There is still much we don’t know about the COVID-19 pandemic, which has brought most of the world to a standstill to slow the spread of the virus. However, no matter where exactly we are headed, we can already say this is no flu.
In France, as of April 7, more than 10,000 people who tested positive for the Wuhan coronavirus had already died. Perhaps even more worrisome, more than 7,000 people are currently in intensive care, where the fatality rate is very high, so many more people are going to die. By comparison, during the worst flu season in the past 10 years, only 2,922 people were admitted to ICUs because of the flu.
Not only are we already way past that with the COVID-19 epidemic, but even in the most plausible optimistic scenario, those figures will be at least 10 times as high despite putting the entire country on lockdown, which will probably last for at least a month and a half. By the time the number of people who require hospitalization reaches a plateau, most French hospitals, and ICUs in particular, will be full, and we’ll have to wait for them to empty before we can slowly go back to a more normal life. At this point it’s still unclear how much worse than the flu this virus is.
The Scientific Models Are Uncertain
On March 16, a team of epidemiologists from Imperial College in London published a report about the results of simulations they had done, which predicted a disaster of epic proportions unless economy-crippling measures were taken. According to their simulations, if nothing were done to slow or suppress the epidemic, more than 2 million people would die in the United States alone.
Even a mitigation strategy, aimed only at slowing down the epidemic without trying to suppress it, would only cut this number in half in the best-case scenario. Only a strategy aimed at suppressing the epidemic, including generalized social distancing, would be able to keep the death toll to a more reasonable number. Those measures, however, would have to stay in place until a vaccine or some other pharmaceutical intervention was available, which could take more than a year.
However, as I argued more at length in an article where I dug into the model they used, the results of those simulations are dubious. Given the amount of uncertainty surrounding some key parameters of the model, such as age-specific infection fatality rates and hospitalization rates, which are bound to have a huge effect on the results, it would be unwise to rely on those simulations for decision-making.
A similar point could be made about much simpler models, although they can be useful to get a better sense of what a worst-case scenario would look like to plan our response. As I argue below, as long as we remain in this state of uncertainty, the right thing to do is to assume the worst and prepare accordingly.
After I published this analysis and J.D. Vance shared it on Twitter, Tom Nichols admonished him on the ground that I wasn’t an expert. It’s ironic that Nichols, a self-appointed expert on expertise who constantly opines on topics he knows nothing about (which I’m afraid is most of them), should criticize someone for talking about something outside his area of expertise.
On the other hand, Nichols’ own area of expertise is supposed to be international affairs, yet he famously supported the invasion of Iraq in 2003, among other calamitous decisions. So perhaps it’s not such a bad thing if he sticks to other issues.
By dismissing my article on the grounds that I’m not an epidemiologist, Nichols was suggesting that someone may not have anything interesting to say about a topic unless he is a professor in that subject or has similar credentials. This form of credentialism is very common, particularly among the educated class. Of course, it’s true that many and probably most non-specialists who talk about something outside their area of expertise say a lot of nonsense, but non-specialists can also make interesting contributions.
Look at the Track Record
Academic credentials are important, if only because they provide a useful heuristic to learn more about a topic. Since you have to start somewhere, it makes sense to listen first to the people who are paid to teach and research that topic. Ultimately, however, the validity of scientific claims depends on the arguments and evidence in their favor.
Thus, if someone makes a claim he can’t substantiate with good arguments and evidence, his credentials should not matter. Conversely, even if someone has no credentials, without evidence of substantive incompetence, this is not a reason to reject his claims.
Of course, non-specialists are often not in a position to determine whether an expert has provided good arguments for his claims or whether a non-expert has demonstrated substantive incompetence. But this is not always the case.
There are many reasons one should not blindly trust experts or dismiss analysis by non-experts. Primarily, experts can be and often are wrong. Everyone makes mistakes, so this should obviously not be disqualifying — otherwise, nobody would be qualified to talk about anything. This is why it’s stupid to dismiss experts just because they have sometimes been wrong, as many people unfortunately do.
But it’s just as stupid to continue to trust experts with a track record of being systematically wrong. On some topics, such as foreign policy, many experts whose input is still taken into account by decision-makers and the media have precisely that kind of track record, so unfortunately this is not theoretical. People like Nichols want you to ignore their track record and continue to take them seriously just because they have the right credentials. You shouldn’t.
In the case of this pandemic, I don’t think we’re in that situation, but it’s hard to deny that many public health experts have been horribly wrong in the past few months. I have no desire to defend President Donald Trump, whose response to this crisis I have found atrocious, but as Zeynep Tufekci recently pointed out, he’s hardly the only one who underestimated the seriousness of the threat.
I have lost count of how many public health experts went on record in the past few months to say the virus was not a big deal and that people in the West had nothing to fear. Meanwhile, China was putting a region containing almost as many people as France on lockdown and bringing the rest of the country to a standstill because of this virus.
Needless to say, this episode will not be remembered as a great moment for expertise. Unfortunately even today, some experts continue to peddle nonsense, such as this statement by the World Health Organization.
If you do not have any respiratory symptoms, such as fever, cough, or runny nose, you do not need to wear a medical mask. When used alone, masks can give you a false feeling of protection and can even be a source of infection when not used correctly. ➡️https://t.co/0nyHux1SgN pic.twitter.com/JDQhnowx3p
— World Health Organization (WHO) Western Pacific (@WHOWPRO) March 26, 2020
Question the Experts
Meanwhile, in my corner of Twitter, many smart people, who have no particular expertise in epidemiology or public health but who do have common sense, were observing what was happening in China and sounding the alarm several weeks or even months ago. This illustrates that sometimes non-experts are right even when they disagree with experts, and while it’s a complicated question exactly how people’s credentials should be weighted in public debate, this should at least rule out the kind of naive appeal to authority that passes as intelligence among people like Nichols.
The fact that many experts even today are still wrong about what is going to happen should also be clear from the fact that, as a recent survey of experts showed, they disagree with each other a lot about that. Since it’s not possible that, for example, the epidemic is going to kill both 5,000 people and 2 million people in the United States, some of them have to be wrong.
Only three out of 18 were able to correctly predict the number of cases in the United States at the end of March, and one of them only did because he chose an absurdly wide confidence interval. A few have already been proved wrong about how many people will die in the United States by the end of the year.
Incidentally, the wide range of estimates that experts have proposed sits pretty well with the conclusion of my article about Imperial College’s simulations, which is that nobody really knows what is going to happen because the data we have is difficult to interpret, generally of poor quality, and difficult to reconcile within itself.
Although the uncertainty is real and should be acknowledged, it is also not a reason not to act. The economic consequences of shutting down a country can’t be ignored, and I disagree with people who claim that as long as lives are at stake, we should not care about the economy. Economic crises also destroy lives, even if not in the same way as epidemics. Nevertheless, we should assume the worst and prepare accordingly.
This virus may turn out to be less dangerous than we fear, but we already have more than enough evidence something very bad is going to happen unless we take strong measures to prevent it. In my opinion, this warrants putting everyone on lockdown for at least two weeks, at which point we can reevaluate the situation in the light of the evidence that will have emerged.
That should give us a better understanding of exactly how dangerous this virus is. If we do that and the virus turns out to be less dangerous than we feared, it will be bad, but not nearly as bad as if we don’t and the virus really is as dangerous as the best current estimates suggest.