The civil war in Syria (2011–present) unleashed the largest displacement crisis in a generation. This chapter analyzes the strategies and tactics employed by the regime of President Bashar al-Assad and its allies, which have, according to the UN and other observers, embarked on a deliberate campaign to uproot civilians through punishing airstrikes, the targeted shelling of civilian infrastructure, besiegement, and forced evacuations.Footnote 1 While many observers describe the government’s use of displacement in Syria as sectarian cleansing against the country’s Sunni population, patterns of violence indicate that most state-induced displacement has taken the form of depopulation. This case therefore permits a closer comparison of the use of these two different displacement strategies.
I show that, while some patterns of violence and displacement orchestrated by pro-government forces indicate cleansing, they occurred early in the conflict in religiously mixed and non-cosectarian enclaves. Cleansing was also employed under certain conditions: when perpetrators were acquiring or consolidating control over a given territory, which allowed them to collectively identify and remove individuals through direct, face-to-face violence. The evidence suggests that collective targeting was often driven not by sectarian animus but rather a tendency to use sectarian identity as an indicator for political loyalty. This is consistent with the idea that cleansing follows a logic of punishment.
Depopulation, however, often occurred in territories that perpetrators did not control. It was triggered by the indiscriminate use of indirect violence – shelling and airstrikes – as opposed to more intimate violence. These trends are consistent with what conflict scholars would expect in general patterns of wartime violence, which tends to be more discriminate and direct in areas where combatants exert some control.Footnote 2 However, since this violence often sought to induce displacement, did depopulation serve the same function and follow the same logic as cleansing? By focusing on depopulation, this chapter provides an opportunity to examine a type of strategic displacement that has received little attention in this book. In this chapter I explore the scope conditions of my theory in a case that poses a difficult test of the argument. I have shown in Chapters 4–6 that sorting displacement tends to be employed in irregular conflicts, but Syria is a conventional civil war, with different sides engaging in artillery battles across clearly defined front lines and controlling significant swathes of territory.
The Syrian government’s efforts to depopulate rebel-held areas seem to have been partly motivated by a logic of denial: To deprive the opposition of the civilians and infrastructure it needed to build a viable counter-state. Yet this explanation paints an incomplete picture, as it overlooks other aspects of the regime’s strategy and fails to account for what happened after people were uprooted. In many instances, after regime forces reasserted control over former rebel-held territories, civilians and combatants who remained were given a choice: either to stay in regime-controlled territory or to relocate to areas that remain under rebel control. This is puzzling. Researchers of civil wars tend to assume that combatants, particularly incumbents, either try to exert control over the entire population of a territory or attempt to expel it in order to eliminate any potential opposition. Why has the government routinely offered people a choice of destination?
I argue that these actions reflect a desire by the Assad regime to sort the population and weed out the disloyal, for which triggering civilian flight has served as a critical tool of differentiation. I demonstrate that, in employing depopulation, the government used displacement not only to remove civilians from specific areas but also to lure them into its territories. While the regime’s cleansing campaigns seemed designed to get rid of the targeted population, its depopulation methods had an assortative element. Pro-government forces made inferences about people’s affiliations and allegiances based on whether they abandoned areas targeted for depopulation, and whether they moved to regime or rebel-held territory. The state also conscripted those who fled to its areas to help bolster the manpower of the depleted Syrian army, while using their arrival as propaganda and a source of legitimization. This challenges the idea that displacement in Syria has primarily been a strategy for demographic change. It also demonstrates that depopulation can share some of the same sorting and capturing logics as forced relocation, at least in certain cases.
7.1 The Syrian Civil War
In March 2011, pro-democracy protests erupted in the southern province of Dera’a, Syria, after authorities tortured a group of teenagers caught spraying anti-regime graffiti. Demonstrations calling for political and economic reform quickly spread to the central, northern, and eastern parts of Syria, in addition to suburbs surrounding the capital, Damascus. The protests largely comprised members of Syria’s Sunni majority, while the regime of Bashar al-Assad is dominated by Alawites, a sect of Shi’a Islam. But the protestors, buoyed by the Arab Spring revolts in Egypt, Tunisia, and Libya, emphasized their democratic, nonsectarian, and largely secular values. Syrian security forces responded to the uprising by arresting and firing on participants, which fueled public resentment and galvanized the opposition. Months of suppression, coupled with only limited promises of political reform by Assad, transformed a peaceful revolution into a violent rebellion intent on overthrowing the regime.
Armed insurgency first broke out in July 2011. A wave of resignations from the ruling Ba’ath Party and the desertion and defection of members of the Syrian Arab Army (SAA) supplied tens of thousands of fighters to the opposition, which mobilized a variety of rebel brigades under the banner of the Free Syrian Army (FSA). By the end of 2011, the FSA and its affiliates were launching regular ambushes against government and military targets around the country. In 2012, the conflict became a conventional war as rebel factions started directly confronting and overrunning SAA positions, taking control of towns and villages in the provinces of Aleppo, Idlib, Rural Damascus, and Homs. The war took on a sectarian character as the regime sought to portray the opposition as foreign-backed Islamist extremists. Assad increasingly stoked fears that Alawite and other minority communities would be subjected to violent reprisals should he be overthrown. Meanwhile, the failure of FSA brigades to unite under a single leadership structure provided an opening for the emergence of Salafi jihadist groups, including Al-Qaeda affiliate Jabhat al-Nusra (JAN).
Despite the growing fractionalization of the insurgency, rebels continued to make advances from early 2012 to mid-2013, often confronting regime forces directly in pitched battles. External actors began to pour money, arms, and fighters into Syria, as Saudi Arabia, Qatar, the United Arab Emirates, Turkey, and the US backed the opposition, while Shi’a fighters from Iran’s Islamic Revolutionary Guards and Lebanese Hezbollah provided battlefield support to the Syrian government. A series of UN-backed peace talks and other meditation initiatives beginning in 2014 eventually failed. Infighting and poor coordination further weakened the FSA, and hardline groups like JAN became the insurgency’s potent fighting forces. In the summer of 2014, ISIS broke away from its affiliates in JAN and seized parts of eastern Syria and western Iraq, declaring the Syrian city of Raqqa the capital of its Islamic caliphate. A suspected chemical weapons attack by regime forces in the suburbs of Damascus in August renewed calls for international military action in Syria.
As a coalition of NATO countries began launching airstrikes against ISIS territories, the rebels made further gains, taking the provincial capital of Idlib City in March 2015. The SAA was forced to strategically withdraw from some towns and seemed increasingly imperiled until the Russian government, at Assad’s request, dispatched its air force in September to help Syrian forces repel rebel advances and retake parts of Aleppo, Homs, and Rural Damascus. The regime subsequently went back on the offensive. In early 2016, the Syrian Kurds and its militia (the People’s Protection Units, or YPG), who had filled the political vacuum in parts of northern Syria by establishing its own autonomous governing areas, joined forces with Arab, Assyrian, and Turkmen groups to form the Syrian Democratic Forces (SDF). With support from an American-led coalition, the SDF began steadily pushing ISIS out of its territories in northern and eastern Syria, recapturing Raqqa in September 2017. Alarmed by the advances of the Kurdish units, Turkey – which continued to fight a rebellion by its own Kurdish separatists, the PKK – launched Operation Euphrates Shield, occupying parts of northern Syria to stem SDF encroachment and fortify the Turkish border.
Meanwhile, the Syrian regime, after expelling rebels from Aleppo City in December 2016, made steady territorial gains throughout 2017. The US began to soften its opposition to Assad. Still, after a chemical weapons attack in rebel-held Khan Shaykhun in April 2017, American forces bombarded a Syrian air base with fifty-nine cruise missiles. One year later, the US, Britain, and France responded to another chemical attack in Rural Damascus by striking alleged chemical weapons facilities near Damascus and Homs. These limited interventions made little difference on the ground, however. The SAA and its allies continued to wipe out rebel strongholds around Damascus, in southern Aleppo, and in northern Hama, prompting Russia to announce its partial withdrawal from the conflict at the end of 2017. In June 2018, having solidified their hold on Damascus and Homs, Syrian forces launched a successful operation to recapture rebel-held territories in the southern provinces of Dar’a and Quneitra. As the US withdrew most of its forces from Syria in 2019, and Turkey continued to launch airstrikes in the Kurdish-controlled northeast, the regime set its sights on the remaining rebel stronghold of Idlib province in the northwest. In 2020, Russia and Turkey brokered a ceasefire that stopped Syrian forces from retaking Idlib, but air strikes and shelling continued. A powerful earthquake in early 2023 compounded the devastation in northwest Syria and aggravated the humanitarian crisis in Idlib. As of 2024, with most of the country back under government control, the conflict had reached a stalemate, having claimed an estimated 500,000 lives.
7.1.1 Overall Patterns of Displacement
The high death toll of the civil war in Syria is matched by massive civilian displacement. Half of the country’s population has been uprooted, including more than five million refugees and another seven million internally displaced (Figure 7.1). While many of the displaced fled spontaneously, others were deliberately forced to leave their homes as part of a calculated strategy by parties to the conflict. According to the IDMC, “the extent of displacement and widespread destructions of homes and infrastructure” in Syria “were not collateral damage from the fighting, but resulted instead from the protagonists’ deliberate actions … [of] targeting civilians with the aim of forcing them to evacuate from certain areas.”Footnote 3 Pro-regime forces have been the primary perpetrators of strategic displacement in Syria. Other actors in the conflict, including ISIS and the Kurdish YPG, have been accused of ethnic cleansing. Yet as with the rest of this book, I focus on displacement by state actors.
Figure 7.1 Forced displacement in Syria, 2010–2022
7.2 Methods and Data Sources
This case study relies on quantitative and qualitative data from a diverse array of sources, including my own field research in Syria, Turkey, and Lebanon. Each set of data has potential biases and limitations. When combined, however, they allow me to triangulate information and make powerful arguments about the strategic logic of displacement in Syria. I confine my analysis to the period between 2011 and 2018 – the decisive phase of the conflict before the government regained control of most of the country and appeared on the cusp of a military victory.
I use the data to analyze patterns of displacement-inducing violence by pro-government forces. In demonstrating that strategic displacement took different forms and served different functions, I focus on three patterns. The first is how displacement was carried out: initially through direct violence and collective targeting, and then through indirect violence and indiscriminate targeting. The second is the demographic result of displacement: It made regime-held areas less ethnically and religious homogenous, not more. Finally, I analyze government responses to displacement by examining how flows of IDPs influenced subsequent regime violence against civilians. In general, these findings are consistent with my argument that displacement has often not amounted to ethno-sectarian cleansing or followed a logic of collective punishment.
After examining these patterns, I evaluate how well different arguments account for them. I use process tracing to explore different stages and dimensions of the conflict and illuminate different aspects of the regime’s strategy. Before proceeding, however, a brief caveat is needed. When this chapter refers to “the regime” or “pro-government forces” it encompasses multiple groups that have comprised, and fought alongside, the government of Syria. This includes the SAA, intelligence agencies (mukhabarat), and state-sponsored militias (shabiha), along with non-Syrian militias from Iran, Iraq, and Lebanese Hezbollah, which have provided ground support. Since the focus of my analysis is on displacement induced through airstrikes and barrel bombs – the general province of the military – it is reasonable to treat the regime as a unitary actor in this instance. I also include in this definition airstrikes launched by Russian forces after their intervention in late 2015, under the assumption that they coordinated these attacks with the Syrian government.
7.2.1 Quantitative Data
Several independent organizations have documented violence in Syria since the onset of the conflict. The largest data collection effort has been conducted by the Violations Documentation Centre (VDC), a network of Syrian activists who maintain an online database of individual victims of violence, detainees, and missing people. While VDC researchers are affiliated with the opposition, they have documented human rights violations by both pro-government and rebel forces, disaggregated by the date and cause of death (shooting, shelling, airstrikes, etc.) between 2011 and 2018. An independent analysis for the UN found that the VDC database contains the most extensive and detailed identifiable records of killings in Syria compared to those compiled by other organizations, such as the Syrian Observatory for Human Rights (SOHR).Footnote 4
A limitation of the VDC data, however, is that it loses accuracy at lower levels of analysis than the province (Administrative Unit or ADM-1). I therefore kept my analysis at the province level to balance the value of subnational disaggregation with data quality. I also built a novel dataset of airstrikes in Syria from mid-2011 to mid-2017. Working with two Syrian researchers, I triangulated deaths from aerial attacks documented by VDC with reports from the Syrian Network for Human Rights, SOHR, Airwars, and the Institute for the Study of War – which produces periodic maps of airstrikes in Syria – along with videos and reports from social media to generate daily airstrike data at the province level.Footnote 5
Recall from Chapter 2 that strategies of depopulation differ from strategies of cleansing in both the type of targeting (indiscriminate versus collective) and the type of violence (indirect versus direct) used to induce displacement. As I explain in more detail later in the chapter, the indiscriminate and indirect nature of air attacks by pro-government forces – which various observers have characterized as intended to uproot civilians – make them a suitable proxy for depopulation. To capture potential episodes of cleansing, I used data on atrocities collected by PITF.Footnote 6 PITF tracks all reported massacres of civilians in Syria, with a death threshold of five people, carried out through direct and deliberate violence. Cleansing is often triggered or accompanied by these kinds of massacres. I focus on a particular subset of killings committed by pro-regime actors: those that, according to PITF, were (1) not categorized as collateral damage and (2) entailed “scorched earth” tactics – which are particularly likely to induce displacement.
For displacement, I pulled data from several sources. For aggregate annual flows, I relied on UNHCR, which tracks the number of registered Syrian refugees worldwide, along with data on IDPs collected by the IDMC.Footnote 7 I also built a subnational dataset on monthly IDP flows using data from the UN Office for the Coordination of Humanitarian Affairs (UNOCHA). The data include the origin and destination province for all recorded internal displacement flows, by month, from January 2016 to December 2018. While the data are limited in temporal scope and (like VDC) are only available at the province level, they are the most fine-grained displacement data available for multiple years of the Syrian conflict.
Finally, I incorporated data on territorial control in Syria from the Carter Center, which analyzes social media to track changes in the control of territory by different actors across the country.Footnote 8 The Carter Center captures monthly shifts in territorial control between January 2014 and December 2018 by four main parties to the conflict: the Syrian government, the armed opposition, Kurdish forces, and ISIS. The data delineate control down to the city or village level (ADM-4). In order to aggregate it up to the province level, I calculated the percentage of communities controlled by the regime in each province in each month. In addition to the raw percentage, I created a scaled variable of territorial control ranging from 1 (little to no regime control) to 5 (full to dominant regime control).Footnote 9
7.2.2 Qualitative Data
Qualitative data from various sources during the Syrian war have been gathered by human rights activists, scholars, journalists, humanitarian organizations, and other observers. Many sources describe patterns and dynamics of violence and displacement, and contain statements from perpetrators and victims – including a number of interviews with government and rebel combatants, regime officials, community leaders, refugees, and other civilians. Some primary documents, including evacuation orders issued by the Syrian army, were published on social media. These disparate sources offer a substantial body of evidence for my analysis.
This case study also draws on original in-depth, semi-structured interviews with thirty-one people, several of whom were interviewed multiple times, that I conducted during fieldwork in Turkey, Germany, Lebanon, and Syria in 2016, 2017, and 2019. I conducted the interviews in English and Arabic, with the assistance of several Syrian translators. My subjects included Syrian activists, journalists, regime defectors, and former combatants, along with knowledgeable experts on the war in Syria such as humanitarian and development aid workers, local and international academics and researchers, and former diplomats. I arranged interviews through several contacts, including a Syrian journalist, a group of former human rights activists, and Syrian and American staff of humanitarian organizations active in Syria. Since I lived in Turkey for two years (2015–2017) and served as a consultant for US government aid programs in Syria, I developed an extensive network of personal and professional contacts that proved critical for accessing informants.
Interviews were conducted in Gaziantep, Istanbul, Beirut, and several cities and villages in northeast Syria. Most Syrian subjects were refugees or asylum seekers. Many of them had spent years living under regime and/or opposition rule during the war, and they maintained constant communication with family, friends, and colleagues inside Syria and closely monitored local developments. Some of these individuals openly express their political views online or as part of their professional endeavors and have even written articles and editorials under their real names. However, given the risks involved, I conducted all interviews on the condition of anonymity. Ensuring confidentiality was essential as some of my respondents are wanted by the regime or other armed groups, or they were in the midst of processing asylum applications.
7.3 Displacement by Pro-government Forces
As noted earlier, multiple observers of the Syrian conflict characterize the regime’s strategic use of displacement as ethno-sectarian cleansing.Footnote 10 If this observation is correct, it has empirical implications regarding the target of displacement (specific identity or political groups), the form it should take (expulsion), and the results of it (ethno-sectarian homogenization). My sorting theory, by contrast, suggests that targeting should be more indiscriminate, displacement should result in more social heterogeneity, and the government should respond to displacement by treating civilians differently based on their movements, not just their identities. I use multiple types of data and research methods to explore these observable implications.
7.3.1 Targeting of Displacement
Some population displacement orchestrated by pro-government forces in Syria has indeed been consistent with cleansing. Early in the conflict – between mid-2011 and mid-2013 – the SAA and affiliated militias systematically expelled residents of Sunni enclaves in the Alawite-majority coastal provinces of Tartous and Latakia. These expulsions occurred during “coordinated clearance” operations in which regime fighters entered villages and proceeded house to house, massacring residents and ordering the rest to flee.Footnote 11 In Damascus between June and September 2012, and in Hama in September 2012 and May 2013, government forces razed large sections of neighborhoods occupied by rebel fighters. While no one was reported injured or killed, the army “used megaphones and told residents they had one hour to pack their things.”Footnote 12
These displacements appeared to stem from a desire to punish the affected populations. They occurred after prolonged battles in which regime forces regained or consolidated control over the towns and neighborhoods in question. For example, the demolitions in Damascus started after both state and opposition sources confirmed that government forces “had largely regained control of” the capital.Footnote 13 Expulsions also concentrated on areas from which rebels had launched attacks and seemed to selectively or collectively target people based more on their political than their sectarian affiliations – or because sect was used as a proxy for opposition affiliation. While the evacuated areas were overwhelmingly Sunni, they had allegedly been used by opposition fighters, and other majority Sunni neighborhoods nearby were not targeted for demolition, such as Al-Midan, Joubar, and Al-Qadam in Damascus. For instance, in 2012, according to Amnesty International, thousands of civilians were forced from villages in Aleppo and Idlib after security forces destroyed 1,500 properties in “an obviously deliberate manner” that indicated “premeditation.” Yet Amnesty reported that “some, possibly many, of the homes that were burned down or otherwise targeted belonged to [anti-regime] activists … or to people who had become fighters with the opposition.”Footnote 14 After demolishing Wadi al-Jouz in Hama in May 2013, the army “warned residents in other neighborhoods that their houses would also be demolished if opposition fighters attacked government forces from these neighborhoods.”Footnote 15 It is also clear that the perpetrators focused on expelling victims on a permanent basis – as indicated by leveling their homes – and made no effort to encourage or order them to relocate to regime territory.
Figure 7.2 shows the geographical distribution of scorched earth massacres perpetrated by pro-regime forces in Syria, according to the PITF data. PITF identifies thirty instances of such killings between June 2011 and May 2013. As the map demonstrates, they mostly occurred in ethnic or religiously mixed areas, or in Sunni enclaves in Alawite strongholds. Local reports indicate that these massacres were intended to uproot the resident population. Those that occurred in the provinces of Tartous and Latakia, for example, were “intended to displace a civilian population that was perceived as supportive of the opposition from an area that a government minister characterized as ‘very sensitive’” since it bordered the Alawite heartland.Footnote 16 Likewise, the massacres in the city of Homs largely occurred in Sunni neighborhoods that bordered Alawite enclaves and appeared “to be conducted as part of a state policy” that was intended “to make civilians flee.”Footnote 17 Similar atrocities were carried out in al-Houla and Qubeir in Homs province – both Sunni villages that bordered Alawite ones.
Figure 7.2 Scorched earth massacres in Syria, 2011–2013
However, according to Eline Bostad, despite the “discriminate nature” of these operations, “it was not their sectarian identity per se that made the Sunni population targets.” Rather, in these heterogeneous territories, “sectarian identities served as particularly potent proxies for the opposition.”Footnote 18 Syrian activist Rifaie Tammas similarly regards these acts of violence as “politicide” because they were meted out against “a particular segment of Syrian society due to its pro-opposition affiliation.”Footnote 19 Indeed, these cleansing campaigns targeted opposition strongholds in militarily strategic areas between Damascus and Hama, a central transport hub linking different parts of Syria This suggests an ex ante determination of guilt. The al-Bayda and Ras al-Nabe areas of Baniyas in Tartous were, for example, “notoriously pro-revolution,”Footnote 20 while Homs and Darayya were among the first to join the Syrian uprising. The town of al-Qusayr in Homs, from which Sunnis were also expelled, had become a well-known haven for regime defectors.
In some cases, operations by Christian and Alawite militias to drive Sunnis from their homes appeared to be idiosyncratic, opportunistic, and uncoordinated. As one analyst has argued, “observers should be careful to distinguish cases like [these] – where local minorities engaged in ethnic cleansing out of revenge or perceived self-preservation – from cases in which central government forces have expelled rebel populations as part of a deliberate military strategy.”Footnote 21 Instances of strategic cleansing, such as those described here, were primarily carried out through direct violence – including individual and mass executions, beatings, physical intimidation, and the bulldozing of homes – and in many ways they reflected the regime’s initial counterinsurgency approach. Early in the conflict, Syrian forces relied on cordon-and-search tactics, conducting one major operation at a time by moving from town to town in a “proactive targeting and detention campaign.”Footnote 22 Cleansing mostly occurred as part of these operations or during incursions by pro-regime militias, and were likely part of a concerted effort to punish agitators and root out armed opponents.Footnote 23 Executing these expulsions required a degree of territorial control and access to the population, and they transpired, as I mentioned earlier, after regime forces had either repelled a rebel advance or retaken a contested area. These micro-level patterns are consistent with the macro-level findings from Chapter 4. Cleansing was typically the result of Syrian forces trying to conquer territory in socially heterogeneous areas and using sectarian identity as an indicator of disloyalty.
7.3.1.1 From Direct to Indirect Violence
Yet the bulk of displacement during the Syrian war was triggered not by direct forms of violence but rather by pro-regime forces’ use of indirect violence, including airstrikes, barrel bombs, and heavy shelling.Footnote 24 The SAA placed greater emphasis on these methods beginning in mid-2012, when it began to regularly deploy helicopter gunships and fighter jets.Footnote 25 This came soon after the rebels opened new fronts in Aleppo and Latakia, which were “beyond the reach of overstretched ground troops” engaging insurgents elsewhere.Footnote 26
The use of barrel bombs – highly imprecise and destructive explosives – was first reported in August 2012. In the ensuing months, as the number of civilians killed by pro-government forces through face-to-face violence – namely, shooting – plummeted, the number killed by airstrikes and shelling increased. This produced a corresponding spike in population displacement, as Figure 7.3 illustrates. While the figure indicates a lag between deaths from indirect violence and displacement, this is consistent with other research on wartime migration, which does not find a significant relationship between violence timing and displacement timing.Footnote 27 The number of registered Syrian refugees worldwide doubled from 250,000 to 500,000 between September and December 2012, then doubled again three months later, reaching almost four million by the end of 2014. The number of Syrians reported to be internally displaced also surged to more than seven million during this period.
Figure 7.3 Monthly deaths from violence by pro-regime forces, 2012–2018
The sharp rise in displacement can be directly linked to the government’s increased reliance on airstrikes and other forms of indirect violence, which became “the most significant instrument in the regime’s efforts to displace populations.”Footnote 28 Indeed, while 94 percent of rebel combatants killed by pro-regime fighters between 2011 and 2018 died by shooting, airstrikes and shelling accounted for 65 percent of civilian casualties inflicted by the government during this period (see Figure 7.4). Such attacks were widespread and indiscriminate, often reflecting the use of “imprecise and unguided munitions with wide-area effects”Footnote 29 that “could not distinguish between civilians and combatants.”Footnote 30 Pro-government forces typically targeted civilian infrastructure – hospitals, bakeries, and schools – instead of military assets. The Syrian air force hit a range of rebel-held locations beyond the most contested or vital strongholds, including places “where there were no reported clashes that day, suggesting that the airstrikes were not in tactical support of Syrian Army units fighting rebels.”Footnote 31 In fact, many bombed towns and villages “had seen no recent ground fighting at the time of the attacks,” according to Human Rights Watch.Footnote 32 As Table 7.1 illustrates, not only did the incidence of airstrikes double between 2013 and 2014, the proportion of airstrikes launched by regime-aligned forces that were accompanied by ground troops declined substantially from 58 percent in 2012 to 20 percent in 2016. Thus, rather than paving the way for subsequent ground offensives, these aerial assaults began to replace ground offensives altogether.Footnote 33
Table 7.1 Regime airstrikes accompanied by ground forces, 2012–2016
| Ground forces | No ground forces | % Ground forces | Total | |
|---|---|---|---|---|
| 2012 | 623 | 454 | 58 | 1,077 |
| 2013 | 515 | 884 | 37 | 1,399 |
| 2014 | 810 | 2,011 | 29 | 2,821 |
| 2015 | 677 | 1,474 | 31 | 2,151 |
| 2016 | 541 | 2,099 | 20 | 2,640 |
| Total | 3,166 | 6,922 | 35 | 10,088 |
Figure 7.4 Deaths due to violence by pro-regime forces, 2011–2018
As the regime sought to confront an increasingly multipronged insurgency across a growing number of front lines, it concentrated its forces on retaking and securing provincial capitals and other strategic population centers. Yet its use of indirect violence persisted in areas where it lacked and made little effort to deploy the troops necessary to capture and hold territory contested or occupied by rebels. Instead, in these areas – particularly the northern border provinces of Aleppo and Idlib – state forces launched rampant and relentless air campaigns designed to trigger civilian flight. Even as the government began to recapture towns and cities from rebels in mid-2013, it continued to engage in pitched battles in strategic areas while relying on airstrikes elsewhere.Footnote 34 Even places considered less strategic for the regime – such as the Hama countryside or eastern provinces such as Deir ez-Zor – were not immune from attack. This indiscriminate violence has not just victimized members of Syria’s Sunni population. For example, the Syrian Network for Human Rights has documented the destruction of many Christian sites of worship by government forces in contested and rebel-held areas. According to the group, while the Assad regime “has portrayed itself as ‘the protector of minorities’ especially Christians … barrel bombs do not differentiate between Christians and non-Christians,” demonstrating that the government “is willing to target anybody … regardless of their religious or ethnic affiliations.”Footnote 35
Taken together, the data presented in this section suggest that between mid and late 2012, the regime altered its approach from one of direct violence to one that prioritized indirect violence. This shift corresponded with (1) a massive increase in population displacement and (2) a seeming decline in ethnic and political cleansing. The regime used direct violence to penetrate and consolidate control over contested territories. But the fact that its indirect violence was widespread, highly indiscriminate, and often unaccompanied by ground forces indicates that this violence – and the massive displacement it produced – served as a substitute for territorial occupation, at least in the short term, rather than a means of achieving it. This shift from direct to indirect violence, and from selective to nonselective targeting, is consistent with existing theories of civil war violence that emphasize the importance of territorial control.Footnote 36 That the regime resorted to more indiscriminate violence as the rebels captured more territory and became a greater threat is therefore not surprising. Yet to the extent that this violence was used to intentionally uproot civilians, how do we understand potential shifts in the function that displacement served for the Syrian government? Was it simply intended to eliminate or punish the targeted population? Or did it serve another purpose – one that differed from the logic underlying the cleansing campaigns described earlier?
7.3.2 Results of Displacement: Mixing or Unmixing?
If the regime’s displacement tactics amounted to sectarian cleansing, we would expect them to result in a significant “unmixing” of the population by sect. Sunnis would comprise the bulk of residents in rebel-held areas and non-Sunnis would comprise the majority of the population in regime-held territory. Before the conflict, Arab Sunnis made up approximately 65 percent of the Syrian population. By 2017, that number had declined modestly to 61 percent. Likewise, the share of Alawites in Syria, the sect to which Assad belongs, increased from roughly 10 percent to 13 percent during the war.Footnote 37 Therefore, despite the fact that Syria has hemorrhaged a substantial number of refugees, any purported effort to drain the country of its Sunni population has not been very effective.
In fact, in areas controlled by the regime, population displacement increased their level of social heterogeneity. Fabrice Balanche estimates that, as of June 2017, Sunnis constituted at least 58 percent of those living in regime territory.Footnote 38 The influx of IDPs into government strongholds has in many instances led to an intermixing of the population. In Latakia, the proportion of Sunnis increased from 40 percent before the conflict to 50 percent in 2013.Footnote 39 Tartous, which was 90 percent Alawite at the start of the uprising – and where thousands of regime loyalists marched in 2011 to chants of “Assad, or we burn the country” – fell to 60 percent Alawite by late summer 2012.Footnote 40 By 2014, IDPs made up 52 percent of the governorate’s population.Footnote 41 As a result, regime areas have “the most diverse sectarian mix in Syria, welcoming IDPs from all denominations.”Footnote 42 So when Assad declares, as he did in a 2017 speech, that Syria has gained “a healthier and more homogenous society” during the war, he appears to be extolling homogeneity along political lines – not sectarian ones.Footnote 43 Or as one Syrian general reportedly told his security officers in 2018, “a Syria with 10 million trustworthy people obedient to the leadership is better than a Syria with 30 million vandals.”Footnote 44
7.3.3 Government Responses to Displacement
If my argument is correct and the regime used displacement in part to draw people to its territories – and not just to expel them – then state forces should kill fewer civilians following IDP inflows. On the other hand, IDP outflows from an area should lead to more government killings because the regime aims to deter flight and perceives those who flee as potential traitors who should be punished. To test these propositions, I combined data from VDC and UNOCHA to create a subnational dataset of civilian fatalities and displacement flows. The dataset features fatalities and IDP inflows and outflows, by month, at the province level from January 2016 to December 2018. While this time period falls within the later stages of the conflict, it still provides an opportunity to test some of my claims more systematically, as Syria has experienced high levels of displacement and lethal violence throughout the war.
For this analysis, the outcome variable of interest is the number of fatalities due to regime violence, both combatant and civilian (regime killings [all]), and the number of civilian fatalities only (regime killings [civilians]), as reported by VDC. The independent variables are the number of IDP inflows into a province and the number of IDP outflows from a province during a given month. I also controlled for the number of within-province IDP flows (in-province flows). Finally, I included two control variables that could also influence the level of displacement and violence against civilians. The first is a measure of the degree of territorial control exercised by the regime in a given province during a given month, according to the Carter Center.Footnote 45 This variable equals the total percentage of communities controlled by the government in a province. The second control variable is the number of combatant fatalities in a given province in a given month, which serves as a measure of battle intensity.
To explore potential relationships between these variables, I conducted a regression analysis, the results of which are displayed in Table 7.2.Footnote 46 Controlling for the degree of territorial control and battle intensity, I find that IDP inflows lead to a decrease in regime killings, including total fatalities (Model 1) and civilian fatalities only (Model 2). IDP outflows, meanwhile, lead to an increase in killings by government forces, and both results are highly statistically significant (p < 0.01). This is consistent with the notion that the regime has (1) discouraged civilians from leaving its territories and punished those who have, and (2) welcomed civilians who have entered, on average.Footnote 47
Table 7.2 Civilian killings in Syria, January 2016–December 2018
| Model 1 Regime killings (all) | Model 2 Regime killings (civilians) | Model 3 Rebel killings (civilians) | Model 4 Regime killings (all, lagged) | Model 5 Regime killings (civilians, lagged) | |
|---|---|---|---|---|---|
| IDP inflows | −0.03Footnote *** (0.01) | −0.04Footnote *** (0.01) | 0.01 (0.02) | −0.04Footnote *** (0.01) | −0.04Footnote *** (0.01) |
| IDP outflows | 0.05Footnote *** (0.02) | 0.06Footnote *** (0.02) | 0.09Footnote *** (0.03) | 0.06Footnote *** (0.02) | 0.05Footnote *** (0.02) |
| Within-province flows | 0.00 (0.01) | −0.01 (0.01) | −0.07Footnote *** (0.03) | 0.00 (0.01) | 0.00 (0.02) |
| Territorial control (%) | −0.00 (0.00) | −0.00 (0.00) | −0.01Footnote *** (0.00) | 0.00 (0.00) | 0.00 (0.00) |
| Combatant fatalities | 0.02Footnote *** (0.00) | 0.02Footnote *** (0.00) | 0.01Footnote *** (0.00) | 0.01Footnote *** (0.00) | 0.01Footnote *** (0.00) |
| Constant | −0.67Footnote *** (0.18) | −0.80Footnote *** (0.19) | −0.90Footnote ** (0.37) | −0.92Footnote *** (0.18) | −0.93Footnote *** (0.20) |
| Observations | 498 | 463 | 390 | 485 | 450 |
Note: Standard errors in parentheses;
** p < 0.05, *** p < 0.01.
It is possible, however, that I have the relationship backward. Since violence obviously affects displacement, it could be that fatalities are leading to more or less IDP flows, rather than the other way around. While I am reassured by existing research finding no significant relationship between the timing of violence and displacement,Footnote 48 I took two steps to address these concerns. First, I lagged the main dependent variables (regime killings [all] and regime killings [civilians]) by one month, reran the analysis, and found similar results.Footnote 49 Second, I conducted a separate regression analysis with the number of civilian killings by rebel forces as the dependent variable (Model 3 in Table 7.2). If my findings simply reflect more violence against civilians producing more displacement, then the results for regime and rebel killings should be similar. Although this is indeed the case for IDP outflows, I find that IDP inflows are positively associated with rebel killings (and statistically insignificant), which diverges from the findings for regime-induced fatalities.
These tests do not fully resolve concerns about reverse causation. The results only indicate a correlation between IDP flows and violence in Syria, and the data have inherent limitations since they only capture monthly trends. But overall, this analysis finds some evidence that government forces tended to kill fewer people following the inflow of IDPs into a province and more people following the outflow of IDPs from a province. The results hold even when accounting for the intensity of fighting and the extent of territorial control exercised by the regime. While these findings should be met with caution, they indicate that the government welcomed displaced civilians moving to its territories and inflicted violence on those who have left to other parts of Syria. This aligns with the expectations of my sorting theory but is not consistent with idea that state-induced displacement has sought to eliminate the targeted population.
7.4 Explaining Depopulation in Syria
Early in the Syrian civil war, strategic displacement by pro-regime forces focused on expelling members of the opposition and their perceived sympathizers. Sectarian identity was sometimes used as a proxy for disloyalty, providing a basis for the collective expulsion of Sunni populations. These expulsions helped the government conquer key territories, primarily Sunni enclaves in Alawite strongholds and religiously mixed cities and towns. Such acts of cleansing were carried out through direct methods of violence – summary executions, threats, or destruction of property – and occurred as the perpetrators solidified control over strategic areas.
But as the battlefield expanded, the number of rebel groups proliferated, and the regime began to suffer from troop shortages, it increasingly relied on indirect violence. As a result, targeted cleansing operations declined in favor of more widespread and indiscriminate campaigns to depopulate rebel-held areas. Observed patterns in the targeting of displacement, its demographic results, and the government’s responses to it suggest that the strategic logic of depopulation differed from the punishment-based logic that drove cleansing. Next, I examine support for three possible explanations for depopulation: ethnic nationalism, where the principal motivator is sectarianism and exclusionary beliefs; rebel threat/desperation, where displacement is a tactic of last resort; and my sorting theory, where displacement seeks to identify disloyal civilians and provide material and symbolic resources to the state.
7.4.1 Ethnic Nationalism
By challenging the characterization of regime-induced displacement in Syria as a uniform strategy of sectarian cleansing, the evidence presented earlier casts doubt on explanations that emphasize ethnic nationalism or religious enmity. Even those episodes that can be described as cleansing appear to have stemmed from the use of sect as an indicator of potential rebel affiliation, as opposed to some deep-seated animosity toward Sunni Arabs. The Assad regime certainly stoked sectarian tensions during the conflict in order to galvanize support among Syria’s minority communities. Some of its allies – namely Iran and Lebanese Hezbollah – harbor a more nakedly sectarian outlook, and Iran has reportedly encouraged the regime to alter the demographics of some areas by providing citizenship and housing to Shi’a foreigners, including militiamen.Footnote 50 But sect is only one of multiple fault lines in the Syrian conflict, along with class, ideology, and tribe; a fact that has caused observers to classify the war as “semi-sectarian.”Footnote 51 According to the International Crisis Group, unlike concomitant conflicts in neighboring Iraq and Lebanon, the opposition in Syria was “a broad-based popular movement predicted on cross-communal solidarity, national revival and collective aspirations.” In this environment, “communal loyalties coexist with local identities, kinship ties, class affinities, ideological preferences, generational cleavages and pan-Arab and pan-Islamic sympathies, as well as strong attachment to the ‘nation-state’ in its existing boundaries.”Footnote 52
Moreover, while the regime has long privileged Alawites – appointing a disproportionate number to senior government positions – Sunnis remain fixtures of the country’s economy and security apparatus.Footnote 53 Syria maintains a nonsectarian, secular constitution, which many people I interviewed made sure to remind me regardless of whether they were government or opposition supporters. Before the war, Bashar al-Assad, who took over the presidency from his father Hafez in 2000, offered significant opportunities to Syria’s Sunni majority. According to Charles Lister, the younger Assad “presided over a partial revival of Sunni Islam within state-accepted circles” and established “friendly and eventually rather cozy relationships with moderate Sunni leaders, who were duly installed in positions of authority.” This resulted in a “gradual integration of Sunni Muslims into the spheres of officialdom.”Footnote 54 Cultivating support across religious sects and tribes in Syria has been a vital governing strategy for the regime; one that has helped ensure its survival.
The war has done little to change that. As explained by Thanassis Cambanis, “many rich Sunni industrialists serve as pillars of the regime,” which “still relies on Sunnis to fill its fighting ranks. There are Sunni Muslim Syrians fighting on the front line for Assad even today.”Footnote 55 Kheder Khaddour similarly points out that Sunni merchants who relocated to Tartous and Latakia “have generally been warmly welcomed and are often allowed to shift their business and employees to new sites in the market.”Footnote 56 Since the early days of the uprising, the primary cleavage in the Syrian conflict has been political, not sectarian, as explained by Syrian activist Rifaie Tammas:
[T]he Syrian population became divided mainly along one political line, in terms of their stance towards the Syrian uprising. Syrians were labeled as either pro-regime or pro-revolution, and this label sometimes meant the difference between life and death … Within the new pro-Assad versus pro-revolution division, people greatly differed on the extent of their support of Assad or the opposition, depending on the person’s ideology, ethnicity, and religious background.Footnote 57
7.4.2 Rebel Threat/Desperation
There is little question that the Syrian opposition posed an existential threat to the Assad regime, at least initially. The escalation in indirect, displacement-inducing violence by government forces was undoubtedly a response to the growing strength of the insurgency in 2012. Between January and August of that year, the FSA took at least partial control of seventeen towns and cities in seven provinces, and its ranks swelled to between 100,000 and 150,000 fighters spread across thirty-three brigades.Footnote 58 The rebels also began receiving substantial assistance from foreign governments. Saudi Arabia, Qatar, the United Arab Emirates, Turkey, the US, the UK, and France funneled military equipment and hundreds of millions of dollars in nonlethal aid to the insurgency.Footnote 59 The US and its European allies ratcheted up pressure on Assad to step down and imposed sanctions on Syrian officials. According to Aron Lund, “even on the loyalist side of the war, many seem to have thought in 2012 that Assad would eventually fall from power or lose control over so much of Syria that his regime could no longer function as a national government.”Footnote 60
Yet claims of the regime’s likely demise quickly proved to be exaggerated. Some territorial gains by rebels were the result of the government strategically withdrawing its forces from the countryside and areas along the country’s periphery in order to ensure it preserved a firm grip on what analysts deemed “useful Syria”: The populous and economically vital corridor connecting Damascus to Homs, Aleppo, and the coastal provinces of Latakia and Tartous.Footnote 61 Observers noted that despite the loss of territory and mass defections from the Syrian army, Assad maintained “an unfounded confidence.”Footnote 62 If the regime became desperate, it appeared to be fleeting: It faced little pressure from its core domestic constituents and was able to recover on the battlefield with the support of its “remarkably dependable” Russian and Iranian allies.”Footnote 63 Moreover, the regime retained “a monopoly over public institutions,” which “kept Syrians dependent on its rule,” since even before the conflict, the state had long been the primary provider of essential food, subsidized fuel, education, and health care.Footnote 64
Thus by late 2013, Assad had “checked opposition advances” and “shored up his hold on power.”Footnote 65 Another turning point came that September. After US President Barack Obama declared that the use of chemical weapons in Syria would be a “red line,” he declined to intervene militarily when government forces used sarin gas to kill nearly 1,500 civilians (including 400 children) in rebel-held suburbs of Damascus. Ten months later, ISIS established its so-called caliphate, which shifted the focus of Western involvement in Syria to combating violent extremism. This meant that the US and its allies were “more reluctant than ever to see Damascus fall, fearing that a government collapse would provide the Islamic State with even more ungoverned space in which to operate.”Footnote 66 Intervention by the Russian air force in September 2015, alongside an Iranian-led expeditionary ground force, repelled additional rebel advances and helped the regime reclaim territory. By this time, “it was acknowledged by most that the rebels couldn’t win without U.S. military air intervention.”Footnote 67 The Syrian army dealt the opposition an existential blow in late 2016 when it retook Aleppo City, “the most valuable of the mainstream opposition’s dwindling assets.”Footnote 68 By 2017, the regime was clearly winning the war.
Contrary to what desperation-related explanations would predict, the Assad regime resorted to depopulation well before the Syrian conflict became a war of attrition. More critically, these methods persisted even after the military situation tilted back in Assad’s favor. As Figures 7.4 and Table 7.1 demonstrate, airstrikes and barrel bombs – the regime’s preferred instruments of depopulation – continued to increase in 2014 and again in 2016, 2017, and 2018, after Assad was no longer in grave danger. It is unclear how desperate the Syrian government actually was, even in 2012. But despite receiving widespread international criticism for helping orchestrate mass displacement, even when it had the upper hand, the government continued its depopulation strategy. A fine-grained analysis of conflict trends in Syria by Andrew Halterman further questions whether desperation can account for the regime’s use of violence against civilians. Halterman finds a positive but weak relationship between contemporaneous US intelligence forecasts about the probability of Assad leaving power – which reflects the likelihood of regime survival during each day of the conflict – and the number of civilian fatalities. This suggests that civilian victimization in Syria was not attributable to increased threats to the regime’s survival.Footnote 69
7.4.3 Denial and Sorting
Depriving Syrian rebels of the resources they needed to be politically and militarily successful appears to have been part of the motivation for regime-induced depopulation. José Ciro Martínez and Brent Eng argue that the use of aerial bombardment by Syrian forces has sought to disrupt the governing capabilities of the opposition by crippling key infrastructure, sowing divisions between rebel groups and the local population, and sabotaging efforts to build an effective counter-state.Footnote 70 Likewise, draining contested and opposition-held areas of their populations may have aimed to deny Syrian rebels supplies and recruits furnished by civilian residents.Footnote 71 Victims of these efforts have explicitly described the government’s actions as tansheef al-bahar (draining the sea).
A logic of denial only tells part of the story, however, as it does not account for those who resisted displacement or for what transpired after people fled their communities. Depopulation in Syria sought to divide the sea, not just to drain it. My sorting theory anticipates that, in perpetrating depopulation, Syrian forces will not only create push factors to drive people out; they will also employ pull factors to lure people into government territory. The argument also expects the regime to treat civilians differently based on their movements, not just their identities. I review the evidence for these propositions in what follows.
During the civil war, most IDPs from opposition territories have fled to government zones of Syria.Footnote 72 Early in the conflict, many of the displaced, including Sunnis, were welcomed and even “considered regime supporters” simply because they moved to areas under state control.Footnote 73 According to one regime defector I interviewed, “the government views all IDPs who choose to live in opposition areas as anti-regime, and those who choose to live in its territory as loyal.”Footnote 74 A member of a local opposition council in rebel-held Idlib also told me that “the regime considers [people] to be terrorists because they stayed in opposition areas.”Footnote 75 Assad himself voiced this sentiment in a BBC interview: “In most areas where the rebels took over, the civilians fled and came to our areas, so in most of the areas that we encircle and attack [there] are only militants.”Footnote 76 The view that where people move indicates who they support has not been limited to the government. A European diplomat, after observing that most civilians were fleeing to regime areas, stated that it “probably shows where their political preferences lie.”Footnote 77
As the war wore on, IDPs who spent longer living under the opposition have been viewed with greater suspicion and investigated when they “defect” to regime territory. But, as mentioned at the beginning of this chapter, following the government’s takeover of rebel-held areas in Homs, Rural Damascus, and Aleppo, most residents have been given a choice of destination: rebel-held territory or regime-controlled areas nearby. These evacuations have often been carried out as part of “reconciliation” agreements or truces between the regime, rebel fighters, and community leaders. These agreements have varied by location, however. According to Raymond Hinnebusch and Omar Imady, “regime negotiators offered different kinds of deals in different areas; for example those that demonstrated high resistance in fighting the regime faced total population removal.”Footnote 78
In Daraya, an early bastion of anti-government protests that became “a platform for rebel attacks on regime-held Damascus,” the entire population was forcibly removed and Syrian troops razed most of the town so they could not return.Footnote 79 In contrast, the neighboring town of Moadamiyah, which “had been more defensive in the conflict, was treated more generously” as residents were given a choice of whether and where to move.Footnote 80
This is consistent with the idea that for communities perceived as disloyal, displacement was used for cleansing purposes, while for those whose allegiances were more ambiguous, displacement was used as a sorting mechanism. The evacuations brokered under local reconciliation deals appeared to have two objectives. The first was to expel unrepentant disloyalists, such as rebel fighters, opposition civilian leaders, and activists.Footnote 81 The second was to identify and give neutral or loyal civilians a chance to return to the state. Hence buses carrying evacuees to rebel-held areas were stopped at regime checkpoints and told by soldiers “that if any [passengers] would like to return to the government-controlled areas, [they] would be welcomed and taken care of.”Footnote 82 Or consider the filtering logic behind the “humanitarian corridors” established by the regime and its Russian backers during the siege of rebel-held eastern Aleppo in 2016 – where four passageways out of the city were opened, three for civilians and one for combatants.Footnote 83 Evacuees who elected regime areas were held in government shelters, where they were screened for rebel ties and provided with aid by humanitarian agencies or local authorities.Footnote 84 Some were arrested and incorporated into the regime’s expansive incarceration system but most were not. Between November and December 2016, 70,000 to 100,000 IDPs fled from rebel-held eastern Aleppo to government-run western Aleppo; 2–4 percent of them were reportedly detained.Footnote 85 After the evacuation of Homs, state forces screened hundreds of men, but “many … were eventually released.”Footnote 86
Fleeing to rebel territory, meanwhile, was treated as an act of defiance or – for those wanted by the government – an admission of guilt. This was the implicit message in the bombing of displacement camps near the Turkish border by pro-regime forces; camps that technically fell within rebel territory but were typically miles away from military front lines and towns ruled by the opposition.Footnote 87 Perceived guilt by location also explains why IDPs from Homs City who resettled in the opposition-held neighborhood of al-Waer faced more persecution from government forces than those who moved to regime areas.Footnote 88 In fact, after Homs was recaptured by Syrian forces, the only IDPs reportedly allowed to return were those who fled to regime areas.Footnote 89 Sorting the populace through physical displacement also concentrated disloyalists in a specific area where they could be easily targeted – specifically, Idlib province. According to the International Crisis Group, the regime “deems [those] who chose passage to Idlib … to be ‘irreconcilable.’”Footnote 90 As a regime defector explained to me, “think of a dumpster where you gather garbage to finally burn it.”Footnote 91 Similar language was used by a regime propagandist in describing the government’s military strategy: “You collect trash, separate it, recycle what can be recycled and bury the rest in the ground.”Footnote 92
Thus in the words of one UN report, forced displacement and community evacuations have allowed government forces “to categorize populations on the basis of allegiance.”Footnote 93 Or as one Syrian journalist remarked to me in an interview, “it’s like they [the government] are trying to filter out the people who are against them.”Footnote 94 This indicates that for the Syrian regime and its allies, the strategic benefits of uprooting civilians go beyond draining restive towns of their residents. If the regime’s ultimate objective were punishment by expulsion or demographic engineering, authorities would discourage or prevent IDPs from entering their territories and would likely not allow them to reside there unmolested. But the regime has actually employed several methods to entice people to its territories, where many residents have benefited from its patronage. The SAA inundated civilians in Homs, Aleppo, and other contested areas with text messages announcing relief distributions and leaflets providing detailed instructions and “passes” for entry into regime territory. In 2012, state authorities urged Syrian refugees in neighboring countries to “cast away humility and hunger and return to the homeland” and emphasized their “readiness” to secure aid and housing for the displaced.Footnote 95 The government operated collective relief centers and ensured that a disproportionate amount of international aid goes to areas it controls, since food availability has been a key attraction for IDPs. Beyond relief assistance, the regime “has concentrated on delivering services from within – and mostly to – areas under its control. Besides drawing more Syrians into areas under its control, the Assad regime has managed to monitor them by embedding Syrian state institutions – universities, hospitals, courts and schools – in or near buildings that house its intelligence agencies.”Footnote 96
During the war, the state remained the country’s largest employer,Footnote 97 and civil servants living under the opposition continued to receive government paychecks, which often required traveling to regime areas. This maintained an opening for more people to return to the state. According to one aid worker I interviewed, the government “wants the people back in its territory … except for fighters or activists who are not reconcilable.”Footnote 98
Indeed, in depopulated areas, the procedures for return have provided further opportunities for the state to screen and collect information from civilians. In the six months following the depopulation of eastern Aleppo in late 2016, some 200,000 IDPs were allowed to return provided they registered with local authorities.Footnote 99 The registration process typically requires Syrians to answer detailed questions about their role in anti-government demonstrations and armed rebellion; provide the names, locations, and activities of rebels operating in their areas; and sign loyalty oaths pledging allegiance to the state. Authorities then cross-check people’s responses with other information collected by government agents in order to assess its credibility.Footnote 100 Subjecting IDPs to this process has therefore made them more legible in the eyes of the regime, while supplying authorities with intelligence on rebel identities and capabilities. Given that the war significantly disrupted the state’s extensive intelligence network, the mukhabarat, these background checks have provided critical sources of information.Footnote 101
Aspiring returnees have also been required to show documentation of property ownership to local police or security forces. In 2018, the Syrian government introduced a new law, Act 10, that permitted the state to redevelop properties not claimed by their owners, providing a strong incentive for residents to return to regime territory, particularly since property claims had to be made in person. Yet according to the Syria Institute, authorities reportedly used these claims “as a method of vetting people instead of vetting ownership” and consulted land registries to identify constituencies allied with the rebels.Footnote 102 These vetting procedures apply to Syrian refugees as well as IDPs. While President Assad has publicly appealed for refugees to come back to areas of the country recaptured by the regime,Footnote 103 his government has used civil registration and property claims to screen returnees and filter out the unapologetically disloyal and irreconcilable elements of the Syrian population.Footnote 104 Such elements are therefore discouraged or explicitly prevented from returning, reducing the risk of future revolts. The entire displacement process, then – from the initial departure, to the move to a temporary destination, to return and resettlement – has served as a tool for the regime to sort the population.
7.4.4 The Extractive and Symbolic Benefits of Displacement
For the Syrian government, enticing people back to the state has stemmed from extractive needs, not just punitive ones. Civilians displaced to Tartous and Latakia have offered capital and labor to the local economy, leading regime officials to facilitate the entry of IDPs from Aleppo into the labor market.Footnote 105 Returnees also provide much-needed revenues to the government: Upon arriving in regime-controlled Syria, refugees returning from neighboring countries have been immediately taken to their local municipalities to pay fees for all utilities they would have paid in the years they were gone.Footnote 106
More importantly, the displaced offered a remedy to one of the most pressing problems the regime faced throughout the conflict: a lack of sufficient manpower. While the SAA claimed some 220,000 soldiers in 2011, by April 2012, 50,000 to 60,000 soldiers had deserted or defected to the opposition.Footnote 107 Of the 65,000 to 75,000 troops the regime deemed politically reliable, as many as 7,000 were killed and 30,000 wounded by the end of 2012. The combination of selective deployments, defections, and battlefield casualties greatly hindered the army’s combat power.Footnote 108 In response to increasingly acute manpower needs, the government turned to more desperate recruitment efforts – mobilizing reservists en masse and imposing new service requirements and regulations to prevent desertions.Footnote 109 As a consequence, compulsory conscription for all men of military service age became “rampant” in Syria.Footnote 110 As the New York Times reported: “It is impossible to live in government-controlled Syria without noticing that there are almost no young men on the street. They are in the army, or they are dead. Veterans must carry their military papers with them or risk on-the-spot re-enlistment.”Footnote 111
Investigations by human rights organizations are replete with reports of the regime conscripting IDPs, often forcefully, to serve as fighters or spies.Footnote 112 For instance, over 4,000 men who fled from eastern to western Aleppo in December 2016 were conscripted into the army and sent to the front line with little training.Footnote 113 Upon entering regime territory, eligible displaced men are immediately placed in a register for the army.Footnote 114 The Peace Research Institute in Oslo further describes how the regime has exploited IDPs:
In areas under its control, the regime’s administrative apparatus – while weakened – retains its capacity to register the displaced … Assad’s regime clearly sees those displaced to areas under its control as part of the pool of people from which it can recruit. The displaced … are more dependent on humanitarian aid than anybody, and clearly, receiving aid is followed by an expectation to support the cause.Footnote 115
Finally, the displaced provide the regime with propaganda fodder. As noted by one Western journalist, the government has “used the fact that most people had [come] to live in regime held areas as ‘evidence’ of its popularity, arguing that it reflected the Syrian people’s support to the regime.”Footnote 116 Visits to IDP shelters by state officials, including Assad himself, have been used to showcase the number of people seeking sanctuary from “terrorism” and receiving assistance from the government, in an attempt to legitimize the regime to both domestic and international audiences.Footnote 117 In mid-2017, when hundreds of families decided to move from northern Syria to parts of Homs retaken by the regime, it was broadcast on state media.Footnote 118 In a statement, the local governor declared that the returnees had fled “persecution and inhumane treatment” in rebel-held areas and that their decision to return was evidence that “the state is the only guarantor of the dignity of the Syrian citizen.”Footnote 119 According to the Beirut-based Daily Star, the return
was a propaganda coup for the Syrian president, who is looking to burnish his image as Syria’s legitimate ruler. His readiness to welcome returnees stands in stark contrast to the indifference in many other places toward the plight of displaced Syrians … as the families disembarked from the buses, they waved placards bearing Assad’s image. Some chanted before the cameras that they would sacrifice “blood and soul” for the president.Footnote 120
This evokes other cases where combatants have portrayed population movement as a reflection of their popularity or authority. Since both governments and rebel groups use symbolic processes to bolster their sovereign claims,Footnote 121 they can use the appearance of the population abandoning a rival and seeking shelter in their territories as evidence of their legitimacy. In the Greek Civil War, government supporters deployed rhetoric and imagery meant to depict displacement from rebel areas as foot-voting: “the 680,000 bandit-stricken refugees, ragged and starving, testified, by their flight, that despite the [rebels’] promises of a better future and of democratic freedom, they could not live under a communist system; and this is the strongest and most unassailable argument against the [rebels].”Footnote 122 Refugees were also used to discredit insurgents in Western Sahara. When displaced Sahrawis returned to Moroccan-controlled territory, they were regularly featured on television news broadcasts, “performatively describing the abuses they had suffered in the camps and expressing their gratitude to be in Morocco.”Footnote 123
In Syria, the government’s exploitation of returning citizens has led the UN and other international organizations to caution countries hosting Syrian refugees against pressuring people to go back to Syria even as the conflict stalls. President Assad and his backers clearly see refugee return as a major boost to the regime’s political legitimacy. Moreover, the issue of return gives the regime a bargaining chip in its quest for international recognition and the reconstruction funds that it will need to rebuild the country, which could exceed $250 billion.Footnote 124
7.4.5 Movements and Loyalties: Perception versus Reality
There is compelling evidence that pro-government forces have drawn inferences about the allegiances of the population based on their flight decisions. But have people’s movements actually reflected their true loyalties or affiliations? It is difficult to answer this question definitively. Studies of civilian flight in Syria have found that IDPs tend to go “where there are people with similar political or ethnic characteristics” as those “supporting the regime generally go to [government] areas” while those opposing the regime “tend to move to opposition-held areas.”Footnote 125 Several people I interviewed in a displacement camp in northeast Syria said that when rebels took over a town or village, residents who stayed typically had “a relationship with” the rebels, while those who were “not with them” left.Footnote 126 A survey of Syrian refugees in Turkey, meanwhile, indicated that those who fled the conflict zone withdrew support from all warring parties and refused to pick a side.Footnote 127
Those who remain in Syria, however, have faced strong pressure to pick a side. The evacuation deals described earlier are a case in point. After regime forces captured rebel-controlled parts of Eastern Ghouta in Rural Damascus in 2018, local insurgent groups forged an agreement with the government. As part of the deal, up to 13,000 people opted to leave for rebel-held Idlib province – most of whom were fighters, their families, and local civil society figures.Footnote 128 In interviews, aid workers told me that evacuees who selected Idlib as their preferred destination tended to be combatants, activists, and other people “in opposition to the government.”Footnote 129 This explains why the regime, according to one media report, “has treated [Idlib] province as a dumping ground for those it does not want in its territory.”Footnote 130
For many of the displaced, fleeing to government areas has likely been motivated more by expediency and survival than by a strong political preference. Indeed, IDPs surveyed in Syria tend to cite safety and security as the primary factors influencing their flight.Footnote 131 Yet seeking sanctuary with the state still amounts to a symbolic act of obedience to a regime that built its authority on outward signs of passive compliance from citizens instead of by cultivating true believers. Lisa Wedeen calls this the “politics of ‘as if’” – for decades, the Assad government required not that its followers believe in the power or sanctity of the regime but merely that they talk and act “as if” they do.Footnote 132 In the eyes of the regime, the element of “choice” exercised by residents of opposition territories likely mitigated the potential that its depopulation tactics would backfire. This is evidenced in the terminology used by Syrian state media, which according to one study has eschewed the Arabic word for “refugees” (al-laji’een) in favor of ones that “translate to ‘those who have fled’ or ‘Syrian migrants’ (muhajirah), euphemisms that place the agency for flight solely on the [displaced] themselves.”Footnote 133 Thus civilians who are neither dedicated to the opposition nor guilty of anti-regime activity should, all else being equal, eventually flee rebel areas when incentivized to do so. Moreover, they should move to government territories given the promise of greater security, livelihood opportunities, relief aid, and other incentives created by the state to attract new arrivals. Failing to take these steps is quite costly unless one is affiliated with, or dedicated to, the opposition. From the regime’s standpoint, displacing the population is not driving people into the arms of the rebels; it is simply forcing them to choose between the rebels and the government.
7.5 Conclusion
Strategic population displacement in the Syrian civil war has gone beyond instances of ethno-sectarian cleansing. I have drawn on a broad range of qualitative and quantitative evidence to demonstrate that, in many instances during the conflict, forced displacement has been employed by the government in order to sort and capture the civilian population – not to expel or eliminate it. Denying Syrian rebels resources seemed to play a role in motivating the regime’s depopulation strategies, but this explanation is insufficient by itself. As cleansing operations by pro-government forces gave way to depopulation, a measure initially intended to penalize and deter the regime’s enemies became a method for helping it figure out who its most dedicated enemies were. Decisions over whether and where to go were seen as increasingly political, with the refusal to move to government areas considered a costly act of noncompliance and thus an indicator of disloyalty. State actors treated people differently based on their movements, not just their religious or ethnic identities; profiling that relies on “guilt by location” and perceives those who flee to – or remain in – rebel areas as traitors. Employing weapons that are inherently indiscriminate, the regime used displacement to separate and differentiate the loyal from the disloyal. It used displacement to improve the “legibility” of local communities. And it used displacement to extract much-needed revenues, military recruits, and symbolic benefits from the population. In many instances, population displacement has therefore increased the social heterogeneity of certain provinces, including government strongholds.
The evidence suggests, then, that the regime’s displacement tactics have been motivated much more by political concerns over loyalty than demographic concerns regarding religious sect. My argument also helps account for an important puzzle of the Syrian war: why the regime has given people in recaptured territories a choice of where to move. In offering an answer to this puzzle, I have shown that strategies of depopulation can also exhibit the sorting logic of strategic displacement, similar to strategies of forced relocation.
But why did the Syrian government employ depopulation instead of relocation? Was it a function of capacity constraints or the strength of the rebels? Did the relatively urbanized battlefield play a role? My research design for the Syria case cannot answer these questions. However, in demonstrating that depopulation can serve an assortative function, this chapter offers some insight into how broadly the theory proposed in this book applies – which I return to in Chapter 8.
