Competitive statebuilding from the demand-side: counter-state services and civilian choice in Kosovo, 1989-1998

ABSTRACT

Counter-state actors often supply services to foster civilian support. Yet little work explores the civilian ‘demand-side’ of this interaction. This paper examines how civilians navigated between overlapping state and counter-state services during a case of non-violent competitive statebuilding. Between 1989 and 1998, a Kosovar-Albanian ‘parallel state’ provided education, healthcare, and justice as part of a strategy to secede from Serbia. It finds that two factors are key: the level of group solidarity individuals are subject to, and the unique characteristics of the services they are receiving. Increased group solidarity constrains how individuals decide between providers, yet the extent to which this impacts choice depends on the characteristics of different services. Education is collectively delivered and tied to nation-building; decisions depend on social norms. Healthcare is individual and immediate; decisions are rooted in trust. Justice varies between discreet civil cases where people can ‘shop around’, and criminal cases, which can comprise highly visible, collective events with significant social pressure. Understanding how solidarity and service characteristics intersect is key to understanding the demand-side of competitive statebuilding.

KEYWORDS:

• Governance
• conflict
• statebuilding
• public services
• Europe

Introduction

A fraught peace existed in Kosovo between 1989 and 1998. Before a violent cycle of guerrilla attacks and counterinsurgent operations led to NATO’s intervention, the province in southern Serbia had effectively split into two competing polities. Following the revocation of Kosovo’s self-governing status within Serbia, a Kosovar Albanian movement organised an underground ‘parallel state’ as an alternative to its formal Serbian counterpart. This ‘state’ organised elections and provided public services ranging from schools, clinics, and courts to sports teams and theatre groups in ‘parallel’ to the Serbian state – all funded by an underground system of taxation. Yet no rebel-held areas signified that a contest over sovereignty was underway. Before 1998 there were no shifting front lines or consolidated pockets of guerilla government; Serbian forces maintained uncontested territorial control, while the parallel state’s leaders were avowedly non-violent, limiting their abilities to compel local support. With two competing suppliers of governance, Kosovars faced a peculiar choice: from whom to be governed?

A growing literature documents the variety of ways that counter-state actors – non-state groups that seek to challenge the political or territorial authority of an existing state – build and manage relations with ordinary civilians. Many hardly govern their constituents at all. Others only provide rudimentary forms of order. Yet a significant minority engage in ‘competitive statebuilding’—the building of local support through the provision of services and institutions that aspire to be perceived as more legitimate and effective than those offered by the incumbent state.¹

The establishment of alternative structures of governance is a common thread across these otherwise heterogeneous actors. Political parties such as Hezbollah managed hospitals, schools, orphanages, rehabilitation centres for the handicapped, supermarkets, gas stations, construction companies, a radio station, and a television station in southern Lebanon.² Rebel groups like FARC provided financial services to Colombian farmers, built roads, and ran health centres.³ Underground movements like Egypt’s Muslim Brotherhood organised disaster response.⁴ Civil resistance committees during the first Palestinian Intifada established education, medical relief, and dispute resolution services.⁵ Unrecognised ‘de-facto’ states like Transnistria or Abkhazia often educate their children and maintain local economies to a higher degree than the recognised states of which they are still notionally a part.⁶

Service provision is costly and complex. Why do counter-state actors bother? Scholars have reached some consensus that fostering civilian support is key. Supplying services increases perceptions of legitimacy among local populations, reduces the amount of information shared with counterinsurgents, prevents defection, aids taxation, and supports recruitment – all factors in a group’s success.⁷ Apart from strategic considerations, secessionist groups must also hew to the norms of the nation-state.⁸ Public goods delivered to a defined population constitutes a central tenet of modern political orders: services show that rulers “care for ‘the people’”.⁹

Yet despite the centrality of civilian support, the actual recipients of counter-state services are often neglected. We know that services are a central interface between rulers and the ruled, vital components of the ‘infrastructural power’ states use to penetrate societies and make their populations ‘legible’.¹⁰ But just as services allow states to ‘see’ their citizens, they also allow citizens to ‘see’ right back at the state. Receiving services – attending schools, obtaining treatment, settling a dispute – allows people to encounter the otherwise abstract state in their everyday lives.¹¹ Exploring civilian interactions with counter-state service provision thus opens a rich field of enquiry. We have a good understanding of the ‘supply side’ of competitive statebuilding – the scope and variety of services provided by counter-state actors to foster civilian support. What we lack is an understanding of the ‘demand side’. How do civilians navigate between competing providers? Do some services foster more support than others? Why might people choose one provider over another?

The answer, I argue, rests on two factors: the level of group solidarity civilians are subject to and the unique characteristics of the services they are receiving. Increased group solidarity constrains how individuals decide between competing service-providers. However, the effect of solidarity varies due to the different characteristics of services that structure and influence how people relate to them. I argue that understanding how the factors of group solidarity and service characteristics intersect can help us understand how people navigate between competing authorities during periods of competitive statebuilding.

Kosovo during the ‘parallel state’ period of 1989–1998 offers a relevant case study with which to explore this process. However, two key distinctions stand out. First, competitive statebuilding in Kosovo was non-violent. Until the rise in guerrilla activity in late 1997, the province was free from violent conflict. This goes against a core assumption in the literature that competitive statebuilding only occurs ‘during the protracted violence and high levels of coercion produced by civil war’.¹² Whether violence is a necessary condition should be an empirical question, not a definitional attribute. The phenomena of competitive statebuilding does not solely occur during violent insurgency, yet studies of civil conflict have long overlooked non-violent action during political crises, resulting in a disproportionate focus on violent actors.¹³ ‘Rebellion’ is thus usually defined by indicators like annual battle deaths, excluding less coercive cases of competitive statebuilding that, while still fundamental contests over authority, are both more likely to achieve their objectives and are far better at attracting support from risk-averse civilians.¹⁴ There are good reasons to examine these alternatives when it comes to questions of agency. Fear of physical harm or death can incapacitate people and skew decision-making towards compliance,¹⁵ leading to impassiveness, withdrawal, and ‘submission’.¹⁶ If violence ‘can bend the people’s posture’,¹⁷ less coercion may broaden their scope of action.

Secondly, Kosovo’s parallel state lacked territory – a break from the assumption that competitive statebuilding requires territorial control.¹⁸ Yet empirical work has increasingly questioned this requirement. In Nepal, Maoist mobile courts were able to hear cases in territory beyond their direct control, offering people an alternative form of justice alongside the formal Nepali judiciary.¹⁹ In Afghanistan, Taliban courts and tax collectors often preceded the establishment of direct territorial control.²⁰ Similarly, in Kosovo people regularly encountered both state and counter-state services in close physical proximity, even the same neighbourhood. Again, such cases are well suited for exploring civilian agency. Not faced with a territorial monopoly by either state or counter-state actors allows civilians significant scope to navigate between competing providers.

This article proceeds in four sections. The first section reviews existing literature before presenting the article’s approach and outlining criteria for empirical exploration. The second, third, and fourth compare how individuals engaged with education, healthcare, and justice during the parallel state period. The fifth concludes with a discussion of the article’s implications.

Literature Review

Counter-state actors aim to replace the state, either outright or in part through secession. Yet their dependence on the local population varies. Some – what Mancur Olson²¹ called ‘roving bandits’—are uninterested in civilian support thanks to pre-existing endowments such as a foreign patron or ‘lootable’ commodities like minerals or timber.²² Others rely on local populations to provide tax revenue, a supply of recruits, or to foster broad-based participation in anti-regime disobedience. These actors invest in governance to facilitate taxation and foster consent among local populations.²³ Marxist or Maoist guerrillas have long emphasised the need to be backed by the people, but empirical work demonstrates that this goes beyond ideology. The more counter-state actors depend on civilian support – whether for money, material, participation, or recognition – the more they provide them with services.²⁴

The ‘supply-side’ of competitive statebuilding is now well-documented. For instance, rebel groups founded governing institutions in a third of all major civil wars between 1950 and 2006,²⁵ ranging from building a tax system, conducting foreign affairs, or holding elections to providing services such as schools, healthcare facilities, police, and a court system.²⁶ Most groups only supply basic services to their fighters or supporters.²⁷ Yet almost 30 per cent since 1950 have provided complex services like education, healthcare, and courts, while a significant minority (13 per cent) went on to establish a comprehensive range of institutions, ‘creating something akin to full blown states-within-states’.²⁸ The investment is often worth it. Service provision may not guarantee a movement’s success, but not providing them is almost certainly a recipe for failure.²⁹

Similarly, civil disobedience campaigns often depend on providing alternative structures to attract a broad base of popular support – a strategy particularly well-suited to non-violent action over insurgency given the lower cost of participation.³⁰ Gene Sharp, one of the most influential theorists of non-violent action, lists almost 200 tactics that counter-state actors can draw upon, including the withdrawal from state institutions and setting up institutions of ‘dual sovereignty and parallel government’.³¹ In pre-1947 India, Gandhi referred to the Congress Party as a ‘parallel state’, a key part of the campaign to deny the British Raj of legitimacy.³² Opposition movements adopted similar strategies in Europe’s waning communist regimes. Czechoslovak dissidents formed a ‘parallel polis’ that encompassed uncensored information, unofficial education, popular music, and the black-market economy.³³ Poland’s ‘quiet revolution’ built parallel social, educational, cultural, and scientific institutions that aimed to ‘sap the communist system of its power’.³⁴

Early studies suggested that counter-state actors provided services to establish a social contract.³⁵ Subsequent research has specified how this works. Effective service provision contributes to perceptions of a group’s performance-based legitimacy,³⁶ as well as its moral and symbolic legitimacy,³⁷ both of which foster civilian support. For example, services can help pull local populations along a continuum of support, from hostility or indifference to active participation.³⁸ Service-providing rebels are more likely to engage in peace negotiations, thanks to their abilities in fostering broad-based social support and deterring spoilers.³⁹ And positive perceptions of counter-state governance compared to the state makes civilians more likely to stay in rebel-held territory.⁴⁰ As an Iraqi taxi driver said of his city under the Islamic State, ‘They have made Mosul better. The water is back. The electricity is back. The prices are lower’.⁴¹

Yet the ‘demand side’ of competitive statebuilding – the ways in which civilians engage with counter-state services – remains a black box for systematic empirical research. In the rebel governance literature, questions of civilian agency examine the various strategies ranging from full cooperation to outright resistance that civilians exercise vis-à-vis rebel groups.⁴² Others have explored decisions to aid rebels,⁴³ or the ways that communities can avoid violence during civil conflict.⁴⁴ These are important in re-framing civilians as agents rather than helpless victims, but they neglect the more nuanced micro-level relationships around service provision. Social movement scholarship, meanwhile, has tended to privilege structural explanations: the shifting political opportunities that allow for mass participation in the first place,⁴⁵ or on the resources that different counter-state organisations can mobilise towards a particular end.⁴⁶ Civilian decision-making, where it is explored, is often assumed to be driven by the strategic goal of toppling an unjust regime. But the apparent worthiness of a goal does not mean we can deduce the individual motivations of participants to that outcome. People have a range of motivations; whether sweeping political change is one of them remains an empirical question.

On the other hand, studies of counter-state service-provision fall into two categories. First are those that examine ‘public services’ in the aggregate.⁴⁷ Variation is explored in terms of scope, but less attention is paid to variation across services – despite indications that this may be significant. For example, in Hegar and Jung’s⁴⁸ study of service-providing rebels and peace negotiations, they briefly note that ‘Educational services appear to have a different effect [than] security or religious services’ before pointing out that ‘Theoretically these differences are entirely underdeveloped’. A second category focuses on individual services, adding nuance to a particular sector but not comparing across sectors. Loyle⁴⁹ highlights how different kinds of counter-state justice are used for different purposes: those that address grievances, such as truth commissions, aim to co-opt a population, while trials aim to project strength and authority. However, she does not assess how this compares to other services commonly provided by counter-state actors, such as healthcare or education.

Argument

This study aims to bridge this gap. It asks why civilians choose one source of governance over another in contexts where both state and counter-state institutions are supplying equivalent and overlapping services. The answer, it argues, depends on the interaction between two factors: the unique characteristics of different public services, and the salience of group solidarity.

Public services vary in ways that shape how people interact with them. First, they differ in the technical characteristics of how they are structured and delivered.⁵⁰ For instance, education and policing are transaction intensive, depend on vast organisational capacity, and require an army of ‘street-level bureaucrats’ to facilitate their delivery.⁵¹ Healthcare depends on high levels of provider expertise and discretion, contributing to wide information asymmetries that make it difficult for recipients to value or judge them. A postal service is best organised and delivered hierarchically.⁵² Primary schooling, in contrast, must be ‘co-produced’—depending not only on providers but also on the collective participation of recipients.⁵³

Providers and recipients face different considerations around a service’s technical characteristics. A critical variable for providers is the degree to which a service is visible to recipients: the more visible the service, the more political incentive there is to deliver it.⁵⁴ For counter-state actors, visibly demonstrating their capacity to ‘get things done’ can result in increased political support.⁵⁵ It also attenuates often violent reputations with more prosaic public management skills, demonstrating ‘that they are qualified and able to govern, not just to make speeches and set off bombs’.⁵⁶

For recipients, though visibility is important for perceptions of service delivery, other technical characteristics define their actual experience. These include the frequency and predictability with which people use a service, the extent to which it is collectively or individually received, and whether it is territorially concentrated or not. For instance, attending school is a quotidian, collective experience with a timespan of a decade or more. Schools are also focal points in a community. By allowing people to routinely meet and exchange information they can be sites of interaction and organisation.⁵⁷ Non-chronic healthcare, in contrast, is an episodic, individual, and often unpredictable one-off encounter, resulting in a much lower capacity for people ‘to gain information, exercise choice, bargain, and collectively organise’.⁵⁸

Beyond these technical considerations, public services also vary in their normative characteristics, or the key social and political values they communicate to a society.⁵⁹ By reaching into people’s lives, services can be concrete statebuilding tools ‘grounded in the idea of national solidarity’.⁶⁰ The ‘national’ framing of history or literature curricula are widely accepted as core elements of the school system and key vehicles to safeguard the national story.⁶¹ Healthcare is a powerful generator of social trust thanks to its importance in people’s lives,⁶² while legal institutions help provide an ‘aura of legitimacy’ to counter-state actors.⁶³ These norms are shaped by pre-existing expectations of what the state ought to provide. In the Democratic Republic of Congo, civilians were suspicious of any state effort to intrude into their lives.⁶⁴ Meanwhile, social welfare legacies in post-communist Europe have resulted in expectations of significant state involvement, raising the bar for potential state challengers.⁶⁵

Technical and normative characteristics endow services with unique political profiles. Yet focusing on characteristics alone would be insufficient to account for individual decision-making. In the politically charged and insecure contexts of competitive statebuilding, the limitations on ‘free’ individual choice are significant, and powerful group norms around who one ought to obey or support are common features. This leads to the second key factor: the relative strength of group solidarity and the social pressure that comes with it.

Solidarity refers to the binding and melding of individuals into a cohesive group or collectivity.⁶⁶ As solidarity increases, groups exert greater influence over their members’ decision-making. Highly bonded groups often exhibit a preferential bias towards other in-group members, more cooperation amongst in-group members than those of an out-group, and a heightened conformity with in-group norms.⁶⁷ These bonds have strong emotional significance. The more people feel attached to a group, the more willing they are to defend it.⁶⁸ This may be due to internal desires and feelings of appropriateness: the ‘pleasure of agency’ Elizabeth Wood⁶⁹ found among rebel-supporting campesinos during El Salvador’s civil war. But groups also control their membership though social pressure. Ethnic groups are particularly well-placed in this respect, often sharing a language, communal traditions, or other attributes that help them monitor individual behaviour, sanction deviants, and generate in-group trust.⁷⁰

However, the extent to which solidarity can influence individual behaviour rests on what I refer to as observability—the degree to which groups can monitor the behaviour of their members. This is where service characteristics come back into the picture. The heightened salience of a social group influences and constrains individual decision-making, but this influence is mediated by a service’s unique characteristics. Services that are collectively experienced and politically salient entail a higher degree of social pressure than those that are individual and infrequent. Below I highlight how these intersections play out regarding three services common to competitive statebuilding: education, healthcare, and justice.

Education

Among public services, education is most closely linked to nation-building. Public schooling equips populations with the linguistic and historical foundations of an ‘imagined community’, giving it a deep political salience for both providers and recipients. Historically, the cultivation of a common language through mass schooling produced more cohesive and homogenous populations.⁷¹ As such, states were (and still are) especially eager to drive the process, not only because inculcating a national identity ‘inoculates the population from external agitation and ensures resistance to alien rule’,⁷² but because providing education directly was often the only feasible way to monitor something so intangible as a patriotic ethos.⁷³

Moreover, the experience of receiving education is a frequent, predictable, and highly collective endeavour, making it uniquely well-disposed to observability. Participating in a physical classroom means that both students and teachers can monitor attendance, share information, gossip, and organise. Attending school may even be akin to an act of defiance or resistance. The salience of an appropriately ‘national’ education, along with its facilitating technical characteristics, are significant for individual choice; it is citizens, not just authorities, who care deeply about the beliefs and values that school curricula seek to promote.

Healthcare

Like public education, healthcare is part of the modern state ‘package’. Expectations of state provision – where they exist – mean that supplying health services can be an enticing prospect for an authority: the service is more closely associated with generating support for governments than any other, thanks to the importance it has for ordinary citizens.⁷⁴ How an authority allocates healthcare resources can also convey its values and priorities to a population. More equitable provision may promote social cohesion and a sense of shared identity.⁷⁵ Counter-state actors are often keen to take credit for the service – even if many leave its actual provision to religious organisations, international NGOs, or remaining elements of the government.⁷⁶ It may not be as tightly associated with fostering and safeguarding a particular national identity as education, but healthcare plays a powerful role in communicating values of equality and universality to its recipients.

Yet healthcare is a less uniform experience than education. Curative healthcare – supplied in hospitals and clinics – involves clear outputs and a connection between patient and provider. In contrast, preventative healthcare, such as a public health campaign, is less direct or visible.⁷⁷ Recipients also have widely different time horizons. One-off cases such as mending a broken arm have short time horizons, often being unpredictable, irregular, and used in moments of urgent need. They are also highly individual, with treatments varying among patients. In contrast, cancer treatments, birthing arrangements, or post-natal care have a longer time horizon, with more frequency and predictability than one-off cases. People will likely interact with healthcare discreetly, with an emphasis on trust, confidentiality, and perceived vulnerability.⁷⁸ In contrast to education, the observability of healthcare is muted due to its characteristics. Not only does healthcare largely sidestep questions of national identity, but it is also more difficult to collectively monitor the individualised and infrequent interactions that define it.

Justice

Legal institutions are a constitutive element of modern statehood and an important source of ‘output legitimacy’ for any authority.⁷⁹ As such, providing a system of law and order is often ‘the highest priority’ for counter-state actors in contexts of competitive statebuilding.⁸⁰ Rebel groups have found the service particularly appealing. One recent dataset records over 200 cases of rebel-initiated justice efforts across 57 different conflicts since 1946, ranging from ad hoc trials to fully fledged court systems with bespoke legal codes.⁸¹

Like healthcare, the characteristics of justice vary, ranging from small-scale disputes to criminal convictions. Unlike healthcare, they rarely involve the same information asymmetries, allowing people to weigh up different options and discreetly ‘shop around’ for the provider that offers the best chance of a favourable outcome.⁸² For example, until the Taliban takeover in 2021, Afghans in urban areas could choose between state courts, Taliban courts, and local forms of community dispute resolution when it came to resolving civil cases.⁸³ But other forms of justice can be highly visible, collective, and politically salient for both providers and recipients, ranging from amnesty agreements and public truth commissions to impromptu show trials.⁸⁴ Because of this variation, civil cases dealing with land, business, or family disputes tend to be ‘pragmatic’—people can ‘shop around’ thanks to the low political salience and individualised interaction. However, aspects of justice that deal with politically salient issues of treason or national identity and are delivered in a collective and highly visible manner will follow a different logic. In these cases, high levels of observability will exert a strong influence on individual decision-making.

Methods

This article uses a qualitative historical case study to highlight variations in how civilians navigated between services provided by state and counter-state actors. It does not aim to refute existing hypotheses, but to propose new ones – a building block procedure of theory development that seeks to fill a ‘space’ in existing accounts of competitive statebuilding.⁸⁵ Data for this study was generated during five separate periods of fieldwork in Kosovo between 2015 and 2018. Altogether 91 semi-structured interviews and two focus group discussions were completed, as well as multiple informal conversations, visits to relevant institutions and field sites, and an examination of primary and secondary materials. Interviewees were divided across Kosovo to ensure balanced geographical representation, covering both urban and rural areas. Most individuals identified as Albanian, as this was the community most realistically able to navigate between providers during the period in question. Informants were selected based on their experience interacting with, or providing, any of the three public services examined. The data gathered during fieldwork were triangulated with relevant academic literature, NGO and UN agency reports, and news articles.

The Kosovo Parallel State, 1989–1998

In March 1989, Serbian politician Slobodan Milosevic helped orchestrate the revocation of Kosovo’s status as an autonomous province within Serbia. Since the adoption of a decentralised Yugoslav constitution in 1974 the Albanian-majority province had enjoyed significant powers. Ethnic Albanians had largely replaced the once-dominant Serb minority in provincial politics and public sector employment, fuelling resentment among the latter and calls for even greater autonomy among the former. Milosevic used rising ethnic tension in Kosovo to build support for a broader ‘reorganisation of Serbia and Yugoslavia’—effectively a re-centralisation of power towards Belgrade and away from the constituent republics.

With the abolition of autonomy came widespread dismissals of ethnic Albanians from public sector employment. Within one year approximately 100,000 were either expelled their jobs or quit in protest.⁸⁶ Any public employee who remained in their position was required to sign a written pledge of loyalty to the Serbian state. Not signing invariably meant dismissal. In response to the mass dismissals, revocation of autonomy, and within the broader context of the ‘democratic spring’ sweeping Eastern Europe, a newly formed political movement called the Democratic League of Kosovo (LDK)⁸⁷ adopted a strategy of non-violent resistance.

Initially aiming to restore Kosovo’s autonomy, the LDK’s goal shifted to independence following the secession of Croatia and Slovenia from Yugoslavia. Along with emphasising human rights, democracy, and the moral superiority of civil disobedience – all meant to contrast with Serbian aggression and authoritarianism – the LDK established a clandestine ‘parallel state’ to govern Albanians in Kosovo.⁸⁸ Underground elections, a constitution, even a ‘presidential’ car helped establish a performance of de-facto statehood upon which it was thought Kosovo’s independence could be recognised by Western powers. This ‘state’ could now rely on unemployed Albanian professionals to provide concrete services like education, healthcare, and access to justice in ‘parallel’ to those provided by Belgrade. To support the effort, an underground taxation system levied funds both inside Kosovo and across the Albanian diaspora.

By 1995 the parallel state had established a network of functioning schools and health clinics, spearheaded an effort to reconcile thousands of customary vendettas known as bloodfeuds, and built up a system of civil dispute resolution – all while maintaining a strict policy of non-violence. That year, after considerable diplomatic and military pressure, Milosevic signed the Dayton Accords that brought an end to the Bosnian war. Kosovo was conspicuously absent from the agreement. Compared to the violence in Bosnia, Kosovo was ‘just not bad enough’ to warrant commensurate international pressure.⁸⁹

After Dayton the LDK’s strategy of non-violence faltered. The lesson appeared to be that ‘International attention can only be obtained through war’.⁹⁰ In late 1997 a group called the Kosovo Liberation Army (UÇK)⁹¹ issued their first public statement. The UÇK sought to achieve independence through a violent insurgency; in the eyes of the UÇK, the LDK’s pacifism only served Milosevic. Scant attention was paid to providing services. By 1998 the group had transformed into a guerrilla army that was capable of seizing and holding territory.⁹² Milosevic responded with a counter-insurgency operation that culminated with Serbian forces killing 45 Albanians at Raçak in January 1999. Western opinion turned sharply against the Serbian leader. Efforts to negotiate a final peace deal failed; NATO intervened and following a 78-day campaign Serbian forces agreed to withdraw from the province on June 9^th, 1999. The next day Kosovo was placed under a United Nations transitional administration.

Solidarity and Survival

The Kosovo parallel state was declared illegal by Serbia, remained unrecognised internationally, and operated underground for almost a decade. In the process, it fostered a powerful solidarity among Albanians. Routine police harassment for ‘irredentist activities’ added to its clandestine nature. Students avoided walking to class in groups and frequently changed their routes to avoid attention. Classrooms routinely changed location. Underground elections had to be carefully planned with polling stations in private homes. Parallel state tax collectors lived with a constant fear of being beaten or arrested if discovered.

Surveys captured the growing salience of ethnic identities. One survey in Kosovo found that Albanians characterised Serbs in sharply categorical terms; they ‘disliked of other nations’, and were seen as ‘sly’, ‘selfish’, and ‘rough mannered’.⁹³ Revealingly, the same survey found that the most common characteristic that Kosovo Serbs ascribed to Albanians was ‘united’, followed by ‘hating other peoples’. Surveys carried out in Serbia proper further illustrated how ethnic groups were seen as monolithic wholes. Over half of the respondents in one poll agreed with the statement that ‘all Albanians are primitive and uncivilised’.⁹⁴ Another found that the best ‘solution’ to Kosovo lay in the expulsion of Albanians.⁹⁵

Growing inter-ethnic cleavages were matched by increased social sanctions within the Albanian community. Words like ‘treason’ or ‘betrayal’ were used to describe Albanians perceived to be supporting the Serbian state or deviating from the LDK line. Not supporting a strike, voting in a boycotted election, or refusing to quit a public sector job could all result in sanctions. ‘The LDK brooked little dissent’ one author wrote at the time, ‘and those that challenged it were howled down in LDK publications and could even be ostracised from the tight-knit Albanian community’.⁹⁶ Underground Albanian-language media helped to spread information about suspected collaborators. ‘They were humiliated’, a former journalist explained,

We would publicly inform people that this guy or that guy is working for the Serbs and that he’s a spy. We would start spreading out rumours, in the newspaper or by word of mouth, that this person is a tool of the Serbs and is against us. It would warn people. The purpose was to create fear among the Albanians so that they wouldn’t work for the enemy, and to tell them that if you were, we would find out and would tell the people.⁹⁷

Increased solidarity was a powerful foundation for the parallel state. Its ‘officials’ went years receiving little more than a symbolic wage. Through them, the underground state provided school and university classes, health clinics, and a nascent system of justice. But they were not supplied in a vacuum; people could still choose between providers. Serbia’s healthcare system may have been harder to access after so many dismissed public employees had lost their social insurance, but it was still available. State education was restricted in several ways, but it could still be accessed. And the Yugoslav legal system continued to operate throughout the 1990s. Moreover, the 1990s ‘post-socialist’ transition saw a broader privatisation of formerly state-run services in Yugoslavia. Market-based private providers – most prominently in healthcare – were available for those who could afford it. Yet unlike the parallel state, private providers were not competing with the state; rather, they were complementing its provision.

I now turn to three services to examine how group solidarity intersected with different services to shape individual choices.

Education

Education was the most developed and visible aspect of the parallel state. Funds raised by LDK tax collectors were primarily spent on paying teachers and providing classrooms. But it was also the service most tightly linked to questions of national identity and resistance. In 1990, after the Serbian parliament adopted a new constitution, centralised school curricula were introduced across Kosovo.⁹⁸ Serbian history and culture were prioritised; Albanian language, music, history, and literature were reduced to a minimum. Soldiers guarded school entrances to ensure that no students or teachers could enter unless they agreed to follow the new curriculum.⁹⁹ The aim was overtly assimilationist: to develop ‘the feeling of permanent membership in the Socialist Republic of Serbia’.¹⁰⁰ Albanian media framed the issue in sharply national terms: ‘an attempt at a reduction of the Albanians’ national identity’ or an ‘insult to Albanian national dignity’. Resisting the imposed curriculum became a matter of ‘the defence of national identity’.¹⁰¹

In response, the LDK identified a network of warehouses, mosques, churches, or houses in which to carry out instruction. Parallel education resumed in these locations by January 1992, though classrooms would frequently relocate to evade police attention. The next year around 300,000 elementary school students and 60,000 high school students were attending classes in the parallel system, taught by nearly 18,000 teachers,¹⁰² while just under 14,000 students continued at the parallel university – a drop from almost 20,000 at the start of the 1991 academic year.¹⁰³ Only credentials from the state system were officially recognised. Nevertheless, diplomas issued by parallel authorities bore the official-looking stamps of the ‘Republic of Kosovo’.

For teachers, the decision to switch to the parallel system was framed in national terms, illustrating the political salience of education. ‘They changed our history books’, one teacher recalled. ‘They would write that the Albanian people were being assimilated. We didn’t want to teach this to Albanian students; it wasn’t right’.¹⁰⁴ Yet the LDK aimed to exert as much control over the educational system as the Serbian state. Teachers who did not opt to switch to the parallel system were threatened with dismissal. ‘We were under pressure because we didn’t accept the [Serbian] curriculum’, another teacher remembered. ‘But it was pressure from our side too, for us to teach in those conditions’.¹⁰⁵ The same combination of motives applied to students, who were a vanguard of setting the normative rules around parallel education. Actual credentials were secondary. ‘At that time, the main aim for a student was not to study, but to be part of this Albanian system we were creating’, a former university student explained. ‘So, you were a soldier in a way; you went to university to prove that you’re Albanian in Kosovo, to prove we can function as a state’.¹⁰⁶

The collective experience of classroom teaching was important as a display of unity, but its observability and salience carried a sharp edge. Classroom education facilitated compliance. ‘It was about coming and being educated and not stopping’, a former teacher recalled. ‘If a teacher didn’t come to teach one day, his students would know; they would say that he was a coward … that he was afraid of the Serbs and that’s why he didn’t come to work’.¹⁰⁷ Children at primary schools also felt pressure. One student from a mixed Croatian-Albanian family recalled chatting to her Serb friends on the school grounds in Serbian until Albanian classmates overheard her and started calling her a derogatory term for Serbs. Such social pressure could be highly effective. ‘I never spoke [to my Serb friends] again’, she remembered.¹⁰⁸ Names of ‘collaborators’ were spread through Albanian media. LDK newspapers published the names of lecturers who continued working at the state-run university or students who sat their exams. Those exposed either fled or were ostracised. ‘Everybody ignored them’, one teacher explained. ‘We didn’t socialise with them or their family members. If someone saw me hanging out with them, they would think I was the same as them’.¹⁰⁹

Healthcare

Parallel healthcare arose partly out of the widespread dismissals following the revocation of autonomy. But it was shaped by a bizarre episode in the spring of 1990 when, in early March of that year, over 200 Albanian students were rushed to hospital after they ‘experienced a strange odour in the school’.¹¹⁰ By the end of the month the number had risen to 7,000. The diagnosis according to one doctor was ‘poisoning with neurotoxic effects’.¹¹¹ A Serbian conspiracy was widely suspected, though international toxicologists blamed it on ‘collective hysteria’.¹¹² State authorities never invited a formal international investigation, nor did they allow Albanian doctors access to blood samples, adding to rumours that there was a Serbian plot to sterilise or poison Albanian children. Increased mistrust of the medical establishment meant parents refused to have their children vaccinated. Measles, polio, tuberculosis, and whooping cough increased. By the end of 1991 Kosovo had the worst health indicators in Europe.¹¹³

The alleged poisoning was seen as anti-Serb ‘propaganda’ by state officials and used as a pretext for the purges of Albanian medical employees that summer. More than half the Albanian doctors and medical staff in Kosovo were dismissed – a total of 1,855 employees, including 403 physicians.¹¹⁴ However, far fewer Albanian doctors willingly left their state jobs than teachers. The national stakes with healthcare were not as high, and social sanctions for doctors were limited. Even if they remained working in state-run hospitals, they were still perceived to be ‘helping their fellow Albanians’.¹¹⁵

The first organised effort to provide alternative medical services outside the formal state system was by an NGO called the Mother Theresa Society (MTS).¹¹⁶ Though formally registered under Yugoslav law, Albanians largely saw the MTS as the parallel state’s ‘Ministry of Health’. Its doctors had been dismissed from the formal system and its leadership was closely connected to the LDK, which was widely understood to be secretly organising the service. It was also clearly competing with the state for constituents. As a former head of the MTS explained, ‘people who were starving might go and cooperate with the Serbians, with the enemies. To prevent that we had to do something so that people would be well fed and well dressed, so that they wouldn’t go to the Serbians for help’.¹¹⁷

The aim of the MTS was at first humanitarian. Efforts focused on distributing monthly food parcels to around 46,000 families.¹¹⁸ In 1992 it began opening primary care clinics. In 1994, Médecins sans Frontières and Catholic Relief Services began supporting the medical arm of the MTS, allowing work to be scaled upwards. Two years later a maternity clinic opened in Pristina with support from a diaspora organisation in the United States. Though a private house with only seven beds, over 13,000 babies would be delivered at the clinic over the next three years. At its peak in 1998, there were 98 MTS clinics across Kosovo, supported by 120 pharmacies, and staffed by 239 general practitioners, 140 specialists, and 423 nurses.¹¹⁹ These services operated at impressive scale: over one million admissions were recorded in 1996 alone.¹²⁰

On the surface, parallel healthcare mirrored parallel education. Both depended on a core of professional ‘street-level bureaucrats’ and offered credible alternatives to the formal system. Yet decision-making around healthcare was markedly different. The cultivation and preservation of national identity took place in schools, not hospitals or clinics; doctors and nurses were never the bedrock of resistance that teachers were. As a result, charges of ‘treason’ or ‘betrayal’ carried less weight. Medical staff rarely felt the enormous social pressure to resign in protest with their colleagues that teachers experienced, nor did they feel the same pressure to participate in the parallel system. Not being on the frontlines of a national cause meant that dismissed doctors were free to set up private practices. Hundreds of fee-paying clinics – both legal and illegal – were established over the 1990s, giving people a private option beyond the state/parallel state dichotomy that characterised education.

Similarly, recipients could access healthcare discreetly. The collective social pressure and observability of classrooms were lacking in the individualised encounters at clinics or hospitals. A deep mistrust of state healthcare had developed following the alleged poisoning in 1990. Yet the trustworthiness of the provider informed calculations of where the ‘best’ service could be obtained, as opposed to the ‘rightfulness’ that framed decisions around schooling. Part of this depended on the level of treatment needed and capacity available. Minor or less-urgent issues could be referred to the parallel clinics. For critical cases requiring specialist treatment, there was little option beyond the formal system, even if it meant travelling to Serbia proper. ‘We didn’t trust them [Serb medical staff]’, one man remembered, ‘but we didn’t have a choice for my uncle’s cancer. These [MTS] clinics couldn’t afford the same facilities’.¹²¹ As an organiser of the parallel healthcare system recalled, ‘We can tell lies, but the truth is people who had money and needed special treatment went to Belgrade’.¹²²

Even so, the quality of the state system could be overplayed. State-run hospitals had relatively advanced operating theatres but were often short on essentials like syringes, needles, and sutures, forcing physicians to avoid giving injections and delay operations.¹²³ Even the quality of pharmaceuticals available was inferior to the MTS clinics, which had access to high-quality Western drugs and the expertise of international NGOs.¹²⁴ The Yugoslav economic crisis of the early 1990s also meant humanitarian services of the MTS were in high demand, even among Serbs. After recounting how he had told a police checkpoint that he had been delivering food aid to Serb families, a former MTS volunteer described how the police went to the families and shouted at them, ‘Never again do you take from Mother Teresa! We are a rich government; we are the ones that will help you’. And the Serb families said, ‘Yes, if you are a rich government then help us. Then I dont have to go to them [the MTS].’’¹²⁵ Even if such recollections are biased, MTS volunteers clearly envisaged themselves outcompeting the Serbian state on its own terms, by providing social welfare that a ‘proper’ state ought to provide its citizens.

Justice

Like education, Albanians were removed from the formal justice system after 1989. In 1990 the police were purged of Albanians. By 1993 over 300 of the 500-plus Albanian judges, district attorneys, and other senior officials had been dismissed. Of the 168 replacement judges appointed to Kosovo by the Serbian parliament, just twenty-five were Albanian; only sixteen were willing to take up office.¹²⁶ Serbian became the sole official language for legal proceedings and publications. Unlike education and healthcare, there was no parallel institution to absorb the dismissed legal professionals. Instead, the first instance of a parallel justice service was spearheaded by ethnographers and religious leaders with the aim of reconciling thousands of bloodfeuds in Kosovo’s rural hinterland.

Bloodfeuds refer to the customary practice of a ‘deliberately limited and carefully counted killing in revenge for a previous homicide, which takes place between two groups based on specific rules for killing, pacification, and compensation’.¹²⁷ In Kosovo, these rules were enshrined in a system of oral codes called kanuns after the Ottoman term for customary self-governance. Both the Ottomans and later Yugoslav authorities had tried to stamp out the practice. From 1977 to 1989, during Kosovo’s period of autonomy, the province’s Penal Code specifically incriminated acts related to feuding, which, beyond bloodfeud killings themselves included assisting a minor in committing a murder or limiting freedom of movement out of revenge. By the end of the 1980s around 2,000 families were involved in bloodfeuds – though because feuds applied to any member of the family or clan, the actual number of people affected was closer to 20,000.¹²⁸

For the LDK, reconciling the bloodfeuds became a demonstration of national solidarity through a strategic reframing of the traditional referent unit of the feud – the family – towards the nation. But it was also a matter of emphasising the ‘modern’ credentials of the parallel state to an international audience. Families ‘in blood’ were visited by councils of high-status men, such as LDK representatives or religious leaders, who would mediate between disputants to persuade the aggrieved party to ‘forgive the blood’—effectively ending cycles of tit-for-tat revenge that had often lasted decades. Appeals were framed in overtly national terms: ‘Isn’t it important to have our people living and unified’, or ‘do you think Europe will accept such barbaric practices?’. The first reconciliations took place among individual families in February 1990. By May they had transformed into ceremonial events, the largest of which attracted crowds numbering in the hundreds of thousands. When the reconciliation movement officially ended in 1992, it claimed to have reconciled 2,952 feuds.¹²⁹

Bloodfeud reconciliations were highly visible events and powerful symbols of the parallel state’s capacity – but not only for attendees. Proceedings were photographed, recorded, and widely disseminated throughout Kosovo. The projection of an ‘imagined community’ marked an indelible experience for viewers. ‘Everyone saw the reconciliations, whether on TV or in the news’, one man explained. ‘It was the first time we saw all of us united in that way; everyone at those places, standing together, watching this happen. It gave us resolve, made us feel strong’.¹³⁰

Reconciliation participants were also highly observable to each other. Even household reconciliations were collective experiences for those in attendance, including disputant families, local dignitaries, and the students and elders making up the visiting councils. And councils consciously used social pressure to facilitate forgiveness. Council members were specifically chosen based on the influence they had over a family: senior LDK officials for party members, religious leaders for those who attended the church or mosque. Councils ensured that the process and outcome of a reconciliation were written down; parties who refused would be publicly recorded. Bureaucratic formality contributed to a sense that what was being done was ‘national’ and ‘official’. Images of inclusion and exclusion were explicit; mediators explained that ‘you are either on this side of the river, or that side’.¹³¹ Those who refused faced exclusion. ‘We called them traitors’, one campaign leader explained, ‘for not overcoming their selfish desires for the sake of national unity against Serbia’.¹³²

At first, reconciliations only provided a choice between customary law and the parallel state. After the bloodfeud reconciliations ended in 1992, councils evolved to provide an alternative legal mechanism to the formal state as well. By 1998, a total of 23,000 cases claimed to have been resolved by these alternative structures.¹³³ Most were civil issues around debt, ownership of land, marriage and engagement, or traffic accidents. All council proceedings were transcribed, with documents mimicking those of the formal Yugoslav court system. For cases that could not be resolved – land ownership claims required the cadastral register, still firmly under state control – the council documented and ‘froze’ the dispute until ‘it would be handed over to the competent court of the [independent] Republic of Kosovo for a decision’.¹³⁴

Civil disputes were politically enigmatic and less visible than the bloodfeud reconciliations. Councils were now much smaller affairs; only a few members would pass decisions on a case. Moreover, the state system still provided a genuine alternative. When it came to civil issues, one business owner explained, ‘we did use the [state] courts, even if they were run by Serbs. Compared to the police – we all thought the police were dangerous – the courts still had some professionalism that people trusted’.¹³⁵

People could now ‘forum shop’ to an extent, though each forum offered different benefits. The ‘informalisation’ of Kosovo’s economy during the post-socialist transition of the 1990s meant that more disputes fell outside the remit of the state. Businesses handling smuggled goods – which many did out of necessity – could only approach a parallel council. However, the formal system continued to produce decisions that were legally binding and enforceable. Parallel civil councils, lacking any enforcement agency, could only offer mediation. The ambiguous ‘national’ stakes of civil disputes also made it difficult to win parties over with appeals to ‘the flag, the martyrs, and the nation’. These were the bedrock of the bloodfeud reconciliations, events that constituted visible, collective manifestations of national solidarity. In contrast, the caseloads of civil councils were precisely the opposite – a daily reminder of the fissiparous disputes Albanians inevitably still dealt with. Rather than seek to make their work visible and national, there was an interest to keep them invisible.

Discussion & Conclusion

Two key implications emerge from the case of the Kosovo parallel state. First, disaggregating competitive statebuilding into distinct service-specific activities can add nuance to the core question of how counter-state actors gain civilian support. A widely held assumption is that the delivery of services is a necessary condition for counter-state actors to win over local populations. However, analysing the ways that people interact with different services demonstrates that some services carry more ‘bang for the buck’ than others.

For authorities, visibility is an important characteristic. Services like education are tightly linked to fostering a collective identity and sense of nationhood. Being seen as guardians of a nation’s history through physical schools staffed with teachers can endow authorities with a powerful sense of purpose in the eyes of civilians. Education is also an important way for authorities to see, making it particularly amenable to the monitoring and enforcement of behaviour. Like education, the delivery of healthcare can convey normative commitments to a population. But visibility is compromised by the individual nature of the healthcare experience and the difficulties arising from asymmetries of information. And services like justice may be multifaceted, some aspects visible and politically salient, others more hidden and individualised.

Second, civilian support is not only a product of vertical relationships between rulers and the ruled. Understanding how support is gained or lost for counter-state authorities also requires examining its horizontal dimension, among local populations. This study lends support to the view that a social contract can be articulated through service delivery, but the ‘relational glue’ holding civilians and (counter-)states together goes beyond a simple supply-demand transaction.¹³⁶ It adds two additional points: the relational glue can be just as important when it holds civilians together through group solidarity; and that this solidarity – and its impact on individual agency – can be highly variable, with certain services tapping into or contributing more than others. People have agency and use it in ways that manage and secure their existence as best they can. But there are conditions under which doing what is individually best is difficult or dangerous. Questions of civilian support are not binary. Instead, choices play out in a complex context, where political desires and ideals rub up against apolitical calculations and tactical considerations.

In 1989, sovereignty in Kosovo effectively split into two irreconcilable halves, each competing to establish itself as the sole authority of the land. Against the centralising policies of the Milosević regime, an Albanian parallel state organised and delivered a range of services to a population that saw it as a non-violent vehicle for independence. This study has sought to examine how the people navigated this complex situation. As described above, it introduced two factors that help understand how people on the demand side of governance pluralism decide between authorities. On the one hand, group solidarity binds and melds people into collectives to different degrees, and higher group solidarity compels support for certain services over others. On the other hand, the characteristics of services shape the ways that individuals interact with them. Combining these factors can highlight different variations of decision-making, helping us to understand why people in contexts of competitive statebuilding choose one authority over another.

Acknowledgements

This work was supported by Oxford University’s Clarendon Fund and the Canadian Social Sciences and Humanities Research Council. The author gratefully acknowledges guidance, comments and feedback from John Gledhill, Joerg Friedrichs, and Alexandra Panman on previous iterations of the article and its arguments.

Disclosure Statement

No potential conflict of interest was reported by the author(s).

Additional information

Notes on contributors

Ian Madison

Ian Madison is an LSE Fellow in International Development and Humanitarian Emergencies (IDHE) at the Department of International Development. His research focuses on the politics of citizenship and public service delivery during periods of conflict or natural disaster. Ian holds a DPhil in International Development from the University of Oxford and an MSc in International Development and Humanitarian Emergencies from LSE.

Notes

1. Revkin, ‘Competitive Governance and Displacement Decisions Under Rebel Rule’, 46–80; Kalyvas, The Logic of Violence in Civil Wars.

2. Heger and Jung ‘Negotiating with Rebels’, 1206.

3. Arjona, Rebelocracy.

4. Grynkewich, ‘Welfare as Warfare’, 350–370.

5. Rigby, Living the Intifada.

6. King, ‘The Benefits of Ethnic War’, 524–552; Caspersen and Stansfield, Unrecognised States in the International System; Bakke et al., ‘Convincing State-Builders?’ 1–15.

7. Weinstein, Inside Rebellion; Berman, Shapiro, and Felter, ‘Hearts and Minds’; Flanigan ‘Charity as Resistance’; Berman and Laitin, ‘Religion, Terrorism and Public Goods’.

8. Stewart, ‘Civil War as State-Making’, 205–226.

9. Wimmer, Waves of War, 117.

10. Mann, The Autonomous Power of the State; Scott, Seeing like a State.

11. Corbridge, Seeing the State.

12. Arjona, Kasfir, and Mampilly, Rebel Governance in Civil War, 15.

13. Dorff, ‘Violent and Non-violent Resistance in Contexts of Prolonged Crisis’, 286–291.

14. Chenoweth and Stephan, Why Civil Resistance Works.

15. Kroonings and Kruijt, Societies of Fear.

16. Arjona, ‘Social Order in Civil War’, 103.

17. Kalyvas, The Logic of Violence in Civil Wars, 124.

18. Kasfir, ‘Rebel Governance – Constructing a Field of Inquiry’.

19. Loyle, ‘Rebel Justice during Armed Conflict’, 108–134.

20. Jackson, Life under the Taliban Shadow Government, 25, https://www.odi.org/publications/11144-life-under-taliban-shadow-government.

21. Olson, ‘Dictatorship, Democracy, and Development’, 567–576.

22. Weinstein, Inside Rebellion.

23. Arjona, Rebelocracy; Weinstein, Inside Rebellion.

24. Huang, The Wartime Origins of Democratisation; Stewart, ‘Civil War as State-Making’.

25. Huang, The Wartime Origins of Democratisation, 71.

26. Heger and Jung, ‘Negotiating with Rebels’.

27. Stewart, ‘Civil War as State-Making’, 206.

28. Huang, The Wartime Origins of Democratisation, 72; Florea, ‘De Facto States in International Politics (1945–2011)’, 788–811, http://doi.org/10.1080/03050629.2014.915543.

29. Stewart, ‘Civil War as State-Making’.

30. Chenoweth and Stephan, Why Civil Resistance Works.

31. Sharp, The Politics of Non-violent Action.

32. Migdal, Strong Societies and Weak States, 164.

33. Benda et al., ‘Parallel Polis, or An Independent Society’, 211–246.

34. Ramet, Thinking About Yugoslavia, 39.

35. Wickham-Crowley, ‘The Rise (and Sometimes Fall) of Guerrilla Governments’, 473–499.

36. Lee, Walter-Drop, and Wiesel, ‘Taking the State (Back) Out?’, 635–654.

37. Schlichte and Schneckener, ‘Armed Groups and the Politics of Legitimacy’, 409–424, http://doi.org/10.1080/13698249.2015.1115573; Mampilly, Rebel Rulers.

38. Flanigan, ‘Non-profit Service Provision by Insurgent Organizations’, 499–519.

39. Heger and Jung, ‘Negotiating with Rebels’.

40. Revkin, ‘Competitive Governance and Displacement Decisions Under Rebel Rule’, 46–80.

41. Revkin, ‘Competitive Governance and Displacement Decisions Under Rebel Rule’, 5.

42. Arjona, ‘Civilian Cooperation and Non-Cooperation with Non-State Armed Groups’, 755–778, http://doi.org/10.1080/09592318.2017.1322328; Barter, Civilian Strategy in Civil War.

43. Wood, Insurgent Collective Action and Civil War in El Salvador.

44. Kaplan, Resisting War.

45. Tarrow, Power in Movement.

46. Weinstein, Inside Rebellion.

47. Berman, and Felter, Can Hearts and Minds Be Bought?; Stewart, ‘Civil War as State-Making’; Huang, The Wartime Origins of Democratisation.

48. Heger and Jung, ‘Negotiating with Rebels’, 1223.

49. Loyle, ‘Rebel Justice during Armed Conflict’.

50. Batley and Mcloughlin, ‘The Politics of Public Services’, 275–285.

51. Lipsky, Street-Level Bureaucracy.

52. Andrews, Pritchett, and Woolcock, Building State Capability, 112.

53. Ostrom, ‘Crossing the Great Divide’.

54. Mani and Mukand, ‘Democracy, Visibility and Public Goods Provision’, 506–529.

55. Flanigan and O’Brien, ‘Service-Seeking Behaviour, Perceptions of Armed Actors, and Preferences Regarding Governance’, 622–651.

56. Szekely, ‘Doing Well by Doing Good’, 276.

57. Putnam, Bowling Alone.

58. Batley and Mcloughlin, ‘The Politics of Public Services’, 280.

59. Mcloughlin, ‘When Does Service Delivery Improve the Legitimacy of a Fragile or Conflict-Affected State?’ 1–19.

60. Béland and Lecours, Nationalism and Social Policy, 200.

61. vom Hau, ‘State Infrastructural Power and Nationalism’, 334–354.

62. Sacks, The Antecedents of Approval.

63. Arjona, Rebelocracy, 51.

64. Stel, Boer, and Hilhorst, Multi-Stakeholder Processes, Service Delivery, and State Institutions.

65. Cook and Dimitrov, ‘The Social Contract Revisited’, 8–26, http://doi.org/10.1080/09668136.2016.1267714.

66. Hechter, Principles of Group Solidarity.

67. Sambanis and Shayo, ‘Social Identification and Ethnic Conflict’, 294–325.

68. Tajfel and Turner, ‘An Integrative Theory of Intergroup Conflict’, 33–47.

69. Wood, Insurgent Collective Action and Civil War in El Salvador.

70. Humphreys and Weinstein, ‘Who Fights? The Determinants of Participation in Civil War’, 436–455.

71. Weber, Peasants into Frenchmen.

72. Darden and Mylonas, ‘Threats to Territorial Integrity, National Mass Schooling, and Linguistic Commonality’, 1447.

73. Pritchett, ‘When Will They Ever Learn’.

74. Sacks, The Antecedents of Approval.

75. Marmot, ‘Achieving Health Equity’, 1153–1163.

76. Mampilly, Rebel Rulers, 205.

77. Harris, Batley, and Wales, The Technical Is Political.

78. Gilson, ‘Trust in Health Care’, 359–375.

79. Krasner and Risse, ‘External Actors, State-Building, and Service Provision’, 545–567.

80. Mampilly, Rebel Rulers, 63.

81. Loyle and Binningsbø, ‘Justice during Armed Conflict’, 442–466.

82. Bierschenk and Olivier de Sardan, ‘Powers in the Village’, 145–173.

83. Jackson, Life under the Taliban Shadow Government, 20.

84. Loyle and Binningsbø, ‘Justice during Armed Conflict’.

85. George and Bennett, Case Studies and Theory Development in the Social Sciences, 78.

86. Mertus and Mihelić, Open Wounds.

87. In Albanian: Lidhja Demokratike e Kosovës.

88. Pula, The Emergence of the Kosovo “Parallel State,” 1988–1992.

89. Russell, ‘The Exclusion of Kosovo from the Dayton Negotiations’, 498, http://www.tandfonline.com/doi/abs/10.1080/14623520903309537.

90. Surroi, ‘The Albanian National Question’, 25.

91. In Albanian: Ushtria Çlirimtare e Kosovës.

92. Pettifer, The Kosova Liberation Army.

93. Nikolić, ‘Ethnic Prejudices and Discrimination’, 53.

94. Ibid.

95. Reitan, ‘Strategic Non-violent Conflict in Kosovo’, 80.

96. Judah, The Serbs, 314.

97. Author interview, Journalist, Pristina, 17 Dec 2015.

98. Kostovicova, Kosovo.

99. Vickers, Between Serb and Albanian, 248.

100. Kostovicova, “‘Shkolla Shqipe’ and Nationhood”, 164.

101. Ibid., 165.

102. Mertus and Mihelić, Open Wounds, 114.

103. ICG, Kosovo Spring (Pristina-Sarajevo, 1998), 23.

104. Author interview, teacher, Štrpce region, 7 March 2016.

105. Author interview, former teacher, Gjakova, 2 March 2016.

106. Author interview, former university student, Pristina, 11 Dec 2015.

107. Author interview, former teacher, Gjakova, 15 Dec 2015.

108. Jouffre, Drums of Resistance.

109. Author interview, former teacher, Gjakova, 31 Oct 2015.

110. Radovanovic, ‘On the Origin of Mass Casualty Incidents’, 103.

111. Mertus, Kosovo, 188.

112. Hay, ‘Yugoslavia’, 1196.

113. Rich, ‘Yugoslavia’, 232.

114. ICG, ‘Kosovo Spring’, 25.

115. Author interview, hospital administrator, Gjakova, 30 Oct 2015.

116. Officially ‘The Charity and Humanitarian Society of Mother Teresa’.

117. Author interview, Pristina, 25 May 2017.

118. Mertus and Mihelić, Open Wounds, 111.

119. ICG, Kosovo Spring, 26.

120. Ibid.

121. Author interview, Klina, 19 March 2016.

122. Author interview, Pristina, 19 July 2016.

123. Saric and Rodwin, ‘Health System in the Former Yugoslavia’.

124. Binet, ‘Violence against Kosovar Albanians’; Mason, ‘Rebuilding Pharmaceutical Services’.

125. Author interview, Pristina, 12 July 2016.

126. Vickers, Between Serb and Albanian, 274.

127. Boehm, Blood Revenge, 194.

128. Elsie, A Dictionary of Albanian Religion, Mythology, and Folk Culture, 43–45.

129. Luci, ‘Seeking Independence’, 101.

130. Author interview, Pristina, 5 Dec 2015.

131. Author interview, former reconciliation leader, Malisheva, 29 May 2017.

132. Author interview, Former council leader, Pristina, 26 May 2017.

133. March and Sil, The “Republic of Kosova” (1989–1998) and the Resolution of Ethno-Separatist Conflict, 1999.

134. Pritchard, ‘Albanian Law and Nation-Building’, 318.

135. Author interview, Pristina, 3 Nov 2015.

136. Migdal, State in Society, 6.