The coronavirus public health crisis is also a political-communication and health-communication crisis. In this article, the authors describe the key communication-related phenomena and evidence of concerning effects manifested in the United States during the initial response to the pandemic. The authors outline the conditions of communication about coronavirus that contribute to deleterious outcomes, including partisan cueing, conflicting science, downplayed threats, emotional arousal, fragmented media, and Trump's messaging. The authors suggest these have contributed to divergent responses by media sources, partisan leaders, and the public alike, leading to different attitudes and beliefs as well as varying protective actions taken by members of the public to reduce their risk. In turn, these divergent communication phenomena will likely amplify geographic variation in and inequities with COVID-19 disease outcomes. The authors conclude with some suggestions for future research, particularly surrounding communication about health inequity and strategies for reducing partisan divergence in views of public health issues in the future.
On February 28, 2020, during a campaign rally, President Donald Trump referred to the Democrats' criticism of his handling of coronavirus as a “new hoax,” claiming that “the Democrats are politicizing the coronavirus—they're politicizing it” (Cook and Choi 2020). As scholars who have studied health and political communication for more than a decade—including the politicization of health issues and how conflicting messages about health and science can produce deleterious consequences—this moment felt ominous. The dynamics that would shape the public's response to the newly emergent coronavirus had been laid down in politics and health policy years before, but these words from the largest megaphone in the country contributed to a cascade of responses that made a catastrophe even more catastrophic.
In this commentary, we outline the key communication-related phenomena and emerging evidence of effects manifested in the early stage of the crisis in the United States (February–April 2020). While coronavirus is of course a novel public health challenge, we apply lessons learned from our analyses of other emergent health issues to draw implications. This is not a comprehensive review, nor do we introduce our own new empirical evidence. Instead, our goal is to cue academics, journalists, and policy makers to the key communication challenges and introduce concepts warranting further research. As of this writing (early May 2020), the coronavirus threat will be with us for months, if not years. The cascade of effects from early communication will also likely persist.
New Issues and the Specter of Motivated Reasoning
Our research indicates that the way a new health issue emerges into media discourse has consequences for the public's response, likely for the long term. The emergence of a new issue will be accompanied by competition to define the problem and shape public understanding (Baumgartner and Jones 1993; Nisbet, Brossard, and Kroepsch 2003). For instance, the human papillomavirus (HPV) vaccine was approved in 2006, and states quickly introduced legislation proposing compulsory vaccination for middle-school girls. A heated debate among politicians, medical experts, pharmaceutical companies, and ideological and religious groups ensued. Our research documented how the framing contest played out in print media, showing that the news frequently mentioned controversy explicitly, significantly more so after states introduced legislative mandates (Fowler et al. 2012). Furthermore, when an issue such as HPV vaccine emerges with partisan competition as a feature—specifically, debate among political actors—this presentation is “sticky” and shapes subsequent public understanding and policy support (Fowler and Gollust 2015). A similar process played out with news coverage of the Affordable Care Act (ACA), an issue defined by partisan conflict immediately after its passage. Local TV news coverage of new health insurance marketplaces in 2013–14 often featured political frames and partisan cues, not substantive policy detail (Gollust et al. 2017). And, in subsequent TV news coverage of health insurance (2018–19), these partisan cues persisted (Gollust, Fowler, and Niederdeppe 2020).
A major consequence of the politicization of health issues—defined here as when political or partisan cues become integrated into the public presentation of the health issue (Fowler and Gollust 2015)—is that it increases the likelihood of the public's interpreting the issue through the lens of partisan identity, known as motivated reasoning (Strickland, Taber, and Lodge 2011). Studies demonstrate that polarization in elite rhetoric about the ACA leads the public to interpret information in ways consistent with motivated reasoning, with Democrats and Republicans having divergent responses to ACA media information (Fowler et al. 2017) and new evidence (James and Van Ryzin 2017).
As is clear from the quote that opens this essay, the conditions for partisan motivated reasoning were baked into the context of the emergent coronavirus. From the earliest alarm, Republican politicians followed Trump's lead in publicly downplaying the threat, while Democratic politicians responded with more concern, signaling different public cues. And indeed, survey evidence from March demonstrated that individuals interpreted the threat in partisan-patterned ways, with Republicans following party leaders in dismissing the threat and taking fewer health-protective actions than did Democrats (Gadarian, Goodman, and Pepinsky 2020). These sequelae of events contrast with earlier time periods: Bethany Albertson and Shana Kushner Gadarian (2015) argued that public health threats did not generate a partisan response because smallpox and H1N1 were viewed as unframed threats (that is, partisans did not already view them differently). While there is some evidence that the public evaluates the management of previous infectious disease outbreaks in ways aligned with their allegiance to the political party in charge (Nyhan 2014), and that this can translate into partisan differences in vaccine uptake (Baum 2011) and in attitudes about immigration (Adida, Dionne, and Platas 2018), the context of coronavirus communication is distinct.
The Distinctive Characteristics of Coronavirus Communication
Distinctive features of the coronavirus context in early 2020 might have shaped the public and political response. First, as SARS-CoV-2 is a genuinely novel viral threat, the public, politicians, and scientists alike have had limited initial information. With limited knowledge, people are likely to rely on media information and especially on cues provided by political elites (Zaller 1992; Cobb 2005). And, after years of polarized messaging around a health issue common in public discourse—the protracted political fight to repeal, replace, and erode the ACA—combined with high levels of entrenched partisan polarization, the conditions were set for the public to receive new information from elites in politically filtered ways.
Second, with new issues comes rapidly evolving science, which can give rise to the appearance of expert disagreement or conflicting information—such as whether masks are helpful, what the case-fatality rate is, and whether or not hydrochloroquine has therapeutic potential. As we describe below, conflicting health messages may have important consequences (Nagler 2014). The speed with which new scientific information is produced and the need for rapid response also enhance the likelihood of misinformation being available in the information environment (Scheufele et al. 2020). Moreover, scientific research is currently disseminated within the context of long-term decay in scientific trust among political conservatives (Gauchat 2012), further contributing to an asymmetric partisan response to scientific information.
Third, one element of conflicting messaging among elites was the severity of the coronavirus threat—whether it was exceptionally threatening or just “like the flu.” Because the core communication conflict concerns threat severity (unlike, say, whether marketplace websites were working, as in the early stages of ACA communication), behavioral science predicts that the public's likelihood of taking behavioral action would be affected. One critical factor leading to individuals' intentions to change their behavior is perceiving a personally resonant and sufficiently severe threat (Jones et al. 2015). A reduced impetus to act is particularly concerning given that large-scale behavioral changes by individuals are critical to pandemic response (Van Bavel et al. 2020).
Fourth, the viral threat provokes strong emotions, particularly fear and anxiety. These emotions can shape individuals' likelihood of taking action directly (Van Bavel et al. 2020); from a communication perspective, they can also drive information-seeking. Media imagery in January and February of overtaxed hospitals and overflowing morgues in China and Italy likely contributed to this strong emotional response. Research demonstrates that when people feel anxious, they tend to seek out information to resolve this feeling (Albertson and Gadarian 2015). Anxiety-fueled searching can be biased; that is, people select information, especially on the Internet, that conforms to their priors (Valentino et al. 2009). Indeed, as of late March, 70% of Americans had searched online for coronavirus information (Anderson and Vogels 2020), potentially increasing their likelihood of exposure to partisan-oriented or inaccurate information.
Fifth, the public is exposed to divergent information environments given long-term trends in media fragmentation and selection of media sources based on one's political worldview (Iyengar and Hahn 2009; Muddiman, Stroud, and McCombs 2014). These trends mean that providing uniform messaging to the public is a challenge, particularly within the localized communication context of COVID-19, where most cases and deaths are clustered in a relatively small number of counties. Furthermore, varying health and political environments contributed to different policy recommendations by locality (such as mandates to wear masks). Local news is thus a particularly important information source, even while local outlets are facing economic collapse that is accelerated during the COVID-19 downturn.
Finally, as we describe in the next section, the communication context includes extraordinary messaging by Trump that defied well-supported risk communication guidance and increased the accessibility of deleterious messaging, including misinformation.
The Context of Trump's Coronavirus Messaging
President Trump's communication choices will surely yield extensive future commentary (see, e.g., Peters, Plott, and Haberman 2020). For now, we highlight a few specific—and likely damaging—features of early presidential messaging. As mentioned above, Trump downplayed the threat by comparing it to the flu and predicting that the virus would “miraculously” go away by April 2020. He also rapidly changed his views within short time periods, contributing to the propagation of conflicting information in public discourse.
In addition to downplaying the threat, Trump also disseminated misinformation. Over a period of weeks, he promoted the drug hydrochloroquine as a COVID-19 treatment in frequent communications and even stockpiled the drug, despite evidence accumulating against its use (Eban 2020). On April 23, he suggested that light or disinfectant agents might be used to cure the disease, prompting swift backlash to these inaccurate and harmful claims (Rogers et al. 2020). And on April 30, he suggested at a White House event that the coronavirus originated in a Chinese lab—a conspiracy theory circulating in right-wing media for weeks—despite scientific consensus that the virus is naturally occurring (Singh, Davidson, and Borger 2020).
Trump also engaged in other politically consequential messaging. To avoid taking responsibility for administration missteps and delays, he blamed others, including labeling the virus a “Chinese virus” in a purposeful attempt to shift responsibility to China. This framing contributes to an understanding of the problem that emphasizes “us”-versus-“them” division rather than shared understanding, stigmatizing both the nation of China and people of Chinese descent. He further promoted the appearance of public division when he called for liberation from shelter-in-place orders, boosting fringe movements despite broad public support for social distancing and stay-at-home recommendations (Kirzinger et al. 2020).
While shirking responsibility for missteps, Trump also went to great lengths to claim credit, even where it was not warranted and could be dangerous. Claiming credit for accomplishments is a common technique in politics (Mayhew 1974), and Trump's insistence on putting his name on CDC coronavirus health information mailers (or on relief checks) could be viewed as just another example. However, in an extremely polarized era, when the stakes of not listening to guidance can threaten everyone's safety, putting a partisan label on a health communication message could promote backlash on both sides. After all, the public trusts medical experts more than politicians during a public health crisis (Albertson and Gadarian 2015).
All of these features starkly contrast to principles of effective risk communication. On March 29, 2020, former US Surgeon General Vivek Murthy tweeted five “tried and true” principles for crisis communication: (1) be transparent and truthful, even if it involves acknowledging bad news or unknowns; (2) be consistent and, if circumstances change, explain why; (3) overcommunicate with the public rather than sharing too little information; (4) lead with scientists and science—with experts, not politicians; and (5) be compassionate and empathetic (Murthy 2020). Additional best practices include using simple plain language (e.g., “stay at home” vs. “social distancing”) and coordinating communication across national, state, and local authorities (CDC 2014).
Past and Emerging Research on Media Messaging and Its Effects on the Public
While coronavirus-specific research is in the early stages, there is already evidence corroborating that the presidential messaging and communication conditions described above have led to concerning outcomes in terms of public attitudes and behavior.
The sequence of events describing partisan-motivated reasoning introduced above has emerged, demonstrating that Republicans and Democrats had divergent responses. In early March, 40% of Democrats compared to 20% of Republicans believed the threat of coronavirus was imminent (Axelrod 2020). In an assessment of the public's views later in March, Shana Kushner Gadarian, Sara Wallace Goodman, and Thomas B. Pepinsky (2020) documented extensive partisan differences in attitudes (e.g., perceptions of threat and personal concern), reported health behaviors (e.g., washing hands, avoiding contact with others), and policy opinion (e.g., payment for treatment and testing). Similarly, KFF polling demonstrated that Democrats were more likely to take precautions such as changing travel or gathering plans and stocking up on supplies than were Republicans (March 11–15, just before most state shelter-in-place and school closures began) (Hamel et al. 2020). Of course, self-reported behaviors have limitations; yet even data sources tracking behavioral patterns discretely (e.g., geolocation data from cell phones) showed that people living in more Republican-leaning areas lagged behind those in Democratic-leaning areas in engaging in social distancing recommendations (Andersen 2020). In mid-April, there was a strong bipartisan consensus on the threat and support for shelter-in-place recommendations (Kirzinger et al. 2020), although small groups of Trump-supporting protesters, stoked by the president himself, created the public appearance of partisan discord. By late April, polling continued to display partisan differences in understanding of the virus even among essential workers and also demonstrated that geography matters: essential workers of all parties living in counties won by Trump were less confident that social distancing saves lives than those living in counties won by Clinton (Rothwell 2020).
Media sources covered the pandemic differently, and this contributed not only to these partisan patterns in response but also to specific news-source effects on public understanding, and misunderstanding, of the pandemic (Jamieson and Albarracin 2020). Previous research has indicated that exposure to certain media outlets (particularly Fox News) contributes to politically relevant outcomes (DellaVigna and Kaplan 2007; Feldman et al. 2012); new research similarly suggests coronavirus-related media-source effects. For instance, Matt Motta, Dominik Stecula, and Christina Farhart (2020) found that right-leaning news sources (Fox News, Breitbart) were more likely than others to disseminate specific pieces of misinformation (e.g., that coronavirus is a conspiracy or was made in a lab), and that survey respondents who consumed more right-leaning news were also more likely to endorse misinformation. Similarly, Kathleen Hall Jamieson and Dolores Albarracin (2020) identified associations between exposure to mainstream broadcast news and the correct belief that COVID-19 is more lethal than the flu, exposure to mainstream print news and more accurate beliefs about infection prevention, and exposure to Fox News and Rush Limbaugh and more endorsement of conspiracy beliefs (that the CDC is exaggerating to undermine President Trump). Emerging research even links viewership of specific programs that downplayed the threat to population health outcomes, based on evidence that threat dismissal leads to delayed protective behavior (Bursztyn et al. 2020).
Finally, while scientific evidence production and dissemination continues to grow, we can predict consequences of expert disagreement and conflicting information. When the public perceives conflict and controversy about health issues—such as contradictory nutrition research (Nagler 2014) or shifting mammography screening guidelines (Nagler, Yzer, and Rothman 2019)—this can generate confusion about and decreased trust in such recommendations. Such effects might even carry over to other unrelated health topics—even those about which there is expert consensus—thus potentially undermining broader trust in health recommendations (Nagler 2014; Nagler, Yzer, and Rothman 2019). We need to consider this possibility in the context of COVID-19. Over the course of just a few weeks, there have been shifts in advice around protective behaviors such as mask wearing. Researchers are well equipped to understand why recommendations shift—given the novelty of this virus, there is much scientists do not yet know—yet whether the public, with limited literacy about scientific research, can negotiate what feels like daily swings in recommendations is unclear. To the extent that people perceive expert disagreement, this can generate not only perceptions of scientific uncertainty (van der Bles et al. 2019) but also potential behavioral consequences, including less compliance with mitigation practices.
Implications for the Future
As of early May 2020, certain states were beginning to “open” again, even while US cases exceeded one million and more than 70,000 Americans had died. It is clear that partisan differences in response to the pandemic among the public and elites alike—as well as media sources—have contributed to fragmented approaches to managing the threat, varying based on proximity to the early cases as well as state and local leadership. These communication patterns will yield substantial differences in health outcomes. Of course, this is not new for health policy—divergent state decisions regarding insurance markets and Medicaid expansion further fragmented an already fragmented health coverage landscape—but these local variations will be exacerbated in the coming months with COVID-19 morbidity and mortality. This has grave consequences for inequities that are already obvious by race (see Bailey and Moon and Alberti, Lantz, and Wilkins in this issue) and that will continue to expand by race and geography.
While rapid social science has documented the partisan dynamics of the crisis, more research is needed at the intersection of health communication and health equity. Within the context of inequality in experiences with COVID-19, communication could have an important role: not only boosting strong and credible health communication messages to promote health-protecting behaviors but also increasing understanding of and empathy toward people's disparate experiences. However, response to health messaging is always predicated on having access to the communication (e.g., broadband access) and capacity to respond (Viswanath et al. 2012). With those most vulnerable to infection the least able to take effective protective actions (such as staying at home for low-income essential workers), health communication alone is insufficient, particularly given the economic crisis that rages alongside the public health crisis. Moreover, as we have described, some political communication—particularly that emanating from the White House—challenges a shared understanding of who is at a risk and why (i.e., propagating conspiracy beliefs and stereotypes). This messaging could contribute to a heightened “us versus them” mentality alongside classic cleavages of race, class, and immigrant status. If this happens, it is less likely there will be broad public enthusiasm for solutions to ameliorate inequity. Communication that bridges, rather than enhances divides, is needed. Finally, we have seen unprecedented attention to racial disparities in COVID-19 infection and death in elite news media. While this attention could elevate health equity on the public and policy agenda, past research tells us that when disparity statistics are not contextualized in terms of systematic causal attributions, the public could rely on simplistic understanding of personal responsibility, stereotyping, and individual blame in interpreting group racial differences (Gollust and Lynch 2011; Niederdeppe et al. 2013).
More research is also needed on how we might overcome the politicized cascade demonstrated in coronavirus politics—partisan cueing, motivating reasoning, and polarized behavioral and attitudinal response. There are certainly roles for journalists who can amplify medical experts and public health guidance, and downplay the horse-race style of political coverage, which promotes political cues. While news coverage of politics and governance is important for political accountability, these stories and framing should be secondary to communicating clear and evidence-based health messaging. To do this, we need well-trained health journalists, particularly in a time of declining media resources. Research can also contribute to practical communication guidance, such as identifying effective ways to communicate about this pandemic that can reduce the likelihood of partisan-filtered responses (Bolsen and Druckman 2015). These might include specific messages, sources, or their combination—such as messages that warn the public not to follow partisan cues or that match unexpected source cues with messages emphasizing personal issue salience (see, e.g., Mullinix 2016; Cook, Lewandowsky, and Ecker 2017).
Finally, future support for public health will depend on overcoming the partisan divisions on display in 2020. Public health must once again become a bipartisan priority. While politicization appears to be “sticky” (i.e., once political dimensions of health issues emerge, they persist; see Fowler and Gollust 2015), and the political environment in 2020 has amplified partisan differences, this is not inevitable. Certainly, a change in federal leadership could shift the powerful bully pulpit messaging of the presidency. At the same time, historical and comparative research can also contribute to our understanding of when and how issues can become depoliticized. As Arthur Lupia (2013: 14050) noted, child labor was “once a contested political issue in American politics because people held, and were willing to publicly voice, different points of view [about whether child labor should be allowed]”—much like today's environment surrounding support for public health guidance. Over time, Lupia added, a social and moral consensus around child labor emerged, laws followed, and ultimately people no longer considered the issue to be politicized. Future research must explore how contested issues can shift out of a political lens to illuminate a path forward to when public health guidance will be a universally shared concern among the public and policy makers alike.
The authors acknowledge support from numerous mentors, collaborators, and funders who have sustained and fostered our interest in studying the intersection of public health and politics for many years. The content of this article is solely the responsibility of the authors and does not represent the official views of any of our past or current funders.