is odd since the groups do not normally agree. The subnational groups they zoom in on are not antidemocratic movements but bureaucracies and organizations whose faculties incline toward desiring and reasoning within narrow confines, rarely moralizing at all. While varied, this literature highlights the inherent displacement of wills that can take place when governance tasks are delegated from the people to their leaders or from leaders to their subordinates. Because agents possess different preferences than their principals and because monitoring and sanctioning agents can be costly, agents have latitude to reinterpret their mandates, pursue parochial interests, and impede rapid change. The result is that collective will is only imperfectly reflected in collective actions, and at the extremes, highly concentrated and highly motivated interests can commandeer politics and snarl communication (Casler and Jost, 2025; Snyder, 1991). The inference is that actions may be rational in a bureaucratic sense, but in no other. The development of robust norms and professional identities may shield or advance the bureaucracy’s collective will. One of the purposes of norms of civil-military relations in democratic societies, such as the principles of apoliticism and deference to civilians, is to prevent military officers from equating their interests with the national interest. Externally, force and fraud are often off the table because most of the organizations studied operate in legal contexts. But this sometimes deceives; what seems to be voluntary deference or technocracy is deeply influenced by potential violence or structural intimidation. Pratt (2018: 561) finds that deference to international organizations (IOs) by other IOs “is shaped both by efficiency concerns and power politics.” And Jost (2024) finds that what is ostensibly rational bureaucracy is often profoundly influenced by fears of coups or conquest.

Regardless of their specific theoretical commitments, the upshot of theories that populate this box is that voluntarism is primarily a question of institutional design, and the hindrances to it are primarily internal. States are better off if their institutions are democratic, the preferences of their elected politicians closely match those of the people, and norms of civic obligation and professionalism restrain the ambitions of those in positions of power. Politics is messy, in the sense that aggregating preferences requires bargaining and the bargains that result are imperfect. But with the right processes in place, collective will can emerge and voluntary action is probable.

A world of dangers: the lower-left box

For still other theorists, external factors are of overriding importance. Structural or neorealists fit most clearly into this quadrant. As with other realists, structural realists tend to be inclusive about what actors have volition, but they put domestic politics in a black box and prefer not to look inside (Waltz, 1979: 88, 96). Internally, states are the key actors capable of willing, and the desiring faculty tends to prevail over morality (Layne, 2006: 7–9; Posen, 2014). Writings on the national interest recognize that states have conflicting domestic interests, but assert that government policy usually advances a stable and representative set of collective preferences, most notably survival, security, or autonomy (Krasner, 1978). Beyond that, states are free to will whatever they like. That may morph into nonrepresentative or self-defeating preferences, of course, but structural realists assume that those who do so indulge fanciful desires will be chastised, being either selected out by competitive pressures or socialized into willing sensibly (Glaser, 2010: 20, 35; Waltz, 1979: 126).

Thus, for structural realists, will is primarily disciplined by external factors. The real risk to voluntarism is the accumulation of disproportionate military power by other states, which can be wielded as a negative sanction. For Waltz (1979: 142, 194), “power provides the means of maintaining one’s autonomy in the face of force that others wield,” though power was not restricted to military power (1979: 189–203). “Pursuing a balance-of-power policy is still a matter of choice,” he (1959: 205) says, “but the alternatives are those of probable suicide on the one hand and the active playing of the power-politics game on the other.” Which is to say: little real choice and not much voluntarism. Perplexingly, while structural realists view external compulsions as exigent, they also see them as indeterminant (Gilpin, 1983: 226; LaRoche and Pratt, 2017: 161–162). Consequently, classical and neorealists part company on voluntarism.

Many liberal institutionalists also cohabit this quadrant. Where realists focus on the state, liberals lean more on rational subnational actors, but both accept that many sorts of actors have volition. While realists focus on external compulsions caused by military force and negative sanctions, liberal institutionalists focus on how failures in collective action can impose collective harms on states and trans- or subnational groups. States need to reason together on a variety of fronts: to ensure a functioning global currency system, to develop a global system to regulate trade, to reduce greenhouse gas emissions, to slow the spread of nuclear weapons, and so on. The challenge is largely an external one: how can actors best engineer a system in which these collective wills are achieved? Preserving voluntary action thus becomes a question of regional or global governance (Buchanan and Keohane, 2011; Gruber, 2000: 4–8). By creating mechanisms to acquire and share information, to monitor behavior and facilitate reciprocity, to manage differences and distribute gains, and to set standards and articulate collective norms, states can manage and minimize compulsions through collective action. In some cases, actors trade off one set of external compulsions for another. To avoid the harms caused by protectionist trade wars or recurring financial crises, states accept a modest reduction in their sovereignty by delegating authority to the World Trade Organization or International Monetary Fund. To secure the benefits of a stable global economy, hegemons voluntarily provide public goods in exchange for voluntary deference.

A world of possibilities: the lower-right box

For a final set of theorists, neither internal nor external obstacles are high. Voluntarism is a frequent modus operandi; political communities largely shape their political worlds. Liberal constructivists fit best in this quadrant, although they are admittedly a diverse club. For them, the key actors in politics are individuals or communities, and moral faculties figure prominently. Internally, voluntarism appears ubiquitous in constructivist thinking. Krebs (2015: 42) finds there are many domestic strictures on how rhetoric can promote foreign policies, but few situations where rhetoric coerces. Wendt (1999: 127) acknowledges that cultures can narrow options, but the future is fluid and can be remade by stepping beyond culture. “Most of the time states do not find themselves in hotel fires,” he (1999: 237) concludes, “in which case a variety of beliefs about how to meet security needs may be compatible with the national interest.” Yet cultures can also help actors achieve better options, and for various reasons (cultures can be contradictory, agents are never perfectly socialized, and exogenous shocks and creativity can loosen a culture’s hold), actors are self-organizing entities with powerful volition (Reus and Smit, 2018: 7; Wendt, 1999: 167, 188). Externally, not only are compulsions comparatively weak for liberal constructivists, but they are getting weaker. Philpott (2001: 4, 253–254) tells the development of sovereignty as an unfolding logic of liberation across history. In contrast with the state, security communities promote peace not by maintaining a “monopoly of force” but through shared collective identities: “the system is voluntary in a way that a state is not.” (Wendt, 2003: 522). The takeaway is Wendt’s famous formulation that “anarchy is what states make of it.”

There are several pathways to express genuine will that alter outcomes in a world of possibilities. Legro (2005: 14) describes how ideational entrepreneurs can exploit moments of crisis and normative uncertainty to inject new ideas into world politics. Keck and Sikkink (1998: 8–18; cf. Terman, 2023) emphasize the transformative power of transnational advocacy networks that can name and shame governments to promote change. Klotz (1995: 9–12) emphasizes the fusion between social movements and political interest groups, which can seize on new normative commitments to shift national policies. While more varied in terms of their ontological commitments, there are similarities between liberal constructivists and the contentious politics literature, which stresses the ways groups can exploit political opportunities to organize politically for social change (Tilly, 2004: 136–143). These movements can learn from one another, cultivate shared memberships, and innovate new repertoires of resistance. Chenoweth and Marks (2022) tell a story of nonviolent movements organizing collectively and outflanking the coercive power of the state. Both liberal constructivist and contentious politics accounts reject the distinction between internal and external constraints, seeing the politics of populist authoritarianism and democratic activism alike as being global, national, and local at the same time. However, actors promote change in practice; a world of middling compulsions provides major opportunities for voluntary action.

Voluntary action and its applications: theory meets practice

How can a better understanding of voluntary action help us navigate theoretical and political disputes? First, having an explicit guide to voluntarism holds scholars accountable, saves them from talking past each other, and opens up space for evidence to change minds. Consider, for example, debates about the US role in the so-called liberal international order. Liberal authors, such as Ikenberry (2011: 10), argue that the establishment of the United States as a liberal hegemon was the product of “the voluntary granting of that status by other states.” Realist critics, such as Glaser (2019: 69), respond that other states “may have little choice but to accept highly asymmetric terms.” The guide we offered earlier helps locate where disagreements are and assess their merits. If scholars make explicit what they think voluntary action is and when they think it is possible, then they can come to see why others think differently. In this case, neither side disputes that states are the chief willing actors, that hegemons could coerce others involuntarily, and that other states could acquiesce to hegemonic demands voluntarily, primarily through rational faculties. The debate gravitates around external conditions: whether hegemonic pressures are sufficient to suborn other states’ wills, which has largely to do with the valence of potential sanctions and which instruments are decisive in creating outcomes of interest. While not solving the debate, our approach advances it by making it more tractable empirically. It may also make claims more nuanced. The conditions for voluntary action under a hegemon may be sharply different from one region to another (cp. LaFeber, 1993; Lundestad, 2003).

Second, a clear account of voluntary action can help inform how actors draw the line between legitimate and illegitimate behavior. All else equal, actions seen as voluntary tend to be perceived as more legitimate. The genuine will of the people likewise tends to be afforded more legitimacy than the counterfeit will of the dictator. Yet we caution that not all willful behavior is inherently legitimate, and that actors often manipulate claims of “voluntarism” for political purposes. Powerful actors often mask naked exercises of power by claiming to respect the voluntary choices of the weak, as in Russia’s use of a plebiscite to justify its annexation of Crimea. Along the same lines, actors often justify illegitimate policies on the basis that they had “no other choice,” as the Bush administration did around its use of torture. Without a framework that distinguishes between genuine and counterfeit will, we lack well-reasoned objections to lupine claims.

Third, without a reasoned account of voluntary action, actors in world politics cannot be held responsible for their actions. To quote Pettit (2007: 172), “No incorporated agency without incorporated responsibility.” If an actor wills an act of commission or omission, it is on the normative and political hook for it. This is why realists fault decision-makers for embracing gratuitously bad policies. It is why critics fault realists for absolving decision-makers of responsibility by likening their decisions to go to war to sleepwalking. It is also the moral foundation for the European Council’s (2025: 2) demand that Russia “to show real political will to end the war [in Ukraine].” Assigning responsibility need not only be about assigning blame, but also praise. Kaufmann and Pape (1999) credit Britain for voluntarily fighting a 60-year campaign against the Atlantic slave trade. Defeat may be an orphan and victory may have a thousand fathers, but only with a clear understanding of voluntarism and responsibility can we accurately assign parentage.

Fourth, a well-specified theory of voluntary action can help us understand whether moral responsibility applies to coporate actors beyond states, such as “coalitions of the willing,” IOs, and the “international community” (Bukovansky et al., 2012: 16). As Erskine (2014: 118) remarks: “judgments of moral responsibility must be directed toward entities capable of responding to ethical imperatives.” She argues convincingly that coalitions of the willing are capable of responding to ethical imperatives. Provided coalitions have at least minimal managerial functions and act in a concerted fashion like unitary egoists, they are entities no morally different than political parties or states. This has far-reaching implications. In the post-Cold War world, coalitions of the willing have earned mixed marks, invading Iraq, campaigning against landmines, intervening in Kosovo, East Timor, and Libya, and aiding the defense of Ukraine. Yet they appear to have a valuable role in world politics as failed states continue to struggle and leading states increasingly show indifference to human rights.

In extreme cases, responsibility may flow up to the international community. Sienknecht and Vetterlein (2023: 18) contend “If a state is unwilling or unable to fulfil this duty, the responsibility shifts to the international community.” In this context, international law may be a yardstick to identify genuine versus corrupted will. There is a growing view that states can be bound to rules they did not consent to, like antigenocide norms, because it represents the will of the international community (Cronin, 2023: 3). Hermansen et al. (2025) find that the European Court of Justice highlights its independence and legitimacy by favoring the weak. Carnegie and Clark (2026) make the case that IOs are willing actors that shape global governance. Whether these corporate actors are unlocking opportunities for state voluntarism or shackling states in legal frameworks against their will requires us to know what constitutes genuine will in the first place.

Conclusion: the voluntary world as will and representation

All theories of IR believe that voluntary action exists and ascribe to it a central role in world politics. They coalesce around the idea that voluntary action represents an expression of genuine will. But they have key disagreements, and those disagreements are largely driven by their different assumptions about who the chief actors are in international politics, how internal faculties interact to produce the will in those actors, and what internal and external factors interfere with voluntary action. Naturally, the degree of will is a consideration (more from the actor versus more from its antagonists), but so is the valence of the potential sanctions (positive versus negative) and the character of the relevant instruments (military vs nonmilitary).

Interacting internal factors to voluntarism with external factors to voluntarism yields four clusters of paradigms. Some view world politics as a world of chains in which internal and external compulsions conspire to undermine voluntary action in all but extraordinary circumstances. Others view world politics as a world of possibilities in which compulsions are few and the potential for change is ever-present. Still others view things somewhere in the middle, pointing either to internal dysfunctions or external dangers as the primary impediments to genuine will.

Given these possibilities, it might be tempting to treat the prospects for voluntarism as a kind of pendulum that swings back and forth between “settled” moments where voluntarism is scarce and “unsettled” moments where it is abundant. That temptation should be resisted. The differences between rival theories are ontological and normative as much as causal. Our conceptions of voluntary action set a baseline for our expectations, which means there is no neutral place to stand. We are responsible for the theories we will.