Introduction

Our objective in this chapter is to analyze how nonpolicy considerations – such as partisanship, candidate characteristics, and the sociodemographic characteristics of voters – affect parties' policy strategies. Our goal here is not to develop formal theorems about party strategies (such theorems are presented in Chapter 4), but to convey the strategic logic of party competition in the unified model via simple, nontechnical examples.We present a series of heuristic arguments that illustrate how, in multiparty or multicandidate elections, incorporating nonpolicy variables into a spatial model can alter the strategic calculus of vote-maximizing candidates, as compared to using the policy-only model in which policies are the only measured influence on the vote. After considering what happens when voters are motivated only by policies and vote in a deterministic fashion, we consider what happens when voters are motivated by a mixture of policy and nonpolicy considerations but still vote in a deterministic fashion. Then we introduce discounting by voters of the platforms presented by candidates as well as probabilistic voting. Finally, we illustrate the central intuitions on party strategies that grow out of the unified spatial model with empirical applications to one national survey – that of the 1988 French presidential election.

Each of our initial illustrations of the importance of nonpolicy factors and other aspects of voter choice involves only a one-step optimal strategy – that is, the vote-maximizing location of a focal candidate when the positions of the other candidates are fixed. Each candidate, however, can be expected to react to the locations of all the other candidates.