Introduction

Political parties have long been treated in both the theoretical political science and economics literature as unitary actors. Take, for example, the Downsian model of political competition (Downs 1957), the multidimensional spatial models (e.g., Enelow and Hinich 1984), and even the more recent models of redistributive politics (e.g., Dixit and Londregan 1996, Lindbeck and Weibull 1987). In many of these models, important results about policy outcomes hinge on specific assumptions about party objectives.

This begs the question of what political parties actually are: What is the notion of party that theorists should be interested in? Are they agglomerations of the policy positions of individual candidates (Osborne and Tourky 2002) or are they informative brand names aimed at voters (Snyder and Ting 2002)? Are they mechanisms to economise on the costs of standing for elections (Riviere 1999; Osborne and Tourky 2002) or simply credible commitment mechanisms (Levy 2004; Morelli 2004)?

One may go further and ask if we should really be looking at “parties” as coalitions of individuals or at coalitions of “parties” or groups of individuals – whether preelection or postelection – because, in many electoral systems, it is such coalitions that affect actual policy outcomes.

If we care about choice in politics, we should worry about the number of parties that a country has, their size, and the platforms they take. Presumably, systems dominated by one party – as in many African dictatorships – are less desirable than a multiparty system in which parties have divergent platforms.