Is the Whole the Product of the Parts? The Institutionalization of European Political Parties and Party Systems since 1848

The paper departs from a theoretical distinction between the party-level and the systemic level analysis of party politics (Randall and Svåsand, 2002). Following a conceptual re-examination of these two phenomena, we first review the existing literature on party and party system institutionalization (PSI), indicating limitations and inconsistencies, before offering a new analytical model which, on the one hand, focuses on the party-specific parameters of the system (party institutionalization or PI) and, on the other, concentrates on the interactions of the political parties (PSI). This will be followed by a contrast between data showing the degree of PI and the data indicating the level of PSI in 65 European democracies between 1848 and 2015. For the latter, we will introduce a new composite index of PSI (Casal Bértoa and Enyedi, 2016) which, looking at the predictability in the structure of inter-party competition at the time of government formation, captures the degree to which the patterns of partisan interaction can be considered to be stable over time. For the former, following Levitsky (1998), and trying to unpack the concept of PI, we will make use of various measures of PI: namely, Dix’s (1992) average party age to measure the dimension of rootedness, the Varieties of Democracy (2015) party organization index, and Lewis’ (2006) index of party stability. This way the paper not only constitutes an attempt to demonstrate to what extent the stability of party relations is helped by the stability of the individual units of the system, but also which of the two dimensions is closely related with the stability of partisan interactions. Moreover, instances when a pattern crystallized at the systemic level in spite of fluctuation at the level of parties will be contrasted with cases when partisan interactions remained inchoate despite the continuity of the political parties.