The Spoils of Agent Office: Determinants of Parliamentarian Behaviour Under Intra-Party Elite Conflict

Control over nomination is often seen as a core tool of party elites for attaining party discipline. Likewise, control over politicians’ career opportunities has been characterized as such a tool. Yet, control over these two tools might be in the hands of different actors with differing preferences. So how should parliamentarians then behave? We provide an answer to this question under a principal-agent perspective, and then employ a mixed electoral system for empirically testing it. We argue that those in in control of nomination within a party also profit from the parliamentarians nominated by them attaining offices. Given that this provides a separate source of utility, these principals have an incentive for trading dis-obedience of parliamentarians as their agents against these agents acquiring offices. Such is not the case for those in control of office allocation – it is only obedience that matters for them. Hence, a parliamentarian will abide by the wishes of the latter rather than the former, because she can count on those in control over her re-nomination allowing her the more leeway the more valuable the offices attained by her are to them. We then relate our argument to the case of Germany. Although most policy responsibility is centralized in Germany, entry into the political system remains decentralized. Nomination as a candidate in elections to the national parliament is either a local or state-level affair. Still, federal parliamentarians abide by the federal party line; and not by the wishes of their state-level principals. Germany lends itself as a test-bed for our empirical claims for two reasons. 1) The German federal electoral system is a mixed one, with a party-list and a district component. 2) Within each state, there is always only a subset of parties whose candidates have a chance at gaining district mandates, most Bundestag MPs of these parties owe their mandate to having won a district mandate, and it has been clear for MPs and party elites who these parties are in each state. Only party-list rank, thus, counts for candidates of all other parties in this state. This double feature allows us to test party-specific hypotheses on the role offices acquired play for the odds of re-election, as opposed to mere "show-case" nominations. We employ two datasets for testing our hypotheses. One of them, on regional-level and federal-level MPs, is a novel one addressing all the ca. 8.500 MPs of the German post-war state parliaments. We match these data with data on all 3.500 Bundestag MPs, and 35.000 Bundestag candidates of the post-war era. We thus cover the electoral and governmental fate of all German legislators of the period 1946-2009. The paper adds to theoretical insights on the intra-party distribution of power in that it goes beyond the enduring juxtaposition of party elites vs. rank-and-file. Empirically, it delivers the first encompassing picture of federal and state-level parliamentarian career tracks in Germany in their post-war entirety.