The Trouble With Allies
America Needs a Playbook for Difficult Friends
The war in Ukraine is being waged with a host of advanced technologies, from remotely operated drones to space-based surveillance, precision weapons, hypersonic missiles, handheld jammers, artificial intelligence, networked communications, and more. Many argue that this array is transforming warfare, with omnipresent surveillance combining with newly lethal weapons to make legacy systems such as the tank obsolete and to make traditional methods such as large-scale offensive action impractical. As the military analyst David Johnson has put it, “What I believe we are witnessing is a pivotal moment in military history: the reascendance of the defense as the decisive form of war.” Drones, artificial intelligence, and rapid adaptation of commercial technologies in Ukraine are creating “a genuine military revolution,” according to military strategist T. X. Hammes. Former Google chief executive and Pentagon adviser Eric Schmidt has argued that Ukraine is showing that “the future of war will be dictated and waged by drones.”
But in many ways, this war seems quite familiar. It features foot soldiers slogging through muddy trenches in scenes that look more like World War I than Star Wars. Its battlegrounds are littered with minefields that resemble those from World War II and feature moonscapes of shell holes that could be mistaken for Flanders in 1917. Conventional artillery has fired millions of unguided shells, so many as to strain the production capacity of the industrial bases in Russia and the West. Images of code writers developing military software accompany scenes of factory floors turning out mass conventional munitions that lack only Rosie the Riveter to pass for images from 1943.
This raises the question of how different this war truly is. How can such cutting-edge technology coexist with such echoes of the distant past? The answer is that although the tools in Ukraine are sometimes new, the results they produce are mostly not. Armies adapt to new threats, and the countermeasures that both sides have adopted in Ukraine have dramatically reduced the net effects of new weapons and equipment, resulting in a war that in many ways looks more like a conflict from the past than one from an imagined high-tech future. U.S. defense planners should understand that the war in Ukraine does not portend a “revolution in military affairs” of the kind that has often been predicted but somehow never quite arrives. Policymakers and analysts should closely study what is happening on the ground in Ukraine, but they should not expect their findings to produce transformational change in U.S. military strategy. Instead, as has often been the case in the past, the best path forward will involve incremental adaptations, not tectonic shifts.
One way to assess the net results of the use of new weapons in Ukraine is to look at the casualties they have inflicted. Those who see a military revolution in Ukraine usually argue that new surveillance techniques, such as coupling drones with precision weapons, have made the modern battlefield radically more lethal. Yet the realized lethality (as opposed to the potential lethality) of Russian and Ukrainian weapons in this war is little different from that seen in previous wars, and in some cases it is actually lower.
Consider, for example, tank losses. Many revolutionists see heavy tank casualties in Ukraine as the key indicator for the tank’s looming obsolescence in the face of newly lethal precision antitank weapons. And tank losses in Ukraine have certainly been heavy: Russia and Ukraine have each lost more than half the tanks with which they entered the war. At the time of the invasion, Russia had about 3,400 tanks in active service. But in the first 350 days of the war, it lost somewhere between 1,688 (the number verified photographically by the open-source organization Oryx) and 3,253 (the number claimed by the Ukrainian Defense Ministry), for a loss rate of somewhere between 50 percent and 96 percent. Ukraine fielded about 900 tanks at the time of the invasion and lost at least 459 (the Oryx figure) in the first 350 days, for a loss rate of at least 51 percent. Both countries have either built or been given additional vehicles as replacements. Russia, especially, has extensive reserves of older vehicles that have been pressed into service. Damaged tanks can sometimes be repaired and returned to battle. So even though the armor fleets in the field have thus not shrunk massively, it is clear that many tanks have been lost in battle.
Yet these are not unusually heavy loss rates for major warfare. In just four days during the Battle of Amiens in 1918, the United Kingdom lost 98 percent of the tanks it had when the fighting began. In 1943, the loss rate for German tanks was 113 percent: Germany lost more tanks than it owned at the beginning of the year. In 1944, Germany lost 122 percent of the tanks with which it started the year. The Soviet Union’s loss rates for tanks in 1943 and 1944 were nearly as high, at 109 percent and 80 percent, respectively. And in a single battle in Normandy (Operation Goodwood, in July 1944), the United Kingdom lost more than 30 percent of all its armor on the continent in just three days of fighting. Few, however, argued that the tank was obsolete in 1918 or 1944.
In war, new technology matters, but adaptations dramatically dampen its effects.
Or consider aircraft losses. Some have suggested that modern antiaircraft missiles are so lethal to traditional piloted aircraft that these, too, are headed for the ash heap of history. And like tanks, aircraft have suffered heavy losses in Ukraine: in almost a year and a half of fighting, the Ukrainian air force has lost at least 68 aircraft, or more than a third of Ukraine’s prewar fleet; the Russian air force has lost more than 80 of its preinvasion inventory of 2,204 military aircraft. Yet this level of destruction is hardly unprecedented. In 1917, the life expectancy of a new British pilot was just 11 days. In 1943, the German Luftwaffe lost 251 percent of the aircraft it had at the beginning of the year. Its loss rate for 1944 was even higher: in the first half of the year alone, it lost 146 percent of its January strength. The Soviet loss rate for aircraft was 77 percent in 1943 and 66 percent in 1944. Yet few argued that the piloted airplane was obsolete in 1917 or 1943.
Or consider artillery. Since at least 1914, artillery has inflicted more casualties in major wars than any other weapon. And today, some observers believe that as many as 80 to 90 percent of Ukrainian casualties have been caused by artillery fire. Many accounts of the fighting in Ukraine feature scenes of the two armies using drones to find enemy targets and then using networked communications to quickly relay the information for precision engagement by guided artillery. Of course, not all artillery in Ukraine is precision guided; most rounds fired by either side are relatively old-fashioned. But the teaming of these unguided rounds with new drone reconnaissance and rapid-targeting systems is often described as a new and profound development in Ukraine. If one assumes, however, that 85 percent of Russian casualties are caused by Ukrainian artillery, that Russia suffered as many as 146,820 casualties in the first year of the invasion (the Ukrainian Defense Ministry’s figure), and that Ukraine fired a total of around 1.65 million rounds of artillery in the first year (as the Brookings Institution has estimated), then drones and the mix of guided and unguided artillery in the Ukrainian army inflicted, on average, about eight Russian casualties per hundred rounds fired in the first year of the invasion.
That rate exceeds the world war rates, but not by much. The historian Trevor Dupuy estimated that in World War II, around 50 percent of casualties were caused by artillery, which means that on average, it inflicted about three casualties per hundred rounds fired. In World War I, the figure was about two soldiers wounded or killed per hundred rounds fired. Casualties per hundred rounds has thus grown since 1914 but at a steady, almost linear annual rate of around an additional 0.05 casualties per hundred rounds. Artillery in Ukraine looks more like an incremental extension of long-standing trends than a revolutionary departure from the past.
Of course, casualty infliction is only one element of warfare—armies also seek to take and hold ground. And many revolutionists think that new equipment has changed the patterns of advance and retreat in Ukraine relative to historical experience. In this view, today’s newly lethal weapons have made offensive maneuver prohibitively costly, inaugurating a new era of defense dominance in which ground is much harder for attackers to take than in previous eras of warfare.
Yet the Ukrainian war to date has been far from a uniform defensive stalemate. Some attacks have indeed failed to gain ground or have done so only at great cost. The Russian offensive at Bakhmut eventually succeeded, but only after ten months of fighting and a casualty toll of perhaps 60,000 to 100,000 Russian soldiers. Russian offensives in the spring of 2022 gained little ground, and the Russian attack on Mariupol in southern Ukraine in February lasted almost three months before an outnumbered defense was overwhelmed and the Russians captured the city. Ukraine’s counteroffensive in Kherson began with weeks of slow, expensive attrition warfare in August and September 2022.
But other attacks have moved much farther and faster. Russia’s initial invasion in February 2022 was poorly executed in many ways, yet it gained over 42,000 square miles of ground in less than a month. Ukraine’s Kyiv counteroffensive then retook over 19,000 square miles in March and early April. Ukraine’s Kherson counteroffensive in August 2022 eventually gained almost 470 square miles, and its Kharkiv counteroffensive in September 2022 retook 2,300 square miles. The war has thus presented a mix of successful offense and successful defense, not a pattern of consistent offensive frustration. And all this—both the breakthroughs and the stalemates—has occurred in the face of new weapons and equipment. Conversely, older legacy systems such as tanks played prominent roles in both the offensive successes and failures. These variations are hard to square with any technologically determined new epoch in war.
This, too, is an important echo of the past. The popular imagination sees World War I as a technologically determined defensive stalemate and World War II as a war of offensive maneuver unleashed by the tank, the airplane, and the radio. This perception encourages observers today to look for another such epochal shift in Ukraine. But in reality, neither world war followed a uniform, technologically determined pattern: the same technologies produced both offensive actions that took ground quickly and defensive stalemates in which battle lines barely moved. Both world wars displayed wide variations in offensive success that correlate poorly with variations in equipment.
In World War I, for example, the trench stalemate of 1915–17 dominates the popular image of the conflict. Yet the initial German invasion of Belgium and France in 1914 advanced more than 200 miles in four weeks in spite of modern machine guns and artillery. The German spring offensives of 1918 broke through Allied lines on the western front three times in succession and took nearly 4,000 square miles of ground using virtually no tanks; the subsequent Allied Hundred Days Offensive then drove the Germans back over open ground on a roughly 180-mile front, capturing more than 9,500 square miles of German-held territory in the process. In fact, 1918 as a whole saw more than 12,500 square miles change hands in some eight months of fighting. World War I also saw many unsuccessful offensives, but stalemate is not the whole story.
Conversely, the popular image of World War II is dominated by tanks and blitzkrieg offensives. And certainly, there were plenty of tank-equipped offensive breakthroughs, whether during the German invasions of France in 1940 or of the Soviet Union in 1941, or during the American offensive in Operation Cobra in Normandy in 1944. But the war also saw some of the most costly offensive failures in military history. The 1943 Battle of Kursk in Russia cost the German attackers more than 160,000 casualties and destroyed more than 700 German armored vehicles but failed to break through Soviet defenses. The failed British offensive at Goodwood in 1944 has been described by the historian Alexander McKee as “the death ride of the armored divisions.” Repeated Allied attacks on the Gothic Line in Italy in 1944 and 1945 produced failure after failure at the cost of more than 40,000 Allied casualties. Like World War I, World War II involved a great deal of variance in outcomes: it was not a simple, uniform story of offensive success. And in Ukraine, both the war’s offensive successes and its defensive stalemates have occurred in the face of drones, precision weapons, hypersonic missiles, and space-based surveillance. In none of these wars have the tools predetermined the results.
The reason technological advances are not more determinative in war is that they are only a part of what shapes outcomes. How combatants use their technology and adapt to their enemy’s equipment is at least as important and often more so.
This has been true since the dawn of the modern era. For over a century, weapons have been lethal enough that armies who mass exposed forces in the open have suffered annihilating loss rates. As early as 1914, as few as four 75-millimeter field guns could saturate an area the size of a football field with lethal shell fragments in a single volley. A French version of this—the 1897 Model Soixante-Quinze—could do this 15 times in one minute with sufficient ammunition. An army that simply charged defenses armed with such weapons would be committing suicide. Even heavily armored tanks can be destroyed en masse by modern antitank weapons if they operate this way: the British tanks that charged German antitank guns at Goodwood and the German tanks that charged Soviet antitank guns at Kursk offer vivid examples.
As a result, most armies adapt in the face of modern firepower. Sometimes this means deploying new tools to counter enemy technology: antitank guns encourage the development of tanks that use heavier armor, which encourages the use of bigger antitank guns, then still heavier armor, and so on. Multiple cycles of these technological measure-countermeasure races have already occurred during the war in Ukraine. For example, expensive, sophisticated drones were countered by guided antiaircraft missiles, encouraging combatants to deploy simpler, cheaper, and more numerous drones, which have been countered by simpler, cheaper antiaircraft artillery and hand-held jammers, and so on. The long-range guided HIMARS missile systems the United States provided to Ukraine in June 2022 use GPS signals for guidance; the Russians now routinely jam the signals, which has dramatically reduced the accuracy of the missiles. Technical countermeasures are ubiquitous in war, and they quickly limit the performance of many new weapons.
Casualty rates in Ukraine have not been unusually high by historical standards.
But the most important adaptations are often not technological but operational and tactical. They involve changes in the way armies use the tools at their disposal. Over a century ago, armies developed tactics that reduced their exposure to enemy fire by exploiting dispersion, cover, concealment, and suppressive fire. The complex topography of the earth’s surface provides many opportunities for cover (impenetrable obstacles such as hillsides) and concealment (opaque obstacles such as foliage) but only if armies disperse by breaking large, massed formations into smaller subunits that can fit into the patches of forest, the interiors of buildings, and the irregular folds in the earth that offer the greatest opportunities to escape hostile fire.
For centuries, armies have augmented such natural cover by digging trenches, bunkers, and fieldworks. And by 1917, armies discovered that by combining suppressive fire with sprints from cover to cover, they could reduce casualties during brief periods of exposure to gunfire and survive forward movement on the battlefield. Attackers learned to combine infantry, tanks, artillery, engineers, aircraft, and more to enable this “fire and movement” style of fighting: infantry who could see concealed enemies, tanks that could bring firepower forward to destroy the enemies, artillery to provide suppressive fire to cover the attackers’ movement, engineers who could clear mines, and aircraft to strike from above and protect troops from enemy airplanes. Defenders learned to distribute dug-in forces into depth to delay offensive advances by such attackers while rearward reserves maneuvered to reinforce defenses at the threatened point. These methods were what broke the trench stalemate in 1918, and continued extensions of these concepts have been in use ever since.
Air forces, unlike ground armies, cannot dig in for cover and still fly combat missions. But air forces can avoid enemy fire in other ways. They can restrict aircraft to altitudes and flight paths designed to evade enemy air defenses. They can coordinate their operations with ground forces or other aircraft in ways that suppress the fire of enemy air defenses during brief periods of aerial exposure. They can move between multiple runways to reduce vulnerability to preemptive attack on the ground. And air forces, too, can reduce their formation density when in flight; the massed thousand-bomber raids of World War II are now a thing of the past. As antiaircraft weapons have grown more lethal, air forces, like ground forces, have increasingly adapted to reduce their vulnerability.
These methods can be extremely effective when used properly. Unhindered by suppressive fire, a single BGM-71 guided antitank missile crew can destroy seven tanks at ranges of over one and a half miles in just five minutes. If forced by suppressive fire to take cover and relocate between shots, its kill rate can be reduced to one tank or fewer. A 100-soldier infantry company massed in the open on a 200-yard front can be wiped out by a single battalion volley from hostile artillery; dispersed over a 1,000-yard front with a depth of 200 yards, the same unit might suffer less than ten percent losses. If the unit has even partially concealed itself and the artillery misses the formation’s center, losses might be reduced to as little as five percent.
Dispersion can also make targets unworthy of engagement. A $100,000 guided 155-millimeter artillery shell is too expensive to fire at a two-man target even if a drone locates the soldiers’ foxhole perfectly. When soldiers spread out on the battlefield, it makes more economic sense to try to hit them with cheaper, unguided rounds. But that has drawbacks, too: artillery risks detection every time it fires, so to fire multiple unguided rounds at a single small target makes the shooter vulnerable to counterfire in exchange for a limited payoff. Aircraft that could be shot down quickly if they overfly enemy air defenses are far less vulnerable if they fly below enemy radar while firing from behind friendly lines.
Such methods can be challenging to implement correctly, however. Most armies can manage dispersion, cover, and concealment at the small-unit level, if only by digging in. This reduces casualty rates, but it also limits what an army can accomplish if this is all it can do. Air forces can restrict themselves to low altitudes in safe rear areas, but this limits their contribution to the fighting.
The Ukraine war is more evolutionary than revolutionary.
To take ground on a large scale and prevent the enemy from doing so requires forces to coordinate deep defenses with mobile reserves; to combine infantry, armor, artillery, engineers, air defense, and more, on the offensive; and to integrate fire and movement on a large scale—and these are much harder tasks. Some militaries have mastered these skills; others have not. When defenses are deep, prepared, and backed by mobile reserves, they have repeatedly proved very hard to break through—regardless of whether the attackers have tanks or precision-guided weapons. But when defenses are shallow, poorly prepared, or inadequately supported by reserves, attackers that can implement combined arms and fire-and-movement methods on a large scale have been able to break through and take ground quickly—even without tanks and even against precision-guided weapons. Think, for example, of the German infantry breakthroughs in 1918 or the Ukrainian gains in the face of Russian drones and precision weapons at Kharkiv in 2022.
New technology does matter, but the adaptations that armies have increasingly adopted since 1917 dramatically dampen its effects on outcomes. Precision weapons that are devastating on the proving ground or against exposed, massed targets yield much lower casualty rates against dispersed, concealed forces. And as weapons have grown more lethal over time, armies’ adaptations have kept pace accordingly. In the nineteenth century, for example, armies typically massed their forces to battlefield concentrations of approximately 2,500 to 25,000 troops per square mile. By 1918, those figures had fallen by a factor of ten. By 1945, they had fallen by another factor of ten. By the time of the 1991 Gulf War, a force the size of Napoleon’s at Waterloo would be spread over an area about 3,000 times as large as the one the French army occupied in 1815.
This combination of ever more lethal technology but ever more dispersed and concealed targets has produced far less net change in realized outcomes over time than one would expect by looking only at the weapons and not at their interaction with human behavior. Better tools always help, and Western assistance to Ukraine has been critical in enabling Ukraine to cope with a numerically superior Russian army. But the actual battlefield impact of technology is shaped powerfully by its users’ behavior, and in Ukraine, as in the last century of great-power warfare, that behavior has usually been a better predictor of outcomes than the tools themselves.
Although the Ukraine war has seen plenty of new equipment, its use has not yet brought transformational results. Casualty rates in Ukraine have not been unusually high by historical standards. Attackers in Ukraine have sometimes been able to advance and sometimes not; there has been no pattern of uniform defensive stalemate. This is because those fighting in Ukraine have responded to newly lethal weapons just as their predecessors did: by adapting with a combination of technical countermeasures and further extensions of centurylong trends toward increased dispersion, cover, concealment, and suppressive fires that have reduced both sides’ exposure to hostile firepower.
Losses are still heavy, as they have often been in major wars, but loss rates in Ukraine have not prevented major ground gains in offensives at Kyiv, Kharkiv, and Kherson. Success on the attack is hard, and it normally requires a combination of offensive skill and defensive error, as it has for generations. In Ukraine, as in the past, when skilled attackers have struck shallow, ill-prepared defenses that have inadequate reserves or logistical support, they have broken through. But in Ukraine, as in the past, when this combination has been absent, the result has usually been stalemate. This is not the result of drones or access to broadband Internet, and it is not anything transformational. It is a marginal extension of long-standing trends and relationships between technology and human adaptation.
If the Ukraine war is more evolutionary than revolutionary, what does that mean for defense planning and policy? Should Western countries abandon the pursuit of modern weapons and equipment and freeze doctrine development? Of course not. Evolutionary change is still change, and the whole point of adaptation is that militaries must adopt new methods and equipment. A 1916 tank would stand little chance on the battlefield of 2023—the stable attrition rates of warfare since World War I are products of continuous, two-sided adaptation in which combatants have always worked to avoid allowing rivals to gain much of an edge.
The crux of the revolution thesis, however, is an argument about the pace and nature of needed change. If warfare is being revolutionized, then the traditional, incremental updating of ideas and equipment is insufficient, and something more radical is needed. Tanks, for example, should be mothballed, not modernized. Robotic systems should quickly replace humans. Preparation for large-scale offensive action should be replaced with a heavy emphasis on defense and injunctions against attack in all but exceptional conditions.
The war in Ukraine, to date, offers little support for such ideas. It is still in progress, evidence is imperfect, and the future course of the fighting could be different. But so far, few of the observable outcomes are consistent with an expectation of revolutionary change in results or a need for radical reequipment or doctrinal transmogrification. This, too, is consistent with previous experience. It has been almost 110 years since the tank was introduced in 1916. Some have argued that the tank is obsolete because of technological improvements in antitank weapons. This argument has been commonplace for over 50 years, or almost half the entire history of the tank. Yet in 2023, both sides in Ukraine continue to rely on tanks and are doing everything they can to get their hands on more of them.
The U.S. Air Force redesigned itself in the 1950s around an assumption that the nuclear revolution had replaced conventional warfare and that future aircraft would be needed primarily for nuclear weapons delivery. The subsequent nonnuclear war in Vietnam was waged with an air force that was designed for a transformational future that never arrived and that proved ill suited for the war it actually fought. Or consider U.S. Army doctrine. This was reshaped in 1976 to reflect a view that precision weapons had made offensive action prohibitively costly under most conditions, yielding a new emphasis on mostly static defense from prepared positions. This “Active Defense” doctrine was highly original but ill conceived and had to be abandoned in favor of the more orthodox “AirLand Battle” concept that the U.S. military used for successful offensive action in Kuwait in 1991.
Calls for revolution and transformation have been commonplace in the defense debate in the generations after World War II. They have mostly not fared well in light of observed experience in that time. After a year and a half of war in Ukraine, there is no reason to think that this time they will be proved right.