Dialing Back the Apocalypse: The Scientific Debate

The predictive model currently dominating public discourse is the one produced a little over a week ago by the UK’s Imperial College (call it “Ferguson et al” if you want to sound like an insider). It delivered four shocking predictions that have greatly shaped governmental and media thinking since, as follows:

1. Unimpeded, COVID-19 will kill over half-a-million in the UK and about 2.2 million in the US between now and late August;
2. Mere mitigation, AKA “flattening the curve,” would only cut the slaughter by half;
3. Urgent and rigorous containment is the only way to greatly reduce possible deaths;
4. Containment, episodically relaxed and tightened, would need to continue until a vaccine was developed, approved, and deployed at scale, or the pandemic would quickly re-escalate.

The modelers did not consider the economic and social implications of #4, a long-running clamp-down on inherently social economic activities such as retail, travel, manufacturing, governance, etc. Others have pointed out that the full implication of such an extended reversion to a pre-medieval economy on a planet with 7.4 billion people could be a catastrophe even crueler that a viral pandemic.

The draconian consequences of an open-ended shutdown have led to proposals to find ways to reboot the economy sooner rather than later (see Thomas Friedman, below), while minimizing the risk to those at greatest risk of dying as much as possible. Implicit in this is the argument that have some people “take a bullet” for the greater good is a better strategy than destroying our world to achieve the very fewest COVID-19 deaths. 

The starting-point for this approach would be a two-week national lockdown—tight as a tick. Most cases that developed in that period would be mild enough for shelter-in-place. We’d do our best to help the high-risk cases. After that, let the uninfected return to work, quarantine everybody else.  There are many more layers to this approach and assumptions that need to be tested, but it seems to assume that Ferguson et al got it wrong when they said the pandemic will inevitably roar back if we take our foot off the brakes more than slightly. Needless to say, favoring the economy over the virally vulnerable has quickly attracted the feral attentions of the President, who seems ready to skip right over the essential first step of a fierce lockdown.

As it happens, a critique of Ferguson et al has just been published by seriously qualified people. It argues that the virus can be much more fully contained after the first wave than the model predicts. This would be very good news if true, so let me tell you who they are and try to unpack their argument, translating awkwardly from the original Nerdese.

The best-known of the three authors is Nassim Nicholas Taleb, the author of “The Black Swan,” “Fooled by Randomness” and “Anti-Fragile.” Taleb is a very smart, mathematically sophisticated explicator of complex systems problems. His two co-authors, Chen Shen and Yaneer Bar-Yam, are also his colleagues at the New England Complex Systems Institute, a concentration of scientists from Harvard, MIT and other schools you’ve heard of. Bar-Yam runs NECSI, and recently authored an opinion piece in USA Today arguing for an immediate 5-week national lockdown in the US (link below). These authors have the credentials to assess Ferguson et al, and here’s what they say:

1. Ferguson’s prediction of the potential scale of the pandemic, if unimpeded, is perfectly valid.
2. Ferguson is also right to argue that mere mitigation will not be sufficient to prevent tragedy, and that a fierce containment effort will be required.
3. Ferguson is WRONG to conclude that only a sustained severe containment regime would work to prevent a rapid resumption of the pandemic prior to the deployment of a vaccine.

Their argument to support #3 goes basically like this:

1. We have learned that the pattern of infection from this virus features low contagion between individuals in single brief encounters. Contagion appears to depend on degree of viral load: you can get it from a bunch of brief encounters in a short period of time, or from a single substantial encounter with an infected person. For the cumulative brief encounters model to sustain a pandemic, the share of the population that is infectious needs to be relatively high: there is a threshold level of infectiousness in a population below which the virus dies out (R0 < 1.0, as we cognoscenti say), rather than growing exponentially, unless there is some contagion mechanism beyond scattered brief encounters. After a sufficiently long hard lockdown, the number of contagious individuals will collapse to below this threshold.
2. In an unprotected society, there IS in fact a contagion mechanism to lift the infection rate high enough to trigger a pandemic: a sufficient number of super-spreader events and people can lift the concentration of contagious people above the threshold, after which frequent brief encounters can be enough to support exponential growth without intervention. However, after a lockdown has forced the infection level down below the critical threshold, a regime of sufficient testing, tracking and isolating, combined with certain travel restrictions and constraints on gatherings can keep the virus sub-exponential, which will mean in practice that it will die off to near non-existence for most people most of the time. If this is true, social and economic life can largely resume, with tolerable modifications in some behaviors.
3. Their key critique of Ferguson et al in this regard is that Ferguson assumed a simple individual-to-individual contagion model (as in those lovely animated graphics last week from the Washington Post), but in fact that’s not how this virus works. It spreads in vectors and clusters, a contagion model which can be efficiently stifled by shutting down super-spreader events, and fast-tracking and isolating super-spreader individuals. Of course, this will require great testing resources.

Needless to say, this outcome would be vastly preferable to an extended return to the economy of the early Middle Ages. It depends on a very, very strong initial lockdown, and then an aggressive, proactive program to prevent and contain super-spreaders. Do that, they argue, and we can get on with our lives until a vaccine arrives.

These authors are hardly Pollyannas: they have argued in the past that a mishandled pandemic could turn into an “extinction event” for Homo sapiens.  I’m far from qualified to judge the merit of their argument, but their qualifications to make the argument mean to me that it should be more visible in the urgent global dialog about what to do next. Unlike nearly all medical doctors and even most epidemiologists, they are very deep in expertise about the potential behavior of complex systems such as, for example, a global pandemic. Here are links to two of their recent publications on this topic, in the original Nerdish.

Explore

• Chen Shen, Nassim Nicholas Taleb and Yaneer Bar-Yam, Review of Ferguson et al “Impact of non-pharmaceutical interventions…”, New England Complex Systems Institute (March 17, 2020).
• Chen Shen and Yaneer Bar-Yam, A linked shared space model for COVID-19 transmission and its prevention by forming closed social circles, New England Complex Systems Institute (March 20, 2020).

Here’s Yaneer Bar-Yam in USA Today:

Here’s Thomas Friedman in the New York Times on restarting the economy quickly:

And here’s a link to the original “Ferguson et al” model, within an explanatory article: