Marginalization, Military Service and Rebellion in Imperial Russia
The “Honorary Revolutionary Weapon” (“Pochetnoe Revoliutsionnoe Oruzhie”) was one of the earliest awards of the Russian Red Army and the only one dedicated to the Russian Civil War. Only 21 Red Army commanders received this award — for exceptional military distinctions displayed during the 1917 revolutionary movement. Notably, 20 of them had previously fought in the Russian Imperial Army and were officially recognized by the state for their bravery in WWI. The recipients came from diverse backgrounds: only eight of them were Russians, followed by three Ukrainians, two Poles, one Belarusian, one Moldovan, one Estonian, one Lithuanian, one Latvian, one Armenian, one Bashkir, and one Cherkess. Why did these once-loyal servicemen of the Imperial Army turn their weapons against the empire? What role, if any, did ethnicity play in these decisions?
Research detailing the relationship between repression and fighting behavior has primarily focused on the effects of wartime discrimination on battlefield dynamics. Wartime discrimination and military inequalities can play a detrimental role in battle outcomes and negatively affect individual experiences, as Connor Huff and Robert Schub highlight in this Broadstreet post. I document this in my own work with Arturas Rozenas and Yuri Zhukov. While questions of wartime behavior are important, studies often overlook how pre-war discriminatory and repressive politics affect the political behavior of marginalized veterans in a post-conflict environment. In my job market paper, I fill this gap by studying the effects of a state’s exclusionary treatment of ethnic minorities on their subsequent behavior that extends beyond the battlefield.
Prominent theories suggest two competing mechanisms through which military service of individual soldiers shapes their post-war behavior. First, military service could increase the sense of national unity and loyalty to the regime through intensive discipline and ideological indoctrination or increased social learning and inter-group contact and, therefore, induce obedience toward the state. Second, military service could either reinforce or further exacerbate existing cleavages, increase aggression toward rivals, and provide a conducive environment where marginalized individuals gain the skills necessary for engaging in collective action.
The form of this relationship between military service and post-service political behavior likely depends on the level of discrimination soldiers were exposed to in their communities prior to enlistment. Exclusionary treatment, such as discrimination and repression during peacetime, might make soldiers resilient to direct inculcation of unitary identity and national propaganda at wartime. Moreover, important organizational and fighting skills gained through military training and on the battlefield, which are difficult to attain outside military institutions and may be denied to individuals from marginalized groups, could bolster these individuals’ ability to challenge the state. These potentially differential effects of military service merit scholarly attention and have important theoretical and policy implications.
Imperial Russia in WWI
To investigate how the military service of soldiers from the marginalized groups impacted their behavior toward the state I focus on the Russian Empire. Unlike many of its multi-ethnic contemporaries, Imperial Russia did not grant administrative and cultural autonomy to its non-Russian constituents. Nor was it particularly keen on enfranchising its minority populations. Political, social, and cultural oppression and persecution were prevalent, and political censorship and cultural assimilation were the dominant state policy choices. Russia’s strategy on the nationality question evolved in two directions: political integration based on loyalty to the empire and Russianness delineated along ethnic and religious categories (read more about imperial minority policies in this Broadstreet post).
I draw on several original multi-level datasets constructed from declassified administrative military personnel records, archival materials and lists, the first and only Russian Imperial Census of 1897, historical-geographical atlases, and contextual data from additional contemporary resources. The recently declassified and digitized archival records on the Russian imperial army conscripts and revolutionary movement allow individual-level analysis with a high degree of spatial and temporal detail.

The case of the Russian Empire in WWI is compelling for several substantive and empirical reasons. The Great War was one of the deadliest conflicts in history, claiming over 35 million military and civilian lives worldwide. The Russian Empire suffered the second-largest military losses among the belligerents, following Germany with over 9.1 million total battlefield casualties — 76% of its mobilized WWI forces. WWI marked the first mass mobilization in Russian history. The military experience of WWI is also unique because the war culminated in “national mobilization” across all combatant countries, particularly in Imperial Russia. The timing, scale, and severity of the Russian Civil War provide a unique opportunity to trace the immediate post-war behavior of WWI veterans. The Russian Civil War immediately followed the Great War and engulfed the country, devastating the society with crime, famine, and disease and claiming five million lives, 90% of which were civilians.
In the summer of 1914, when Tsar Nicholas dragged his empire into one of the deadliest wars in human history, he commanded the largest standing army in the world. Within a few months following Germany’s declaration of war, more than a million individuals belonging to various ethnic and religious minority groups were conscripted into the Imperial Army. In desperate need of manpower, the Tsar undertook additional military reforms to ensure full mobilization across the Russian Empire, even as he questioned the loyalty of his non-Russian subjects.
A few years later, as a bitter civil war engulfed the Russian Empire, the prime movers behind the revolutionary process turned out to be the members of the minority groups that had served in the Imperial Army during WWI. Roughly 30% of the newly formed Red Army of Workers and Peasants consisted of ethnic and religious minorities. In a short span of time, thousands from the elite Latvian riflemen (“Latishskie strelki’‘) division sided with the Bolsheviks. Not long ago fighting for the same side on the battlefronts of WWI, the individual Imperial Army soldiers ended up fighting against each other, split between the Bolshevik forces and the pro-government White Guard.
Legacies of Targeted Repression
To understand whether prewar ethnic marginalization played an important role in the revolutionary behavior of war veterans, I collected data on 2.3 million Imperial soldiers of WWI, 350,000 soldiers of the state-led White Guard, and 50,000 soldiers of the revolutionary Red Guard, as well as detailed demographic and socio-economic data for approximately 824 imperial districts (101 provinces). I match the three individual-level administrative datasets with the 1897 district data based on each soldiers’ birth location.
The administrative records do not contain information on soldiers’ ethnicity. To address this shortcoming, I develop three different measures that capture both the ethnic background of a given individual and the ethnic composition of the district corresponding to the birth location. The first and second measures are direct measures of an individual’s ethnicity, while the last is an index measure capturing the ethnic heterogeneity of the birth location.

I begin by testing the association between the ethnicity of individual veterans of WWI and their post-war loyalty during the revolutionary period. All empirical specifications include a range of controls for socio-economic, demographic, and geographic characteristics of districts, province-level fixed effects, and spatial splines for smooth local spatial interpolation. To causally identify the observed effect, I exploit ethnic dissimilarities across the imperial district borders in a spatial regression discontinuity design, focusing on a smaller area (western European provinces) with 25 provinces corresponding to 221 unique districts. The district borders in the European section did not correspond to any preexisting differences, unlike the southern and southeastern frontiers, some of which correspond to the borders of independent nation-states before they were occupied by the Russian Empire.
The findings suggest that soldiers from marginalized ethnic groups were more likely to join the Red Army to fight against the crumbling empire. Ethnic Russian WWI veterans, on the other hand, were more likely to join the White Army to suppress the revolutionary movement. The probability of an ethnic minority veteran joining the Revolutionary Red Army was 72% higher than that of an ethnic Russian veteran. I also find that these identity-driven effects have geospatial salience. Inhabitants of ethnically heterogeneous districts were more likely to join revolutionary forces, while soldiers from districts with less ethnic diversity continued to fight on behalf of the imperial forces.
The Role of Combat Experience
I also examine the variation in soldiers’ combat experience to isolate the effect of military service. I focus on two major military engagements of the Imperial Army: the Brusilov Offensive, the deadliest in military history and the most expansive Russian military operation of WWI, and the Galician Operation, one of the largest frontline operations of WWI. These operations involved more than 25 separate battles with hundreds of unique Russian military regiments.
Using daily-level unit movement data from WWI, I identify individual soldiers who fought in these battles. I find that ethnic minority soldiers who participated in these frontline battles were more likely to join the revolutionary movement than those who fought in the rear. Further analysis reveals that assignment to active combat duties, as opposed to “dirty tasks,” such as sanitary, cooking, construction, is positively associated with post-war unrest. I consider a set of alternative tests, investigating differential exposure to casualties, systematic discrimination in unit assignment, the timing of conscription, various spatial adjustments, and alternative measures of district diversity.
Lessons for the Authoritarian Rulers of Multiethnic States
This study establishes three main results. First, the effects of pre-war social cleavages travel beyond soldiers’ battlefield performance. Veterans turn against the discriminatory state when an opportunity to challenge state institutions arises. These longer-term effects matter, particularly in autocracies, because the state resorts to its military to ensure regime survival when internal security agencies fail in the face of domestic unrest. Second, these individual-level experiences have spillover effects in areas with a high diversity of marginalized groups. In heterogeneous societies, where territorial boundaries do not mirror existing cleavages, even individuals who do not experience discrimination might support their neighbors in their fight against the state. Third, the results capture a consequential trade-off for the state: in times of war, it may need to rely on marginalized groups to defend itself from the external enemies, but doing so jeopardizes state survival in the long run.
Taken together, these findings suggest that, when given a choice to demonstrate their loyalty to the throne or their people, army recruits from historically marginalized communities will choose the latter. The main explanation supporting these results is consistent with the literature on military performance: citizens who become soldiers rarely lose their civilian identities.