Seeing this claim more and more - that models were the main basis for shelter-in-place orders, so any revision in the model means the orders were overkill.

It's total rubbish. The basis for those orders wasn't mainly models - it was observed evidence from other countries. https://twitter.com/brithume/status/1248286548916867073

And this isn't only coming from right-wing commenters (though mostly). @FareedZakaria made a similar argument, on uncharacteristically thin grounds, in this column.

Let's get into it. https://www.washingtonpost.com/opinions/without-mass-testing-were-flying-blind-through-this-crisis/2020/04/09/bf61e178-7a9b-11ea-a130-df573469f094_story.html

We did not go to shelter-in-place mode just “because models.” We did it because we observed what happened elsewhere.

Places that locked down early have done better. Those that delayed have gotten hammered. The difference is mere days or weeks. That's not modeling, it's evidence.

First - China. -However skeptical you may be of their numbers, in January they had out-of-control spread in Wuhan and sparks shooting elsewhere. They are also a county that – to put it mildly – prioritizes econ growth. Still, they looked at this risk and shut down their economy.

And doing so – worked. The lockdowns plus intensive public health measures (which, incidentally, we’re not replicating here) brought case counts down rapidly. The curve began plateauing after a few weeks and they got the virus down to a manageable level.

Skeptical of China? Look at Singapore. They have some of the most robust test/trace/isolate infrastructure in the world, yet still see distancing as necessary. Even with case levels much lower than the US, they've triggered a "circuit breaker" lockdown. https://twitter.com/EricTopol/status/1248737155653894144?s=20

Or look at Hong Kong. They triggered intensive distancing very early on - like closing schools, bars, restaurants - and it has paid off. They have (so far) been able to keep cases to a low and manageable level. https://www.reuters.com/article/us-health-coronavirus-hongkong/hong-kong-extends-social-distancing-restrictions-to-contain-coronavirus-idUSKBN21Q08D

Or more interestingly, look at South Korea. They have been able to avoid the level of economic disruption we’ve seen in China and in the US/Europe because they enacted sensible distancing EARLY while investing intensively in test/trace/isolate (which we still aren't).

The point here is that EVERY country that has managed to reduce and suppress transmission of the virus has made ample use of social distancing and shutdowns. And the later the decision, the harsher the measures required.

That’s not models. That’s hard evidence.

And there are experiments on the other side of the ledger too. Countries (and US states!) that took a more blasé approach, or tried to ride it out.

Look at the difference between Seattle and NYC – both large metro areas but with quite different outbreak trajectories.

Seattle (and large local employers like Microsoft and Amazon) pulled the trigger early on social distancing. NYC delayed; DeBlasio had to be overruled on cancelling the St Pat's Parade, and was urging people to go out to bars as cases were growing.

The UK initially tried to ride this out via a herd-immunity strategy that would leave the country mostly open and operating. They changed that before long, but not soon enough. They're now on track for a worse death toll than Italy.

So listen – ignore the models if you like. They’re a math exercise, not a crystal ball. They can be useful if you understand their limitations; they can be misleading if you don’t.

But the observed evidence both from other countries and from variances within the US is pretty consistent. Everybody ends up using stay-at-home & widespread distancing eventually. It's just that the ones who delay face much larger outbreaks by the time they end up there.

The shelter-in-place measures rolled out across our country are not fundamentally about models. They’re about seeing what has worked elsewhere, what has backfired elsewhere, and applying the lessons from that. Now is not the moment to begin backtracking.