Regionalist Challenges to European States: A Quantitative Assessment

Regional autonomy demands are widespread in Western Europe and are often seen as a significant challenge to the authority of the European ‘nation states’. This article formulates and tests possible explanations for variation in such demands in six Western European countries. As suggested by the existing literature, several cultural and economic characteristics of regions are associated with autonomy demands. To account for further variation and differences in specific demands, however, it is necessary to focus on the nature of party competition in a region. Autonomy demands will be strong if several regionally organized parties compete against each other and more restrained if regional and national parties are in competition. This suggests that regionalism is generally less consequential for existing state structures than often presumed. It seriously challenges state structures only under certain political conditions.

Acknowledgements

I am grateful to Fiona Adamson, Paolo Dardanello, James Fearon, David Laitin, Steven Wilkinson, the audiences at the various occasions where I presented this paper and several anonymous referees for comments on earlier drafts of this article.

Notes

1. For some overviews of these debates, see Caporaso (1996), Keating (1998), Kahler (2002) and Spruyt (2002).

2. Sorens (2005) is an exception. His study focused on support for secessionist parties, while this article focuses on variation in demands by regional political actors and offers a different argument for the role of party competition. Our conclusions are, however, largely complementary. Interestingly, similar studies as presented in this article have been undertaken for the Soviet Union and Russia (Emizet and Hesli, 1995; Treisman, 1997; Hale, 2000).

3. Strictly speaking, only the latter are autonomy demands. In fact, this third type partly cross-cuts the other two types (i.e. it constitutes an additional dimension), as cultural and economic autonomy demands can be accompanied by any of the forms of this third type, depending on the proposed financing if the demands were to be successful. However, tax autonomy demands can also be autonomy demands in their own right, as they are not necessarily accompanied by cultural or economic autonomy demands.

4. It is also worth noting that financial and fiscal decentralization are widely used in empirical work as proxies for decentralization more generally (Rodden, 2004).

5. This discussion defines globalization in economic terms, i.e. as international economic integration and considers European integration as one aspect of this phenomenon (Kahler and Lake, 2003).

6. This scenario has important similarities to Rabushka and Shepsle's (1972) argument about ‘ethnic outbidding’ in plural societies.

7. With the exception of the English regions in the UK, which only have administrative bodies.

8. Another complicating factor in defining a measure of regional autonomy demands, which is at most partially circumvented here, is how to account for existing regional powers in a country.

9. Selected sources: Deschouwer (1999); Flemish Government, Discussienota voor een verdere staatshervorming (1996); Christelijke Volkspartij (CVP) election programmes 1995 and 1999; Vlaamse Liberalen en Demokraten (VLD) election programme 1999; Socialistische Partij (SP) party resolution, June 1998 and interview with Ivo Van Den Bossche (advisor to the Flemish prime minister, Brussels, July 1998).

10. Selected sources: McRoberts (2001); CiU programmes (1993 and 1996 national elections and 1992 and 1995 regional elections); ERC website (wwwesquerra.org) and interview with Antoni Castells (Barcelona, April 2002).

11. Selected sources: Renzsch (1998); Bavarian Government programme (1998); Bavarian Government, Föderaler Wettbewerb: Deutschlands Stärke – Bayerns Chance (February 1998) and Financial Times, 29 December 1997.

12. Selected sources: NRW Government coalition agreement (2000); CDU-NRW press report (5 December 1998); interviews with Herbert Jakoby (NRW Ministry of Economic Affairs, Düsseldorf, November 1998) and Frank Littwin (NRW Ministry of Finance, Bonn, October 1998).

13. Selected source: ‘Aquitaine battles to win own identity’, Financial Times, 6 March 1998.

14. Selected sources: Bukowski (2003); interview with Ana Rodriguez (Andalucia office to the EU, Brussels, July 1998) and PA website (www.partidoandalucista.org).

15. Based on language codings in Grimes (1996). For a more extensive discussion of this measure, see Laitin (2000, pp. 148–149).

16. This variable is coded 1 if the religion with the most regional adherents differs from the one with the most adherents in the capital region of the country and 0 otherwise. Sources: Barrett (1982) and Lane and Ersson (1999, pp. 44–53).

17. Another suggested indicator is a measure of regional identities or ‘feelings of belonging’, as indicated by public survey data (such as Eurobarometer surveys). There is, however, likely to be too much endogeneity between such indicators and the dependent variable in this study.

18. See Eurostat's REGIO database and its statistical yearbooks for regions.

19. Sources: online publications of the statistical agencies of Germany (www.statistik-bund.de), Italy (www.istat.it), the UK (www.statistics.gov.uk), the Basque Country (www.eustat.es), the French Ministry of Foreign Trade (www.commerce-exterieur.gouv.fr) and the Belgian National Bank (www.nbb.be).

20. For an alternative definition, which requires regionalist issues to be the primary aspect of a party's platform, see Müller-Rommel (1998). De Winter (1998) provides a comparative, mostly qualitative assessment of a number of such parties. According to the present article's definition, a party does not need to see itself as ‘regionalist’ to be coded as regional, but it does need to focus its organizational efforts on a subset of all regions in a country.

21. The German Partei des Demokratischen Sozialismus (PDS) is a difficult case. In the 1990s, it contested national elections in the whole country, but regional elections only in the regions of former Eastern Germany and made little effort to establish organizations in the other regions. Consequently, it is here coded as a regional party. Only in recent years, when it has merged into the Linkspartei, has it become a genuine national party. Note, however, that recoding it as a national party does not substantively change the conclusions of the empirical analysis.

22. Sources: De Winter and Türsan (1998), Dupoirier (1998), Alcántara and Martínez (1998) and various internet sites with election data (www.vub.ac.be/belgianelections, elections.figaro.net, www.bundeswahlleiter.de, csa.berkeley.edu:7502/cattest.html, www.eleweb.net and www.election.demon.co.uk).

23. As an illustration of how to read Table 4, this corresponds to the first row and the second, fourth and sixth columns.

24. The other variables indicated in Table 2 were not statistically significant. As expected, the indicator for religious differences between a region and the centre cannot account for any of the variation in autonomy demands in these six Western European countries. The same conclusion applies to economic variables such as regional unemployment rates, regional employment by sectors and research and development expenditures. Moreover, the amount of European subsidies to a region is not significantly correlated with autonomy demands.

25. An analysis of the data set with regional parties as dependent variable reveals that the language differences and regional share of GDP (regional GDP divided by country GDP) can account for some but only a small part of the observed variation.

26. Moreover, it is debatable whether a loss of competencies necessarily means a ‘weakening’ of the central or national state, if we mean by this a loss of legitimacy or popular support (cf. Moravscik, 1994). This article, however, focuses solely on the acting capacity of the state, so does not discuss this interpretation.