While the United States and Europe have taken significant action to assist Ukraine and pressure Russia, there is increasing pressure to “do more.” With Russian war crimes in plain sight and Ukraine unable to trade with the world to sustain its economy, many feel a moral imperative to intervene and end the war. Simultaneously, continued Ukrainian success on the battlefield has some believing that with more support the Ukrainians can outright defeat Russia, pushing its forces back over the border without making any concessions.
Together, these motivations have led many to advocate for a more muscular U.S. policy. Gérard Araud, the former French ambassador to Washington, recently suggested that NATO inform Moscow that if Odesa is attacked (which it already has been, it should be noted), the alliance will intervene militarily. Michael McFaul tweeted that the best way to avoid future conflict with Russia is to help Ukraine stop Putin now, on the battlefield. Anne Applebaum forcefully argued that the goal in Ukraine should not be a truce or long-term resistance, but an outright Ukrainian victory. Eliot Cohen agrees, arguing that a Ukrainian victory would lead to a new balance in Europe and that “the West must aim to break Russia’s societal will through the grinding up of its army and the devastation of its economy.” Charli Carpenter sounded a clarion call for Western military intervention as the only way to protect civilians from further war crimes. High-ranking Biden administration officials have begun to talk about dramatically weakening Russia rather than just defending Ukraine, and while President Joe Biden says he wants to limit escalation, his messaging is undercut by provision of powerful weapon systems intended to target Russian forward command posts.
We sympathize with what animates these calls for action. They want to protect Ukraine and defeat Russia because that is the right thing to do morally, politically, and geo-strategically. Letting Ukraine burn as Russia bleeds itself white through attrition and economic crisis is indeed monstrous, and a Ukrainian victory over Russia would garner significant long-term upsides. However, we should be clear-eyed about the risks of escalation as we seek that victory. While militaries on all sides consider escalation to be a tactical or strategic choice to coerce or defeat their opponents, it also represents an undesirable risk inherent in continued conflict, one that is not entirely under anyone’s control. It would behoove anyone advocating escalation to think seriously about the risks they would be running, and to do so we must first make them explicit. One of us has written a book on trends in the initiation and escalation of wars and a summary of the current state of knowledge, and we are currently collaborating on a new project on escalation. Two lessons stand out starkly from our research: War escalation is extremely unpredictable, and most people don’t appreciate just how easily and quickly wars can escalate to shocking levels of lethality.
The first point to understand is that wars are not like most ordinary phenomena, in that their size doesn’t cluster closely around an average value in the way that many measures do. Consider the following: American men average about 5’10” in height and American women average 5’5”, with almost all people falling within nine inches of those averages. This information tells us quite a bit about the height of the population. For example, at 8’11”, Robert Wadlow, the tallest man known to have ever lived, is quite extraordinary: One in a trillion is a wild overestimation of the odds that anyone would ever be that tall.
The distribution of the size of wars, by contrast, does not conform to a tidy, bell-shaped curve. Most wars remain relatively small, but some get unbelievably large, and the magnitude of the difference is difficult to comprehend. As a result, the size of an “average” war tells us surprisingly little about the size of wars more generally. A war 50 percent larger than the mean, far from being extraordinary, is unremarkable and would not get a second glance. People might even breathe a sigh of relief that the war remained so small.
How we think about probabilities and the fatality counts in the deadliest wars are difficult to grapple with in the abstract, so instead we will use a more concrete analogy. Imagine a small town, with 95 buildings of varying heights. Each of the interstate wars documented by the Correlates of War project since 1812 represents a building, and the lethality of the war, measured in battle deaths, is proportional to the building’s height. The smallest of the 95 buildings, those representing conflicts like the 1823 Franco-Spanish War and the Falkland Islands War of 1982, will be a single story tall. The wars they represent barely cross the thousand-death threshold for inclusion in the data. Ten of the 95 buildings are about this tall.
What about the next 10? The tallest of those, representing the Assam War of 1962, is about two stories tall. And the 10 after that? The tallest of that group, representing the Boxer Rebellion of 1900, is three stories, maybe the height of a small apartment building. That doesn’t seem too bad, so far: We’ve covered about a third of the wars, and the worst of the group is only about three times as lethal as the least deadly.
The numbers accelerate from there. The median war, number 47 (the 1948 Arab-Israeli War), is eight stories tall, which is large but a common enough building height in many cities. True skyscrapers, 40 stories tall, start at war number 71. Such a height is still well within the realm of normal. Beyond that point, however, our small town quickly leaves the realm of human construction. The tallest building on Earth, the Burj Kalifa at 200 stories, does not quite crack the top 10, and from there the heights explode into absurdity. The Korean War is represented by a building that’s 900 stories tall. World War I’s building is around 8,500 stories in height, and World War II reaches well above the stratosphere at 16,600 stories tall — about the height of eight Mount Everests stacked on top of one another.
Is there any evidence that the lethality of war is in decline? Decision-makers and the public might find solace in Steven Pinker’s claim that warfare in general has been in decline for some time, and that as a result escalation is far less likely than it used to be. Rigorous data analysis paints a less rosy picture. While some scholars do find evidence of a slight decline in the lethality of interstate wars around 1950, most do not. Even in the best case, though, wars can still become horrifically lethal.
With this analogy in mind, we can see why statistics like mean and median are not particularly useful and may even be misleading. Most wars will either be far less lethal or far more lethal than the median. The bottom 50 percent of wars have an average of about 2,900 battle deaths, while the top 50 percent have an average of 653,000, and it is effectively a coin-flip which half any given war will end up in. In Ukraine, after three months and with no end in sight, Western analysts estimate at least 20,000 fatalities, putting this war well into the top half of conflicts.
One might argue that the statistical evidence tells us little about any given war, and we might have a much better idea as to the escalation risks in a given conflict than it might appear by looking at the entire dataset. No one thought the Falkland Islands War was going to engulf all of Europe. Skeptical American analysts were more concerned with the political ramifications for Margaret Thatcher should the counter-invasion fail and spent little, if any, time considering broader military implications. But did they know, or did they simply assume?
The Bush administration assumed that the Iraq War would be quick and produce few casualties. The Biden administration assumed the Afghan government could make the Taliban fight for every mile instead of collapsing before U.S. forces even finished pulling out. Saddam Hussein was convinced the United States would not intervene in his invasion of Kuwait. Kaiser Wilhelm II is often quoted as having said, in August 1914, that his soldiers “would be home before the leaves fall from the trees.” Russia thought its invasion of Ukraine would be a simple operation against a much weaker foe. It should be no surprise that the outcome of a war is hard to predict even in the most lopsided of circumstances: Consider how many wars have been lost by the aggressor!
One might be tempted to think that our estimates have improved over time, but we already have evidence of highly unpredictable developments in the Ukraine war. Russia’s shocking ineptitude, Ukraine’s surprising resistance, and the extent to which the entire world responded with unprecedented economic sanctions on Russia all confounded expectations. At the same time, Kremlin spokesman Dmitry Peskov’s indication that Russia would use nuclear weapons in the event of an “existential threat,” the arrival of U.S. troops in Poland, and President Joe Biden’s statement that Vladimir Putin “cannot remain in power” have all contributed to growing uncertainty regarding the course and outcome of the war.
Indeed, predicting escalation may be largely impossible a priori because so much depends on individual, idiosyncratic decisions. Key events in wars — like President Harry Truman deciding to order U.S. forces to cross the 38th parallel during the Korean War or Adolf Hitler ordering the capture of Kyiv over Moscow against the advice of German High Command in World War II — are difficult if not impossible to anticipate, and they can make the difference between thousands versus millions of fatalities. Especially considering the statements noted above, it’s not at all clear why we should believe that political and military leaders are better at limiting escalation now than they were in the past.
Others might argue that the shadow of nuclear weapons inhibits escalation, so the truly catastrophic outcomes are less likely than they used to be. In response, we would first note that Ukraine is not a nuclear-armed state, so nuclear-deterrence dynamics do not apply, and some have even suggested that Russia might use low-yield nuclear weapons to achieve some semblance of victory in Ukraine. But even if the United States were to get more directly involved, the logic of nuclear weapons is not as clear cut and de-escalatory as many believe it to be. If deterrence is highly effective, it creates what is known as the stability-instability paradox: Because all states are deterred from using nuclear weapons, warfare becomes more pervasive at lower levels of escalation. Russia and NATO could fight for years in Ukraine, safe in the knowledge that no one is likely to use nuclear weapons. There is evidence from the Korean War that U.S. and Soviet aircraft were shooting at each other, and while neither superpower used nuclear weapons, no one would call that war small. Finally, the argument that nuclear weapons pacify assumes that they would only be used intentionally, while we know that the risk of a nuclear exchange arising from an accident is far from trivial.
What does all this tell us about the war in Ukraine? Prior to the war, our best guess would be that the building that represents it would stand as tall as one drawn from the town of 95 that we described earlier. Not even four months in, we are already well into the upper 50 percent of that category. Where do we expect it to stop? Many analysts thought the war was unlikely to begin with. Once it began, the common wisdom was that Ukraine could not possibly hold off Russia for long. But the war is still ongoing and casualties continue to mount. In the deadlier half of wars, 10 percent have over a million battle deaths — what is stopping the Ukraine conflict from reaching that number?
In The Prince, Machiavelli wrote that despite mercy and generosity being virtues worthy of emulation, a political leader must be aware that doing the virtuous thing can lead to worse outcomes, both for yourself and for those you intended to help. Many across the United States and Europe, the authors included, have a deep desire to help Ukraine, a nation that is suffering a monstrous injustice. Doing all they can to end the conflict is the virtuous choice, and we are certainly not arguing against a robust Western response. We are, however, concerned with calls to seek total victory without a realistic analysis of the costs and risks involved. A call to do more without thinking about the risks leaves the door open for “virtuous escalation.” We want to underscore the unpredictability and danger of this path. If this conflict does escalate, if Russia decides reducing cities to rubble is its only path to victory, or if NATO countries decide to intervene directly, the potential for catastrophe is far higher than most people realize, and vastly too high for comfort.
Michael Lopate is a doctoral candidate in political science at The Ohio State University.
Bear Braumoeller is a fellow of the American Association for the Advancement of Science and a professor of political science at The Ohio State University.
Image: Ukrainian Ministry of Defense via ArmyInform