(iPoster) Strategic Nominations and Party Discipline under Maximal Voter Rules

• In Event: Poster Sessions: Formal Political Theory & Foreign Policy II – Alliances, Coercion, and Foreign Policy Decision-Making
  In Poster Session: Poster Session: Formal Theory

Abstract

This paper endogenizes candidate type in a dynamic model of electoral promises by embedding nomination and intra-party discipline into a repeated two-party competition with reputation. The environment is an infinite-horizon electoral game on a one-dimensional policy space with a representative median voter. Each period begins with a nomination stage: party i in {L,R} selects a nominee, and this choice determines the publicly observed type xi that summarizes the party brand-nominee bundle faced by voters. Nomination institutions are modeled by a parameter that captures elite control versus open primaries: greater party control yields nominees closer to the party’s electorally optimal type, while more permissive primaries shift nominations toward the party base. After nominees are selected, parties announce platforms; the election is held; the winner implements policy; and reputations update as a function of whether the implemented policy matches the announced promise. Voters coordinate on a refinement-maximal discipline under which a reputable candidate’s platform is treated as credible only if it is the most median-directed promise among those that are (i) incentive compatible for the nominee and (ii) electorally winning against the opponent. Intuitively, if a candidate could credibly and successfully promise a policy closer to the median, then any weaker (slack) promise is discounted.

I characterize Markov perfect equilibria and derive a credibility frontier mapping patience, nomination institutions, and intra-party discipline into the maximal sustainable set of credible promises. A maximal credibility radius summarizes the frontier: the greatest distance a reputable nominee can credibly move toward the median (relative to her ideal point) without inducing profitable reneging after election. The analysis yields three main results. First, maximal discipline selects slackless on-path platforms but can contract the credibility frontier in contested high-trust states (when both parties are reputable and electorally competitive), because tightening the on-path promise compresses the continuation-value wedge that sustains reputation. Second, anticipating this, parties optimally substitute from platform "slack" to candidate selection: more permissive nomination rules induce more extreme nominees and strictly reduce equilibrium credible moderation even when on-path platforms are slackless. Third, internal party discipline—modeled as a career-control relational contract contingent on promise keeping (e.g., renomination, resources, and intra-party career rents)—relaxes the nominee’s incentive constraint and can fully offset the credibility contraction induced by maximal discipline.

Formally, the state in each period consists of the pair of party reputations (sL,sR) with si in {G,B}, the publicly observed types (xL,xR), and the electoral environment. A reputable party (G) can sustain commitment only within the credibility radius, while a bad reputation (B) implies that all promises are ignored and implemented policy reverts to the winner’s ideal point. I provide conditions under which maximal discipline yields a unique on-path platform mapping (xL,xR, sL,sR) -> (pL,pR), and I characterize how the credibility frontier shifts with the nomination-control parameter and with internal party sanctions. In particular, the nomination margin is governed by a tradeoff between electoral advantage from moderating the nominee and intra-party costs of selecting against the base; the discipline margin is governed by how much the party can increase the nominee’s continuation payoff conditional on compliance. Comparative statics deliver threshold results: for intermediate levels of patience, increasing public coordination can reduce the maximal sustainable credibility radius in contested G-G states, but sufficiently strong internal party discipline restores (and can expand) the sustainable set of credible promises.

The model also delivers sharp asymmetry results. Cross-party differences in nomination control, office rents, or reputational fragility generate one-sided moderation and non-monotone polarization effects of primary reforms: a party with stronger internal discipline or more centralized nomination can credibly moderate toward the median even when its opponent cannot. These mechanisms yield empirically separable predictions for median-directed moderation versus two-party convergence and for heterogeneous responses to changes in public coordination capacity. In particular, increases in public coordination technologies reduce platform slack but need not increase convergence in contested high-trust environments; opening primaries increases polarization where both parties are trusted, while improvements in voter coordination reduce broken promises and can amplify asymmetric convergence when parties differ in discipline capacity.

Author

• Shiladitya Kumar, University of Houston