Democratic patterns and party systems in the Western Balkans

The article explores the role of political parties in hybrid regimes in the Western Balkans. The Western Balkans have been marked by considerable variations in democracy ratings over the past decades, altogether informed by the number of government turnovers and degree of political consolidation since the 1990s. Yet all countries in the region feature deeply entrenched political parties, that draw on either their socialist legacies, anti-communist profile, or the representation of a particular demographic, as well as their control over state institutions. Rather than differentiating between democratic and authoritarian systems in the region, the article stresses their commonalities. It understands parties and their extended networks as the central gatekeepers that mediate citizens’ access to economic and societal resources, effectively reversing the accountability between parties and voters. The article categorises party systems along two axes, based on the criteria of competitiveness and consolidation. Instead of adhering to the common dichotomy of authoritarian versus democratic systems, it suggests to understand plural and dominant systems as different manifestations of the same principle. It concludes that the main difference between plural and dominant systems lies in the fact whether power is monopolised or dispersed.


Introduction
The first multiparty elections in Southeastern Europe did not bring forth democratic and pluralist systems, but instead resulted in hybrid regimes that used nationalism and patronage to retain control, often to the detriment of minorities and the opposition. Wars and state collapse rendered democratic consolidation especially difficult in the countries that emerged from socialist Yugoslavia and Albania. Since the end of the wars and the collapse of the semi-authoritarian regimes of the 1990s, these countries seem to have followed the path of other Central European states in their apparent commitment to democratic procedures and EU integration. This perspective seems to have been confirmed by the repeated alternations of power that have taken place in most of the countries since the early 2000s. Yet this impression is deceptive. Not only has EU integration faltered, but so has democratic consolidation. Dominant parties monopolising state institutions and strong patronage-based party networks continue to curtail the consolidation of democracy. This occurs against a backdrop of a global trend of democratic decline, affecting neighbouring countries as well, such as Turkey and Hungary. perspective this essay takes on competitiveness and political parties in the Western Balkans. In Sects. 3 and 4, we analyse the seven country cases of this study according to their degree of competitiveness and consolidation.

Competitiveness and patronage-based parties
Political competition lies at the heart of every definition of democracy. However, as the concepts 'electoral autocracy' (Schedler 2006) or 'competitive authoritarianism' (Levitsky and Way 2002) already imply by their names, competition alone cannot qualify as a sufficient criterium. The decisive feature rather appears to be whether competition translates into a change of power. Or to use Adam Przeworski's poignant definition, "Democracy is a system in which parties lose elections" (Przeworski 1991, p. 10). There is a shared understanding that regular rotation is a necessary precondition to routinise interaction between parties and entrench democratic principles in social and political institutions. It indicates that incumbents are willing to hand over power and that political actors agree to settle their conflicts through the democratic process. Once democratic procedures have become "the only game in town" (Linz and Stepan 1996, p. 5), challenges to the political system and, hence, a revision to autocracy become less likely.
As Matthijs Bogaards (2000) argues in his analysis of party systems in Africa, this conviction is rooted in the prevalent focus on "the tasks of the opposition in the transition to democracy" rather than "the role that the opposition can play in the consolidation in democracy" (171). Referring to Sartori's seminal definition, Bogaards differentiates between competition and competitiveness specifying the former as "a structure, or a rule of the game", and the latter as "a particular state of the game" (Sartori 1976, p. 218;cited in Bogaards 2000, p. 174). According to this definition, the quality of democracy cannot be judged by the presence of competitors alone; but rather by the fact if the conditions in place allow for a regular turnover in power, thereby increasing actors' accountability and commitment to democratic procedures (Bogaards 2000, p. 175).
Over the past three decades since the end of state socialism, the seven cases under investigation have largely varied in their degree of competitiveness. In some cases, multi-party elections initially reaffirmed the socialist parties' superiority (Serbia 1990, Montenegro 1990. Others brought new dominant parties onto the scene (Croatia 1990, North Macedonia 2006-2017, or ushered in yet another period of party predominance following initial pluralisation (Serbia 2012-). Again, others have been marked by a regular rotation in power (Albania 1992/7-) or at least seen periods of repeated turnovers (Croatia 2000-). In order to account for the apparent divergence in outcomes scholars have scrutinised a number of (interrelated) structural conditions, such as economic development (Boduszyński 2010), experience of armed conflict (Fink-Hafner and Hafner-Fink 2009), nation-state building (Bieber and Ristić 2012;Zakošek 2008), EU integration (Bieber and Ristić 2012), or actor-centred factors, such as elites' susceptibility to western leverage (Boduszyński 2010), or the mode and sequence of transition (e.g. Bunce 2015; Džihić and Segert 2012;Fink-Hafner and Hafner-Fink 2009). Yet despite their different de-grees of competitiveness, all countries have remained susceptible to corruption and clientelism, reflected in institutional deficits, and, in most cases, a lack of rule of law. Apart from Croatia, none of these countries has lived up to a more substantial definition of democracy. According to the degree of competitiveness, scholars have placed these countries either more along the democratic or autocratic end of the regime continuum. Thus, the common scholarly approach of dealing with this disparity has been either 'conceptual stretching' (Sartori 1970) or conceptual innovations, such as autocracies or 'democracies with adjectives' (Collier and Levitsky 1997).
With this article we suggest a reversed perspective. Instead of focusing on the differences among the countries in this study, we emphasise their commonalities. At the core of our analysis stand political parties as the main actors in politics. Political parties are the central building blocks of representative democracy. They aggregate, articulate, and represent societal interests, and provide citizens with organisational platforms for participation (Lipset 2000). Yet recent research has substantiated that parties in the Western Balkans demonstrate substantial deficits in terms of their vertical responsiveness and accountability (Dolenec 2013, p. 126). In contrast to westerntype party systems 2 , party-voter relations in these countries are not so much programme-based, than they are rooted in patron-client exchange. Clientelist linkages render the "horizontal market place of elections" (ibid.) primarily a competition over resources and power. They install dependencies in the relationship between citizens and parties, thereby distorting voters' preferences and subverting representative institutions. Parties in the Western Balkans are overall highly centralised, hierarchical, and leadership driven. Intra-party democracy is rare, as are open leadership contests in which both challenger and incumbent remain in the party (Passarelli 2019). Whereas some parties have high membership numbers, these are based on patronage rather than signalling high levels of political participation. The ruling party of Serbia SNS, for example, claims to have over 700,000 members, or around 9% of the population. The DPS in Montenegro has a similar percentage of members (55,000) among the population (CZDS 2019).
This article echoes the notion of parties as 'patronage networks'. In doing so, it also draws on a body of work that deals with the entanglement of informal and formal politics in post-Soviet Eurasia, labelled neo-patrimonial (e.g. Lewis 2012) or patronage or patronal politics (Hale 2015). 3 We understand political parties in patronage-based societies as extended networks that are closely interlinked with informal elite groupings, local patrons, economic stakeholders, or criminal cartels. They act as gatekeepers through which clients pursue their political and economic 2 This is, of course, not to say that party patronage is absent in western democracies (see for example, Kopecky and Scherlis 2008). 3 While this essay draws on the underlying logic of the two concepts, we consider that they are not entirely transferable to the countries of this study. Neopatrimonialism commonly denotes authoritarian, that is dominant systems only. Hale's concept of patronage politics is more dynamic and mindful to the interrelation between political power and the societal equilibrium that results from clientelism. Yet his notion of monolithic power pyramids appears less applicable to the Western Balkans. We understand parties in this study as centrally organised conglomerates or networks of patrons and sub-patrons that are neither necessarily monolithic nor dominant actors.

Dominant Systems
Serbia ( 1990-2000, 2014-) Montenegro (1990) North Macedonia (2008-2017) Croatia (1990 Albania (1992)(1993)(1994)(1995)(1996)(1997) Serbia (2012-2014) North Macedonia  goals based on the exchange of concrete rewards and punishment, rather than abstract and impersonal principles (Hale 2015, p. 10). Although patronage-based societies share structural similarities, they differ in terms of the number and persistence of the patronage networks that compete for power. Whereas some countries may feature one or two main networks, others are characterised by several fragmented networks of which none has succeeded to establish predominance. The concept of patronage politics is all the more fruitful as it accounts for cycles of political opening and closure (Hale 2016). For example, a period of closure, or predominance, may be initiated by one network replacing or absorbing another. In contrast, a phase of opening, or pluralisation, may be introduced by the fragmentation or implosion of a previously dominant network. Thus, framing the ousting of one dominant player as "democratic breakthroughs" can be misleading, given the strong tendency for a new network to arise that is just as patronage-based as the previous one (Hale 2015, p. 87). Drawing on this notion, this article engages with the arbitrary differentiation of regime types based on the degree of competitiveness. Instead of tautologically identifying both regime type and regime quality based on the presence or absence of political turnovers, we suggest to understand predominance and pluralisation as oscillations in terms of power concentration and dispersion.
We propose two dimensions against which we analyse party systems. First, we differentiate between dominant and plural systems. Dominant systems are characterised by one dominant network monopolising the largest share of political and economic resources. Dominant networks reinforce themselves by absorbing or marginalising ascendant challengers. In contrast, power in plural systems is shared among several competing networks. We further distinguish plural systems as polarised systems, featuring two equally strong networks competing for power, and fragmented systems, marked by several smaller networks, none of which has established complete predominance. Second, we differentiate between consolidated and unconsolidated systems. We consider a system as consolidated if the main competing networks remain persistent over a period of two general elections. In contrast, unstable systems see the constant disintegration, re-appearance, and absorption of networks.
In the following sections we will explore four types of party systems in the Western Balkans, based on a distinction between dominant and multiple network-based systems on the one hand and consolidated versus unconsolidated systems on the other.

Dominant versus multi-network systems
The degree of power centralisation and dispersion greatly varies across the cases and period of analysis (table 1). More than half of the countries in this study have experienced at least one episode of party predominance since the fall of state socialism. In Croatia (1990Croatia ( -2000, Serbia (1990Serbia ( -2000, and Montenegro (1990Montenegro ( -2020, the disintegration of Yugoslavia was immediately followed by the rise of dominant parties that exploited the disarray of the transition years to consolidate power. North Macedonia experienced a concentration of power between 2008 and 2017 under VMRO-DPMNE, whereas Serbia has seen the return of a dominant party in 2014. Croatia's first multi-party elections in 1990 brought the anti-communist and nationalistic Croatian Democratic Union (Hrvatska demokratska zajednica, HDZ) to power. The HDZ of the 1990s is often described more as a movement than a political party. It "lacked a tight organization, and it was full of ideological inconsistencies and contradictions" (Ottaway 2003, p. 111). The party's extended network included dissidents as well as political and managerial elites of the predecessor regime (Zakošek 2008, p. 600). Power was strongly centred around the party leader, Franjo Tudman (also state president from 1990 to 1999). The HDZ established tight control over Croatia's state institutions, economy, media, security, and state apparatus, impeding the opposition's access to power and resources. Privatisations under the HDZ favoured individuals "who had close ties to the new government, either through party membership or through informal networks of kinship and regional or local origin" (Pauković and Raos 2015, p. 39). The end of the so-called 'homeland war' deprived the HDZ of an important source of legitimacy and instrument to demobilise the opposition. This led the party increasingly to shore up support through its clientelist networks, partly emerging from volunteers and war veteran associations, displaced persons, war victims, and 'patriotic forces' in the police and army (Zakošek 2008, p. 600). Tudman's death in 1999 heralded the disintegration of the highly personalistic regime and the electoral victory of a broad pro-European coalition.
Serbia's 1990s were characterised by the predominance of the Socialist Party of Serbia (Socijalistička partija Srbije, SPS). As the successor to the League of Communists of Serbia, the SPS inherited all economic assets, as well as the entire state, party, and media infrastructure of the socialist regime, which it further centralised. Power was highly personalised under the SPS' leader Slobodan Milošević (also president of Serbia from 1989 to 1997 and president of the Federal Republic of Yugoslavia from 1997 to 2000), evident by the high turnover in the party's main board (Bieber 2005, p. 238). Milošević heavily invested in building a loyal police apparatus while cutting finances to the comparatively more independent army (Cohen 2002, p. 178). Serbia's State Security assumed an instrumental role in circumventing international sanctions, that were imposed on the Federal Republic of Yugoslavia first in 1992 and again in 1998. It turned into a mafia-like structure, organising money launder-ing, and smuggling in cooperation with organised crime and paramilitary groups in Croatia and Bosnia and Herzegovina (Hajdinjak 2002, pp. 13-14, 25). The SPS stalled Serbia's privatisation to create a parastate economic elite, whose privileges and power were subject to complete partisan loyalty (Cohen 2002, p. 175). Since 1992, the SPS needed minor coalition partners to sustain its rule on the federal, national, and local level, of which the Yugoslav Left (Jugoslovenska levica, JUL), a satellite party formed in 1994 by Milošević's wife Mira Marković, remained the only constant. On the one hand, the SPS' erratic co-optation hampered the coordination of the already fragmented opposition. But it also impeded the formation of reliable alliances between the SPS and its changing coalition partners, on the other. By the end of the decade, Serbia's state assets were depleted by the Yugoslav dissolution wars, two rounds of draining sanctions, and the regime's parasitic elite, boosting the opposition's unification on the common goal to oust the SPS.
The most striking example of predominance is Montenegro's Democratic Party of Socialists (Democratska partija socijalista, DPS). Like Serbia's SPS, the DPS is the successor of the League of Communists of Montenegro. Amid its victory in the first multi-party elections, it retained its access to the socialist state's infrastructure, which it used to pack state institutions, public administration, the media, and police apparatus with party favourites (Darmanović 2003, p. 147). The DPS initially remained aligned with Milošević (until 1997), forming a joint state with Serbia until 2006. Like the SPS, the DPS sponsored large-scale smuggling organised by the republic's State Security in cooperation with crime cartels. The DPS was also favourable to partial privatisations and the flourishing of the (grey) economy (Ðurić 1999). Privileged individuals accumulated wealth through illicit businesses, which they were able to legalise with dubious privatisation schemes once Montenegro gained independence in 2006. Largescale privatisations, especially of coastal land, were one of the DPS-government's greatest incomes. It was not until the DPS' break with Milošević, accompanied by an internal party split in 1997, that the DPS lost its absolute majority in parliament. Between 1998 and 2020, the party regularly formed governments with minority and pro-Montenegrin parties, allowing each to maintain their own clientelist networks. At the same time, the DPS excluded pro-Serbian parties; although anecdotal evidence points to a partial toleration of independent networks maintained by the opposition and the Serbian Orthodox Church. Over time, highlevel corruption, a stagnation of rents, and the government's antagonisation of the Serbian Orthodox Church led some DPS' voters to opt out of the clientelist cycle, encouraging the otherwise fragmented opposition to form a coalition government after the 2020 elections.
In North Macedonia, the 2006 general elections ushered in a decade of predominance by the nationalist, anti-communist, and centre-right Internal Macedonian Revolutionary Organization-Democratic Party for Macedonian National Unity (Vnatrešna Makedonska Revolucionerna Organizacija-Demokratska Partija za Makedonsko Nacionalno Edinstvo, VMRO-DPMNE), following a period of government turnovers between the socialist successor, the Social Democratic Union of Macedonia (Socijaldemokratski Sojuz na Makedonija, SDSM), and the VMRO-DPMNE. During their alternation in office (1990)(1991)(1992)(1993)(1994)(1995)(1996)(1997)(1998)(1999)(2000)(2001)(2002)(2003)(2004)(2005)(2006), both parties had crafted their respective patronage machines and refrained from installing strong institutional safe-K guards. Thus, VMRO-DPMNE was astonishingly quick to capture state institutions, including the judiciary branch and security service, and utilise public resources for party gains. Its constant practice of scheduling early elections was both a means to reinvigorate clientelist voter linkages ahead of election cycles and disorient the opposition (Gjuzelov and Hadjievska 2020, p. 46). VMRO-DPMNE formed coalitions with Albanian parties, first the Democratic Party of Albanians (Partia Demokratike Shqiptare, PDSh) and after 2008 the Democratic Union for Integration (Bashkimi Demokratik për Integrim, BDI), as had become the norm in the country since 1992. The VMRO-DPMNE's junior partners were allowed to employ their clients in public administration and direct state resources to their constituencies. The SDSM' organisational strength and its alliance with a broad anti-corruption protest movement eventually brought the party back into office in 2017.
In Serbia's 2012 elections, the Serbian Progressive Party (Srpska napredna stranka, SNS) took centre stage. The SNS was formed in 2008 as a break away party from the far right nationalist Serbian Radical Party (Srpska radikalna stranka, SRS). Its ascent was owed to the disintegration of the post-Milošević coalitions. The SNS took over the largest share of the SRS infrastructure, while strategically rebranding itself to target median voters (Bieber 2020, p. 44). Serbia's weak institutions, a legacy of the continuing corruption during the post-Milošević era, presented the SNS with fertile soil to ingrain its patronage network (Pavlović 2020, p. 23). The party uses employment in the public administration and state-owned companies as an instrument to bolster its support base and pressure voters. The SNS also extracts public funds for maintaining its patronage machinery, for example through construction projects, such as Belgrade Waterfront (Pavlović 2020, pp. 29-31). The party has governed with several junior partners, from the far left to the far right, including the SPS, as well as minority parties. Routinely held early elections are an additional means to reinvigorate pre-electoral distributions of perks and keep the anyhow fragmented opposition from coordinating. Some countries within this study are marked by (periods of) regular power alternations. These countries can be subdivided into two groups: those featuring two major patronage-networks and those characterised by a fragmented competitive environment.
Albania (1992/7-), North Macedonia (1990Macedonia ( -2006, and Croatia (2000-) are all characterised by a strong polarisation between the socialist successor parties and their respective conservative antagonists. Albania is the most striking example of this pattern. In 1992 the Democratic Party of Albania (Partia Demokratike e Shqipërisë, PDSh) took over power, ruling as a dominant actor until the short civil war and state collapse in 1997. The successor to the communist Party of Labor, the Socialist Party of Albania (Partia Socialiste e Shqipërisë, PSSh), formed in 1991, reentered office from 1997 to 2005, and again since 2013. The two parties formally represent two different ends of the ideological spectrum, though they primarily mobilise votes through clientelist exchange (Bieber 2020, p. 78). The parties' support bases are partly regionally centred, the PDSh maintaining its base in the north and the PSSh in the south of the country. Although the regional divide is not entirely clear-cut, government alternations are evidently accompanied by regional shifts in the allocation of social assistance and appointments of public officials (Gërxhani and Schram 2009, p. 308). Both parties are extremely leadership-centred and disciplined. Almost 80% of minor parties emerged as splinters from either the PDSh or PSSh, usually resulting from disciplinary expulsions of defiant party factions (Krasniqi 2017, pp. 59-60, 65). The two major camps display a partial cooperation in maintaining their "duopoly [...], both in freezing other parties out of the system and ensuring strong party control over the state" (Bieber 2020, p. 76). Only the Freedom Party (Partia e Lirisë, PL), formerly known as the Socialist Movement for Integration (Lëvizja Socialiste për Integrim, LSI), a splinter of the PSSh, managed to establish itself as a third player, functioning as junior partner to whichever of the two major parties wins the elections (Krasniqi 2017, p. 65).
North Macedonia's transition to multi-party politics was administered by an allparty government (1990)(1991)(1992), followed by the return of the socialist successor SDSM into government (1992)(1993)(1994)(1995)(1996)(1997)(1998)(2002)(2003)(2004)(2005)(2006), briefly interrupted by one term in office of the VMPRO-DPMNE (1998. Both parties employed similar methods to bolster their respective clientele, such as employing voters in the country's bloated public administration or awarding loyal economic elites with public tenders in rigged privatisation and public procurement processes (Gjuzelov and Hadjievska 2020, p. 44). In the 1990s, private company owners could attain considerable wealth through smuggling during the international sanctions against Serbia and Montenegro and Greece's unilateral trade embargo against North Macedonia-a result of the country's name dispute with Greece (Hajdinjak 2002, pp. 17-18). Both SDSM and VMRO-DPMNE have regularly formed coalitions with Albanian minority parties, granting their junior parters their own patronage networks (Bieber 2020, p. 55). Both party camps have developed significant organisational capacity and cohesiveness. Despite their similarities, the parties serve clearly demarcated electorates, underpinning the polarisation of North Macedonia's system (Laštro and Bieber 2021, p. 622).
Since 2000, the HDZ and the socialist successor party, the Social Democratic Party (Socijaldemokratska partija Hrvatske, SDP), have taken turns at the helm of Croatia's government, with the SDP taking office from 2000 to 2003, and again from 2011 to 2015. The first SDP-led government undertook some efforts to de-politicise the police force and army. But it also continued in the HDZ' tradition of high-level corruption and abuse of public office, leaving much of the political and economic structures of the predecessor regime intact (Dolenec 2013, pp. 148-149). The process of EU accession compelled the HDZ to restructure and commit to reforms as well. Nonetheless, the party has remained highly personalised and clientelistic to this day. An example of this is Ivo Sander's resignation in 2009 as HDZ president and prime minister due to his entanglement in high-level corruption (ibid., p. 153, 156). In contrast to North Macedonia, Croatia's EU accession process has encouraged the two main parties to move towards the political centre, competing over a similar electorate (Raos 2015, 169). Bosnia and Herzegovina (1997-), Kosovo (2003-), Serbia (2000-2012, and Montenegro (since 2020) feature fragmented party systems. In Bosnia and Herzegovina and Kosovo the main competing parties have their roots in the wartime period, having transferred their networks into state institutions.
In Bosnia and Herzegovina political competition is fragmented between ethnonationalist parties, each catering to their own electorate. The segmented structure of the state and the geographic distribution of national constituencies encourage patron-client linkages along ethnic lines. Hulsey and Keil (2019) therefore suggest to "think of Bosnia as having three party systems" (406). Some even referred to Bosnia as a system with competitive authoritarian sub-systems (Kapidžić 2020b), such as the Republika Srpska and Croat majority areas. Still, even these sub-systems feature substantial competition. In Republika Srpska competition takes place between the two equally nationalist and patronage-based Serb Democratic Party (Srpska demokratska stranka, SDS), in government until 2005, and the Alliance of the Independent Social Democrats (Savez nezavisnih socijaldemokrata, SNSD), governing since 2005. The mainstay Bosniak party is the Party of Democratic Action (Stranka demokratske akcije, SDA) standing in competition with a number of Bosniak splinter parties and some non-nationalist alternatives, such as the Social Democratic Party (Socijaldemokratska partija, SDP). The Croatian Democratic Community (Hrvatska demokratska zajednica, HDZ), sister party of the Croatian HDZ, enjoys the strongest support in regions with a Croatian majority population, while it also faces competition by minor parties again of little ideological distinction. The parties within the respective national camps mainly differ in terms of the personal networks of their leaders that compete over access to power and assets (Hulsey and Keil 2019, p. 408;Hulsey 2018, p. 27). This dynamic is additionally supported by the absence of a single economic space, locating decision making on industrial, agricultural, social policies, and taxation at the entity level (Efendić and Hadžiahmetović 2016, p. 116). This has supported the 'oligarchisation' within the respective communities, promoting voters' dependencies on those parties and their extended networks that have access to resources. This also means that smaller parties that are either not part or trying to opt out of the clientelist distribution cycles are less likely to succeed (Hulsey 2018, p. 28).
Until 2017, the two main parties in Kosovo were the Democratic League of Kosovo (Lidhja Demokratike e Kosovës, LDK) and the Democratic Party of Kosovo (Partia Demokratike e Kosovës, PDK). The LDK was the leading force of the parallel state structure in Kosovo in the 1990s, whereas the PDK emerged from the Kosovo Liberation Army (KLA). None of the two parties ever achieved complete predominance, regularly forming coalitions with junior partners or with one another. Nonetheless, both parties, PDK and LDK, have established strong control over the state institutions and the economy, feeding their patronage networks (Bieber 2020, p. 70;Tadić and Elbasani 2018, p. 185). Political power in Kosovo is also intertwined with organised crime, particularly with structures of the former KLA. After the demilitarisation process, parts of the KLA entered into the newly established Kosovo Police and Information Service. The most prominent example is the former PDK Prime Minister (2008-2014) Hashim Thaçi who has been accused of involvement in organised crime (Proksik 2015, pp. 408-209). The PDK also used to draw on the support of international actors. This relationship, however, went sour once the PDK revoked investigation into crimes connected to the KLA in the 1990s (Elbasani 2018, p. 152). Growing corruption and international tutelage undermined Kosovo's established parties, paving the way in 2020 and again 2021 for a government headed by the movement party Lëvizja Vetëvendosje (Self-determination Movement, LVV).
In Serbia and Montenegro, the party systems fragmented following the electoral defeat of the dominant parties. In Serbia, the post-Milošević coalition already ruptured in 2001 with no party strong enough to take up the mantle. Although this period saw crucial reforms and a shared commitment to free and fair electoral conditions, institutions remained structurally weak and the heritage of the predeceding regime (Pavlović 2020), including organised crime and paramilitary structures, remained. In 2003, Serbia's former dominant party, the SPS, regained influence by supporting a minority government. Also, the SRS, a former ally of Milošević's SPS, emerged as the largest single party from 2003 until 2008, even if confined to the opposition.
In Montenegro, the defeat of the DPS in 2020 catapulted a slim majority of three different party lists, each with a distinct profile and constituency, into government. Conflicts over offices, resources, and the government's policy led to the coalitions dissolution after only 14 months. The next government, a minority government, included some of the DPS' former coalition partners and drew on support from the DPS. Yet also this government was toppled in a vote of no-confidence in 2022. Although the DPS itself has remained cohesive, it has lost significant electoral support in municipal elections held since 2020 and the presidential election in 2023.

Consolidated vs. unconsolidated systems
Both plural and dominant systems, as outlined above, can be qualified as consolidated or unconsolidated (see table). Since the 1990s, the countries of the Western Balkans have seen a variety of dominant and plural systems. None of the countries have been characterised by either variant throughout the period of multi-party politics. Instead, we have seen several shifts in both directions over the past decades. Thus, the overall level of consolidation varies.
To illustrate this, we can consider several cases. Croatia and Albania feature the most durable plural systems. In Croatia, the alternation of the two main camps was preceded by a consolidated dominant system headed by the conservative, anticommunist HDZ from the founding elections in 1990 until 2000. Albania is more difficult to categorise. The anti-Communist PDSh equally came to power in the first multiparty elections in 1992, quickly installing an authoritarian regime. This regime, however, collapsed only five years later amidst the civil war triggered by the state-wide breakdown of pyramid schemes that impoverished many citizens. Thus, Albania could be classified as a consolidated plural systems either since 1992 or 1997, featuring two main parties cartels on the centre-left and centre-right that have regularly alternated in power. In Albania, a change in government took place in 2005and 2013, in Croatia in 2003, 2011. Although the conservative HDZ in Croatian and the Socialist Party in Albania have overall served longer terms in office than their respective contenders, they are each confronted with a stable and relatively coherent opposition party.
North Macedonia falls into a similar category featuring two equally strong parties, the centre-left SDSM and the nationalist VMRO-DPMNE. Both parties have been, like their counterparts in Croatia and Albania, the largest parties since the early 1990s, regularly alternating in power. However, the authoritarian period of Nikola Gruevski (2006Gruevski ( -2016 constitutes a rupture. Especially in the second term (2008), which saw an increased control over the media and state institutions, the political system moved away from a consolidated plural system towards a concentration of power. The previous pattern was restored after 2017.
Kosovo and Bosnia can also be characterised as consolidated plural systems, though they follow a different pattern. Both have seen several strong parties competing over extended periods of time. Unlike in Croatia, Albania, or North Macedonia, these parties did not take turns in power, but governed in coalitions with each other. In Bosnia, the largest ethnonationalist parties formed regular coalitions, with brief interruptions, as a result of the state's consociational and federal design that prescribes power sharing between its constituent peoples. In Kosovo, the two largest parties, the LDK and the PDK, were never able to govern on their own and thus regularly formed coalitions with one another until 2017 (2008)(2009)(2010)(2011)(2012)(2013)(2014)(2015)(2016)(2017). Kosovo has moved to an unconsolidated plural system with the rise of LVV, a radical critic of the established elite which entered government first with the LDK in 2020 and once again in 2022 after a brief interlude.
Consolidated dominant systems are characterised by the enduring dominance of one party (and its satellites). The prime example is Montenegro, which was dominated by the centre-left DPS from 1990 until 2020. While the DPS had to rely on coalition partners after 1997, it remained the dominant party and faced a variety of changing and often highly fragmented opposition parties. This pattern was only disrupted once the DPS lost its majority in 2020. Its defeat in elections mirrors, with a delay of two decades, the decline of other consolidated dominant systems, such as Serbia under Milošević's SPS and Croatia under Tudjman's HDZ, both ruling from 1990 to 2000. Serbia saw a renaissance of a consolidated dominant system led by the centre-right SNS und Aleksandar Vučić in 2014.
Unconsolidated systems have been overall transitional in the Western Balkans, moving from dominant to a plural system or vice versa. Between 2000 and 2014, Serbia experienced an unconsolidated period, throughout which no single party dominated the government. From 2008 and 2012, the centre-left Democratic Party (Demokratska stranka, DS) clearly prevailed against other parties; yet this period is too brief to qualify as dominant according to our definition. At the same time, the political opposition was highly fragmented, with the SRS being the largest opposition party since 2003, but its influence declined in 2008, when the SNS was established. Between 2012-2014, the SNS gained office, but only established itself as a dominant actor after 2014, when it won an outright majority of 158 out of 250 seats in parliament. During this transitional period, comparable to the first mandate of Nikola Gruevski in North Macedonia between 2006 and 2008, the opposition remained strong, thus the imminent authoritarian shift was not yet conceivable. Finally, Montenegro since 2020 constitutes an example of an unconsolidated plural system. The defeat of the DPS in 2020 did not give rise to another dominant party, but rather several parties vying for power. As such, it rather resembles Serbia than Croatia after the fall of the dominant parties in 2000.
The regional trends thus indicate that consolidated systems prevail, and unconsolidated systems are generally transient. Overall, there is a high level of party continuity and new parties are largely the result of splits. The rise of a new party strong enough to challenge established parties (either in dominant or plural systems) is rare. The only new party that has gained considerable executive power as a result of elections and not of a coalition or party split, is Kosovo's anti-establishment and populist Vetëvendosje! (Yabanci 2016). Throughout the region, the main parties have either remained the same or emerged as splinters or mergers of existing parties. Serbia's SNS, for example, entered the political scene as a splinter of the SRS and does therefore not qualify as newcomer. New challengers such as progressive parties like Naša stranka in Bosnia and Herzegovina, URA in Montenegro, or Moramo in Serbia, have achieved only modest electoral successes, even if some were able to shape government policies. In Croatia, Možemo, a left-green group managed to enter parliament in 2020 and gain control of the capital Zagreb in 2021. Yet, Vetëvendosje! and Možemo are rather exceptions to the rule. As a general trend, both plural and dominant systems are fairly closed to newcomers. This phenomenon can be explained by the new parties' lack of institutional and economic resources, which impedes their ability to compete against the patronage networks of the established parties.

Conclusion
While all countries of the Western Balkans have experienced alternations of power since 1990, in the case of Montenegro just once in 2020, two regional patterns have emerged: dominant and plural systems. Both types are fairly stable, whereas instability commonly marks the transition from one type to another. Serbia is the only example in which instability lasted for as much as twelve years (2000)(2001)(2002)(2003)(2004)(2005)(2006)(2007)(2008)(2009)(2010)(2011)(2012). It would be simplistic to denote plural systems as democratic and dominant systems as authoritarian. This ambiguity can be illustrated by the Montenegrin case. In the last decade of the dominant DPS' rule, Montenegro's media system was considerably freer and more diverse than in some plural systems. This was mostly due to the fact that, amid the rise of private TV stations and newspapers, the ruling party was unable to control the entire media landscape. Yet whereever possible, the DPS quelled critical voices and impeded pluralism.
In comparison, a plural environment does not mean the absence of repression, or even the use of lethal force. Albania's DP-led government, for example, employed police violence against the opposition in 2011, resulting in the death of three anti-government protestors. Equally, plural systems, even those that are fairly fragmented, do not provide a safeguard against state capture, as the cases of Bosnia and Herzegovina and Kosovo exemplify.
Thus, this essay has suggested to conceptualise plural and dominant systems as different manifestations of the same principle. As we have shown, all countries in this study are marked by deeply entrenched, patronage-based parties that generate support through clientelist exchange. Given that parties and their extended networks are the central gatekeepers that regulate citizens' access to economic, political, and societal resources, the accountability between parties and voters is fundamentally reversed. The main difference between plural and dominant systems is whether power is monopolised or dispersed. In plural systems, resources are divided between several patronage-based parties. The incumbent has a clear advantage of controlling state resources-essential for employment, media control, and economic power. But still, resources are partly owned by other patronage-based parties that act ambivalently both as challengers, as well as predictable and oftentimes complicit partners. In contrast, dominant systems feature one main patronage-based network that has realised comprehensive control from the municipal to the state level.
Funding This project has received funding from the European Union's Erasmus+ programme -Jean Monnet Chairs under grant agreement 20191622. This publication was produced with the financial support of the European Union. Its contents are the sole responsibility of the authors and do not necessarily reflect the views of the European Union.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4. 0/.