Lessons for the Next Plague

not what we did wrong but why

Looking back at the pandemic it is clear we made a lot of mistakes, that many of the costs of the pandemic were self-inflicted. The Swedish example, no lockdowns and excess mortality on the low end of the range, suggests that the costs of lockdown may have been unnecessary. Extensive precautions against surface transmission, based on what turned out to be an early mistaken theory of contagion, were continued to the end. Vaccine effectiveness was tested with a controlled double-blind study, scientifically ideal but slow, instead of by vaccinating and deliberately infecting volunteers which would have reduced by several months the delay before vaccines became available

The next pandemic will be a different disease so what were mistakes this time might not be next — but we might be able to learn some of the reasons for the mistakes we made. For example:

Plague and Politics

“I don't see how it can be decoupled from politics. It's safety vs freedom. One side always prefers safety and therefore gets neither.” (Marcus on DSL¹)

It was inevitable that questions of vaccine policy would end up as political issues. It may have been inevitable, in a politically polarized society, that they would link up to existing divisions, that factual questions (Are the covid vaccines safe? Does masking work?) became linked to tribal identity . Once that happened people stopped thinking — no need to ask a question if you know the answer.

If our politics remain as polarized as it is now are there is probably no way of preventing that from happening again, which is one reason to hope it doesn’t.

Politicization may have been inevitable but the particular linkages were not.² Trump pushed for vaccine development and use but somehow vaccine became a left/right issue the other way around. In the early days restrictions on movement were supported by the right, opposed by the left. Later, when the issue shifted from restrictions on international travel to lockdowns, that flipped.

Unjustified Certainty

Quote from: The Ideal Feds

"Look, this is a brand new disease and we honestly know very little about it. But it might be really bad. Here are some measures we think might help, so it's probably a good idea to do those things until we know more. We'll study this and get back to you as soon as we can."

Quote from: The Actual Feds

"We know exactly what this is and how to fight it, and if you don't follow every official recommendation exactly you're almost a murderer." (Samwisethebold on DSL)

The reason for the pretense of certainty is obvious — to persuade people to do what you think they should do, if necessary justify forcing them to do it. There are three costs. The first is that it makes it hard to change what you are saying on the basis of new evidence; doing so requires you to admit not only that you were wrong but that you lied about how much you knew. The second is that, if you do admit error, people will conclude, correctly, that you cannot be trusted to tell the truth and discount your future advice accordingly. The third is that treating your current policy as known with certainty implies that those who do not go along are evil, should be neither tolerated nor listened to, making it harder to find out if you are wrong and contributing to political polarization.

Tunnel Vision

One problem with the response to Covid was the focus on a single variable, mortality from Covid, ignoring all other costs of the policy being considered. Some of those costs are also mortal — shifting medical resources to dealing with Covid shifts them away from other things.

Patients who had heart attacks during the first COVID-19 lockdown in the UK and Spain are predicted to live 1.5 and 2 years less, respectively, than their pre-COVID counterparts, according to a study published in European Heart Journal – Quality of Care and Clinical Outcomes. (Respiratory Therapy)

My father died because his cancer, which could have been found earlier and could have been treatable/operable, became metastatic by the time the hospitals reopened. (Ana on DSL)

There are other costs. Closing schools down resulted in a large loss to education, arguably also to socialization. There were large costs to the economies of countries that engaged in extended lockdowns. Those costs were for the most part ignored in decision making.

We Must Do Something

It is hard, especially for someone in a position of authority, to admit that there is nothing to be done.

This is something. Do it.

Untrusted Science

Many policy issues depended on factual questions such as whether mask mandates reduced infection rates or the effect of vaccination on rates of transmission. There was no way that a random voter, even an influential political figure such as a mayor or governor, could get answers he could trust because there is no trusted source of scientific information.

At one point some group, probably the CDC, released an article “showing” that mask mandates worked to slow the spread of infection. It was based not on a controlled experiment — under the circumstances couldn’t be — but on a statistical analysis of county level data. It was, so far as I could tell, honest but it was quite easy for anyone familiar with the problems of such analysis to come up with possible links between the pattern of infection and the decision to impose mask mandates that could have produced the result reported even if the mandates had no effect. It was also easy to see that, if different analyses of the data had produced different conclusions, which got published would have been determined by what the authorities publishing them wanted people to believe not by which were most likely to be correct.

Reading that particular study increased my estimate of the probability that mask mandates worked, but not by very much. If I had not known enough statistics to be able to make sense of the analysis, the fact that the article existed would have been, for me, very nearly no evidence at all that its conclusion was true.

Ideally, the solution is to have trusted sources of scientific information, organizations with the ability and willingness to look over the complicated evidence for and against a controversial conclusion and give an honest report. The replication crisis provided good evidence that the system of peer reviewed journals did not do it. Other evidence has been produced by controversies over issues such as the origin of covid or the costs of climate change. The Lancet published a short piece describing the lab leak conjecture as a conspiracy theory; the piece contained the claim that the authors had no competing interests, did not mention that one of the authors was Peter Daszak, the head of a nonprofit that had funded bat virus research at the Wuhan Institute of Virology, the suspected source of the leak. Nature published an article on the costs of carbon whose implicit assumptions, all tending to increase the estimated cost, included no medical progress for the next three centuries.

It is possible that there exist organizations whose reports on scientific controversies that people care about can be trusted, but there is no way that anyone outside of the field the controversy is in can identify them. Fixing that problem would require a reform of the scientific enterprise, probably impossible as long as the output of that enterprise matters in political controversy.

At the individual level it is sometimes possible, with enough effort, to evaluate the evidence on both sides of a controversy and reach at least a tentative conclusion³ but there is no way of making the process easy enough to determine political outcomes in a democratic system.⁴

---

My web page, with the full text of multiple books and articles and much else

Past posts, sorted by topic

A search bar for past posts and much of my other writing

1. I started a thread on the topic of this post on the Internet forum Data Series Lox in order to mine the group for ideas, as I have done for some other posts. Several of the responses I thought worth quoting.

2. Scott Alexander discussed at length, in the context of an earlier and milder disease scare, both links between political ideologies and the way in which the presentation of an issue might determine which faction took which side.

3. A number of my past posts are on the subject of discovering truth. There is no easy answer.

4. This is an argument in favor of benevolent dictatorship. The argument against is the lack of any mechanism to make sure that the winner of the competition for power is both competent — at something other than gaining power — and benevolent.