Party games: How political party change results from internal conflict

Existing explanations of party change do not explain why parties might take long time to adapt, or why they make choices that may not be "rational" at all. Part of the problem lies in treating the party as a unitary actor pursuing one primary goal, be it vote getting, office-getting or policy-getting. Such objections have been addressed by work that views parties as a collection of different groups. The model in this project explains party change as the result of rational choices made by competing internal groups seeking different goals within the same party. The main argument is that internal party groups pursue their own goals, their own rationality within the party, according to their hierarchy of preferences. Following Panebianco, the project maintains that change in the party as a whole follows competition between internal groups. A new coalition of internal groups may form in the party, and environmental conditions will shape whether that coalition becomes dominant. Adopting Kitschelt's typology of groups in social democratic parties, the model focuses centrists, ideologues and workers, each acting rationally to improve its preferred combination of electoral, ideological and economic payoffs. Each group has a different hierarchy of preferences, and each group dominates specific resources that help improve payoffs for the others. External electoral, economic and institutional conditions shape the priorities of each party grouping, and also the relative value of the resources controlled by each group. The model therefore helps predict the choice of coalition partners, as each group seeks to improve its combination of payoffs. The internal party context (party centralization and leadership autonomy) will then shape which coalitions become dominant, leading to party change. The model is tested by examining seventeen major conflicts (centered on party identity, policy and organization) in three large social democratic parties (PSOE, SPD and Labour), over a period from 1959 to 1989. The model is found to be quite useful in explaining the outcomes of party conflict, and points to further research on party structure, membership and goals.