Harming and Shaming through Naming: Examining Why Calling the Coronavirus the “COVID-19 Virus,” Not the “Chinese Virus,” Matters

ABSTRACT

In an effort to stave off racial and geographic prejudices, The World Health Organization (WHO) has advised against naming pathogens for geographic regions. Despite this, some media reports, and prominent politicians, have persisted in calling the coronavirus the “Chinese Virus,” or some variant thereof. In this pre-registered online experiment (N = 614), we examined the effect media frames and individual factors (i.e., political affiliation and party) had on perceptions of COVID-19, prejudice against Chinese- and Asian-Americans, and attributing blame to China for the coronavirus. We found that articles that labeled the coronavirus the “Chinese Virus” (versus the “COVID-19 Virus”) were perceived less favorably by Democrats and liberals, in comparison to Republicans and conservatives, who were also more likely to express greater racial prejudice against Chinese- and Asian-Americans and to blame China for the pandemic. Findings from this study suggest that amongst a host of other variables, media framing has an effect on the public’s attitudes and feelings of blame for the pandemic.

KEYWORDS:

• COVID-19
• framing
• China
• Chinese-Americans
• “Chinese virus”
• “Wuhan virus”
• “Kung flu”
• prejudice

On several levels, the coronavirus could be the most vexing problem to the hit the world in recent memory. Health officials scrambled to find ways to treat patients in the first global pandemic to hit the world in more than a century. Businesses closed in the wake of both employee shortages and fears of becoming incubators for the contagion; universities, churches, and schools followed suit (Fox, 2021; Lumpkin & Svrluga, 2020). For media, saddled with the task of communicating information to help people make sense of this new reality, the problem was particularly arduous.

To many, media serve as their veritable window onto the world (Gerbner, Gross, Morgan, Signorelli, & Shanahan, 2002). Thus, the way media frame, depict, and describe events can have a profound influence on the public’s perception of the world, people, and events (Bolsen, Palm, & Kingsland, 2020; Entman, 1993; Gerbner et al., 2002; Ogbodo et al., 2020). In this study we examine if media sources that call the coronavirus the “COVID-19 virus” evoke vastly different responses than those that call it the “Chinese virus,” which has been associated with antagonisms toward persons of Asian descent (Liao, 2020; Yang, 2020). Media messages matter, in that tiny nuances, turns of phrases, or the ways media couch, emphasize, or even negate certain aspects of a story, can shape how individuals feel about an issue (Entman, 1993; Iyengar, 1990; Scheufele, 2000). These connotations undoubtedly played a role in World Health Organization’s Assistant Director-General for Health, Security, and Environment, Dr. Keiji Fukada, underscoring the organization’s previous directive against naming pathogens for geographic regions (WHO, 2015). In reinforcing the call against such behaviors, the WHO stated:

We have seen certain disease names provoke a backlash against members of particular religious or ethnic communities, create unjustified barriers to travel, commerce and trade, and trigger needless slaughtering of food animals. This can have serious consequences for peoples’ lives and livelihoods. (Para 2)

As a result, the WHO warned against labels connecting viruses to geographic regions or groups and officially named the coronavirus disease “COVID-19” (WHO, 2020).

Not all persons are affected by media equally. Different news sources could frame information about the coronavirus very differently, potentially leading to very different beliefs about the coronavirus. The traits of the person receiving the message also matters. The same message could have very different effects, depending on the traits (e.g., political orientation, race, preexisting beliefs) of the recipient (Holt, 2018; Holt, Ellithorpe, & Ralston, 2017). These factors could be partly why many are responding to news during this pandemic in very different ways. Some, perhaps spurred by media broadcasting messages from political leaders who claimed the coronavirus was a “hoax,” a “conspiracy,” or that hydroxychloroquine or bleach could somehow kill the coronavirus, act in in ways that align with the beliefs of the source speaking to the media. In Nigeria, for example, hospitals reported an increase in overdose cases involving hydroxychloroquine after former U.S. President Donald Trump said that the malaria drug could somehow treat or cure COVID-19 (Ogbodo et al., 2020). Conversely, others, perhaps in response to scientists whom they believe are credible, became fully vaccinated, received booster shots, engaged in social distancing, wore masks, and stayed home to stop the spread of the coronavirus.

Although it is impossible to draw a direct connection between media accounts and action, media do wield considerable influence, especially during times of crises (Ogbodo et al., 2020; Van Bavel et al., 2020). Research on framing and blame attribution, for example, finds that exposure to conspiracy theories and misinformation linking China to the spread and creation of the coronavirus, is associated with an increase in anti-Chinese sentiment and xenophobia (Dhanani & Franz, 2021). During the pandemic, violent crimes against Asians have increased dramatically. The advocacy group Stop AAPI Hate said it received more than 2,800 reports of hate incidents against Asian Americans and Pacific Islanders nationwide in 2020 (Cabral, 2021). New York City alone saw a nine-fold increase in crimes against Asians and Asian Americans in 2020 (Cabral, 2021). These facts, scholars say, are fanned by some U.S. media reports linking China with the coronavirus and some reports that characterized China – and even American-born Chinese – as dirty and living dangerous lifestyles (Bolsen et al., 2020). Discrimination against Asians is also on the rise. Across New York state, unemployment for Asians has risen 6,900%, compared to 1,890% for Whites due to widespread fear of interacting with Asians (Liao, 2020).

These responses to persons of Asian descent beg several questions: How did we get here? How are people, presumably moored in their homes with primarily nothing but media as their main source of information, affected in such different ways by news about the same reality? In this study, we expand the domain of framing theory and apply it to an important health-related topic that is very germane today. Specifically, we examine what effect media messages have on perceptions of COVID-19 to see if the effects of these messages vary based on the traits of the individual receiving the message, and on how the message is framed by the media.

Theoretical Foundation: framing

One theory that could help provide answers to these important questions is framing. To frame is to, “select some aspects of a perceived reality and make them more salient in a communicating text, in such a way as to promote a particular problem definition, causal interpretation, moral evaluation and/or treatment recommendation for the item being described” (Entman, 1993, p. 52). Frames acknowledge the ability of media messages to define a situation (Ogbodo et al., 2020). Specifically, media not only establish the issue in the minds of the public, but they also determine the parameters of what is acceptable and unacceptable in terms of the way it is discussed and should be handled. As one scholar observed, media frames provide an “interpretive storyline that sets a specific train of thought in motion, communicating why an issue might be a problem, or what might be responsible for it, and what should be done about it” (Nisbet, 2009, p. 15).

Frames can just as easily be defined by what they exclude as by what they include (Entman, 1993). The emphasis on certain aspects of an issue – at the expense of others – can exert tremendous influence and can push opinions in the direction of the emphasized frame (Bolsen & Druckman, 2018; Bolsen, Palm, & Kingsland, 2019; Bolsen & Shapiro, 2018). In the present context, the media’s framing of COVID-19 has been shown to exert a powerful influence on its audience, as well as considerable “downstream effects” on beliefs about the coronavirus (Bolsen et al., 2020). Information about the origins of COVID-19, whether scientific or conspiratorial, are transmitted through media frames and can influence beliefs, actions, and perceptions about the coronavirus (Bolsen et al., 2020).

Individual Attitudes and Receptiveness to Particular Frames

Knowingly or unknowingly, media frames have helped the public construct their perception of reality regarding COVID-19 (Ogbodo et al., 2020). Moreover, messages that accentuate the public or personal health benefits of behaviors that stop the spread of the coronavirus have been shown to increase individuals’ intentions to engage in such behaviors (Jordan, Yoeli, & Rand, 2020). Conversely, exposure to conspiratorial claims have been shown to result in behaviors that are less likely to reduce the spread, including individuals being less likely to become vaccinated, wash their hands, engage in social distancing, and wear masks in public (Bolsen et al., 2020). Scholarly evidence shows that the way media present crises such as COVID-19, goes a long way in determining how people react to it (Ogbodo et al., 2020).

The frames media choose, and the issues to which viewers are attuned, can also influence perception. One study, for example, found that individuals gave more weight to aspects of the frame made most salient when forming their opinion (Chong & Druckman, 2007). Which considerations are taken into account, and which will be ignored, depends on how accessible, or easily understood the information being emphasized is to the perceiver (Iyengar, 1990). Exposure to information stating that global warming is a hoax, for example, has been found to decrease individuals’ willingness to engage in behaviors that will reduce their carbon footprint (Jolley & Douglas, 2014; van der Linden, 2015). Similarly, exposure to media linking the coronavirus with China has been shown to increase negative attitudes toward Asian Americans, a belief that Americans should be prioritized over immigrants, and increased feelings of xenophobia in general (Dhanani & Franz, 2021).

Although a direct connection between messaging and action cannot be established, research on framing effects shows that media do affect public perception based on how media contextualize an issue. Individuals cannot be attuned to all media at all times. Thus, decisions are made by individuals as to what media they will select, and ultimately, what media frames will have the potential to influence their interpretations of reality. To test the assumption that framing the coronavirus as the “COVID-19 virus,” rather than the “Chinese virus” influences attitudes, we test the following hypotheses:

H1: Participants who read the “Chinese virus” article will have more prejudiced attitudes toward Chinese- and Asian-Americans than will participants who read the “COVID-19 virus” article.

Scholars have long known that media do not tell people what to think, but mainly what to think about (Cohen, 1963), which reaffirms earlier work showing that audiences are prone to interpret media stories along the lines of their preexisting biases, especially in relation to racial groups (Holt et al., 2017). Moreover, in today’s environment of increasingly negative news stories, audience members favor information about their own group, and oppose information about other groups (Holt & Carnahan, 2020). In short, it might not be media that helped generate negative attitudes about Chinese- and Asian-Americans; those with such beliefs might have been attuned to such beliefs prior to the pandemic and the sources of media and media frames were just a byproduct of these beliefs.

Political Attitudes and Media Orientation

It is also our assertion that political affiliation and attitudes also matters. Being a Republican and having politically conservative views has been correlated with higher levels of anti-Asian sentiment (Tan, Lee, & Ruppanner, 2021). Also, conservative media outlets uptick in calling the coronavirus he “China virus” has also shown to be correlated with an increased beliefs that Asians are anti-American (Darling-Hammond et al., 2020). Thus, to test the influence party affiliations and political leanings have on perception, we test the following hypotheses:

H2: Republicans will have more prejudiced attitudes toward Chinese- and Asian-Americans than will Democrats.

H3: The more conservative people are, the stronger will be their prejudiced attitudes toward Chinese- and Asian-Americans.

H4: The “Chinese virus” article will have a stronger effect on prejudiced attitudes toward Chinese- and Asian-Americans among Republicans than among Democrats.

H5: The “Chinese virus” article will have a stronger effect on prejudiced attitudes toward Chinese- and Asian-Americans among more conservative people.

We also examined the effects of stories that pair the coronavirus with China will have on feelings of blame for the pandemic and predict that media frames will have an influence on the public’s perception in general, but these effects will be more pronounced for certain groups than others. Thus, we also hypothesize that:

H6: Participants who read the “Chinese virus” article will be more likely to blame China for the pandemic than will people who read the “COVID-19 virus” article.

Previous research has shown that in an era of almost endless media choices, the news individuals choose is often what they find most interesting and/or appealing to the groups to which they most identify, especially when that news is negative (Holt & Carnahan, 2020). Given how politicized the handling of the coronavirus has become, perhaps not the least of these influencers is political orientation. Frequently, the sources that have capitalized on these fears are far-right conservative and extremist groups (Ogbodo et al., 2020).

During periods of great uncertainly – in which the pandemic could undoubtedly qualify – research has shown that the America public often rely on political leaders for guidance and information (Shao & Hao, 2020). Moreover, messages from sources perceived as being credible tend to be more accepted by individuals than messages from sources perceived as non-credible (Druckman, 2001). As undoubtedly the most-covered news source in the U.S., and given his popularity amongst conservatives and Republicans, labels from the president could be more resonant with some parts of the public and could have an influence on his followers’ perception of COVID-19.

Despite measures to prevent labels connecting blame for the coronavirus with China, conservative leaders have still insisted on labeling the “COVID-19 virus,” the “Chinese virus,” the “Wuhan virus,” “Kung flu,” or some variant thereof (Riechmann & Tang, 2020). Often, it is not just what is said, but how it is said, and by whom, that matters (Druckman, 2001; Merolla, Pantoja, Cargile, & Mora, 2012; Ogbodo et al., 2020). The ubiquity and prominence of former U.S. President Trump has been shown to override even expertise on COVID-19. Moreover, as the most prominent member of the Republican Party, and the face of the “new conservative movement” (Alberta, 2017; Obeidallah, 2021), this framing of the virus by officials some persons deem credible, and the media rebroadcasting such terms, could have some influence on public attitudes, especially for Republicans and conservatives.

Although at least half of Americans do not support the Republicans or conservative ideology, those who do appear to be ardent supporters as the divide between the two major political parties appears to be widening. Some studies have found the political divide between political parties is the greatest it has been since 1952, with voters no longer voting for candidates as much as they are voting against the other side (Abramowitz & Webster, 2015). Frames are especially potent when they mesh with the existing schema and beliefs of the recipient (Holt et al., 2017; Shoemaker & Reese, 1996). As several scholars have noted, public acceptance of conspiracy narratives can be harmful, not only because they can lead people to dismiss credible evidence, but also because they can also lead them to engage in risky and potentially dangerous behaviors (Lewandowski, Oberauer, & Cignac, 2013; Oliver & Wood, 2014; Uscinski, Douglas, & Lewandowski, 2017).

However, attenuation to media frames could also operate differently based on political ideology. Persons who are liberal might find that information to be untrustworthy and do just the opposite of the recommendations that are given. Notably, being liberal has been shown to be the strongest predictor of attitudes about COVID-19, according to recent studies. U.S. counties who voted Democrat in the 2016 presidential election exhibited 14% more social distancing than those who voted Republican (Gollwitzer et al., 2020). Other research has shown conservatism to be associated with less perceived personal vulnerability to the coronavirus, and a greater belief that the media exaggerated the impact of the coronavirus, and the belief that the spread of the coronavirus is a grand conspiracy (Calvillo, Ross, Garcia, Smelter, & Rutchick, 2020).

We postulate that political attitudes have influenced attitudes toward media frames of COVID-19 in that:

H7: Republicans will perceive news stories that frame the “COVID-19 virus” as the “Chinese virus” as more authoritative, believable, credible, informative, and persuasive than will Democrats.

H8: The more conservative respondents view themselves, the more authoritative, believable, credible, informative, and persuasive they will rate the “Chinese virus” article as being.

H9: Republicans will be more likely to blame China for the pandemic than will Democrats.

H10: The more conservative people are, the more blame they will assign to China for the pandemic.

H11: The “Chinese virus” article will lead to assigning more blame to China for the pandemic among Republicans than among Democrats.

H12: The “Chinese virus” article will lead to assigning more blame to China for the pandemic among more conservative people.

Method

Participants

This online experiment was pre-registered (Open Science Framework,

https://osf.io/hmeb5/). It received IRB approval from The Ohio State University (Protocol 2020B0124). The sample size for this study was determined using a power analysis (Cohen, 1988). We wanted to detect an effect size of d = 0.25 with power = .80 at the .05 (two-sided) significance level. The required sample size is 284 participants per group, for a total of 568 participants in the two groups. According to Promising Practices Network (2014), terms such as significant, important, notable, and consequential are typically reserved for effect sizes at least d = 0.25. We tested some extra participants (n = 46) in case any had to be discarded due to missing data.

Participants were recruited and paid by Dynata, an online survey software company, to complete a survey programmed in Qualtrics. Dynata used the following age and race quotas to obtain a nationally representative sample of American adults in terms of age and race/ethnicity. For age, the quotas were: 18–24 (17.73%), 25–34 (16.82%), 35–44 (16.82%), 45–54 (18.28%), 55–64 (14.08%), 65 and older (16.27%). For race/ethnicity, the quotas were: White/Caucasian (63.80%), Black/African American (12.07%), Hispanic (16.27%), Native American/Inuit or Aleut, Native Hawaiian/Pacific Islander, Other (7.86%). Asian-Americans were excluded from the sample because we were interested in reactions from other Americans toward Asian-Americans. However, not all participants who received invitations completed the survey. Thus, our sample deviated in significant ways from quotas. Participants were 614 American adults (48.05% male, 49.74% females, 1.63% Other, and 0.65% unreported). For age, we obtained the following percentages: 18–24 (1.48%), 25–34 (5.90%), 35–44 (12.79%), 45–54 (16.89%), 55–64 (25.08%), 65 and older (37.87%). Overall, participants ranged in age from 18–90 (M_age = 57.98, SD = 14.55). For race/ethnicity, we obtained the following percentages: White/Caucasian (83.84%), Black/African American (7.00%), Hispanic (5.70%), Native American/Inuit or Aleut, Native Hawaiian/Pacific Islander, Other (3.42%). Thus, it is better to consider our sample to be a diverse sample of American adults rather than a nationally representative sample.

Data were collected from April 15th to May 21st, 2020, when stay-at-home orders were in place in many states in the U.S. to avoid the spread of the coronavirus.

Procedure

Consent and cover story

After giving their consent, participants were given a cover story that read: “This survey consists of two separate studies, which have been combined for efficiency.” Participants reported their gender, age, and race. They were told, “The first study is about how people evaluate news sources.”

News article

By random assignment, participants read a fabricated National Public Radio (NPR) news story about the coronavirus that was described using the term “COVID-19 virus,” or the term “Chinese virus” (see Appendix A). Both articles discuss two possible theories about the origins of the coronavirus, as well as the development of a vaccine to curb its spread. Both articles include quotes from infectious disease expert Dr. Anthony Fauci and Chinese Military Spokesperson Zhao Lijian. The articles primarily differ in their use of the label “COVID-19 virus” versus “Chinese virus.” In addition, the “COVID-19 virus” version says, “Regardless of its’ origin, the World Health Organization [WHO] advises against naming pathogens for geographic regions. Thus, the WHO uses the term coronavirus disease of 2019 for this deadly virus, or COVID-19 virus for short.” In contrast, the “Chinese virus” version says, “Both theories suggest that the proper term for this deadly virus should be the ‘Chinese virus.’ Many other diseases have been named after their country of origin, such as ‘German measles’ and the ‘Spanish flu.’” Afterward, participants rated how much they agreed that the article was authoritative, believable, credible, informative, and persuasive (1 = strongly disagree to 5 = strongly agree). Because they were designed to measure the same construct and loaded on a single factor (Eigenvalue = 3.60, which explained 72.05% of the variance), we averaged the five individual ratings to form an overall favorability rating for the article (Cronbach α = .90, McDonald ω = .90).

We did not include an article manipulation check because we did not want it to contaminate our treatment by drawing attention to the term used for the coronavirus (e.g., Kidd, 1976; Hauser, et al., 2018).

Prejudice measures

Next, participants were told, “The second study is about your attitude about different groups of people.” Using feelings thermometers, participants rated their feelings about the following groups of Americans (in a randomized order): African-Americans, ArabAmericans, Asian-Americans, Chinese-Americans, European-Americans, Hispanic-Americans, Native-Americans (0 = very cold/not at all positive to 100 = very warm/extremely positive).

Participants then responded to four scales designed to measure prejudice against Chinese- and Asian-Americans (1 = strongly disagree to 5 = strongly agree). The first scale was the Social Distance Scale (Mather, et al., 2017; Cronbach α = .97, McDonald ω = .97), which consists of nine items used to measure participant willingness to engage in various social contacts with Chinese-Americans (e.g., “I would be willing to accept a Chinese-American as a close relative by marriage,” “I would be willing to accept Chinese-American as a coworker”). The second scale was the Scale of Anti-Asian American Stereotypes (Lin, et al., 2005; Cronbach α = .85, McDonald ω = .90), which consists of 25 items that are divided into two subscales: (1) a competence scale that contains 12 items (e.g., “constantly in pursuit of more power,” “striving to become number one”; Cronbach α = .84, McDonald ω = .86), and a sociability scale that contains 13 items (e.g., “commit less time to socializing than others do,” “do not interact smoothly in social situations”; Cronbach α = .76, McDonald ω = .78). The third scale was the Attitudes Toward Asians Scale (Ho & Jackson, 2001; Cronbach α = .91, McDonald ω = .92), which contains 28 items that are divided into two subscales: (1) a positive stereotype subscale that contains 11 items (e.g., “Asian Americans increase the ‘brain power’ of the United States,” “Generally, Asian Americans are smart”; Cronbach α = .92, McDonald ω = .92), and (2) a negative stereotype subscale that contains 17 items (e.g., “There are too many Asian Americans in this country,” “Asian Americans are overly competitive”; Cronbach α = .97, McDonald ω = .97). The fourth scale contained three items (“It is okay to ask someone of Asian descent where they are really from,” “It is okay to tell someone of Asian descent that they speak English very well,” “It is okay to ask someone of Asian descent what their real name is”) from the 28-item Asian American Racism Related Stress Inventory (Liang, Li, & Kim, 2004; Cronbach α = .87, McDonald ω = .88).

Because they were designed to measure the same construct and loaded on a single factor (Eigenvalue = 3.28, which explained 54.67% of the variance), we standardized and averaged the six prejudice scales to form a single prejudice measure (Cronbach α = .78, McDonald ω = .91). Scales were reverse scored when required so that higher scores on all scales indicated higher levels of prejudice toward Chinese- and Asian-Americans.

Political party and ideology

In addition, participants indicated their political party: Republican (35.74%), Democrat (35.25%), Independent (26.89%), or other (2.13%), and their political ideology (1 = very liberal to 7 = very conservative, M = 4.14, SD = 1.68). We measured both political party and political ideology because they are not the same thing. A political party refers to a group of people that acquire and exercise political power, whereas a political ideology refers to a set of political beliefs and values.

Blaming China for pandemic measure

Finally, participants were asked whether they agreed with the statement “China is responsible for the current global pandemic” (1 = strongly disagree to 5 = strongly agree).

Debriefing

Because there was deception involved in the study, a debriefing was provided at the end of the survey, and participants were provided the opportunity to withdraw their data at that point.

Results

The main statistical analyses for all hypotheses are in Supplemental Material. We used analysis of variance to test the independent and joint influence of news article (i.e., “COVID-19 virus” vs. “Chinese virus”) and political party (i.e., Democrat vs. Republican) on the outcome variables (i.e., article ratings, prejudice toward Chinese- and Asian-Americans, blaming China for the pandemic). We used regression analysis to test the independent and joint influence of news article and political ideology (i.e., 1 = very liberal to 7 = very conservative) on the outcome variables.

Demographic Variables

There were no significant effects for gender, age, or race on any of the outcome variables. Thus, the data for males and females of all ages and races were combined for subsequent analyses.

Favorability of the News Articles

Political party

We expected Republicans to view the “Chinese virus” article more favorably than Democrats. To test this hypothesis (H1), we excluded Independents and Others. As expected, there was a significant interaction between type of article and political party for the overall article rating (see Table SM1 in SM). As expected, Democrats evaluated the “Chinese virus” article less favorably (M = 3.38, SD = 0.97) than did Republicans (M = 3.79, SD = 0.80), t(210) = 3.39, p < .001, d = 0.46. Democrats (M = 3.74, SD = 0.76) and Republicans (M = 3.81, SD = 0.69) did not differ in their evaluations of the “COVID-19 virus” article, t(221) = 0.65, p = .52, d = 0.10. Put differently, Democrats viewed the “Chinese virus” article more negatively than the “COVID-19 virus” article, t(213) = 3.72, p < .001, d = 0.42. In contrast, Republicans rated the “Chinese virus” and “COVID-19 virus” articles about the same, t(218) = 0.48, p = .63, d = 0.027. Thus, there was support for hypothesis H1.

Political ideology

We expected more conservative people to view the “Chinese virus” article more favorably. As expected, the interaction between type of article and the overall article rating was significant (see Table SM2 in SM). As expected, the correlation between political ideology and evaluation of the “Chinese virus” article was positive and significant, r(300) = .12, p = .034. In contrast, the correlation between political ideology and evaluation of the “COVID-19 virus” article was nonsignificant, r(310) = .04, p = .49. Thus, hypothesis H2 was supported.

Prejudice Attitudes toward Asian- and Chinese Americans

There were no main effects of type of article on the prejudice measures, regardless of whether the overall article rating was included as a covariate in the models (see Tables SM6 and SM7 in SM). Thus, hypothesis H3 was not supported.

As expected, Republicans had more prejudiced attitudes toward Chinese- and Asian-Americans than did Democrats, regardless of whether article ratings were included as covariates in the models (see Tables SM3 and SM4 in SM). Thus, hypothesis H4 was supported.

As expected, the more conservative people were, the more prejudiced they were toward Chinese- and Asian-Americans, regardless of whether article ratings were included as covariates in the models (see Tables SM6 to SM8 in SM). Thus, hypothesis H5 was supported.

Contrary to expectations, there were no significant interactions between article and political party on the prejudice measure, regardless of whether article ratings were included as covariates in the models (see Tables SM3 and SM4 in SM). Thus, hypothesis H6 was not supported. Likewise, there were no significant interactions between article and political ideology on the prejudice measure, regardless of whether article ratings were included as covariates in the models (see Tables SM6 and SM7 in SM). Thus, hypothesis H7 was not supported.

Blaming China for the Pandemic

As expected, participants who read the “Chinese virus” article were more likely to blame China for the pandemic (M = 3.75, SD = 1.15) than those who read the “COVID-19 virus” article (M = 3.59, SD = 1.09), but only in the model with the overall article rating included as a covariate (see Table SM12 in SM). Thus, hypothesis H8 was supported in that model.

As expected, Republicans were more likely (M = 4.19, SD = 0.94) than Democrats (M = 3.27, SD = 1.13) to blame China for the pandemic, regardless of whether article ratings were included as covariates in the models (see Tables SM9 to SM11 in SM). Thus, hypothesis H9 was supported. Likewise, the more conservative people were, the more likely they were to blame China for the pandemic, r(612) = .38, p < .001, regardless of whether the overall article rating was included as a covariate in the models (see Tables SM12 to SM14 in SM). Thus, hypothesis H10 was supported.

Contrary to expectations, there were no significant interactions between article and political party on blaming China for the pandemic, regardless of whether article ratings were included as covariates in the models (see Tables SM9 to SM11 in SM). Thus, hypothesis H11 was not supported. Likewise, there were no significant interactions between article and political ideology on blaming China for the pandemic, regardless of whether article ratings were included as covariates in the models (see Table SM12 to SM14 in Sm). Thus, hypothesis H12 was not supported.

Discussion

The entire world population has been challenged by the coronavirus pandemic. It has left health officials, national economies, and businesses scrambling to tackle the effects of the growing infection rates. Media have been struggling to provide the most recent and relevant information to the public. For many, media have served as the primary source of what they believe to be the important, correct, and relevant information (Gerbner et al., 2002). Media cannot be directly connected to peoples’ actions, but they can have a considerable impact on peoples’ perception of the coronavirus (Bolsen et al., 2020). Research on framing effect supports this in demonstrating links between how media contextualize issues and public perception of those issues. The distinct frames that news sources deploy when discussing the coronavirus can influence a variety of viewer beliefs about the coronavirus and who is responsible for its existence and spread. (Dhanani & Franz, 2021).

Our analysis finds that there is a considerable difference between Democrats and Republicans in their perception of the “Chinese virus” article, with Democrats viewing it more negatively than Republicans. The result was similar for political ideology, as more conservative individuals viewed the “Chinese virus” article more favorably than did more liberal individuals. These findings suggest that a single word in a report can signal partisan framing, impacting perceptions of the favorability of that report along political ideology. Furthermore, the outcomes suggest that the impact of a shared language frame, particularly one that demonizes a foreign Other, can influence perception of an entire group (i.e., Chinese- and Asian-Americans).

Overall, individuals who read the news article in which the coronavirus was called the “Chinese Virus” were more likely to blame China for the pandemic, than those who read the same news article in which the coronavirus was called the “COVID-19 Virus,” but only when the overall article rating was used as a covariate. Thus, the effect of the “Chinese Virus” article on blaming China for the pandemic was most pronounced when controlling for how negatively people judged that article to be. This effect, albeit small, was not limited to Republicans and conservatives, despite these partisans bringing more self-reported anti-Asian prejudice to the table.

Though small, this is perhaps our most troubling finding. It suggests that despite supposedly more egalitarian beliefs, even Democrats harbor biases against Asians that have not gone away despite media and prominent officials disavowing the onus for the virus being on China. This finding seems to underscore previous research showing that frames, like media messages themselves, are polysemic, in that audiences bring with them preexisting attitudes and can interpret the framing to mean whatever they so choose, based on their traditionally held beliefs (Holt et al., 2017). Although it is not solely the frame that influences what people think of media messages, with an individual’s ideologies also providing support for the direction of the story, the frame can have broader impact in times of uncertainty. Perhaps what is happening is that people are seeking out media that underscores what they already believe, and they are using that media to reinforce their prejudices against Chinese- and Asian-Americans.

Although some Democrats demonstrated negative attitudes, across all findings in general, Republicans were more likely than Democrats to blame China for the pandemic. Likewise, the more conservative people were, the more likely they were to blame China for the pandemic. This could explain why Republicans and conservatives did not view the “Chinese virus” label negatively. However, the two news stories had similar effects regardless of political party and political ideology. These findings underscore earlier findings that preexisting beliefs might have worked in concert with media messages to determine the effectiveness of framing. In other words, it is not solely the media messages about the coronavirus and how these messages are framed, that shapes peoples’ perceptions and opinions of Chinese- and Asian-Americans. These perceptions and opinions are also shaped by what people wanted to believe, or what their political ideology or party believe.

Limitations and Future Research

In this study, we did not manipulate the news source, which was NPR for both articles. Although NPR is a public news source rated as relatively neutral and unbiased, it might not be perceived as a credible source for Republicans and conservatives (Mitchell, et al., 2014). Likewise, we did not attribute a source to the “Chinese virus” label in the news article. The “Chinese virus” label might have been more impactful to Republicans and conservatives if we had attributed it to former President Donald Trump.

To reduce suspicion, we did not include any pretest measurements on our outcome variables. Thus, we cannot test whether people (e.g., Republicans, Democrats, conservatives, liberals) changed their attitudes as a function of the version of the article they read. This remains a topic for future research.

It might not be surprising that exposure to a single news story did not move relatively stable prejudice attitudes. We included a “feeling thermometer” state measures, but the news story manipulation did not affect them either. It could be that feelings toward different outgroups are relatively stable. Another limitation is that we used several one-item measures, which are less reliable than multiple item measures.

Although Anti-Asian prejudice and discrimination is occurring in many other Western countries (e.g., “The coronavirus,” 2020), this study was conducted in the U.S. Thus, it is unclear whether these results would generalize to other countries.

Conclusions

In this analysis, we find that words matter, but they matter differently for different groups, based on both the news recipient’s political party and political ideology. In short, attitudinal factors audiences bring with them long before they encounter the news matter. Media are unlikely to change an individual’s attitude. However, words spoken by prominent political sources received through media channels could trigger negative thoughts about groups along stereotypical lines, with the potential to reinforce or even heighten racist and/or xenophobic animus toward these groups. Hence, it is wise to note not just what media say, or how they say it, but to whom, and about whom.

Open Scholarship

This article has earned the Center for Open Science badges for Open Data, Open Materials and Preregistered. The data and materials are openly accessible at https://osf.io/hmeb5/

Acknowledgments

We would like to thank Golnoosh Behrouzian, Chris Crandall, Emily Davidson, and Kate Helmstetter for their help with this study.

Disclosure Statement

No potential conflict of interest was reported by the author(s).

Data availability statement

The data described in this article are openly available in the Open Science Framework at https://osf.io/hmeb5/

Supplementary Material

Supplemental data for this article can be accessed on the publisher’s website.

Correction Statement

This article has been republished with minor changes. These changes do not impact the academic content of the article.

Appendix A:

Coronavirus News Stories

Note. The “COVID-19” vs. “Chinese” virus manipulation is in parentheses.

National Public Radio

The United States and the world remain in the crushing grip of the (COVID-19/Chinese) virus, which has infected more than 2 million people worldwide and claimed over 120,000 lives since creeping out of China, spreading across Europe and is now taking lives across the U.S.

New York has become the new epicenter for cases of the (COVID-19/Chinese) virus, having more than 200,000 infections and over 10,000 deaths. These numbers are highest in the world, surpassing even more densely populated regions of southeast Asia. Across the U.S., all aspects of life are affected. Restaurants and bars are closed. All major sporting events are canceled; the Olympics have been postponed until 2021. Residents in every major U.S. city are quarantined. Universities and other schools are requiring students to stay home. Millions are out of work, putting many Americans at risk of losing their homes.

The (COVID-19/Chinese) virus is a once-in-a-generation virus that spreads quickly and with it, so have stories about its’ origin. Two prominent theories have emerged. The first is that it originated at a Chinese wet market where live animals like bats, dogs, cats, snakes and koala bears are slaughtered in filthy and unsanitary conditions. These wet markets are breeding grounds for animal-to-human virus transmission. The second theory is that the virus emerged from a hazardous level 4 bio-weapons facility in Wuhan, China located adjacent to the wet market. The virus managed to migrate from the lab, perhaps from an infected bat, to the wet market. (Regardless of its’ origin, the World Health Organization [WHO] advises against naming pathogens for geographic regions. Thus, the WHO uses the term coronavirus disease of 2019 for this deadly virus, or COVID-19 virus for short/Both theories suggest that the proper term for this deadly virus should be the “Chinese virus.” Many other diseases have been named after their country of origin, such as “German measles” and the “Spanish flu.”)

Along with questions of the (COVID-19/Chinese) virus’s backstory, so have questions emerged as to who is to blame for its’ global spread. Some point to China’s totalitarian regime. They point to the fact that in December – before the virus ever left their country – Chinese health officials knew the virus could spread from person to person. Yet, On January 18, Chinese President Xi Jinping allowed a Lunar New Year banquet to go forward in Wuhan where tens of thousands of families shared food, and then let millions travel out of Wuhan, allowing the disease to spread across the world. It was not until January 23 that the Chinese government enacted a quarantine in Wuhan.

“If the regime had taken action as soon as human-to-human transmission was detected, it might have contained the virus and prevented a global pandemic,” infectious disease expert Dr. Anthony Fauci told National Public Radio. “Instead, they tried to cover it up because leaders who abuse their people are less concerned with saving lives than they are in making sure the world does not discover the deadly inefficiency of their system.”

Chinese Ministry Spokesperson Zhao Lijian has since countered saying the U.S. Army, which has operations all over the world, is actually ground zero for the virus.

“Why are there so many more people infected in the U.S. than anywhere else?” he asked. “It might be the U.S. Army that brought the epidemic to Wuhan. When did patient zero first get sick in the U.S.? The U.S. owes us an explanation! Be transparent!”

Regardless of blame or origin, the (COVID-19/Chinese) virus still continues to spread at an astounding rate. Scientists and other health officials estimate even with quarantines and social distancing the death toll in the U.S. alone could easily top 240,000, in their best-case scenarios.

A vaccine for the (COVID-19/Chinese) virus is not expected for at least a year to 18 months.