Scholarship on nationalism has long emphasized the modern and political nature of nations as “imagined communities” constructed through state institutions, mass education and symbolic repertoires (Gellner, 1983, 1987; Anderson, 1991; Smith, 1987; Brubaker, 1996; Mylonas and Tudor, 2021). Classic accounts conceptualize nations not as primordial entities, but as products of industrialization, print capitalism and selective narratives of the past. Building on this constructivist tradition, this article understands Russian nationalism as a historically contingent project that redefines the boundaries of the political community and hierarchizes groups within it (Riasanovsky, 1969, 2005; Pipes, 1974; Rich, 1976; Suny, 1993; Marsili, 2016). Crucially, nationalism is treated here not only as an ideology but also as a repertoire of state practices that target populations and territories (Weber, 1948; Wedeen, 2008; Mylonas and Tudor, 2021).

Within this perspective, the Soviet Union can be seen as a paradigmatic case of what has been described as an “empire of nations”: a state that simultaneously institutionalized ethnicity through ethnofederal arrangements and sought to cultivate a common political identity (Bromley, 1983; Suny, 1993; Brubaker, 1996; Derluguian, 1999; Raffass, 2012). Nationality was codified in law, registered in internal passports and territorialized through the creation of union and autonomous republics (Comrie, 1981; Riasanovsky, 2005; Rosser and Barkley, 2003). Such institutionalized ethnicity created powerful incentives for political actors to mobilize along national lines, while also providing the centre with instruments to reward, punish or reconfigure groups through changes in status and borders (Bunce, 1998; White, 2000; Derluguian, 1999; Raffass, 2012; Marsili, 2016).

The concept of ethnic engineering is central to the analysis. It refers to deliberate state interventions that seek to reshape the demographic and territorial distribution of groups, including through deportations, forced resettlement, colonization campaigns and the redrawing of administrative boundaries (Martin, 1998; Bugaĭ, 1996; Pohl, 1999; Polian, 2004; Naimark, 2010; Perovic, 2018). In the Soviet case, these practices were justified through a mixture of security discourse – depicting targeted groups as “unreliable” or “enemy” nations – ideological arguments and developmental claims (Burds, 2007; Ellman, 2002; Weiner, 2002; Naimark, 2010). Yet, beyond their immediate objectives, they produced enduring legacies: new majorities and minorities, overlapping claims to territory, and competing memories of injustice that could later be activated in moments of crisis (Allworth, 1998; Marshall, 2010; Williams, 2015; Marsili, 2024).

These interventions interacted with the ethnofederal structure of the USSR. By assigning titular status to certain nationalities and embedding them in quasi-state institutions, ethnofederalism both recognized and reified national difference (Bromley, 1983; Brubaker, 1996; Bunce, 1998; White, 2000; Raffass, 2012). When combined with large-scale deportations and selective rehabilitation, this institutional design generated what might be called asymmetric ethnopolitical configurations: some groups enjoyed a republic of their own, others were scattered across several units, while others still were removed from their homelands and reinserted as minorities elsewhere (Martin, 1998; Pohl, 1999; Polian, 2004; Perovic, 2018). The argument advanced in this article is that such configurations formed part of the structural “infrastructure” of many post-Soviet conflicts (Yamskov, 1991; Dunlop, 1998; Fowkes, 1998; Lordkipanidze and Otkhmezuri, 2007; Tüysüzoğlu and Özkan, 2022).

Finally, the analysis draws on research on forced migration and political violence, which has shown that displacement can be both a consequence and a driver of conflict (Weiner, 2002; Mishali-Ram, 2006; Naimark, 2010; Marshall, 2010; Marsili, 2024). Forced migration transforms local power relations, alters economic opportunities and reshapes the symbolic hierarchy of victimhood and entitlement (Yamskov, 1991; Allworth, 1998; Fisher, 2014; Williams, 2015). By tracing how Soviet-era deportations and resettlements configured these dimensions, the article contributes to a more general understanding of how state-led population politics can generate long-term patterns of contentious politics and war (Smith, 2003; Mylonas and Tudor, 2021).

The study employs a qualitative, historical-comparative research design aimed at identifying and tracing mechanisms connecting Soviet nationalities policy – particularly deportations and other forms of ethnic engineering – to the later emergence of ethnic conflicts in the post-Soviet space (Bugaĭ, 1996; Martin, 1998; Pohl, 1999; Polian, 2004; Perovic, 2018). Rather than testing a single hypothesis on a large number of cases, the research focuses on in-depth analysis of a limited set of empirically rich episodes that illuminate the broader argument, in line with comparative-historical approaches to nationalism and state-building (Brubaker, 1996; Suny, 1993; Mylonas and Tudor, 2021).

The empirical material consists of three main types of sources. First, official documents, including Soviet constitutional texts, decrees of the Presidium of the Supreme Soviet, party resolutions and legal acts regulating nationality status and territorial-administrative changes (Constitutions of the USSR, 1924, 1936, 1977; Decree of the Presidium of the Supreme Soviet about the Transfer of the Crimean Oblast, 1954). These documents are used to reconstruct the formal justifications, categories and procedures through which deportations and resettlements were planned and implemented (Ellman, 2002; Weiner, 2002; Naimark, 2010). Second, quantitative and cartographic data, such as census figures and historical maps, inform the analysis of demographic and territorial transformations in the affected regions (Comrie, 1981; Polian, 2004; Marshall, 2010). Third, a wide range of secondary literature, including historical monographs, regional studies and earlier analyses of specific deportations and conflicts, provides context, empirical detail and interpretive frameworks (Allworth, 1998; Dunlop, 1998; Fowkes, 1998; Fisher, 2014; Perovic, 2018; Tüysüzoğlu and Özkan, 2022).

Data collection followed a systematic strategy. Relevant legal and constitutional texts were retrieved from published collections of Soviet legislation and archival compilations, while secondary sources were identified through keyword searches in academic databases and bibliographic snowballing, focusing on works that explicitly address deportations, nationalities policy and post-Soviet conflicts in the Caucasus, Crimea/Donbas and Central Asia (Allworth, 1998; Dunlop, 1998; Fisher, 2014; Marshall, 2010; Perovic, 2018; Tüysüzoğlu and Özkan, 2022). Where available, archival materials and contemporary reports were consulted to triangulate different accounts of the same events (Ellman, 2002; Weiner, 2002).

Data analysis proceeded in two steps. First, all sources were subjected to thematic coding. Recurrent themes were identified inductively and then organized into a coding scheme that included, among others, categories such as “collective punishment,” “security discourse,” “demographic engineering,” “border manipulation,” “ethnofederal restructuring” and “rehabilitation policies” (Bugaĭ, 1996; Martin, 1998; Pohl, 1999; Polian, 2004; Naimark, 2010). This coding allowed for systematic comparison across different deportations and regions, highlighting both common patterns and specific variations. Second, the study used structured, focused comparison and elements of process tracing to reconstruct the sequences linking Soviet-era interventions to post-Soviet conflicts, asking for each selected case the same set of questions regarding pre-existing demographic and institutional conditions, the nature and timing of deportations and resettlements, and the ways in which these legacies were mobilized after 1991 (Yamskov, 1991; Dunlop, 1998; Fowkes, 1998; Marshall, 2010; Marsili, 2024).

Case selection is theory-driven and strategic. The article focuses on (a) the deportation of Chechens and Ingush, Crimean Tatars and Volga Germans as emblematic instances of ethnic engineering that combined collective punishment with territorial reordering (Allworth, 1998; Bugaĭ, 1996; Pohl, 1999; Polian, 2004; Fisher, 2014; Williams, 2015); and (b) post-Soviet conflicts in the North Caucasus, Crimea/Donbas and selected Central Asian sites where these earlier interventions had a demonstrable impact on demographic balances and territorial claims (Yamskov, 1991; Dunlop, 1998; Fowkes, 1998; Lordkipanidze and Otkhmezuri, 2007; Tüysüzoğlu and Özkan, 2022). These cases do not exhaust the universe of Soviet deportations or post-Soviet conflicts, but they offer analytical leverage by covering different types of nationalities, institutional statuses and regional contexts.

As with all qualitative historical research, the study faces limitations. Archival gaps, inconsistent demographic data and the retrospective reconstruction of motives and perceptions constrain the degree of certainty with which causal claims can be made (Ellman, 2002; Naimark, 2010). Moreover, the focus on selected cases entails that some potentially relevant regions and groups receive less attention. The analysis therefore refrains from claiming statistical generalization. Its contribution lies instead in specifying plausible mechanisms and providing empirically grounded interpretations that can inform further comparative and quantitative work on the long-term effects of ethnic engineering and forced migration (Weiner, 2002; Mishali-Ram, 2006; Mylonas and Tudor, 2021; Marsili, 2024).

Bolshevik discourse famously denounced the Russian Empire as a “prison of nations” and promised a new order based on the self-determination of peoples (Lenin, cit. in Suny, 1993; Brubaker, 1996). In practice, however, the emerging Soviet state combined elements of decolonization with a strong drive to recentralize political authority. The early policy of korenizatsiia promoted the development of national languages and elites in the non-Russian republics, while union and autonomous republics were created as ethnically defined territorial units (Bromley, 1983; Suny, 1993; Brubaker, 1996). Nationality was codified in law, registered in internal passports and mapped onto administrative borders (Comrie, 1981; Rosser and Barkley, 2003; Raffass, 2012).

This institutionalization of ethnicity has often been interpreted as a concession to non-Russian groups. Yet it also gave the centre powerful instruments to manage and reorder ethnic hierarchies. By defining “titular” nations and embedding them in quasi-state institutions, the Soviet leadership both recognized and reified national difference (Bunce, 1998; White, 2000; Raffass, 2012). The result was an “empire of nations” in which ethnicity became the primary language of political representation, but ultimate authority remained concentrated in Moscow (Bromley, 1983; Derluguian, 1999; Marsili, 2016).

Within this framework, Russian nationalism was not openly celebrated, but it was strategically embedded in the narrative of Soviet statehood and in the composition of central institutions. Russians dominated the party and security apparatus, and the Russian language functioned as the main medium of administration, education and inter-ethnic communication (Riasanovsky, 2005; Bryan, 1984; Bonnell, 1999; Brandenberger, 2002; Boer, 2017). As a result, the formal equality of nations coexisted with a de facto hierarchy in which Russian culture and political leadership occupied a privileged position (Allison, 2008; Marsili, 2016).

The evolution of Soviet nationalities policy from the 1920s to the 1950s illustrates the tensions between these principles. In the 1920s, korenizatsiia sought to empower non-Russian elites, foster local languages and integrate peripheral regions through affirmative action policies (Suny, 1993; Brubaker, 1996; Martin, 1998). National schools, publishing houses and cultural institutions proliferated, and local cadres were promoted within party and administrative structures (Bromley, 1983; Brandenberger, 2002).

However, by the early 1930s, this experiment increasingly came to be seen by Stalin and his circle as a potential threat to Soviet unity, especially in border regions and among groups with transnational ties (Paczkowski, 2003; Parrish, 1996; Morris, 2004). The Great Terror entailed not only the elimination of political rivals but also purges of national elites accused of “bourgeois nationalism” or collaboration with foreign powers (Rosefielde, 2009; White, 2012). In this context, policies shifted toward a renewed emphasis on Russian language and culture as the “elder brother” of the Soviet family of nations (Riasanovsky, 2005; Waller, 2015).

The promotion of Russian history, symbols and narratives intensified in the 1930s and 1940s, particularly during the Second World War, when appeals to patriotic traditions became central to mobilization (Tuminez, 2000; Petro, 1995; Brandenberger, 2002; Allison, 2008). This did not mean an outright abandonment of the ethnofederal framework, but rather its recasting under a more explicitly Russocentric ideological umbrella. Russian nationalism thus became a key resource for legitimizing both the domestic order and the Soviet Union’s expansionist policies in Eastern Europe and the Caucasus (Lukin, 1992; Margolis, 2008; Marshall, 2010; Marsili, 2021).

It is in this context that mass deportations and forced resettlements acquired a central role as instruments of ethnic engineering. Between the late 1930s and the late 1940s, entire populations were uprooted and relocated to remote regions, often under conditions of extreme violence and deprivation (Bugaĭ, 1996; Pohl, 1999; Polian, 2004; Naimark, 2010). The targets included, among others, Poles, Germans, Koreans, Chechens, Ingush, Crimean Tatars, Volga Germans, Karachays, Balkars and Meskhetian Turks (Allworth, 1998; Bugaĭ, 1996; Pohl, 1999; Polian, 2004; Naimark, 2010; Perovic, 2018).

Officially, these operations were justified as security measures against “enemy nations” suspected of collaboration with invading armies or foreign intelligence services (Burds, 2007; Ellman, 2002; Weiner, 2002; Naimark, 2010). Yet their scope, selectivity and timing reveal a deeper logic. Deportations allowed the Soviet leadership to neutralize real or imagined opposition in strategically sensitive areas, to clear borderlands of groups considered unreliable, and to redistribute populations in ways that altered local balances of power (Martin, 1998; Pohl, 1999; Polian, 2004; Perovic, 2018).

At the same time, deportations were embedded in broader projects of territorial and demographic reordering. The removal of certain groups created opportunities to settle others in their place, including Russians and other Slavs, thereby reinforcing central control and facilitating exploitation of resources (Weiner, 2002; Marshall, 2010; Perovic, 2018). Administrative boundaries were redrawn to reflect new demographic configurations, or conversely, to maintain institutional claims over territories from which titular populations had been expelled (Allworth, 1998; Yamskov, 1991; Lordkipanidze and Otkhmezuri, 2007; Rayfield, 2012).

The deportations of Chechens and Ingush, Crimean Tatars and Volga Germans are particularly illustrative. In the North Caucasus, the abolition of the Chechen-Ingush Autonomous Soviet Socialist Republic (ASSR) after the 1944 deportations and the redistribution of its territory among neighbouring units transformed local ethnic and political relations, laying the ground for subsequent disputes over borders and return (Askerov, 2015; Perovic, 2018; Williams, 2015). In Crimea, the expulsion of Crimean Tatars and other minorities, followed by the settlement of Slavic populations and the later transfer of the oblast to the Ukrainian SSR in 1954, created a complex layering of claims that would become central after 1991 (Allworth, 1998; Fisher, 2014; Williams, 2015; Yarmysh and Cherviatsova, 2016). The case of the Volga Germans, whose autonomous republic was abolished and whose population was scattered across Kazakhstan and Siberia, illustrates how deportation could permanently dismantle institutionalized forms of national autonomy (Pohl, 1999; Polian, 2004; Perovic, 2018).

Although partial rehabilitation occurred in the Khrushchev era and some deported peoples were allowed to return to their homelands, the process was uneven and incomplete (Bugaĭ, 1996; Polian, 2004; Naimark, 2010). Certain groups, such as the Crimean Tatars and Meskhetian Turks, faced substantial obstacles in reclaiming land and political representation, while others never recovered their former institutional status (Allworth, 1998; Williams, 2015; Fisher, 2014). These asymmetries generated enduring grievances and competing narratives of victimhood, which later fed into mobilization in the late Soviet and post-Soviet periods (Yamskov, 1991; Dunlop, 1998; Fowkes, 1998; Marshall, 2010; Marsili, 2024) (Table 1).

Taken together, the institutionalization of ethnicity, the shift from korenizatsiia to Russification and the extensive use of deportations and territorial rearrangements amounted to a far-reaching project of ethnic engineering. While framed in the language of socialist modernization and security, these policies systematically reshaped the demographic and territorial landscape of the Soviet Union in ways that embedded fault lines into the very structure of the state. The next section examines how these Soviet-era practices generated mechanisms that later contributed to the emergence and persistence of ethnic conflicts in the post-Soviet space.

Mass deportations radically altered local demographic structures. The removal of entire populations – such as Chechens, Ingush, Crimean Tatars or Volga Germans – created immediate vacuums that were often filled by settlers from other regions, including Russians and other Slavs (Bugaĭ, 1996; Pohl, 1999; Polian, 2004; Weiner, 2002; Marshall, 2010). In some cases, deportations were followed by deliberate colonization campaigns, which further consolidated new majorities. When partial rehabilitation allowed some deported groups to return, they found their homelands populated by others, and their previous status as local majorities or titular nations had frequently been eroded (Allworth, 1998; Fisher, 2014; Williams, 2015). These majority–minority reversals generated structural tensions that later fuelled disputes over land, resources and political representation.

Deportations were closely intertwined with changes in administrative borders and the institutional status of territories. The abolition of autonomous republics, the redistribution of their lands among neighbouring entities and, later, the creation or restoration of units with altered boundaries produced what might be called institutional “mismatches” between populations and territories (Martin, 1998; Raffass, 2012; Rayfield, 2012; Perovic, 2018). Some groups lost their autonomous status altogether, others saw their historical territories fragmented, and still others were granted autonomy in areas where they no longer constituted a clear majority. These arrangements were largely manageable under a highly centralized one-party state, but they became focal points of contention once the Soviet Union disintegrated and ethnofederal units turned into potential or actual states (Derluguian, 1999; Bunce, 1998; Brubaker, 1996).

Deportations also generated powerful symbolic legacies. The trauma of forced removal, high mortality during transport and exile, and the difficulties of return left deep marks on collective memory (Allworth, 1998; Naimark, 2010; Marshall, 2010; Williams, 2015). In the late Soviet and post-Soviet periods, these experiences were articulated in narratives of historical injustice that demanded recognition, restitution and, in some cases, territorial reconstitution (Yamskov, 1991; Dunlop, 1998; Fowkes, 1998; Fisher, 2014). At the same time, other local populations developed their own counter-narratives, portraying deported groups as collaborators or external intruders. The result was a hierarchy of victimhood claims that could be mobilized by elites to legitimize secession, resistance to returnees or demands for protective intervention by the Russian Federation (Marsili, 2016, 2021, 2024).

From the outset, deportations were framed through a security discourse that constructed targeted populations as “enemy nations” or potentially disloyal groups in strategically sensitive borderlands (Ellman, 2002; Burds, 2007; Weiner, 2002; Naimark, 2010). This discourse did not disappear with the end of Stalinism. It was periodically reactivated in the late Soviet and post-Soviet eras, particularly in the North Caucasus and Crimea, where local mobilization could be presented as a threat to state integrity or to the rights of Russian-speaking populations (Dunlop, 1998; Fowkes, 1998; Gammer, 2006; Williams, 2015; Tüysüzoğlu and Özkan, 2022). The enduring association of certain groups with treason or extremism facilitated harsh counterinsurgency campaigns and justified the militarization of conflicts that might otherwise have been addressed through negotiation.

The partial return of deported groups after Khrushchev’s denunciation of Stalin’s crimes introduced additional mechanisms linking past deportations to contemporary conflict. Returnees often faced legal and practical obstacles in reclaiming land and housing, which had long since been occupied by other populations (Bugaĭ, 1996; Polian, 2004; Allworth, 1998; Fisher, 2014). Local authorities were frequently unwilling or unable to accommodate competing claims, leading to protracted disputes and sporadic violence. In some cases, such as the North Caucasus and Crimea, these tensions intersected with broader struggles over the status of autonomous republics and the redistribution of power within the Soviet and post-Soviet federal systems (Yamskov, 1991; Dunlop, 1998; Fowkes, 1998; Perovic, 2018; Williams, 2015).

Taken together, these mechanisms suggest that Soviet deportations and related forms of ethnic engineering did not simply remove threats in the short term. They created new demographic and institutional configurations, entrenched conflicting memories of injustice and established security narratives that could be reactivated decades later. The following comparative case studies illustrate how these mechanisms operated in specific regional contexts and how they contributed to the emergence and persistence of ethnic conflicts after 1991 (Table 2).

The North Caucasus offers a paradigmatic illustration of how Soviet-era deportations and territorial reordering generated legacies that later fuelled violent conflict. The region had long been a site of intense resistance to Russian imperial expansion, most notably during the Caucasian War of 1817–1864 (King, 2008; Dowling, 2014). Soviet rule did not resolve these tensions; rather, it reconfigured them within the framework of ethnofederal institutions and security-driven population policies (Derluguian, 1999; Perovic, 2018).

In 1944, the Chechens and Ingush were collectively accused of collaboration with Nazi Germany and deported en masse to Central Asia and Siberia. The Chechen-Ingush Autonomous Soviet Socialist Republic (ASSR) was abolished, its territory redistributed among neighbouring republics and regions, and even place names were Russified (Bugaĭ, 1996; Pohl, 1999; Polian, 2004; Askerov, 2015; Perovic, 2018). Mortality during deportation and exile was extremely high, and the experience became a central element in Chechen and Ingush collective memory (Polian, 2004; Marshall, 2010).

The partial rehabilitation launched under Khrushchev allowed Chechens and Ingush to return and led to the restoration of the Chechen-Ingush ASSR in 1957, but the process was far from smooth (Bugaĭ, 1996; Pohl, 1999; Polian, 2004). Returnees found their former homes and lands occupied by settlers, including people from neighbouring regions who had been encouraged or compelled to move in during their absence (Marshall, 2010; Perovic, 2018). Property disputes, competition over scarce resources and resentment among both returnees and current inhabitants generated an atmosphere of latent conflict.

These tensions interacted with broader institutional and political dynamics. The boundaries of the restored ASSR did not fully coincide with pre-deportation borders, and some territories with mixed or disputed populations remained outside its jurisdiction (Rayfield, 2012; Rezvan, 2010; Perovic, 2018). As a result, competing claims over land and status persisted between Chechens, Ingush, Ossetians and other groups, contributing to violent clashes in places such as the Prigorodny district in the early 1990s (Rezvan, 2010; Tüysüzoğlu and Özkan, 2022).

With the disintegration of the Soviet Union, these structural and symbolic legacies were mobilized in new ways. Chechen elites framed independence demands in part through references to historical injustices, including the deportations and brutal counterinsurgency campaigns of the Soviet period (Dunlop, 1998; Fowkes, 1998; Gammer, 2006). Moscow, in turn, portrayed Chechen separatism as a security threat with potential spillover effects across the Caucasus and into Russia’s Muslim regions, drawing on longstanding discourses of the “unreliable” or “dangerous” North Caucasus (Marshall, 2010; Gammer, 2014; Perovic, 2018). The two Chechen wars of 1994–2009 thus unfolded against a backdrop shaped by earlier ethnic engineering: demographic restructuring, territorial ambiguities, unresolved property disputes and entrenched narratives of mutual victimhood (Dunlop, 1998; Fowkes, 1998; Gammer, 2006; Askerov, 2015; Marsili, 2016, 2024).

In this case, the mechanisms identified above are clearly visible. Deportations and the abolition of the Chechen-Ingush ASSR created majority–minority reversals and institutional mismatches; return migration after 1957 generated intense local competition over land; and security discourses depicting Chechens as inherently rebellious or extremist facilitated the escalation from protest to full-scale war. The North Caucasus thus exemplifies how Soviet ethnic engineering could transform ethno-national diversity into an enduring infrastructure of violence.

Crimea and the Donbas region illustrate a different, though related, configuration of legacies. Here, deportations, resettlement and administrative transfers produced a complex layering of demographic realities and legal claims that became central to post-Soviet conflicts (Allworth, 1998; Fisher, 2014; Williams, 2015; Yarmysh and Cherviatsova, 2016).

In 1944, Crimean Tatars, along with other minority groups such as Greeks, Bulgarians and Armenians, were deported from Crimea to Central Asia on charges of collaboration with the German occupiers (Allworth, 1998; Bugaĭ, 1996; Pohl, 1999; Polian, 2004; Naimark, 2010). Their property was confiscated, and the peninsula was rapidly repopulated by settlers from the Russian heartland and neighbouring regions, leading to a sharp increase in the share of ethnic Russians in Crimea’s population (Fisher, 2014; Williams, 2015). In 1954, the Crimean oblast was administratively transferred from the Russian SFSR to the Ukrainian SSR, a decision that later acquired outsized political significance (Decree of the Presidium of the Supreme Soviet about the Transfer of the Crimean Oblast, 1954; Yarmysh and Cherviatsova, 2016).

Although formal restrictions on Crimean Tatars were gradually relaxed from the late 1960s onwards, large-scale return only began in the late Soviet and early post-Soviet periods. Returnees faced substantial obstacles in reclaiming land and housing, as their former properties had long been occupied by others and local authorities were often reluctant to recognise their claims (Allworth, 1998; Fisher, 2014; Williams, 2015). Many settled in peripheral or less desirable areas, often without full legal status. These conditions fuelled grievances not only against the Soviet and later Ukrainian state, but also against Russian-speaking populations perceived as beneficiaries of earlier injustices (Allworth, 1998; Fisher, 2014; Williams, 2015).

At the same time, the demographic and symbolic weight of the Russian population in Crimea provided fertile ground for narratives portraying the peninsula as historically and culturally “Russian,” despite its administrative inclusion in Ukraine (Fisher, 2014; Williams, 2015; Marsili, 2021). Memories of the 1954 transfer and debates over its legality – including the interpretation of Soviet legislation on territorial changes – became central to political struggles in the 1990s and 2000s (Yarmysh and Cherviatsova, 2016). After 1991, local elites and Moscow-based actors increasingly invoked the protection of Russian speakers as a justification for challenging Kyiv’s authority, culminating in the 2014 annexation (Fisher, 2014; Williams, 2015; Marsili, 2021, 2024).

The Donbas region, although not directly shaped by mass deportations to the same extent as Crimea, was also marked by Soviet-era population movements and industrialization policies that created a high concentration of Russian-speaking workers and a strong identification with Soviet industrial culture (Rosser and Barkley, 2003; Marshall, 2010; Tüysüzoğlu and Özkan, 2022). In both Crimea and Donbas, the combination of demographic legacies, contested memories of the Soviet past and ambiguities in legal and administrative status provided a fertile ground for post-Soviet conflicts in which appeals to history and victimhood played a central role (Allworth, 1998; Fisher, 2014; Williams, 2015; Marsili, 2021, 2024).

Here too, the mechanisms linking Soviet ethnic engineering to post-Soviet conflict are evident. Deportations and resettlement reconfigured the ethnic composition of Crimea; the 1954 transfer introduced an institutional mismatch between demographic realities and administrative affiliation; and the partial, contested return of deported populations generated localised conflicts over land and status. These dynamics interacted with broader geopolitical shifts and with the Russian Federation’s evolving doctrine of protecting compatriots abroad, illustrating how long-term population politics can shape the fault lines of contemporary war.

Central Asia illustrates a different constellation of legacies linking Soviet ethnic engineering to post-Soviet tensions. Here, the combination of deportations, planned settlement and national-territorial delimitation produced intricate border configurations – including enclaves and exclaves – that became flashpoints of interethnic violence after 1991 (Crews, 2006; Weiner, 2002; Tüysüzoğlu and Özkan, 2022).

During the Stalinist period, large numbers of deported populations were resettled in Kazakhstan and other Central Asian republics, including North Caucasian groups, Volga Germans and Crimean Tatars (Bryan, 1984; Martin, 1998; Polian, 2004; Naimark, 2010). At the same time, national-territorial delimitation in the 1920s and 1930s carved the region into union and autonomous republics whose borders reflected a mixture of ethnic, economic and strategic considerations (Hirsch, 2000; Weiner, 2002; Crews, 2006). The result was a patchwork of interlaced territories, where ethnic communities did not always coincide with administrative units and where transport routes and resources often crossed republican boundaries (Tüysüzoğlu and Özkan, 2022).

The dissolution of the USSR transformed these internal administrative lines into international borders, hardening what had previously been permeable frontiers. In places such as the Fergana Valley – shared by Uzbekistan, Kyrgyzstan and Tajikistan – Soviet-era decisions on district boundaries and land allocation left behind a dense mosaic of mixed settlements, enclaves and exclaves (Chang, 2019; Tüysüzoğlu and Özkan, 2022). The Kyrgyz exclave of Barak, a tiny village in the Fergana Valley surrounded by Uzbek territory, and several Uzbek enclaves within Kyrgyzstan such as Sokh and Shokhimardon, together with Tajik enclaves in Kyrgyzstan, became recurrent sources of tension and intermittent violence in the post-Soviet period (Lachert and Kamiński, 2019; Tüysüzoğlu and Özkan, 2022).

Ethnic clashes in the Osh oblast in 1990 and again in 2010, where Uzbeks and Kyrgyz competed over land, political power and economic opportunities, illustrate how Soviet-era demographic engineering and border-drawing continued to shape local conflict dynamics. Under Soviet rule, industrial and agricultural development policies had attracted labour migrants and reshaped settlement patterns, while local elites were integrated into republican and district-level institutions (Crews, 2006; Weiner, 2002). After 1991, the same demographic and institutional configurations – now embedded in independent states with contested borders and uneven state capacity – produced overlapping claims to territory, resources and representation (Tüysüzoğlu and Özkan, 2022).

These Central Asian cases differ from the high-intensity wars of the Caucasus or the large-scale secessionist conflicts in Crimea and Donbas. Violence has often been episodic and localized, and external intervention less direct. Yet the underlying mechanisms are similar: demographic restructuring through deportation and resettlement; institutional mismatches between populations and borders; and competing narratives of victimhood and entitlement rooted in Soviet-era policies. Recent efforts by Uzbekistan and Kyrgyzstan to resolve some border disputes demonstrate that these legacies can be partially managed through negotiation, but they also highlight the lasting impact of Soviet ethnic engineering on contemporary conflict risk (Lachert and Kamiński, 2019; Tüysüzoğlu and Özkan, 2022) (Table 3).