Barriers to Reintegration after Ethnic Civil Wars: Lessons from Minority Returns and Restitution in the Balkans ERIN K. JENNE This article evaluates the record of minority return in Croatia, Bosnia and Kosovo to assess the viability of ethnic reintegration in the wake of protracted sectarian violence. Comparative analysis reveals that the logic of post-war ethnic spoils has greatly limited the success of such programmes. What success has been achieved is largely due to third party efforts to disrupt patronage networks and challenge post-war authorities. I conclude that these factors are more significant barriers to reintegration than inexorable ethnic hatreds and fears derived from memories of war. Because such barriers are more readily overcome than entrenched grassroots hostilities, there may be more hope for reintegration than previously thought. However, the systematic failure of the international community to protect and assist prospective minority returnees suggests that continued scepticism of post-war reintegration is in order. Since the end of the Cold War, there has been a growing interest in ethnic reintegration as a means of post-war reconstruction, the first step of which is returning displaced minorities to their local communities.1 Accordingly, large-scale returns have been attempted in the aftermath of nearly every recent civil conflict.2 An estimated 604,000 refugees were voluntarily repatriated in 2008 alone; the main countries of return were Afghanistan, Burundi, Sudan, Democratic Republic of Congo, Iraq and Angola - all sites of recent or ongoing ethnic warfare. In the past decade alone, over 11 million refugees - many victims of wartime ethnic cleansing -have returned to their home countries if not to their homes.3 The increased use of returns and restitution is motivated partly out of a commitment to assist people displaced by civil war and partly out of pressures from western governments to staunch the tide of refugees arriving on their shores. Academics, meanwhile, have begun to explore the peace-building potential of such programmes. The logic follows that societies riven by war can be rebuilt by returning combatants and ethnic minorities to their pre-war communities and by promoting ethnic cooperation in local institutions. Over time, daily contact between the former combatants is believed to rebuild cross-ethnic bonds and gradually consolidate ethnic peace. Although plausible, a full test of this proposition has yet to be made as post-war reintegration has rarely gone beyond the returns and restitution phase. Civil Wars, Vol.12, No.4 (December 2010), pp.370-394 ISSN 1369-8249 print/ISSN 1743-968X online DOI: 10.1080/13698249.2010.534622 © 2010 Taylor & Francis BARRIERS TO REINTEGRATION AFTER ETHNIC CIVIL WARS 371 This poor record of success begs the question as to what stands in the way of minority returns. I argue that the principal barrier is the logic of ethnic spoils, which holds that the wartime ethnic entrepreneurs have incentives to maintain ethnically homogenous enclaves in the wake of sectarian conflict, and that co-ethnics in their patronage networks have incentives to assist them. This paper lays out the ethnic spoils argument, which is illustrated using the cases of post-war Croatia, Bosnia and Kosovo.4 I have chosen to compare these cases for three principal reasons. First, there was massive internecine conflict and ethnic cleansing in all three cases. The challenges of reintegration, therefore, resemble those of other post-war settings with which peacebuilders are concerned. Second, the Balkan programmes of returns and restitution have been the most comprehensive to date, providing good tests of the limits of post-war reintegration. Finally, despite their overall poor record, these programs have met with variable success over time and across cases. The similarity of background conditions permits a controlled comparison for the purposes of identifying the determinant(s) of successful reintegration. The paper proceeds in four parts. First, I outline the strategy of post-war reintegration and hypothesize preconditions of success. The next section lays out the case against reintegration and in favor of ethnic partition. The third examines programs of return and restitution in Croatia, Bosnia, and Kosovo to identify the primary impediments to success in this first and most critical stage of reintegration. Finally, the Bosnian cities of Brcko and Prijedor are examined as 'least likely' cases of success, as sites of extreme sectarian violence and thus considerable ethnic hostilities. The moderate success of minority returns in Bosnia in particular suggests that post-war reintegration has potential so long as ethnic spoil systems are effectively disrupted and minority returnees receive prompt resettlement assistance. I conclude with summary observations concerning the feasibility of reintegration in recent war-torn states such as Iraq. POST-WAR REINTEGRATION AND ETHNIC SPOILS Post-war reintegration5 aims to reconstruct multi-ethnic communities divided by ethnic war, beginning with the return and resettlement of displaced minorities who may have fled the country.6 The peace-building effects of reintegration are premised on the notion that former combatants tend to reach practical accommodations when they live and work in close proximity. Diverse electoral districts are also thought to promote ethnic tolerance in the political sphere because politicians have incentives to appeal to voters across ethnic lines.7 When this occurs, politicians move from ethnic exclusivism to the moderate center where inter-ethnic alliances can be formed. In these and other ways, returnees 'represent an important opportunity to strengthen the fragile peace.'8 Over the long run, daily contact across ethnic boundaries promotes ethnic peace by muting mutual recriminations and undermining group stereotypes.9 A growing body of research demonstrates that inter-group contact can help build ethnic tolerance in divided societies. Surveys of students in integrated schools in northern Ireland show, for example, that friendships 372 CIVIL WARS across confessional lines increased from 41 per cent prior to attending an integrated school to 67 per cent afterward; moreover, more than half of those who had attended integrated schools had a partner from a different background as their own, against eight per cent in northern Ireland as a whole.10 Other survey research in northern Ireland showed that inter-group contact is a strong predictor of positive out-group attitudes and the propensity to trust members of the other group.11 A study of mixed Jewish-Arab settlements in Central Galilee revealed that Jews in these settlements had more moderate attitudes toward Jewish-Arab relations than either Jews or Arabs outside these settlements.12 Similar survey work in South African universities revealed a strong positive relationship between inter-racial contact and tolerance toward the other group.13 Does reintegration mean that once-hostile groups are 'reconciled' in the sense that they have forgiven one another's past transgressions?14 Not necessarily - any number of factors can impede the progress of reconciliation once the displaced have returned to their pre-war communities. These include significant group inequalities, external intervention, and other environmental stressors that intensify ethnic cleavages. Nevertheless, reintegrated communities are less likely than divided communities to foment violence in the absence of external provocation.15 Thus, even if reintegration does not guarantee reconciliation in the short run, it can stabilize ethnic relations and lay the foundation for reconciliation down the road. In the same way that minority returns are the first step on the way to reintegration, reintegration is the first step toward reconciliation. Assuming post-war reintegration is possible, what are the conditions that favor its emergence? Most importantly, local institutions and authorities must be supportive of efforts to reintegrate. Experimental research has shown that '[t]he effectiveness of interracial contact [in promoting tolerance] is greatly increased if the contact is sanctioned by institutional support,' which can come in the form of 'the law, a custom, a spokesperson for the community, or any authority that is accepted by the interacting groups.'16 Heterogeneous civic organizations can also foster shared interests among antagonistic groups, inducing cooperation across ethnic lines.17 The corollary is that reintegration is wnlikely to succeed under a system of ethnic spoils. Ethnic spoils systems are a natural outgrowth of internecine conflict, wherein one group gains control over a community's resources through war and uses ethnic discrimination to exclude other groups from a share of these resources. Such policies may be as overt as violent expulsions or as subtle as employment discrimination or police harassment. A careful student of ethnic conflicts in Africa, Bates, observed that 'ethnic groups are, in short, a form of minimum winning coalition, large enough to secure benefits in the competition for spoils but also small enough to maximize the per capita value of these benefits.' This logic was later formalized by Caselli and Coleman, who concluded that 'ethnicity provides a technology for group membership and exclusion which is used to avoid indiscriminate access to the spoils of conflict. Without such a technology, groups become porous and the spoils of conflict are dissipated."8 The more people are excluded from a share in these BARRIERS TO REINTEGRATION AFTER ETHNIC CIVIL WARS 373 resources, in other words, the greater the share enjoyed by members of the winning coalition. Ethnic entrepreneurs exploit this logic to shore up their political base in post-war settings. Having secured control over a territory through ethnic cleansing, warlords distribute the spoils of war to members of the in-group - particularly to those in their patronage networks. This enables them to buy the support of the in-group, which keeps them in power. Such systems may be disrupted or weakened when out-groups return to reclaim their property, effectively reducing the size of the spoils. Ethnic entrepreneurs, therefore, have incentives to deter out-groups from returning to reclaim their homes, property, and other community resources. For these reasons, post-war authorities and those who benefit from the spoils system represent a formidable barrier to post-war reintegration. To enhance the success of such programs, interveners must target not only obstructionist elites, but the ethnic spoils system itself. This means disrupting the local patronage networks that make up the elites' power base. To do so, the spoils of war must be redistributed from members of the patronage network (usually members of the local majority) to those from whom the resources were appropriated (usually displaced minorities). This is why property restitution and reconstruction assistance are critical to sustainable minority return. Displaced minorities are unlikely to return to their homes so long as their residences or property is occupied by beneficiaries of the ethnic spoils system. Returnees also require reconstruction assistance to rebuild houses and property that have been destroyed in their absence.19 Finally, sustainable return necessitates assurances of a minimum standard of living. Indeed, practitioners and scholars consistently argue that minorities will not return if they face discrimination in housing, employment, and other social services.20 The speed of this assistance is critical, as the likelihood of return decreases with each year spent in exile.21 THE REALIST CHALLENGE International Theory (IR) scholars from the realist perspective have come out strongly against post-war reintegration under any circumstances. They argue that 'restoring civil politics in multi-ethnic states shattered by war is impossible because the war itself destroys the possibilities for ethnic cooperation.' In this view, the dynamics of ethnic war itself 'create such intense fears and distrust among combatant groups that rebuilding peaceful multi-ethnic societies is bound to fail, no matter what caused the conflict in the first place.22 There are two reasons for this. First, unlike ideological identities, ethnic identities are relatively fixed and salient, so that individuals cannot choose sides in an ethnic civil war, nor can they easily reconcile once the conflict is over. Second, the atrocities and hypernationalist rhetoric that accompany ethnic conflict tend to 'harden[.] ethnic identities to the point that cross-ethnic appeals are unlikely to be made, and even less likely to be heard.'23 Because the hyper-nationalist rhetoric used to mobilize fighters and the military technology common to guerrilla warfare can be employed either defensively or offensively, the combatant parties cannot tell whether their opponents are aggressors. Under this 374 CIVIL WARS so-called ethnic security dilemma, both sides operate on worst-case assumptions concerning the other's motives, creating incentives to strike preemptively to avoid victimization.24 In this context, ethnically mixed settlements present irresistible windows of vulnerability and opportunity for the combatant groups. When surrounded by enemy territory, 'ethnic islands' must strike first to avoid annihilation. Stranded minorities may also 'tempt potential rescuers [national homelands] to jump through any windows of opportunity that arise' to rescue diasporas stranded in enemy territory.25 The resulting spirals of communal violence will not be resolved '[ujnless outsiders are willing to provide permanent security guarantees.'26 In short, ethnic combatants cannot cohabitate after sectarian violence because neither side can trust that the other will not attempt to annihilate them once they have disarmed. Minority returns can themselves trigger violence because 'repatriating any substantial number of refugees back to territory held by the other group risks making control of that territory once again uncertain, thus re-creating the same security dilemma that will help to escalate the conflict in the first place.'27 Furthermore, ethnic combatants are unlikely to disarm because doing so 'forfeits their leverage over their rival, thereby making themselves vulnerable.'28 Even if the rival groups are merged into a common security force, they will maintain their ethnic loyalties for many years to come, a situation that can easily devolve into violence. The only stable long-term solution is to partition the groups into defensible enclaves, thereby permanently resolving the ethnic security dilemma. * * * There is little doubt that post-war reintegration is a difficult proposition - post-war communities cannot be reintegrated if displaced groups do not return to their homes, and return programs have a disappointing record of success. However, from a policy point of view, it pays to understand why such programs founder. If it is obstructionist elites and the logic of ethnic spoils, then post-war reintegration is difficult but not impossible. If it is intractable grassroots fears and hostilities, then reintegration is not only impossible, but also dangerous. Although both mechanisms may be at play, I argue that the world in which ethnic patronage networks are the principal barriers to ethnic co-existence is very different from the one in which grassroots fears and hatreds are the primary obstacle. Ethnic spoils logic implies that returns will be higher when patronage networks are disrupted, allowing the displaced minorities to reclaim their pre-war property. They should also be higher when obstructionist authorities are sanctioned, permitting ethnic minorities to resettle in their homes. Finally, returns programs will be more likely to succeed when returning minorities are protected from harassment and have some assurance of employment and access to social services. By contrast, the ethnic security dilemma holds that reintegration is unlikely to succeed under any circumstances and is least likely to succeed in communities that suffered the greatest atrocities due to the intractable fears and hostilities arising from horrific wartime memories. The security dilemma model also predicts that resistance to reintegration BARRIERS TO REINTEGRATION AFTER ETHNIC CIVIL WARS 375 (in the form of minority returns) will be spontaneous and bottom-up rather than orchestrated by local or state elites. Finally, violence should be most pronounced in areas of mixed ethnic settlement because ethnic 'islands' are vulnerable to attack and, therefore, have incentives to strike out preemptively. The success of minority returns will be measured as the number of registered ethnic minority returns to a locality; this is the first and most critical stage of post-war reintegration. Additional measures of reintegration include minority (re)incorporation into local government institutions - including legislative assemblies, law enforcement, and schools. Post-war reintegration is judged to be higher where the rate of minority returns is higher and where community institutions enjoy a greater degree of minority participation. For figures on minority returns, I rely on statistics provided by the UN High Commissioner for Refugees (UNHCR) - the most reliable repository for such information. For measures of minority participation and sustainability of minority returns, I use qualitative assessments gathered in field interviews with participants and observers on the ground, as well as opinion polls and reports by scholars and journalists who have interviewed minority returnees as well as members of the local majorities. Elite obstruction is measured as organized intimidation of minority returnees, refusing to evict the occupants of returnee properties, and using procedural or other obstacles to impede return. Efforts to combat elite obstruction consist of sanctioning or removing obstructionist authorities from office or conditioning economic assistance upon demonstrated progress in minority returns. It is impossible to say how much elite obstruction must be countered, how much aid must be provided, or how timely this assistance must be to ensure sustainable minority returns. However, I expect to see a rise in registered returns when authorities are severely sanctioned for obstructing return and restitution. Similarly, where property is restituted within the first few years after the war or when returnees receive substantial assistance to resettle in their pre-war communities, we should see higher and more sustainable minority returns. It should be noted that several theories of refugee return are not tested in this paper. For example, individuals are more likely to return when they have a network of support in their pre-war community.29 Individuals are also more likely to return when they have job opportunities in their pre-war community or when the relative standard of living in the pre-war community is higher than that of their host country.30 Although many of these factors can explain variation in returns on the individual level, they cannot explain variation in returns at the group level. Since there is no reason to assume that the demographic profile of refugees varies systematically across the three cases, differences in program success cannot be attributed to individual-level factors. Also, while variation in repatriation31 across the three cases might be explained by variable repatriation rates by the primary host country or by the relative standard of living between the host and home countries, these factors cannot explain variable minority returns across these cases. This is because repatriated refugees do not necessarily return to their pre-war homes - forcibly repatriated refugees are more likely to go to collective centers or to areas dominated 376 CIVIL WARS by their co-ethnics. Variable rates of minority return across communities are, therefore, best explained by factors in the community itself. THE EMPIRICS Croatia When Croatia declared independence in June 1991, Serb rebels in the Krajina region launched a secessionist bid of their own. In the early stages of the conflict, the Serbs (which made up 12 per cent of Croatia's population of four million) enjoyed the upper hand. With the help of the Yugoslav People's Army (JNA), the separatists gained control of Krajina and eastern and western Slavonia (one-third of Croatia's territory), expelling 220,000 ethnic Croats from the region.32 The rebels thereupon declared an independent Republic of Serbian Krajina. Croatian forces began to beat back the separatists in late 1991, and a peace agreement was concluded between Croatia and Serbia in 1992 that established the UN Protection Force (UNPROFOR) in Krajina with a mandate to protect Serb civilians from reprisals and facilitate the return of Croat refugees. In 1995, the Croatian army reclaimed the remaining territories in Operations Flash and Storm. This time it was Serb civilians who were forced to take flight. From 1991 to 1995, over 300,000 minority Serbs escaped to Bosnia, Serbia, and eastern Slavonia (still controlled by Serb rebels).33 At the end of the war, Franjo Tudman's government institutionalized and consolidated a return and restitution regime that discriminated against displaced Serbs in favor of Croats in the areas of property repossession, reconstruction, and access to social services and pensions.34 This contributed directly to the ethnic homogenization of Krajina; less than one year after Operations Flash and Storm, fully half of all Croat refugees had spontaneously returned to their homes.35 By contrast, only a few thousand minority Serbs had returned to their pre-war residences; most of these were elderly or had family ties to the community. When Tudman died in 1999, a more moderate government came to power with a promise to improve the situation of ethnic minorities. In a newspaper interview, the new Croatian president, Stipe Mesic, called for the return of all Serbs who had fled during the war. To demonstrate his sincerity, the government announced a 55 million USD program to assist minority returns. These actions were lauded by the international community, and Zagreb was promised funding for the program through under the Stability Pact.36 Croatia was also admitted to the World Trade Organization in July 2000 and NATO's Partnership for Peace program in spring 2000. The European Union (EU) announced that it would begin accession negotiations with the country in mid-2000. The new government sought to make good on its promises by amending the system of property restitution, removing much of its discriminatory character. The claims process was regularized, making restitution more transparent and fair. The ruling coalition also adopted a more conciliatory tone toward ethnic Serbs, increased social assistance to minority returnees, and invited minority representatives into the government. In 2002, the Croatian parliament passed the BARRIERS TO REINTEGRATION AFTER ETHNIC CIVIL WARS 377 FIGURE 1 MINORITY RETURNS IN CROATIA 80000 70000 60000 50000 40000 30000 20000 10000 0 ■ Minority returns , Peak at 20,716 in 2000 CO 00 CT> O T— CM CO CO CO CT> CT> CT> CX> o o O o o o O o O CT> CT> CT> CX> o o O o o o O o O 1— 1— 1— 1— CM CM CM CM CM CM CM CM CM Constitutional Law on the Rights of National Minorities, mandating proportional representation of minorities in the civil service.37 Due to the emergence of a more moderate leadership, minority returns doubled from 1999 to 2000.38 However, minority returns declined every year thereafter (see Figure 1); by 2005 only about one-third of the Serbian refugees had returned to stay. At the same time, all but a few thousand ethnic Croats had returned to their homes.39 The 2001 census revealed that ethnic Serbs now made up 4.5 per cent of Croatia's population - a two-thirds drop from 1991.40 The low rate of return is partly due to destroyed houses and infrastructure that disproportionately affected Serbian villages; moreover, economic assistance to returnees was limited and had come far too late to ensure substantial and sustainable minority return. There is every indication that returns would have been higher had there been adequate employment opportunities and reconstruction aid. A survey conducted in 2003 (see Figure 2) reported that 'up to 42 per cent of Serb refugees in Serbia and Montenegro and Bosnia and Herzegovina might return if there were access to housing and improvements in the economy.'41 One Serb refugee explained her decision to remain in Belgrade: 'The assistance we would get would barely cover the rent, let alone let us buy up the flat, as we would have done had we not been forced to leave.'42 Employment discrimination was another significant barrier to sustainable return, and this was also largely attributable to the actions of state and local authorities. Public sector jobs in schools, hospitals, law enforcement, national parks, post offices, and other areas of civil administration were effectively closed to Serbs. Human Rights Watch reported that communities governed exclusively by Croatian parties had almost no Serbs working in the public sector, in contrast to communities with mixed-ethnicity governments, which hired civil servants from both ethnicities.43 Indeed, the fact that few Serbs have returned besides the rural and elderly is better explained by the lack of jobs and cancellation of Serb occupancy rights in urban areas than the purportedly minimal security threat posed by elderly, rural returnees.44 378 CIVIL WARS According to ethnic security dilemma logic, the reintegration of Croats and Serbs should have been next to impossible due to irresolvable ethnic fears and hostilities generated by memories of war. Contrary to this expectation, over one hundred thousand Serbs did return to their homes once they had the opportunity to claim their homes. However, many returnees left once again due to job discrimination, the lack of economic and housing assistance, and harassment by extremist groups. According to one report, 'although the improvement of the political climate is undeniable, such progress has come quite late in the process and many Internally Displaced Person(s) (IDPs) have already rebuilt their lives elsewhere and given up on return.'45 The poor record of returns in Croatia is largely due to Tudman's ethnic spoils system and the international community's failure to impose conditions on Croatia's accession to international organizations. Five years after the peace agreement, wartime authorities remained in power, consolidating a system of ethnic spoils that discriminated against Serb returnees. In 2000, a more conciliatory government began to dismantle the returns and restitution system, and the rate of minority returns shot up almost immediately. Later, many Serbs left, however, due to job discrimination and harassment at the local level. As a result, the initial successes of the early 2000s quickly evaporated, and the number of Serb returnees declined every year thereafter. Poor ethnic reintegration could also be seen in law enforcement, as ethnic Serbs made up only three per cent of Croatia's police officers.46 The exception is eastern Slavonia, where the internationally brokered 1995 Erdut Agreement required the government to maintain proportionality in the number of Serbs and Croats on the police force; the result has been high minority participation in the region's police force. This suggests that in law enforcement as well, the low level of reintegration had less to do with grassroots fears and hatreds than with hiring discrimination on the part of local and state authorities. Bosnia-Hercegovina As late as 1992, Bosnia was a peaceful multi-ethnic republic in which Muslims, Croats, and Serbs cohabitated and even intermarried.47 In March 1992, an independence referendum garnered broad popular support among Croats and Muslims, but was vehemently opposed by the Serbs.48 With the Serbo-Croatian war in full swing, Bosnia's president, Alija Izetbegovič, tried to protect the country from external intervention, but pitched battles soon escalated into a full-blown war between the Bosnian Muslims, the Belgrade-backed Bosnian Serbs, and the Zagreb-backed Bosnian Croats. Ethnic cleansing by paramilitaries gradually transformed the once heterogeneous country into a collection of ethnically homogeneous cantons. Under the Dayton Peace Agreement, these territories gained autonomous status as the Federation for Croats and Muslims (the Federation) and the Republika Srpska (RS). By the time the dust had settled, over 100,000 people had been killed and 2.3 million displaced.49 Of those displaced, 1.3 million were refugees and 1 million IDPs.50 The Dayton Accords established the Office of High Representative (OHR) to implement the terms of the peace agreement; a NATO force of 60,000 peacekeepers BARRIERS TO REINTEGRATION AFTER ETHNIC CIVIL WARS 379 (the Implementation Force or IFOR) was put in place to enforce it. The UNHCR was tasked with returns and property restitution.51 According to Annex 7, '[t]he early return of refugees and displaced persons is an important objective of the settlement of the conflict in Bosnia and Herzegovina.'52 Despite the clear mandate for refugee return, the UNHCR had no enforcement powers and was only able to achieve returns that were unopposed by local authorities. This only applied to majority returnees, who made up a small proportion of the displaced population. Some 250,000 (predominantly majority) returns took place in the first two years.53 The vast majority of wnresolved claims involved minorities who had been ethnically cleansed - Serbs driven out of the Federation and Croats or Bosniaks pushed out of RS. The failure of reintegration in the early post-war period was directly attributable to local warlords seeking to consolidate their wartime spoils. Now in the post-war administration, these officials used their control over housing stock, social services, public jobs, and education to entice co-ethnics to migrate to their enclaves while preventing minority returns.54 One study of Bosnian peacekeeping noted that [h]ard-line nationalist political leaders frequently urged fugitives of a given ethnic group to settle in communities captured by their side in the war. This served both to concentrate their own ethnic population and consolidate their political grip on territory acquired during the war.55 RS authorities were among the worst offenders in this regard. By 1999, they had resolved only 1.6 per cent of registered property claims and had not yet reversed the wartime firings of Bosniak and Croat civil servants.56 Similarpatterns could be observed in the Federation. A member of the Croat minority in Bihac observed that '[ojrdinary people do not have a problem with my language. It's the [local] government who has the problem; it's trying to make me go.'57 The barrier to returns was not so much ethnic animosities on the ground as local officials who 'discouraged returnees in the interests of creating ethnically pure blocs' as a means of shoring up their power base.58 International administrators tried to combat elite obstructionism with 'a mix of bribes, threats, and other leverage available.'59 The problem was that they 'had no authority to coerce returns, [they] needed a legal system to implement these rights.'60 A former Organization for Security and Cooperation in Europe (OSCE) representative concurred: 'Until 1997, people didn't even have the right to reclaim [their property] because the municipalities had fixed the law.'61 According to an NGO assisting refugees and IDPs, 'Certain municipal authorities were very obstructionist; they said there were too few officials, too many requests, or that [the case] was "not in their jurisdiction." It was not clear who was in charge of what.'62 The reluctance of the OHR to confront wartime authorities meant that there were very few minority returns in the early postwar period. The Turning Point As Germany and other west European governments prepared to repatriate as many as 200,000 Bosnian refugees, it was feared that continued elite obstructionism would 380 CIVIL WARS cement post-war divisions once and for all.63 The Peace Implementation Council (PIC), therefore, gave the OHR the power to sanction or even replace local officials who were blocking the process. Equipped with these so-called 'Bonn powers,' the OHR adopted a more coercive approach to returns and restitution. In 1997-99, the OHR-led Reconstruction and Return Task Force (RRTF) and the new Property Law Implementation Plan (PLIP) began to enforce property restitution, which greatly assisted with minority returns: '[ojfficials were dismissed who refused to implement the new property laws. Implementation was monitored on the municipal level and local authorities were pressured to resolve outstanding cases. The International Police Task Force (IPTF) supervised local police to ensure that the necessary evictions were carried out.'64 PLIP officials also kept a list of obstructionist officials who were threatened with removal if they continued to block returns. Minority returns were further facilitated by a symbolic show of force. Since the main obstacles to return were local authorities, all that was usually required was an SFOR escort to assist returnees in reclaiming their homes. This simple show of force signaled to local leaders that the international community intended to enforce returns. According to Moratti, 'the fact that SFOR was around created a general [environment of] deterrence' against elite obstructionism. Once local elites had fallen into line, there was little grassroots resistance with which to contend.65 The rate of minority returns increased dramatically as a result of these policies. In 2000, minority returns doubled over the previous year and peaked in 2001-02.66 By June 2008, fully 1 million of the original 2.2 million refugees and IDPs had returned to their homes, including an estimated 446,000 minority returnees.67 This represents fully one-half of all displaced minorities in the Bosnian war; meanwhile, over 90 per cent of property claims had been adjudicated. Rather than incite ethnic violence, minority returns were accompanied by a dramatic increase in popular support for reconciliation, suggesting that increased contact with members of the other group may have depressed ethnic animosities. Surveys showed a steady rise in FIGURE 2 MINORITY RETURNS IN BOSNIA AND HERCEGOVINA Peak at 80,480 in 2002 80000 70000 60000 50000 40000 30000 20000 10000 BARRIERS TO REINTEGRATION AFTER ETHNIC CIVIL WARS 381 support for minority returns across Bosnia, even in RS. The most radical shift in public opinion could be seen in Serb-majority areas, where popular support for reintegration shot up from 18 to 81 per cent between 1995 and 2004.68 According to another survey in 2005, only 18 per cent of Bosnians had friends exclusively within their own ethnic group. Serbs, Bosniaks, and Croats most often said that their friends were 'from mixed nationalities,' and nearly half of all respondents claimed to want more friends from other ethnicities.69 Interestingly, the strongest predictor of ethnic tolerance was the level of daily contact across ethnic lines (confirming the contact hypothesis), while the level of wartime trauma was a weak or insignificant predictor of attitudes toward other groups (contradicting the ethnic security dilemma hypothesis).70 Ethnic reintegration could also be observed in the area of law enforcement. At the end of the war, Bosnia's police were little more than ethnic paramilitaries that were ill-equipped for civilian protection; little changed in this respect through the end of the 1990s. Under the 2000s reforms, however, the UN and later the EU actively recruited minority officers for all 12 territorial police administrations. The result was that minority police officers increased from 1.5 to 15.5 per cent for the Federation and from a handful to five per cent in RS from 1999 to 2002.71 The entity armies also agreed to merge under a unified command structure as a condition of EU accession. Although the lower ranks of the military and much of the police force remained ethnically divided, these events suggested that former combatants could be induced to form a unified force. Contrary to security dilemma expectations, minority returns in Bosnia were accompanied not by increased strife, but by increased tolerance at the grassroots level. Also against security dilemma predictions, the spike in minority returns did not occur due to an enhanced external security guarantee. In fact, the number of SFOR troops in Bosnia was halved to 30,000 in the late 1990s, with NATO announcing further force reductions.72 The surge in minority returns occurred despite indications of a diminishing external security guarantee, primarily because OHR had begun to sanction and remove obstructionist officials. Overall, the return programs in Bosnia met with moderate success, largely due to delays in enforcement and insufficient aid to minority returnees. In the intervening years, many displaced minorities put down roots elsewhere, despite increased levels of ethnic tolerance in both entities. According to one analyst, 'the international community should have made a much stronger stand right after the war to show that they were going to use their full mandate. [Had they begun in this way,] the pre-war [demographics] could have been restored to a much greater degree.'73 This suggests that reintegration was a viable alternative to partition and may have enjoyed greater success had the international community sanctioned elite obstructionism and assisted minority returnees from the very beginning. Kosovo NATO launched an air war against the Federal Republic of Yugoslavia (FRY) in spring 1999 after negotiations to reach a peaceful settlement between Belgrade and the Kosovo Liberation Army (KLA) broke down. Under the cover of NATO 382 CIVIL WARS bombing, Yugoslavia began to ethnically cleanse the Kosovo region - burning villages and terrorizing the ethnic Albanian population. Around 863,000 Albanians fled across the border into Macedonia and Albania where they were housed in UN-administered refugee camps; another 590,000 became internally displaced.74 After three months of bombing raids, Milosevic agreed to withdraw his forces from Kosovo. UN Resolution 1244 set out the terms of the peace, establishing the UN Interim Administration Mission in Kosovo (UNMIK) as the transitional government. The resolution mandated that the international community '[ensure] a secure environment in which refugees and displaced persons can return home in safety' (Art. 9c). Security was to be provided by the NATO-led Kosovo Force (KFOR). Meanwhile, the KLA signed a demilitarization agreement under which it agreed to hand over its weapons to KFOR. Nearly, all the Kosovar Albanian refugees returned to their homes shortly after the cease-fire.75 At the same time, however, Serbs began to flee the province in response to threats and intimidation by Albanian extremists. Following early spontaneous purges of Serbs, the cleansing of minorities took on a certain patterned logic 'suggestive of a period of planning or of terror cells operating in certain areas.'76 In all, some 200,000 Kosovar Serbs and Roma fled or were driven out after UNMIK assumed control of the province, in full view of the UN, the OSCE, the EU, and 40,000 peacekeeping troops.77 Few Serb refugees later returned to their homes. Minority returnees made up a paltry 1,906 in 2000 and peaked at 3,801 in 2003. In May 2004, riots broke out, and Albanian gangs and paramilitaries drove an additional 4,000 Serb civilians from their homes - looting and burning their houses and property (Figure 3). The failure of the international community to protect minorities from violence convinced prospective returnees that it was unsafe to return to their communities, significantly FIGURE 3 MINORITY RETURNS IN KOSOVO 80000 70000 60000 ■ Minority returns 50000 40000 30000 20000 10000 Peak at 3,801 in 2003 0 o o o CM O O CM —I— CM O O CM —I— CO o o CM —r~ ■* o o CM -1— o o CM -1— CO o o CM —I— o o CM 00 O O CM BARRIERS TO REINTEGRATION AFTER ETHNIC CIVIL WARS 383 depressing minority returns. Overall, 15,280 minority returns were registered from June 2000 to June 2006 - less than one-tenth of the displaced minority population.78 Wartime elites and extremist elements played a key role in obstructing minority returns. With backing from Belgrade, Serb leaders quickly established a Serbian parastate in northern Kosovo, while Kosovo Albanian leaders - many of them former KLA commanders - homogenized the territory south of the Ibar River. The Serbs who remained in the province were ghettoized in northern districts and southern enclaves. International administrators failed to prevent the de facto partition due to their unclear mandate and unwillingness to forcefully confront either side. UNMIK National Political Affairs Officer Ardian Arifaj explained, There was not a lot of planning for a mission on the scale of UNMIK.. .There was no clear guidance as to what they wanted to do or what they wanted to achieve; they came not to create a country but rather to keep the two sides from killing each other.79 Minority returns were manipulated by both sides in a struggle over Kosovo territory. UNMIK Head, Lamberto Zannier, told the UN Security Council that the sharp fall-off in minority returns was directly attributable to the policies of Prishtina and Belgrade.80 In the run-up to the declaration of independence in 2008, Prishtina slashed funding to Serbian enclaves in the south and cut the annual budget for minority returns. Albanian majority municipalities with the fewest returns were run by officials who 'were at best lukewarm' and sometimes 'openly hostile' to the prospect of organized returns. For these authorities, 'displacement and return [was] viewed as a political bargaining chip for independence.'81 For their part, Serb separatists used minority returns to shore up control over northern Kosovo. Belgrade supported the separatist government by paying the salaries and pensions of civil servants, financing social services and infrastructural development, and controlling movement in and out of the area.82 In collusion with Belgrade, the Serb coalition in the Kosovo Assembly called for the 'massive return of Serbs' (in some cases to strategic locations where they had not lived previously) as the 'Serb parallel system integrate[d] itself more closely into Belgrade's institutions.' While the majority of would-be Serb returnees wanted to return to Kosovo, most concluded that they would be 'pawns in a political game' and opted to resettle elsewhere.83 Another barrier to minority returns was the ineffectual system of property restitution. The UNMIK and the Provisional Institutions of Self-Government (PISG) had created a hybrid international-domestic system of property restitution, the Housing and Property Directorate and Claims Commission (HPD/CC), which had a mandate to restitute property to refugees and IDPs.84 Almost immediately, the system was undermined by delays in authorizing and organizing special commissions to hear the cases, inadequate funding, and a lack of qualified and/or unbiased local staff - slowing the claims process down to a trickle.85 By the 384 CIVIL WARS mid-2000s, many of these problems had been rectified, and nearly all of the 30,000 cases of contested residential property were resolved. However, even with the decisions in hand, the system lacked enforcement power, leading to widespread cynicism over HPD adjudications.86 One prospective returnee observed that 'HPD is just handing out pieces of paper - everyone already knows who owns which house; the issue is how we can return to live there?'87 The director of a local NGO confirmed that in Kosovo 'there is a much softer, more hands-off approach [than in Bosnia]. There are no sanctions for non-cooperation by municipal authorities.'88 Matters improved somewhat with the introduction of USAID-funded Municipal Infrastructure and Support Initiatives (MISI-1 and MISI-2), under which local governments received funding in exchange for encouraging minority returns and participation in local government. However, the system had no sanctions for non-compliance, greatly limiting its effectiveness. At the time of this writing, minority returns in Kosovo have been negligible. The flow of Serb returnees slowed to a trickle after the 2004 riots and practically ground to a halt after Kosovo's declaration of independence in 2008. The reasons displaced Serbs have not returned are straightforward - ethnic elites have systematically blocked minority returns and the international community failed to impose sanctions against such policies. International peacekeepers stood by as the KLA and other extremist elements pushed out hundreds of thousands of Serbs at the end of the war. During the March 2004 riots, the peacekeepers failed to prevent the burning and looting of thousands of Serbian homes. Minority returns in post-war Kosovo have foundered not because of grassroots hatreds and fears from wartime memories, but because Serb separatists used ethnic engineering to gain de facto control over territory in the north, while Albanian authorities blocked minority returns to consolidate control over the south. The international community permitted the operation of ethnic spoils systems on both sides. Perhaps surprisingly, the reintegration of law enforcement has met with greater success. As of 2010, the Kosovo Police Service (KPS) had a total of over 7,000 officers, nine per cent of which were Serbs and seven per cent other minorities, thus meeting their goal for minority recruitment.89 This modest success was due to the fact that the international community, and not local authorities, headed up the KPS where it actively pursued Serb recruitment. Against security dilemma expectations, the ethnic integration of police did not lead to ethnic divisiveness in the units. Despite a handful of incidents during the 2004 ethnic riots and the 2008 declaration of independence, the multi-ethnic police force has acted professionally. A survey of the cohesiveness of multi-ethnic police units in the Balkans concluded that: [i]n spite of the significantly higher level of inter-ethnic violence during the Kosovo conflict, the inter-ethnic social climate at the KPSS and in the multi-ethnic units was judged to be almost as good as in South Serbia and Macedonia. This is the result of the extensive use made by UNMIK and OMIK of their capacity for intervention.90 BARRIERS TO REINTEGRATION AFTER ETHNIC CIVIL WARS 385 Although Serbian and Albanian police officers are still largely segregated because the ethnic groups themselves are segregated, the Kosovo case suggests that reintegration of combatant groups is not the impossibility that partition advocates claim it to be. I now provide a snapshot of two (relative) success stories at the local level. These cases demonstrate that even in the wake of extreme internecine violence, post-war reintegration is possible so long as ethnic spoils systems are effectively challenged. Success Stories? Brčko and Prijedor The Bosnian city of Brčko was the site of some of the worst wartime atrocities. Prior to the conflict, the town was a multi-ethnic town of Muslims, Croats, and Serbs. When Serb forces captured the city, it was ethnically cleansed of non-Serbs, and as many as 3,000 mostly Muslim men, women, and children were tortured and executed at the Luka-Brčko camp in 1992.91 At the end of the war, it was decided that Brčko would be divided between the RS and the Federation and governed internationally. Although minority returns were an important element of the Final Arbitral Award, RS officials blocked returns and restitution at every turn - resettling displaced Serbs in minority homes and using them as 'human shields.' They also restricted movement into Brčko by levying illegal tolls and demanding 'Republika Srpska visas' from motorists. The displaced minorities that did return to their homes were often met by hostile Serbs in demonstrations that were orchestrated by RS officials and extremist leaders.92 In contrast to OHR, Brčko Supervisor Robert Farrand had a special mandate to enforce minority returns using international police monitors and an ethnically mixed police force. He used his authority to confront obstructionist elites by banning relocation orders that did not have his prior authorization. He also ordered the police to discontinue illegal tolls, a directive with which RS officials were ultimately forced to comply. As a consequence of these actions, Brčko was the only area of RS with 'substantial minority returns' in 1997.93 By 2008, the district registered almost 20,000 minority returns.94 Because the minorities constituted a substantial voting bloc, politicians were forced to appeal across ethnic lines, leading to a marked decline in nationalist voting and the emergence of inter-ethnic political coalitions at the local level.95 Reintegration progressed in other areas as well; Brčko had the first fully integrated school system, police force, and trade union in post-war Bosnia. This was accomplished not by repressing grassroots hostilities, but through targeted sanctions against RS and local wartime authorities. Besides enforced returns, the economic aid flowing into Brčko also facilitated ethnic reintegration. Minority returnees received economic assistance to resettle in their communities, including substantial aid packages to repair their homes. All of this helped to promote ethnic tolerance on the ground. According to one local vender, 'When you have work to do and can plan your future, there's no place for other things in your mind.'96 Despite significant steps toward reintegration, reconciliation is a long way off, as ethnic mistrust 386 CIVIL WARS remains at the grassroots level.97 At any rate, violence has not re-emerged, and if the contact hypothesis is to be believed, daily inter-group engagement may help ease ethnic tensions over the long run. * * * Prijedor, a Bosnian city in RS, was another site of wartime atrocities. Before its takeover by Bosnian Serb militias, the town was a mix of ethnicities, primarily Serbs and Muslims. Over the course of the war, tens of thousands of non-Serbs were deported, and thousands of civilians were interned in war camps where they endured systematic torture, rape, and executions; thousands were estimated to have been killed. As in Brčko and elsewhere, the very warlords who had ordered mass executions, deportations, and detentions assumed leadership positions in the post-war administration. The Prijedor mayor and chief of police were two of the principal masterminds behind the ethnic cleansing campaigns. After the war, they sought to consolidate their ethnic enclaves by organizing displays of violence during the visits of prospective Bosniak returnees and by publicly vowing to block minority returns. At first, the international community sought to avoid a confrontation by blocking tens of thousands of displaced Bosniaks who were planning to return to their homes in Prijedor. In response, 40,000 Bosniaks moved to a town just over the entity border and waited for their opportunity to reclaim their homes.98 The opportunity came in mid-1997, when NATO forces moved to arrest indicted war criminals in Bosnia. In Prijedor, they captured one and killed another (the chief of police); SFOR also removed police officers who had been implicated in wartime atrocities. NATO thus 'beheaded the wartime authority,' paving the way for tens of thousands of Bosniak returnees.99 The Bosniak returnees ultimately had 'a moderating impact on Prijedor's political life' because their elected representatives were able to form a governing alliance with the moderate Serbs, effectively ousting the hardliner Serbs from leadership positions.100 This ushered in a succession of inter-ethnic local governments. Despite substantial minority returns and the emergence of inter-ethnic governance, Prijedor did not enjoy the same degree of reintegration as in Brčko due to the direct international administration of the latter. The local police force, for example, was still dominated by Serb officers because RS authorities maintained control over hiring. This underscored the critical importance of international involvement in all areas of post-war reintegration. EVALUATING THE CASES The Balkan cases shed important light on the viability of post-war reintegration. According to the ethnic security dilemma, reintegration should have been least successful in areas with the greatest wartime violence. However, Bosnia - with the most intense and protracted sectarian violence of the three - experienced the greatest success in reintegration. It is also notable that Brčko and Prijedor, sites of extreme wartime atrocities, achieved relative success with post-war reintegration. BARRIERS TO REINTEGRATION AFTER ETHNIC CIVIL WARS 387 It can be argued that the reintegration in Brčko and Prijedor was only possible because of significant security guarantees, which suppressed but did not eliminate post-war hostilities. However, suppression alone cannot account for moderation in every sphere of public life in Brčko, which has boasted inter-ethnic governments, declining support for nationalist politicians, and ethnically integrated schools. Also, minority returns were not successful in Bosnia until the late 1990s, by which time the international peacekeeping force had been significantly downsized with further planned reductions. It is worth noting that Bosnia and Kosovo - with the most and least successful programs, respectively - had roughly equivalent external security guarantees. What made the difference was not the size of security forces or level of violence (which predicts the reverse outcome), but targeted sanctions against wartime elites who were seeking to preserve their ethnic enclaves. The success of the returns programs also varied over time. After Dayton, the peacekeepers and police in Bosnia were reluctant to use sanctions to counter the wartime authorities and declined to intervene in returns and property restitution except to escort OSCE personnel on field visits. The turning point came in 1997, whentheOHR began to remove obstructionist authorities and enforce evictions of illegal occupants. In Croatia, by contrast, most of the displaced Serbs did not return because the discriminatory legislation passed by the Tudman government had created a climate of extreme insecurity for the Serb minority. Although there was a marked uptick in Serb returns when a pro-minority government came to power in 2000, many returnees later left because ethnic patronage continued to operate on the local level. The importance of protecting returnees was particularly evident in Kosovo, where the repeated failure of KFOR and UNMIK to shield ethnic Serbs from harassment and intimidation by ex-KLA and other extremist groups signaled to the Serbs that it was unsafe to live in Albanian majority areas. Economic aid was also critical to the success of minority return programs. Financial aid was made available to returnees sooner in Bosnia than in Croatia, stating that many more minorities were able to return to their homes because they had not yet settled elsewhere. Minority returnees in Brčko were assisted at all stages of property restitution and received generous grants to repair their homes. By contrast, the lack of enforced returns and aid to returnees in Kosovo meant that minorities there were unable to reclaim their homes; they, therefore, resettled in Serbia or northern Kosovo. The record of reintegration in other areas of community life, such as law enforcement, testifies to the critical role played by the international community in post-war settings (Table 1). Mixed-ethnicity police have emerged not in places with the least wartime violence, but where international administrators played an active role in recruiting minority officers - Kosovo, eastern Slavonia, Brčko, and Bosnia as a whole (particularly the Federation). In fact, eastern Slavonia, which suffered the worst atrocities of the Serb-Croatian War, now has the most integrated police force in Croatia due to the terms of the internationally brokered Erdut Agreement that reincorporated the region into Croatia. As one scholar notes, 'of the three components of the "peace-building triangle" - local capacities, international intervention, and level of war-related hostilities - the third has proved to be the least consequential."01 TABLE 1 COMPARING POST-WAR REINTEGRATION IN THE BALKANS Property Restitution External Security Wartime Enforcement of Returns and Assistance Guarantee Atrocities Outcome Croatia Low/medium (EU/OSCE monitoring; limited membership conditionality) Low/delayed (ad hoc government assistance programs. post-2000 reforms to property restitution laws) None3 Medium Low minority returns (~ 109,000, about one-third of total); minimal minority participation in police (about three per cent) Bosnia (1996- -97) Low (voluntary returns only; Low (entity governments in charge High High Almost no minority no penalties for non-compliance) of property restitution; no assistance to returnees) (SFOR/IFOR)b returns (less than 10,000 or about one per cent); almost no participation in local police Bosnia (1998- -2008) High (IPTF sanctions against uncooperative officials) High (PLIP; RRTF; reconstruction assistance from UNHCR, World Bank and other development agencies) Medium/low High Moderate minority returns (-436,000 minority returns, about half of total); modest participation in local police (about 10 per cent) Kosovo Low (no penalties for Minimal (ad hoc municipal Medium/high Low Minimal minority returns non-compliance) aid programs; ad hoc programs of returnee assistance) (KFOR) (15,000-16,000 or eight per cent of total); modest participation in local police (around 15 per cent) aWith the exception of eastern Slavonia. ''Most minority returns occurred after the external security guarantee was significantly downsized. BARRIERS TO REINTEGRATION AFTER ETHNIC CIVIL WARS 389 CONCLUSION These cases suggest that challenging ethnic spoils systems and providing timely assistance to minority returnees is critical to the success of these returns programs. In the absence of these conditions, minority return is unsustainable at best and catastrophic at worst. This implies that the international community should honestly assess whether it can commit to ensuring these conditions before embarking on a large-scale program of reintegration. If these conditions are present, then reintegration is clearly preferable to ethnic partition, which rewards ethnic cleansing and may lead to a destabilizing proliferation of lawless ethnic enclaves. If such conditions are not present, however, ethno-territorial partition and/or the resettlement of evacuees are the only viable alternatives. These lessons have obvious applications outside the Balkans. The Iraq war, for example, has produced refugee flows of crisis proportions. Over the course of the insurgency, Sunni and Shi'a militias ethnically cleansed neighborhoods across the country in a bid to 'consolidate "their" territory by expelling the "others."'102 Over two million Iraqis fled to neighboring countries (with Syria and Jordan sheltering over 1 million and 450,000, respectively), while 2.8 million Iraqis became internally displaced.103 As the international community struggles to deal with the worst refugee crisis in the Middle East in over 60 years, the question remains whether ethnic reintegration is a realistic alternative to ethnic partition in Iraq. Clearly, recreating a peaceful multi-ethnic state necessitates the return of millions of minority refugees to their pre-war homes. The central question is whether this can be done safely and sustainably. In the meantime, the international community must act quickly to ease the burden of caring for refugees from Jordan and Syria, whose systems have been stretched to the breaking point by a quantum leap in their refugee populations. The alternative to reintegration in Iraq is a 'soft' partition of the country along the lines of what has happened in Kosovo, permitting refugees to resettle in areas where they belong to the local majority. This carries risks of its own, namely that those residing in the 'wrong' region may be targeted or even killed. It may also generate centrifugal pressures as rival sectarian groups consolidate power over their respective regions at the expense of the central government. This could lead to a wider regional conflagration if Iran, Syria, or Saudi Arabia initiates proxy wars on Iraqi territory to gain control of Iraq's ethnic cantons and by extension its oil wealth. Ethnic partition would also provide no immediate relief for the millions of Iraqis who are presently displaced both inside and outside of the country. As of this writing, the security situation has improved significantly in Iraq, and in recent years, many refugees and IDPs have begun to return to their homes. However, large-scale returns have not yet taken place, and the majority of 1.5 million IDPs remain in Iraq with no solution in sight.104 This is largely because, as in the Balkans, rival sectarian leaders have settled members of their own ethnicity in the homes of minorities in a bid to consolidate their hold over territory and are actively blocking returns. This situation is unlikely to change unless the international community provides the forces and funding necessary to protect minority returnees from hostile local militias. 390 CIVIL WARS Sustainable return in Iraq will most likely require a long-term troop presence to ensure that externally backed forces are kept in check. It may also necessitate a comprehensive jobs program, housing and public services, repairs of infrastructure, and so on. The third option is maintaining the status quo. This means staying the course as the refugee populations progressively destabilize neighboring countries, leaving the refugees themselves in a permanent state of limbo. As with the Palestinian diaspora, displaced Iraqis may serve to perpetuate conflict in Iraq, providing a potentially limitless source of fighters and weapons from over the border.105 Given the low probability that the resources necessary for ethnic reintegration in Iraq will be provided by the international community, repatriation along the lines of what is being attempted in Afghanistan would constitute a disaster for all involved. The best medium-term solution would be for international agencies to resettle Iraqi refugees in countries outside the Middle East. Substantial minority returns should not be attempted unless and until the world community commits to paying the steep costs of sustainable return. In lieu of this, a peaceful multi-ethnic Iraq remains illusory. ACKNOWLEDGMENTS The author gratefully acknowledges the support of the Carnegie Corporation, which provided generous funding for the research on which this paper is based. She also thanks Patrick James, Stephen Saideman, Idean Salehyan, Florian Beiber, Rhodri Williams, and Charles Philpott for extensive feedback on earlier drafts of the paper. NOTES 1. Majority returnees are those who return to communities where they belonged to the local majority, while minority returnees are those who return to communities where they belonged to the local minority. 2. 'Civil conflict' is defined here as an armed combat between two or more groups or between the government and one or more groups at the sub-state level. 'Ethnic conflict' refers more specifically to civil conflict waged along sectarian or communal lines rather than ideological or class lines. 3. UNHCR. 2008. Global Trends: Refugees, Asylum-seekers, Returnees, Internally Displaced and Stateless Persons. 16 Jun. 2009, p. 11, online at , accessed 8 Nov. 2009. 4. One might object that these cases are not independent observations, since the programs were implemented in quick succession and featured the same interveners, who could, therefore, learn from past mistakes. However, many of my informants observed that the post-war administrations learned little, if anything, from their earlier experiences in the Balkans - making each intervention effectively sui generis. 5. I make a distinction here between 'ethnic reintegration' and 'reintegration' as defined by international law. The UN Guiding Principles of Internally Displaced Persons defines reintegration as (1) the return of displaced persons to their original homes and (2) the resettlement of displaced persons in another part of the country (Section 5, Principles 28-29). This paper uses ethnic reintegration to refer to the return of ethnic minorities pushed out during war as well as to the (re)incorporation of minorities into community structures, including administration, local assemblies, education, and law enforcement. Since minority returns is the necessary first step in the process of reintegration - without which minority reincorporation into community life is not possible - I focus primarily on what explains the variable (and relatively low) success of return programs in these cases. I am grateful to Rhodri Williams for raising this point. 6. This paper does not address ethical or humanitarian debates over refugee return. For more details, see Barbara E. Harrell-Bond, 'Repatriation: Under what Conditions is it the Most Desirable Solution for Refugees? An Agenda for Research', African Studies Review (1989) pp.41-69; Michael Barnett, 'Humanitarianism with a Sovereign Face: UNHCR in the Global Undertow', International Migration BARRIERS TO REINTEGRATION AFTER ETHNIC CIVIL WARS 391 Review 35/1 (2001) pp.244-77; BillFrelick, 'Preventing Refugee Flows: Protection or Peril?', World Refugee Survey (1993) pp.5-13; Daniel Warner, 'Voluntary Repatriation and the Meaning of Return to Home: a Critique of Liberal Mathematics', Journal of Refugee Studies 7/2-3 (1994) pp.160-74. 7. Donald Horowitz, Constitutional South Africa? Democratic Engineering in a Divided Society (Berkeley, CA: University of California Press 1991); Donald Horowitz, Ethnic Groups in Conflict (Berkeley, CA: University of California Press 1985). 8. Barry N. Stein and Frederick C. Cuny, 'Repatriation During Conflict: Protection and Post-Return Assistance', Development in Practice 4/3 (1994) p.180. 9. A survey of ethnic attitudes in Vukovar (inhabited by Serbs and Croats) and Prijedor (inhabited by Bosniaks and Serbs) found that the strongest predictor of conciliatory attitudes was not wartime experiences (which had no effect), but rather the degree of positive contact the individual had with members of the opposing group Miklos Biro and Petar Milin, 'Traumatic Experience and the Process of Reconciliation', Psihologija 38/2 (2005) pp.133-48. 10. C. W. McGlynn, The Impact of Post Primary Integrated Education in Northern Ireland on Past Pupils: A Study, PhD Thesis, University of Ulster (2001). 11. Miles Hewstone, Ed Cairns, Alberto Voci, Juergen Hamberger, and Ulrike Niens, 'Intergroup Contact, Forgiveness, and Experience of "The Troubles" in Northern Ireland', Journal of Social Issues 62/1 (Mar. 2006) pp.99-120. 12. Oren Yiftachel and Naomi Carmon, 'Socio-Spatial Mix and Inter-ethnic Attitudes: Jewish Newcomers and Arab-Jewish Issues in the Galilee', European Planning Studies 5/2 (1997) pp.215-37. 13. J. L. Gibson, 'Does Truth Lead to Reconciliation? Testing the Causal Assumptions of the South African Truth and Reconciliation Process', American Journal of Political Science 48 (2004) pp.201-17; G. Finchelescu, C. Tredoux, L. Muianga, J. Mynhardt, and J. Pillay, Testing Contact Theory in South Africa: A Study of Four Universities, Paper presented at 'Contact and Intergroup Relations: 50 Years On' conference, Ithala Game Lodge, South Africa, 6-9 Jul. 2006. 14. Reconciliation is defined as 'the process of developing a mutual conciliatory accommodation between antagonistic or formerly antagonistic persons or groups' Louis Kriesberg, 'Coexistence and the Reconciliation of Communal Conflicts' in Eugene Weiner (ed.) The Handbook of Interethnic Coexistence (New York: Continuum Publishing 1998) pp.182-98. 15. Indeed, multi-ethnic Bosnian cities such as Sarajevo and Tuzla held peace rallies to actively oppose the march toward war. 16. Yehuda Amir, 'Contact Hypothesis in Ethnic Relations' in Eugene Weiner (ed.) The Handbook of Interethnic Coexistence (New York: Continuum Publishing 1998) p.174. 17. Ashutosh Varshney, Ethnic Conflict and Civic Life: Hindus and Muslims in India (New Haven, CT: Yale University Press 2002). 18. Francesco Caselli and Wilbur John Coleman II, 'On the Theory of Ethnic Conflict', CEPR Discussion Paper No. 5622 (Apr. 2006) p.30. 19. Elizabeth G. Ferris, 'After the Wars are Over: US Policy in Reconstruction' in Aristide R. Zolberg and Peter Benda (eds) Global Migrants, Global Refugees: Problems and Solutions (New York and Oxford: Berghahn Books 2001) pp.322-49; John R. Rogge, 'Repatriation of Refugees: A Not So Simple "Optimum" Solution' in Tim Allen and Hubert Morsink (eds) When Refugees go Home: African Experiences (London: United Nations Research Institute for Social Development 1994) pp. 14-49. 20. Rhodri C. Williams, 'Post-Conflict Property Restitution and Refugee Return in Bosnia and Herzegovina: Implications for International Standard-Setting and Practice', New York University Journal of International Law and Politics 37 (2004) pp.441-553. 21. Richard Black, 'Return and Reconstruction in Bosnia-Herzegovina: Missing Link, or Mistaken Priority?', SAIS Review 21/2 (2001) pp.177-99; Richard Black and Saskia Gent, 'Sustainable Return in the Balkans: Beyond Property Restitution and Policy', International Migration 44/3 (2006) pp. 16-38. 22. Some of the key academic works include Carter Johnson, 'Partitioning to Peace: Sovereignty, Demography, and Ethnic Civil Wars', International Security 32/4 (2008) pp.140-70; Chapman, Thomas and Philip G. Roeder, 'Partition as a Solution to Wars of Nationalism: The Importance of Institutions', American Political Science Review 101/4 (2007) pp.677-91. Alexander B. Downes, 'The Problem with Negotiated Settlements to Ethnic Civil Wars', Security Studies 13/4 (2004) pp.230-79; Chaim D. Kaufmann, 'Possible and Impossible Solutions to Ethnic Civil Wars', International Security (1996) pp. 136-75. John J. Mearsheimer and Stephen Van Evera, 'When Peace Means War', New 392 CIVIL WARS Republic 18 (1995) pp. 16-21. See also the Security Studies special issue on partition in Summer 2004. For critiques of partition, see Nicholas Sambanis, 'Partition as a Solution to Ethnic War: An Empirical Critique of the Theoretical Literature', World Politics 52/4 (2000) pp.437-83; Radha Kumar, 'The Troubled History of Partition', Foreign Affairs 76/1 (1997) pp.22-34; Robert Schaeffer, Warpaths: The Politics of Partition (New York: Hill and Wang 1990). 23. Kaufmann (note 22) p.137. 24. Stephen Van Evera, 'Hypotheses on Nationalism and War', International Security 18/4 (1994) pp.5-39; Barry R. Posen, 'The Security Dilemma and Ethnic Conflict', Survival 35/1 (1993) pp.27 -47. Under the classic security dilemma, one side arms itself to defend against external attacks, inducing others to arm in response. This leads the first side to arm still further, yielding an arms spiral that paradoxically renders all sides less secure. 25. Van Evera (note 22) p.20. 26. Kaufmann (note 22) p.161. 27. Chaim D. Kaufmann, 'When All Else Fails: Ethnic Population Transfers and Partitions in the Twentieth Century', International Security 23/2 (Autumn 1998) pp.120-56. 28. Downes (note 22) p.238. 29. Marita Eastmond and Joakim Ojendal, 'Revisiting a Repatriation Success. The Case of Cambodia' in The End of the Refugee Cycle: Refugee Repatriation and Reconstruction (New York and Oxford: Berghahn Books 1999) pp.38-55. 30. Tanya Basok, 'Repatriation of Nicaraguan Refugees from Honduras and Costa Rica', Journal of Refugee Studies 3 (1990) p.283. 31. Repatriation entails sending refugees back to their country of origin, whereas reintegration involves returning the displaced to their pre-war communities. 32. Walter Kalin. Specific Groups and Individuals, Mass Exoduses and Displaced Persons, Report by the Representative to the Secretary-General on the Human Rights of Internally Displaced Persons. Mission to Croatia (29 Dec. 2005) p.7, online at , accessed 9 Sept. 2009. 33. International Center for Transitional Justice, 'Croatia: Selected Developments in Transitional Justice', p.4. 34. IDMC, Croatia: Reforms Come Too Late for Remaining Serb IDPs, A Profile of the Internal Displacement Situation, 18 Apr. 2006, p.10. 35. RFE/RL Report, 'Refugees and the "Dayton" Map', 26 Jan. 1996. 36. Human Rights Watch World Report 2001 - Croatia, online at . 37. Minority Rights Group (MRG) Report. Croatia: Challenges for Sustainable Return of Ethnic Serb Refugees, Jul. 2005, online at , accessed 18 Nov. 2009. 38. UNHCR Global Report 2000, Croatia, online at , accessed 18 Nov. 2009. 39. The number of minority returnees reached a peak of 20,716 in 2000. From this point of view, the number of returnees dropped sharply (UNHCR, 30 Jun. 2009). 40. OSCE Human Dimension Implementation Meeting, Warsaw, Poland, 6-7 Oct. 2003, online at , accessed 18 Nov. 2009. 41. MRG Report (2005) p.l. 42. As quoted by Anna McTaggart, 'Croatian Serbs Look to EU to Regain Homes', Balkan Insight. 26 Oct. 2006. 43. Human Rights Watch, 'Croatia: A Decade of Disappointment: Continuing Obstacles to the Reintegration of Serb Returnees', 5 Sept. 2006. 44. Downes (note 22) p.263. 45. IDMC (note 34) pp.7, 9. 46. U.S. State Department 2004 Country Report on Croatia, February 28, 2005, online at . 47. Mikhael Barutciski, 'The Reinforcement of Non-Admission Policies and the Subversion of UNHCR: Displacement and Internal Assistance in Bosnia-Hercegovina (1992-94)', International Journal of Refugee Law 3/1 -2 (1996) pp.49-110, p.63. The ethnic breakdown in Bosnia according to the 1981 census was 20 per cent Croat, 32 per cent Serb, and 40 per cent Muslim. 48. The referendum was boycotted by Bosnia's Serbs. BARRIERS TO REINTEGRATION AFTER ETHNIC CIVIL WARS 393 49. These are estimates from the Demographic Unit of the UN International Criminal Tribunal for the former Yugoslavia (ICTY). 50. Daniela Heimerl, 'The Return of Refugees and Internally Displaced Persons: From Coercion to Sustainability?', International Peacekeeping 12/3 (2005) p.377. 51. The initial NATO force was called the Implementation Force (IFOR) and had a one-year mandate. At the end of that year, the force was scaled back substantially, renamed the Stabilization Force (SFOR), and given an indefinite mandate. In 2008, SFOR was replaced with an EU Force or EUFOR. 52. Dayton Agreement (Annex 7, chapter 1, art. 1). 53. Author interview, Halisa Skopljak, Legal Assistant, Human Rights Dept., OSCE Mission to Bosnia and Hercegovina, Sarajevo, Bosnia, 29 May 2006. 54. Paula M. Pickering, Peacebuilding in the Balkans: The View from the Ground Floor (Ithaca, NY: Cornell University Press 2007) pp.26, 34-36; Sumantra Bose, Bosnia after Dayton: Nationalist Partition and International Intervention (Oxford: Oxford University Press 2002). 55. Robert F. Baumann, George W. Gawrych, and Walter E. Kretchik, Armed Peacekeepers in Bosnia (Fort Leavenworth: Combat Studies Institute Press 2004) p.220. 56. Pickering (note 54) pp.42-43. 57. As quoted in Pickering (note 54) p.4. 58. Andrew Gumbel, 'Mutual Mistrust Mires Refugees in the Misery of Homelessness', The Independent, 6 Aug. 1995. 59. Heimerl (note 50) p.379. 60. Author interview with Halisa Skopljak, 29 May 2006. 61. Author interview with Massimo Moratti, Sarajevo, Bosnia, 31 May 2006. 62. Author interview with Una Bejtoviae Mehmedoviace, Vaša Prava Spokesperson, Sarajevo, Bosnia, 5 Jun. 2006. 63. Heimerl (note 50) pp.381-82. 64. Ibid, pp.382-83. 65. Author interview with Moratti, 31 May 2006. 66. The number of Serb returnees to the Bosniak-Croatian Federation peaked in the years following the implementation of PLIP and the Bonn Powers. Only 1,731 Serb refugees and IDPs returned to the Federation in 1996. This number rose to a peak of 39,187 in 2002. From this point of view, the rate began to fall off sharply. The trajectory of minority returns of Bosniaks and Croats to Republika Srpska was similar. UNHCR Representation in Bosnia and Hercegovina, Statistics Package, 30 Jun. 2008, online at , accessed 18 Nov. 2009. 67. UNHCR Representation in Bosnia and Hercegovina (note 66). 68. UN Information Agency (USIA) reports, 1995-99, and UN Development Program (UNDP) reports, 2001-05, as cited in Pickering (2007) pp. 141-42. 69. John O'Loughlin, 'Inter-ethnic Friendships in Postwar Bosnia-Herzegovina: Socio-demographic and Place Influences', Unpublished manuscript. Institute of Behavioral Science and Department of Geography (Boulder: University of Colorado 2005) p.29. 70. Ibid., pp.14-15. 71. Florián Bieber, Post-War Bosnia (Basingstoke: Palgrave 2006) pp.73-74, 82-83. 72. See SFOR website, online at , accessed 18 Nov. 2009. 73. Author interview with Ivan Barbalic, former OSCE staff member and President of the Alumni Association of the Center for Interdisciplinary Postgraduate Studies (ACIPS), University of Sarajevo, Bosnia, 6 Jun. 2006. 74. Independent International Commission on Kosovo, The Kosovo Report (Oxford: Oxford University Press 2000) p.2. 75. UNHCR. Kosovo Crisis Update, 4 Aug. 1999, online at , accessed 18 Nov. 2009. 76. Eileen Simpson, Kosovo/Kosova, unpublished MA Thesis, University of Cambridge (2003), as cited in Iain King and Whit Mason, Peace at Any Price: How the World Failed Kosovo (Ithaca, NY: Cornell University Press 2006) p.53. 77. UNHCR. Kosovo Crisis Update, 1 Sep. 1999, online at , accessed 18 Nov. 2009. 394 CIVIL WARS 78. Guido Ambroso, Balkans at a Crossroads: Progress and Challenges in Finding Durable Solutions for Refugees and Displaced Persons. UNHCR research paper 133, 8-9 Nov. 2006, online at , accessed 18 Nov. 2009. 79. Author interview with Ardian Arifaj, UNMIK National Political Affairs Officer, Priština, Kosovo, 19 Aug. 2006. 80. Briefing to the United Nations Security Council (SC/9623). 6097th Meeting, 23 Mar. 2009, online at , accessed 18 Nov. 2009. 81. International Crisis Group, 'Return to Uncertainty: Kosovo's Internally Displaced and the Return Process', ICG Europe Report 139 (13 Dec. 2002) p.18. 82. Author interview with Jeff Bieley, UNMIK Political Advisor, Priština, Kosovo, 18 Sep. 2006. 83. The Urban Institute. Kosovo Returns Assessment. Prepared for the United States Aid for International Development (USAID), 20, 28 Jun. 2003, online at , accessed 4 Apr. 2009. 84. Leopold Von Carlowitz, 'Crossing the Boundary from the International to the Domestic Legal Realm: UNMIK Lawmaking and Property Rights in Kosovo', Global Governance 10/3 (2004) pp.307-32. 85. Author interview with UNHCR official, Priština, Kosovo, 15 Sep. 2006. 86. Anneke R. Smit, 'Housing and Property Restitution and IDP Return in Kosovo', International Migration 44/3 (2006) pp.63-88. 87. Anonymous informant, 2005, cited in Smit (note 86) p.75. 88. Author interview with Kristin Griffith, MercyCorps Mission Director, Priština, Kosovo, 19 Sept. 2006. 89. Florian Bieber, 'Policing the Peace after Yugoslavia: Police Reform between External Imposition and Domestic Reform', GRIPS Policy Research Center Discussion Paper, online at , accessed Jan. 2010, p.7. 90. Thorsten Stodiek and Wolfgang Zellner, 'The Creation of Multi-ethnic Police Services in the Western Balkans: A Record of Mixed Success', Deutsche Stiftung Friedensforschung (2007) p. 23, online at . 91. United States Department of State, Third Report on War Crimes in Yugoslavia-Supplemental United States Submission of Information to UN Security Council in Accordance with Paragraph 5 of Resolution 771 (1992) and Paragraph 1 of Resolution 780 (6 Nov. 1992), online at , accessed 18 Nov. 2009. 92. ICG, 'Brčko: What Bosnia Could Be', ICG Bosnia Project - Report 31 (10 Feb. 1998) p.9. 93. Ibid., p.40. 94. UNHCR, 30 Jun. 2008. 95. Florian Bieber, 'Local Institutional Engineering: A Tale of Two Cities, Mostar and Brčko', International Peacekeeping 12/3 (2005) p.429. 96. Vesna Peric Zimonijic, Brčko: An Example for the Rest of Bosnia. IPS (22 Dec. 2005), online at , accessed 18 Nov. 2009. 97. Janine Natalya Clark, 'Bosnia's Success Story? Brčko District and the "View from Below'", International Peacekeeping 17/1 (Feb. 2010) pp.67-79. 98. Roberto Belloni, 'Rebuilding at the Local Level: Refugee Return to Prijedor', International Peacekeeping 12/3 (2005) pp.434-47. 99. Author interview with Moratti, 31 May 2006. 100. Belloni (note 98) pp.443-44. 101. Ibid, p.445. 102. Ashraf al-Khalidi and Victor Tanner, 'Iraq Bleeds: Millions Displaced by Conflict, Persecution and Violence', Refugees Magazine 146, Apr. 2007. 103. Andrew Harper, Iraq's Refugees: Ignored and Unwanted 90/869, Mar. 2008, online at , accessed 18 Nov. 2009; UNHCR, Iraq Operation: At a Glance - Jan. 2009, online at , accessed 18 Nov. 2009. 104. UNHCR Country Operations Profile - Iraq, 2010, online at . 105. Mean Salehyan, 'No Shelter Here: Rebel Sanctuaries and International Conflict', The Journal of Politics 70/1 (2008) pp.54-66; Mean Salehyan and Kristián Skrede Gleditsch, 'Refugees and the Spread of Civil War', International Organization 60/2 (2006) pp.335-66. Copyright of Civil Wars is the property of Routledge and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.