It’s not often one gets the chance and obligation to say this. Brad DeLong :
Note to Self: Is there anything wrong with this analysis? With 14 deaths in the U.S., a 1% death rate, and 4 weeks between infection and death, that means that as of Feb 8 there were 1400 coronavirus cases in the United States. If it is doubling every seven days, then now about 22,000 people have and in the next week about 44,000 people in the U.S. will catch coronavirus. These numbers could be five times too big. These numbers could be five times too small. But with only 1 in 10,000 currently affected, it seems 4 or 5 weeks early to start imposing serious geographical quarantines …
No, no, no. R is not a function of the number of cases. It is only a function of herd immunity and the individual chance of transmission.
The condition for the decline and fall of the epidemic is lowering R below 1. There are two paths to this. Call them the Trump policy and the Xi policy.
Under the Trump policy of malign neglect, the virus spreads until most of the potential transmittees of the virus have recovered from it and are immune. Meanwhile, the cemeteries have filled up with those who didn’t recover. 200,000? 480,000? 1.7 million, if the hospitals collapse and the treatment is back to Black Death standards? The epidemic expires from satiety.
Under the Xi policy (also now the Moon, Conte, Sanchez, and Merkel policy), the state cuts the opportunities for transmission, including quarantines as well as contact tracing and mass preventive screening. The cost of this, direct and indirect, is fixed and independent of the number of cases. Imagine a perfect lockdown in which everybody stays in their house or flat, living out of tins by candlelight, for a fortnight: end of epidemic. This can’t be done perfectly of course, so real outcomes are a risk distribution, but you can get pretty close, as Singapore and Taiwan have shown. The death toll is still 1% or 1.6% or whatever of those infected.
It is lower the earlier you start the policy. Starting when the diagnosed cases are in the thousands, as seems to be the political trigger, looks as if it might limit ultimate deaths also to the thousands or tens of thousands. QED.
A more picturesque way of looking at this is from the point of view of the virus. It’s a clone army of dumb replicants with no leader and a single mission: reproduce. For the Virus Army as a whole, a pandemic is a death ride. At the end of it they will all be dead, apart from the small source population living quietly in non-fatal parasitism with its animal hosts. The fun part is how many non-standard hosts they can kill along the way.
No zoom in on the virus commando that has infiltrated a single human host, you. The commando is doomed. In 14 days, either you are dead (end of viruses) or your immune system has destroyed them all and you are recovered (end of viruses). Their only hope for reproduction is for some members of the commando to jump ship and invade another disarmed host. The jumping ship is nearly always fatal, as the viruses can only survive a few hours outside a host. They have no independent motility and are dependent on cooperation by the hosts: handshakes, kisses, cough aerosols, unwiped door handles. Reduce that cooperation, and the survival odds for SEAL Team Virus drop to almost nothing. That, in essence, is the Xi policy: and it works.