Partisan 'Strongholds' in Comparative Perspective: Territorial Concentrations of Partisan Support Under Different Institutional Settings

The "nationalization" of party politics thesis claims territorial interests are increasingly being replaced by national issues, while the distribution of preferences among voters for policies/parties is becoming homogeneous across territorial units. Yet, territoriality in politics has not gone away, but appears resurgent in many established democracies. For example, trends of regionalization and decentralization since the 1970s has accelerated through Western Europe, with resulting increases in the territorial heterogeneity of party competition in some states. Under new majoritarian electoral rules introduced to Japan in 1994, some researchers claim trends of greater nationalization of Japan’s party system. Yet there is also contrary evidence. Conservative party candidates in certain regions have dominated individual and clusters of single-member districts continuously for nearly a quarter of a century. Elsewhere, new parties have emerged with strong regional biases of support. In subnational election results and party organization structures, the national parties’ strengths are highly heterogeneous. There thus remain clear differences in the competitiveness and rootedness of Japan's major parties across territories. These barriers to party nationalization in Japan need to be investigated from a comparative perspective. This paper seeks to do this by focusing on territorial concentrations of support for specific parties: i.e. what are commonly called partisan “strongholds”. How should such strongholds be defined and measured? Where do they emerge and under what contexts? What are their characteristics? Despite common references to parties having region-specific "strongholds"/"heartlands", there is no scientific measure to capture the extent and duration of these strongholds. This paper aims to create an index that can measure the extent of "regional strongholds" for statewide parties and use this index to map where and when such strongholds exist for major parties in selected states. Having identified these strongholds, the paper will then seek to categorize different types of “strongholds” in differing institutional contexts (centralized vs. decentralized state structure and proportional representation vs. majoritarian electoral systems). We investigate the existence and extent of strongholds in states which provide variation on the dimensions of interest (Japan, UK, Sweden, Canada, and Belgium), with a particular focus on multi-level features that underpin such non-competitive electoral regions/districts.