What is Wrong with Mearsheimer’s Offensive Realism?

1 Introduction

John J. Mearsheimer’s offensive realism, developed most systematically and elaborately in his magnum opus The Tragedy of Great Power Politics (2001), is doubtlessly one of the most impactful International Relations theories out there.^[1] Not only does it attract tremendous scholarly attention but it also inspires much of the contemporary discourse on international politics (Toft 2005, 381–2), in particular with respect to the Sino-American rivalry (Mearsheimer 2001, chap.10; 2006), the Middle East conflict (Mearsheimer and Walt 2007), and the Russo-Ukrainian war (Mearsheimer 2014; 2022). The purpose of the present article is to examine the merits of offensive realism (hereinafter: OR). Stated more precisely, the following poses the question whether Mearsheimer’s theory withstands the double test of logical validity and substantive soundness. The verdict reads that if it fulfils one condition, then it fails to meet the other. More specifically, three of the notorious five ‘bedrock assumptions’ of OR – the uncertainty of international politics, the primacy of survival, and the rationality of statesmen – will be challenged as either far-fetched or insufficient to support the theory’s claims. It will be shown that when taken as they stand in The Tragedy, they do not generate Mearsheimer’s world of relentless hegemony-seeking. And when interpreted assiduously through the prism of his broader analyses, what they actually entail is a surprisingly utopian and idealistic theory resting on the implausible notion of absolute national security.

In the literature, there already exist quite a few instructive criticisms of OR. Some of them come from within the broad realist camp (Kirshner 2012, 2015, 2019, 2022; Rathbun 2008; Snyder 2002; Tang 2008, 2010). Others are external critiques leveled by advocates of non-realist schools in IR (Donaldson 2024; Donnelly 2000; Hanania 2022, 26–35; Pashahkhanlou 2013; Specter 2022; Steinsson 2014; Trantidis 2023). Some of the critics point to the inadequacy of structural realism in general, arguing that a cogent IR theory must take account of a variety of unit-level (domestic) factors at play in international politics (Donnelly 2000; Hanania 2022; Kirshner 2012; Pashahkhanlou 2013; Trantidis 2023). Many pinpoint logical flaws in the derivation of conclusions from OR’s assumptions (Kirshner 2012; Pashahkhanlou 2013; Steinsson 2014; Toft 2005). The validity of Mearsheimer’s use of empirical data in corroborating his predictions has also been questioned (Kirshner 2012; Pashahkhanlou 2013; Toft 2005). Beyond that, it is sometimes objected that OR, and structural realism more broadly, build on misconstrued analogies to, or premises divorced from, economic theory (Hanania 2022, 30–5; Kirshner 2015; Snyder 2002). Recently, Mearsheimer’s stand on the sources of Russia’s invasion of Ukraine has also sparked heated controversy (Donaldson 2024; Driedger 2023; Edinger 2022; Person and McFaul 2022; Specter 2022; Ross Smith and Dawson 2022).

In some measure, this article brings together all these lines of critique. Although offensive realism is the main target, the paper also touches on the flaws of the neorealist school as such, especially its reliance on the unitary actor model (or the exclusion of unit-level variables from the scope of IR theory). The main thrust of the criticism lies in economic analysis. From Kenneth Waltz’s seminal Theory of International Politics (1979) on, neorealism (or structural realism) has grounded its claim to scientificity partly in ostensible parallels to economics (Kirshner 2015). According to Waltz (1979, 72), just as theories of the market economy boil down to invisible hand-based structural explanations and are kept in separation from theories of the firm, so IR theory must be centered around the anarchic structure of international politics while leaving unit-level variables on the cutting-room floor.^[2] IR theory as such has been profoundly influenced by the dismal science outside the realist camp as well (Brown 2018, 553–4; Walt 1999). The charge of not taking economics seriously enough must therefore be a consequential one.

As the remainder of the article will show, OR is at odds with a number of well-established economic theorems. Some of them belong in the very core of modern economic science (e.g., the uncertainty and scarcity theorems or the law of diminishing marginal utility). Others are contributions of the Austrian and public choice schools (on the concept of modern political economy and the importance of the Austrian school and public choice in laying the groundwork for it, see Boettke 1998; Boettke and Lopez 2002; Machaj 2014; Pennington 2011). In particular, OR runs afoul of what Mark Pennington (2011) terms ‘robust’ political economy: an economic theory taking full account of the phenomenon of government failure and its ramifications for public policy. Still more specifically OR turns out to gloss over the deficiencies of governments in terms of rationality, motivation, and incentives (for example, Barnett 2014; Pennington 2011).^[3] A cutting-edge contribution to the analysis of realist IR theory from this perspective has been made by Richard Hanania (2022). This author, however, focused solely on the criticism of the unitary actor model, thereby targeting the neorealist school as a whole. While drawing on Hanania’s insights, the burden of this essay is the discussion of OR, with all its shortcomings that go far beyond the corollaries of the unitary actor model. Other realist schools, in turn, are at times positively contrasted with its pitfalls. As will be seen, of all major varieties of realism, it is OR that is the most susceptible to attacks from the standpoint of the political economist.

The occasional accolades for rival realist schools should nevertheless not be taken to imply too much. True enough, the alternative varieties of realism that, like classical and neoclassical realism, do take stock of unit-level factors in international politics (see Aron 1966; Kirshner 2012, 2022; Schweller 1996; Rathbun 2008; Ripsman et al 2014), are less vulnerable to the argument advanced herein than any kind of structural realism is. However, inasmuch as classical realism hinges upon the concept of the national interest defined along the lines of the unitary actor model (as is the case with the canonical formulation in the work of Hans Morgenthau 1951, 1985), it is fair game for the students of government failure just as neorealism is.

Neoclassical realism, in turn, is a heterogeneous phenomenon. Norrin Ripsman et al. (2014, 12), discern three types of this outlook. Type I seeks “merely to fix structural realism by using domestic-level intervening variables to explain away empirical anomalies for structural realist theories.” Type II deploys “systemic stimuli, moderated by domestic-level intervening variables, to inform an approach to foreign policy more generally.” Finally, the aim of type III is to propound a Waltz-style, grand theory of international politics. Type I’s explicit reliance on the core neorealist assumptions plainly makes it untenable from the vantage point of a robust political economy. As Brian Rathbun (2008) explains, this approach is but a “necessary extension of structural realism,” the starting point being “that ideal state behavior is that which conforms with the unitary actor and objectivity premises of neorealism” (Rathbun 2008, 312). This article, on the other hand, follows Austrian and public choice theorists in contending that the neorealist ideal itself is fanciful, and there is nothing inherently irrational (from the decision-maker’s viewpoint) in pursuing policies deviating from the neorealist precepts. They are features, not bugs.

The other two varieties of neoclassical realism do not necessarily take the domestic determinants of international politics to be but sources of erroneous behavior. They nevertheless still maintain the basic neorealist framework by regarding them merely as intervening variables, with the anarchy of international politics occupying the position of the independent variable (Ripsman et al 2014, chap.2). By contrast, the present study consciously advances a branch of liberal IR theory, as long as the latter is defined as positing that “the fundamental actors in international politics are individuals and private groups” and that “states … represent some subset of domestic society, on the basis of whose interests state officials define state preferences and act purposively in world politics” (Moravcsik 1997, 516, 518). The robust political economy perspective distinguishes itself from the mainstream liberal IR theories by applying these assumptions in a singularly consistent fashion. Unlike the standard liberal peace theories, which tacitly embrace the unitary actor model in envisioning states cooperating with one another for the benefit of their respective societies (Hanania 2022, 20; Ripsman et al 2014, 6), a robust political economy predicts that those “individuals and private groups” that get to “define state preferences” will not infrequently do so for the benefit of their own, and to the cost of the rest of the population. It is in this sense that this kind of political economy is not realist but profoundly realistic.^[4]

Although the article concentrates on OR as a grand theory of international politics, the work of a scholar as active in the public debate as Mearsheimer is difficult to engage with without alluding to his political commentary at all. Hence, Mearsheimer’s applied scholarship, especially his incendiary stance on the Ukraine war, is invoked on sundry occasions as an illustration of OR’s perils. Sometimes (e.g., Mearsheimer’s influential work on the Middle East conflict and the Israel lobby), that scholarship deviates from OR precisely because OR itself fails to account for the situation in question. Other times (e.g., much of Mearsheimer’s position on Ukraine), it is OR whose corollaries or contradictions produce flawed foreign policy analysis. Simultaneously, the article leaves aside those views of Mearsheimer’s (e.g., his findings about lying in international politics [Mearsheimer 2011] or the recent conception of homo theoreticus [Mearsheimer and Rosato 2023]) that are neither informed by OR nor resultant from its failures.

The argument unfolds in the following order: Section 2 briefly reconstructs Mearsheimer’s theory. Section 3 exposes OR’s grave fallacy: the failure to distinguish between hegemony and the bid for hegemony. This sets the stage for the original input of the article: the substantive criticism of Mearsheimer’s assumptions. Section 4 shows how OR plays fast and loose with uncertainty as an analytic category and demonstrates that the uncertainty assumption is ultimately parasitic upon the notion of survival as the principal (in fact, absolute) objective of states. Section 5, which thereby carries most of the present argument’s weight, questions that very notion. Lastly, Section 6 dissects the vision of leaders’ rationality that underpins OR and hints at ways IR theory could be illuminated by robust political economy in this respect. The final section concludes.

2 Offensive Realism

The “five bedrock assumptions” (Mearsheimer 2001, 30) of offensive realism are as follows:

1. The international system is anarchic in the sense that “there is no higher ruling body,” no “government over governments” (30).

2. States, and in particular great powers, possess some offensive capability, which makes them “potentially dangerous to each other” (30).

3. International relations are pervaded by uncertainty: “[s]tates can never be sure about other states’ intentions.” Most importantly, they cannot know for certain what other actors are planning on doing with their military capability (31).

4. “[S]urvival is the primary goal of great powers. Specifically, states seek to maintain their territorial integrity and the autonomy of their domestic political order.” The primacy of survival stems from the fact that it is a prerequisite for the attainment of any other goals states may wish to pursue (31).

5. “[G]reat powers are rational actors.” This means that they understand the nature of their environment, “think strategically,” consider the preferences of other actors, and are attentive to both short- and long-term ramifications of foreign policy choices (31).

These five premises, holds Mearsheimer, jointly entail that power-maximization is the most viable strategy states may undertake. As Mearsheimer (33) points out, “[a]pprehensive about the ultimate intentions of other states, and aware that they operate in a self-help system, states quickly understand that the best way to ensure their survival is to be the most powerful state in the system.” Briefly put, it is imperative that states seek hegemony, i.e., the status of “a state that is so powerful that it dominates all the other states in the system” (40). To this end, states will, and are well-advised to, deploy policies of expansion unless prevented from doing so by some objective constraints. Becoming the global hegemon would require the wherewithal to project military power all around the world, often simultaneously and across massive bodies of water. Hence, great powers must settle for a somewhat less ambitious objective: regional hegemony, preferably coupled with the absence of hegemons in other parts of the world so as to avoid their meddling in one’s own sphere of influence (40–2, chap.4).^[5]

Even so, Mearsheimer (20) believes that his theory improves on Waltz’s defensive realism by overcoming its notorious ‘status quo bias’ (Schweller 1996). As Mearsheimer avers, in an anarchic, self-help reality where actors possess the capacity to wreck one another, the balance of power may not be enough. For while survival is the principal goal of all states, some of them may have aggressive plans (Mearsheimer 2001, 414). In virtue of the uncertainty permeating international relations, no state leader can know whether his peers do not harbor purely hostile intentions that will prompt them to challenge the existing equilibrium. To secure themselves from such a contingency, states will then strive for hegemony rather than for the balance of power (29). This pattern constitutes the eponymous tragedy of great power politics: even absent predatory intentions in any of the actors, the epistemic barrier that separates their minds from each other incentivizes them to adopt offensive strategies (Pashahkhanlou 2013, 205).

What, on the face of it, can be said of Mearsheimer’s argument? The first two premises are indeed rather uncontroversial. The proposition that anarchy is the ordering principle of international politics has assumed the status of a received wisdom in IR, espoused nearly unanimously across the field (see Nye 2007, 3).^[6] That states, especially great powers, have some offensive capability is, in turn, an analytic truth that can be inferred from the very definitions of great power and the state as the Weberian monopoly of force more generally. The same, however, does not apply to the remaining three assumptions. But before the substantive critique of Mearsheimer’s outlook on uncertainty, survival, and rationality begins, let us first inspect whether his conclusions actually follow from the premises as they have been cited above.

3 The Elusiveness of Hegemony

Various objections have been raised as regards the validity of Mearsheimer’s reasoning (see Toft 2005 for an extensive overview). Among other things, it has been observed that OR’s five bedrock assumptions do not in the least bit go beyond those of Waltz’s defensive realism (Kirshner 2012, 60; Snyder 2002, 154; Tang 2010, 35; Toft 2005, 389). How come, then, that completely opposite conclusions are derived? An elephant in the room is the jump from the relatively uncontroversial assertion that hegemony is in and of itself a desirable thing to the far more contentious conclusion that it is reasonable to embark on a bid for it (Kirshner 2012, 61). In reality, four out of five such attempts invoked by Mearsheimer (Napoleonic France, Germany under Wilhelm II and again under Adolf Hitler, and Japan in the 1930s and 40s) failed miserably, usually resulting in the destruction of the contender state (which, recall, is exactly what state leaders fear the most in Mearsheimer’s account). The only successful bid, i.e., that of the US in the Western Hemisphere, can in turn be explained by the exceptional weakness of regional competitors (see Kirshner 2012, 62; Pashahkhanlou 2013, 215; Snyder 1991; Tang 2010, 45). This meager score appears to be no coincidence. As predicted by classic balancing theories, aggressive behavior triggers counter-armaments and balancing coalitions (see seminal treatments in Aron 1966; Walt 1990; Waltz 1979). Furthermore, it cannot be held that the gains brought about by hegemony are worth the risk. To the contrary, great powers are, by definition, states that have enormous military resources, which translates into robust deterrence and defensive capabilities (Kirshner 2012, 61–2).

The pointlessness of bidding for hegemony has been articulated perhaps most perspicuously by Steinsson (2014): “If states were irrational, if leaders of states were less concerned about a state’s survival than their own bank accounts or re-election prospects, if states were assumed to be aggressive …, one could understand the tendency to risk it all for hegemony.”

As will be seen, Mearsheimer’s theory offers an implicit rebuttal against this objection. Although it seemingly elaborates on the concept of uncertainty, the argument actually reduces to a radical view on states’ survival. This conception is centered around something which may be labelled absolute security, that is, security as the highest, ungradable, and non-exchangeable good. A careful inspection of this idea, cognizant of a robust political economy, furnishes plausible reasons to believe that at least some characteristics suggested in the above quote from Steinsson fare better in explaining states’ bellicose conduct than the security-seeking underscored by neorealism.

4 Uncertainty

The notion of uncertainty is critical for Mearsheimer’s theory. Mearsheimer and Rosato (2023, 19) articulate this point unequivocally: uncertainty is “the defining feature of international politics”. As Mearsheimer (2001, 45) explains in a well-known passage from The Tragedy, “intentions are ultimately unknowable, so states worried about their survival must make worst-case assumptions about their rivals’ intentions.”

At first glance, Mearsheimer might seem to stand on a firm ground. As Frank Knight (1921) taught us, uncertainty is all-pervasive in human affairs (see also a detailed exposition of Knightian uncertainty in Mearsheimer and Rosato 2023, 23–5). Unlike risk, which is quantifiable and thereby predictable, uncertainty cannot be articulated in numerical terms. Both economy and politics are invariably marked by uncertainty for the simple reason that, in stark contrast to natural events, they are not (at least from the observer’s vantage point) predetermined but instead hinge on individual choices of human actors. Moreover, as Ludwig von Mises (1998, 105–18) observes when expounding upon his twin distinction between case and class probability (Knightian uncertainty and risk, respectively), whereas uncertainty (case probability) arises from the scrutiny of individual cases, (class) probability has to do with homogenous classes of phenomena with no case-specific information available (as in dice rolling).

This, however, by no means entails that the social reality is unpredictable, period. ‘Uncertain’ does not equal ‘unknowable.’ Indeed, Mises’s term ‘case probability’ may be more handy in grasping this fact than Knight’s ‘uncertainty,’ as case probability is still some probability. Doubtless, there exist various factors ameliorating the uncertainty of human conduct. One is instrumental rationality: we surmise people will do such-and-such because, given the ends they presumably have in mind, it would be unwise of them not to do so. The other sources, like psychology, historical understanding (Verstehen), or personal acquaintance, are in turn supposed to provide knowledge of the goals that actors pursue (Mises 1998, 30–71). If those sources did not exist, any peaceful social interaction would be impossible (Keohane 1993; Wendt 1992). And yet it is possible, often in immensely complex forms, many of which require highly advanced forecasting skills. Perhaps the most instructive example is the free market: entrepreneurs and capital owners get to make profit, thereby demonstrating correct judgment regarding future preferences of consumers; workers make choices that prove advantageous on the future labor market, and so forth (Kirzner 1973).

The same is true of international politics. Policymakers make guesses about the intentions and future actions of their peers based on their personal relationships, the current behavior of the states they run, the history of bilateral relations, or intelligence – and the guesses are far from wild. The Canadians, for example, do know, even if without certainty, that they will not be overrun by the Americans any time soon despite the glaring imbalance of power between the two nations.

In particular, as per the fifth assumption of OR, states are instrumentally rational. It is also difficult to fathom how a concept of instrumental rationality could not comprise the actor’s ability to learn. In particular, as we know from the basics of game theory, any strategic considerations presuppose that actors are capable of understanding and responding to the preferences of others, which in turn are pieces of private information that must somehow be signaled rather than shown directly (Fudenberg and Tirole 1991, chap.8 and 9). In Mearsheimer’s (2001, 31) own words, rational states “consider the preferences of other states and how their own behavior is likely to affect the behavior of those other states, and how the behavior of those other states is likely to affect their own strategy for survival” [italics added]. Several pages later, Mearsheimer contradicts himself when he writes: “When a state surveys its environment …, it focuses mainly on the offensive capabilities of potential rivals, not their intentions” (45). The offensive realist is therefore presented with a dilemma: either the preferences of other states are unknowable, in which case no rational strategic thinking is possible, or they are knowable, in which case benign states can successfully signal their peaceable attitude.

Ponder Mearsheimer’s views on the causes of the Ukraine war (Mearsheimer 2014; 2018, 171–9; 2022). Did he believe that the intentions of other states are unknowable, he would not be in a position to hold that the West was primarily responsible for Russia’s aggression by defecting on its earlier declarations not to expand NATO eastward. It is only assuming that states learn about each other’s intentions that Mearsheimer’s argument about the causes of the Ukraine war can even be made (Edinger 2022). Moreover, not only does Mearsheimer’s own work evince that he embraces the uncontroversial proposition that others’ intentions are somehow knowable, but he is also forced to believe so as a matter of logic, in virtue of the rationality assumption of OR. That is to say, it is not that in his treatment of the Ukraine war, Mearsheimer simply fails to follow his own theory: it is the theory itself, with its incompatible assumptions of rationality and irrelevance of intentions, that generates the inconsistency.

As a matter of fact, Mearsheimer usually shies away from explicitly making the – strikingly fallacious – inference that since the course of human action is uncertain, statesmen know nothing about what their counterparts are going to do. Rather, he settles with the far less contestable – in fact, incontrovertibly correct – claim that “[s]tates can never be sure about other states’ intentions” (Mearsheimer 2001, 31). Well, of course they cannot be sure: after all, human action is uncertain, so capitalists, entrepreneurs, and laborers are never sure about their judgments either. Apparently, then, what OR really says is that the levels of uncertainty that we learn to tolerate and manage in other spheres of life are unacceptable in the realm of international politics.

In actuality, therefore, Mearsheimer’s theoretical argument is not really about uncertainty – as it has been defined by great economists such as Knight and Mises. His case boils down to an equivocation: the reasoning explicitly starts out with the standard Knightian notion of uncertainty, thereby coming across as less contentious than it is, but then implicitly ends up anchored in something completely different. According to Shiping Tang (2008), the key to make sense of OR is Mearsheimer’s (2001, 45) assertion (absent from defensive realism) that states “must make worst-case assumptions about their rivals’ intentions.” Plainly, this desideratum does not follow from Mearsheimer’s uncertainty premise, even if “uncertain” were interchangeable with “unknowable” (which it is not). It is, on Tang’s (2008, 458) interpretation, Mearsheimer’s “sixth assumption’, his ‘Sixth Element’.” Whence does this element come from? Tang (454) interprets Mearsheimer to the effect that “states become inherently aggressive due to anarchy.” That is to say, even absent any natural, built-in propensity to aggression in states (as posited by classical realism), states are quick to adopt an aggressive posture. Tang’s exegesis, however, still begs the question: what exactly compels them to do so? The recourse to anarchy alone will not do: even if the plans of other states were unknowable, anarchy itself would justify the worst-case assumption no more than any other motivational assumption. As shall presently be seen, to Mearsheimer, the real driving force behind states’ expansionist undertakings is the premium they put on survival, to which the whole uncertainty (or better put: assuming-the-worst) assumption ultimately reduces.^[7]

5 Survival

6 States’ Rationality

As has been shown, Mearsheimer’s notions of uncertainty and survival, married with the assumption of states’ rationality, do not suffice to prove that a rational state will pursue hegemony to secure its survival in an uncertain world. But perhaps something is amiss in Mearsheimer’s concept of rationality as well? Some critics charge that the rational actor assumption, which sits underneath the structural realist theories, is oversimplistic and psychologically untenable in and of itself (Freyberg-Inan 2004, 111). Mearsheimer and Rosato (2023), however, are right when they hold that to rid oneself of this assumption entirely would be to lay the entire enterprise of IR theory in its grave. While acknowledging the limitations of human rationality, robust political economy suggests that instead of throwing the baby out with bath water, one should rather compare different institutional frameworks with an eye on the constraints on rationality they produce (Caplan 2007; Pennington 2011).^[16]

To start with, recall that in The Tragedy, Mearsheimer (2001, 31) characterizes his rationality assumption as follows: states “are aware of their external environment and they think strategically about how to survive in it.” Thus, Mearsheimer introduces survival to the very notion of rationality as a substantive, motivational facet (cf. Donnelly 2000, 63). This view has been reiterated explicitly in How States Think, where Mearsheimer and Rosato (2023, 211) defend what they refer to as states’ goal rationality: “[a]lthough rational states invariably have many goals, they rank survival as the most important,” they write.

To elevate survival’s priority to the rank of rationality’s defining feature is a very bold move. For sure, far more argumentative effort is needed here than Mearsheimer’s standard “survival is the prerequisite of all other goals” talk. In addition to what has already been said on that matter in the previous section, several further observations regarding states’ rationality are worth making.

When elaborating on the fifth bedrock assumption of OR, Mearsheimer adds one more substantive element to his idea of rationality. He (2001, 32) asserts: “[s]tates pay attention to the long term as well as the immediate consequences of their actions” (italics added). This, too, is by no means self-evident. Some people are future-oriented, some are not. Depending on multifarious variables such as personality, age, wealth, or institutional arrangements, the level of what Austrian school economists call time preference (i.e., the premium placed on the current satisfaction of wants) varies across the population (on the concept of time preference, see Herbener 2011; Mises 1998, 480–6; Rothbard 2009, 15). Likewise, the time preference exhibited by one person can change significantly under different incentives. For instance, an increase in one’s money stock helps save a greater part of one’s income, which reveals lower time preference (Block, Barnett II, and Salerno 2006; Rothbard 2009, 410–16). On the other hand, there are factors contributing to high time preference. One of them may be the political system. Given the temporal limitations of the democratic term, democracy incentivizes decision-makers to act in a myopic fashion, i.e., to display high-time preference. For if one’s continued stay in office hinges upon the outcome of the next election, it is the time horizon of this election, not the distant ramifications of present decisions, that matters the most. Even if leaders were as public-spirited as Mearsheimer tacitly assumes them to be, they would still lack the incentives to pursue the long-term interests of the country whenever that goal collided with the need to win elections (which must be won if the public interest, as politicians see it, is to be sought at all). Things are not so much different in dictatorships. Unlike hereditary monarchs, dictators do not hold a legal title to pass their office down to the future generations of the dynasty, so the link between the long-term interest of the country and their own remains tenuous (Hoppe 2007).

By contrast, OR posits relatively low time preference on the part of state leaders, irrespective of systemic conditions. Admittedly, when negating the long-term benefits of institutions envisioned by liberal institutionalism, Mearsheimer (1995, 82) asserts that “[i]nstitutions cannot get states to stop behaving as short-term power maximizers” (italics added). This is hardly surprising, since it is a well-founded insight of liberal institutionalism that iterated games favor cooperative strategies over defection (Axelrod 1984): a highly unwelcome result for someone who has little faith in international cooperation.

Nevertheless, Mearsheimer’s own theory as well is premised on the assumption of politicians being long-term strategic players, only that it oddly ignores the prospects for interstate cooperation that come with such a posture. As Tang (2008, 460) notes, in Mearsheimer, states “must have a very long time-horizon when preparing for conflict. At the same time, however, states must also have a very short time-horizon when considering whether to forge cooperation.”

Consider three points made by Mearsheimer time and again. He argues that even if great powers are safe now, they must make sure that things stay that way in the future: they must be alert to the future changes of other states’ intentions as well as in the distribution of power (Mearsheimer 2001, 34–5). Consequently, the balance of power is not enough. Instead, states pursue regional hegemony and see to it that a competing hegemon not emerge in another part of the world (Mearsheimer 2001, chap.2). Mearsheimer (2014; 2018, 171–9; 2022) applies this contention to Russia’s foreign policy: as a great power protecting its sphere of influence, Russia cannot but go to great lengths to prevent Ukraine from joining NATO. Otherwise, Ukraine could one day be used by the West as a stationing ground for an invasion of Russia. By the same token, whereas the PRC should attempt to become a regional hegemon, the US must do (and should have done decades ago) everything it can to prevent that from coming to pass (Mearsheimer 2001, chap.X; 2006).

Observe that each of these claims is informed by the assumption that leaders devise their plans having the long run in mind. Indeed, China subjugating the American allies in its neighborhood and then aligning with Mexico to threaten the US with nuclear warheads planted on Mexico’s territory, or the US Army using the plains of Ukraine to initiate Operation Russian Freedom, are highly hypothetical possibilities that in all likelihood will not materialize in decades, if whenever.^[17] It is debatable whether leaders of a liberal democracy and two post-communist dictatorships even think that far ahead, let alone make such developments their principal concern. While this is certainly not impossible, a general theory presupposing this degree of far-sightedness in politicians rests on very shaky foundations.^[18]

In summary, OR again betrays a good deal of idealistic thinking. Not only are statesmen deeply concerned about the interest of their nation but, so we are being told, they are also busy securing it for generations to come. Recall now the discussion of the concept of survival in Mearsheimer’s theory. The conclusions of the present section supply an additional reason for Mearsheimer’s adopting a nationalist rather than a self-interest-based account of that concept. Were statesmen thought to be driven by narrow self-interest, the suggestion that they care about their states’ survival in the distant future would become strikingly groundless, perhaps save for hereditary monarchs. It is only provided that leaders’ intentions were benign that they could possibly devise their plans in such a far-sighted manner.

It is worth noting that this far-sightedness assumption once more makes OR even less plausible than defensive realism is. The struggle for the balance of power may well proceed with the politicians involved thinking within the confines of their normal electoral business and the related time horizon. Hardly any leader wants to be attacked by a superior power on their own watch. As mentioned, when a country is being threatened by a foreign power, its behavior may begin to resemble more closely the unitary actor model and to follow the precepts of balance of power politics. Offensive realism, on the other hand, posits that those are followed all the time and with regard to the distant future. By the same token, the assumption of far-sightedness aggravates OR’s uniquely acute problems with opportunity cost and diminishing marginal utility: the theory presupposes that not only the present but also the future utility of security never decreases in comparison to alternative uses of scarce resources, including present uses.

Embracing statesmen’s narrow self-interestedness by the structural realist would have one more ramification germane to the theory of rationality in politics. If the chief concern of foreign policymakers is to stay in power, then, in most cases, the most pressing threats to that endeavor usually lurk at home, not abroad. These are posed by domestic competitors and the public. Yet with this fact acknowledged, the concept of states’ rationality has to be cleansed from elements of the national interest such as the primacy of state survival or long-term planning. The question thus becomes not if policy X was rational insofar as the national interest is concerned, but whether it was rational given the presumed purposes of the individual actors and groups they form below the state level.^[19] Accordingly, the researcher has to attend to the specific configuration of organized interests in a state; they have to figure out what those interests are, how powerful they are relative to one another, and how likely the policymakers are to respond positively to them given their own purposes and the incentive structures they operate in. In a word, the national interest must be disaggregated.

As it happens, political economists have long been exploring the topic of how government policies tend to serve the special interests of politicians and pressure groups at the expense of the majority of citizens (for classic discussions, see Buchanan and Tullock 1987; Caplan 2007; Downs 1957; for applications in IR, see Coyne 2008, 2023; Hanania 2022; Huemer 2013; Trantidis 2023). One of the questions that have never been properly addressed by neorealism is why foreign policy should be any different. If it is not, then policies of aggression and bids for hegemony, ultimately left unexplained by structural realism, could be understood as both rational and deleterious to the commonweal. Curiously enough, one of the persistent themes in Mearsheimer’s own scholarship is his contention that the US continued commitment in the Middle East conflict on the side of Israel does not stem from careful security calculations but instead from the prevalence of the ‘Israel lobby’ in American domestic politics (Mearsheimer and Walt 2007). From the structural realist perspective, this situation can only be seen as an anomaly, to be tackled by an analysis along the lines of type I neoclassical realism. In a manner of speaking, the Mearsheimer who penned The Tragedy and the one that co-authored The Israel Lobby are two different scholars: the former set forth the theory of OR, whereas the latter was pushed to deviate from it by the theory’s inability to account for a major phenomenon of our day. The follower of a robust political economy, in turn, will consider the US commitment to Israel anything but vexing. For him, it is only natural that foreign policy serves organized interests, whereas the public is (rationally) too ignorant and passive to keep them in check.

7 Conclusions

This article has subjected Mearsheimer’s theory of offensive realism to a thorough-going critique. Upon examination, it transpired that Mearsheimer’s conceptions of uncertainty, survival, and rationality are either insufficient to substantiate the claims of OR or seriously strained. As the paper submits, this lack of soundness is in no small part down to the neglect of economics and political economy.

The inquiry found Mearsheimer’s reasoning not only wanting but also surprisingly idealistic and utopian, especially for a theory advertised as realist. As regards idealism OR apparently cannot get off the ground without positing (a) that policymakers and citizens are motivated primarily by the common good defined in terms of the ideology of nationalism, and (b) that both groups are capable of overcoming numerous disincentives that stand in the way of such altruistic conduct. These presuppositions are implicit in Mearsheimer’s absolutist account of national security, thought of as an ungradable, non-exchangeable good that must be assured irrespective of the cost involved. It is this notion that justifies the tremendously hazardous and costly policies advised by OR. The absolutist account of national survival is also the real key to understand OR: as has been shown, the other two contentious premises of it, that is, the ideas about uncertainty and rationality, are, in the last analysis, parasitic upon that conception. Its inherent idealism contrasts sharply with the feature that has long been a hallmark of political realism, especially in the classical tradition: the emphasis on human imperfection (see Donnelly 2000, 6–42; Kimla 2009, 147–60).

In thinking that way, the offensive realist resembles E.H. Carr’s (1946, 6) personification of the utopian: “The thought of the alchemist was purely purposive. He did not stop to enquire whether the properties of lead were such as to make it transmutable into gold. He assumed that the end was absolute (i.e. that gold must be produced), and that means and material must somehow be adopted to it.” Mearsheimer’s gold is absolute security guaranteed by regional hegemony and the absence of another hegemon anywhere else in the world. This, as Kirshner (2012, 66–8) aptly notes, may run counter to yet another view commonly associated with IR realism: the acknowledgement of power and its limits. The lack thereof in OR, combined with the downplaying of political goals other than national survival and the embrace of a peculiar, non-biological and nationalist understanding of the latter, makes the theory not only erroneous but also potentially dangerous in the time when great power rivalry is back on the table. Despite Mearsheimer’s repetitive calls for ‘restraint’ regarding enemies he finds insignificant for the US national interest, OR proves singularly instrumental as a rationale for threat inflation and imperialism (Kurowska 2023; Specter 2022). The very same theory that dovishly advises the US to appease Russia and sever its military ties to Israel simultaneously shows its hawkish face when it comes to justifying Russia’s own adventurism or the rising tensions in the US-China relations. Robust political economy, on the other hand, supplies much more consistent grounds on which to doubt the benignity of state leaders claiming to wage ‘necessary’ wars in the public interest.^[20]