Refugees are fleeing Syria in such astonishing numbers because armed groups continue to target civilians with violence.
That’s what we heard in September when the U.N. Human Rights Council discussed the most recent report of the Commission of Inquiry on Syria. The commission’s chair, Paulo Sérgio Pinheiro, made a plea for international action to end the conflict, pointing to gross violations of the laws of war by all the warring parties: indiscriminate bombing of civilian homes, the deliberate torture and murder of civilians, and widespread rape and sexual violence as acts of war.
On Nov. 19, the U.N. General Assembly’s Third Committee approved a draft resolution based on the commission’s report, which strongly condemned these human rights violations and called for greater accountability.
The commission cites the Syrian Armed Forces for many of these atrocities, as well as the plethora of pro-government militias like the shabiha, the Popular Committees and the many loosely organized militias referred to collectively as the National Defense.
[You might think the Geneva Conventions protect civilians, or that the Red Cross does. Think again.]
Despite human rights treaties that are supposed to protect civilians, governments often abuse civilians during wars. Civilian atrocities are more common, research shows, when a civil war involves pro-government militias. Many scholars argue that governments deliberately outsource brutal violence to militia groups, allowing the state plausible deniability for breaking the laws of war.
If that common wisdom were correct – if governments hand off the most egregious violence to militias – then we would expect to see a pattern of substitution. In other words, governments and militias would not both commit the worst atrocities. And if governments were delegating their atrocities to militias, we’d see government forces committing fewer human rights violations once militias come on the scene.
Is it true?
What’s the evidence on governments, militias and brutal violence in civil wars?
While the delegation logic makes sense on paper, it is not supported by the evidence. In fact, militia violence and government violence are closely correlated, as two recent studies show.
Data on government and militia attacks against civilians in civil wars from 1989 to 2010 show that when governments target civilians — whether through massacres, ethnic cleansing or deliberate bombing and shelling — they generally do so through both their regular military forces and militia forces. And when states decide not to target civilians, militias generally hold back as well.
More broadly, there’s no evidence that militias target civilians even when the government does not. Governments can – and often do – control militia behavior during civil war. Of the 130 militia groups fighting in the civil wars that we studied, only seven reportedly attacked civilians while government forces refrained.
So, no, governments generally do not outsource brutal violence to militias. Rather, they use militias to complement their repression of civilians.
[In two charts, this is what refugees say about why they are leaving Syria now]
Remarkably, these patterns hold even for rape and sexual violence. One might assume that government forces would be particularly interested in evading responsibility for sexual attacks. But our analysis of sexual violence in all armed conflicts involving a militia group in the years 1989-2009 shows that the opposite is true. When militias are first reported to rape and sexually assault civilians, state forces reportedly increase their own sexual violence.
Governments don’t outsource violence to militias. They model it.
In other words, governments don’t outsource violence against civilians; they model it. They may influence militia behavior through training or through more informal diffusion — or both. Studies show that when governments train militias, militias are more likely to target civilians both with sexual violence and other kinds of violence.
That’s consistent with the idea that governments try to complement their regular forces’ violence with violence by militias to escalate and extend counterinsurgent attacks.
Militias that attack different ethnic groups are more likely to target civilians.
Militia violence is also influenced by how — and from which groups — militias recruit their fighters. Militias that recruit from different ethnic or religious groups than the insurgents are three to four times more likely to target civilians.
[Five things you need to know about militias in civil wars]
Because these militias are not as familiar with these other ethnic or religious communities, they often lack the information necessary to distinguish rebel collaborators from ordinary civilians — and so are more willing to attack and punish whole communities for the transgressions of a few.
Further, militia groups that recruit children, which mean they are also likely to recruit by force, tend to commit much more sexual violence against civilians. What can explain this consistency is that groups that recruit this way tend to use sexual violence as an organization-building tool for making fighting comrades out of initially unwilling strangers.
So how can atrocities be prevented?
What does all this mean for policy? How can atrocities be anticipated and even prevented?
First, stop it early. Research shows that it’s much harder to stop atrocities that have become pervasive; early interventions are most effective. That means taking seriously any initial reports of sexual violence or civilian targeting by militias — and, since we know that governments and militias often carry out such attacks at the same time, responding to both immediately.
Policymakers can use this information to anticipate violence. Reports of sexual violence or other forms of civilian targeting by militias should serve as a warning sign to policymakers that the regular government forces are also likely engaged in such atrocities. (See a previous post on what everyone should know about wartime sexual violence here). And of course, once government forces are targeting civilians, policymakers should be on the alert for militias to follow suit.
[What everyone should know about wartime sexual violence]
Second, to prevent atrocities, policy and advocacy should hold governments accountable for atrocities committed by their own forces and for atrocities committed by the militia groups that support them. Although governments may directly order or command militias to brutalize civilians, it is more likely that governments indirectly encourage or enable such militia violence, if only by failing to stop it. Any government denials of responsibility for militia violence should be met with serious skepticism.
As the head of the Commission of Inquiry for Syria noted with regard to the Syrian conflict, international actors must end impunity for targeting civilians. Governments must be urged not only to reign in violence within the regular armed forces, but also to investigate and prosecute militia groups that violate the laws of war.
If governments refuse to do so, international actors should penalize them severely for violating the laws of war, for example, by cutting off access to weapons, supplies, and financing.
Jessica Stanton is an assistant professor of political science at the University of Pennsylvania. Her book “Violence and Restraint in Civil War: Civilian Targeting in the Shadow of International Law” is forthcoming from Cambridge University Press.
Ragnhild Nordås is a senior researcher at the Peace Research Institute Oslo (PRIO), Norway.
Dara Kay Cohen is an assistant professor of public policy at the John F. Kennedy School of Government at Harvard University. Her book “Rape During Civil War” is forthcoming from Cornell University Press in 2016.