Party Change in Western Democracies, 1965-2015: From a Generalization to the Overlooked Country-Level Variance

One of the most prominent developments in contemporary democratic politics is the change in party-society linkage. This change was widely recognized in several studies. Some scholars generalized this development under the title party decline while others saw it as a change in party-society linkage that was somewhat compensated by other developments. These studies pointed, among other things, to the decline in the number of party members, the weakening of party-interest group relationship and various aspects of voter behavior (e.g. increase in electoral volatility and decline in voter turnout) as indicators for party change/decline. In this Paper, we seek to demonstrate that this generalization might overlook variance in the development of the state of parties in various countries. Surely, in some countries parties declined, crashed or “just” lost a significant amount of their powers. But in other cases political parties adapted and performed rather well. In other words, we argue that there is a cross country variance that was often overlooked. Our analysis looks at country-level variance of party change. Twenty-five countries will be examined, using 13 indicators for the party-society linkage. These include, beyond the “usual suspects” (membership, volatility, party identification, turnout) additional indicators such as the relationship between parties and interest groups, the performance of parties in sub-national (regional and local) levels and party continuity. We start with a brief literature review of party change and show that it tends to generalize and pays little attention to variance. We then present the 13 indicators that we used and share the methodological challenges we faced while collecting and working with the data. The findings, processed into calibrated and standardized partisanship index, are presented in the main section. The findings indicate that within the broad phenomenon of party decline there is substantial county-specific variance.