Models of Multiparty Electoral Competition

K. Shepsle

• Publisher: Routledge
• Year: 2012
• ISBN:
  9781135646158
• Edition: 1
• Topic: Business
• Subtopic: Business General
• Index: Business

Kenneth A. Shepsle surveys the formal literature on multiparty electoral competition.

Models of Multiparty Electoral
Competition

Kenneth A. Shepsle

Harvard University, Cambridge, Massachusetts

Part I. Introduction

1. Theoretical Orientation

Rational choice theories of politics, in their infancy as late as the 1960s, have matured rapidly. In political science, rational choice has moved from minor tributary to main stream, from back alley to main street. In the 1960s rational choice theories (then called ‘economic approaches to politics’) had to fight for a place at the intellectual table, and had to dodge the sharp elbows of sociologically-and psychologically-based theories. Today, papers based on the rational choice paradigm are amply represented in the pages of leading journals and on the programs of major professional meetings. Indeed, today there are a number of different schools of thought all based on some version of the rationality hypothesis.

One of the unexpected by-products of this development and growth has been the rediscovery, both in economics and political science, of the importance of institutions. Aided and abetted by the noncooperative revolution in game theory, a focus on rational behavior has forced scholars to model social processes explicitly and to tie equilibrium concepts to particular institutional arrangements. Different rules of participation, methods of delineating the alternatives available for choice, and procedures for aggregating the expressed preferences of eligible participants over available alternatives — in short, differences in institutional practices — have led to different hypotheses about rational behavior.

It is, therefore, something of an irony that rational theories of politics are principally an American affair. First, the scholarship is overwhelmingly American, although there are growing sections of European political science and economics engaged in rational choice modeling of politics, and an increasing receptivity among European journals to their output. Second, and more significant, rational theories of politics are principally about politics as practiced in America. Models of elections are typically based on the existence of two parties (sometimes notationally referred to as D and R!). Models of legislatures often assume the characteristics of the American House of Representatives and Senate (a charge to which I, personally, plead guilty). Yet surely America is a sufficiently ‘exceptional’ political system to discourage its use as the principal basis for scientific generalization.

This theoretical ethnocentrism is no more apparent than in the study of elections. It is convenient to date the formal study of this topic with Downs’ (1957) seminal work. Downs’ spatial model of electoral competition is essentially Hotelling’s (1929) spatial model of duo- polistic competition as applied to politics. Although Downs did devote some casual attention to multiparty circumstances, his legacy is a model of dualistic (duelistic?) competition in the Anglo-American mold. For more than three decades, a research program has enriched and embellished the original Hotelling-Downs formulation (see Davis, Hinich, and Ordeshook, 1970, and Enelow and Hinich, 1984, for surveys).¹ Only very recently has this foundation been appropriated for the formal analysis of electoral politics in settings different from the Anglo-American. It is this work that constitutes the focus of the present volume.

During the same period that Downs and his successors were elaborating a political version of Hotelling’s spatial model, economists were extending the original model of duopoly under the rubric of ‘spatial economic competition’ or ‘location theory’ (see Gabszewicz and Thiesse, 1986, Tirole, 1988, and Greenhut, Norman, and Hung, 1987, for surveys). To a considerable degree these two bodies of research have traveled down parallel tracks with only an occasional cross reference. Thus, a second purpose of this volume is, if not thoroughly to integrate these two research traditions, to lay them down side by side to determine if insights from one can inform research in the other.

Before moving directly to these two tasks, there are some preliminaries to mention briefly. Central to the Hotelling-Downs research program, as well as to the more recent extensions in political science and economics, is the spatial framework. Accordingly we must describe that.

It is assumed in the bodies of literature reviewed here that the foremost basis of competition among agents is spatial. For economic agents spatial location affects customer price, since products must not only be produced but also delivered. The spatial distance between production and consumption must be traversed, its cost borne by the producer (shifting its marginal costs upward) or directly by the customer (who pays a mill price at the production center and then pays to transport it to his or her consumption site).

Most models of economic location are unidimensional, though there have been some multidimensional extensions (see Section 2). Customers are distributed on this dimension — a main street through town or a beltway around it — and are assumed to take seller location into account in making their purchasing decisions. Anticipating this, sellers choose locations to maximize their returns.

An alternative interpretation of spatial competition may be offered in terms of product attributes. Here, by analogy (a formulation especially suggestive of political competition) producers ‘locate’ in a product quality space over which customers have preferences. Thus, location may be taken either literally or figuratively.

In models of electoral competition, candidates or parties (called electoral agents) choose locations in an ideological or policy space over which potential voters have preferences. Like producers, electoral agents seek to maximize their ‘returns,’ though what it is they maximize is subject to controversy. In the political science literature, there is much intellectual hand-wringing over the substance of objective functions, but no consensus has been reached.² Also in the political science literature, much attention is given to multidimensional formulations, since the assumption of a single ideological dimension as the basis for competition strikes most as implausible. Here, too, there are difficulties (associated with the fact that, generically, voter preferences are collectively intransitive).

In both the economic and the political science literatures which I review, the preponderance of work is restricted to unidimensional competition. Typically, competition is assumed to take place on the unit interval. Thus, in the Downs model there is a single ideological continuum over which voters have (single-peaked) preferences and parties adopt locations.³ Parties, of which there are only two, are assumed to be vote maximizers. A voter supports the spatially closer party. This binary-contest specification, with vote maximizing and ‘sincere’ voter preference revelation, strains credulity when there are more than two electoral agents. Vote maximizing in multicandidate contests may not advance either the office or policy objectives of electoral agents. Sincere preference revelation may not advance voter objectives. Each of these issues will loom large in what follows.

With these preliminaries behind us, we may proceed to the crux of this volume. Part II examines the research on spatial economic and spatial political competition for a given set of producers and electoral agents, respectively. Part III explores the work that seeks to endogenize these sets of agents. Finally, in Part IV I discuss some special topics in multiparty electoral competition around which theoretical literatures have developed.

One final note: although this volume appears in a series on pure and applied economies, I would hope that interest in it spans the division between economics and political science. This poses a writing dilemma, since the readership is bound to be more heterogeneous than that for most research monographs in this series. I have tried to report theoretical results in a manner that is accessible to students of the subject, wherever their disciplinary home. This has unfortunately required occasional casualness of language and, with it, some lack of economy in argument for which I know no simple solution. For the reader bothered by these lapses, I suggest repairing to the original articles where ambiguity is the problem, and skimming where redundancy is objectionable.

Part II. Spatial Competition with Exogenous Agents

2. Spatial Economic Competition: Exogenous Agents

In the introductory section I have casually described the spatial framework. In...

Table of contents

• Cover Page
• Half Title page
• Title Page
• Copyright Page
• Original Title Page
• Original Copyright Page
• Contents
• Introduction to the Series
• Acknowledgements
• Dedication
• Part I. Introduction
• Part II. Spatial Competition with Exogenous Agents
• Part III. Spatial Competition with Endogenous Agents
• Part IV. Extensions and Conclusions
• Reference
• Index