As communal life comes screeching to a collective halt for the indefinite future (including not-quite-Italy-but-close bans on movement in San Francisco), and as public health officials go into “more lockdown is better” mode, it seems important to underline at least three things: (1) no one knows what they are doing, hence (2) these measures are a rough approximation of acting according to a maximin principle, and (3) they are not socially sustainable. I'll conclude with what strike me as the only two currently-plausible off-ramps.
Let’s start with (1). John Ioannides makes the best case I’ve seen for the uncertainty behind what we’re doing in a recent piece on StatNews. The depth of our collective ignorance is astonishing. Before we even get to Ioannides’ argument, note that (a) Everyone knows about the catastrophic, Trump-induced failure to get testing and surveillance going early on, and that this failure cannot ever be remedied because we will never again be in those early stages. Trump’s and FOX News’ collective failures here should be treated as cases of criminal negligence. But notice that this means that we have absolutely no idea about basic facts about the disease prevalence. (b) Even if we had good testing, recent data says that lots and lots of COVID-19 is escaping detection. How many cases? We don’t know, of course, because these are undetected cases. They may not be as infectious. But whatever the number, it profoundly affects everything from the case fatality rate to how long a hypothetical “herd immunity” would take to develop. (c) As Ioannides points out, a lot of the data we do have isn’t worth much. For example, the mortality data is meaningless. I’ll let Ioannides give you a sense of why:
“The one situation where an entire, closed population was tested was the Diamond Princess cruise ship and its quarantine passengers. The case fatality rate there was 1.0%, but this was a largely elderly population, in which the death rate from Covid-19 is much higher. Projecting the Diamond Princess mortality rate onto the age structure of the U.S. population, the death rate among people infected with Covid-19 would be 0.125%. But since this estimate is based on extremely thin data — there were just seven deaths among the 700 infected passengers and crew — the real death rate could stretch from five times lower (0.025%) to five times higher (0.625%). It is also possible that some of the passengers who were infected might die later, and that tourists may have different frequencies of chronic diseases — a risk factor for worse outcomes with SARS-CoV-2 infection — than the general population. Adding these extra sources of uncertainty, reasonable estimates for the case fatality ratio in the general U.S. population vary from 0.05% to 1%”
Helpful! COVID-19 can be half as deadly as the flu, or up to ten times as deadly, and we don’t have a good way to know where in that range it falls. Also, as Ioannides points out later, we don’t know the CFR of other forms of coronavirus, though some studies suggest it might be quite high. Here the presence of the test makes COVID-19 stand out.
(2) When risk calculations don’t work well, one strategy is some form of non-risk-based thinking. The version that’s ascendant with public health authorities is basically some form of maximin, where you want to avoid the worst outcome. Economics-minded folks don’t like this, of course – John Harsanyi famously criticized Rawls’ use of it, arguing that it would lead to absurd conclusions like not taking a great job because you were afraid of the plane trip to interview for it. Of course, the Harsanyi-style objection gets its foot in the door by comparing a specific, definite loss with a tiny risk of catastrophe. Here, the catastrophe seems inevitable unless we exercise some sort of reduction strategy.
The catastrophe that seems most pressing to avoid is overwhelming the healthcare system. This is the point about flattening the curve. There’s now a model for hospitals to model coronavirus demand on their systems, including the all-important ICU-bed and ventilator demand. It’s of course all projections, but if you read through to the end you learn that even relatively small decreases in the spread of the disease will make enormous differences in what the hospitals are dealing with. It’s a really concrete demonstration of the curve-flattening point. The reason that matters is that most of the hospital systems in the country will be overwhelmed with a rapid spike in COVID-19 cases. If you keep the number of cases down, and drag them out over 18 months, then the numbers start to look somewhat manageable.
Overwhelmed hospital systems, of course, will cause enormous disruption and a much higher CFR. We’ve all seen the accounts from Italy. In the best case, flattening the curve both reduces the number of total cases and the burden on the healthcare system. In the worst, it reduces the burden on the healthcare system at any given time. Either outcome will save lots of lives. And so at least for now, social distancing and methods of flattening the curve are a good idea.
It is probably worth underscoring that there is considerable uncertainty here, too. For one, social distancing has been unevenly practiced: the criminally negligent messaging from Fox News and Trump have caused Republicans in particular to continue not to take the disease seriously, which means that large numbers of people aren’t attempting social distancing or other moderate measures to contain it. Trump is now lying and claiming that he always took coronavirus seriously, so hopefully his supporters will join him in the about-face.
More generally, Ioannides points to a paper that argues that we don’t actually know if social distancing measures work against respiratory infections (of course, we don’t have any research that’s COVID-19 specific). Moreover, we don’t know how much incremental good any particular kind of social distancing does. Do we need to stop schools? Do we need to ban groups of 50 or more? Shut down bars? Restaurants? Non-essential retail? Do we need to ban all travel outside of the home for any reason, and have the army deliver food rations? What’s more, we don’t know if social distancing that we’ve been trying has worked (and thus whether more intensive measures are needed) because (i) the incubation period of the virus is up to two weeks, so you’d need a solid two weeks to have any idea if you were making a dent in the transmission rate (probably more, since there’s going to be a lag between symptoms and a hospital admission or a test), and (ii) increasing testing rates are a major confounding variable – by increasing testing, we are going to see massive increases in the number of cases, simply as an artifact of catching more cases.
Perhaps this uncertainty is part of what is driving the demand to simply ratchet up the social distancing into full social isolation. It seems obviously right that the cancellation of March Madness prevented thousands of cases of COVID-19. So what about stopping crowds at a bar? Neighbors mixing? The logic gets very quickly to Foucault’s example of a town buttoned-down against the plague: “it is a segmented, immobile, frozen space. Each individual is fixed in his place. And, if he moves, he does so at the risk of his life, contagion, or punishment” (Discipline and Punish, 195).
The problem is that our entire socio-economic order is based on mobility, even if modulated and regularized, which is why (3) this strategy simply is not sustainable in the long term. There’s two obvious reasons. One is that social isolation kills people; not only that, the negative effects of isolation are most borne by the already vulnerable. For example, social distancing will likely make the opioid crisis worse. The other is that extreme social distancing for long periods will blow up the economy completely. There is a debate over whether mortality rates fall during a recession and if so, why. But the sorts of recessions modeled in those studies don’t come close to the economic meltdown that months or a year of total social isolation would cause. We already need to bail out everything and everyone (no, seriously: there’s a good argument there, and the only downside risk is inflation, which seems like a pretty small price at the moment, if the alternative is no economy). What’s more, decades of rapacious capitalism have left us without a good social fabric for enduring this, as have more general trends like recent erosions of social trust. So it’s not even clear that we can manage to pull it off.
I don’t have a lot of hope to offer here, but I do think one way forward is to characterize the current strategy as stalling for time. Not for a vaccine – that’s probably a year to 18 months away, which is about how long the pandemic would probably take anyway. So like the Messiah, the Vaccine will come only after we no longer need it. But we can plausibly be stalling for two things. One is viable treatment. If the mortality rate could drop to that of the seasonal flu, then presumably there would be less cause to panic. If we could keep people out of the hospital, then the burden on the healthcare system would be lessened. The grim data showing the hospital system being overwhelmed assumes that patients have a twelve day hospital stay (this is based on data from Wuhan). If you could cut that in half, if would make an enormous difference. There is a lot of work on treatments now, in varying stages of clinical trial. The WHO is trying to coordinate a massive effort to improve the research. If you poke around looking for COVID-19 therapies, there’s some glimmers of hope. I’ve seen some optimistic analysis suggesting that there might be some viable treatments by sometime over the summer. But treatment offers a way out of the current situation.
The other thing to stall for is data. We need to understand better how the virus spreads, the role of asymptomatic transmission, how long it lasts on surfaces, the role of children in its transmission, the efficacy of various methods to reduce transmission, and so on. Better data offers us the chance to modulate and refine social distancing efforts, with a goal toward making them sustainable. The current maximin strategy is primarily the product of ignorance.
Recent Comments