All posts by Joachim J. Savelsberg

ICC’s Effectiveness and the Explanatory Black Box: Deterrence or Cultural Prevention?

Joachim J. Savelsberg is a Professor of Sociology and Law and Arsham and Charlotte Ohanessian Chair at the University of Minnesota. He has authored several books on the representation and collective memory of atrocity, bridging the gap between criminology and genocide studies. His most recent book is Representing Mass Violence: Conflicting Responses to Human Rights Violations in Darfur (University of California Press).

“In the past rebel leaders would have directed me to nearby villages where they had left piles of corpses behind. They would have shown off their child soldiers. Now they know they may be held accountable. They have become more cautious.” These words, paraphrased from an interview I conducted with an Africa correspondent of a prominent European newspaper in 2011, resonate with empirical patterns revealed in Hyeran Jo and Beth A. Simmons’ paper (“Can the International Criminal Court Deter Atrocity”)? But they constitute anecdotal evidence, while Jo and Simmons fill a void of knowledge through systematic empirical investigation. The need for such data is urgent because the institutions of international criminal justice, specifically the ICC, are so new, and because we strongly desire effective intervention against those crimes over which the ICC has jurisdiction: war crimes, crimes against humanity, and genocide.

Jo and Simmons argue that the ICC and its prosecutions have a deterrent effect on both state actors and rebel leaders; that deterrence works along two lines: “prosecutorial deterrence” and “social deterrence;” and that these effects are contingent. These theses, and the evidence the authors provide, must be taken seriously. They are thorough and well presented. Also, their work is courageous: it steps beyond the traditional division of academic labor between scholars of criminal justice versus foreign policy and international relations. Doing so is an appropriate response to changing institutional realities, as Kathryn Sikkink (2011) made clear when, for good reasons, she subtitled her book on The Justice Cascade with How Human Rights Prosecutions are Changing World Politics. Appropriately, Jo and Simmons are less concerned with law on the books than with law in action, following Roscoe Pound’s early 20th century imperative and a century of law and society scholarship. Approaching the issue from the other side, law and society scholars and sociological criminologists have begun to address international relations (e.g., Hagan 2003; Savelsberg 2015).

Jo and Simmon’s paper yields crucial insights, but limits to their argument demand a reformulation of their central thesis. The authors show that the time following the introduction of the Rome Statute and the ICC and the onset of prosecution witnesses a reduction in killings by state actors, especially those who have supported the ICC and who depend on the world community. Also, rebel leaders kill less, especially those who lead secessionist movements that strive for recognition by the world community. These are important findings.

My critical commentary focuses on one issue: the causal interpretation of the statistical patterns Jo and Simmons identify. Even if the patterns observed can be attributed to ICC intervention, there is little evidence that they can be attributed to a deterrence mechanism. The authors fill the black box between court intervention and trends in violent deaths with the magic formula of “deterrence,” but they do not prove that this is indeed the causal mechanism at work.

The notion of deterrence is, of course, the first modern utilitarian justification of punishment, beyond the Kantian notion of retribution. Jeremy Bentham proposed the idea in his rational actor model, suggesting that criminal offenders weigh crime’s benefits against the costs of punishment (multiplied by the risk of detection). We know today that deterrence does work for certain offenders and under specific circumstances. More specifically, Jo and Simmons correctly cite criminological literature that attributes its effectiveness more to the certainty than to the severity of punishment. And indeed, they provide suggestive evidence that at least part of the effect they measure may result from deterrence experienced by rational actors (e.g., more reduction of killings in aid-dependent countries).

Yet, the logic of the argument is insufficient as the black box Jo and Simmons fill with “deterrence” contains at least one other, possibly more powerful, mechanism: a cultural effect of criminal justice intervention, the radical delegitimization of atrocious violence. Sikkink acknowledges this mechanism in The Justice Cascade (pp. 173f). While she takes exception to a too limited focus on collective memory, one aspect of culture, her finding that human rights records improve especially when trials are combined with truth commissions provides support for the cultural effect. Importantly, cultural transmission is – while potentially effective per se – also a precondition for deterrence to work: rational actors must know about past and current punishment before they can consider them. Prosecutions must have become part of the collective conscience, possibly of the memory, of political and military elites. Then, strategies involving mass atrocities may have disappeared from the decision tree of political and military leaders.

The cultural effectiveness of criminal justice intervention is suggested by a powerful line of neo-Durkheimian thought, and recent empirical work has demonstrated it for the ICC (Savelsberg 2015; Savelsberg and Nyseth Brehm 2015). Analyses of a comprehensive data set on Darfur resulting from content analysis of 3,400 media reports, and of interviews with experts in foreign ministries, human rights and humanitarian NGOs and Africa correspondents from eight Western countries, show that criminal justice intervention – from the International Commission of Inquiry on Darfur via UNSC decisions to ICC prosecution – produce a powerful public representation of those as criminal perpetrators who bear responsibility for the violence. Interventions keep the violence in public view, and they enhance its framing as human rights crimes and genocide. Qualitative analyses illustrate and multi-level, multivariate analyses confirm this effect for each of the eight countries, even if levels of receptivity for the criminal justice frame vary among them. Following ICC intervention (and the cumulative cultural effect of the ad hoc courts of the 1990s), powerful state leaders with responsibility for mass atrocity certainly no longer appear as heroes (as they did for much of human history), and their deeds may be less likely to be denied.

Does it matter if reduction of violence results from deterrence or cultural mechanisms? It certainly does in scholarly terms as we seek to avoid misinterpretations of empirical reality. It also matters for practice as the concern with consequences should inform how international justice institutions communicate their decisions and reasoning to the world public.

In conclusion, Jo and Simmons’ contribution is important and timely. The authors, however, do not prove deterrence to be the central causal mechanism. It would serve future scholarship (and practice) well to take seriously the cultural effect of new interventions in grave human rights crimes by criminal law and justice.


Hagan, John. 2003. Justice in the Balkans. Chicago: University of Chicago Press.

Savelsberg, Joachim J. 2015. Representing Mass Violence: Conflicting Responses to Human Rights Violations in Darfur. Oakland: University of California Press (open access online at: <>.

Savelsberg, Joachim J. and Hollie Nyseth Brehm. “Representing Human Rights Violations in Darfur: Global Responses, National Distinctions.” American Journal of Sociology 121(2):564-603.

Sikkink, Kathryn. 2011. The Justice Cascade: How Human Rights Prosecutions are Changing World Politics. New York: W.W. Norton.