IMPORTANT NOTE: I wrote this email to Amy Harmon on January 25, when it wasn't yet clear to me that a Wuhan coronavirus pandemic was likelier than not. I focused far more on how to reassure U.S. citizens who were "overreacting" than I would today. Now I think it is far more important to prepare the U.S. public for the not-unlikely prospect of considerable illness and disruption. (I am writing this note on February 6; it's possible that now I'm overreacting!)
I decided to post this email anyway, because what I had to say about risk communication basics, over-reassurance, overconfidence, stigma, etc. is at least as relevant today as it was when I wrote it 12 days ago.
Risk Communication Basics
I know you want to write about Wuhan CoV, not risk communication theory. But whether or not you can find a way to include any theory in your article, I feel an ex-professor’s need to start with the big picture. I promise to get to story-telling. So bear with me.
The fundamental truth of risk communication is the exceedingly low correlation between how dangerous a risk is and how upsetting it is. Because this correlation is so low, there are three kinds of risk communication:
- When people aren’t upset enough – think flu – the task is to get them more upset. The reason to get people more upset is to motivate them to take appropriate precautions, so I call this kind of risk communication precaution advocacy.
- When people are too upset – a vaccine-hesitant parent in the pediatrician’s office, say; or the worried neighbors of a properly managed waste site – the task is to calm them down, so they won’t take or demand precautions that aren’t wise. I call this kind of risk communication outrage management.
- When people are rightly upset – as the populace of Wuhan, China is right now – the task is to guide them through the situation, to help them bear what has to be borne and make wise decisions about which precautions to take or demand. That’s crisis communication.
Which kind of risk communication officials ought to be doing depends on how serious those officials think the risk is and how upset they think the public is. Obviously, that can vary from one audience to another, and the situation can change over time.
Wuhan CoV Outrage Management
Right now, Wuhan CoV presents all three risk communication challenges at the same time. In Wuhan itself, the principal task is crisis communication: guiding the public through an awful situation. In other Chinese cities, where sizable outbreaks are considered likely or even inevitable, the main task is probably precaution advocacy: getting the public ready for what’s probably coming. As soon as people realize what’s probably coming, or as soon as it comes, the task shifts from precaution advocacy to crisis communication.
But in the U.S. there is certainly no crisis now – so there is zero need here for crisis communication. And based on their current assessment, most experts do not expect a Wuhan CoV crisis in the U.S. (Note however that the current assessment could change rapidly and dramatically.) There will be more imported cases, and there may well be small local outbreaks, perhaps even a handful of U.S. deaths. Outbreaks among healthcare workers or transmission to other patients would certainly represent local crises. But New York and Chicago and Los Angeles are unlikely to go the way of Wuhan. So the main risk communication task here is outrage management – aimed at the subset of the American public who are overreacting or likely to overreact.
We don’t know yet how big that subset is. But even if it’s tiny, it will be significant. The most worried members of the public will, as always, get the most attention from media, from politicians, and even from public health officials. During the 2014 Ebola epidemics, for example, many reporters built their stories around one lady in an airport wearing homemade aluminum foil “protective” gear. (For an example with photo, see https://www.smh.com.au/world/picture-of-woman-in-homemade-hazmat-suit-goes-viral-20141018-1182td.html.) The Internet exacerbates the excessive visibility of the over-concerned minority. No need to make up quotes; you can find as many apparent hysterics and paranoids as you want online.
And the over-concerned minority may not be that tiny. Wuhan CoV has characteristics that make it susceptible to a temporary overreaction on the part of normal people who are not routinely hysterical or paranoid. Many of these characteristics will lose their potency over time. People have an “adjustment reaction” to scary new diseases, and then they get used to the new normal (the new disease). The long-term risk communication challenge is usually precaution advocacy, not outrage management.
But at the start, there’s going to be need for outrage management about Wuhan CoV because of factors like these:
- It’s new – literally a pathogen that never existed before in the human population (as far as we know).
- It’s mysterious – lots of scientific unknowns.
- It has its origins in a distant place with unfamiliar customs – how many Americans eat wild animals or patronize wet markets?
- There’s no vaccine and no cure – nothing much that people can do to assert control.
- There are abiding issues of trust – grounded only partly in Chinese dishonesty about SARS in 2002-2003.
- Officials and the media often mock and ridicule people whom they think are excessively upset, insulting the very people they are trying to calm – thereby unintentionally prolonging and exacerbating the adjustment reaction.
Wuhan CoV Precaution Advocacy (Really Preventive Outrage Management)
The less obvious U.S. risk communication challenge about Wuhan CoV is a weird sort of precaution advocacy. It’s weird because with so little Wuhan CoV in the U.S. so far, there is not much precautionary action for most people to take. But it’s important to forewarn the public that the situation is fluid – that there will almost certainly be additional imported cases; that there may be some small local outbreaks; and that there’s a small but nonzero possibility that Wuhan CoV will turn serious in the U.S. It is even remotely conceivable that we’re in the early days of a worldwide pandemic crisis that will engulf the U.S. along with everyone else. So the precaution to be advocated is watchful waiting and learning.
Perhaps this is also a teachable moment for all-hazards preparedness. For instance, recommend that people gradually stockpile basic supplies that may be hard to get when a hurricane/ice storm/scary outbreak gets closer.
One main purpose of Wuhan CoV precaution advocacy is to get the public emotionally and cognitively ready for the unlikely possibility that the situation turns into a crisis in the U.S. But there’s another purpose that’s just as important: to help people avoid overreacting to the small increases in seriousness that are likely. So this kind of precaution advocacy is actually preventive outrage management. You warn people about the ways in which Wuhan CoV is likely to get worse so that when it happens, people are less prone to imagine the very worst and thus less inclined to take or demand excessive precautions.
Most emerging infectious diseases don’t become worldwide crises, much less worldwide disasters. A few times in my career, I thought I was working on one that probably would. Jody and I even wrote articles outlining the key crisis communication messages to use if H5N1 (bird flu) went pandemic … which (so far) it never did.
Over-Reassurance
I think the single most common error officials make when talking with the public about emerging infectious diseases is their failure to prepare people for likely small increases in seriousness and thereby inoculate them against overreacting. So the key U.S. message ought to be that we will probably see more cases and even some local transmission, but we probably won’t see anything like what Wuhan is enduring.
Importantly, this key message needs to come across as more warning than reassurance. That’s a central paradox of outrage management. If people feel that officials are trying too hard to “calm” them, they become less calm. Over-reassurance backfires; people feel patronized and infantilized, and trust declines. “Don’t worry, we’re not likely to have more than a few cases and maybe some local transmission” has the right facts but the wrong tone. “Unfortunately, we are likely to have at least a few cases, and maybe some local transmission” is better.
The goal is to give people the information they need to understand why they shouldn’t be too upset (at least not yet), while simultaneously giving them the impression that you’re taking the risk very seriously indeed. If you get it exactly right, most of your public thinks you’re overreacting just a bit. A very concrete version of this rule-of-thumb: Put reassuring information in a subordinate clause of an otherwise alarming sentence. “Even though all the cases of Wuhan CoV in the U.S. so far have been travelers from China, we are preparing for the possibility of local transmission.”
In addition to that key message, officials should also warn people now that way short of becoming Wuhan, we could conceivably have ghastly hospital outbreaks like Toronto’s during SARS. When Ebola came to Dallas in 2014, then CDC head Tom Frieden said some regrettable things about how well-prepared U.S. hospitals were, calling Ebola “easy to stop by using gloves and barrier precautions.” Frieden pretty much claimed that regular hospital infection control procedures were sufficient to keep the disease from spreading. So when two nurses who treated an infected Liberian patient were infected, it was a very big deal. People felt not just frightened but blindsided and angry.
Frieden’s bigger point, that the U.S. was vanishingly unlikely to have an Ebola epidemic, was solid and important. But it was lost in his failure to sound alarmist enough about the likelihood that the U.S. would have a few cases – which it did – and that those cases wouldn’t necessarily go smoothly.
So when I read news stories quoting U.S. officials (especially hospital officials) about how ready we are for Wuhan CoV, I worry that they may be falling into this trap.
Of course sounding too alarmist also has downsides when you’re trying to calm a public you think is already too alarmed. But in the dialectic between being over-alarming and being over-reassuring, it’s not damned if you do and damned if you don’t. It’s damned if you do and darned if you don’t. Over-reassurance is the cardinal sin of outrage management – the one you’re damned if you do. Over-alarm is far more survivable for public officials.
When officials make over-reassuring statements, it’s typically because they’re trying to calm an overly alarmed public. This isn’t just a strategic error, since over-reassurance usually backfires. Quite often it is also a diagnostic error, since the public is likely to be a lot less alarmed than officials imagine. Officials routinely perceive the public as panicking and hysterical when the public is merely paying a huge amount of attention to a novel situation with a lot of uncertainties and unknowns. This diagnostic error is behind a lot of risk communication missteps. It leads not only to over-reassurance, but also to overconfidence, to expressions of disdain and contempt for the public, and to cover-ups of potential or actual bad news.
Officials usually fall into the over-reassurance trap out of good intentions. They want to calm what they perceive as excessive public fears. The key point to remember is that even when the public is genuinely more fearful than it ought to be, over-reassurance doesn’t calm excessive public fears. It leaves people alone with their fears, and thus exacerbates their fears.
It is more empathic and paradoxically more reassuring to validate people’s fears than to disparage them. Compare these two:
- “Of course many people are worried about Wuhan CoV. It’s new and mysterious. And nobody knows how bad it might get.”
- “Any American who worries about Wuhan CoV but doesn’t bother to get a flu shot has his or her priorities totally wrong.”
They’re both true. But I’d lay odds the second example will leave people more anxious about Wuhan CoV than the first – and no likelier to get their flu shots.
Not all over-reassurance originates with the good intention of calming excessive fears. Sometimes the goal is to protect the reputations of local officials or the profits of local industries – tourism, for instance. These goals, too, are better served by empathic candor than by over-reassurance.
Toronto was one of the four cities in the world hardest hit by SARS. A number of high-ranking Toronto officials struggled to pretend that the outbreak was less severe than it was. James Young, Ontario’s Commissioner of Public Security, said, “The streets of Toronto are safe from SARS. They are as safe as the streets of London, Paris, or Washington.” The SARS risk in Toronto wasn’t mostly on the streets. It was in the hospitals. At the height of Toronto’s SARS outbreak, the rational thing to do if you broke your leg in Toronto was to get into a taxi and head to another city for medical care.
Toronto’s mayor made such an enraged protest against the World Health Organization’s travel advisory – recommending against unnecessary travel to that city – that WHO prematurely withdrew the advisory. As a result, visitors and conventioners returned to Toronto while it was incubating its second wave of SARS. Some of them caught SARS as a result.
Singapore was also one of the four hardest-hit cities, but it handled the travel advisory situation differently. The same day WHO lifted Toronto’s travel warning, it also said that the worst of Singapore’s SARS outbreak seemed to be over. Singapore Health Ministry spokeswoman Eunice Teo responded with a statement that Jody and I found stunning, the absolute opposite of over-reassurance. “The WHO said the peak is over in Singapore,” she noted, “but our minister has said it is too early to tell.” While Canada celebrated the lifting of the travel advisory, Singapore was advising potential visitors to be more cautious than the World Health Organization.
Toronto cared about its tourism, while Singapore cared about its tourists. I haven’t verified whether tourism in Singapore rebounded more quickly than tourism in Toronto. But I think so. And I certainly hope so.
Overconfidence
Overconfidence is almost as bad outrage management as over-reassurance. Virtually everything the experts think they know about an emerging infectious disease is tentative. Some of what they believe to be true will turn out wrong, and as a result, some of what they decide to do will turn out ill-advised (or at least suboptimal). Wise experts emphasize the uncertainty of their preliminary judgments and decisions. David Heymann, who managed SARS for the World Health Organization, captured this brilliantly when he said “We are building our boat and sailing it at the same time.” Really wise experts go a step further and predict their own errors and missteps. Jeffrey Koplan, CDC director during the 2001 anthrax scare, said at one point, “We will learn things in the coming weeks that we will then wish we had known when we started.”
Both Heymann and Koplan made their points about uncertainty and tentativeness with authority. They didn’t sound unable to cope; they sounded like coping with uncertainty and making decisions that might turn out wrong was their stock-in-trade (as indeed it was).
This is the goal: confidently expressing how unconfident you are. The opposite combination is the one that most erodes public trust: sounding unsure of yourself as you claim to know what you’re doing.
Acting CDC Director Richard Besser achieved a virtually ideal merger of empathic candor and authoritative tentativeness in the early days of the 2009–2010 swine flu pandemic. Swine flu turned out quite mild, but when it first emerged in April 2009 it was scary. Besser’s press briefing began, “First, I want to recognize that people are concerned … and we are worried as well. Our concern has grown since yesterday in light of what we’ve learned since then.” Then he turned to uncertainty. “At the early stages of an outbreak, there’s much uncertainty, and probably more than everyone would like. Our guidelines and advice are likely to be interim and fluid, subject to change as we learn more.” Bravo!
Official statements about emerging infectious diseases traditionally sound overconfident for two reasons. The first is the officials’ fault: They say overconfident things. But a big piece of the problem is everybody else’s fault. Journalists systematically simplify official statements because they must – but one of the main things they “simplify out” is any acknowledgment of uncertainty. And what uncertainty makes it into news stories tends to get simplified out by readers and viewers, who systematically remember what they read or saw as more definitively one-sided than it actually was.
I am watching this mistake happen right now with official and expert statements about Wuhan CoV. Some academics have issued preliminary estimates of key Wuhan CoV metrics – its case fatality rate, its contagiousness, its incubation period, etc. The original statements are clear that the estimates are highly uncertain. The best journalists report the uncertainty. But most write stories that make the experts sound a lot more definite than they sounded in real time. And within a day or two these uncertain, preliminary estimates have been reified, and I’m reading and viewing news stories that claim to tell me what the real numbers are.
So reluctantly acknowledging uncertainty usually doesn’t do the job. Officials must proclaim their uncertainty. Better yet: They should exhort reporters not to misrepresent today’s tentative preliminary guaranteed-to-change numbers as definite.
Local Decision-Making
Your email asks in particular about local decision-makers diverging from CDC “dictates,” and gives the example of school closures. So let me turn to this cluster of issues.
First of all, as I suspect you know, there are no CDC “dictates” vis-à-vis the public health decisions of states and localities. For better or worse, the federal government has very little power over these decisions. The CDC informs, helps, funds, and advises – but governors and state legislatures, mayors and city health departments make the actual decisions. I see this local control as a feature, not a bug. The fact that different state and local governments respond differently to the same set of facts is the essence of our constitutional republic. Most of the time, state and local governments decide to do what the CDC recommends. The exceptions may turn out well or badly, and everyone gets to see what works.
When U.S. volunteers started returning from the 2013–2016 West African Ebola epidemic, the CDC recommended a fairly lax protocol to watch for symptoms of Ebola. Mostly, volunteers were instructed to “self-monitor,” and to notify the authorities if they started to feel sick. A few states took a more cautious approach. Most famously, New Jersey isolated returning nurse Kaci Hickox in a hospital for several days after a Newark Airport screening initially registered a fever that she insisted was illusory – and that indisputably turned out not to be Ebola. The controversy over whether returning volunteers should be quarantined became highly politicized. Most experts took the view that quarantine was unnecessary, arguing that volunteers could be trusted to recognize and report their symptoms and that they were extremely unlikely to be infectious before those symptoms began. The pro-quarantine position was widely ridiculed as right-wing unscientific demagoguery.
Interestingly, however, many hospitals and medical schools would not let returning volunteers return to patient care until they had been back and symptom-free for several weeks – a very quarantine-like policy. And the CDC’s recommended protocol for returning volunteers went through several revisions, all in the direction of greater caution. The CDC ended up with guidelines closer to New Jersey’s policy than to its own original lax protocol.
In short, local divergence from CDC recommendations is a right, a tradition, and a valuable source of new information.
Precautions are themselves a useful kind of risk communication. When the public is insufficiently concerned, officials can take precautions as a role model – for example, by publicly getting a flu shot every year. And when the public is excessively concerned, officials can take precautions as a way to calm the public’s concern. By itself, reassurance theater isn’t an acceptable reason to restrict someone’s civil liberties. But if the medical wisdom of a precaution is debatable, as I think was the case for Ebola quarantine, then it’s worth noting that a quarantine-like policy can reassure returning volunteers’ neighbors or patients, and thus ease their reintegration. This is largely why many returning volunteers self-quarantined, often on the advice of the organizations they had volunteered for.
School Closings and Giving People Options
School closings illustrate many of these points. The decision to close local schools is quite properly a local decision, grounded in local realities, some of them nonmedical – for example, how many students and teachers are going to be absent anyway, either because they’re ill or because they’re anxious. Even medically unnecessary school closings can be profoundly empathic. At the height of Singapore’s SARS outbreak, Prime Minister Goh Chok Tong told his people that he had decided to close the schools. His health minister, he openly explained, was quite confident that there was no medical reason to do so. But many of his other ministers said that as parents and grandparents they would be uncomfortable sending their kids to school when there was so much SARS around. To Goh, that was reason enough to close the schools. And while the schools were closed, the government developed a SARS precaution education curriculum for every age group, ready to roll when the schools reopened.
I have long advised clients to bracket their precautionary recommendations with options that are more and less precautionary – especially in situations where uncertainty is high. “At least do X. We recommend Y. If you want to go further, consider Z.” I think it’s useful to have fallback advice (X and Z) for people who choose not to heed your core advice (Y). People are likelier to comply with a recommendation if it’s bracketed between a less burdensome precaution and a more protective one. And if they pick one of the other two, you haven’t defined them as misbehaving, so they’re probably likelier to pay attention to your future warnings.
I wish the CDC would do this more often, on virtually every issue – from the childhood vaccination schedule to Ebola quarantine.
Z’s – additional precautions people can take if they’re more concerned than you think necessary – are invaluable tools of outrage management. Here’s another good example from Singapore’s SARS experience. The health ministry thought wearing masks during ordinary activities wasn’t necessary or even especially desirable. But many people wanted to wear masks. The ministry told people it didn’t recommend masks in public, but it also explained how masks should properly be worn, for the benefit of those who chose to wear them anyway. That was so respectful! In many other countries, officials (and reporters) ridiculed people who wore masks. Canadian Airlines even forbade flight attendants from wearing them on trans-Pacific flights because it might scare the passengers.
If you want people to stay calm, why would you put unnecessary barriers in the way of their doing things that help them stay calm?
Returning College Students and Stigma
The issue of returning college students is more like the Ebola quarantine issue than it is like K-12 school closings. Inevitably, some people are going to feel anxious around people who have recently arrived from Wuhan (or from any epidemic epicenter). We can call that stigma, but we can’t claim it’s completely irrational. True, there’s no evidence (yet, anyway) that asymptomatic people can transmit the disease. But who knows? And since nobody wants to be quarantined, there is ample reason to think that some people won’t accurately report their early symptoms – especially since the early symptoms of Wuhan CoV are cold-and-flu-like. Students will be inclined to believe, usually correctly, that they don’t have Wuhan CoV.
The risk of catching Wuhan CoV from an apparently healthy fellow student just back from visiting family in Wuhan is pretty small. But it’s obviously not as small as the risk of catching Wuhan CoV from a fellow student who’s never been to Wuhan. And in other contexts we support precautions against small risks; the risk of catching measles from an asymptomatic fellow student who isn’t vaccinated is also small, even during a measles outbreak.
Nonetheless, it is unfair to expect one student to stay away from campus for weeks because some other students (or their parents) are worried – more worried than their actual, small risk justifies.
I don’t have a solution to this dilemma. But risk communication does have a piece of the solution: Stigma gets worse when the stigma itself is stigmatized.
Even though the U.S. experienced very few SARS cases, many Chinese restaurants in the U.S. experienced a noticeable temporary decline in business during SARS. People figured, “Okay, maybe the risk is small, but it will be even smaller if we eat Mexican tonight.” The decline didn’t last long. I think things would have gotten back to normal even sooner if this temporary decline had been accepted with less vituperation – not endorsed, just tolerated as a natural and fortunately evanescent overreaction. Calling people irrational – or worse, calling them racists – makes them dig in. In their own lived experience, they’re simply taking a reasonable precaution.
By the way, at the same time the general public was being castigated for avoiding Chinese restaurants, one subset of the public was doing the same thing: residents of Chinatowns around the country. They knew better than anyone that many local people had recently returned from visits home to China.
On balance, I don’t think it would be wise for colleges to restrict the movement of students who have traveled to Wuhan. I’d be okay with some kind of surveillance – texting them every afternoon for a quick symptom check, maybe. But I suspect the bigger issue won’t be how colleges treat students recently back from Wuhan. It will be how colleges treat those who are worried about students recently back from Wuhan. I urge empathy, not ridicule.
Of course if Wuhan remains shut down, in a few weeks there won’t be a lot of students recently back from Wuhan. And if CoV starts spreading significantly in the U.S., colleges will face tougher problems, and so will the rest of us.
Postscript on Zika Precaution Advocacy
You may notice that I have said nothing about Zika in this email. That’s because I am focusing on the Wuhan CoV risk communication task I expect to be paramount in the U.S. – outrage management, the effort to prevent or ameliorate excessive concern. In the judgment of the CDC and most public health professionals, Zika wasn’t an outrage management problem. They judged it to be a precaution advocacy problem. Their task as they saw it was to arouse more concern in an unduly apathetic public.
The dominant risk communication challenges are completely different when you’re trying to scare people than when you’re trying to calm people.
For what it’s worth, I never understood the judgment that Zika called for a precaution advocacy approach. Zika was a very serious disease indeed for the baby of any pregnant woman who caught it. But not many pregnant women in the continental U.S. caught Zika. And not many were ever likely to catch Zika. I saw the need for Zika precaution advocacy in Puerto Rico and the Virgin Islands, where the disease was widespread. I saw the value of Zika precaution advocacy aimed at people contemplating travel to Latin America or the Caribbean, where the disease was widespread. But elsewhere in the U.S., I didn’t see the point.
A great deal of Zika precaution advocacy in the U.S. struck me as overly alarmist and not entirely honest. In New York City, for example, a virtually nonexistent local Zika threat was used to justify a mosquito spraying program, the real goal of which was to tamp down West Nile Virus. At the CDC, intentionally misleading maps were produced that suggested Zika-carrying mosquitoes were a significant risk in parts of the country that rarely saw even a single mosquito of the right species. And in Congress, the battle over Zika funding was far more a political exercise than a public health effort.
But these are different issues not likely to apply to Wuhan CoV.
Copyright © 2020 by Peter M. Sandman
