Voting behavior under proportional representation
| Author | Peter Buisseret,Carlo Prato |
| Date | 01 January 2020 |
| Published date | 01 January 2020 |
| DOI | 10.1177/0951629819892342 |
| Subject Matter | Articles |
Article
Journal of Theoretical Politics
2020, Vol.32(1) 96–111
ÓThe Author(s) 2020
Article reuse guidelines:
sagepub.com/journals-permissions
DOI: 10.1177/0951629819892342
journals.sagepub.com/home/jtp
Voting behavior under
proportional representation
Peter Buisseret
Harris School of Public Policy, University of Chicago, Chicago, IL, USA
Carlo Prato
Department of PoliticalScience, Columbia University, NewYork, NY, USA
Abstract
Understanding voter preferences in proportional representation (PR) systems is an essential first
step towards understanding candidate selection and behavior. We unearth conceptual challenges
of constructing a theory of sincere voting in PR elections. Using a rich but tractable framework,
we propose a foundational theory of voting behavior in list PR systems. We show how voters’
expectations about the election outcome shape the relative salience of different candidates on a
party’s ballot and ultimately determine their induced preferences over a party’s list. Our frame-
work shows how voters allocate their attentionacross different ranks on a party’s ballot, and how
this depends on each party’s relative popularity.
Keywords
Proportional representation;voting behavior; political preferences
1. Introduction
Proportional representation (PR) systems account for about two thirds of the
world’s legislatures. Nonetheless, a decisive obstacle to studying representation
under these systems has been the difficulty of properly understanding the electoral
incentives faced by candidates and political parties. A key issue is that a description
of the ‘electoral connection’ (Mayhew, 1974) between voters and politicians
requires a clear theory of how voters formulate preferences over different political
Corresponding author:
Carlo Prato, Columbia University, 420 West 118th Street, New York, NY 10027, USA.
Email: cp2928@columbia.edu
alternatives, e.g., party lists and possibly the candidates within those lists, and how
these preferences are shaped by the actions and characteristics of the candidates
competing for office.
In this article, we propose a tractable model of instrumental voting under PR
systems that (i) clarifies the conceptual challenges of extending the notion of sin-
cere voting from single-member districts to multi-member districts, (ii) develops an
analytical framework to characterize voters’ preferences over different party lists
and candidates within lists, and (iii) generates testable implications about the rela-
tionship between candidates’ attributes and electoral outcomes.
Despite the existence of a large and influential formal theoretical literature on PR,
in almost all models voters’ preferences over lists are driven by party-level characteris-
tics such as the party’s location on a left–right scale (Austen-Smith and Banks, 1988;
Matakos et al., 2016) or the party’s policy commitments over public good provision
and/or redistribution (Austen-Smith, 2000; Iversen and Soskice, 2006; Lizzeri and
Persico, 2001; Persson and Tabellini, 2016). A handful of studies endogenize the con-
nection between these party-level outcomes and the behavior of individual legislators
(Krasa and Polborn, 2018; Kselman, 2020). To the best of the authors’ knowledge,
however, there is no tractable model that derives voters’ induced preferences over lists
from the attributes of individual candidates within party lists (for a model where vot-
ers’ preferences are driven by representatives’ behavior, see Buisseret and Prato, 2018).
1.1 Statement of the problem
Consider a geographic constituency represented by two members in a national leg-
islative body, elected according to a closed-list PR system. Suppose, moreover, that
a voter is deciding whether to cast her ballot for one of two parties, Aand B; this
voter has preferences over the individual candidates on the parties’ lists. Letting i
denote a voter i’s strict preference, suppose that the following holds:
First-ranked candidateon party A0s list
iSecond-ranked candidateon party B0s list
iFirst-ranked candidateon party B0s list
iSecond-ranked candidateon party A0s list
ð1Þ
How should this voter cast her ballot? Under a proportional rule system, party
Acould win zero, one, or two seats. A voter’s evaluation of which of these circum-
stances is relatively more likely is a crucial determinant of her voting decision.
Suppose that she thinks party Ais likely to do very badly in the election, so that it
is near-certain that the second-ranked candidate on party A’s list will fail to win a seat.
In fact, what is more likely to be at stake for party Aiswhetheritwinsasingleseat
versus no seats. Equivalently, in this situation party Bis likely to do very well in the
election: because its first-ranked candidate is nearly certain to be elected, what is more
likely to be at stake for party Bis whether it wins two seats versus a single seat.
In this scenario, the voter deciding between party Aand party Bshould weigh
the relative value of electing the first-ranked candidate on party A’s list, versus the
Buisseret and Prato 97
To continue reading
Request your trial