Relative Gratitude and the Great Plague of 2020

In the depths of despair over not just having lost the Category War, but having lost it harder and at higher cost than I can even yet say (having not yet applied for clearance from the victors as to how much is my story to tell), I'm actually pretty impressed with how competently my filter bubble is handling the pandemic. When the stakes of getting the right answer for the right reasons, in public is measured in the hundreds of thousands of horrible suffocation deaths, you can see the discourse usefully move forward on the timescale of days.

In the simplest epidemiology models, the main parameter of interest is called R0, the basic reproduction number: the number of further infections caused by every new infection (at the start of the epidemic, when no one is yet immune). R0 isn't just a property of the disease itself, but also of the population's behavior. If R0 is above 1, the ranks of the infected grow exponentially; if R0 is less the 1, the outbreak peters out.

So first the narrative was "flatten the curve": until a vaccine is developed, we can't stop the virus, but with social distancing, frequent handwashing, not touching your face, &c. we can at least lower R0 to slow down the course of the epidemic, making the graph of curent infections at time t flatter and wider: if fewer people are sick at the same time, then the hospital system won't be overloaded, and fewer people will die.

The thing is, the various "flatten the curve" propaganda charts illustrating the idea didn't label their axes and depicted the "hospital system capacity" horizontal line above, or at most slightly below, the peak of the flattened curve, suggesting a scenario where mitigation efforts that merely slowed down the spread of the virus through the population would be enough to avoid disaster. Turns out, when you run the numbers, that's too optimistic: at the peak of a merely mitigated epidemic, there will be many times over more people who need intensive care, than ICU beds for them to get it. These cold equations suggest a more ambitious goal of "containment": lock everything down as hard as we need to in order to get R0 below 1, and scurry to get enough testing, contract-tracing, and quarantining infrastructure in place to support gradually restarting the economy without restarting the outbreak.

The discussion goes on (is it feasible to callibrate the response that finely?—what of the economic cost? &c.)—and that's what impresses me; that's what I'm grateful for. The discussion goes on. Sure, there's lots of the usual innumeracy, cognitive biases, and sheer wishful thinking, but when there's no strategic advantage to "playing dumb"—there's no pro-virus coalition that might gain an advantage if we admit out loud that they said something true—you can see people actually engage each other with the full beauty of our weapons, and, sometimes, change their mind in response to new information. The "flatten the curve" argument isn't "false" exactly (quantitatively slowing down the outbreak will, in fact, quantitatively make the overload on hospitals less bad), but the pretty charts portraying the flattened curve safely below the hospital capacity line were substantively misleading, and it was possible for someone to spend a bounded and small amount of effort to explain, "Hey, this is substantively misleading because ..." and be heard, to the extent that the people who made one of the most popular "flatten the curve" charts published an updated version reflecting the new argument.

This level of performance is ... not to be taken for granted. Take it from me.

Submit to Reddit

(Post revision history)

Comments permit Markdown or HTML formatting.