Your conclusion does not follow from the premise. Perfect timing is always producing better result, but do we have a way to establish perfect timing in the moment, lacking the benefit of hindsight? It's very easy to overfit the answer.

If you do it too late then more people will die than could have, if you do it too early you will end up overreacting and sending economy into a tailspin every other year without a good reason. You will also lose trust of the people, killing your chances at effective containment in the future.

Also, sorry about the downvotes you're getting. I think you're raising a good point and in a good manner.

You're assuming that if all states had taken the same approach, they would have all taken the best approach. You're correct that one hypothetical outcome is that 5500-8800 more people died. However, another outcome is that the entire nation could have responded in a way similar to New York in which case there would have been many more deaths.

> You're assuming that if all states had taken the same approach, they would have all taken the best approach.

I'm really not. It makes sense that different states should try different things, and I imagine that if all politicians listened to the scientists there would have been different answers proposed for different states.

This isn't just playing around with abstract new things that we have no experience with in public health. The same situation is happening in Sweden's "experiment" as compared to other countries with similar health systems in Europe. Sweden is expected to have about 10,000 deaths and they should have about 2,000 deaths if their health system works about as well as Switzerland. See https://covid19.healthdata.org/sweden. For reference Sweden has about 10 mil pop, Switzerland has 8.5 mil. Sweden is expected to kill about 8,000 people unnecessarily. We know this can happen because of experience with many previous epidemics.

Another question, though, is whether any lives would have been saved in the long run.

A vaccine showing up before we have to relax controls is not a given. If we have to rely on population immunity, a certain number of people need to get sick for us to attain it. Of course, we want the least vulnerable people to do that, and for it to happen at a rate that doesn't cause healthcare overload and excess mortality (which NYC came perilously close to).

>A vaccine showing up before we have to relax controls is not a given.

The opposite is probably a given. In the US, opening won't be initially full throttle but I think it's probably a given that significant re-opening will happen over the next 30-45 days.

More local control might actually have helped in New York, as Governor Cuomo initially blocked stronger measures proposed by Mayor DeBlasio.

I don't think that's actually true. On the 17th de Blasio announced that it might happen and he'd release the decision within 48 hours. Cuomo smacked that down but then issued a shelter in place order on the 20th. So there's up to three days variance there. Realistically it's one day. But the point in the article is that both politicians should have been considering shelter in place ten days before.

I didn't mean to imply that DeBlasio might have locked down NY City 10 days earlier; just that he probably would have done so a couple days earlier. And even a couple of days might have helped.

I think he would have because DeBlasio implemented quite a few restrictions earlier. On the 13th, he cancelled large gatherings and limited smaller venues to 50% capacity.

https://www1.nyc.gov/office-of-the-mayor/news/138-20/mayor-d...

On the 15th, he issued an order closing bars and restaurants,

https://www1.nyc.gov/office-of-the-mayor/news/152-20/stateme...

and closed schools.

https://www1.nyc.gov/office-of-the-mayor/news/151-20/new-yor...

Even San Francisco didn't issue a lockdown order until the 16th, and Washington state was even later than NY. I don't remember anyone considering a lockdown on the 10th.