Party Nationalization in Kosovo*

by Artak Galyan, Olimpija Hristova Zaevska, Roland Schmidt**

Expectations were running high for the November 3 municipal elections in Kosovo. Following the Brussels Agreement brokered by the European Union between the Kosovo and Serbian governments, Belgrade exerted considerable pressure on Kosovo Serbs and called on to their participation in elections for the first time since Kosovo’s independence in 2008. High turnout of Kosovo Serbs was expected to signal readiness for the integration of the four northern municipalities into Kosovo’s state institutions and as an endorsement of the creation of an association of Serb municipalities as envisioned by the Agreement.

Against this background, it should come as no surprise that much of the international news reporting on the election was dominated by the incidences of violence in northern municipalities, which triggered the closure of polling stations and required a re-run (see e.g. here, here, and here). The focus on the violent incidents and the disruption of elections in the north has casted a shadow on an otherwise strong participation of Serbs in other parts of the country. Municipalities with Serb majorities registered a higher turnout than areas predominantly populated by Albanians. Their high turnout was viewed as a positive evaluation of the efforts during the last four years to engage with this community.

Irrespective of the above, the recent elections were also of considerable importance with regard to Kosovo’s overall political development and consolidation of its political landscape. On a more descriptive level, how much the political landscape has changed becomes apparent by graph 1. While the first municipal elections in 2000 resulted in a widespread dominance of LDK, Kosovo’s party system today presents a much more fragmented picture. Despite retaining an overall dominant role with relative majorities in most municipalities the three biggest parties LDK, PDK, and AAK have lost almost all of their absolute majorities. Only PDK continues to have an absolute control of the municipal assemblies in Gllogovc/Glogovac and Skenderaj/Srbica, the birthplace of the party’s leader and incumbent Prime Minister Hashim Thaqi. This development reflects an increasing contestation within the party system. It is against this background that new parties such as ORA, LDD, AKR, and Vetevendosje have emerged throughout the years alongside the three big parties, challenging the latter’s absolute dominance across different municipalities.

GRAPH 1: Dominance of AAK, LDK, and PDK

Taking a closer look, and shifting the perspective from the municipal to the party level, the degree of party nationalization of the PDK, LDK, and AAK becomes of particular interest. The phenomenon of party nationalization refers to the geographical distribution of party support across a country’s territory. A nationalized party is equally strong (or weak) across the entire territory of a country, while a denationalized party’s support is concentrated in a particular region, with unequal distribution across the country (Jones and Mainwaring 2003; Bochsler 2010). From the Lorenz curves in graph 2, it becomes immediately visible how weakly AAK has been nationalized in comparison to PDK and LDK. AAK’s support is strongly concentrated in the western municipalities of Gjakove/Dakovica, Peje/Pec, and Decan/Decani, the latter being the home of Ramush Haradinaj, the party’s leader and former Prime Minister. At a closer look, however, it becomes also visible that the relative contribution of these three municipalities decreased over time (from 50 % in 2002 to 26 % in 2013). The party therefore balanced its relative geographic reach and increased its nationalization, but still falls short of the levels of LDK and PDK. Note that this observation stands valid irrespective of AAK’s loss in some municipalities in the 2013 elections.

GRAPH 2: Lorenz Curves - Distribution of absolute Votes

Contrary to AAK’s process of increasing nationalization, however, most parties in Kosovo experienced to a lesser or greater degree a denationalization of their support, with an increasing vote concentration in few municipalities. Remarkably, the new parties which have been entering competition in 2002, 2007, and 2009 have all denationalized at their second electoral cycle. Interestingly, medium and small Albanian parties such as ORA, LKCK, and PLK have left competition following a considerable move towards denationalization at their last electoral cycle. Graph 3 depicts these dynamics with the Party Nationalization Scores (PNS) (1) which measure the equality of the distribution of votes across municipalities. The higher the score the more evenly the distribution and the more the party is nationalized.

GRAPH 3: Party Nationalization Scores - LKCK, ORA, PLK

The low level of nationalization has been distinct for political parties representing Kosovo’s minorities (e.g. Turks, Bosniaks, Roma, Gora). The concentration of support for these parties stems from their geographically concentrated settlement (graph 4a and 4b). In some cases, it has also been the result of the redrawing of municipal boundaries , rather than reflecting a factual increase or decrease of their geographical reach (graph 4c). Turkish parties have had concentrated vote support in Prizren, as well as in Mamusha municipality since 2009. The Bosniak parties have had concentrated vote support in Dragash and Prizren municipalities, while the Roma, Ashkali and Egyptian community has been lacking such concentration due to its dispersion throughout Kosovo.

GRAPH 4a - 4c: Relative Distribution of Votes

The phenomenon of party nationalization is generally considered to be important in relation to at least three aspects: the structuring role of underlying cleavages; its nature for the contestation in the political system; and the consolidation of the party system. Cleavages along e.g. ethnic, regional or ideological lines divide the electorate into distinct groups. Party nationalization scores provide a useful tool to measure the degree of geographic concentration of electoral support and the prevalence of territorially distinct voting behaviour. Thus, they provide information on a party’s ability to appeal to voters across the entire territory of a country, rather than to only specific sub-groups which are territorially concentrated. Taking this aspect a step further, it should become also clear why party nationalization can have implications for the overall Quality of Democracy in a country. Democracies which feature predominantly regionally concentrated (= denationalized) parties, are likely to be confronted with a particularly strong pressure to balance these particularistic interests.

Similarly, party nationalization has important implications for the nature of the contestation within a political system. Countries with weakly nationalized party systems are almost synonymous to a country confronted with a dominant inter-regional political competition. Rather than featuring a competition which cuts across the municipalities it pits larger geographic areas against each other. While the presence of regional parties is a feature of many party systems, a party system that is dominated by de-nationalized parties can put considerable strain on a democracy's ability to strike compromises (this of course also depends on many other factors such as the political culture and institutional arrangements). Finally, the dynamics of party nationalization provide important information on the state of a country’s party system consolidation. In this context, the phenomenon is closely linked to the other core dimensions such as a system’s fragmentation, volatility and effective number of parties.


* The above presented data are the preliminary results of an ongoing project initiated by the Center of the Imperfections of Democracies (DISC) at Central European University on party nationalization in Bosnia and Herzegovina, Kosovo and Macedonia. The project is based on a comprehensive dataset of all general and local elections and aims at exploring patterns of party nationalization within and across these countries. These patterns promise to provide important insights on a number of issues relevant to the consolidation of democracy in these post-conflict societies.

** GALYAN, Artak. PhD. Candidate, Doctoral School of Political Science, Public Policy, and International Relations, CEU Budapest. Email: HRISTOVA ZAEVSKA, Olimpija. Research Assistant, Balkan Institute for Faith and Culture, Skopje. Email: SCHMIDT, Roland. PhD. Candidate, Doctoral School of Political Science, Public Policy, and International Relations, CEU Budapest. Email:

(1) Conceptually Party Nationalization Scores are almost identically to the standard Gini - coefficients. Mainly as a matter of convention, the only difference is that the Gini coefficient is subtracted from 1 for the PNS.