An Empirical Model of Issue Evolution and Partisan Realignment in a Multiparty System

• DOI: 10.1177/1065912917722234
• Date: 01 March 2018
• Published date: 01 March 2018
• Subject Matter: Articles

https://doi.org/10.1177/1065912917722234

Political Research Quarterly

2018, Vol. 71(1) 59 –74

© 2017 University of Utah

Reprints and permissions:

sagepub.com/journalsPermissions.nav

DOI: 10.1177/1065912917722234

journals.sagepub.com/home/prq

Article

Introduction

Since Carmines and Stimson’s (1986, 1989) seminal

work, the concept of issue evolution has become a com-

monly used theoretical toolkit to examine polarization

around cultural issues and partisan realignment in the

United States. Issue evolution occurs if a highly salient

new issue cross-cuts the existing party alignment and

then, after party elite polarization on the issue, the public

reorganizes itself around the new cleavage. Political

change and polarization in the United States since 1964

has followed an issue evolution process, with issues such

as abortion, civil rights, race, or religion leading to pro-

found realignments among American voters (Adams

1997; Carmines and Stimson 1986, 1989; Carmines and

Woods 2002; Carsey and Layman 2006; Layman 2001;

Lindaman and Haider-Markel 2002).

More recently, research has begun to apply the issue

evolution concept to political change outside the United

States (Stevens 2013 for the United Kingdom, Raymond

and Feltch 2014 for Chile, and Stimson, Thiébaut, and

Tiberj 2012 for France). Most studies of issue evolution

outside the United States, however, have been limited to

other two-party systems or the effect of Euroscepticism

on vote choice in European Parliament elections. De

Vries and Hobolt show that the impact of Euroscepticism

on vote choice in European Parliament elections followed

an issue evolution process as niche parties and losers of

previous elections acted as issue entrepreneurs (de Vries

and Hobolt 2012; Hobolt and de Vries 2015). Niche par-

ties in particular, such as nationalists, greens, or new right

parties, induced an elite-level polarization on European

integration, to which the mass public reacted by reward-

ing the niche parties’ more Eurosceptical platforms.

Nevertheless, only Stevens’s study provides a real

cross-temporal examination of issue evolution and

realignment using both micro-level data and data on elite

polarization outside the United States, which neverthe-

less excludes the Liberal Democrats and other third par-

ties. Yet, we lack a thorough application of the issue

evolution concept to a real multiparty system over time,

722234PRQXXX10.1177/1065912917722234Political Research QuarterlyArndt

research-article2017

1Aarhus University, Denmark

Corresponding Author:

Christoph Arndt, Department of Political Science, Aarhus University,

Bartholins Alle 7, 8000 Aarhus C, Denmark.

Email: arndt@ps.au.dk

An Empirical Model of Issue

Evolution and Partisan Realignment

in a Multiparty System

Christoph Arndt1

Abstract

Since Carmines and Stimson’s seminal work, the concept of issue evolution has become a common theoretical toolkit

to examine and explain polarization around cultural issues and partisan realignment in the United States. However,

very few studies outside the United States have applied the concept of issue evolution to explain electoral change and

realignment around new issues at national elections over time. Analyzing whether and how Carmines and Stimson’s

concept travels to electoral change in a multiparty system would provide more theoretical leverage and create

empirical knowledge on the logic of issue evolution outside the United States and also whether the logic of electoral

change differs from U.S.-style two-party systems. This article applies the issue evolution concept to a multidimensional

multiparty system using micro-level data from 1971 to 2011 in combination with data on elite-level polarization to

demonstrate that partisan realignment in Denmark follows an issue evolution process with niche parties as main

drivers. Moreover, niche party polarization on the new dimension of conflict has a stronger effect on vote choice than

mainstream party polarization.

Keywords

issue evolution, realignment, immigration, multiparty system, elite polarization, issue salience

60 Political Research Quarterly 71(1)

which incorporates both the role of mainstream and new

parties in the issue evolution process and their effects on

voting behavior and electoral change over time.

Consequently, this article applies the issue evolution

concept to a multidimensional multiparty system using

micro-level data from Denmark in the period 1971–2011

in combination with data on elite-level polarization.

Denmark represents a European multiparty system, where

cultural issues such as immigration, law and order, or the

environment became more salient at the expense of eco-

nomic issues (Arndt 2016; Green-Pedersen 2006, 2011;

Stubager 2010). Furthermore, the balance of power has

shifted substantially: until 1973, 90 percent of all govern-

ments were formed by the Social Democrats, while the

majority of governments afterward were formed by the

center-right. In this respect, my contribution is fourfold. I

first argue and demonstrate that only niche party polariza-

tion but not mainstream party polarization on the new

dimension of conflict had significant effects on vote

choice since 1971. Second, and in line with Carmines and

Stimson’s (1989) clarity and affect argument, I show that

issue preferences on the new dimension of conflict had a

stronger effect on party choice the more salient the new

dimension had become on the elite level. Third, this effect

is considerably stronger for the choice of niche parties

than for the choice of mainstream parties. Finally, models

of voters’ migration reveal that partisan realignment in

Denmark followed an issue evolution process, where

niche party polarization conditioned the effect of cultural

conservatism on realigning with a right-wing party. In

sum, the key finding is that issue evolution in a multi-

party system in the sense of Carmines and Stimson is not

driven by mainstream competitors, but by new and usu-

ally niche or challenger parties.

Issue Evolution and Partisan

Realignment in Multiparty Systems

Carmines and Stimson (1986, 1989) established the

sequence of issue evolution in the United States, which

starts with elite polarization on a new issue dimension

and ends with mass partisan realignment. Clarity and

affect provide two intervening steps in this causal chain.

Clarity means that partisans have become aware of the

increasing elite polarization on the new dimension. Affect

means the electorate has reacted to the increasing elite

polarization: “The public must not only perceive a differ-

ence in party issue stands, but must also care about this

difference” (Carmines and Stimson 1989, 161). If these

steps are fulfilled, voters’ mass realignment is the ulti-

mate consequence.

Carmines and Stimson’s sequence of issue evolution

has been widely applied in explaining political change in

the United States since the nomination of Barry Goldwater

as GOP candidate in 1964. Various contributions have

shown that cultural issues such as civil rights, race, or

religion have led to profound realignments among

American voters. After deliberate polarization of new

issues by political elites, the public showed strong reac-

tions and finally changed party affiliations, shifting the

balance of power to the benefit of the Republican Party in

the post-1968 party system (Adams 1997; Carmines and

Stimson 1986, 1989; Carsey and Layman 2006; Layman

2001; Lindaman and Haider-Markel 2002). At the same

time, Lindaman and Haider-Markel (2002) identified

examples in which elite polarization failed to produce an

issue evolution due to insufficient reactions at the mass

level. This relates to the nature of the issue as issues that

are “easy,” in the sense that voters can easily understand

their meaning, the elite framing around them, and the

conflicting positions, have the potential to induce an issue

evolution, whereas too technical issues will fail to do so

(Carmines 1991; Carmines and Stimson 1989, 11–12).

While Carmines and Stimson developed their theory

for a pure two-party system with the very rare emergence

and success of third parties or independents, an analysis

of issue evolution in a multiparty system must incorpo-

rate more than two parties. The most important difference

is the presence of niche (or new) parties who have differ-

ent incentives to compete and create polarization over

new issue dimensions.

The goal here is not to engage in an extensive debate

about niche party definition and conceptualization. Rather,

I follow one crucial defining criterion of niche parties,

namely, that they predominantly compete on a new issue

dimension neglected by their mainstream competitors

(Bischof 2017). This definition concerns green, national-

ist, new right, and some radical left parties for the purpose

of this paper. I further assume in line with Bischof’s (2017,

224) review of the literature that niche parties at least in

their early phase of existence attempt “to construct novel

conflict lines from the periphery of party systems on

issues with less competition,” and that they in contrast to

mainstream parties act as first movers.

In a two-party system, the losing mainstream party has

an incentive to compete on new issues, especially if it has

lost elections consecutively and it is clear that the current

alignments within the electorate provide a structural dis-

advantage for the party (Carmines 1991; Riker 1982). It

can therefore try to expand its electoral base by compet-

ing on new issues which potentially divide the opponent’s

electorate to regain votes and office. This is different

under multiparty competition, where mainstream parties

such as Conservatives, Christian Democrats, Liberals,

and Social Democrats can still govern as long as their

position on the established issue dimension permits coali-

tions with other mainstream parties, provided no party

has a structural majority in the electorate.