Addressing the Coronavirus Threat From a Risk Manager's Perspective
One political leader in the world has delivered an honest speech about a rational policy to respond to the novel coronavirus. Boris Johnson’s approach may prove to be wrong, but it’s the only one so far proposed that isn’t doomed from the start. Of course he’s being pilloried for it, and he gives his critics ammunition by not releasing the detailed assumptions behind it.
The coronavirus is a major danger with gigantic uncertainties. In this circumstance there is a temptation to do nothing on the grounds that we don’t know enough to choose helpful policies from perverse ones. Another temptation is to take dramatic action of uncertain value because, “after all, we can’t stand aside and do nothing.” This temptation is irresistible to politicians, because “wait until we have a clear enough understanding to know what’s helpful,” undercuts the myth that they are handling the crisis and gives insecure voters the impression that politicians do not care or are incompetent.
The result is that most places adopt a raft of expensive reactions with vague rationales. These policies may help or hurt, but they don’t constitute a rational response. They can conflict with each other. They always include a bunch of popular stuff—mainly giving people money—not clearly related to the problem. Because the assumptions and goals are not specified in advance, after the arrow lands proponents will paint a target around it and claim success; while opponents will paint a faraway target and claim failure. We don’t learn much about what works and what doesn’t, but we will increase political divisiveness.
I don’t know the best response to the novel coronavirus. I'm a risk manager, not a public health expert. But as risk manager I can say the first step is to decide the size of the expected problem. At this point, a medium-case outcome seems to be that about half the US population is infected by the virus over the next 12 months, and that a million die as a result. Things could be much better or much worse, but we have little control over that. It depends on the virus, and how it mutates, and how it responds to warming weather, how much immunity infection confers and other matters of biology, not policy. In the longer term vaccines and improved treatments can matter, but these are unlikely to make much difference in 2020. Incidentally, the distribution is skewed, meaning that the outcome will probably be better than a million deaths, but it could plausibly be much worse, with two or ten or more times that number.
So you can’t say a million deaths is unthinkable. It has to be thinkable for rational planning. That means we should be willing to consider policies that will likely lead to hundreds of thousands of deaths, something we would never do in normal times. And we have to give serious weight to policies with high costs that reduce the probability of unlikely extreme scenarios.
The second step is to focus on what we can control. Attempts to reduce the gross exposure rate to the virus--how many people are exposed eventually--are probably futile given the kinds of solutions likely to be implemented and enforced for a year or more. From what we know, the virus is already too widespread, and carriers remain infectious for too long, and society is too free, to contain it. Attempts to improve treatment to reduce the death rate are important, of course, and should be funded; but relying on them in the short-term is more “wishin’ and hopin’” than a plan. And waiting for warmer weather or the virus to mutate to a less-lethal form that provides immunity is pure wishin’ and hopin’.
Much of the virus response seems to be based on the unspoken goal of containment and elimination--as happened with SARS and MERS. This seems unrealistic for SARS-CoV-19. Strict measures to control transmission that are not likely to be maintained for six months or longer, can be likened to taking an antibiotic until symptoms disappear and then stopping. This does nothing to change the eventual outcome of the disease in an individual, and leads to the evolution of antibiotic-resistant strains that transmit more readily and remain nonsymptomatic longer. The analogy is not exact, but the point is valid, that measures to slow infectious diseases are often counterproductive when considered more broadly. What Emerson wrote about kings, "When you strike at a King, you must kill him," has some application to public health as well.
Closing schools is a particularly problematic response. It can make sense for influenza, because children are superspreaders, and a brief shutdown can outlast the outbreak. But from what we know of SARS-CoV-19, it seems to mostly bypass children and is likely to remain a threat for many months if not indefinitely. While we have no real idea, it seems at least as likely that closing schools will expose children to more danger than less, and we know the disruption in parents' lives has negative consequences that can lead to more infections. On top of that is the cost in money and education. It could be a good idea, but no one promoting it has provided clear assumptions to justify the policy, nor even clear goals of the policy.
In plausible models, about half the deaths occur among at-risk individuals—older or with other health issues—who become symptomatic during a peak in which the healthcare system is overwhelmed. At the moment, this seems likeliest to occur around June and July. This is something we can influence, mainly by infecting more at-risk people now. Infecting healthcare workers now could also help by providing more healthy immune healthcare workers when we need them.
Should we be running “Coronavirus Cruises” for the elderly and adequate healthcare workers to care for them? Should be invite at-risk people to “You Bet Your Life” parties at the currently empty casinos? No leader is going to remain in office with these kinds of proposals, but they’re not crazy. Increasing the population of immune people, especially immune healthcare workers, keeping infected people quarantined together to reduce general population spread, and reducing the peak load on the healthcare system are all good things. They may not outweigh the bad things, especially because we don’t know how much immunity is acquired by being exposed to the virus, and how the virus is going to mutate. But the point is efforts to speed the spread of the virus in controlled circumstances might be good, and therefore broad-based efforts to slow the spread might be counterproductive.
Consider it from the individual’s point of view. I am a marginal at-risk person who might a 1% chance of death if infected with good healthcare, and perhaps has about a 50% chance of getting the virus in the next year. If my chance of death is higher than 2% if I am infected during a peak, I might rationally decide to deliberately infect myself now. It’s like a dangerous vaccine.
I’m not going to opine on the best policy reaction to the novel coronavirus as I lack the epidemiological knowledge. But I do claim expertise on how to organize reaction to big uncertain dangers. You start with clear thinking about the expected case, you focus on what you can control, and you make rational decisions based on clearly stated assumptions and goals—which you monitor skeptically and update quickly as events unfold. This doesn’t guarantee success, it may not even help much, but it makes you part of the solution rather than part of the problem.
Unfortunately, we have a political system in which power is achieved by promising unrealistic foresight and control. “I will keep you safe,” is a winning slogan. “I’ll evaluate all the evidence and take steps when I’m confident of doing more good than harm,” is a losing one. Dramatic steps and money giveaways are popular, much more so than incremental improvements, inconvenient rules and counterintuitive solutions. Boris Johnson is a lonely beacon of rational hope in a world of dueling unrealistic promises and random policies. His policy choices may prove to be poor ones, but at least he's not pandering to the crowd.
