American Attitudes Toward COVID-19: More Trumpism Than Partisanship

• Author: Robert Y. Shapiro,Neeraj Kaushal,Yao Lu,Jennifer So
• DOI: 10.1177/1532673X211046251
• Date: 01 January 2022
• Published date: 01 January 2022
• Subject Matter: Articles

2022, Vol. 50(1) 67 –82

https://doi.org/10.1177/1532673X211046251

American Politics Research

© The Author(s) 2021

Article reuse guidelines:

sagepub.com/journals-permissions

DOI: 10.1177/1532673X211046251

journals.sagepub.com/home/apr

Article

Introduction

President Trump and partisan conflict were front and center

as COVID-19 engulfed the United States—and the world.

The United States suffered worse than most other countries

from COVID-19 in terms of reported cases and mortality.1

Critics argue that this was at least partly due to President

Donald Trump’s leadership, calling COVID-19 “no worse

than seasonal flu,” and “a political conspiracy to destroy the

Trump presidency,” and his public mocking of individuals

who wore masks, which weakened public compliance of

measures that would have restricted COVID-19’s spread

(Coppins, 2020; Kristof, 2020; Krugman, 2020). Empirical

research emphasizes the role of rising political partisanship

on public attitudes and action concerning COVID-19 (Allcott

et al., 2020; Barrios & Hochberg, 2020; Bursztyn et al.,

2020; Gadarian et al., 2020; Milosh et al., 2020; Platzman &

Shapiro, 2020; Simonov et al., 2020).2 Republicans were

generally more cavalier about the pandemic, less likely to

practice social distancing and mask-use in public, and more

likely to oppose business closures and lockdowns that were

designed to contain the pandemic (Clinton et al., 2021; Fan

et al., 2020).3

President Trump’s response to the pandemic in public

debates was associated with political differences in attitudes

and perceptual biases—including the evoking of “motivated

reasoning” on steroids (e.g., Garrett, 2009; Redlawsk et al.,

2010; Taber & Lodge, 2006; Taber et al., 2009)—in how

each side saw the reality and threat of the outbreak, and how

the ideologically divided nation and states responded. Yet,

the question remains: what was the greatest cue to public

opinion and action on COVID-19: the party or the

president?

In this paper, we investigate how support for President

Donald Trump, beyond partisanship, guided Americans’ atti-

tudes toward COVID-19. This speaks to not just how

“Trumpism” and the Trump administration’s handling of the

pandemic influenced public attitudes but to the larger issue

of how Trump’s hold on voters within and beyond the

Republican Party provides evidence that leaders surpass the

role of parties in influencing public opinion. Our research

adds to the literature on the role of parties versus leaders in

American democracy (Barber & Pope, 2019; Key, 1961;

1046251APRXXX10.1177/1532673X211046251American Politics ResearchKaushal et al.

research-article2021

1Columbia University, New York, NY, USA

Corresponding Author:

Neeraj Kaushal, Columbia University, 1255 Amsterdam Avenue, New

York, NY 10027-6902, USA.

Email: nk464@columbia.edu

American Attitudes Toward COVID-19:

More Trumpism Than Partisanship

Neeraj Kaushal1, Yao Lu1, Robert Y. Shapiro1,

and Jennifer So1

Abstract

We investigate how support for President Donald Trump, beyond partisanship, guided Americans’ attitudes toward

COVID-19. This speaks to not just how “Trumpism” and the Trump administration’s handling of the pandemic influenced

public attitudes but to the larger issue of how Trump’s hold on voters within and beyond the Republican Party provides

further evidence that leaders surpass the role of parties in influencing public opinion. Using longitudinal data with individual

fixed-effects, we find that from the start of the pandemic, support for Trump above and beyond partisanship drove public

attitudes capturing skepticism toward COVID-19, fears of personal vulnerability, compliance with public-safety measures,

and viewing the pandemic in racist terms. Between March and August 2020, this gulf in attitudes between Trump voters and

non-supporters, and between Republicans and Democrats, widened; the widening was more pronounced between Trump

voters and non-supporters. Trump’s influence on Independents and non-voters also grew over the same period. While the

use of terms like “China virus” was related to partisanship and support for Trump, we find an increase in awareness across

groups that these terms were racist.

Keywords

partisanship, political polarization, racism, COVID-19

68 American Politics Research 50(1)

2 American Politics Research 00(0)

Page & Shapiro, 1992; Page et al., 1987; Zaller, 1992), yet

differs in an important respect from previous research in that

we investigate public attitudes on COVID-19 related issues.

These issues are new and, a-priori, non-ideological in that

they do not conform to the core ideology of the Republican

Party. We investigate how Trump’s personal views and rheto-

ric, that quickly became partisan, impacted public opinion on

COVID-19 among Trump supporters, Republicans,

Democrats, and Independents. We argue that because

Trump’s views and rhetoric on COVID-19 do not emanate

from the core ideology of the Republican party, its impact

would not necessarily be limited to Republicans, but may be

greater among Trump’s non-Republican supporters. Further,

Trump’s rhetoric to minimize the dangers of the pandemic

grew louder as the election came closer and COVID-19

emerged as an electoral threat to the President. Whether its

impact increased is an empirical issue that we investigate in

the paper.

We use two waves of a longitudinal survey to investigate,

first, how did public attitudes (skepticism, fears, prejudices,

and response) toward COVID-19 change among individuals

with different political affiliations between March and

August as the pandemic spread across the country? Did the

partisan and Trump divide in attitudes widen or narrow?

Second, if the political influences were dominant, was the

conflict found in public opinion toward COVID-19 driven

primarily by partisanship or did Trump evoke and amplify

them further?

We define partisanship and support for Trump as follows:

(i) self-identified party identification as Republican (includ-

ing Independent “leaners”), Democrat (also including “lean-

ers”), and Independent/other (hereafter Independents); and

(ii) self-reported Trump voters in 2016, voters for candidates

other than Trump, and non-voters. For ease of exposition,

hereafter, we use the term Trump supporters to describe

Trump voters in 2016; Trump opponents to describe those

who voted for other candidates in 2016. In the empirical

analysis, we estimate multivariate statistical models using

partisan affiliation versus support for Trump separately.

Further, to examine the extent to which “Trumpism,” so to

speak, increased political conflict in the U.S., we create a

third measure that combines the first two measures in five

categories: Republican Trump supporters; Republican Trump

opponents; Non-Republican (mostly Independent) Trump

supporters; Democrat Trump opponents and Independent

Trump opponents.

Note that our measures of partisanship and Trumpism are

based on data that were collected before the first survey

wave. These data are part of the information that YouGov,

our survey agency, routinely collects on the survey panel.

This reduces the possibility of endogeneity between our

measures of partisanship and Trumpism and public attitudes

toward COVID-19. Further, use of longitudinal data with

individual fixed effects strengthens causal interpretations.

To preview, as expected based on extant research, we find

that from the start of the pandemic, public attitudes toward

COVID-19 were deeply divided along party lines as well as

among Trump supporters and opponents. These gaps wid-

ened over the next 5 months. Importantly, at the start of the

pandemic, non-Republican (mostly Independents) Trump

supporters were generally more skeptical of the pandemic,

less likely to practice social distancing, and more racist than

Republican Trump supporters, suggesting that the reach of

Trumpism on COVID-related attitudes advocated by

President Trump went beyond the partisan divide. Over the

5 months that followed, the skepticism toward COVID-19

increased among all groups, including Independents, with

the exception of Democrats Trump opponents. Such an

increase in political conflict exacerbated the already difficult

political environment for managing the COVID-19 public

health crisis.

Parties, Leaders, and Democracy

Studying public opinion is central to the study of democracy

because of how the public can influence what governments

do directly or indirectly through the election of its leaders

who are held accountable at election time (see Shapiro,

2011). The public, however, is not autonomous and all-

knowing, and it depends heavily on having information and

guidance available. Leaders, broadly defined to include those

inside and outside of government positions and party poli-

tics, are central to this, and what they say and do is commu-

nicated to the public through the mass media (see Key, 1961;

Page & Shapiro, 1992; Page et al., 1987; Zaller, 1992).

There is tension, however, in emphasizing the effect of

formal political leadership, especially persuasive leadership,

on public opinion. That the public can be influenced by lead-

ers is seemingly at odds with the expectation for democratic

accountability: that leaders should be responding to the pub-

lic and not the reverse (see Lenz, 2012). Whether leaders

should lead or follow public opinion is a dilemma, and how

this plays out can vary by issues and political context (cf.

Shapiro, 2011). But there is no getting around that the public

needs and depends on information and guidance from leaders

(Key, 1961).

This dependence on leadership, however, does not mean

that the public as a whole is easily persuaded or led by politi-

cal leaders, with presidents as prominent ones. Efforts by

presidents to change public opinion have historically had

limited persuasive effects (e.g., Edwards, 2003) or depended

on the support and trust presidents have had at particular

times (see Page & Shapiro, 1984, 1992). On the other hand,

while leaders may not affect the public as a whole there are

theoretical reasons and increasing evidence that partisan or

ideological leaders can influence partisan subgroups. This is

consistent with Zaller’s (1992) two-message theoretical

model, which postulates that when partisan leaders or other