Introduction

The Central Asian states of the former Soviet Union gained renewed global attention during the Syrian conflict as the source of several thousand conflict volunteers, often referred to as ‘foreign fighters’. Combining with other parts of the former Soviet Union – including the Caucasus region torn by civil war for much of the 1990s and early 2000s – this broader region of which Central Asia is a part surprisingly made up the single-largest source of foreign volunteers who traveled to conflict zones in Syria and Iraq.¹

The outflow was the largest migration to a foreign conflict in the history of the Central Asian states. It was unique in that in many cases, those who traveled to the conflict zone were not ‘foreign terrorist fighters’ at all, but women and children. In Kazakhstan for example, according to official figures released by the state security services in 2018, women and children outnumbered men five-to-one in the conflict zone.

Following the recapture of large parts of Syria and Iraq at least hundreds and as many as several thousand women and children from Central Asia were detained in camps like al-Hol in northern Syria and within the Iraqi prison and orphanage system, alongside citizens of between 80 and 110 other countries.² Addressing both the humanitarian and potential security crises that this situation presented, the governments of Kazakhstan and Uzbekistan have become world leaders in repatriation of their citizens following the conflict, organizing state programs to repatriate and re-integrate 654 and 512 people respectively. They have been joined by Tajikistan and Kyrgyzstan in smaller pilot programs scheduled to continue.³

The long-term success of repatriation programs has come under increased challenge in recent months with the unexpected collapse of the government of Afghanistan – with which Uzbekistan, Tajikistan and Turkmenistan share direct borders – and the rapid expansion of territorial control by the Taliban. With recruiting efforts by both Taliban-affiliated Central Asian-led militias and isis Khorosan, fears of a new outflow of citizens or domestic insurgency have raised the stakes for innovative approaches taken to repatriation programs.

This paper will examine the recent history of Central Asian migration to foreign conflicts and insurgencies, offer a brief discussion potential for a new wave of migration that the fall of the Afghan government could present for the region, and argue that new developments have increased the importance of understanding the causes of conflict migration from Central Asia. This essay will argue, based on recent research, that a one-dimensional focus on ideological or theological motivations for past waves of conflict migration is a poor explanatory mechanism for the broader conflict migration that occurred. Independent academic research has instead long presented consistent evidence that a complex, localized, and multi-factor approach provides a much better explanatory model for mobilization to both local violence and foreign conflict. More recent research from the Syrian conflict indicates that community-level focus on factors that drive mobilization needs to be prioritized to ensure the best possible outcomes for re-integration and to avoid a new wave of conflict migration, in addition to the broader toolkit for Preventing and Countering Violent Extremism (p/cve) and individual support.

Foreign Terrorist Fighters or Conflict Migrants? A Historical Perspective

Central Asians, particularly ethnic Uzbeks and former citizens of Uzbekistan, have gained notoriety as leaders of small militant Islamist movements like the Islamic Movement of Uzbekistan that trained and fought in Afghanistan and Pakistan in an alliance with both the Taliban and al Qaida beginning in the late 1990s. While those groups waned sharply after the collapse of the first Taliban government in Afghanistan, with the rise of Daesh, Central Asian émigrés became particularly well-known as perpetrators of small-scale but resonant terror assaults in the US, Europe, and Turkey. In 2017, the most intense period of these attacks, former citizens of Uzbekistan carried out attacks inspired by Daesh in Istanbul, Stockholm, and New York – the deadliest terror attack in the city since 2001 – and also a bombing in the St Petersburg metro that year was attributed by Russian authorities to an ethnic Uzbek citizen of Kyrgyzstan.

While the involvement of Central Asians in violent movements is the exception rather than the rule, the one-off or lone-actor attacks like these that involve only a few individuals among the estimated five million Central Asian labor migrants obscure a much larger problem. Historically, a large proportion of Central Asians drawn into conflict zones that involve Islamist insurgencies – in Afghanistan, in Syria, in Iraq – are families or even extended families from marginalized groups or regions seeking better prospects for themselves and their children in a new state and social structure promised by the groups they join.

This is not to say that no Central Asians participate in ideologically motivated terrorism, or to downplay the seriousness or impact of tragic attacks in recent years. But an overfocus on only one aspect of a broader problem has frequently obscured an important aspect of the problem of mobilization to foreign conflicts. This paper will contribute to an argument that a better understanding of the broader problem of conflict migration can help prevent mobilization of both militants and non-combatants at home and abroad. This approach is suggested by the wide disparity outlined by Edward Lemon between the perception – and often the politicization – of Central Asian involvement in international groups and conflicts and the relative rarity of terror attacks at home. From 2008–2018, a database that Lemon gathered using a broad definition of ‘terrorism’ contains only nineteen attacks, many of which were targeted assassinations and attacks on police. They are described by state officials as terrorism but have no externally verifiable link to a larger group or ideology.⁴

Attacks like the Daesh-inspired New Year’s Eve mass shooting in an Istanbul nightclub in 2017 and the brutal murder of four foreign cyclists in Tajikistan the following year have drawn attention again to Central Asian militants, but they had long before already drawn significant security response from many of the world’s most powerful states. But focus on these relatively few cases of violent actors and mobilizers obscures several cycles of much larger migration by people who believed often that a better life and new opportunities were awaiting them in a new state or territory controlled by groups they supposed would welcome them.

The first of these cycles of conflict migration was relatively short and smaller scale. Following the founding of the imu in the late 1990s, hundreds of people – up to two thousand at the peak⁵ – from Uzbekistan, Kyrgyzstan and Tajikistan joined the movement in Afghanistan torn by civil war during the Taliban’s first period in power. This cycle coincided with the beginning of a period of intense discrimination and persecution directed at practicing Muslims and followers of a number of popular religious teachers in Uzbekistan under former President Islam Karimov (d. 2016).⁶ While few contemporary accounts of life within the group are still extant from that period, recollections from non-combatant former members were published in an extended video series formerly available in Uzbek on YouTube and have been collected within a field interview project in Istanbul (2019). These anecdotal accounts indicate that following the collapse of the first Emirate in Afghanistan, a large proportion of the imu families who fled into Pakistan left the group and voluntarily demobilized, rather than continuing to fight with the imu and its splinter group, the Islamic Jihad Union.

While this first cycle never fully ended, by 2012 and early 2013 it became clear that recruits, facilitators, and fundraisers were far more interested in the growing conflict in Syria.⁷ This was also true of many of the fighters and some high-ranking commanders of those groups, many of whom began to leave the Afghanistan/Pakistan theater and relocated to Syria, sometimes forming new groups that competed – mostly very successfully – for both funds and recruits.

This second and much larger cycle that began with the conflict in Syria had already by 2015 become the largest mass migration to a conflict in a Muslim majority country since the Second World War.⁸ Not surprisingly, the Central Asian migration to Syria was far larger than the earlier wave to Afghanistan. In all states except Kazakhstan, which is a receiving country for labor migrants from its neighbors, the available evidence indicates that a large proportion of those who eventually migrated to conflict zones in Syria were already at least temporarily displaced from their home country by economic, political, and religious factors. The frequency with which conflict migrants brought their whole families suggests they hoped for a better future for their children in the Caliphate. Secondary interviews with family members and with demobilized individuals indicate that in many cases those who went, hoped for a better future for their children, a place where Islamic education – illegal for minors in Uzbekistan and Tajikistan – was supported and where they would not face discrimination or prosecution for their religious beliefs.⁹

Narratives vary significantly though: some cite economic factors and promised opportunities, housing, new citizenship, the chance to do meaningful work. One of the most jarring Daesh recruitment videos targeted at Central Asians featured a congenitally blind young man from the desert mining city of Satpaev in Kazakhstan who had struggled to live off a disability stipend at home. His local mosque community gathered donations to pay for him to be trained as a massage therapist. But he had failed to find enough clients to support himself, after which he disappeared, leaving his community hoping he had moved to another city to seek work.¹⁰ In the recruiting video, he is shown providing massage therapy to wounded militants in a combat hospital, and is presented as someone who found dignity, purpose, and belonging in the Caliphate, something they claim he could never find at home.¹¹

The mobilization of Central Asians to the Syrian conflict and the level of recruiting directed toward them has significantly decreased. The defeat of Daesh in Syria and Iraq, serious losses of territory and fighters for other groups operating still in northern Syria, in combination with a far less permissive environment for Central Asian recruiters and facilitators in Turkey following the 2017 nightclub attack and the global pandemic, have all but halted observable recruiting to that conflict. Groups themselves that continue to operate in Syria have shifted messaging from recruiting to trumpeting successes of the Taliban in Afghanistan or the struggle of isis Khorosan to counter them, signaling perhaps the beginning of a new wave that will be discussed in more detail below.

Lessons of the Syrian Mobilization: Complex Problems Require Complex Solutions

While evidence indicates a wide spectrum of factors that pushed and pulled conflict migrants into armed groups and the territories they controlled, public and elite discourse as well as policy approaches within Central Asia have remained consistent since the mid-1990s. As well established already by scholars analyzing and often critiquing these approaches (Heathershaw and Megoran,¹² Montgomery and Heathershaw,¹³ Lemon,¹⁴ Thibault,¹⁵ Nasritdinov and Esenemanova,¹⁶ Beissembayev and Sharipova¹⁷ and many others), state-driven discourses have almost exclusively focused on individual-level ‘pathways to radicalization’, governed by the assumption that ‘fanatical’ or extreme religious beliefs are the core engine of radicalization. These approaches have frequently resulted in discrimination against people of faith who had no association with violent groups, and in sometimes bizarre policies like Tajikistan’s campaign against beards that led to thousands of young men detained by police and forcibly shaved, sometimes becoming the occasion of serious abuse.¹⁸

The scale of the migration to Syria provided practical evidence that policies to police religious expression were simply not effective. Rather than resulting in a drop in outflow to armed groups or insurgencies associated with Islamist ideology, consistent application of these policies and their progressive intensification in the years following the Arab Spring and leading up to the outbreak of the Syrian conflict, saw instead a sharp rise in successful recruitment by armed groups, with per capita rates among the highest in the world.¹⁹

The decision by four of the five Central Asian states to repatriate and rehabilitate conflict migrants or former members of armed groups in the Syrian conflict indicates a sharp change in approaches and a willingness to work to integrate disaffected or separated citizens, rather than focus on penalizing and isolating them. This is a cardinal change that indicates a willingness to be guided by evidence-based approaches. But what does the evidence tell us about the last wave of conflict migration, and how are those factors potentially affected by the emergence of a victorious Islamist insurgency just across the border?

First, evidence available from across the region indicates that the conflict migration is diverse. Those who arrived in the conflict zone in Syria cannot be lumped into any cohesive profile or generalization. As Almakan Orozobekova established, based on a large field project in Southern Kyrgyzstan, we see diversity and also a wide spectrum of agency or ‘dependency’ among those who migrated, men and women alike: from ‘true believers’ who recruited other women to support Daesh ideology to those who followed friends or family members with little idea what sort of cause they were joining or what awaited them there, to those whose gender, age, or other factors severely limited their agency.²⁰ Thus, we immediately encounter serious problems in the evidence with the overwhelming focus on the individual radicalization model. No matter what may happen in the process of socialization inside a Daesh or Jabhat al-Nusra camp, case studies and family interviews from across the region indicate that a large proportion of conflict migrants were not radicalized in any meaningful sense before deciding to emigrate.

Second, evidence from multiple studies across the region and publicly available data released by states indicate that while mobilization was diverse in many ways; it was not widely or evenly dispersed geographically. Rather, it was highly concentrated in a few ‘hotspots’. One small rural district in Southern Kyrgyzstan, for example, called Aravan was home to around a third of all officially recorded cases of mobilization from the entire country. While not a random sample, this small district that makes up less than 2 percent of the country’s population is home to 40 percent of the first wave of returnee children repatriated recently by the Kyrgyz government.²¹

There are analogous hotspots in each of the neighboring countries, and while the demographic and political context among them differs in important ways, all of them share a third important general characteristic: political and economic marginalization compared to other areas of each country. They are often the site of unresolved political conflicts within authoritarian states that lack any political mechanism to resolve those issues.²² In many cases within these areas, ethnic and/or religious discrimination play a large role in those unresolved political conflicts.²³

Importantly, perceptions of inequality and perceived discrimination appear to likely play a key role in explaining why demographically or economically similar districts differ sharply in their rate of mobilization. None of the key hotspots are in the poorest areas or among the poorest demographics of any state in the region. Rather, they are frequently cities or districts that enjoyed relative prosperity and prestige in the Soviet period but have suffered a sharp drop in opportunities because of political or economic changes after independence. In some cases, like the western oil fields and central mining districts of Kazakhstan, they are regions that produce significant wealth, but enjoy very few of its benefits. In these ‘factory towns’ citizens and workers have even more limited self-government or mechanisms for conflict resolution because they are in special autonomous regions (a legacy of the Soviet period) in which powerful corporations are given wide latitude in local politics and labor unions have been effectively outlawed. Beissembayev and Sharipova (2021) found in their analysis of prison and family interviews with those convicted of extremism or participation in domestic attacks from 2011–2016 frequent mentions of inequality, limited opportunities and a perceived lack of possibility for future improvement among motivating factors.²⁴

Fourth, studies of mobilization in multiple states within the region have indicated that personal networks played a major role in decisions to emigrate to the conflict zone or join armed groups operating there. Boxing clubs, extended families, labor migrants from the same city or district who rely on one another’s common identity abroad for social support, classmates and neighbors often influenced one another’s decisions or traveled together in clusters. As Orozobekova noted, members of these groups often have varying levels of dependency or agency in the decision, even when all end up in the same situation.²⁵

A Third Wave? Scenarios for Afghanistan and Central Asia

With a brief examination of the factors that we can observe within the last wave of conflict migration, we are better placed to extrapolate whether recent events risk exacerbating these factors and whether strategies to prevent that are likely to succeed. A successful insurgency in Afghanistan and an internal conflict driven by a splinter from an armed group that in the recent past has successfully recruited thousands of Central Asians make the potential for a third wave of conflict migration a possibility worth serious consideration.

Afghanistan is significantly geographically closer than Syria, and shares common languages in border regions (Uzbek, Tajik, Dari, and Turkmen). Although the rigidity of the Taliban’s fundamentalism has little to nothing in common with Islam as it has been understood historically in the region, the same cannot be said of Islam as historically practiced in Afghanistan. And while the borders with Afghanistan have long been militarized and are far less easy to cross than the Turkish border with Syria was during the peak period of conflict, opium, and other contraband smuggling are a multi-million-dollar economy in border districts and support both organized crime and corrupt state officials.

Assuming commercial travel resumes, porous borders and smugglers may become largely irrelevant for both foreign fighter recruits and migrants seeking to build a new life in the Islamic Emirate. Overall, from a superficial overview migration of either disaffected citizens or self-styled mujahedeen to Afghanistan is in many ways more feasible to the nearby Islamic Emirate than it was to the far-away Caliphate. Recent history, however, led to the opposite trend as outlined above, and suggests we should use caution in making assumptions about future behavior.

At present we have very limited data but enough to suggest that the defeat of the Daesh Caliphate and losses in territory and prestige to other groups have not exhausted the potential interest of a very small proportion of Central Asians. Governments, security services, and local researchers have continually argued that at least small groups of latent supporters exist within their states and pose a threat to domestic security and potential base of support for facilitating out-migration to another conflict zone.²⁶

Given the well-documented history of the authoritarian states of the region inflating domestic security threats to justify political repression, these warnings must always be considered in that context.²⁷ But other research approaches can offer some additional insight independent of local officials or their allies. Fieldwork conducted by the author in 2017 and 2018, for example, in areas of Central Kazakhstan that were sites of significant mobilization to Syria corroborated with anecdotal evidence that small groups of residents continued to support Daesh and threatened neighbors who spoke out against Daesh with violence (though at that time there was no evidence those threats were ever carried out). Research partners in Southern Kyrgyzstan who live and work in the most affected communities have relayed similar anecdotal accounts.

Similarly, online research indicates that messaging and support for Central Asian groups like the al-Bukhoriy Brigade or Daesh, as well as networks that show at least passive interest for their narratives continue to exist. Secondary supporters, like the Uzbek émigré ‘Shaykh’ Abdulloh Zufar, who shot to fame in the late portions of the Syrian conflict as effectively the only trained religious scholar from the region who publicly supported Daesh, continues to draw thousands of views on YouTube and has sermons shared across a variety of Telegram channels and other networks. These no longer advocate for Daesh or for armed groups, however, and focus primarily on rigid personal and family morals and other standard issues addressed in fundamentalist religious teaching. Analyses produced by some commercial data firms have argued that online extremist groups in Central Asian languages are ‘dug in’ or even ‘growing.’²⁸

There are reasons to put claims of growth into context, however. Analysis that relies on systematic data collection only since 2019 has important comparative limits. In 2012 or 2013 tech companies that hosted recruiting messages like YouTube or Facebook (the platforms most heavily used at the time) were scarcely aware of what was happening in Uzbek, Tajik, or Kazakh on their massive, global platforms. Recruiters for Daesh or other groups working in those languages were largely free to post as much content as they wanted, no matter how graphic or how badly they violated the terms of service of the providers. In those early years of the conflict during the peak of active online recruiting, messages from armed groups in Syria were so widely dispersed that they regularly appeared in popular general interest groups online, reaching a much broader audience than those who sought them out or were in direct communication with recruiters.

As content providers became aware of the issue and began to enforce terms of service even in less common language groups, profiles or groups promoting violence were rapidly deleted and became far more difficult to find. Recruiters moved in part to encrypted applications like Telegram and WhatsApp, making them more difficult to remove but also far less able to reach a new audience. Even on Telegram, accounts that openly promote armed groups or their messages are often quickly shut down, forcing their media activists to create (often several) new channels and work to transfer or rebuild their former audience. Thus, while many new online outlets continue to be created, often just as many or more are simultaneously shut down or deleted, making it very difficult to assess how many exist at any one time or how broad their audience of discrete individuals might be. When beginning to update research on current messaging from armed groups for this paper, the author found that none of the Telegram channels monitored in 2020 continued to function in 2021. Each time a core group of channels and accounts for monitoring must be reconstructed, it has become more difficult to find even official messaging outlets of armed groups that continue to exist such as the al Bukhoriy Brigade or Uzbek-language isis-K messaging.

These outlets do still exist, however, and they find an audience. Conflict fatigue or the collapse of groups in Syria does not seem to have exhausted interest or enthusiasm from a small number of Central Asians for insurgent groups and their promised utopias. It perhaps should not be surprising that a millenarian movement like Daesh would live on past the failure of its core prophecies or promises. The circumstances creating grievance and unresolved political or economic conflict – often expressed as justice issues – that resonated so widely in the Syrian mobilization have in many cases not meaningfully improved in most of the Central Asian states since that wave of mobilization ended in 2016. There are important exceptions, however. The death of former president Islam Karimov in Uzbekistan that year and his succession by a reform-minded longtime prime minister Shavkat Mirziyoyev has changed everyday life in Uzbekistan for the better for many. Similarly, the abrupt changes in messaging to welcome back returnees from the Syrian conflict and encourage the public to treat them as full citizens and re-integrate them into normal life is both positive and important since marginalization appears to have been a contributing factor to the initial Syrian migration.

Conclusion: Takeaways for Preventing a ‘Third Wave’ of Conflict Migration

Since independence, Central Asian regimes have consistently treated foreign Islamist insurgent groups as an existential security threat, often exaggerating or overestimating the physical security risk posed by externally based groups and often using the alleged threat of Islamist terrorists as a political cover for delegitimizing perceived rivals in internal disputes and even electoral competition. As Rustam Burnashev argued when isaf withdrawal from Afghanistan was initially planned, these external groups have likely never posed a serious security threat to any of the states in the region.²⁹ But a wide array of unresolved internal problems – particularly in specific hotspot regions in each state – have made some citizens vulnerable to calls by armed groups to leave their lives at home or in migration and join utopian projects these groups claim to represent. Until recently, each state in the region has treated this conflict migration or mobilization to insurgent groups as an ideological and identity issue that can be fixed essentially by policing and preventing diversity: by banning ‘bad Islam’ or what it sees as ‘excessive’ or ‘foreign’ displays of religious faith and promoting national identity and a nationalized version of Islam in its place.

These policies and practices were instituted in the first wave of conflict migration and mobilization to armed insurgencies in the late 1990s, but despite relatively continuous application they failed to prevent a much larger and wider out-migration to armed conflict zones in Syria even from states that had never experienced it before. While the international community has often referred to this migrant group as ‘foreign terrorist fighters’, in Central Asia in particular this term fails to capture the significant proportion – in some cases the majority – of conflict migrants who were women and children, often with limited or no agency, or acting on misinformation. With many of them returning and four of the states of the region committed to reintegration of former conflict migrants, the lessons of these two waves of past conflict migration should be applied to prevent another humanitarian tragedy and the potential for a strengthened insurgency based just across the Uzbek and Tajik borders.

The first of these lessons, already well reflected in the re-integration program designed and implemented by civil society groups in Uzbekistan,³⁰ is that conflict migrants were a diverse and heterogenous group who acted on a broad spectrum of motivations, dependency, and agency. Preventing further mobilization requires a localized and context-driven approach that responds to the needs of different groups and communities rather than securitizing those needs and further marginalizing them. This diversity of causal factors also demands that states and groups working to prevent future mobilization need to go beyond a one-size-fits-all approach that seeks to substitute one identity for another, or insist that citizens who derive meaning and value from Islam give priority to an ethnonationalist political identity which they may feel excludes them.

Rather than denying that these challenges exist and portraying those who sought a better life or a different kind of government as lazy, crazy, or made into ‘zombies’ by all-powerful recruiters, a more effective strategy is to acknowledge these underlying problems occur and offer citizens a mechanism they can participate in to improve them. The international consensus is that democracy and the respect for human rights and basic dignity of all citizens is exactly such a mechanism and the best one we have available. Putting political power into the hands of ordinary citizens is not an option Central Asian elites are likely to consider. It is perhaps not even possible within the political institutions that currently exist in any of the region’s states, no matter how earnestly any president or political leader advocates reform. Even so, taking a responsive and positive approach to the citizens who experience marginalization and injustice is a far better solution than policing religious diversity or punishing dissent.

The humanitarian and dignified approach taken toward returnees from the Syrian conflict by Kazakhstan, Kyrgyzstan, Tajikistan, and Uzbekistan are strong first steps in this direction and worth significant recognition. By treating the conflict migrants as a human security issue rather than a threat to regime security, each country has made a critical policy change that makes future efforts at prevention more likely to succeed. The next step is to acknowledge and address community-level factors – rather than an exclusive focus on individual pathways – that helped create the crisis in the first place and can help counter the potential for a new wave of conflict migrants to a much closer phantom utopia next door.