Party–Group Ambivalence and Voter Loyalty: Results From Three Experiments

• Author: Gregory Shufeldt
• DOI: 10.1177/1532673X17705854
• Published date: 01 January 2018
• Date: 01 January 2018
• Subject Matter: Articles

https://doi.org/10.1177/1532673X17705854

American Politics Research

2018, Vol. 46(1) 132 –168

© The Author(s) 2017

Reprints and permissions:

sagepub.com/journalsPermissions.nav

DOI: 10.1177/1532673X17705854

journals.sagepub.com/home/apr

Article

Party–Group

Ambivalence and Voter

Loyalty: Results From

Three Experiments

Gregory Shufeldt1

Abstract

Renewed emphasis on the group-based nature of political parties makes

understanding the relationship between partisan and group identities

essential. How do citizens respond to the internal disconnect between their

partisan identity and their other politically salient identities? In addition,

do differences in the group-based nature of each party lead to asymmetric

effects of party–group ambivalence? Using data from an original survey

experiment across three samples—the 2012 Cooperative Congressional

Election Study, the 2012 Convention Delegate Study, and a 2015 sample from

Amazon Mechanical Turk—I find that party–group ambivalence diminishes

party loyalty, making respondents less likely to vote for, contribute to, or

volunteer for their political party’s candidate. Moreover, the strength of this

impact is consistently larger among Republican identifiers than Democratic

identifiers. These results suggest that party asymmetry in party coalitions

may have an impact on party loyalty.

Keywords

ambivalence, political parties, voting behavior

1Butler University, Indianapolis, IN, USA

Corresponding Author:

Gregory Shufeldt, Butler University, 4600 Sunset Ave., Jordan Hall 345, Indianapolis,

IN 46208, USA.

Email: gshufeld@butler.edu

705854APRXXX10.1177/1532673X17705854American Politics ResearchShufeldt

research-article2017

Shufeldt 133

The group-based nature of political parties makes understanding the relation-

ship between party and group essential. Yet, as renewed emphasis is placed

on the group-based nature of political parties, there remains a need to test

how citizens respond to ambivalence between their partisan identity and

other nonpartisan, politically salient identities, such as ideology or social

identity. How does party–group ambivalence, a disconnect between party

and group evaluations, affect partisan political behavior?

Moreover, recent research has pointed to asymmetries between the parties

in the role that groups play (Shufeldt, 2014; Grossman & Hopkins, 2015,

2016). If Democrats are a party that promotes group benefits while

Republicans promote ideological goals, it stands to reason that party–group

ambivalence may affect the parties differently. The standard conventional

wisdom has long posited Democrats to be the disorganized, fractured party

with Republicans a more disciplined, homogeneous party. Recent Republican

turmoil is particularly newsworthy because it flies in the face of what we

have come to expect. Paul Waldman (2014) wrote in The American Prospect:

For the longest time, Democrats were the party of infighting and disunity,

whose squabbling never failed to find its way into the news. It’s a grim inside

joke among liberals that the most common headline in the political media is

“Democrats in Disarray.” But it hasn’t been that way for a while. In fact,

perhaps the most important political dynamic of the current era is the conflict

within the previously monolithic Republican party.

To test whether party–group ambivalence has asymmetric effects for

Democratic and Republican identifiers, I conducted an original experiment

across three samples: the 2012 Cooperative Congressional Election Study

(CCES), the 2012 Convention Delegate Study (CDS), and a 2015 sample

recruited from Amazon Mechanical Turk (mTurk). In the experiment, voters

are exposed to fictitious candidate vignettes where the endorsement of their

political party and political groups most important to them are varied ran-

domly. Respondents then are asked about their likelihood of voting for, con-

tributing to, and volunteering for, either candidate.

The findings suggest that the relationship between partisan and group iden-

tities can strengthen party loyalty. Reinforcing cues maximize support as

respondents demonstrate more party loyalty by becoming more likely to vote

for, contribute to, and volunteer for their party. However, when voters experi-

ence ambivalence between their partisanship and another politically salient

group identity, partisan support is diminished. Explicit conflict between cues

and the absence of reinforcing cues have a more pronounced effect among

Republican identifiers—diminishing party loyalty. Taken together, these results

134 American Politics Research 46(1)

suggest higher rates of party–group ambivalence may dampen any perceived

advantage Republicans might have once enjoyed in party loyalty. This provides

support for the idea that Republican appeals based on broad ideological goals

at the expense of emphasizing group benefits is a sound electoral strategy.

Party–Group Ambivalence and Partisan Behavior

Our political identities are the product of multiple competing identities. As peo-

ple identify with these various groups, they form an in-group versus out-group

bias, or more simply an “us versus them” bias (Kinder & Kam, 2009; Tajfel &

Turner, 1979; Turner, Hogg, Oakes, Reicher, & Wetherell, 1987). As competi-

tion between the parties grows fiercer, this us versus them dynamic drives higher

levels of polarization, animosity, and bias directed at members of the opposing

party (Iyengar, Sood, & Lelkes, 2012; Iyengar & Westwood, 2015). In spite of

the strength of party identification, many voters still are able to identify both

positive and negative aspects of their party (Basinger & Lavine, 2005). Voters

are frequently ambivalent, defined as “an individual’s endorsement of compet-

ing considerations relevant to evaluating an attitude object” (Lavine, 2001, p.

915), between the parties. More specifically, partisan ambivalence is defined as

a disconnect between party identification and party evaluations (Lavine,

Johnston, & Steenbergen, 2012). Ambivalent partisans are the voters who expe-

rience this disconnect. Univalent partisans are voters who do not experience

ambivalence between the parties. More recent studies in political psychology

have sought to bring attention back to ambivalence (e.g., Greene, 2005;

Groenendyk, 2013; Lavine et al., 2012), arguing that ambivalence has a clear

effect on political attitudes (Alvarez & Brehm, 2002) and political behavior

(Basinger & Lavine, 2005; see Thornton, 2013, for a recent summary).1

Here is where the work of Liliana Mason and colleagues has been most

illuminating. Her work on political identity alignment has found that “cor-

rectly sorted” partisans are more likely to be motivated to engage in political

participation, will sense larger differences between the parties, and have

lower levels of tolerance than partisans who are incorrectly sorted (Huddy,

Mason, & Aaroe, 2015; Huddy, Mason, & Horwitz, 2016; L. Mason, 2013,

2015, 2016). When identities are correctly aligned, in-group bias and out-

group intolerance are heightened (Roccas & Brewer, 2002). As a result,

stronger partisan identifiers are driven to participate via enthusiasm for their

own groups and anger directed at opposing groups (Groenendyk & Banks,

2014; P. R. Miller & Conover, 2015). This body of work largely confirms the

early research finding of how voters respond to cross-pressure (Berelson,

Lazarsfeld, & McPhee, 1954; A. Campbell, Converse, Miller, & Stokes,

1960).