Non-Religious Identity Salience for Candidate Choice

• Published date: 01 May 2024
• DOI: http://doi.org/10.1177/1532673X241230053
• Author: Spencer Kiesel
• Date: 01 May 2024

Article

American Politics Research

2024, Vol. 52(3) 279–289

© The Author(s) 2024

Article reuse guidelines:

sagepub.com/journals-permissions

DOI: 10.1177/1532673X241230053

journals.sagepub.com/home/apr

Non-Religious Identity Salience for Candidate

Choice

Spencer Kiesel

1



Abstract

Religion is on the decline in the United States. Americans increasingly report low religiosity,have less attachment to religion, and

a rapidly growing number identify as nonreligious. In Congress, the story is different. While a quarter of the public identif‌ies as

nonreligious, only one member of Congress does. Why are the nonreligious vastly underrepresented in government? I use a

conjoint candidate choice experiment to causally link religious voters’bias against nonreligious candidates to reduced support

for them in electoral settings. I demonstrate that bias against the nonreligious affects electoral decisions and is causally linked to

the exclusion of the nonreligious from government. Furthermore, I show that nonreligious voters only exhibit in-group support

for candidates who explicitly identify as Atheists, not agnostic or candidates that merely lack a religious identity.

Keywords

religion and politics, conjoint, nonreligious

Introduction

The nonreligious: atheists, agnostics, and nones

1

comprise

between 20 and 25% of the population in the United States,

making them the second-largest religious group, behind

Christians, (Pew Research Center, 2019). This large and

rapidly growing group emerged in the 1990s as Christian

identif‌ication plummeted and Church attendance collapsed

(Pew Research Center, 2019). However, this enormous in-

crease in the number of nonreligious identif‌iers has not

translated into representation. The nonreligious are almost

totally excluded from government in the United States as only

one sitting off‌icial at the Federal level explicitly identif‌ies as

nonreligious (Sandstrom, 2019).

2

Nearly one quarter of the

population is nonreligious but only 1/535th of the present

Congress publicly identif‌ies as nonreligious. Here, a para-

digm contrary to the core tenets of our society exists. In a

democratic nation that often prides itself on representing

diverse constituencies descriptively, a large and rapidly

growing religious group is underrepresented by a factor

exceeding one hundred. Were this occurring with a racial

group or even a non-Christian minority religious group, such

glaring underrepresentation would result in condemnation

and panicked attempts to elect descriptive representatives.

However, no such concern exists for the unrepresented

nonreligious masses of America; the issue is barely known

about and scholars have only begun examining the issue as

one of representation.

Traditionally, legislative bodies and governments more

broadly have been dominated by men from dominant racial,

ethnic, and religious groups (Hughes, 2013). Descriptive

representation or the actual presence of minorities in legis-

lative bodies is a thriving area of research and a proven

method of minority protection. More recently, the importance

of having groups actually represented in legislatures by

members of their own has been widely embraced. There are

two reasons to prefer descriptive representation. The f‌irst

argument for descriptive representation is that substantive

policy outcomes may differ if previously excluded groups are

represented by their actual presence in policy making bodies.

For example there is signif‌icant evidence that substantive

policy gains are made by women via descriptive represen-

tation (Chattopadhyay & Duf‌lo, 2004) and evidence that this

argument holds for both religious (Tatari, 2010), ethnic

(Dunning & Nilekani, 2013), and racial (Cameron et al.,

1996) minorities. For nonreligious groups, these arguments

are theoretical and based on assumed beliefs in secular in-

stitutions because no empirical work has been done to date

(Yates, 2007). However, the theoretical argument is rather

1

UC Davis, Davis, CA, USA

Corresponding Author:

Spencer Kiesel, Department of Political Science, UC Davis, Kerr Hall, Davis,

CA 95616-5270, USA.

Email: skiesel@ucdavis.edu