The worst people to trust — except for everyone else.
 | | A burial in New Jersey last week.Todd Heisler/The New York Times |
|
For the past couple of months one epidemiological model — the IHME model from the University of Washington’s Institute for Health Metrics and Evaluation — has played an outsized role in public discussion of Covid-19. |
It’s not at all clear that it deserved this role. Among other things, its predictions have been highly unstable, sometimes revised sharply downward and sometimes sharply upward. Many epidemiologists have criticized the model as simplistic. But its very simplicity let it offer state-by-state predictions other models couldn’t. And the White House liked it, at least better than many other models, because it generally predicted a lower death toll than its rivals. |
But the White House probably likes IHME less today than it did yesterday: the institute just drastically revised its projected death total upward, from 72,000 to 134,000. Documents obtained by The New York Times suggest that modelers within the U.S. government have also revised death projections sharply upward. |
This is terrible news, and makes the push from Trump and many others on the right to relax social distancing look even more irresponsible than it already did. But it also tells us something about the field of epidemiology. It turns out that epidemiologists often disagree, sometimes by a lot. Their forecasts are often wrong, sometimes very wrong indeed. They are, in fact, the worst people to rely on in a crisis — except for everyone else. |
In other words, they’re a lot like economists. |
In a great essay published early in the Great Depression, John Maynard Keynes lamented that economics is a technical and difficult subject, but “no one will believe it.” He didn’t mean that economists are a priesthood possessed of unique and arcane knowledge, let alone that they are always right, but simply that even making educated guesses about the economy requires both hard thinking and knowing a lot about what smart people have learned over previous decades. |
Pro tip: If you or someone you listen to claims to have a deep insight all those stuck-up economics professors have missed, it’s overwhelmingly likely that the insight is either (a) something economists have known about for decades if not generations or (b) a well-known fallacy. |
And above all, you shouldn’t trust economic assertions from people who combine ignorance of the subject with politically motivated desires to believe certain things. |
Well, here we are in a pandemic, a complex phenomenon that depends on human behavior as well as biology. Like financial crises, different pandemics share many common features but differ in detail, in ways that can create huge uncertainty. Nobody can forecast their course especially well, but you do much better listening to the professional epidemiologists than to law professors, politicians, or, yes, economists who claim to know better. |
And you should rely more, not less, on the epidemiologists because pandemic prediction and response has become such a politically charged issue. Motivated reasoning — believing things because they’re what you want to be true, not because they’re really true — is a temptation for everyone. But researchers with a professional reputation to maintain are less susceptible than most. |
So let me give a shout-out to the hard-working, much-criticized epidemiologists trying to get this pandemic right. You may take a lot of abuse when you get it wrong, which you unavoidably will on occasion. But you’re doing what must be done. Also, welcome to my world. |
If you’re enjoying what you’re reading, please consider recommending it to friends. They can sign up here. If you want to share your thoughts on an item in this week’s newsletter or on the newsletter in general, please email me at krugman-newsletter@nytimes.com. |
 | | Flamenco meets heavy metal?YouTube |
|
Rodrigo y Gabriela have basically created their own musical genre; it’s amazing, and here it is from quarantine. |
歡迎蒞臨：https://ofa588.com/
娛樂推薦：https://www.ofa86.com/
沒有留言:
張貼留言