Decentralization as a National-Level Electoral Strategy in Western Europe

• In Event: Peripheral Politics, Central Problems: Local Politics in Comparative Perspective

Abstract

Why do existing governments choose to give away some of their powers to subnational authorities? Given that power maximization is assumed to be a central goal of political actors, it may seem puzzling that parties controlling the vast administrative, political, and fiscal competencies of a central government would be willing to transfer some of them to other levels of government. And yet, over the past forty years, waves of decentralization have swept across Latin America, Africa, Asia and Europe. In Western Europe, for instance, levels of decentralization have increased since the 1970s. The nature and extent of the reforms has though varied across and within these countries, from the maintenance of centralized rule in Iceland to the creation of subnational assemblies with directly elected officeholders and significant administrative, policy-making and tax-raising powers in Belgium, Spain, and Italy and the United Kingdom.

Explanations offered for why national governments voluntarily weaken their own power have often focused on decentralization’s effect as a multiplier of powerful subnational offices (i.e., office multiplication theories of decentralization). Scholars have argued that governments decentralize either because it benefits the country, in terms of economics or political stability (e.g., Alesina et al. 1999; Panizza 1999; Stepan 1999, 2001; Garrett and Rodden 2003; Brancati 2009; Hooghe et al. 2010; Hooghe and Marks 2012), or because it benefits the governing party, specifically by increasing the party’s power at the newly strengthened subnational level (O’Neill 2003, 2005; Escobar-Lemmon 2003; Sorens 2009).

In treating decentralization as just one of many possible forms of institutional rearrangement that parties can use to benefit themselves or the country, these theories have downplayed the importance of decentralization to voters. Yet, especially in countries with ethnoterritorial parties, decentralization is also an explicit policy demand of those parties and their voters. To help account for the pattern – degree and timing – of decentralization reforms, this paper tests an alternative strategic explanation rooted in the exigencies of national-level party competition. Consistent with the observations of other scholars, including Meguid (2008), Hopkin and van Houten (2009), Maddens and Swenden (2009), Toubeau (2011) and Lublin (2014), I maintain that decentralization is a strategy of policy appeasement. In this paper, I develop the argument that governing parties facing threatening ethnoterritorial parties in national-level elections will implement decentralization reforms to attract their pro-decentralization voters. Thus, in contrast to Sorens (2009) and others, the focus of the governing party is on bolstering its national, as opposed to subnational, electoral support. I furthermore argue that the degree of decentralization enacted depends upon the policy demands of the ethnoterritorial party, the degree of governing party electoral vulnerability and the level of ethnoterritorial party threat in a given region.

Statistical analyses of an original dataset of decentralization levels across the subnational regions of Western European countries from 1970 to 2010 – the first time-series cross-sectional analyses of its kind – confirm that governing parties transfer additional competencies to regions when the party is electorally vulnerable in national elections, but only where and when there is an ethnoterritorial party threat. Conversely, if a party is not dependent upon a region for national parliamentary control, the presence of a strong ethnoterritorial opponent in that region actually has a negative effect on decentralization levels. Consistent with a view of decentralization as a voter appeasement strategy, the analyses also indicate that governing parties tailor their use of decentralization to the demands of an ethnoterritorial party; decentralization is not extended by even vulnerable governments when the threatening ethnoterritorial party is demanding regional independence (i.e., secessionist party) as opposed to regional autonomy (i.e., autonomist party).

This paper has numerous important implications. For instance, this paper allows us to understand why we see parties create and strengthen subnational governments that they do not always expect to control. It also speaks to the larger debate on the causes versus effects of decentralization. The results here provide additional support for the claim that ethnoterritorial party vote share is a cause of, rather than simply the effect of, greater decentralization (Rokkan and Urwin 1982; de Winter and Türsan 1998). Lastly, the paper confirms the larger conclusion in the endogenous institutions literature that governments will only change the rules under which they were elected when their ability to maintain electoral control is threatened.

Author

• Bonnie M. Meguid, University of Rochester