> I don’t get it. How can social distancing NOT help during a pandemic?

It mostly seems to be for the same reason that several other things didn't work: half the population doesn't do it and ruins it for everyone else.

In that case, I'd expect to see more obvious correlation between areas where people were fairly observant of masking rules in public and a lower incidence of disease. AFAIK, that correlation is weak at best in general.

The places high in observence of safety protocols are cities, which are more dense, and they have a decent proportion of Republicans anyway. The different cities you might be able to compare (eg between different states that had different messaging and rules) are in different climates, so that ruins comparison as well.

But if we compare California to Florida/Texas, for example, we can at least say pretty conclusively though that the effect wasn't strong enough to overcome climate differences or whatever other confounding variables might be involved. So while we can't say for sure they aren't effective at all, we can place a limit on how effective they are based on the data, and as far as I'm aware, that points to them being marginal at best. They certainly aren't, in themselves, any kind of panacea for stopping a pandemic--that much has been proven.

Does the comparison even tell us that? There's more confounding variables than there is signal IMO.

Well that's my point. If these measures were really so effective, then there should have been a strong signal in spite of the other variables, but we didn't see it.

The variables are all over the place though, why should the signal be strong?

At some point then though, it's about faith. The data is noisy but we believe this hard to discern signal to be true even in the absence of statistical evidence because it's complicated.

Faith in what exactly?

For many, that mitigation measures that so much blood, sweat, and tears were poured into at significant cost and strife must have been worthwhile even given scant evidence.

It should be strong if the effect is strong. We find strong signal in noisy data in all kinds of contexts.

If all the states with masks and social distancing mandates had, on average, 30% less cases per capita than states without, that would be a strong signal even if it's not conclusive given all the other variables involved. I doubt any other variable could be found that would show that much signal. But we haven't see anything like that.

> It should be strong if the effect is strong. We find strong signal in noisy data in all kinds of contexts

And we fail to find signal, or can find whatever "signal" we want to find in all kinds of other noisy data. What says any kind of useful info is here to find?

Exactly. I've not done clinical research but I have done market research. And, while you can cherry-pick apparent differences (or lack of differences), construct stories around those, and just conclude some results are weird anomalies, at some point you need to conclude there's very little evidence there's a lot of consistent signal in some particular policy/demographic/cultural difference.

My own experience is more like 1 in 5000 people followed all the rules.

"All" is doing a lot of work there. My experience in Massachusetts, for example, is that there was quite a high rate of compliance with public masking rules in general. Yes, some people were sloppy about wearing them and many wore less effective masks. But I saw very little outright refusal in indoor public places.

Followed every rule perfectly? That sounds about right. That's really not that important though.

But half of the US at least is actively hostile to doing anything to help anyone. Our society is broken and doomed.

I think many people were practicing performative masking and social distancing in public spaces, while at the same time spending lots of time with relatives and friends in private. If the latter was what drove the majority of the spread, then masks and social distancing policies could only have marginal impact.

Sounds like a self-fulfilling prophecy. If people were observant of some rules in public, but not in private, it would almost certainly make private transmission more prevalent.

That doesn't mean masks and distancing didn't work. It's quite possible that if they weren't being observed at all in public, the spread might have been a lot higher in this space.

Assuming the above is true, it's quite possible that public + private spread would have driven the pandemic to much higher numbers.

> How can social distancing NOT help during a pandemic?

It's not that it doesn't help at all. It's that it doesn't help enough to be worth the cost. As an analogy, consider that a wholesale ban on trains would help prevent things like what happened in East Palestine, but nobody thinks doing that would be a good idea.

It does help. This is the first sentence of the Introduction (p. 21):

> Social distancing works.

The paper concedes this point, but wants to argue that government mandates are not worth it. The Introduction continues:

> If you keep distance from others, your risk of being infected with a communicable disease is reduced. However, the fact that social distancing works does not imply that compulsory non-pharmaceutical interventions (NPIs), commonly known as ‘lockdowns’ – policies that restrict internal movement, close schools and businesses, ban international travel and/or other activities – work.

They actually find that many of these activities do work. They just argue that compulsory government intervention saved many fewer lives than the actions were anticipated to save at the time and had other negative effects.