Would anyone better at statistics than I care to help me debunk this article?

[Music Man]

It was shared with me today. The title "Lockdowns Don't Work" is a little clickbaity as his point seems more to be "we don't have evidence that lockdowns work."

https://www.thepublicdiscourse.com/2020/04/62572/?fbclid=IwAR2GLzXD57hMW6MjHrgUVLs2WJyuwNvTFhzaopFororZ9RPXeDQ_lBd73xs

Primarily, he argues that--given the typical timeframe from contracting Covid-19 to death--countries like Spain and Italy began to see a downward trend in deaths earlier than is possible for the lockdown to be the reason. If that's true, my thinking is that many nations were seeing decreased social activity even before formal lockdowns, and thus stricter lockdowns are reinforcing distancing behaviors and hastening a decline of the virus that would have lingered if distancing measures were voluntary or moderated.

Anyway, many of you have better statistical and epidemiological minds than I do and can parse this out pretty quickly.

[onecaliberal]

1. California proves it DOES work.

[sheshe2]

2. NY proves it as well.

[thesquanderer]

9. Correlation cannot prove causation, i.e. that something worked. But LACK of correlation...

...can show that something does NOT conclusively work. So countering an example of where something didn't show the expected results with another example of where it did does not actually refute the premise.

I'm not saying the article's premise about lockdowns is right, only that showing that the same pattern did not appear somewhere else does not prove the article's premise is wrong, either.

Part of what the author does here, though, is come up with his own definition for lockdown, i.e. "Stay-at-home orders, low assembly thresholds, and business closures together constitute a lockdown. Without those three features, it’s not a lockdown." Which also means that other rules that may or may not have been announced at the same time are also irrelevant. Making up your own definition will tend to make it easier to make sure the evidence supports your conclusion.

With so many additional variables, it really is hard to establish specific correlations down to individual elements like that. IOW, even IF he is right in his statistical analysis, what OTHER variables may exist? Like differences in population densities? Whether there were also requirements for masks on those occasions when people DID have to leave their homes, and if so, whether there was ready availability of the right kinds of masks? Whether 6-foot social distancing was also being aggressively observed? Were these changes all instituted at once, or were some being followed for longer than others?

Logically, anything that keeps people apart helps. Anything that does that matters, but we have no control group to isolate different elements when you're doing multiple things at once. So for example, is it possible that NY and CA would have seen the drop in cases even without the stay-at-home orders, as long as they did everything else mentioned here (close many businesses, low assembly thresholds, 6 foot distancing, and masks for everyone)? We can't be sure, we don't have control groups. But it's also kind of moot, since proper protective masks are not widely available. (The author DOES recommend masks, but doesn't explain where we were all supposed to get them, when even our heath care workers can't get enough.)

[Music Man]

3. The other issue as I see it:

The danger of coronavirus is that (in addition to no vaccine) it spreads quickly, overrunning health care systems. That was certainly the case in Italy during the depths of their crisis. Lack of adequate care and equipment might as well be the cause of death. Thus, the argument all along has been that intense social distancing measures are intended to avoid major strains on hospitals, thereby reducing deaths. So the average time from contracting Covid-19 to death is irrelevant, as the lockdown reduces cases that need to be treated, freeing up ventilators, beds, doctors, PPE, etc.

[jberryhill]

4. We isolated quite a bit early

My assistant was making fun of me when I told her to stop coming in and work from home.

There were plenty of people who saw what was coming and didn’t need to be told what to do.

[lapfog_1]

5. I was able to "self-isolate" on or around Jan 30th

and I haven't been anyplace but to 2 runs to Safeway and 1 to Wallmart, plus 1 run to my favorite burrito place since then. Last couple od weeks in January I started buy essentials to live self contained for 60 days... unfortunately that wasn't nearly enough.

WFH ever since. I'm lucky to have a job where that is possible.

That said falling sales of things we make hang over me like the Sword of Damocles

[canetoad]

6. Yes they do work

Compare Australia to the USA. We are still locked down, borders closed.

[Igel]

7. US is still pretty much locked down.

Georgia's not reopened yet, a few red states never locked down but they contribute (in total) maybe 10-15% of what just NY does to the total. They're a trivial fraction of that curve.


Close the borders? Only racist xenophobes do that.

[dawg]

8. My family and business implemented a lockdown weeks before our stupid governor had anything to say.

I would imagine that would be true for well-informed people everywhere.

[LeftInTX]

10. The author does not define "lockdown"

He states "lockdown", then uses "strict lockdown" interchangeably

It's a bit of a word salad.

What is a "lockdown"?

What is a "strict lockdown"?