Portal de Periódicos da CAPES

Targeting Civilians in Ethno-Territorial Wars: Power- and Preference-Based Sources of Ethnic Cleansing and Mass Killing Strategies

2013; Taylor & Francis; Volume: 36; Issue: 5 Linguagem: Inglês

10.1080/1057610x.2013.775455

1521-0731

Shale Horowitz, Min Ye,

Political Conflict and Governance

Abstract In internal ethno-territorial conflicts, what explains why state or rebel group leaderships use civilian-targeting strategies—expulsion or mass killing strategies designed to punish enemy civilians or to decimate the enemy civilian presence on contested territory? One argument is that those living under the worst initial conditions—defined in terms of collective goods such as weak collective autonomy, policy outcomes, and material conditions—are most likely to target enemy group civilians. Another approach focuses on relative power—arguing that the enemy civilian population is targeted either because of weaker or stronger relative power. A third approach argues that differences in leadership preferences—in particular, more ideologically extreme or power-seeking preferences—are likely to drive direct assaults on enemy civilians. We examine these proposed mechanisms in terms of expected effects on benefits and costs in a simple ethno-territorial bargaining framework. We argue that relative power advantages and more extreme nationalist preferences seem most likely to predict decisions to target enemy civilian populations. We expect strongly power-seeking preferences to lead to civilian targeting more conditionally—where there is a greater internal political threat along with either greater relative power or a more moderate enemy. Last, we do not expect that variation in initial conditions will have a significant direct effect. We apply the framework to explain patterns of civilian targeting following the collapse of Yugoslavia in 1991. [Supplementary material is available for this article. Go to the publisher's online edition of Studies in Conflict & Terrorism for the following free supplemental resource: online appendix.] Notes 1. There are also cases where more than one dominant ethnic group is associated with the state, or where the group fighting the state for greater territorial self-determination is the largest ethnic group. 2. Following the Genocide Convention, “genocide” is usually defined as involving intent to destroy all or part of a group per se. In an ethno-territorial conflict, this may or may not be a goal of intentional mass killing. 3. The term “designed” refers to intent rather than outcome. Thus, measuring such a threshold depends on observing what was done on territory under sufficient control to implement the desired strategy. Choosing a particular threshold is difficult in the way that choosing a threshold for “genocide” has been difficult. It is certainly possible that mass killings and expulsions of less than 10 percent of the enemy population may still reflect a more gradual or limited strategy of removal rather than one of subjugation alone. We tentatively choose 10 percent because it seems a reasonable rough dividing line between large-scale removal and pure subjugation strategies. However, better understanding and data-gathering on removal strategies may make it reasonable to alter the threshold in future research. In discussing scale differences, it is also possible to distinguish more than two levels of mass killings and expulsions. We distinguish only two levels for the sake of theoretical clarity and simplicity as well as ease of measurement. As discussed below, the explanations we offer have a continuous logic, so that the analysis of scale differences generalizes to more than two levels. 4. MacArtan Humphreys and Jeremy Weinstein, “Handling and Manhandling Civilians in Civil War,” American Political Science Review 100 (2006), pp. 429–447. 5. Laia Balcells, “Rivalry and Revenge: Violence against Civilians in Conventional Civil Wars,” International Studies Quarterly 54 (2010), pp. 291–313. 6. For example, Max Abrahms, “Why Terrorism Does Not Work,” International Security 31 (2006), pp. 42–78; Martha Crenshaw, “The Causes of Terrorism,” Comparative Politics 13 (1981), pp. 379–399; Robert Pape, “The Strategic Logic of Suicide Terrorism,” American Political Science Review 97 (2003), pp. 343–361; and Benjamin Valentino, Paul Huth, and Dylan Balch-Lindsay, “‘Draining the Sea': Mass Killing and Guerrilla Warfare,” International Organization 58 (2004), pp. 375–407. 7. Studies that have sought to explain mass killings or expulsions of civilians in general or across different conflict-types include Andrew Bell-Fialkoff, “A Brief History of Ethnic Cleansing,” Foreign Affairs 72 (1993), pp. 110–121; Barbara Harff, “No Lessons Learned from the Holocaust? Assessing Risks of Genocide and Political Mass Murder since 1955,” American Political Science Review 97 (2003), pp. 57–73; Matthew Krain, “State-Sponsored Mass Murder: The Onset and Severity of Genocides and Politicides,” Journal of Conflict Resolution 41 (1997), pp. 331–60; Benjamin Valentino, “Final Solutions: The Causes of Mass Killings and Genocides,” Security Studies 9 (2000), pp. 1–59; and Reed Wood, “Rebel Capability and Strategic Violence against Civilians,” Journal of Peace Research 47 (2010), pp. 601–614. 8. We discuss a one-period model. The same logic applies in two- or infinite-period models. Proofs are available from the authors upon request. The basic model was developed by Ariel Rubinstein, “Perfect Equilibrium in a Bargaining Model,” Econometrica 50 (1982), pp. 97–109. Well-known applications include James D. Fearon, “Rationalist Explanations for War,” International Organization 49 (1995), pp. 379–414; and Robert Powell, In the Shadow of Power: States and Strategies in International Politics (Princeton, NJ: Princeton University Press, 1999). 9. The all-or-nothing outcomes are a simplifying assumption. The assessed probability of victory can also be interpreted as the mean of a range of possible outcomes, including intermediate, “stalemate” outcomes. 10. A sufficiently large adverse change in the status quo may also lead the party made worse off to initiate crisis and war. But such a significant change in the status quo is typically an effect of exogenous shocks to relative power or preferences, and thus would be considered an effect rather than a cause of crisis. 11. See the discussion in Barry Posen, “The Security Dilemma and Ethnic Conflict,” Survival 35 (1993), pp. 27–47; Chaim Kaufmann, “Possible and Impossible Solutions to Ethnic Civil Wars,” International Security 20 (1996), pp. 136–175, views the situation from an ex post angle. 12. See Valentino et al., “Draining the Sea.” 13. For example, Slavoljub Djukić, Milošević and Marković: A Lust for Power (Montreal: McGill-Queen's University Press, 2001), emphasizes power-seeking motives. John B. Dunlop, Russia Confronts Chechnya: Roots of a Separatist Conflict (Cambridge: Cambridge University Press, 1998), and Jagath Senaratne, Political Violence in Sri Lanka, 1977–1990: Riots, Insurrections, Counterinsurgencies, Foreign Intervention (Amsterdam: VU University Press, 1997), point to nationalist ideological motives. 14. Pape, “Strategic Logic of Suicide Terrorism,” emphasizes the bargaining leverage gained from imposing such costs. 15. Hatred or dehumanization of the enemy, along with leadership preferences more generally, are emphasized in much of the genocide literature (e.g., Harff, “No Lessons Learned from the Holocaust?” and Straus, “Second-Generation Comparative Research on Genocide,” World Politics 59 (2007): 476–501. 16. For example, Abrahms, “Why Terrorism Does Not Work.” 17. Michael Walzer, Just and Unjust Wars: A Moral Argument with Historical Illustrations (New York: Basic Books, 1992). 18. For example, Ada Huibregtse, “External Intervention in Ethnic Conflict,” International Interactions 36 (2010), pp. 265–293. 19. Targeting civilians also tarnishes the international reputation of the nationalist causes. Such reputational costs, in turn, may change third-party behaviors in ways that strengthen the enemy. 20. The change in payoffs to the “ordinary nationalist” player of targeting civilians can be written as , where the subscripts stand for the type of benefit b or cost c. As discussed, all of these benefit and cost types are expected to be increasing or unchanged in relative power r. (Note that the payoffs derived from polarization benefits b3 and normative costs c1 are not expected to be affected by relative power.) Such linear preferences are standard in the general conflict bargaining literature. Fully specified bargaining models are available from the authors. 21. The state will not always have greater power. For example, state capacity may be extremely low; there might be state failure; powerful external actors may support internal secessionist groups; and so on. 22. An extreme nationalist's change in payoffs from targeting civilians can be written as . E*>1 indicates the greater relative weight, as compared to an ordinary nationalist, placed on upside outcome benefits and polarization benefits, and E indicates the likely decrease in bargaining benefits (due to lower credibility of any commitment to stop imposing costs in exchange for concessions) and in normative costs. As a leader becomes more extreme, the E* and E terms have larger effects. As before, r indicates the effect of relative power on benefits and costs. 23. Symmetrically to an extreme nationalist, a moderate nationalist's change in payoffs from targeting civilians can be written as . M* 1 indicates the greater relative weight, as compared to an ordinary nationalist, placed on polarization benefits and enemy response costs, and P indicates the likely decrease in bargaining benefits (due to lower credibility of any commitment to stop imposing costs in exchange for concessions) and in normative costs. The d term indicates that polarization benefits are increasing in the extent of the internal political threat. As a leader becomes more strongly power-seeking, the P*and P terms have larger effects. As before, r indicates the effect of relative power on benefits and costs. If the level of internal threat is very low, then the stronger power-seeker likely has a weaker incentive to target civilians than an ordinary nationalist. As the level of internal political threat grows, the polarization benefits create a stronger incentive to target civilians. But whether this is sufficient to trigger the decision depends most on whether polarization benefits are strongly offset by enemy response costs, which in turn depend on relative power and on enemy preferences. 26. If we think of qualitative escalation of civilian targeting as a more continuous choice, strongly power-seeking leaders seem most likely to capture what they consider to be the largest net benefits with very limited harassment or suppression---“token” killings that are enough to capture the internal polarization benefits. We thank an anonymous reviewer for this point. 27. Another similarity is that both are expected to discount war costs. 28. As compared to more extreme nationalists, there is at least one notable difference in how stronger power-seekers are expected to pursue polarization benefits. Strong power-seekers are expected to pursue them more strongly where there is a greater internal opposition threat, so that the polarization strategies are expected to be more episodic—to rise and fall with the level of internal opposition threat. In contrast, the more extreme nationalist is more likely to seek polarization benefits throughout the conflict period—to prevent own-group moderates from compromising away any of the extremists' ideal goals. 29. See the Appendix for details. 30. For Bosnia and Herzegovina, we use the abbreviated “Bosnia.” 31. Burg and Shoup, War in Bosnia-Herzegovina, pp. 68, 131–185; Goldstein, Croatia, pp. 228–233, 242–243, 246–247, 253–254; Silber and Little, Death of Yugoslavia, pp. 170–174, 180, 244–251, 358–360, 368; Shrader, Muslim-Croat War, pp. 70–162; Tanner, Croatia, pp. 253–255, 266–267, 289–290. 32. Steven L. Burg and Paul S. Shoup, The War in Bosnia-Herzegovina: Ethnic Conflict and International Intervention (Armonk, NY: M.E. Sharpe, 1999), pp. 84–96; Laura Silber and Allan Little, The Death of Yugoslavia (London: Penguin, 1996), pp. 198–201, 352–374; Marcus Tanner, Croatia: A Nation Forged in War (New Haven, CT: Yale University Press, 2001), pp. 253–254, 273–274; Robert Thomas, Serbia under Milošević: Politics in the 1990s (London: Hurst, 1999), pp. 238–239, 251. 33. Silber and Little, Death of Yugoslavia, pp. 109, 170; Tanner, Croatia, pp. 249–251. 34. Burg and Shoup, War in Bosnia-Herzegovina, p. 91; Silber and Little, Death of Yugoslavia, pp. 188, 353–360; Tanner, Croatia, pp. 279–280, 284, 296–298. 35. Djukić, Milošević and Marković; Thomas, Serbia under Milošević. 36. Burg and Shoup, War in Bosnia-Herzegovina, pp. 132, 175; Silber and Little, Death of Yugoslavia, pp. 258–259; Tanner, Croatia, p. 285. 37. Burg and Shoup, War in Bosnia-Herzegovina, pp. 67–73, 108–117, 126–127, 194–197. 38. Ibid., pp. 68, 137–139, 178–180. 39. Ibid., p. 362; Silber and Little, Death of Yugoslavia, p. 369. 40. Burg and Shoup, War in Bosnia-Herzegovina, pp. 331–337. 41. Ibid., pp. 134–136, 292–294; Charles R. Shrader, The Muslim-Croat War in Central Bosnia: A History, 1992–1994 (College Station: Texas A&M University Press, 2003). 42. Ivo Goldstein, Croatia: A History (London: Hurst, 1999), pp. 239–248; Tanner, Croatia, pp. 286–292. 43. Burg and Shoup, War in Bosnia-Herzegovina, pp. 215–298; Shrader, Muslim-Croat War, pp. 71–72, 159–162. 44. Tudjman also encouraged Ibrahim Rugova, leader of the Kovoso Albanians, to open a third front against the Serbs. But Rugova, particularly in view of the Bosnian Muslims' fate, refused. See Tim Judah, Kosovo: What Everyone Needs to Know (Oxford: Oxford University Press, 2008), pp. 69–74. 45. Burg and Shoup, War in Bosnia-Herzegovina, pp. 66, 292–294; Goldstein, Croatia, pp. 243–248; Tanner, Croatia, pp. 286–293. 46. If “terrorism” is defined as intentional killing of civilians to achieve political objectives, this contradicts the well-known argument that terrorism is a “weapon of the weak.” 47. The same thing occurred again in the Kosovo war of the late 1990s, but that is beyond the scope of this article.