Home > News > How is the ICC supposed to work?
296 views 11 min 0 Comment

How is the ICC supposed to work?

- April 2, 2013

I just got back from a fascinating week in Rwanda with a great group of graduate students in Stanford’s International Policy Studies program.  One of the hot conversation topics was Bosco Ntaganda’s surprise appearance at the US Embassy in Kigali on March 18 and his subsequent departure for the International Criminal Court.  In 2006, the ICC indicted  Ntaganda for war crimes he is alleged to have committed in eastern DRC in 2002 and 2003.  The ICC has no police force, however, and Ntaganda remained at large continuing to fight and commit egregious human rights abuses.  So his arrest and appearance at the Hague has been widely commented on as a big win for the ICC.  A “Victory for Justice” as Human Rights Watch put it.

In an abstract sense — or thinking about “justice” independent of the consequences of the ICC as an international institution — I can see the argument that, yes, this is a victory for justice.  Typical one-line summary statements of the rationale for the ICC is that its purpose is to “end impunity” for big-time human rights abusers, including government leaders and officials.  The ICC is said to be about creating “accountability” for a class of criminals who have not faced prosecution and punishment in the past.

One could argue for such an institution independent of whether it has any good consequences beyond satisfaction that victims and others take at seeing people like Ntaganda in the dock.  But in the next breath supporters of the ICC usually make the argument that “ending impunity” and creating accountability will have the good effect of reducing human rights abuses by increasing deterrence.   The model that advocates have in mind is exactly that of policing and domestic crime:  If you increase the probability of prosecution and punishment for committing certain crimes, the total amount of these crimes should go down.

But the domestic analogy doesn’t work here because as noted the ICC has no police force.  Instead it relies on states that are party to the ICC to hand over any indictees they happen to come across, or, as in the Ntaganda case, on voluntary surrenders.  The implications of this difference aren’t immediately obvious.  But there are good reasons to think that they include the possibility that the ICC will have no effect on reducing abuses or might even increase them.  There are two main ways this can happen.

First, the ICC option can actually limit the liability of human right abusers, to the extent that they can choose whether to turn themselves in.  Ntaganda’s case is a great example of this.  He almost surely turned himself in because his option in the DRC had recently become worse.  He lost a fight with another faction of his M23 group and would very likely have been killed had he stayed in DRC.  (It’s possible that this decision was made for him by some higher powers — that is, “go to the ICC and keep your mouth shut or we will kill you here” — but the implication is basically the same for limited liability.)

So is that a great victory for the ICC?  His surrender ends the embarrassment of having Ntaganda on the loose committing more crimes, but if the general principle is, The ICC is a way to save yourself from death if things get too hot, then the institution is potentially encouraging, not deterring, human rights abuses.  

The second way that the ICC may increase human rights abuses is through what you might call a conflict prolongation effect:  If ICC prosecution is seen as a bad thing by some indicted leaders in the midst of some civil conflict, then they may be more inclined to continue fighting and abusing than to end the conflict by cutting a deal or going into exile.  In this story, the ICC is basically a way for states to tie their hands against offering amnesties and exiles to criminals like Omar al-Bashir in Sudan in order to end a conflict or facilitate a political transition.  If this is right (or in the class of cases where it applies), then the ICC would have an ex ante deterrence effect — rulers would be less likely to start committing abuses since the exit options later are now worse.  But it would also prolong conflict and human rights abuses in cases where they decide to start down the road anyway.  And there is no good reason to suppose that the ex ante deterrence effect on total abuses would be greater than the ex post conflict prolongation effect.  In the aggregate it could go either way.

This ICG report on Sudan suggests this effect is at work and is a problem in the case of Bashir:

Such cooperation [with Bashir and his party, the NCP, including UN Security Council deferrals of ICC prosecution] might be unpalatable to many who hold Bashir responsible for atrocity crimes, but it would be necessary to prevent further conflict and continued humanitarian crises in Sudan as well as South Sudan. (page i)

President Bashir is a key impediment to resolving the crisis, though certainly not the only one. He cannot step down or cede too much power, in part for fear he will end up in The Hague, where he has been indicted by the ICC. (page 25)

Back in DRC, it may have been Western pressure to arrest Ntaganda early in 2012 that led Kabila to go after him, breaking a prior amnesty agreement and stimulating the formation of the rebel group M23 in May 2012, which proceeded to fight, rape, pillage, and threaten to take Goma last November.  In Kenya (a disaster for the ICC that deserves a separate post), the ICC indictments of Kenyatta and Ruto for allegedly fomenting violence against each others’ supporters after the 2007 election greatly increased the stakes in the election this last month, driving these two former adversaries into coalition together and probably helping to get them elected.  (Although perhaps it is an ironical, unintended ICC success that it has effected an elite-level reconciliation between ethnic leaders of the two main fighting sides from 2007.)  Finally, Uganda is another case where some have argued that ICC indictments of Joseph Kony and his lieutenants prolonged or increased a conflict that otherwise might have been settled sooner.

Navi Pillay, the UN High Commissioner for Human Rights, and several others  have been calling for the Security Council to refer the horrible conflict in Syria to the ICC, which would open up Bashar al-Assad and other officials (and rebel leaders) to potential ICC prosecution.  (Syria is not a state party to the ICC so this is the only way the ICC could get jurisdiction.)  This would certainly feel good, but how exactly would it help matters?  Either the limited liability or the conflict prolongation effects might come into play in the case of Assad and his regime, depending on how the conflict evolves.

The more general question is, How is the ICC supposed to work to reduce the worst sorts of atrocities and human rights abuses?  The analogy to the domestic crime and punishment model isn’t right.  “Ending impunity” for commission of atrocities is a great objective, but so is reducing the incidence of atrocities and it’s not clear how the ICC can do both given the way it is set up and the realities it has to work with.  To me it looks like a well-intentioned but not fully thought out institutional experiment that will tend to be used primarily as a way to make rich countries feel better about cases whether they aren’t willing to intervene, while the institution itself sometimes has consequences that contradict its avowed purpose.  The ICC is still a young institution and maybe it can evolve to work out some of these problems, but it’s not there yet.

(There are some excellent papers in the political science literature proposing, in effect, rationales for the ICC, in particular Beth Simmons and Allison Danner’s 2010 article in International Organization and Michael Gilligan’s 2006 piece in the same journal.   And there are other ways you can tell a story about how the ICC might be creating deterrence without having conflict prolongation effects.  This already being a long post, I’ll just say that I find the considerations given above more compelling at this point.  A bit more, and more references in the slides for a talk on which this post is based, here.)