Nationalisation of regional elections in Central and Eastern Europe

Abstract

Territory is a salient issue in post-communist countries in Europe, yet subnational elections have received surprisingly little attention. This article analyses congruence between national and regional elections in Croatia, the Czech Republic, Hungary, Poland, Romania and the Slovak Republic held between 1993 and 2010. The findings suggest that election incongruence can be partly explained by approaches which work well in the West European context – territorial cleavages, regional authority and the second-order election model – but in order to fully grasp regional election outcomes in post-communist countries one also needs to take into account electoral alliance strategies and party supply change.

Keywords:

• nationalisation
• regional election
• congruence
• Central and Eastern Europe

Acknowledgements

A previous version of this article was presented at the workshop “Territorial Politics in Western and Eastern Europe” held at the University of Edinburg on 14–15 June 2012, which was organised by Valentyna Romanova. I wish to thank the organiser and discussants for their comments. I am very grateful to Ivan Koprić for help in obtaining regional election data for Croatia. I am also indebted to Gabór Dobos and István Gergő Székely, who provided support in classifying parties, identifying electoral alliances, and cleaning up data for Hungarian and Romanian elections, respectively. The research in this article has been supported by the Royal Society through a Newton Fellowship alumni award and by a fellowship award from the Hanse-Wissenschaftskolleg, Institute for Advanced Study, Delmenhorst, Germany. Both are gratefully acknowledged.

Disclosure statement

No potential conflict of interest was reported by the author.

Notes on contributor

Arjan H. Schakel is Assistant Professor in Research Methods at Maastricht University. His research interests include federalism, multilevel governance, regional elections and regional parties and his work has appeared in journals such as Comparative Political Studies, European Journal of Political Research, Party Politics, Regional Studies, and West European Politics. He is co-author of the book The Rise of Regional Authority (Routledge, 2010) and co-editor of the book Regional and National Elections in Western Europe (Palgrave Macmillan, 2013).

Notes

1 Estonia, Latvia, Macedonia, Montenegro and Slovenia have no regional government according to this definition. Albania (since 2000), Bulgaria and Serbia have regional government which consist of deconcentrated offices of the central state administration. Lithuania (between 1995 and 2010) has regional assemblies which consist of indirectly elected politicians. The federal and regionalised countries in Central and Eastern Europe have been excluded because of international interference in domestic politics (Bosnia and Herzegovina), because of particular confederal institutions (Serbia and Montenegro until 2006), and because of lack of democratically free and fair elections (Russia and Ukraine).

2 Before the introduction of regional elections, there were 77 okres in the Czech Republic, 49 województwa in Poland and 38 okresy in the Slovak Republic which functioned as deconcentrated central state administrations with no directly elected assemblies, but the województwa in Poland had an advisory council composed of delegates from municipalities from 1990 to 1998.

3 Bochsler (2010b) finds stronger nationalisation for CEECs with majoritarian electoral systems. None of the six countries included in this study employ a majoritarian or plurality rule except for regional elections in the Slovak Republic. In addition, I am mainly interested in the dissimilarity between regional and national vote shares, and electoral systems may contribute to this dissimilarity especially when electoral rules are different between the national and regional levels. This is the case for Hungary (mixed electoral system in national elections but proportional rule in regional elections) and the Slovak Republic (proportional electoral system in national elections but majoritarian rule in regional elections). I do not “control” for the difference in electoral systems in the models presented below, because this variable would only vary at the country level and thereby “absorb” large parts of the variation between countries and regions. In addition, the two CEEC-specific approaches discussed below (party supply change and electoral alliances) may be (and probably are) closely related to the electoral system (this issue is addressed in the discussion).

4 In the six CEECs, many ethnoregional parties which participate in national elections do so in an electoral alliance with a larger state-wide party but often compete in regional elections on their own. This is also the reason why this variable is considered under party supply change rather than under territorial cleavages.

5 With the category ethnoregional parties, I aim to include three kinds of parties. First, it includes parties which obtain their vote share in one (institutional) region only. Second, it may include parties which obtain their vote share in more than one (institutional) region but whose ideology advocates decentralization to a “region” which includes multiple (institutional) regions. Third, the label may include parties which claim to represent ethnic minorities and which obtain (significant) vote shares in a restricted number of (institutional) regions, because the ethnic minority is territorially concentrated.

6 The results are not driven by multi-collinearity between these variables. The Pearson correlations are below 0.5 in the region data-set and below 0.35 in the party data-set. In addition, the variance inflation factors for these variables are smaller than 2.

7 A proper analysis on “anti-government” and “pro-opposition” swings requires an analysis in which parties are coded according to their governmental status and vote share swings are not analysed according to their absolute values. In addition, next to a “pro-opposition” swing, the analysis should also explore vote share differences for new and ethnoregional parties which may benefit from the “anti-government” swing instead of opposition parties.