Russia’s Repeat Failures
Moscow’s New Strategy in Ukraine Is Just as Bad as the Old One
Russia’s botched invasion of Ukraine, along with unexpectedly dogged Ukrainian resistance, has sparked a debate about whether the current ways of measuring military capabilities are flawed. Indeed, U.S. intelligence analysts have been shocked by the rapid disintegration of the Afghan military in 2021, the collapse of Iraqi units in the face of the so-called Islamic State (or ISIS) in 2014, and Ethiopia’s shambolic performance against Tigrayan rebels in the last two years. U.S. analysts correctly predicted that Russia would invade in February, but their assumption that Ukraine would fall in a matter of days has only fueled the perception that something is amiss in how the intelligence community thinks about and measures military power. Prodded by angry lawmakers, the U.S. intelligence community has launched a sweeping internal review of how it assesses foreign military power amid its missteps in Afghanistan, Ukraine, and elsewhere.
These intelligence failures have a common taproot: they neglect the nonmaterial drivers of military power. Most important, they overlook how social inequalities within armies shape battlefield performance. Current models of net assessment, the framework used to measure the relative strength of a military, privilege quantifiable indicators, such as numbers of tanks and soldiers. These metrics have proved to be poor predictors of how armies fight and whether they win wars. Afghan forces outnumbered Taliban fighters but could not translate their material advantages into victory. Even the largest and best-equipped fighting forces can fail on the battlefield if they lack the cohesion and determination of their foes. Armies, after all, are not just the sum of their personnel and their equipment; they can be riven with ethnic, racial, class, and other social divisions that invariably shape their capacity to fight. The intelligence community ignores these dynamics at its peril.
States spend fortunes sizing one another up. Better insights can tip the scales, but net assessment is a fraught business. States practice elaborate deception, seeking to hide their true military strength from prying eyes. Some capabilities can be properly appraised only when combat begins. Context, too, can shape battlefield dynamics in unpredictable ways. Analysts and organizations alike are fallible, forced to make judgments with incomplete information. Cognitive biases and internal politics can distort assessments. In the face of these overlapping problems, many analysts choose to be cautious and err on the side of exaggerating enemy capabilities to avoid battlefield surprises.
This conservative bias leads analysts to privilege quantitative measures of military capabilities. They dwell on so-called objective factors, including the number of soldiers under arms, annual defense spending, per capita income, and the acquisition of new technologies, to create estimates of relative military power. Much effort is also devoted to divining the capabilities of adversaries from published doctrines. Analysts estimate quantitative indicators of military strength using a sprawling infrastructure of satellites, open-source data collection, spies, and electronic eavesdropping.
Such efforts have scored successes—Russia’s plan to invade Ukraine was published even before its tanks crossed the border. But analysts using objective measures rely on three assumptions that create dangerous blind spots.
Armies reflect their societies and are subject to the same social divisions.
First, analysts should not assume that material strength translates into battlefield victories. In his 2004 book, Military Power, the international relations scholar Stephen Biddle found that material factors such as GDP, population, and military spending have had, at best, a weak connection to victory in wars since 1900. In my own work examining 252 wars since 1800, I found virtually no statistical correlation between the size of opposing armies and battlefield outcomes, including relative casualties, the likelihood of desertion and defection, or who eventually won. Given these results, it is perhaps unsurprising that much of the current debate among scholars over the sources of military effectiveness focuses on nonmaterial factors such as regime type, ideology, and culture, not relative material strength. Yet public debates outside the academy, including ones on the outcome of the war in Ukraine, still privilege traditional indicators of relative material strength.
Second, today’s net assessment assumes that armies are efficient killing machines. This, too, is mistaken. Nearly all armies reflect the societal divisions—whether ethnic, racial, ideological, or class based—of the broader country. Left unmanaged, these divisions create friction that saps military strength. In some cases, armies draw heavily from marginalized groups to spare the regime’s supporters from the costs of war. Some armies maintain cohesion through intimidation, forcing their own soldiers to fight. These armies might appear formidable on paper but typically lurch into battle hobbled by discord. The alienation of the Sunnis in Iraq inflamed sectarian tensions within the Iraqi army. That led many Sunni soldiers to desert in 2014, allowing ISIS to take Mosul with ease despite being outnumbered and outgunned.
Third, analysts assume that material indicators are more objective, leading them to downplay intangible factors, such as soldier morale or unit cohesion. And when analysts do consider these intangibles, they tend to treat them as uniform across an entire army. Doing so, however, omits the human side of war. Soldiers are not homogenous. Some might rally to the colors. But others may be driven by more mercenary motives. Still others may be forced to fight on behalf of a government they despise. Stitching together a coherent narrative that inspires all soldiers to fight equally hard is beyond the power of many governments. Russia is trying to form volunteer battalions to fight in Ukraine from prisoners, non-Russian Muslim populations, pro-Russian Ukrainian soldiers, and draft dodgers. Motivating soldiers in this patchwork quilt, especially given their limited training time, will be difficult. Analysts must recognize that for many armies, maintaining order and discipline is as challenging as fighting enemy forces.
Improving net assessment requires opening the black box of armies to explore the nature and severity of the inequality that lies within. The intuition here is simple: armies reflect their societies and are subject to the same social divisions. Analysts need to gather two pieces of information to estimate a military’s level of inequality. They first need to map the size and composition of the social groups that make up the army. Historically, ethnicity has been a powerful source of potential division. Since 1800, the average army has gone into battle with five different ethnic groups in its ranks. In 1812, Napoleon’s Grande Armée marched on Moscow with enlisted soldiers from at least nine different nationalities, with French soldiers in the minority. During World War II, the Soviet Union’s Red Army, often imagined as ethnically Russian, cobbled together rifle divisions with soldiers from 28 different ethnic groups. Analysts next need to know how the state treats each group within its military. Some groups may be granted full rights and opportunities. Others, however, could be considered second-class citizens or, worse, could be subject to violent political oppression. France and the United Kingdom routinely levied colonial armies from marginalized populations to fight wars at home and abroad. The more an army is drawn from marginalized groups, and the more harshly those groups are treated by the state, the more unequal the force becomes—and the worse it performs on the battlefield.
Diversity is not destiny. Battlefield performance is driven by how the state treats each social group, not by the total number of groups within an army. Inclusive armies, ones that recruit among groups that enjoy full citizenship, can rise above crippling social divisions. For these armies, patriotism can be a powerful motivator. So, too, can be the promise of greater inclusion. During both World War I and World War II, Black Americans fought a double war: one against external foes and one for equality when the war ended. Even authoritarian powers, including Joseph Stalin’s Soviet Union, dangled greater inclusion to repressed minorities in Central Asia to bolster recruitment. But history carries a weight. Wartime adjustments to increase inclusion can improve battlefield performance on the margins but cannot overcome the legacy of the state’s oppression before the war.
States cannot easily hide the structures of their societies or the fact of repression.
Unlike standard indicators of national power, military inequality offers a clear connection to how and whether armies will fight. The greater the share of soldiers from marginalized groups in a given army, the worse that army will perform on the battlefield. Prewar inequalities wend themselves through the military like a poison, corroding combat power before a shot is even fired. Soldiers who have faced discrimination at home can be less willing to fight on the field. Inequality also sows distrust between privileged and marginalized groups, eroding bonds between soldiers and, in many cases, between officers and the rank and file. Past injustices create common cause among targeted groups to resist military authorities. The army of the Austro-Hungarian Empire unraveled during World War I as repressed Magyar and Slavic soldiers chose desertion over duty, fleeing the battlefield.
Unequal armies tend to be coercive toward their troops. Fearing mass indiscipline, such armies will adopt rigid command structures. They threaten or use violence to manufacture cohesion. Left without convincing ideological appeals, highly unequal armies may encourage soldiers to pillage and rape as rewards for continued service. Inequality also forces commanders to simplify their tactics. Sophisticated combined arms operations are too complicated for halfhearted soldiers to execute. Commanders are left facing a war within the war as they scramble to impose discipline and muster sufficient combat power to fight. In many cases, they fail. Over the past 200 years, highly unequal armies have suffered higher casualty rates and more frequent outbreaks of mass desertion than their more inclusive counterparts. The Mahdist Army, fighting in Sudan, was bitterly divided along ethnic and tribal lines and suffered one of the most lopsided defeats in recorded history. At Omdurman in 1898, it lost over 12,000 soldiers while killing only 48 of the opposing Anglo-Egyptian forces.
Military inequality offers several advantages as a guide to battlefield performance. For one, it is more visible. States cannot easily hide the structures of their societies or the fact of repression. Analysts can use open-source data as well as historical research to determine just how unequal and fractious a given military might be. Social divisions are slow to change, creating stable baselines for comparison. Given the gradual and hard-to-conceal nature of social structures, estimates of inequality are more likely to be accurate than those assessing new military technologies or doctrines. Measures of inequality can also be tailored for context. In some settings, ethnic or racial identities might divide a military. Class, ideology, or gender might prove more salient in other armies. Perhaps most important, military inequality is a scalable indicator. It can predict the battlefield performance of entire armies, small formations, or even individual units.
Useful data on morale and cohesion abound. Soldiers air their grievances on social media, including Facebook, Instagram, Telegram, TikTok, and Twitter. Such griping can be used to assess overall morale, distrust within the ranks, and support for the government. Soldiers’ posts can be geotagged to specific bases. Surveys can also be effective tools for assessing client armies. Polling within the Afghan military provided early clues of ethnic tensions, dwindling support for the regime, and an unwillingness among many soldiers to die in defense of the government.
The intelligence community ignores these dynamics of inequality within military forces at its peril.
Inequality can warp how armies recruit and deploy their soldiers. Highly unequal armies will often recruit, forcibly or otherwise, from marginalized communities as a way of insulating the regime from domestic antiwar sentiment. Russia’s invasion of Ukraine stumbled in part because it relied on soldiers from ambivalent non-Russian minority groups, opportunistic Kyrgyz contract soldiers, and duped Russian conscripts plucked from poor, distant regions. Frontal, often uncoordinated, assaults by these “cannon fodder” forces were a direct result of an army staffed largely by soldiers drawn from populations considered to be expendable. Russia tucked these poorly trained conscripts into the army’s logistical corps, leading to snarled supply lines and columns of abandoned tanks. Identifying where armies deploy their marginalized soldiers or where they miss recruitment quotas offers clues about potential vulnerabilities.
Military inequality can cause rancor between soldiers and their superiors. Unequal armies typically draw their officers from politically reliable groups, relegating less trustworthy ones to the enlisted ranks. Divided armies are plagued by hazing and abuse. Officers rule through intimidation and violence. And unfair military legal systems and extrajudicial punishment flourish. Corruption, too, is often a symptom of inequality, as officers abuse their power to steal from their own soldiers. It is easy, after all, to scrimp on maintenance and equipment or to steal wages if officers view their soldiers as beneath them. Such impropriety cripples military effectiveness and can be spotted before a war begins.
Tracking inequality within armies is not a silver bullet. Analysts still need to account for military capabilities. Friction, uncertainty, and the fortunes of war will continue to confound even the best prewar assessments. But adding military inequality to net assessment will improve accuracy at a fraction of the cost of big-ticket technology, such as new satellite capabilities or clandestine signals intelligence. The ways in which social divisions and identities shape battlefield performance cannot be ignored. Military models must include the human elements of war or risk unwelcome surprises on the battlefield.
A Conversation With Lawrence Freedman