The Rationale Behind Party (In-)Consistency in Multi-level Systems

The assumption that political parties want to behave in a consistent manner is prevalent in the literature. Most prominently, and probably most aptly, it is made with regard to the cohesion of parliamentary groups in parliamentary regimes. However, the scope of this assumption has been substantially broadened. For instance, the literature about parties in multi-level systems draws on it in order to illuminate why national parties constrain their regional branches in terms of coalition formation strategies (e.g. Ştefuriuc 2009; Downs 1998) or programmatic position taking (e.g. van Houten 2009; Swenden/Maddens 2009). The assumption posits a two-step logic. First, if voters perceive parties to behave in conflicting ways on different levels of the system, they will punish them electorally because of the lack of credibility induced by inconsistency. Second, because party elites realize this, they will try to foster consistency across levels. While we can motivate the assumption very well from an information-theoretical point of view (cf. Snyder/Ting 2002), it is seldom empirically scrutinized. In this paper, we explicitly aim at the latter. Do voters perceive inconsistencies between the national party and its regional branches? What are the electoral consequences of programmatic inconsistency across levels? We draw on survey data and relate it to party manifestos in order to answer these questions.