nav-left cat-right

Debunking the Debunkers

Israeli Professor Debunks Government Claims That Lockdowns Slow Coronavirus Infection, Proves Worldwide Decline In Spread

The word that caught my attention was “debunks.” Somehow, these days that type of language triggers my conspiracy theory radar. It practically screams, “motivated reasoning” by someone who is about to go to great lengths to willfully ignore the scientific consensus–or the article they are reporting on. In this case I have the advantage of fifteen days. Fifteen days more data to see whether the claim is holding up. And surprise, it is not.

To be fair to Professor Isaac Ben-Israel who is referenced in the headline, his study does not actually say what the headline claims. The paper’s claim is that there is little observed difference between the curve-flattening effects total lockdowns versus moderate social distancing measures. And to people who lack the cultural context, what is happening in most of the US counts as moderate social distancing measures as compared to the total lockdowns some countries (such as Israel) are doing. Professor Ben-Israel is advocating that Israel be more like the US. He’s not advocating for no measures at all.

The article was was released on 4/21. The study on which the article is based was released 4/8 and updated 4/16. Even if the article was a completely accurate representation of the study (which it is not) my concerns are twofold. Number one, there is really not enough precision in the data around global infection rates to draw conclusions this definite. There are enormous difference between countries in the testing rates, there are concerns about the accuracy of the tests, and there are differences in test protocols between countries. With data this imprecise, you can contort it to support any conclusion you want. The big exception cited in the article–the country that supposedly showed that it didn’t matter how you respond–Sweden–has not been faring too well in the six days since the study was released. It’s infection and death rates appear to be climbing as the standard model predicted, rather than going down as required by article’s claims.

And then there are the recent antibody sample studies which attempt to reconstruct after the fact what the disease was doing during the time when we were unable to measure its spread due to lack of tests. These, too, are imprecise and subject to criticism about both accuracy and how representative they really are. But they seem to show that SARS-Cov-2 was spreading both earlier and more widely in the US than previously recognized, and that possibly as much as 20% of NY State had already gotten it by the time they shut things down, causing (after a four week delay) a big drop in the number of deaths. The new infection numbers remain essentially meaningless, since it remains extraordinarily difficult to get a test if you are not hospitalized (or very rich). But in all, the outcome is tracking to the standard model; efforts to flatten the curve succeeded in flattening the curve.

That is worth repeating. What the study actually shows is that measures to flatten the curve really flatten the curve. Other than that the data is not precise enough to tell us how much of a difference extreme measures make versus moderate measures versus minimal measures. Even Sweden has taken minimal measures (high schools and universities closed, gatherings of more than 50 people prohibited). So even Sweden’s curve has been flattened to some degree (though the past few days show that Sweden is still on the upslope).

Once again the predictive power of conspiracy-thinking-aware linguistic analysis is validated. “Debunks” means “can be mis-represented to appear to buttress my preconceived notions”. Now how likely do you think it is that anyone will change their minds just by being corrected by facts?

Leave a Reply

Your email address will not be published. Required fields are marked *