Terrorism, Civil War Outcomes, and Post-war Stability: Hypotheses and (very) Preliminary Findings

How does the use of terror tactics by one or both sides affect the likelihood of negotiations or the chances for their success? Are terrorist wars more likely to end with the defeat of one side rather than a compromise outcome? Are conflicts that involve terrorism any more or less likely to resume than others? The use of terror seem to make peace close to impossible in some cases (the Palestinian Intifada), but in other conflicts peace is achieved despite the use of terrorist tactics (as by the ANC in South Africa). This paper takes a first cut at answering these questions. Conflicts in which terrorist tactics are used by rebels are compared to those in which rebels chose other strategies. This comparison allows for a systematic study of the effects of terrorism on important questions of war termination and post-war reconstruction. The paper defines and discusses measurement of terrorist rebel groups, enumerates several hypotheses about the effects of terrorism on the outcome of civil wars and the stability of peace in their wake, and presents preliminary findings from statistical analyses. The paper (tentatively) concludes that terrorism does make settlement harder to achieve, but that it has no large effect on the stability of peace (due to selection effects). It debunks the hypothesis that terrorism is effective for rebels – rebels who choose terrorist tactics fare no better in terms of the outcomes achieved on the battlefield or at the negotiating table than those who eschew such tactics. Acknowledgments Earlier versions of this paper were presented at a Festschrift panel in honor of Martha Crenshaw at APSA 2008, and at the Security Studies Program at MIT. Thanks to participants at these presentations for very useful comments. Special thanks to Jessica Stanton for sharing her data on rebel treatment of civilians, to Jake Shapiro for comments on available data sets on terrorism, to David Laitin for comments on the paper, and to Jim Fearon for sharing new data on war outcomes. Last but not least, thanks to Martha Crenshaw for comments on early thoughts for this paper, and especially for her continuing mentoring and friendship. 1 April 4, 2002, the Rose Garden. http://www.whitehouse.gov/news/releases/2002/04/ 200204041.html (accessed July 25, 2008). 2 January 19, 2006, at the White House. http://www.whitehouse.gov/news/releases/2006/01/ 20060119-6.html (accessed July 25, 2008). 3 Hurricane Harry, HH Blowhard, blog “How to Negotiate with Terrorists” August 12, 2006. http://hurricaneharry.blogspot.com/2006/08/how-to-negotiate-with-terrorists.html (accessed July 25, 2008). 2 No nation can negotiate with terrorists. For there is no way to make peace with those whose only goal is death. President George W. Bush We do not negotiate with terrorists. We put them out of business. US Press Secretary, Scott McClellan A peacemaker walks up to the left side of a line. A terrorist walks up to the right side of the line. The peacemaker introduces himself. The terrorist kills him. A peacemaker walks up to the left side of the line. A terrorist walks up to the right side of the line. The peacemaker asks, "why did you kill my friend?" The terrorist kills him and rapes his wife. ... A peacemaker walks up to the left side of the line. A terrorist walks up to the right side of the line. The peacemaker says, "I'll pay you $1000 if you stop attacking us." The terrorist agrees to the deal, takes the $1000, and kills him. ... [and so on]. Many governments have a stated policy never to negotiate with terrorists. The rationale for this position is obvious – terrorism must never be rewarded with concessions, negotiations would set a dangerous precedent, and even sitting down to talk with terrorists can grant them political legitimacy. But, of course, what governments say in this regard and what they actually do are very different. Examples of negotiations between governments and terrorists are plentiful: the South African government negotiated a deal with the African National Congress (ANC), the British government negotiated with the Irish Republican Army (IRA), Spain has 4 Neumann 2007; Moran 2006. Much of the literature on bargaining with terrorists examines negotiations over hostages. See for example, Atkinson, Sandler, and Tschirhart 1987; Bapat 2006; Lapan and Sandler 1988. 5 One exception is Sambanis 2008. Sambanis notes that “most terrorist events take place in countries affected by civil war” (p.2), though he goes on to draw distinctions between terrorism and civil wars. 6 The paper was first presented at a festschrift panel in honor of Martha Crenshaw. As a scholar of civil wars, but a former student of Martha’s, the panel gave me an opportunity to address the impact of terrorism in civil wars. Note that I am guilty of Audrey Cronin’s charge of conducting research on terrorism as a newcomer to the field with only a beginner’s knowledge of the existing scholarship. Cronin 2006, p.7. I would thus be particularly grateful for suggestions from those of you who have worked on terrorism longer than I as to read and where to look as I move forward with this project. 3 negotiated with Euzkadi Ta Askatasuna (ETA), Israel has negotiated with the Palestinian Liberation Organization (PLO), Sri Lanka with the Liberation Tigers of Tamil Eelam (LTTE), and so on. As these examples suggest, much of the use of terrorism as a tactic takes place in the context of civil wars, yet the study of terrorism and the study of civil wars have generally proceeded in isolation from one another. This paper compares civil wars that involved terrorist rebel groups with those that did not as a way of trying to merge insights from the two literatures. It is motivated by a series of questions about terrorism and the potential for the stable termination of civil wars: how does the use of terrorism affect the likelihood of negotiations or the chances for their success? Are terrorist wars more likely to end with the defeat of one side rather than a compromise outcome? Are they more likely to yield favorable outcomes for the rebel side as the “terrorism works” literature would suggest? Are conflicts that involve terrorism any more or less likely to resume than others? The use of terror seems to make peace close to impossible in some cases (the Palestinian Intifada and the Sri Lankan civil war), but in other