Response to Scott Sumner on Covid Caution
SEE IMPORTANT UPDATE BELOW
Co-blogger Scott Sumner recently responded to co-blogger Bryan Caplan’s critique of Scott’s earlier post in which Scott stated “On a list of regulatory overreaction, these mandates [on masks, vaccines, etc.] don’t even make my ‘top 100’.”
Bryan agreed with Scott that the ratio of costs to benefits for many other regulations is higher than the cost/benefit ratio for Covid regulations in the United States. But, argued Bryan, the difference between costs of Covid regulation and benefits of Covid regulations was much greater than the difference between costs and benefits for the other regulations Scott referred to.
Scott’s recent response meets Bryan on this territory and he argues that Bryan substantially understates the benefits of Covid regulations and overstates the costs.
My view is that Bryan is closer to being right than Scott, although, like Scott, I don’t have Bryan’s confidence that he can use survey data on how people value years under Covid regulations versus years under Covid life. (By the way, a criticism Scott could have made, and maybe implicitly made, is that that’s the wrong comparison. We had “Covid life” with various degrees of regulation. I think the survey, to be useful for this discussion, should have been years of life under Covid with no regulation versus years of life under Covid with some degree of regulation. Admittedly, this is hard to get at because you’re getting not just people’s subjective valuations of life but also people’s subjective estimates of probabilities.)
This will not be a complete response to Scott but here are the major areas on which I agree and disagree with Scott.
I have one major agreement with Scott on his statement about risks:
When I read critics on Covid caution (not Bryan), I see a lot of innumeracy. People talk about a 1% chance of dying as if it’s a small risk.
I agree and I thought that as early as March 2020. I did my Ph.D. dissertation on the economics of safety legislation in underground coal mines. In the late 1960s, if I recall correctly, the risk of dying for coal miners in underground mines was about 1 in 10,000. That’s a 0.01 percent chance of dying in a year. And we thought (and thought correctly) that coal mining was a very dangerous occupation. Not like logging, by the way. When I worked in an underground nickel mine in 1969, older fellow workers (and they were all older than me) told me never to work as a logger. So a 1% chance of dying is huge.
A key number in Scott’s analysis is the infection fatality rate. Scott claims that it’s 0.6, but the number I recall (I can’t find the source offhand) over the whole U.S. population is less than half as much, at about 0.25. That’s a big difference.
Another area of contention is the fatality rate for people with pre-existing conditions. Scott writes:
Some talk about the risk for the under 65 group being merely people with pre-existing conditions, as if those people are sickly invalids. But I often see pictures in the media of healthy looking cops who have died of Covid. On closer inspection, some of them looked a bit overweight. And of course obesity is a major a pre-existing condition. I’m rather thin, and play tennis three times a week. So I’m healthy, right? Actually I’ve had crappy lungs my entire life, with several bad cases of pneumonia in my 30s. (If I’d been born before antibiotics, I doubt I would have lived to age 40.) So am I at higher risk? I honestly don’t know. But I really don’t see the point of people saying Covid is only a problem for the old and those with pre-existing conditions. Lots of people have at least one pre-existing condition. Obesity is not exactly rare in America.
Scott notes that he has a pre-existing condition, namely “crappy lungs.” But that’s just one. My reading of the Italian data in March 2020 was that disproportionately the people who died of Covid-19 were not just old but also had 2 or more pre-existing conditions.
One big difference between Scott, on the one hand, and Bryan and me, on the other, is over how hard it is to be masked. Scott writes:
Wearing a mask is a pain? All I can say is if you think that’s a major problem, I wish I could have your life!!
But that’s not analysis; that’s just Scott telling us his own subjective valuation. As I noted above, Scott rightly is skeptical of Bryan’s use of survey data to measure people’s attitudes to life under Covid. But at least Bryan had a sample size of 476. That’s 475 more than Scott’s sample size. In a comment responding to “DeservingPorcupine,” Scott says, “And when people talk about the awful suffering involved in wearing a mask, all I can do is roll my eyes.” In other words, Scott admits that he really doesn’t take seriously people’s thoughts and feelings about wearing masks. What matters is his subjective valuation.
One commenter on Scott’s recent post, Mark Bahner, writes:
Not only are most of the 800,000 dead old and sickly, but many of the deaths (approximately one-third, per the NY Times in June 2020) have occurred in nursing homes, where COVID restrictions did nothing to help them.
This is a relevant point. And understated. Not only did Covid restrictions do nothing to help them, but also governors in New Jersey, New York, and Pennsylvania actively hurt them by requiring that people with Covid be placed in nursing homes. Scott doesn’t reply to this point.
Scott also seems to vacillate about what his topic is. I took it to be about both regulation and private responses to Covid. After all, Bryan was addressing Scott’s original post about regulation. But in a response to Todd Kreider, Scott writes:
This has absolutely no bearing on anything in my post, as I’m sure you must know. My post is not about government policies.
And then Scott elaborates further to Todd:
You still don’t get it. You are talking about no government actions, I was talking about no attempts by the private sector to avoid Covid. (That’s also what Bryan Caplan was talking about.)
But that’s not what Bryan Caplan was talking about. He was talking about both private sector actions and government actions. Bryan referred to “America’s strange experiment in federalist dictatorship,” for example. And he mentions Scott’s earlier statement, “I’m surprised the regulations aren’t far worse.”
This is a discussion of government actions.
UPDATE: I’m glad I don’t have to rely on my memory of data I gathered 44 years ago. I was off by an order of magnitude in my note above about coal mine fatalities in the U.S. in the late 1960s. Coal mining was far more dangerous than I claimed above. Take 1964, a fairly typical year. There were 150,761 miners and 242 fatalities. That’s a fatality rate of 1 in 623, not, as I claimed above, 1 in 10,000. Big difference. It’s true that it’s not as much as 1 percent, but it’s 1/6 of 1 percent, not the 0.01 percent that I implicitly claimed.