Why Smart Leaders Do Stupid Things

From an essay on foreignaffairs.com by Karen Yarhi-Milo headlined “Why Smart Leaders Do Stupid Things”:

To many observers, Russia’s decision to invade Ukraine was, obviously, irrational. Ukraine is Europe’s largest country, and Russian President Vladimir Putin was short on both good troops and quality weapons. No other state (aside from Belarus) was sympathetic to the idea that Moscow should control Kyiv, and the United States had uncovered Putin’s invasion plans and then released them to the entire planet. Most of the world’s biggest economies threatened to hit Russia with sanctions if it went ahead with its attack, and NATO countries made it clear that they would arm Kyiv.

But to the political scientists John Mearsheimer and Sebastian Rosato, Putin’s decision still makes sense. In their new book, How States Think: The Rationality of Foreign Policy, they argue that Putin and his advisers “thought in terms of straightforward balance-of-power theory,” viewing Ukraine as a bulwark against NATO and Kyiv’s possible membership in the organization as a “redline.”

Keeping Ukraine in Russia’s column, the authors write, was a “matter of life and death” for the Kremlin. If Russia goes on to lose the war in Ukraine or if Putin loses power because of the conflict, the authors contend, it will therefore not be because the invasion was irrational. It will, instead, be a result of Russia’s military incompetence and NATO’s effort to help Ukraine balance against Russia.

There was a time when Mearsheimer and Rosato’s beliefs would have been conventional wisdom, or at least quite popular, among scholars of international relations. For much of the twentieth century, “realism”—a theory made broadly popular by figures such as Kenneth Waltz, Henry Kissinger, and George Kennan—dominated the discourse. Realist scholars believe that states behave according to the same, inexorable logic. They behave rationally, working to maximize their power and protect themselves from attack in an anarchic world. To these experts, the psychology of leaders mattered little. It was the shape, contour, and distribution of power of international systems that dictated how states behaved.

But over the last several decades, the field has seen a quiet revolution. Political scientists began studying how leaders think, what biases they hold, and how these characteristics shape decision-making. They found that, overwhelmingly, psychology has an enormous effect on leaders’ behavior on the international stage. Leaders frequently rely on heuristics to make choices—especially during crises. Leaders’ beliefs, their personalities, and their impressions of their peers influence how they see the world. And their feelings shape how they approach different problems and situations. Putin’s emotional fixation on controlling Ukraine, for example, is frequently cited as the reason he invaded the country.

How States Think tries to undermine these claims and resurrect the older way of thinking. The authors argue that most international decisions are, in fact, rational. They work to poke holes in the scholarship of various political psychologists but also of rational choice theorists, whose definition of rationality (that leaders make decisions based on the expected value of the outcome) differs from the authors’ more circumscribed definition. The very title of the book is a play on a pathbreaking volume of political psychology—How Statesmen Think—edited by the late Robert Jervis. (Full disclosure: Jervis was my mentor.) Personalities, Mearsheimer and Rosato argue, may not be entirely irrelevant in international politics, but they do not matter nearly as much as scholars think.

Mearsheimer and Rosato’s book is an important entry into the debate over rationality in international relations, and it does a decent job of showing why academics struggle to determine which decisions can be considered rational. But the book ultimately fails to establish that countries behave in rational ways. The authors cannot come up with a compelling definition of rationality. They do not explain why what is rational for a leader is also rational for a state. They ignore vast primary and archival data that cuts against their arguments. They come up with post hoc explanations for what they deem rational, highlighting their own bias. And the examples they use to prove their claims frequently undermine them—including the invasion of Ukraine.


Mearsheimer and Rosato are not upstart insurgents in their discipline. Mearsheimer is one of the most famous political scientists in history. His seminal 2001 book, The Tragedy of Great Power Politics, established the idea of offensive realism, which argues that states always maximize their power to guarantee their survival. Rosato, Mearsheimer’s former pupil, made his name by logically dismantling democratic peace theory: the idea that democracies tend not to go to war with each other.

Their new book is sweeping. In How States Think, Mearsheimer and Rosato examine policymakers’ collective choices from World War I to the present. They revisit major choices from the past that have often been considered nonsensical and argue that they were, in fact, quite rational. Even Germany’s disastrous 1941 invasion of the Soviet Union and Japan’s attack on Pearl Harbor that same year are cited as rational decisions.

Some of Mearsheimer and Rosato’s criticisms are valid. The authors correctly point out that “rationality” is a foggy term, and they are rightfully dismissive of commentators who decide whether an action was a success “and then reason backward” to determine whether it was rational. The authors also effectively pick apart overly narrow definitions of rationality that make almost every leader come off as a cartoonish madman.

But when Mearsheimer and Rosato begin to spell out their own theory of rationality, the book’s claims unravel. They write that rationality is “making sense of the world for the purpose of navigating it in the pursuit of desired goals” and that rational decisions are those based on theories supported by “realistic assumptions,” “compelling causal logic,” and “evidentiary support.” It is a formulation that is just as squishy as the definitions they take down. All leaders, after all, think their theories, ideas, and choices are consistent, logical, and well supported—and there is rarely an objective test that can prove otherwise or that does not rely on post hoc reasoning.

The authors unintentionally illustrate this problem when they explain which theories they consider credible and which ones they do not. They dismiss as unconvincing the domino theory, which holds that if one country becomes a democracy or a communist dictatorship, then its neighbors would quickly make the same switch. Yet they argue that Putin’s belief that Russia and Ukraine are part of one country is credible because, historically, Ukraine has been Moscow’s strategic buffer against the rest of Europe. There is no objective reason why decisions guided by domino theory are irrational and Putin’s attack on Ukraine is not. But there is a subjective one. Mearsheimer and Rosato are both realists, and according to their version of realism, Putin’s decision was the natural response to NATO expansion. A theory’s credibility, in other words, is in the eye of the beholder.

In trying to illustrate their argument, Mearsheimer and Rosato also ignore a strong body of literature in international relations on the topic of how leaders think—literature that draws from psychology and behavioral economics, uses primary source materials, and features experimental data on elites. Even as they try to pick apart opposing research, Mearsheimer and Rosato barely mention foundational studies that demonstrate how leaders are motivated by emotions, preexisting beliefs, concerns about reputation, and other factors.

And even if Mearsheimer and Rosato are right about which decisions are rational, it does not mean that leaders are making them for rational reasons. Putin’s decision to invade Ukraine, for example, may have had little to do with balance-of-power calculations. Instead, Russia’s president could have invaded because he perceived himself as being in a domain of losses, making him less risk averse, or because he wanted to avoid a domino effect in which Ukraine’s admission into NATO would prompt more countries along Russia’s border to join the organization. Both are explanations the authors would not consider rational. Leaders can hold multiple, sometimes competing, theories at one time. The Iraqi leader Saddam Hussein, for instance, believed the United States was too worried about casualties to invade his country. But he still feared the possibility of invasion, so he hinted that he might have weapons of mass destruction in an effort to deter attacks. Scholars simply do not have enough information to determine which theories leaders are following or whether they are doing so in ways experts would predict.

British Prime Minister Neville Chamberlain’s infamous appeasement of Adolf Hitler provides another case in point. Mearsheimer and Rosato say that Chamberlain’s decision to let the Nazi leader annex large swaths of Czechoslovakia in 1938, rather than fight against the German war machine, was rational and driven by balance-of-power reasoning. Appeasement, they argue, was consistent with the British government’s theory that Hitler’s expansionist intentions were limited and that Berlin wanted to avoid war. But when he flew to Munich, Chamberlain had mounting evidence that Germany wanted far more territory and would use force to get it. He was, therefore, aware of a different theory about Hitler’s behavior, one holding that Berlin was a revisionist power that would not stop expanding of its own volition. The prime minister nonetheless clung to his belief that he could personally dissuade Hitler.


Mearsheimer and Rosato acknowledge that people can be irrational and can be guided by psychological biases. But they argue that personal capriciousness is rarely a problem in foreign policy. “When the stakes are high, as they are in matters of national security,” they write, leaders “have powerful incentives to think in theoretical terms.” It is a simple argument: when put under pressure, people tend to be rational.

But this claim does not hold up to scrutiny. In fact, one could just as easily argue the opposite: when the stakes are high and policymakers are in a crunch, they are more likely to succumb to cognitive shortcuts, emotions, and other nonrational behavior. This is especially true if leaders do not have enough data, or do not have time to sift through and deliberate on data, to make an informed decision. Israel, for example, had human intelligence signaling that Arab states were planning to attack it in 1973. But the Israeli government believed its neighbors were not foolish enough to invade without air superiority. It ignored the evidence and was therefore caught off-guard when Egypt attacked.

Even if leaders do make rational choices when the stakes are high, it does not mean that the country will behave rationally; there is often a difference between what is rational for a state and what is rational for its leaders. A desire to stay in power, for example, may motivate leaders to carry out diversionary wars or other costly actions that undermine their state’s interests. Research shows that Argentina invaded the Falkland Islands in 1982 partly because the military junta, facing rising unpopularity at home, theorized that the invasion would create a rally-around-the-flag effect that would shore up their support. At first, it did, staving off the toppling of the military. But the war was clearly not in the interests of Argentina, even when hopes for a negotiated settlement on the British territory faded and even though the junta mistakenly believed that the United Kingdom would not intervene. Buenos Aires quickly lost, and shortly after, the junta fell.

Mearsheimer and Rosato try to get around the difference between a leader’s interests and the national interest by emphasizing the role of deliberation in decision-making. For a choice to be rational, they write, leaders must both adhere to a reasonable theory and make their decision after consultations. But the authors’ definition of deliberation is flawed. It simply requires that relevant policymakers enter a room and engage in “robust” debate, with the principal decision-maker breaking any deadlocks. But much like the authors’ description of what makes a theory credible, this criterion is both vague and difficult to conclusively determine, particularly in autocracies.

Part of the reason why the book fails to convince readers about whether deliberation is present or absent is because of its methodology. The authors rely on analytical narratives, not primary data, and do no real process tracing (studying a sequence of events over time and ruling out different explanations) for the cases they highlight. As a result, they miss clear evidence that contradicts their conclusions. When readers go through many of the cases the authors cite as deliberative, they will find not careful debates but leaders engaging in performative discussions. Advisers either spin their arguments to make them seem compatible with the leader’s beliefs or they simply endorse what the leader has already decreed. This behavior is especially prevalent in autocracies, in which leaders rarely search for new information or alternative viewpoints. When dictators run their ideas by others, they are not asking for a gut check. They just want to be told that they are right.

Consider, again, Russia’s invasion. Mearsheimer and Rosato conclude that the process by which Moscow decided to invade was deliberative because Sergey Lavrov, Russia’s foreign minister, told reporters that Moscow’s “decision-making mechanism” was “fully employed.” But the facts show that no such process exists. According to reporting by The Washington Post and The New York Times, the U.S. intelligence community believed Putin made the decision to invade Ukraine as early as March 2021, but he told most of his senior advisers about it only days before the invasion. Dissenting ministers and military officers were shown the door, went into exile, or disappeared. Putin did hold a televised meeting with his advisers before starting the invasion, ostensibly to discuss whether Russia should recognize the independence of Ukraine’s easternmost provinces. But it was clearly just for display. “I would like to underline that I did not discuss anything in advance with any of you,” Putin said to his officials, who looked visibly nervous. They then rose, one by one, to endorse their president’s plan. When one deviated from the script by saying that Russia should annex the territories, Putin snapped at him, and the adviser quickly corrected himself.

The decision to invade Ukraine is hardly the only one Mearsheimer and Rosato misrepresent. The authors code the decision to invade Iraq as nondeliberative, arguing that U.S. President George W. Bush “was not deeply involved in the relevant debates inside his administration.” The war in Iraq may well have been irrational; it certainly ended poorly. But entire shelves of books have been written about the botched decision to invade, and they all show that Bush and his team had real conversations. The president met and discussed the idea with his advisers before making up his mind. The U.S. military planned it out well in advance, and it did not hide its intentions from top commanders. And the administration followed a clear theory: that it needed a preventive war to stop Iraq from acquiring nuclear weapons. They did it based on U.S. intelligence, however flawed, and a perception that Saddam was once again deceiving the United States. The reason Mearsheimer and Rosato saw the invasion of Iraq as nonrational, then, is not because it fails to tick off all the items on their checklist. It is because the invasion cannot be explained by their realist theory of the balance of power.


Despite its logical flaws and lack of compelling evidence, Mearsheimer and Rosato’s book holds important value for scholars and policymakers. Their work shows that international politics is an important discipline by proving that leaders rely on theories, both credible and not, to help them make decisions. American leaders’ belief in liberal hegemony drove U.S. foreign policy in the aftermath of the Cold War. Similarly, the West’s decision to expand NATO was driven, in part, by democratic peace theory. The book also proves the importance of process, something overlooked by scholars, in determining whether a leader or a state made a rational decision. Deliberation, the authors write, is essential if policymakers are to avoid engaging in groupthink or falling prey to biases and misperceptions. It is why democratic societies typically have a geopolitical advantage.

Washington should remember this fact as it figures out what to do about Beijing. There is still a debate in the United States about how to handle China, but increasingly, American discourse on China is dominated by aggressive rhetoric. Today, as Jessica Chen Weiss noted in these pages, “individuals feel the need to out-hawk one another” when it comes to Beijing. “The result,” she said, “is groupthink.”

Given how central U.S.-Chinese relations are to modern-day politics, one would expect a book about “how states think” to discuss the subject at length. And yet present-day China is curiously absent from Mearsheimer and Rosato’s work. Presumably, as realists, the authors believe Washington’s growing efforts to check Beijing are logical. But determining whether the current course of action is rational requires knowing whether China is a selectively opportunist or an expansionist power, which, in turn, requires divining Beijing’s intentions. It is something the authors probably do not want to admit. If, after all, the country is purely defensive, then Washington’s aggressive stance makes little sense, and what it needs to do instead is provide reassurance that it will not try to weaken Beijing.

Unfortunately, the reality is that U.S. policymakers and the American intelligence community know little about how Chinese President Xi Jinping actually thinks, making it difficult for them to use theories to predict Beijing’s behavior. Without such information, American leaders must instead default to other measures: selecting evidence consistent with their own views, using mental shortcuts, or relying on U.S. President Joe Biden’s personal impressions of Xi. (The same is true for U.S. assessments of many other autocracies, including Russia.) They must remember that because the stakes are high, great powers and their mercurial leaders may miscalculate or act in irrational and neurotic ways.

The war in Ukraine made this point readily evident. The United States should keep it in mind when it considers Taiwan, as well. As with Russia and Ukraine, Taiwan is bound up in historical grievances that may prevent Xi from thinking clearly before launching an invasion. (Beijing sees the island as a renegade province.) Taiwan is, relatedly, a deeply emotional issue for China’s leaders. In fact, Xi seems to view taking the island as his personal mission. He has declared that seizing Taiwan is essential to “the great rejuvenation of the Chinese nation”—which he wants to achieve before leaving office.

Xi, therefore, is unlikely to apply cold, hard logic when it comes to Taipei. In fact, it may be wishful thinking to expect that he will be rational about Taiwan at all. He is, instead, more likely to decide what to do based on his emotional state of mind, his subjective assessment of China’s strength, or his reading of U.S. resolve. He is likely to disregard evidence suggesting that his goal is unattainable or that the costs of his actions would be astronomical—just as Putin did with Ukraine. That is the real tragedy of great-power politics.

KEREN YARHI-MILO is Dean of the School of International and Public Affairs and Adlai E. Stevenson Professor of International Relations at Columbia University. She is the author of Who Fights for Reputation.

Speak Your Mind