I haven't written about COVID19 for a while. Here's some random updates on things I think we've learned and how my best guesses have changed:

a. I suspect official opposition to masks will be looked back on as both unjustified on the evidence we had, and very harmful. 1/n

In retrospect one could have figured this out at the time by giving more weight to:

i) the common-sense notion that physical barriers reduce the spread of germs, & ii) the views of authorities in the countries that were already doing the best job of controlling the disease. 2/n

But the issue is not entirely settled, and there's still a ~15% chance masks will be shown not to have helped, or even to have hurt.

That said, please for heaven's sake wear a face-mask everywhere you can, especially if you have to go indoors. 3/n

b. The view held by 90%+ of economists in March — that we had to go hard to control the virus in order to protect the economy — looks sound in retrospect.

80-90% of the decline in econ activity was due to the choices of individuals and businesses rather than govt orders. 4/n

... We didn't need to sacrifice the economy to save lives. Rather the way to save the economy was to control the virus. 5/n

c. The theory that China was telling huge lies about the virus' fatality rate, its prevalence in China, or its number of dead, now seems mistaken.

They have probably been misleading us, but nothing too crazy. 6/n

d. In March some thought it would be impossible to 'crush the curve'.

Others thought that doing so would require extreme measures like forcing the sick out of their homes an into quarantine special facilities.

With the benefit of hindsight we can now see that was mistaken. 7/n

e. Looking at Asia, it appears that organised and law-abiding countries can prevent or control outbreaks of COVID19, on an ongoing basis, and at an acceptable cost.

Many people were (understandably) skeptical this would be practical. ... 8/n

... One could have gotten the right answer on this one by focusing on what China/SK/Singapore/Taiwan/Vietnam/Mongolia/Thailand/etc already seemed to be successfully doing as early as March. ... 9/n

Whether those of us in Europe/US/CANZUK/etc will be able to keep control of COVID like Asia has, with our weaker governance and lower compliance with social norms, remains to be seen.

I think it's more likely than not that we will, with the exception of the United States. 10/n

f. The world continues to spend vast amounts on C19, but the great majority is 'economic stimulus.'

A larger fraction should be going into controlling the virus, and speeding the development of vaccines, tests & treatments, using, for instance, Advanced Market Commitments. 11/n

g. We need to focus relatively less on cleaning surfaces & more on fresh air.

We ought to move everything we possibly can outdoors. Open all our windows. Make open-plan offices into closed-plan offices & blow air out. Etc.

It's weird this isn't a top topic of conversation. 12/n

h. The debate over whether sending kids back to schools will spread the virus seems very peculiar.

We know children who go to school are incredible vectors for contagious diseases in normal times.

And we have germ theory, so we understand perfectly well why that is. ... 13/n

... We should have a strong prior that kids in school will spread COVID19 a lot too, unless and until overwhelming evidence shows otherwise. 14/n

i. Very early expert estimates of the IFR were impressively accurate (0.3%-1.2%) for something that's so hard to figure out. 15/n

j. One of the worst mistakes made by some countries was allowing the disease to start spreading within care homes for the elderly.

That alone seems to 2x or 3x the number of people who die in your country. 16/n

k. Some experts estimate that, if the UK had locked down just a week earlier, half as many people would have died (~30k rather than ~60k) and the UK would have looked more like Germany.

The claim also makes intuitive sense, and jumps out in a simple spreadsheet model. ... 17/n

... In retrospect it appears that being indecisive in early March was a big mistake.

If you were ever going to lock down, sooner was almost always better. 18/n

l. I think my chance of dying from C19 if I caught it would be under 1 in 1,000 — a bit lower than my estimate in April.

But I really worry about all the credible reports of people suffering ongoing symptoms months after apparently clearing the infection.

A big X factor! 19/n

m. I doubt Trump really told his nominees at the FDA and CDC to "slow the testing down" in March.

But if he did, I'm not sure what they would have done differently.

The FDA continues to slow-walk approval for innovative ways of testing more people more cheaply. Sad! 20/n

n. I haven't studied them closely, but Government recommendations often seem weirdly inconsistent.

In theory if I meet a friend at a park I'm not allowed to get within 1 metre of them.

But I'm allowed to fly to Germany for a holiday and return without self-quarantining. 21/n

... For that matter, while I'm not allowed within 1 meter of people outdoors, I'm allowed to drink indoors in a bar.

And I'm even being subsidised to go eat and talk inside a restaurant, surrounded by strangers! 22/n

o. I've still heard of almost no shortages in supermarkets. Four cheers for supermarkets and online retailers! 23/n

p. Research on vaccines is advancing faster than I predicted it would in May. 24/n

q. In my mind the biomedical community comes out of this looking best. Economists & central banks performed pretty well too.

But I've lowered my estimate of the capabilities of many governments to react intelligently to fast-moving disasters... 25/n

... And the quality of policy advice governments around the world were getting early on was pretty mixed.

The OECD should feel somewhat humbled, and aim to learn from those poor and middle-income countries who, on its face, have handled this crisis better than them. 26/n

r. Too many self-important 'experts' were confidently wrong in January, February and March, along the lines of this dismayingly disrespectful video: https://twitter.com/JeremyVineOn5/status/1239512626134224896 ... 27/n

... First it was "the flu is a bigger threat than COVID", then "it's not here yet", then "face-masks don't work you dummy", then "lockdowns can't stop a respiratory pandemic" then "we'll just get a second wave anyway". ... 28/n

In a situation like one one — where so little is known — humility, complete honesty and sincere respect for the public should be what public figures aspire to.

That's true even when you're right, but the lesson is most conspicuous when you've just gotten things wrong. 29/n

s. People should have access to data on COVID19 prevalence not just in their country or city, but their suburb or even street. Spread is often very local.

We get more bang for buck by targeting renewed lockdowns on just those local areas that have active outbreaks. 30/n

t. @1daysooner seems to have won the intellectual debate: once we have candidate vaccines, we should do voluntary human challenge trials, if they'll speed things up.

But I was surprised to see such a vast number of brilliant & famous people come out in support of the cause. 31/n

u. In March a few people predicted the rise of 'green zones' and 'red zones' around the world.

People would be free to travel within green zones, where the virus is controlled, but would face be restrictions moving from red to green.

This prediction looks on point today. 32/n

v. Despite my irritation with people who don't wear masks, or host parties, or buy into idiotic conspiracy theories, it looks to me like 90% of people have been doing the right thing 90% of the time.

What was asked of everyone was huge, and we mostly did it. ... 33/n

... So personally, COVID19 has made me upgrade my estimate of the general public's common sense and how much people are willing to pull together in disasters. 34/n

w. I continue to be surprised that if someone you live with gets C19, your chance of catching it from them is probably <20%. Higher for partners, lower for mere housemates.

Surveying 13 papers the median estimate was 14% & average 18%.

I'd love someone to explain that. 35/n

x. Many people predicted an increase in other social ills like crime, suicide & alcoholism. My impression is that most haven't happened or there's no data to show they have.

A shared disaster promotes social cohesion and being stuck at home makes it harder to get into mischief.