Matthew Kroenig’s provocative piece for Foreign Policy claimed that “realism has become unrealistic” but his entire argument rests on a flawed conceptual foundation: he treats Realism as though it were a monolithic theory with a single set of predictions. It is not. Realism is better understood as a paradigm, a school of thought or a family of theories. In addition to classical realism, it encompasses multiple distinct approaches, including neorealism with its defensive and offensive variants, and neoclassical realism. These theories share important assumptions about power, security, competition, and the constraints imposed by international anarchy, but they differ substantially in what they seek to explain, how they conceptualize causation, and what they predict about state behavior. This distinction matters because Kroenig’s central criticism that Realism has repeatedly failed to explain the U.S. foreign policy behavior depends on aggregating the predictions of different realist theories and treating their explanatory power as a failure of Realism as a whole. However, that is not a satisfactory way to evaluate the validity of a paradigm.
Get E-International Relations delivered to your inbox, free. As you sign up, consider becoming a paid subscriber, or make a donation, to support our work.
Kroenig is right about one thing: theories should be confronted with empirical evidence. If a theory repeatedly generates predictions that are contradicted by observable behavior, scholars should reconsider its assumptions. But the appropriate unit of evaluation is the theory or model that actually generates the prediction. One cannot discredit an entire paradigm simply by demonstrating that particular theories within that paradigm have made particular predictions that turned out to be wrong.
Consider neorealism. Kenneth Waltz’s structural realism emphasizes the international system, particularly the distribution of capabilities among states, as the principal determinant of international outcomes. It deliberately abstracts away from many domestic-level variables and stipulates that it is not a theory of foreign policy but rather a theory of international politics that primarily concerns itself with explaining how systemic balances of power form recurrently.
Offensive realism, associated most prominently with John Mearsheimer, makes a stronger and different claim: because states exist in an anarchic system and cannot be certain about one another’s intentions, great powers have powerful incentives to maximize their relative power and ultimately seek regional hegemony. Defensive realism reaches different conclusions. Scholars such as Charles Glaser, Jeffrey Taliaferro, and Stephen Walt have emphasized that security does not necessarily require unlimited accumulation of power. Excessive expansion can provoke balancing coalitions and thereby make states less secure. From this perspective, the optimal strategy for a state depends on the security environment, the offense-defense balance, the intentions and capabilities of other states, and the costs of expansion. These are not semantic variations on one theory. While inhabiting the big tent of Realism, they are competing theoretical propositions that result in a distinctly diverse set of predictions about world politics in general and particular states altogether.
And then there is neoclassical realism, which is particularly important for understanding U.S. foreign policy nowadays, according to Kroenig’s key yardstick. Neoclassical realists accept the importance of systemic pressures but argue that the international distribution of power does not translate automatically into foreign policy behavior. The effects of systemic incentives are mediated through domestic political institutions, leaders’ perceptions, state capacity, elite consensus, and other intervening variables.
This is precisely where Kroenig’s argument becomes vulnerable. A state does not simply observe the international distribution of power and mechanically produce the foreign policy that a structural realist model predicts. International incentives must be interpreted by human beings operating within domestic political systems. Leaders perceive threats differently. Political institutions filter external pressures; states possess different strategic cultures, varying abilities to extract resources and translate them into national power, and domestic coalitions influence which strategic options are politically feasible or not. Neoclassical realism was developed in significant part to explain this gap between systemic incentives and actual foreign policy behavior. In other words, neoclassical realists recognized the empirical anomalies that Kroenig rightfully identified and responded by revising the theoretical explanation to account for such irregularities. A neoclassical realist explanation would be severely challenged if deviations from systemic expectations could not be consistently traced to domestic-level intervening variables or if states with similar systemic pressures and domestic conditions nonetheless behaved in systematically different ways for reasons the theory cannot account for.
Contrary to what Kroenig claims, the broad and intricate history of American foreign policy does not make Realism irrelevant, and the United States’ postwar creation and maintenance of alliances do not automatically refute Realism either. Nor do NATO enlargement, nuclear modernization, support for Taiwan, aid to Israel, or support for Ukraine. Each of these policies can be interpreted through different theoretical Realist lenses that would emphasize balancing, deterrence, relative power, strategic geography, or competition among great powers.
The case of Donald Trump illustrates the point particularly well. Trump poses an obvious challenge to Realist theories that assume a relatively stable relationship between the international system and U.S. grand strategy. His foreign and security policy has contained seemingly contradictory elements: skepticism toward traditional alliances alongside concern about Chinese power; hostility toward multilateral institutions alongside aggressive economic statecraft; demands that allies spend more on defense alongside continued recognition of the importance of U.S. military primacy; transactional diplomacy alongside threats of coercion; and a willingness to question longstanding foreign policy orthodoxies while simultaneously pursuing policies designed to preserve American leverage.
Trump is difficult to explain through a simple structural-realist model because the international distribution of power did not suddenly change when Trump entered the White House. China was rising, Russia possessed substantial military capabilities, the United States remained extraordinarily powerful, and the basic structure of the international system was broadly continuous with the preceding administration. What changed, to a large extent, was the domestic and perceptual intervening variable: the president and American politics, especially during his second term. This is precisely the sort of phenomenon neoclassical realism is designed to accommodate. Trump’s foreign policy cannot be understood merely by asking what the international system “requires.” We must also ask how Trump perceived the international environment, how he conceptualized American power, how domestic political pressures shaped his preferences, how bureaucratic institutions constrained or enabled him, and how his administration translated his preferences into policy.
Kroenig argues that Realists failed to predict the American decision to remain committed to NATO in the aftermath of the Cold War. Generally speaking, Realists believe that international institutions can facilitate cooperation while simultaneously serving the strategic interests of powerful states. Trump’s approach to NATO and European security provides a particularly useful example. He repeatedly criticized NATO allies for insufficient defense spending and questioned aspects of the traditional American commitment but did not simply dismantle the alliance. Instead, Trump frequently sought to extract greater contributions from allies while preserving the strategic advantages of the alliance itself. Recent data suggest that Trump was successful in forcing key NATO members to increase their defense spending as he wanted and assume more responsibility over European security, including in Ukraine, while freeing Washington to address other, and possibly more urgent matters, such as the rise of China in East Asia.
A neoclassical realist interpretation can make sense of this apparent contradiction too. The international environment creates incentives for the United States to maintain a favorable balance of power in Europe with an eye on other strategically critical areas. But the precise form that policy takes depends on how the president interprets the value of alliances, how domestic constituencies evaluate alliance burdens, and how the executive branch translates those preferences into bargaining strategies.
Trump’s recent war with Iran in an effort to destroy its nuclear program, and the ensuing Strait of Hormuz crisis, further illustrate why Kroenig is too quick to treat Realism as a single theory or cherry-pick some realists who may have ideologically opposed the war as the only voices of that expansive school of thought. The conflict actually reflects familiar realist concerns such as denying Iran strategic leverage, protecting a vital maritime chokepoint, and preserving U.S. power in the Persian Gulf, but Trump’s response cannot be explained by systemic power alone. His transactional approach, emphasis on coercive bargaining, sensitivity to domestic economic pressures, and willingness to alternate between escalation and negotiation point to the importance of leaders’ perceptions and domestic politics. This is precisely what neoclassical realism is designed to capture: systemic pressures shape the strategic environment, but domestic and perceptual variables determine how states respond. Rather than demonstrating that Realism is “unrealistic,” Trump’s Iran policy illustrates why the paradigm requires multiple theories to explain different dimensions of American foreign and security policy.
Kroenig’s argument therefore risks turning a theoretical debate into an argument over whether American foreign policy is appropriate, or if certain realists-turned-pundits got it right. But those are different questions. An international relations theory is not necessarily falsified because policymakers make decisions that the theorist considers unwise. Nor is a theory validated because policymakers supposedly follow its recommendations. The relevant question is whether the theory explains and predicts observable patterns of international behavior better than its competitors. Kroenig is correct in criticizing some realists for their failed theoretical predictions. The proper response to anomalies, however, is not to declare Realism dead, but to determine which realist theory failed, under what conditions, and why.
Perhaps structural realism explains broad strategic constraints but is insufficient for explaining variation in foreign policy behavior within the same international system. Perhaps offensive realism overstates the tendency of states to maximize power. Perhaps defensive realism provides a better account of situations in which security rather than power maximization is the objective. Perhaps neoclassical realism provides the better account of that variation because it incorporates domestic and perceptual intervening variables. Those are meaningful theoretical propositions that can be tested against evidence and criticized when appropriate.
Indeed, this is one of the strengths of the Realist tradition since antiquity: it has continually evolved in response to empirical and theoretical challenges or innovations. Classical realism, neorealism, defensive realism, offensive realism, and neoclassical realism are not interchangeable labels. They represent different attempts to identify the mechanisms connecting power and security to international outcomes, albeit under the overarching Realist tent.
The irony, then, is that Kroenig’s criticism of Realism may itself overlook one of its most important insights: states do not respond to power distributions in a mechanical fashion. Power matters, but how states perceive power matters too. Structural constraints matter, but domestic institutions determine how states respond to those constraints. Security incentives matter, but leaders disagree about what security requires. The international system creates pressures, but those pressures are filtered through states before they become foreign policy. That is not a rejection of Realism. It is the paradigm’s built-in mechanism of theoretical progression.
The appropriate conclusion from the empirical record is therefore not that “Realism has become unrealistic.” A more valid conclusion is that no single realist theory can explain every dimension of international politics. That should hardly be surprising. The existence of multiple realist theories reflects an effort to specify different causal mechanisms rather than an intellectual failure of the paradigm.
Kroenig is correct that Realists should take empirical anomalies seriously and be more attentive to the connection between theory and the policy world. But doing so requires more theoretical precision than his critique provides. Before declaring Realism empirically bankrupt, we should ask which Realism we are talking about. We should also carefully differentiate between the scholarly contribution of realist thinkers and the punditry some of them generate. The former needs to be evaluated by using the common tools of scientific inquiry and the latter through a more ideological or political prism.
If the question is why the international system produces incentives for balancing, deterrence, competition, and concern about relative power, neorealism remains indispensable. If the question is why states sometimes respond differently to similar systemic conditions, defensive realism and offensive realism offer different answers. And if the question is why American foreign policy can vary so dramatically under presidents confronting essentially the same international distribution of power, neoclassical realism may be particularly valuable.
Donald Trump is not necessarily the anomaly that disproves Realism. He may instead be the anomaly that demonstrates why Realism requires more than one theory. As students return to university this semester and enroll in international relations (IR) courses, to use Kroenig’s opening from his Foreign Policy article, professors should make it clear that Realism encompasses a diverse range of approaches, and that it is not a theoretical monolith.
Ilai Z. Saltzman is a Professor of Israel Studies and the Director of the Joseph and Alma Gildenhorn Institute for Israel Studies at the University of Maryland, College Park, and a board member at Mitvim – the Israel Institute of Foreign Regional Policy. He is the author of Securitizing Balance of Power Theory: A Polymorphic Reconceptualization.