University of Birmingham Institutional Outcomes of Territorial Contestation

Since1989, the countries of Central and Eastern Europe (CEE) have experienced major institutional transformations. As part of that process, territorial contestations between states and ethnic minorities engendered three outcomes: negotiated territorial self-government (TSG) arrangements; the denial of such arrangements; and the emergence of de-facto states.Through a qualitative comparative analysis of twenty-four minorityTSG claims in seventeen post-communist CEE states, we find that: (i) TSG arrangements emerged as externally facilitated instruments for managing or preventing violent conflict inpredominantly low-capacity, only partially democratic states; (ii) peacefully pursued TSG claims were most likely to be denied in high-capacity consolidated democracies; and (iii) de-facto states emerged where patron-states intervened in violent conflicts in low-capacity states. These findings defy widely held expectations about the influence of Europeanization, coupled with democratic consolidation, on the accommodation of minority claims; and they offer new insights into the significance of external intervention for the institutional outcomes of ethnic minorityTSG claims.

Territorial self-government (TSG) claims by ethnic minority groups present significant challenges to states around the world. Conflicts over such claims are likely to be violent (Fearon and Laitin 1999;Weidmann et al. 2010) and protracted (Walter 2003). The failure to accommodate ethnic minority populations through broadly acceptable and sustainable institutional arrangements is thus a potentially costly miscalculation for governments to make. Although a rich body of empirical scholarship has emerged about minority territorial claims, systematic comparative research into the conditions under which such institutional arrangements result from ethnic minority TSG claims remains scarce. Much of the literature focuses on the initial and final phases of territorial contestation. This literature is primarily concerned with two issues: why and how minority groups make territorial claims in the first place; and whether, once established, specific territorial arrangements-such as ethno-federation or federacy-induce further conflict or help to resolve it. The literature on the causes of territorial contestation covers a range of explanations, including feasibility (Buhaug et al. 2009;Collier et al. 2009;Weidmann 2009), economic opportunity (Collier and Hoeffler 2005;Fearon 2005), structural factors (Benedikter 2007;Lapidoth 1996), and economic, cultural, security, or political grievances (Gurr 1993;Horowitz 1985;Siroky and Cuffe 2015;Theuerkauf 2010).
Situated between the initiation of a territorial claim and its ultimate effect on stability and conflict, there are two important questions that have received surprisingly little attention: (i) What makes governments accommodate or deny minority self-government claims? (ii) Under what conditions do conflicts over such claims result in broadly acceptable institutional outcomes? Among the few comparative studies addressing the first question, Svensson and Lindgren (2011) focus on the outcomes of separatist claims. Shaykhutdinov (2010) includes a more comprehensive range of territorial claims, asking whether violent or nonviolent strategies are more likely to succeed. Sorens (2009) focuses on partisan political calculations as sources of government concessions in post-1970 Western European democracies. In an analysis focused on parliamentary representation, Alonso and Ruiz-Rufino (2007) show that this mitigation mechanism has no effect on secessionist conflicts. On the second question, Coakley (1990) provides a comparative account of six newly-independent European states after World War I.
There is a pressing need to advance the general understanding about when and how legitimate and sustainable institutional outcomes are likely to emerge from state-minority contestations over territorial claims-not only in situations involving violence or secessionism, but also in peaceful and non-secessionist settings. Our aim is to contribute to this understanding by developing an argument about the conditions under which states accommodate or deny minorities' claims for TSG, in the context of significant institutional opportunities for governments and minorities to renegotiate such claims. Thus, instead of asking why claims for minority TSG arrangements emerge (Cunningham 2013;Jenne et al. 2007), what strategies minority groups use (Shaykhutdinov 2010), or how TSG impacts stability, conflict and violence (Brancati 2006;Iff 2013;Sambanis and Milanovic 2014), we ask: under what conditions are TSG claims accommodated or denied when opportunities arise for the renegotiation of state-minority relations?
Faced with intra-state territorial claims, ruling elites can either grant some form of TSG or deny the accommodation of such claims. Previous research indicates that, from a global perspective, peacefully pursued claims (both secessionist and non-secessionist territorial claims) have a higher chance of achieving new institutional outcomes (Shaykhutdinov 2010). Comparative case studies also indicate that, when a minority claimant group cannot negotiate an acceptable institutional arrangement, and the minority remains part of the existing political system, the legitimacy of that system becomes weaker in the eyes of the minority population (Székely 2014). In other cases of nonaccommodation, minorities exit from the political system and unilaterally declare independence, resulting in either a ''de-facto state'' (Caspersen and Stansfield 2010) or a more broadly recognized new state (Wolff 2013a). As these differences in institutional outcomes have major consequences for long-term peace and democratic development (Chapman and Roeder 2007), it is important to understand the different pathways through which they emerge. Consequently, our analysis of state-minority territorial contestations in Central and Eastern Europe (CEE) between 1989 and 2012 seeks to provide critical insights into the conditions under which states are likely to grant or deny minority TSG claims.
We present our argument in four parts. First, we describe our general approach and case selection. Second, we explain our categories for institutional outcomes of state-minority territorial contestation and define the concept of TSG. Third, we map the variation in institutional outcomes in post-communist Europe from 1989 to 2012. Fourth, we evaluate the impact of domestic and external conditions commonly considered important in explaining the evolution of territorial contestation, and the way they combine to produce multiple pathways toward institutional outcomes. Our principal findings derived from the CEE institutional experience are as follows. Consolidated EU Member States do not grant TSG to minorities as part of the democratic claim-making process. Instead, TSG emerges as an externally facilitated arrangement in violence-affected weak states on the European periphery. In such states, status change and patron-state intervention also leads to the emergence of de-facto states. We conclude with lessons drawn from experiences in post-communist Europe, which highlight the significance of direct international involvement in the management of state-minority territorial contestations. 1

Approach and Case Selection
To contribute empirically and analytically to the question of why particular and distinct outcomes emerge from territorial contestations, we focus on a typological subset of such cases. The temporal and spatial setting of post-communist Europe is particularly well-suited for exploring the conditions under which institutionalized TSG arrangements are more likely to emerge from state-minority contestations during periods of major institutional transformation.
Our understanding of ''minority'' is similar to the definition most broadly cited in international law, formulated by Francesco Capotorti: ''a group numerically inferior to the rest of the population of a state, in a non-dominant position, whose members -being nationals of the State -possess ethnic, religious or linguistic characteristics differing from those of the rest of the population and show, if only implicitly, a sense of solidarity, directed toward preserving their culture, traditions, Institutional Outcomes of Territorial Contestation religion or language'' (Capotorti 1979: pararaph 568). By ''state'' we mean the state's central government, controlled in large measure by a demographically and electorally dominant group.
The dramatic institutional transformations of the post-1989 era created unprecedented opportunities for the institutional renegotiation of state-society relations generally in this region, and a significant number of minority groups made TSG claims that resulted in state-minority contestations (Csergo 00 2007; Jenne et al. 2007;Koinova 2010;Vermeersch 2003). The period between 1989 and 2012 in CEE brought the collapse and reconstruction of states, as well as the transnational project of ''Europeanization''-encompassing EU enlargement and its alternative, the European Neighbourhood Policy, both of which encouraged minority accommodation. These processes fundamentally reconfigured the domestic and international incentives and resources available for minorities to present collective claims, and for states to respond to them.
Evidence from around the world indicates that minority TSG claims are often accommodated during major institutional transformations. By way of illustration, in Italy, five regions with special autonomy statutes were created after World War II (Bull 1999), while in Spain autonomous communities were established after the end of the Franco dictatorship (Börzel 2000;Sala 2013 (Lecours 2013), and post-Somoza Nicaragua (Brunegger 2007) can similarly serve as examples of the accommodation of minority TSG claims in times of transition. Additionally, there are examples of democratically negotiated TSG arrangements in Denmark (Greenland), the UK (devolution in Scotland and Wales), and Belgium (the federalization process since the 1980s). 2 Consequently, we could reasonably expect the governments of democratizing CEE countries with a European policy outlook at a time of transformation to engage in a negotiated accommodation of TSG claims. Yet we find no evidence for this. Quite the contrary: a survey of TSG arrangements in CEE indicates that the more democratic and closer to the EU governments are (geographically and in terms of policy orientation), the less likely they are to accommodate TSG claims. Examining this puzzle is thus interesting and relevant both analytically and empirically.
Our cases were selected according to the presence of the following scope conditions: (i) significant opportunity to reform state institutions during the post-1989 transformation of former socialist regimes; (ii) participation in the Europeanization process (EU accession, European Neighbourhood Policy/Eastern Partnership); and (iii) mobilization by at least one ethnic minority group in the state in pursuit of TSG during the period we cover. Thus, we do not include minority groups that make claims only to cultural and linguistic autonomy, or guaranteed representation in national parliaments. For example, in Poland, we include Opole Germans but not Silesians or Kashubians. With these criteria in mind, we generated a set of twenty-four claimant groups in seventeen countries in CEE, including Kosovo and excluding Russia. 3 There are several grounds for the exclusion of the Russian federation. First, while Russia was affected by the overall Europeanization process, it at no point sought to engage with any of the two types of Europeanization processes we identified, thus not satisfying our second scope condition. Second, from a methodological standpoint, with thirty-three TSGclaiming groups of which twenty-five have received some form of it, the Russian federation would severely skew our analysis toward one country in which minority TSG as a meaningful practice of accommodating ethnic diversity has been in doubt since the early 2000s (Küpper 2013). Territorial contestation in the cases included resulted in some form of TSG for eight minorities in seven countries; no TSG is available for the remainder of eleven groups in eleven countries; and five groups in four countries have established de-facto states.
To assess the influence of key domestic and international factors shaping the institutional outcomes of territorial contestation, we employed qualitative comparative analysis (QCA) (Schneider and Wagemann 2006). This methodology is appropriate for our sample of twenty-four cases for evaluating the impact of indicators that capture broad trends rather than discrete events (Rihoux and Ragin 2009). QCA is also particularly appropriate for the analysis of phenomena that involve ''equifinality,'' i.e., multiple distinct pathways leading to similar outcomes (Schneider and Wagemann 2006: 753). We constructed a dataset for minorities that have made TSG claims in CEE (including the South Caucasus) since the first part of the twentieth century, focusing on the post-1989 period, drawing on the Minorities at Risk Project (2009), Minority Rights Group International (2009), and the Ethnic Power Relations Dataset (Wimmer et al. 2009 (Marshall et al. 2015), the Fund for Peace Fragile States Index (Fund for Peace 2014). 4 We also relied on our own fieldwork-obtained qualitative information derived from previous and ongoing research projects on a large subset of the cases addressed here. 5 Conceptualizing Outcomes of Territorial Contestation: TSG, Its Absence, and De-Facto Statehood There are considerable conceptual and empirical disagreements in the literature about what constitutes institutionalized intra-state TSG, further complicated by a predominant focus on just two forms of TSG: autonomy (or federacy) and Institutional Outcomes of Territorial Contestation federation. 6 The difficulty in conceptualizing TSG has long been recognized in political science and international law (Potier 2001: 54). The one common feature most definitions of TSG share is that the self-governing entity exercises certain powers independently from other levels of government. Such arrangements can incorporate executive, legislative, and judicial powers to varying degrees, but they fall short of full sovereignty. Following Wolff and Weller (2005), TSG is the legally entrenched power of territorially delimited entities within the internationally recognized boundaries of existing states to exercise public policy functions independently of other sources of authority in this state, but subject to its overall legal order. Accommodating TSG claims entails more than statewide regionalization or decentralization, which might provide de-facto control over institutions for minorities that compose local or regional majorities. Rather, it means the deliberate use of state construction for the management of territorial contestation by recognizing ethnically defined ''homelands'' as legitimate legal personalities for the exercise of TSG.
Based on empirical observation, we identify three distinct TSG arrangements in contemporary CEE: federation, federacy, and local self-government (table 1). 7 Federation: extensive self-rule with institutionalized shared rule. This implies a constitutionally entrenched structure in which the entire territory of a state is divided into separate political units, all of which enjoy certain exclusive executive, legislative and judicial powers independent of the federal government, 8 such as in Bosnia and Herzegovina. Federacy arrangement: constitutionally entrenched extensive self-rule for specific entities. The main distinction between a federacy arrangement and a federation is that the former enjoys similar powers and constitutional protection as federal Local self-government: executive and administrative powers at the local level.
Guided by the application of the principle of subsidiarity, local self-government involves the exercise of executive and administrative powers by lower levels of government, including regional and municipal authorities. It does not include legislative competences but provides the power to pass by-laws. Recent examples of the application of this form of TSG as a mechanism of conflict resolution in divided societies include Macedonia 9 (under the 2001 Ohrid Agreement) and Kosovo (under the terms of its 2008 constitution and related ''Ahtisaari legislation,'' as well as the April 2013 agreement over the creation of a Union of Serbian Municipalities in Kosovo). 10

Outcomes of Territorial Contestation in Post-Communist Europe
In our study, the presence of TSG means that one of these institutional forms is available for substate ethnic communities to exercise a degree of self-government in ''their'' homeland, while state majorities retain overall control of the same territory that they consider part of ''their'' state. We code ''TSG availability'' where any form of TSG identified above is present. The underlying assumption is that all forms of TSG commit the state and minorities to recognize the validity of each other's claims and to respect the parameters of an institutional setup designed to accommodate them. The state accepts limitations on its authority to exercise public policy functions within part of its territory. Minorities accept the overall legal constitutional order of the state and its existing borders, and they use TSG neither as a stepping-stone toward unilateral secession nor as a way to limit within the selfgoverning territory the constitutional rights of all of the state's citizens. We define the absence of TSG as an outcome of contestation in which a claimant group seeks TSG, the state refuses to grant it, and the claimant group remains part of the political system. This category refers only to situations where a state has denied TSG claims by a significant minority political party or equivalent organization. The absence of TSG does not mean that the claimant group abandons its demand: claims may persist or subside.
A different outcome of territorial contestation is the establishment of de-facto states. These are entities in which a separate government has been established which is de facto sovereign but lacks international recognition, i.e., de jure sovereignty (Caspersen 2011;Kolstø 2006;Lynch 2007). De-facto states can emerge along distinct pathways with different starting points, e.g., the state denies a TSG claim or Institutional Outcomes of Territorial Contestation an existing TSG arrangement is abrogated, and in response the claimant group exits the political system due to its failure to gain or retain a desired degree of TSG.

Explaining the Outcomes of TSG Claims in Post-Communist CEE, 1989^2012
We derive two main logics of explanation from the existing comparative literature on minority territorial claims: one that links the outcome of state-minority territorial contestation to a broad set of state-level and minority-centric domestic factors; and another that links them to external intervention by international organizations or a patron-state. The first set of factors relate to the power differential between minorities and majorities in the state; while the second set relates to external pressures and counterpressures in the context of actual or threatened violence. We first discuss the significance, and analyze the impact, of individual determinants of TSG outcomes. We then use QCA to examine which conditions combine in what way to create distinct pathways toward TSG, the denial of TSG, and de-facto statehood.

Domestic Conditions
In the analysis that follows, we explore the significance of four domestic factors for the outcomes of TSG claims: (i) state capacity and regime type; (ii) minority status loss in the context of post-1989 transformation; (iii) minority demographic and socioeconomic resources; and (iv) violence in the territorial contestation process. Table 2 lists all TSG-claiming minorities and summarizes key features at the domestic level.

Regime type and state capacity
Regime type and state capacity (state strength) are commonly used as independent variables in both quantitative and qualitative studies of inter-group conflict. Our question is whether these general state characteristics help to explain the institutional outcomes of territorial contestation in the post-communist context. Thus, we need to determine whether and how regime type and state capacity shape governments' preparedness to accommodate minority TSG claims. Among scholars that address similar questions, Walter hypothesizes that ''democratic regimes are more likely to negotiate with challengers than non-democracies,'' because ''democratic governments are thought to face greater domestic constraints on the use of force, be more sensitive to the rights of individuals seeking selfdetermination, and have a greater range of possible compromise solutions to offer ethnic groups'' (Walter 2003: 144f). Walter's findings confirm that democratic regimes are more likely to accommodate minority TSG claims. In the context of our focus on post-1989 CEE, we should expect similar findings in line with the literature on Europeanization: as countries in this region become more democratic and ''European,'' they should adopt more accommodating responses to territorial contestation similar to established patterns across the western part of the continent (Keating et al. 2009;Kymlicka 2007;Wolff 2013b). Regarding state capacity, Walter (2003) uses defense expenditure and the number of military personnel as indicators for state strength, hypothesizing that ''[c]ountries with relatively small defense expenditures and relatively small armies were expected to be more likely to accommodate challengers than countries with large defense expenditures and large armies'' (Walter 2003: 144f.) Most studies of state-minority territorial contestation also rely on GDP per capita scores as a proxy for state strength (Cunningham 2013;Fearon 2004), as well as on a variety of state strength indices that aggregate socioeconomic and military variables.
In this study, we use Freedom House's three-point ordinal regime type to categorize cases by regime; and scores from the Failed States Index (Fund for Peace 2014), Brookings's Index of State Weakness (Rice and Stewart 2008), and GNI per capita to measure state strength. 11 Our results highlight three broad patterns. First, consolidated CEE democracies with high state capacity are unlikely to grant TSG to minorities. In fact, no state that satisfied the requirements for high state capacity and democratic government has accommodated minority TSG demands. This pattern explains eight of eleven cases in which TSG has been denied, and it reinforces our finding that EU membership coupled with democratic consolidation did not make TSG claims more negotiable (see below). Only the remaining three denials of TSG occurred in low-capacity, partial democracies.
Second, the fact that TSG granting occurs in low-capacity, partial democracies (in seven of eight cases in which TSG claims were accommodated) suggest that TSG became a coping mechanism in CEE, indicative of a state's weak bargaining power in relation to minority claimants. Serbia is the only relatively consolidated democracy that granted a form of TSG (Vojvodina) following Milošević's removal from power in 2000. However, Serbia remains a relatively low-capacity state, the regional self-government is not exclusively associated with the Hungarian minority, and implementation of the TSG regime remains incomplete. 12 Third, all five cases of de-facto statehood (Armenians in Azerbaijan, Abkhaz and South Ossetians in Georgia, Northern Serbs in Kosovo and Transnistrians in Moldova) emerged in low-capacity states, including three partial democracies and an authoritarian regime.

Status loss in the context of state transformation
Status loss has been identified as significant in inter-group conflict, but not specifically in relation to TSG availability (Koinova 2013;Petersen 2011). The Institutional Outcomes of Territorial Contestation collapse of communist regimes after 1989, in a large number of cases, resulted in status loss for ethnic groups (i.e., for eleven of twenty-four minority claimants). This process created new or post-imperial minorities (e.g., ethnic Russians and Russian-speakers in Estonia, Moldova, and Ukraine) and led to status reversals for previously dominant groups, such as Serbs in Bosnia-Herzegovina, Croatia, and Kosovo.
While research shows status loss to be a major factor in mobilization for violent conflict (Cederman et al. 2010;Gurr 1993;Siroky and Cuffe 2015), there is little systematic research on how it affects the outcomes of territorial contestation. On the one hand, we could expect greater accommodation of claims when status loss occurs in a democratizing context, where governments are ''more sensitive to the rights of individuals seeking self-determination'' (Walter 2003). Alternatively, existing scholarship has also demonstrated that support from powerful external actors makes the violent pursuit of TSG claims more likely-shifting the balance of power such that governments are more likely to concede demands (Nome 2013;Walter 2003). We could thus expect minorities backed by powerful kin-states or similar patrons to be in a stronger bargaining position vis-à-vis their governments and thus able to pursue their TSG demands more successfully. On the other hand, we could also assume that newly-independent states and states that have broken free from Soviet domination would pursue nationalizing policies vis-à-vis minority groups (Brubaker 1996), and thus be less likely to grant TSG.
While we do find some confirmation of these expectations, our analysis does not reveal easily generalizable patterns across CEE. States are disinclined to grant TSG in the absence of minority status loss, but status loss is neither sufficient nor necessary for TSG. Situations where minority TSG was associated with status loss (i.e., Russians in Crimea; Serbs in Bosnia and Herzegovina and Kosovo) also involved significant violence or a strong threat thereof, as well as significant external intervention forcing state elites to accommodate TSG demands to preserve the territorial integrity of a new state. The clearest pattern we found was that all five cases of de-facto statehood are associated with collective status loss, or evident threat thereof, after 1989.

Collective resources for mobilization
Scholarship on ethnic mobilization, and particularly mobilization for violence, indicates that differences in the collective resources available to ethnic communities play an important role in shaping the motivation for, and feasibility of, violent conflict, and that they do so differently for conflicts over territory and government. Buhaug et al. (2009) find that absolute and relative distance from the capital, proximity to international borders, and availability of lootable resources increase the likelihood of longer-lasting conflicts. The likely explanation is that favorable locations directly and indirectly increase rebel fighting capacity (and motivation). Similarly, Fearon (2004) points out that so-called sons-of-the-soil conflicts typically last longer than conflicts over government (coups or revolutions) because ''a stable regional autonomy deal is harder to construct when the political centre's stakes in the region are greater . . . [and] when the rebel force can extract more from a region during the course of a war'' (Fearon 2004: 297). Toft also makes the point that ethnic groups' settlement patterns determine the legitimacy and capability of a group's mobilization: ''where both capability and legitimacy are high -as they are for a group that is concentrated in a region, especially if that region is its homelandan ethnic group is likely to consider control over disputed territory an indivisible issue and demand independence'' (Toft 2001: 3). Thus, accommodation of TSG demands is only likely ''[i]f an ethnic group is willing to accept an outcome short of full independence, or if the state sees its territory as divisible'' and does not ''fear precedent-setting'' vis-à-vis other groups with potentially similar demands (Toft 2003). Walter makes a similar argument, noting that ''the economic, strategic, or psychological value of land does not fully explain government decisions to settle or fight when territory is at stake;'' rather, ''governments consistently refuse to negotiate not because land is especially valuable as many people have argued, but because they fear that a concession to one state or one separatist group will encourage other parties to seek their own share of a limited pie'' (Walter 2003: 138).
To test whether these general findings also hold for our sample of post-1989 cases of territorial contestation in CEE, we draw on several indicators to account for collective resources that can be mobilized in territorial contestation. We first look at demographic factors such as relative group size, territorial concentration, and the proportion of the minority living in its historical homeland. We also assess socioeconomic and political indicators of minority collective resources, such as relative economic status and economic discrimination using data from the Minorities at Risk Project (2009). The former measures barriers to minority economic activity and social mobility while the latter measures the type of organization available to the minority to mobilize. Our evaluation reveals no clear pattern regarding the impact of group size or territorial concentration on the success of TSG claims. Concerning socioeconomic status, it appears that economic discrimination is higher in the case of minorities without TSG than it is in the case of those that have achieved TSG or established de-facto statehood. Finally, using data about the level of Group Organization for Joint Political Action (Minorities at Risk 2009), we find no clear pattern between minorities' level of organization and the granting of TSG claims. The overwhelming number of minorities in CEE organized political parties and civil society organizations, and TSG granting did not correspond to more robust forms of organization. On the other hand, all cases of de-facto states involve minorities organized in both parties and violent militant organizations.

Inter-ethnic violence in the context of state-/regime transformation
Violence appears as a key factor in the literature on inter-group conflict, but its impact on the institutional outcomes of minority TSG claims is complex. On the one hand, significant violence in a contested region may force minority claims onto the agenda (Hartzell and Hoddie 2003) or trigger third-party pressure or intervention (Kuperman 2005). On the other hand, violence may prompt states to reject plans for granting TSG (Horowitz 2003), rescind existing TSG arrangements (Siroky and Cuffe 2015), or respond with disproportionate violence to suppress minority claims (Walter 2006).
In CEE, territorial contestations between current majorities and minorities involved significant instances of violence since 1989 in just over half of our cases (thirteen of twenty-four). Among eight cases of existing TSG, six involved violence. In the same timeframe, Ukraine and Serbia established federacy arrangements in the absence of violence (to accommodate demands by ethnic Russians and Russianspeakers in Crimea and ethnic Hungarians in Vojvodina, respectively). However, the threat of violence was important in Crimea; and violence was significant in other regions of Serbia, as well as to a certain level in Vojvodina proper (Jenne 2004). Thus, violence was part of the context of accommodation in both cases. Where TSG was denied, violence was also usually absent (ten of eleven cases); and only one violent conflict has not yielded TSG (Serbs in Croatia 13 ). Finally, all five cases of de-facto statehood involved violence.
Our analysis therefore suggests that post-1989 patterns of violent conflict have significantly shaped the granting of TSG arrangements in CEE, but violence was neither necessary nor sufficient for TSG granting. The occurrence of violence is strongly intertwined with external intervention, indicating the importance of analyzing the combination of these factors more closely.

The Power of External Intervention
To explore the role of external factors that the extant literature identifies as significant for either reducing and preventing conflict or exacerbating existing conflict by encouraging or discouraging particular institutional designs, 14 we first evaluate the influence of ''Europeanization'' in broad terms. Then, we analyze patterns of more direct intervention (military and diplomatic) by two types of external actors: (i) international organizations, such as the United Nations (UN), the European Union (EU), and the Organization for Security and Cooperation in Europe (OSCE); and (ii) patron-states involved in the contestation.

The influence of ''Europeanization'' on the emergence of TSG arrangements
The literature on the effects of Europeanization identifies several processes by which new members and candidate countries might be swayed into accommodating minority TSG claims-from the diffusion of European norms to ''harder'' forms of leverage through accession conditionality. 15 The EU is expected to ''de-securitize'' state-minority and minority-patron state relations, cases of which abound in CEE. 16 Although minority-patron state relationships can generate significant security concerns, the European framework is expected to decrease the relevance of conflict over state borders-given that relevant actors form part of a larger political entity that encourages increased regional mobility and a more accommodating approach to minority issues based on an acceptance of common norms and values (Cordell and Wolff 2005;Csergo 00 and Goldgeier 2004;Kelley 2004;Sasse et al. 2004). As an aspect of expectations about the influence of ''Europeanization,'' Western European models of institutionalized minority TSG can serve as ''good practice'' examples in CEE (Kymlicka 2007). The majority of Western European states have accommodated minority TSG claims in various forms, including federations in Belgium and Switzerland, devolution in the United Kingdom, and regional autonomy in Denmark and Finland (Kymlicka and Opalski 2001: 29-31). 17 Building on pre-1945 legacies, many of these arrangements emerged in the context of ''Europeanization'' which in its broadest sense implies the establishment of a democratic security community in post-Cold War Europe (Keating 2008). According to this logic, we should expect that issues of TSG become matters of peaceful democratic contestation within actual and aspiring EU member states, and that the existing pattern of accommodation of minority TSG claims in Western Europe would facilitate similar responses to territorial contestation in CEE (Galbreath and McEvoy 2012;Jutila 2009).
Our findings indicate, however, that CEE states remain reluctant to grant minorities TSG-in line with arguments that fear of secession, reputational effects on other minorities' claims, and the economic, strategic or symbolic values of contested territories for majorities override whatever effect the EU has on minority policies in new and aspiring Member States (Bochsler and Szöcsik 2013;Toft 2001;Walter 2003). Thus, Europeanization has had little impact on state majorities' inclination to allow TSG arrangements to emerge ''democratically'' in processes of post-communist institutional restructuring.
TSG arrangements were not granted in any state that is considered a consolidated democracy and is an EU member (Bulgaria, Croatia, Estonia, Lithuania, Poland, Romania, and Slovakia). TSG arrangements emerged only among less-consolidated democracies outside the EU. Of the six current applicant countries, four provide TSG to at least one claimant group (Bosnia and Herzegovina, Kosovo, Macedonia, and Serbia) and two do not (Albania, Montenegro). Kosovo, moreover, during part of the period under consideration (2008-2012) included a de-facto state in the Mitrovica region. Of the remaining states covered by the EU's Eastern Partnership, Georgia, Moldova, and Ukraine provided TSG arrangements on a selective basis from 1989 to 2012. Thus, the closer states are tied to the EU and the more consolidated their democratic regime is, the less likely they are to accommodate TSG claims. This runs counter to widely-held expectations that (i) the EU's influence will make TSG issues matters of peaceful democratic contestation; and (ii) Europeanization engenders similar institutional outcomes across the continent. The question arises of what other forms of international intervention might account for cross-regional differences in the institutional outcomes of TSG claims.

International military and non-military intervention in conflict settings
Extant literature establishes significant links between intra-state inter-group conflict and third-party intervention. Intervention can affect conflict onset, dynamics and resolution, and empirical research suggests that third-party action can either pacify or exacerbate existing conflict (Carment, James, and Taydas 2006;Jenne et al. 2007;. Post-communist CEE provides evidence for both patterns (table 2). To understand these dynamics, we explored the impact of two types of direct external intervention. We distinguish between (i) international intervention-by which we mean the intervention of international and regional organizations, as well as major international powers that are not parties to the territorial contestation; and (ii) patron-state intervention-which in CEE usually means intervention by a kin-state in support of ethnic kin in a neighboring state. Empirical research indicates that the Hungarian government's patron-state activism has had both moderating and exacerbating effects on local Hungarian minority claims in Serbia, Slovakia and Romania. Hungarian minority claims were pursued peacefully in all cases (Csergo 00 and Goldgeier 2004; Jenne et al. 2007;Waterbury 2010). Russia has played an ambiguous and changing role as a patron-state. In the 1990s, the Russian government and the OSCE (through its High Commissioner on National Minorities) facilitated a negotiated agreement between Crimean separatists and the Ukrainian government. In 2004, when Adjaria (Georgia) was on the brink of major violence after the re-imposition of Georgian authority on that region, Russia's intervention helped to prevent violent escalation. In relation to de-facto states created within post-Soviet states-i.e., Abkhazia and South Ossetia (Georgia), Nagorno-Karabakh (Azerbaijan), and Transnistria (Moldova)-in contrast, Russia supported the separatists, but it also mediated cease-fire negotiations and participated in international mediation efforts. Serbian patron-state activism has been most often unambiguously conflict-generating from 1989 to 2012 (Caspersen 2008). Germany, in contrast, has consistently been a moderating force (Cordell and Wolff 2005).
We found a close link between the emergence of TSG and international intervention (both military and non-military) in violent contexts, while the lack of intervention was consistently associated with the denial of TSG in nonviolent conflicts. International intervention in territorial contestations typically occurred in response to violence. Nonmilitary involvement facilitated TSG more than military intervention. International non-military intervention occurred in seven of eight cases resulting in TSG (Croats and Serbs in BiH, Adjars in Georgia, Central and Southern Kosovo Serbs, Albanians in Macedonia, Gagauz in Moldova and Crimean Russians). International military intervention occurred in four cases where TSG was granted (Croats and Serbs in Bosnia Herzegovina, as well as Albanians in Macedonia and Serbs in Central and Southern Kosovo). However, in three of those cases (Croats and Serbs in Bosnia-Herzegovina, as well as Albanians in Macedonia) international non-military intervention was also present.
Patron-state intervention had a mixed effect. Military intervention by patronstates occurred in a minority of cases that yielded TSG (Croats and Serbs in Bosnia-Herzegovina and Albanians in Macedonia), while non-military intervention by a patron-state occurred in half of the cases where TSG exists (Adjars in Georgia, Central and Southern Kosovo Serbs, Hungarians in Vojvodina, and Russians in Crimea); but also in five of eleven cases where TSG was denied (Turks in Bulgaria, Russians in Estonia, Opole Germans in Poland, and Hungarians in Romania and Slovakia). The impact of patron-state military intervention is stronger and unambiguous on de-facto statehood-as evidenced by all five cases in the region (Armenians in Nagorno-Karabakh; Abkhaz and South Ossetians in Georgia, Serbs in Northern Kosovo, and Transnistria in Moldova).
Overall, our findings regarding the impact of external intervention on the institutional outcomes of TSG claims are three-fold. First, external intervention occurred in all cases where minority TSG demands were granted. In the majority of such cases, non-military international intervention occurred in the context of violent conflicts. Hungarians in Serbia and Russians in Crimea represent exceptions, where external involvement facilitated TSG in the absence of violence. In Crimea, the involvement of external actors (e.g., OSCE and the Russian patron-state) happened in the context of a significant threat of violence, which it prevented from materializing. For Hungarians in Serbia, EU member Hungary lobbied actively in international fora; and the Serbian government had strong incentives to reinstate Vojvodina's regional autonomy as part of an effort to demonstrate democratic credentials in pursuit of EU membership after more than a decade of wars.
Second, where minority TSG claims were denied, yet claimant groups remained in the political system, international intervention was absent or weak, and patronstate activism was also either absent or occurred as non-military support. TSG was denied after violent conflict in only one case, Serbs in Croatia, which lacked significant external support.
Third, where patron-states intervened militarily, in most cases the outcome was de-facto statehood, unless there was an international military counter-intervention to prevent that outcome, as in Bosnia and Herzegovina, Croatia and Kosovo. Patron-state military intervention thus appears as a necessary, albeit not sufficient, condition of de-facto statehood.

Possible Pathways toward TSG, Its Denial and De-Facto StatehoodçQCA Analysis
Our analysis of individual conditions of TSG outcomes revealed sharp differences between consolidated democracies in CEE that are part of the EU (which all deny TSG to their minorities) and other, low-capacity states in the region, where TSG is available dependent on third-party responses to violent conflict. Individual conditions, however, combine in different ways to make TSG available, unavailable, or to produce situations of de-facto statehood. Our analysis suggests that two main pathways have led to the granting of TSG, while two others have led to the denial of TSG. A single pathway has led to de-facto statehood. To identify and summarize combinations of factors that affect the outcome of TSG claims, we performed a crisp set QCA of our dataset of twenty-four cases, as developed by Ragin (2000), using fs/QCA3.0 (Ragin and Davey 2014). QCA is a highly relevant technique for the analysis of institutional change or continuity, as it captures broad factors that operate over time. It is also well suited to the analysis of conditions that impact territorial outcomes only in combination with other conditions, such as intra-state violence and third-party intervention.
Instead of estimating the net causal effect of discrete variables, QCA specifies the combinational logic of necessary and sufficient conditions needed to obtain a given outcome. The analysis aggregates individual cases into a manageable number of typical ''scenarios.'' These scenarios, in turn, encapsulate the causal mechanisms that link certain sets of conditions to alternative institutional outcomes (e.g., TSG availability, its denial, and in some cases the establishment of a de-facto state). 18 Table 3 highlights the minimal conditions for our three outcomes and reveals the typical combination and sequencing of conditions leading to TSG, its denial, and de-facto statehood.
Two sets of conditions lead to the denial of minority TSG claims. In the most common pattern, high-capacity democratic EU Member States deny TSG claims. In the absence of inter-group violence, no international intervention occurs to generate a negotiated TSG outcome. Patron-state lobbying on behalf of the minority is the typical form of external involvement, such as in the case of Poles in Lithuania or Hungarians in Romania. In the second pattern, low-capacity non-EU  states also deny TSG in the absence of inter-group violence, which receive neither international nor patron-state support, as in the case of Crimean Tatars or Sandžak Muslims. Two sets of conditions lead to the granting of TSG. A first pathway involves international (military or non-military) intervention in response to actual intergroup violence in a weak non-EU state. TSG subsequently emerges as an internationally facilitated, negotiated outcome. This pattern was typical of conflicts in the former Yugoslavia. A second, less common, pattern involves the granting of TSG by low-capacity non-EU states, in the absence of violence and with nonmilitary intervention by a patron-state. This pattern covers the cases of Crimea in the mid-1990s and of Hungarians in Vojvodina.
Finally, a single set of conditions leads to de-facto statehood, suggesting that these conditions are collectively sufficient to bring about this outcome. The set includes military patron-state intervention in low-capacity states where the minority experienced status loss or threat thereof in the transitional period (1989)(1990)(1991)(1992), and inter-group violence ensued in that context. While military patron-state intervention between 1989 and 2012 occurred only in violent cases, it is important to note that violence alone is a necessary but not sufficient condition for de-facto statehood.

Conclusion
Our study provides important insights into the institutional outcomes of minority TSG claims, based on a comparative analysis of cases in post-communist CEE from 1989 to 2012. These cases enable us to identify conditions under which TSG is granted or denied in the context of territorial contestations during institutional transformations that allow for comprehensive renegotiation of state-society relations; and what outcomes can be expected when TSG claims are denied. Our findings have important empirical and theoretical implications.
Empirically, our most significant finding is that the overwhelming majority of TSG arrangements emerging in post-Cold War CEE resulted from externally facilitated conflict management (i) in the context of major state crises involving violence or a significant threat thereof, (ii) in low-capacity states, (iii) most of which were non-democracies or partial democracies. This finding adds to a growing literature on the benefits of TSG as a conflict-mitigating mechanism (Bakke 2015;Cederman et al. 2015). It also provides a corrective to a body of scholarship that is critical of TSG arrangements to the extent of alleging it has a conflict-inducing rather than conflict-mitigating effect (Chapman and Roeder 2007;Cornell 2002).
Theoretically, our findings suggest the need for reevaluating the utility of Realpolitik, balance-of-power considerations in explaining the outcomes of intra-Institutional Outcomes of Territorial Contestation state territorial contestations. The fact that high-capacity, consolidated democracies are able to deny minority TSG claims and maintain their sovereignty and territorial integrity; while low-capacity, non-consolidated democracies are generally unable to do so (because they also face external pressure or even military intervention) points in this general direction. Put differently, external intervention appears to tip the balance in favor of negotiated TSG outcomes or de-facto statehood, while its absence seems to allow high-and low-capacity states alike to deny TSG claims without triggering violent responses from minority TSG claimants.
Such a plausible ''hunch'' generates new avenues for future research into the actors and mechanisms of an overarching Realpolitik framework that encapsulates the distinct pathways we have identified toward the three outcomes of intra-state territorial contestation in CEE. First, our findings highlight the need for more research about the conditions under which conflict mitigation by way of accommodating TSG claims is a feasible strategy for international actors to pursue in transitional and post-war contexts. Further research on directive mediation strategies (Bercovitch and Houston 2000) can explore which strategies might produce broadly acceptable and sustainable outcomes. For example, the EUmediated agreement between Serbia and Kosovo in April 2013 indicates that nonmilitary international involvement may result in negotiated minority TSG even in a case where a de-facto state has already been established. Kosovo represents perhaps the clearest indication of the possibility that Europeanization may lead to negotiated agreements, albeit in relation to relatively low-capacity (host and kin) states susceptible to the same (i.e., EU) external leverage.
Second, our finding that stronger, more consolidated democratic states in CEE that joined the European Union did not accommodate minority TSG claims requires further investigation. While minority interest in TSG has not disappeared, none of these countries has experienced violent contestation. More research is necessary to explore whether (and if so what) it is specifically about Europeanization and democracy that has prevented the violent escalation of conflicts despite the denial of TSG claims. We cannot rule out that EU membership itself is a form of ''external intervention'' that tips the balance in favor of states when it comes to denying minority TSG claims; while only offering relatively minimal assurances to minorities, through guarantees of basic civil and political rights and liberties within a democratic institutional framework.
More research is necessary also into the way external intervention shapes domestic ''balances of power.'' The Russian annexation of Crimea in March 2014, and the ongoing conflict in eastern Ukraine, raise the possibility that initially unambiguously non-secessionist claims or TSG arrangements can change to demands for secession or irredentism in the context of decisive external intervention. This scenario might become relevant for Bosnia and Herzegovina (Republika Srpska, Croat cantons in the Federation), as well as for Macedonia and Montenegro in relation to their Albanian communities; and it raises the possibility of a Kosovo and/or Albania-driven irredentism. Here, as elsewhere in CEE and possibly in Western Europe (e.g., Catalonia, Scotland), much will depend on the viability of European integration and the degree to which the European Union can sustain its current role as a viable regional/international actor capable of projecting power and influence within and beyond the borders of its Member States.
The experiences in post-Cold War Europe provide important lessons for multiethnic societies in other regions where opportunities arise for renegotiating state-society relations. Clearly, peacefully negotiated institutional outcomes in states that include politically resourceful territorialized minorities that seek TSG arrangements also require external actors that are ready to facilitate peaceful outcomes. Domestic political (state and minority) elites, in turn, share a great responsibility in negotiating broadly acceptable, legitimate and sustainable institutional arrangements that enable democratic consolidation and state stability. If they are unable or unwilling to do so, our findings tentatively suggest a potential return to the Realpolitik calculations of ethnopolitics that have bedeviled Europe in the past at tremendous human cost.

Note
We are grateful to the three anonymous reviewers of this article, as well as to numerous colleagues who have provided invaluable feedback on earlier drafts of this article, including Henry  3. Our categorization of outcomes reflects conditions at the end of 2012. We thus count Crimean Russians as a case of federacy, Crimean Tatars as a case of no TSG, and we include Kosovo as a state. The status of Crimea changed significantly after its Russian annexation in March 2014, and institutional outcomes in Eastern Ukraine are still developing at the time of writing. Kosovo was under international administration between 1999 and 2008, then declared independence from Serbia. Although Serbia contests Kosovo's independence, its recognition by over 100 UN Member States, in our view, justifies treating Kosovo as a state. The constitution of Kosovo makes specific provisions for local governance units with significant non-Albanian populations. During the period on which we focus , ethnic Serb organizations in the Mitrovica region of northern Kosovo demanded self-government and established a de-facto state. However, Serb minorities in other regions (Central and Southern Kosovo) remained under the Kosovo constitution. 4. Our complete dataset is available in online Appendix 1. 5. This fieldwork consisted of semistructured interviews with state officials and minority political elites, as well as participant observation, in the following subregions: Central and Eastern European states that joined the EU before 2012 (Estonia, Lithuania, Poland, Hungary, Romania, Slovakia); states in the Western Balkans that were EU candidate and potential candidate states before 2012 (Croatia, Serbia, Bosnia-Herzegovina); post-Soviet states with an EU Association Agreement (Moldova, Georgia) and without one (Azerbaijan). 6. For an overview of this debate, see Wolff (2001). On the distinction between the ideologies of autonomism and federalism, see Lluch (2012). 7. Each of these arrangements can be applied with their territorial boundaries cutting either across or around the settlement areas of ethnic or national minorities. Our analysis focuses on arrangements with the latter kind of territorial boundaries. 8. There are common exceptions to this entire-territory rule. Capital cities, for example, unless they are federal entities of themselves, often have special status (Washington, D.C., vs. the German capital Berlin, which has Bundesland status). 9. We refer to the Former Yugoslav Republic of Macedonia (FYRoM) by its constitutional name. 10. According to our definition, de-facto power gained by an ethnic minority in local councils through the regular electoral process in an otherwise significantly centralized state does not qualify as TSG. Examples are: Russophone domination in city councils in Riga and Tallinn; Hungarian domination of local councils in the Szekler region (Bochsler and Szöcsik 2013). 11. Freedom House's three-point scale of regime type is widely used in comparative research. We distinguish between low-and high-state capacity based on per capita GDP and scores from the Fund for Peace Fragile States Index (Fund for Peace 2014). Seminal studies on ethnonational contestation using per capita GDP as a proxy for state capacity include Fearon and Laitin (2003) and Hoeffler (1998, 2005). 12. Personal communication from Petar Teofilovic, 14 April 2016. 13. Serbs in Eastern Slavonia (Croatia) briefly attained de-facto statehood during the Yugoslav wars; then had a brief self-governing status under the United Nations