Technology Threats to Employment, Issues, and Candidate and Party Preferences in the United States

• Author: Tobias Heinrich,Christopher Witko
• DOI: http://doi.org/10.1177/10659129211027658
• Published date: 01 September 2022
• Date: 01 September 2022
• Subject Matter: Articles

https://doi.org/10.1177/10659129211027658

Political Research Quarterly

© 2021 University of Utah

Article reuse guidelines:

sagepub.com/journals-permissions

DOI: 10.1177/10659129211027658

journals.sagepub.com/home/prq

Across affluent democracies, technologies, like automa-

tion and artificial intelligence, are imperiling many jobs

(Acemoglu and Restrepo 2020; Brynjolfsson and McAfee

2014; Haldane 2015; Muro 2016), nearly 50 percent of in

the United States, according to one estimate (Frey and

Osborne 2017).1 It is not surprising, then, that some stud-

ies find that greater vulnerability to technology-induced

job loss, or “TIJL,” is associated with candidate and party

preferences (Anelli, Colantone, and Stanig 2019; Frey,

Berger, and Chen 2018; Gingrich 2019; Im et al. 2019;

Thewissen and Rueda 2017). We still lack a good under-

standing of whether these associations are causal and

what exactly drives them, however.

How the public responds to TIJL is an important ques-

tion. Many studies examine how labor market threats

shape policy preferences and voting (Hacker, Rehm, and

Schlesinger 2013; Iversen 2001; Rehm 2011, 2016; Singer

2011; Wright 2012). In Europe, those with high exposure

to TIJL prefer mainstream left parties (Gingrich 2019) or

populist and far-right parties (Anelli, Colantone, and

Stanig 2019; Gingrich 2019; Im et al. 2019; Kurer 2020).

Frey, Berger, and Chen (2018) find that U.S. areas with

more workplace technology penetration disproportionately

supported Republican Donald Trump for president in

2016. Heinrich and Witko (2021) report that Europeans

who think technology can replace them prefer that the

government addresses economic over noneconomic

issues. Given this, it seems probable that individuals

exposed to technology will be more supportive of candi-

dates that prioritize the economy, unemployment, and

welfare state programs in their current campaigns

(Hacker, Rehm, and Schlesinger 2013; Singer 2011), or

candidates from parties with reputations for doing so

(Hibbs 1977; Wright 2012). Does the support for main-

stream left or right populist and far-right parties among

those exposed to technology reflect these parties’ empha-

sis on issues prioritized by vulnerable workers?

Most of the evidence of TIJL shaping priorities and

voting comes from observational studies in Europe, and

1027658PRQXXX10.1177/10659129211027658Political Research QuarterlyHeinrich and Witko

research-article2021

1University of South Carolina, Columbia, USA

2The Pennsylvania State University, University Park, USA

Corresponding Author:

Tobias Heinrich, Department of Political Science, University of South

Carolina, 817 Henderson Street, Columbia, SC 29208, USA.

Email: heinrict@mailbox.sc.edu

Technology Threats to Employment,

Issues, and Candidate and Party

Preferences in the United States

Tobias Heinrich1 and Christopher Witko2

Abstract

Research indicates that susceptibility to having one’s job replaced by technology is associated with candidate and

party preferences in affluent democracies, but there is little understanding of why. We investigate whether workers

exposed to technology are more supportive of candidates and parties that prioritize the economy, unemployment, and

welfare state programs, currently or with historical reputations for doing so. In the United States which we examine

here, however, technology threatening jobs has not been politically salient and partisan attachments are strong,

which could limit issue-based conversions in candidate preferences. To examine these possibilities, we use survey

experiments randomizing exposure to information about the individual-level threat of losing one’s job to technology

and hypothetical candidate issue priorities in experiments with and without party labels. We find that there are

some differences in “issue premiums” among high-exposure individuals in the party labels experiment, but whether

individuals are made aware of their exposure does not explain any issue-based variation in candidate preferences.

We also find that when made aware of their exposure to technology, high- and low-exposure Republicans and low-

exposure independents become slightly more supportive of Republican candidates.

Keywords

technology, unemployment, automation, voting, elections, parties

2022, Vol. 75(3) 797–811

Article