Violence Against Civilians in Civil Wars: Causes and Effects

Introduction

Human history knows few wars that have resulted in no casualties – and most of these were wars in name only. Most conflicts feature casualties that may range from not particularly significant in the grand scheme of things and almost negligible when compared to populations involved to colossal. Yet the magnitude is not the only aspect in which casualties differ: conflicts of different types mainly feature a higher or lower proportion of civilian casualties. While a relatively recent development in historical terms, the distinction between combatants and non-combatants as legitimate and illegitimate targets for armed violence has been a staple of the Modern Age. However, not all wars are interstate conflicts, and not every political actor views itself as subject to the established rules of warfare. In particular, civil wars consistently demonstrate a much higher probability of violence against civilians. Yet even though the violence itself manifests clearly, it is not necessarily evident why the warring parties deliberately target non-combatants. Research suggests that violence against civilians serves as a means of genocide as a war goal, reprisal, or coercion, but in all cases involves certain restraint and risk-benefit evaluation.

Prevalence and Usage

While violence against the civilian population in wars is as old as the concept of civilian population itself, there is no denying that civil wars feature a disproportionally high proportion of civilian casualties. This is especially true for the late 20^th century, when civil wars have become, by far, the most frequent type of conflict worldwide. Empirical evidence suggests that the proportion of civilian losses to the overall casualties in a conflict is the highest in the civil wars that happened during and after the Cold War. According to Tilly, the relative number of civilian casualties had grown from 50 percent in World War II to 90 percent in the civil wars of the end of the century.¹ In other words, civil wars, especially those of later time, tend to be almost twice as hard on the civilian population as the largest, most indiscriminate interstate war of Modern time. One should also remember that killing only constitutes a part of all threats to non-combatants in civil wars. Thus, one can safely assume that violence against civilians is far more common in civil wars rather than in other military conflicts.

Another important point to consider is that violence against civilian populations is not the exclusive prerogative of the rebels. One would be forgiven to think that the rebels, being the illegal non-government actors they are not bound by international law, would be responsible for the majority of violence against non-combatants. However, it is not necessarily the case, as the established political regimes that should supposedly follow the rules of engagement can persecute the civilian population just as vigorously. Based on their study drawing on multiple African examples, Michalopoulos and Papaioannou point out that the government forces are responsible for just as many cases of violence against civilians as the rebels.² Consequently, it would be wrong to assume that the higher proportion of civilian casualties in civil wars is the sole responsibility of irregular anti-government actors not bound by institutional rules. On the contrary, high rates of violence against non-combatants is a characteristic feature of civil wars per se rather than any specific actor type. Therefore, while assessing the reasons why rebels engage in violence against civilians, it is necessary to remember that other actors can do the same.

Case Study: 1972 Genocide and Burundian Civil War

After achieving its independence in 1962, Burundi went through a tumultuous period of political instability characterized by frequent government changes, political assassinations, and widespread civil violence. In 1966, officers of the armed forces deposed the king and declared Burundi a republic, although it was a de facto military dictatorship. The abolition of the country’s democratic institutions and the fact that the majority of the Army officers were ethnic Tutsi provoked resentment and fears among the Hutu population.³ In 1972, Hutu rebels began attacking the government’s administrative buildings and armories, Tutsi civilians, and their own coethnics, whom they saw as supporters of the Tutsi regime.⁴ The government retaliated in force, resulting in mass killings of Hutu, sometimes characterized as genocide. These events secured the Tutsi regime for a while, but after the election of the country’s first Hutu president in 1993, ethnic tensions resurfaced with increased strength. Burundian Civil War lasted from 1993 to 2005 and, just as the events of 1972, featured extensive violence against non-combatants. As such, the civil war in Burundi is a suitable case study to explore the reasons behind committing violence against civilians in civil wars.

One obvious answer to the questions of why political actors – including but not limited to rebels – attack civilians is fairly obvious: it can be among the primary goals of the conflict. As Tilly noted, “violence against civilians, especially whole categories of the population stigmatized for their religious, ethnic, and/or political identities” is a staple of the civil wars of the second half of the 20^th century.⁵ Genocide of politicide can be one of the foremost goals in a civil war when political actors seek to undermine or outright destroy demographics that oppose them. This was precisely the case in the genocide of 1972 when a Tutsi-controlled military regime faced the armed opposition that framed its case in ethnic terms of Hutu supremacy. Admittedly, the majority of civilian casualties in 1972 were Hutu killed by the government forces.⁶ Still, the rebel Hutu supremacists’ attempts at ethnic cleansing were mainly limited by the lack of means rather than humanitarian concerns. While it would be an oversimplification to paint the events in Burundi as a purely ethnic conflict, ethnicity still offered an easy-to-use criterion for genocide and politicide as a war goal.

Another fairly obvious motivation for violence against civilians is retaliation for anti-government actions. While ethnic cleansing or genocide aims to destroy the group and prevent it from being a factor in politics, retaliation attacks serve to scare the population into submission. They seek to achieve that by increasing the costs of supporting anti-government forces and, thus, solidify the power of the existing political regime in another way. This was partially the case with the 1972 genocide – while Hutu bore the brunt of civilian casualties, it is also true that the hostilities began with attacks by Hutu rebels against perceived government supporters.⁷ In this respect, the Hutu attacks that sparked the conflict were partially motivated by notions of reprisal. It is certainly arguable to which degree the killings were a genuine attempt to retaliate and to which degree they were merely a pretext to decimate the population that the rebels perceived as hostile or unreliable. Yet one can reasonably assume that physical destruction of the potentially disloyal groups and retaliation for ethnic defection, whether real or perceived, claiming to represent said ethnicity were both factors motivating violence against civilians.

One should also remember that it would be erroneous to describe the events in Burundi with a purely binary narrative of Tutsi-controlled government forces against Hutu-controlled anti-government rebels. While the conflict between these two political actors was the basis for much of the hostilities, the groups not affiliated directly with either were also responsible for a considerable proportion of the killings. According to Michalopoulos and Papaioannou, both government and anti-government forces in African civil wars are each responsible for approximately 20 percent of violence against civilians.⁸ The ones to commit more than half of all anti-civilian violence are actually militias – armed groups formed in local communities and not affiliated with either government or its opponents. Militias were particularly prevalent in the Burundian Civil War of 1993-2005, and analyzing violence against civilians requires understanding their actions as well. They were still rebels in the sense that they acted independently from and without the sanction of the government, but their aims did not go as far as overthrowing the existing regime.

The best way to conceptualize violence against civilians committed by local militias is Posen’s conflict of the security dilemma. According to Posen, in the absence of a universally acknowledged sovereign authority, any action by a given actor to improve its security may lead to similar responses of other actors.⁹ Consequently, even when neither actor has hostile or aggressive intent, the overall tension will rise, likely prompting someone to attack sooner or later. This was the case with local militias created during the Burundian Civil War, as they were essentially defensive in their purpose. Yet it is hard to ensure security through passive defense in the atmosphere of overall political instability and violence, and the side on the offensive usually has the advantage of the initiative. Posen noted that superiority of offense over defense encourages “pre-emptive war in the event of a political crisis,” and Burundian militias demonstrated this reasoning in the civil war.¹⁰ Attacking the potential enemies immediately and weakening them in advance rather than risking being attacked by them in the future was a preferable strategy and, as such, one more motivation for violence against civilians.

Case Study: The Lord’s Resistance Army

As mentioned above, not every civil war is easily outlined in ethnic or ideological terms, and the Lord’s Resistance Army (LRA) is a clear example of this thesis. A militant group that has been active in Uganda and the Central African Republic for decades is hard to define in specific ideological terms due to its puzzling blend of religious, ethnic, and political motivations. Led by a self-proclaimed prophet Joseph Kony and combining the notions of Acholi nationalism, Christian fundamentalism, and Kony’s cult of personality. Scholarly studies reflect this complexity in the absence of any clear defining vision of LRA as an organization. The only prevailing consensus is that LRA’s activities have a profoundly negative impact on the region’s civilian populations. Even though the group’s inner workings “remain shrouded in mystery and supposition,” and, thus, the scholars still have an incomplete understanding of its tactics, there is a firm association between LRA and abductions.¹¹ This association makes LRA a suitable case study to assess violence against civilians in civil wars in yet another respect and uncover a different set of motivations behind it.

Violence perpetrated by the LRA is often coercion intended to increase the costs of opposing the group or refraining from supporting it. Unlike the main players in the Burundian Civil War, LRA has no clear and easily identified ethnic or ideological allegiance despite its proclaimed adherence to Acholi nationalism. In the words of Blattman and Annan, LRA demonstrates a “patchwork of motives, methods and structure” rather than a structured and well-defined political agenda.¹² As a result of this ideological ambiguity, it is harder for the group to attract and maintain supporters based on an ethnic or ideological allegiance fostered through corresponding appeals. Consequently, coercion becomes all the more important for LRA as the means of ensuring the loyalty of the population and the acquirement of necessary resources. Cramer points out that even in clearly “ethnic” or “ideological” conflicts, such as the wars in Rwanda and Zimbabwe, respectively, ideology and ethnicity were “typically insufficient to make war work,” thus necessitating coercion.¹³ In the case of LRA, with its convoluted political agenda, coercion should play an even greater role and, thus, be one of the main reasons behind violence against civilians.

Another essential rationale for such violence that is particularly prominent in the case of LRA is recruitment. Using adolescent soldiers is not an unheard practice in civil wars, and LRA presents one of the most known and prominent examples of this tendency. It is hard to establish which proportion of youth soldiers is forced into service and which proportion joins voluntarily – and whether one can speak about free choice for adolescents in a war zone at all. Still, there is no denying that abductions are a case of conscription that occurs “by pure force” and, as such, constitutes a subtype of violence against civilians.¹⁴ Empirical research suggests that LRA focuses its forceful recruitment effort “primarily on the abduction of adolescents,” who are already physically fit enough to be soldiers yet whose social scripts are still malleable.¹⁵ In some cases, recruitment may involve tests, such as the forced killing of relatives, but such cases seem to be an exception rather than a rule.¹⁶ Thus, one more reason why political actors may use violence against civilians as a part of their war effort is to bolster their own numbers through forced conscription.

Adverse Effects and Limiting Factors

Any discussion of why participants in civil wars commit violence against civilians would be incomplete without at least a brief reflection on the limiting factors involved. The paragraphs above outlined multiple and varying reasons to extort, coerce, abduct, and outright kill civilians in a war zone. Yet while there are potential advantages to committing such violence, it would be wrong to assume that there are no costs to it either. Research suggests that political actors involved in violence against civilians are aware of these costs and weigh them carefully against the potential gains when making decisions.

One limiting factor in the use of violence against civilians is the adverse effect it can have on popular sentiment and allegiance in the long run. Using violence against civilians as a means of retaliation or coercion is a widely spread approach exemplified by the 1972 killings performed by both sides in the Burundi conflict.¹⁷ It can dissuade the population from supporting certain groups out of fear of imminent reprisals. Yet the very violence used for the short-term increase in security also causes adverse long-term effects for the side that perpetuates violence. Historical evidence suggests that revenge can be a primary motivation for joining a cause in a civil war, which means that targeting civilians may actually bolster the enemy numbers instead of undermining them.¹⁸ The violence perpetrated by the Tutsi-controlled military against the predominantly Hutu civilian population in 1972 resulted in a similar wave of killings committed by Hutu against predominantly Tutsi civilians in 1993, starting the new war. Thus, if a given political actor targets civilians in a civil war, it means that it either presumes swift victory or sees no opportunity to attract the targeted group anyway.

One should also be aware that revenge for the violence experienced may overcome other potential or actual motivations for choosing an allegiance. As mentioned before, political actors engaged in the civil war often seek to mobilize support by outlining their case in ethnic or ideological terms to support their war effort. Burundian rebels in 1972 proclaimed the ideas of Hutu supremacy while the ruling military regime claimed to protect the interests of Tutsi.¹⁹ Even the LRA, for all the ambiguity of its cause, uses the elements of Acholi nationalism and claims of spiritual enlightenment in its political propaganda designed to attract new recruits and keep those already present.²⁰ Yet hostility toward the perpetrators of violence against civilians may overcome ethnic, class, and ideological bonds and cause people to actively support the enemies of their coethnics, coreligionists, and same class members. As Kalyvas points out, localities that have experienced violence by one side in the civil war are much more likely to supply fighters to the opposing side.²¹ Once again, it suggests that actors can use violence against civilians because they have no intention of enlisting the support of the targeted group.

When a political actor relies on the civilian population at least to a certain degree, this factor acts as an important deterrent against violence targeting non-combatants. LRA is an important example: while the organization is not shy to use abductions and executions, research suggests that it still views it as a last resort. When recruiting new members, LRA seeks to use political propaganda, promises of spiritual enlightenment, and disorientation as opposed to direct violence.²² Thus, even the group as brutal and seemingly irrational as the LRA still uses a rational cost-benefit analysis when evaluating violence against civilians. With this in mind, one should remember that when political actors sanction violence against civilians in a civil war, they do so based on a rational assessment of the situation rather than a momentary impulse.

Conclusion

As one can see, groups involved in a civil war can sanction and execute violence against civilians for a variety of reasons. First of all, physical destruction of a certain group may be the foremost goal of the conflict, in which cases it matters little whether the members of said groups bear arms. Even if a given political actor does not pursue complete destruction of a group, violence against non-combatants may be part of a reprisal campaign. Apart from that, targeting civilians may be the result of the security dilemma when multiple actors convinced of the superiority of offense launch preemptive attacks to decimate their potential enemies as thoroughly as possible. Finally, violence against civilians may be a method of coercion and, in particular, forced recruitment to bolster one’s numbers. Still, one should remember that targeting civilians incurs a number of costs that can serve as deterrents and limiting factors even for the more brutal political actors.

References

Blattman, Christopher, and Jeanne Annan. “On the Nature and Causes of LRA Abduction: What the Abductees Say.” The Lord’s Resistance Army: Myth and Reality, edited by Tim Allen and Koen Vlassenroot, Zed Books, 2010, pp. 132-155.

Cramer, C. “Homo Economicus Goes to War: Methodological Individualism, Rational Choice and the Political Economy of War.” World Development, vol. 30, no. 11, 2002, pp. 1845–1864.

Kalyvas, Stathis N. “Ethnic Defection in Civil War.” Comparative Political Studies, vol. 41, no. 8, 2008, pp. 1043-1068.

Koren, Ore, and Anoop K. Sarbahi. “State Capacity, Insurgency, and Civil War: A Disaggregated Analysis.” International Studies Quarterly, vol. 62, no. 2, 2018, pp. 274-288.

Michalopoulos, Stelios, and Elias Papaioannou. “The Long-Run Effects of the Scramble for Africa.” National Bureau of Economic Research, Working Paper 17620, 2011.

Peters, Krijn. Re-examining Voluntarism: Youth Combatants in Sierra Leone. Pretoria Institute for Security Studies, 2004.

Posen, Barry. “The Security Dilemma and Ethnic Conflict.” Survival: Global Politics and Strategy, vol. 35, no. 1, 1993, pp. 27-47.

Tilly, Charles. The Politics of Collective Violence. Cambridge University Press, 2003.

Footnotes

1. Tilly 4.
2. Michalopoulos and Papaioannou 20.
3. Koren and Sarbahi 285.
4. Koren and Sarbahi 285.
5. Tilly 5.
6. Koren and Sarbahi 285.
7. Koren and Sarbahi 285.
8. Michalopoulos and Papaioannou 20.
9. Posen 28.
10. Posen 28.
11. Blattman and Annan 132.
12. Blattman and Annan 132.
13. Cramer 1852.
14. Peters 6.
15. Blattman and Annan 154.
16. Blattman and Annan 154.
17. Koren and Sarbahi 285.
18. Kalyvas 1053.
19. Koren and Sarbahi 285.
20. Blattman and Annan 154.
21. Kalyvas 1060.
22. Blattman and Annan 154.