Here's some simple and grim math for our covid future: At the moment, the number of new cases each day has been roughly flat for the past 3-4 weeks. 
That the number of new cases is roughly flat means that each person who catches the disease and recovers infects roughly one other. (ie R=1)
That they're not infecting two or three others — as occurred earlier in the pandemic — is because of social distancing.
That they're not infecting two or three others — as occurred earlier in the pandemic — is because of social distancing.
A one-to-one transmission rate is a tightrope. If transmission rates rise then the number of cases will quickly rise.
It's the point at which you want to push even harder to reduce transmission. That's our only chance to avoid mass infections and thousands of deaths each day
It's the point at which you want to push even harder to reduce transmission. That's our only chance to avoid mass infections and thousands of deaths each day
But if we lighten up—even only a little bit—then things get grim quickly.
Say the transmission rate rises a smidge 1.03. Then 3 weeks later, we have *doubled* the number of cases and only a few weeks later, we'll have doubled the daily death rate.
Say the transmission rate rises a smidge 1.03. Then 3 weeks later, we have *doubled* the number of cases and only a few weeks later, we'll have doubled the daily death rate.
We can quibble about whether the number of new cases is flat or declining. Let's say it's declining, and R=.98. Point is, it's still only a smidge away from 1.03. That smidge would cause the number of infections (& deaths) to rise dramatically.
Exponential(ish) growth is a bitch
Exponential(ish) growth is a bitch
All of this is to say: The observation that the Administration is forecasting deaths to rise to 3,000 per day by the end of the month as states start relaxing their restrictions is pretty much obvious.
nytimes.com/2020/05/04/us/…
nytimes.com/2020/05/04/us/…
Of course this forecast—like any forecast—could be wrong.
But the consequences are very asymmetric. If the transmission rate in those states opening up is pushed even a little bit higher, then much much higher death counts are plausible.
But the consequences are very asymmetric. If the transmission rate in those states opening up is pushed even a little bit higher, then much much higher death counts are plausible.
Here's the latest data. Extrapolate away, realizing that loosening restrictions steepens these curves.
A few important states have decreasing caseloads.
Many others are flat or rising. There, more rapid transmission could cause tens of thousands to die.
nytimes.com/interactive/20…
A few important states have decreasing caseloads.
Many others are flat or rising. There, more rapid transmission could cause tens of thousands to die.
nytimes.com/interactive/20…
My advice (and plea) to those state governors looking at easing restrictions: Create the political room to reverse course if it turns out that your decisions lead the infection rate to rise rapidly.
It's not all or nothing.
Call it an experiment, and respond to incoming data.
It's not all or nothing.
Call it an experiment, and respond to incoming data.
Getting the transmission rate below one must be our first goal. Ask an economist.
