Thread by @willchamberlain

A lot of people are sharing out this piece by @aginnt on #COVID19 that I think is profoundly misguided

It combines a pile of statistics (fine) with a very poor policy argument that I'll rebut here, in a thread

medium.com/six-four-six-n…

This is the beginning of the argument

Let's start his analogy about the first rule of medicine being "do no harm"

This is an asinine analogy: every policy has trade-offs and causes "harm" to someone

In a pandemic, the first rule is to "survive": avoid risk of ruin

Ginn complains that "local governments and politicians are inflicting massive harm and disruption with little evidence to support their draconian edicts."

The problem? You don't wait for your house to burn down before you buy homeowner's insurance

Our knowledge about the contagiousness and severity of #COVID19 is limited at this stage

We have some statistics and can make guesses as to the possible range, but we don't know

What we know is that it can cause absolutely unacceptable outcomes: "ruin"

theguardian.com/world/2020/mar…

The demand for better "evidence" before acting gets it completely backwards

If you are in a position of uncertainty - with some risk of ruin - you take *precautions* until you get better evidence

If there's *uncertainty* about your pilot's skill, do you get off the plane?

Next, Ginn talks about basic hygiene

All well and good, until he talks about how masks don't work, because people touch their face

That misses the point: masks are less about protecting yourself and more about making sure you don't transmit the disease to others

Next, Ginn talks about the value of more data (true!)

Better testing will allow us to go back to work, confident that we can avoid risk of ruin

But again he is too aggressive

We shouldn't "open public life" just to have more data - remember the first rule, avoid risk of ruin!

Next, he opposes closing schools

First, he says it's expensive. That's something to care about when risk of ruin is mitigated

Second, he talks about the reduction in medical personnel, which is relevant

We do need to make sure medical personnel can get child care

Here's where the arguments get really asinine

1) Evidence is really limited at this point - we don't know just *how* likely children are to get infected in various places

2) Even if children are more likely to get infected at home, reducing possible transmission events is good

3) We don't need schools as a single point of testing - we can do drive-through testing like in South Korea that doesn't require mass gatherings of students

4) Discounts that it's not just students at school - but teachers and parents as well

Here's where Ginn's naive empiricism is just appalling

"Shuttering the local economy" is not a distraction

There is certainly not overwhelming data that community-based spread is not a threat, Italy's ICUs are overwhelmed

"We don't have specific examples of it spreading through restaurants and gyms"

Well, we're pretty confident that the virus lives on surfaces, and there sure are a lot of surfaces in those places

Absence of evidence is not evidence of absence

"Isolating every family member in their home is a perfect situation for infection/transmission among family members."

This is just a ridiculous argument: the point is reducing spread between families

China wouldn't have beat the bug, by this logic

"Evidence from South Korea/Singapore shows you can continue on with life..."

South Korea and Singapore have 1) tons of masks and 2) tons of testing

Mask-wearing prevents spread, testing means infected individuals get quarantined

We don't have either, yet

My word, the statistical naivety

"The data shows that the overwhelming majority of working people will not be personally impacted..."

WAY too early to be making definitive statements like that

Oh, and yes, "we [do] need money for this fight."

Money printer go BRRRRRRR

Ginn points out how expensive shelter-in-place would be and that it would be cheaper to rebuild our acute care capacity

Indeed, shelter-in-place is expensive, a full travel ban would have been cheaper

But it's an asinine comparison, we need to avoid short-term risk of ruin

If we need more volunteers at the hospital, let's set up a program to do it

We're the wealthiest country in the world, money to fund PPE production/hospital expansion is not an issue

And early government inaction doesn't mean we don't need aggressive governmental action now

This is all irrelevant to the question of how government should respond

Our food supply chain is remarkably robust, toilet paper hoarding is not a serious problem

The big problem is the potential crunch on ICUs and the overall uncertainty about the disease

Ginn says CDC's worst-case projection "sounds implausible."

Yes, that's because CDC is accounting for uncertainty when coming up with a "worst-case" projection

In that sense they are being FAR more sophisticated than Ginn who is flippantly discounting the risk of ruin

Yes, there are a lot of reasons that we will probably be better off than China

Probably

But there's still enormous uncertainty at this point

Prepare for the worst, hope for the best

Here we have it, the worst argument in the piece

"Does stopping air travel have a greater impact than closing all restaurants?"

We don't know, because we can't know

The question is how to proceed in the face of uncertainty

Ginn would have us do nothing

That's what Italy did and now military trucks are driving coffins to regions with more cemetery capacity

In the face of uncertainty, where there is risk of ruin, you take *precautions*

There's no "law" about caution, it's just the precautionary principle

@nntaleb, @normonics, and @yaneerbaryam basically rebutted Ginn's analysis back in January

https://twitter.com/nntaleb/status/1221410676540936194?s=20

This is an argument for more MONEY PRINTER GO BRRR

It doesn't answer the risk of ruin issues

There are a ton of levers we can pull when it comes to saving businesses and putting money in people's pockets

Not so many when it comes to reducing the spread of Coronavirus

Here's a false choice that reveals the problems with Ginn's piece

We absolutely should be building hospitals as fast as possible

But TIME is the problem, we need to take precautions ASAP to avoid near-term risk of ruin, and you can't just snap your fingers and build a hospital

"It's only affected .004% of the population so far!"

Again - you don't wait to buy insurance until the house is on fire

This is the same approach the left uses with regard to terrorism: "Why do you care! Way more people die in car accidents!"

RISK OF RUIN

In sum: this article is just profoundly unsophisticated when it comes to risk

This kind of thinking led to the housing crisis, discounts the risk of terrorism, ignores fat tails

If there is uncertainty and near-term risk of ruin, TAKE PRECAUTIONS

FIN