Groups, Behaviors, and Issues as Cues of Partisan Attachments in the Public

• Author: Michael Barber,Jeremy Pope
• Published date: 01 September 2022
• Date: 01 September 2022
• DOI: http://doi.org/10.1177/1532673X221083831
• Subject Matter: Articles

Article

American Politics Research

2022, Vol. 50(5) 603–608

© The Author(s) 2022

Article reuse guidelines:

sagepub.com/journals-permissions

DOI: 10.1177/1532673X221083831

journals.sagepub.com/home/apr

Groups, Behaviors, and Issues as Cues of

Partisan Attachments in the Public

Michael Barber

1

and Jeremy Pope

1



Abstract

What factors do people most associate with the partisan identity of others: group identity, political issue positions, or social

behaviors? In this research note, we report the results of a conjoint experiment in which we test the predictive power of

descriptive identities against other attributes such as social behaviors and issue positions. We f‌ind that when presented with a

randomized biography to predict partisanship, people rely on issue positions over descriptive gro up identities or behaviors.

Most issues outperform group aff‌iliations and behaviors, with sexual orientation as the partial exception. We then compared

these results to the correlation between the same factors in respondents’own biographies and their own partisan identif‌ication.

We f‌ind that political issues are far less important to people’s own partisan aff‌iliations, while group identity is more predictive.

We conclude that an understanding or perception of ideological concepts and their association with the political parties in

others should be distinguished from adoption of such concepts by individuals themselves.

Keywords

partisanship, conjoint analysis, issues, group aff‌iliations

In contemporary America, people’s partisan aff‌iliations are

correlated with group aff‌iliations, personal behaviors, and

issue positions. However, among this expansive list, what

specif‌ic group identities, behaviors, and issues positions do

people most strongly associate with a person’s partisan

identity? In this note, we measure the correlation between a

host of potential cues and a person’s partisan aff‌iliation using

f‌irst a conjoint experiment that presents randomly generated

prof‌iles of individuals and asks respondents to guess this

person’s partisan aff‌iliation. Wethen compare this correlation

to the correlation between these same factors and the parti-

sanship of the respondents completing the survey. In this way,

we can measure both people’s perceptions of others’partisan

aff‌iliation as well as how their own partisan loyalties are

related to these factors.

Our results strongly suggest that the key factor people use

to draw inferences about partisanship (for both political

candidates and voters) is people’s political stances on various

issues. This suggests that while the public may be ideolog-

ically innocent in the sense of adopting a rigid framework of

issue positions (Campbell et al., 1960;Converse, 1964;

Kinder & Kalmoe, 2017), they are not innocent with respect

to understanding how those issue positions connect to par-

tisanship in the abstract.

1

This pattern is true despite the

importance of group aff‌iliations (Achen & Bartels, 2016;

Huddy et al., 2015) and behaviors associated with world

views (Hetherington & Weiler, 2018) in understanding

people’s actual partisan aff‌iliations. In other words, while

people understand the language of ideology, they are not apt

to adopt this language for their own views—instead relying

on their own shared group aff‌iliations and behaviors in their

partisan aff‌iliations.

There is growing evidence that citizens have some issue

commitments (Costa, forthcoming;Goggin et al., 2020;Orr

& Huber., 2019) that seem to heavily inf‌luence their ability to

draw inferences about partisanship, but views about how

much this matters vary. Mason (2018) claims that “More

often than not, citizens do not choose which party to support

based on policy opinion; they alter their policy opinion ac-

cording to which party they support (20-21).”In contrast,

Fowler (2020) argues that “voter decisions are inf‌luenced by

substantive considerations including policy preferences and .

. . we do not have any compelling evidence to support the

claim that a meaningful share of voters’decisions is inf‌lu-

enced by party identity or any other kind of identity”(p. 219,

emphasis added). Our contribution is twofold. First, most of

the research on this topic deals with how groups, issues, and

1

Department of Political Science, Brigham Young University, Provo, UT, USA

Corresponding Author:

Jeremy Pope, Department of Political Science, Brigham Young University,

745 Spencer W. Kimball Tower, Provo, UT 84602, USA.

Email: jpope@byu.edu