Grounds for Continued Skepticism about the ICC’s Deterrence

Julian Ku is the Maurice A. Deane Distinguished Professor of Constitutional Law at Hofstra Law School. Jide Nzelibe is the Harry R. Horrow Professor in International Law at Northwestern Law School. They are the authors of Do International Courts Deter or Exacerbate Humanitarian Atrocities?


We want to thank Professor Stewart for inviting us to participate in this symposium on the new and very important paper by Professors Jo and Simmons.  The reason that the paper is important is that it fills a very large gap in the literature on international criminal tribunals (ICTs).  This gap is the lack of empirical data that would inform debates over the effectiveness of ICTs. In our 2006 survey of the academic literature on ICTs, we found that many proponents of the International Criminal Court claimed that ratification of the ICC would deter future crimes.  Yet there was, until now, almost no serious empirical study of this question in the academic or policy literature.

So Professors Jo and Simmons have done everyone a great service by establishing that there is some evidence for the claim that joining the ICC will deter the commission of war crimes against civilians.  Their claim is both careful and measured. It does not claim that the effect of joining the ICC is dramatic, but (when controlled for other factors), it makes a more modest claim that the effect is nonetheless non-trivial.

While we welcome the arrival of real empirical data on the deterrent effects of the ICC, we remain skeptical of some of the theoretical assumptions underlying their findings.  We also doubt that the small deterrent effect found by Professors Jo and Simmons will continue in future years.

First, we believe that the Jo-Simmons findings fail to establish the correct, or at least the most plausible, baseline for comparing the deterrent effect of joining the ICC.

As we argued in our 2006 paper, no claim of deterrence effects for criminal punishment can ignore the baseline sanctions that the potential perpetrator likely faces without the ICC.  We drew upon the substantial academic literature that has questioned the deterrent effect of capital punishment on potential murderers since an actual execution is rare given the number of appeals permitted by the U.S. legal system.  Given that the alternative to capital punishment is usually a life sentence, the additional (statistically unlikely) chance of being executed adds very little additional deterrent effect to potential killers.

So how does that apply to the Jo-Simmons analysis?  When one develops framework for deterrence, one has to have a plausible empirical baseline against which to make the comparison.  Jo-Simmons concede that prosecution before the ICC is a fairly distant possibility for any individual perpetrator (like the death penalty above), but they maintain it should make a difference because they also assume that the default baseline is impunity.   See Jo and Simmons, at 16 (“The absolute risk of punishment by the ICC remains small, but it is not negligible and is much higher than was the case when impunity was the default.”) & 16  (“ICC investigations, indictments and convictions or those triggered by complementarity are likely to encourage actual or potential perpetrators to reassess the risks of punishment – relative to the status quo, which is often impunity – and to moderate their behavior.”)

But there are good reasons to doubt that complete impunity is often the default.   It simply is not the case that individuals who engage in genocidal violence or ICC level atrocities and then get caught/captured tend to roam freely in the absence of ICC prosecution.   For instance, if Joseph Kony of Uganda’s Lord Resistance’s Army gets captured by the Ugandan army anytime soon, his realistic options will not be prosecution by the ICC or roaming free; in all likelihood, his choices will be to face long imprisonment or a firing squad by the Ugandan army or getting prosecuted by the ICC.  And notice that the ICC indictment does not change the likelihood of his capture because that is pretty much the responsibility of the Ugandan government.

Another more recent illustration is the case of the former President of Ivory Coast, Laurent Gbabgo. He is currently facing charges before the ICC for election-related atrocities; his wife (Simone Gbagbo) is not so fortunate because although she is under indictment by the ICC, she already has been convicted by an Ivorian court for “undermining state security” and is currently serving out her 20-year-sentence in a local prison. If Gbagbo were not before the ICC, he would pretty much being facing the same fate as his wife. As an interesting side note, Amnesty International actually implored the Ivorian government to transfer her case to the ICC rather than prosecute her locally. One can imagine that Amnesty International was more motivated by her welfare than the risk that she will be left to roam free.

One slight nuance that was not captured by much of the Jo-Simmons analysis is that the realistic relevant subsample of individuals who will ever face prosecution by the ICC are individuals who have been forced from power or captured.  If leaders who commit atrocities have not been forced from power, for instance, their chances of getting successfully prosecuted before the ICC are negligible or next to zero.  And when they are forced from powers or captured, the choices they face outside of ICC prosecution are not pretty.

In addition, the reasons why leaders are forced from power or captured has almost nothing to with the ICC.  Gbagbo was not forced from power in Ivory Coast because his political enemies wanted to have him arrested before the ICC; on the contrary, they forced him from power because they thought they were entitled to hold power instead of him.   Paradoxically, the real role of the ICC in such cases might not be to increase the penalty of perpetrators who do get caught, but to decrease and bring it more in line with “civilized” standards.

Thus, we remain doubtful that the Jo-Simmons study establishes the correct baseline from which to measure the ICC’s deterrence effect.

Second, Professors Jo and Simmons do not isolate the deterrent effects of actual ICC prosecutions from the deterrent effects of joining the ICC and adjusting domestic law to conform to the Rome Statute.  Given the comparatively few ICC prosecutions that have actually been brought as well as its very spotty record on obtaining  timely convictions, it is likely that much of the deterrence effect is a result of the changes in domestic law.  As Professors Jo and Simmons point out, the change in domestic law to conform to the Rome Statute, and the principle of complementarity, places the primary burden for prosecuting or investigating Rome Statute violations on state parties to the ICC.   Indeed, as Professors Jo and Simmons suggest in their theoretical framework for analyzing deterrence, it may be that the effect of joining the Rome Statute is to mobilize local groups in favor of criminal punishment for war crimes and that these political forces also operate to deter state actors.

If it is true that conforming domestic law to the Rome Statute is doing much of the deterrence work, it seems possible, or even likely, that the deterrence effect will substantially fade as time passes. 123 states are members of the Rome Statute, but (not surprisingly) the majority of them joined the treaty between 1999 and 2002.    While numerous other states have joined, only two states have joined since 2013, and only 10 since 2011.

Joining the Rome Statute, and adjusting domestic laws, probably provides the most deterrent bang for the buck.  But as time passes, the awareness and significance of the Rome Statute could fade and it is less likely that local groups can mobilize political activity in favor of it.

Finally, we also wonder if the momentum and euphoria generated by joining the ICC can withstand movements to withdraw from it.   In February of 2016, the African Union endorsed a resolution encouraging its members to withdraw from the ICC to protest perceived bias in the ICC.   Indeed, the ICC’s actual record of prosecutions have been both slow and narrow in geographic focus.  Almost every serious investigation has involved Africa.   While not all AU members will necessarily withdraw, the denigration of the ICC’s political reputation domestically will naturally weaken groups that had mobilized around ICC ratification.

In conclusion, while we believe the Jo-Simmons analysis is a much needed and important contribution to the academic literature on ICTs, we remain skeptical as to the significance of its conclusions.  The baseline for many key ICC perpetrators is not impunity, and so the claimed deterrence effect might be too strong.  Morevoer, given the passage of time and the rise of political forces lining up against the ICC in key countries, we think skepticism about the ICC’s ability to maintain whatever small deterrence effect it has remains warranted.