Comment by __blockcipher__ | Hacker News Reader

Comment by __blockcipher__ | Hacker News Reader

Parent Full thread

__blockcipher__

What's the upper bound on mortality attributable to lockdown and an environment of unprecedented fear and hysteria?

Maybe start with the impending global food shortage.

And the fact that we've went backwards on poverty, went backwards on vaccines for everything that isn't SARS-2, etc.

And the fact that the physiological and psychological state lockdown puts us in, almost certainly makes us more likely to die of COVID-19 if we get it?

I don't carelessly accept the upper bound. I rationally set an upper bound and ask myself, "is this worst case something I would rather live with, rather than the worst case of the lockdown scenario?".

Then I compared the average cases.

What I found was that the worst case scenario of lockdown is worse, and the average-case scenario is worse. Only the true "best-case scenario" of "we magically get a vaccine in a month, the vaccine works for everyone without any significant repercussions and we manufacture 8 billion doses and the entire globe voluntarily accepts it" is better (although the corresponding best-case scenario of no lockdown is also similarly absurdly positive so best-case scenarios are useless anyway).

--

Tell me, what is the logic in practicing "indefinite postponement" (containment), which accrues on-going, non-linear costs (as well as the obvious linear component), and only ends when we develop a game-changer treatment or vaccine, which is a temporally-unbounded future event.

When has it ever been good public policy to hinge everything upon an uncertain future event? That's just bad strategy.

To use a starcraft analogy, we're trying to sit on one base and tech up while a skilled opponent would be on 4 bases, have a strong econ and then out-tech us anyway since they, you know, actually have a sustainable, functioning economy.

There is no "impending global food shortage". It's also baseless speculation that "psychological and physical consequences of lockdown" make you more or less likely to die.

There is ample precedent, however, of doing lockdown right and basically being done with it: Asia and Europe have many countries with drastically lower infection rates than the US. And most of them are further along the way of opening back up than the US.

> What I found was that the worst case scenario of lockdown is worse, and the average-case scenario is worse.

That is assuming that prolonged lockdown continues to be our response. It clearly doesn't have to as other first-world countries cases have actually been floored.

Aside from complaining about our first run at this, we could/should probably just all wear masks and then spreading is mitigated such that new cases approach zero.

You can't just lockdown and then stop, unless you eradicate the virus entirely, which is impossible.

It only stops in the absence of lockdown when the herd immunity threshold is reached. Period.

The only question is what that threshold is. Without these findings of heterogenous susceptibility, that threshold would be 70-90% depending on your estimate of the effective reproduction number.

With these findings, that number drops to a much lower threshold. Much, much lower.

--

If I can ask you one question, what is your explanation as to how cases have floored in other first-world countries, in the absence of herd immunity, if they are not practicing lockdown?

Your perspective is a very weird American one where the response has been one giant failure, whether you think lockdowns are great or not.

Others are doing much better. There are hardly any new cases in, say, Germany, with by now extremely minimal measures and no lockdown.

Your epic texts seem to come from a weird parallel world.

This is just circular reasoning. Countries with lower rates of cases/deaths "did well" and countries with higher rates "did poorly." The total heterogeneity across the types and timing of responses in countries across both groups doesn't ever seem to raise any questions.

In fact, that's exactly the implication of the Nature paper that's ostensibly the reason for this thread. The more we learn about pre-existing resistance to this coronavirus, the more we can figure out whether measures taken made a difference or were simply coincident with a less-vulnerable population.

But I don’t think scientists think that previous common cold exposure actually makes you immune?

The hypothesis I heard is that this immune response due to previous common cold corona virus exposure could explain the high proportion of asymptomatic cases or cases with very mild symptoms.

Those are already priced in, so we already know about their existence, it would just be an explanation as to why they exist.

I‘m also not sure why you think that even neighboring and very similar countries are so wildly different based on wildly different levels of pre-existing immunity. That doesn’t seem very plausible. Why should, say, Germany and Italy be so different? Or Germany and France?

> But I don’t think scientists think that previous common cold exposure actually makes you immune?

That's the subject of the paper being discussed, it's worth going back and looking at it. The researchers demonstrated T-cell reactivity against it not just in blood from people who had SARS-Cov-1/2 but those who had neither. So something else is going on -- maybe one of the common cold coronaviruses, maybe some other infective but not dangerous coronavirus, maybe something else.

> I‘m also not sure why you think that even neighboring and very similar countries are so wildly different based on wildly different levels of pre-existing immunity. That doesn’t seem very plausible. Why should, say, Germany and Italy be so different? Or Germany and France?

Since we don't even know the source of this pre-existing immunity, much less how it varies between countries, it's impossible to say. But it's clear that there are many differences between what happens in countries that aren't just lockdown timing/severity. Immunity, density, demographics, climate, sheer luck, etc.

See, you haven't actually understood what I'm talking about, so no wonder you think my position/view is contradictory.

I'm not saying that as an insult BTW, just a statement of fact.

Countries stop getting new cases without lockdown measures when they approach the herd immunity threshold. So, if we look at New York, which got absolutely ravaged (relatively speaking), they don't have significant new cases anymore. Not because Cuomo is a genius - he's the opposite - but because there's almost nobody left to get infected.

So, no, while I have been citing American examples because I am quite clearly an American, what I'm saying is relevant to every country in the world. Following a course of lockdown, social distancing and universal masking is simply an exercise in lunacy. At best it has no effect, and at worse it massively increases mortality.

In short, I really don't understand the point you're making. My goal is the maximum aggregate well-being of society, thus I oppose lockdown. If my goal were a more myopic measure like "maximum lives saved" (ignoring that a life is not a life is not a life), lockdown would still be the wrong call. Now if my goal were "minimize COVID-19 mortality regardless of the cost to all-cause mortality", then lockdown may or may not be effective, I personally think that it almost certainly makes COVID-19 mortality worse in the long run full stop, but that point is more debatable. I think the first two are inarguable, though.

That is complete nonsense. Germany is nowhere near herd immunity threshold. Taiwan and S. Korea and Japan are no where near herd immunity threshold.

Please elaborate more. First you need to tell me what you think the threshold is; the findings in this paper suggest it could easily be like 15-25% as opposed to the previous figures of 70-90%.

Germany has below 2 percent seroprevalence.

keyword suggest keyword could be keyword like

Yes, if you take the most wildly optimistic assumption for every single variable, when literally millions of lives are on the line, it might be 15 to 25%. Still Germany, S. Korea, Taiwan, and Japan are no where near that number.

I don't have any idea what the threshold is. No one does. Based on what it tends to be with coronaviruses in general, 65% to 80% is expected. Could be way more or way less. That is why it simply isn't worth the risk of finding out. If it is 65%, then the costs of you being wrong about all this are astronomically bad. If you were 99% likely to be correct, this would be a different conversation. But you are not 99% likely to be correct.

Okay, so on mobile and I’ve already talked too much in this thread so I won’t go in-depth, but you need to read the paper that this thread is about. Your estimates of the HIT are way too high because you are not accounting for the proportion of the population that is immune to COVID-19 infection despite never having come into contact with a SARS-type virus.

Also instead of saying “based on coronaviruses in general” you should just say “based on the basic reproduction number”.

Anyway even if HIT were 65-90% my argument is the same. But the fact that you gave that 65-80% estimate means you don’t understand the math. Factor in the preexisting t cell cross reactivity. Factor in heterogenous susceptibility (risk of infection even controlling for # of exposure events/social interactions varies heavily in the population, for example old people are much more likely to get infected)

TL;DR: you need to do more background reading and update your priors. you’re operating in very stale info. I also think you’ve failed to understand my general argument outlined across this thread but I’m too tired to repeat it so I’d urge you to re-read the comments where I outline my case

TL;DR: you are wrong.

There is some tantalizing evidence that maybe just maybe there might possibly be some degree of latent t cell cross reactivity. This is not actionable information, because it very well could be false.

"based on the basic reproduction number" is total nonsense. I am not basing anything on the basic reproduction number. I don't even know what you are even referring to here.

I understand the math just fine. It is you who keep doing the math wrong. You are "factoring in" 3 different variables, each of which are controversial and may not even exist at all. Then you are taking the most optimistic scenario for each one of these variables, compounding your bad math three times over.

Your argument is even more murderous if HIT is 65%. 65% of 7.5 billion people is 5 billion. IFR of 0.3% (which you admittedly completely made up) is 14 million dead? I'm on mobile but I think that's right. You want to kill 14 million people. Even though Japan, S. Korea, Singapore, Canada, Taiwan, Germany, Italy, Spain, heck arguably even NYC right now have proven that this can be controlled.

Reply on news.ycombinator.com