Editor’s Note: This article was published on September 23, 2020, at NEJM.org.
In late April 2020, Ethan Weiss, a San Francisco cardiologist, joined colleagues in New York to care for critically ill patients with Covid-19. On Weiss’s last day there, a relatively young man was dying of Covid, as his wife lay ill with Covid down the hall. They had been in the hospital nearly a month but had been unable to see one another, though the wife had been consenting for the husband’s procedures by phone. When the husband’s death was clearly imminent, Weiss went to the wife’s bedside and had a conversation that was among the most difficult of his career. As she sobbed, Weiss found himself crying into his own mask and goggles.
When the husband died a few hours later, Weiss was struck not just by the brutality of this disease that left people to die alone, but also by the difficulty of appreciating its destructive potential until you experience it firsthand. So when he saw a Facebook post from his high school friend Ms. R. objecting to universal mask wearing, he wanted to respond in a way that might change her mind.
Masks have become a flash point in our culture wars: as a symbol of either a commitment to public health or an infringement on basic liberties, the mask encapsulates the politicization of science. But since human behavior — including wearing or shunning masks — will determine the pandemic’s ultimate toll, communication strategies that bridge our partisan divide over science may prove as important as any novel therapeutic.
Distrust in Science
Beyond the near complete failure of U.S. federal leadership in combating the pandemic, one significant problem, according to Harvard epidemiologist Marc Lipsitch, has been the absence of consistent communication from nonpartisan experts. During the 2009 H1N1 influenza epidemic, he recalled, scientists and public health officials communicated daily with the public, offering nonpartisan advice. “The sidelining of all nonpartisan technical experts…has made it very hard for anyone to know what they should do,” Lipsitch said. That the administration has not just marginalized experts but has actively attempted to undermine their credibility has sown further confusion and distrust, a problem magnified by the many uncertainties surrounding SARS-CoV-2. But though President Donald Trump weaponizes scientific uncertainty and dismisses evidence of the virus’s widespread destruction, he is also exploiting a distrust in scientists that long predated his presidency.
Though such distrust is now more common on the political right,1 for decades conservatives viewed science favorably, seeing it as a means of production and an economic stimulant. But when scientists began studying how some of the technologies that had led to economic success, such as fossil fuels and nuclear power, could have detrimental effects, particularly on the environment, the resultant regulations alienated people committed to free-market principles. Trust in science concomitantly declined. Examining survey data from 1974 through 2010, University of Wisconsin sociologist Gordon Gauchat found that conservatives expressed greater trust in science than liberals or moderates at the beginning of that period but the least trust by the end.2 Moreover, rather counterintuitively, Gauchat found that trust in science decreased the most among the most educated conservatives: greater scientific literacy enabled people to find the limitations in the data or to exploit inevitable uncertainties.
Another contributor is a growing strain of anti-intellectualism, a generalized distrust of experts that is resistant to facts, though relatively independent of political ideology. One recent study, for instance, showed that when people with strong anti-intellectual tendencies were exposed to expert consensus on matters such as genetically modified organisms or water fluoridation, they actually doubled down on their opposition to the scientific message.3 The inevitable and necessary self-corrections involved in the scientific process merely reinforce this skepticism. The pandemic, with its myriad uncertainties, well-publicized retractions,4,5 and shifting recommendations, has most likely exacerbated this distrust.
The mask debacle is a case in point. The American public was initially advised against wearing masks, only to learn that this advice was based on a need to preserve masks for health care workers. This messaging created understandable confusion (why would masks protect health care workers but not others?) and reinforced skepticism among people already distrustful of health authorities. Though we would soon understand the relatively high degree of asymptomatic transmission, and thus the benefits of universal masking for virus containment, by then the debate had been politicized, leaving little room for the scientific flexibility demanded by a rapidly moving pandemic. Whereas the hallmark of scientific reasoning is the capacity to change your mind when the evidence evolves, the nature of dialogue in our polluted information environment generally rewards the opposite: make up your mind and then find evidence to support it. As easy as it is, then, to retrospectively criticize the initial lack of transparency regarding mask wearing, could even the most crisis-savvy communication have changed the outcome? Probably not.
Tom Nichols, who describes anti-intellectualism in his book The Death of Expertise, told me that, particularly in the United States, rejecting science has become a proxy for personal empowerment and autonomy.6 Noting that masks immediately became part of a partisan controversy over whether to believe in science and trust experts, Nichols lamented the growing narcissistic tenor of a society whose battle cry is “You are not the boss of me!” Though Nichols has argued that a disaster, such as an economic depression, war, or pandemic, might boost respect for knowledge and ability, the persistence of science denialism despite the presence of two of the three has changed his thinking. “Some people would rather die than wear a mask,” he said. “Once beliefs become fused to your sense of personal identity, they become very difficult to shake.”
This fusion also contributes to what sociologists call “affective polarization”: our current political divides are characterized not only by disagreement with the opposing party’s views, but also by frank contempt for the people espousing those views. Social media platforms, rather than facilitating exchanges of viewpoints, tend to thrive on these divisions and make reasoned debate impossible.
That’s why, as she explained to me when we first spoke in June, Ms. R. had to delete her Facebook post expressing her mask skepticism. “The conversation just became evil,” she said. “People attacking me. People attacking each other. My whole feed got clogged.” The pile-on, according to Weiss, primarily involved people telling Ms. R. she was a moron. While acknowledging that the controversy could perhaps have been avoided had Trump simply worn a mask at the outset, Weiss pointed out that the human instinct to be tribalistic means that once we pick a team, it’s very hard to switch. Likening the mask to the jersey of our favorite sports team, he said, “The mask became the uniform of the left.”
Ms. R.’s recognition of the politicized nature of the debate didn’t diminish her reservations about mask wearing. “Is it even effective?” she asked. “I feel suffocated. I feel like my oxygen levels are declining.” But I sensed that her reservations had less to do with masks per se than with a generalized confusion regarding the virus and risk-mitigation strategies. For instance, she asked, why could people go to liquor stores and Walmart but not to schools or churches? Why couldn’t kids play sports if people could go to rallies? And how lethal was the virus, really? She said there had been so many mixed messages and disinformation that she no longer knew whom to trust. “I think the medical community needs to put aside their politics and what they think about Donald Trump,” she said. “I don’t hate any of these experts. They just don’t know what’s going on.” For Ms. R., who works two jobs but was out of work for months — and will be again this fall because her daughters’ school is closed — the economic devastation is the most palpable. She doesn’t know anyone who has died of Covid, but she knows countless people who’ve had their livelihoods destroyed.
In that sense, the pandemic may have further alienated the many Americans who already felt that the “expert” or “elite” class didn’t understand their lives. For me, as for many science believers, watching more than 190,000 people die as millions of Americans defy expert advice elicits horror and rage. But though I can’t imagine reacting otherwise, an enlightening piece about class divides recorded by CNN correspondent Fareed Zakaria in mid-May at least helped me broaden my perspective.7
“Imagine you are an American who works with his hands,” Zakaria says, “a truck driver or a construction worker…and you’ve just lost your job because of the lockdowns.” What’s it like, he wonders, to be one of these 36 million jobless Americans and to turn on your TV only to hear “the medical experts, technocrats, journalists explain that we must keep the economy closed”? These experts who’ve advocated shutdowns, Zakaria points out, not only have jobs but have been in greater demand because of the pandemic. Emphasizing how worthless and scared the newly jobless might feel, he asks, “Is it so hard to understand why people like this might be skeptical of the experts?”
To some extent, yes. Given that economic recovery depends on containment of the virus, it seems irrational to defy public health advice for the sake of job preservation. But belief is not rational. Though the foundation of all scientific communication must be facts, insofar as people who distrust science are motivated by the perception that experts view them as idiotic, pointing out the idiocy of their behavior may simply reinforce their alienation. On some level, the medical profession recognizes that shaming people is often counterproductive when the aim is to foster trust or behavioral change. Most of us, for instance, don’t walk into the room of a patient who’s had a large myocardial infarction and begin by telling him how stupid it was to have spent his whole life smoking. It’s much easier, however, to approach people’s self-destructive behaviors with empathy and patience when those behaviors aren’t contributing to others’ deaths, societal unraveling, and paralysis of our own lives. But although disgust is a natural response to blatant disregard for others’ well-being, the collective nature of the consequences doesn’t make contempt any more likely to inspire the behavioral changes we seek.
Emphasizing the futility of trying to shame people into changing their behavior, Julia Marcus, a Harvard epidemiologist, remarked on the lack of empathy in much Covid-related public health messaging. Having studied HIV-related harm-reduction strategies, Marcus likened absolutism in Covid risk mitigation to insistence on lifelong condom use for combating HIV. Some clinicians, she explained, assuming that their patients will use condoms forever, refrain from prescribing preexposure prophylaxis (PrEP) because they don’t want to encourage condomless sex. Then, people who have condomless sex anyway are afraid to tell their clinicians and so don’t receive other risk-mitigation interventions such as PrEP or frequent testing. Similarly, much Covid messaging has “failed to recognize how unrealistic it is to expect people to abstain from the pleasures of life.” Successful harm reduction, says Marcus, requires accepting some level of risk. For instance, rather than repeating “Stay home!” when people disparagingly post pictures of large outdoor gatherings, Marcus’s instinct is to acknowledge that people will gather, and it’s safer outdoors than indoors. After all, she notes, “health is about more than being free of coronavirus. It’s about helping people live their lives safely and sustainably.”
That’s not to say that shaming never works. A large body of evidence in social psychology suggests that people respond to social norms, often out of fear of punishment or condemnation.8 My own experience with mask wearing is consistent with this evidence. As much as I wish I wore a mask purely for the sake of public health, I’ve received enough hateful glares after absentmindedly pulling it down to know that fear of others’ wrath is a powerful motivator. But for people who are resistant to the norm in the first place, shaming can paradoxically backfire. As MIT behavioral scientists David Rand and Erez Yoeli explained, images of people partying in a crowded bar may simply normalize such behavior for people who are skeptical of the precautions anyway. Such doubters may wonder: If others aren’t taking the virus seriously, why should I? Expressing moral outrage still has a role: Rand and colleagues have found that signaling such outrage makes people more trusted by their peers.9 But if we are all shouting into our own echo chambers about risky behaviors, shaming may better serve our own reputations than the collective welfare.
As incensed as Weiss felt when he read Ms. R.’s post, this intuition — that shaming her might do more to satisfy his own emotional needs than to persuade her to wear a mask — made him pause. So instead, he replied by describing the devastation he’d witnessed in New York, the two patients he’d seen die the previous day, and how, as much as he hated wearing masks, they were critical to reopening the world. “Trust me,” Weiss wrote. “It’s much worse to cry into a mask than it is to wear one.”
Though Ms. R., under attack, deleted the entire thread, she reached out to Weiss privately. Weiss, emphasizing that their commonalities exceeded their differences, noted that the defining focus of our era was not on our shared humanity, but on a relatively few sharp disagreements. Ms. R. expressed regret about the devastation Weiss had witnessed but noted her frustration that “doctors keep changing their opinions about what we should do.” While acknowledging that they would never agree politically, Ms. R. told Weiss, “If you actually told me that I really needed to wear that mask everywhere, I would.”
When I spoke to Ms. R. in July, she was in Delaware, where Covid cases were increasing, and she said she’d come around, though she still refuses to wear a mask on the beach. “I will wear a mask if it’s going to help,” she said. “But I’m not going to feel silenced.” It’s impossible to say to what extent Weiss — rather than evolving social norms — persuaded her, but she stressed that she trusted Weiss far more than other experts because he had an open mind and no agenda. When I asked what would increase her trust in medical expertise more generally, she said part of the problem was that “People think doctors know everything. But you’re just people.” Given the virus’s novelty and the persistent uncertainties surrounding it, Ms. R. didn’t expect experts to know everything. Under these circumstances, however, she wished they would admit when they were wrong and more readily acknowledge uncertainty. “I just wish the experts would say, ‘I don’t know.’”
In early August, a colleague of mine tweeted about the death of his father, a physicist, from Covid-19. “He was a man of science,” my colleague wrote, “and his death is the consequence of out of control viral spread from the rejection of it.” While I felt sad for my colleague and frustrated that so many deaths could have been prevented, I also felt uneasy. Was I betraying him by writing a piece attempting to understand, and thus humanize, science deniers?
I’d felt this discomfort before. During our July conversation, I asked Ms. R. if she wanted to remain anonymous. “Do you think I’ll come off like an idiot?” she asked. I was surprised by the question — if she recognized the potential idiocy of her views, why not modify them? — but I’ve struggled more with my hesitance to answer it. I had the power, based on our conversations, to make her seem idiotic, which would win me approval from my tribe. If, on the other hand, I chose parts of our conversations that made her seem reasonable, I risked alienating people, like my colleague, who not only believe in science but have been hurt by others’ rejection of it. Though I haven’t resolved it in my own mind, this conflict — between the desire to belong and the need to understand — seems central to our ongoing divides. For science to lead us out of this pandemic, however, we will need more than just the believers to follow.
Funding and Disclosures
Disclosure forms provided by the author are available at NEJM.org.
Identifying details have been changed to protect people’s privacy.
This article was published on September 23, 2020, at NEJM.org.
1. Funk C, Kennedy B, Johnson C. Trust in medical scientists has grown in U.S., but mainly among Democrats. Pew Research Center. May 21, 2020 (https://www.pewresearch.org/science/2020/05/21/trust-in-medical-scientists-has-grown-in-u-s-but-mainly-among-democrats/).
2. Gauchat G. Politicization of science in the public sphere: a study of public trust in the United States, 1974 to 2010. Am Sociol Rev 2012;77:167-187.
3. Merkley E. Anti-intellectualism, populism, and motivated resistance to expert consensus. Public Opin Q 2020;84:24-48.
4. Mehra MR, Ruschitzka F, Patel AN. Retraction — hydroxychloroquine or chloroquine with or without a macrolide for treatment of COVID-19: a multinational registry analysis. Lancet 2020;395:1820-1820.
5. Mehra MR, Desai SS, Kuy S, Henry TD, Patel AN. Retraction: cardiovascular disease, drug therapy, and mortality in Covid-19. N Engl J Med. DOI: 10.1056/NEJMoa2007621. N Engl J Med 2020;382:2582-2582.
6. Nichols T. The death of expertise: the campaign against established knowledge and why it matters. New York: Oxford University Press, 2017.
7. Zakaria F. Fareed’s take: the Covid-19 divide is a class divide. CNN. 2020 (https://www.cnn.com/videos/tv/2020/05/17/exp-gps-0517-fareeds-take.cnn).
8. Ross L, Nisbett RE. The person and the situation: perspectives of social psychology. Kindle ed. New York: McGraw Hill, 1991.
9. Jordan JJ, Hoffman M, Bloom P, Rand DG. Third-party punishment as a costly signal of trustworthiness. Nature 2016;530:473-476.