Yes, I think they do. This doesn’t mean shutdowns are a good thing, on balance. Maybe, at least in the case of an extended shutdown, the reduction in coronavirus deaths is offset by other health related fatalities. If not, maybe economic considerations tilt the balance against shutdowns.
I think it’s also true that the health related benefits of a shutdown diminish sharply, almost to the point of zero, when the shutdown is extremely stringent — e.g., when it bans all outdoor activity. And we also have to consider the possibility that lives saved in the short term thanks to shutdowns may be lost later due to a resurgence of the virus and the absence of new shutdowns, which governments understandably may be very reluctant to impose.
However, I believe that both common sense and the evidence support the view that, so far, shutdowns have reduced coronavirus deaths significantly.
Let’s start with common sense. The Wuhan coronavirus is a communicable disease. It spreads through close human contact. Reduce the level of close human contact and you reduce the number of infections. Reduce the number of infections and you reduce the number of deaths.
Shutdowns, if enforced, mean a low level of close human contact. Thus, unless the level of close human contact would be the same with or without shutdowns, it seems to me that shutdowns reduce coronavirus deaths.
And if shutdowns don’t cause people to behave differently in a pandemic than they otherwise would, then shutdowns don’t affect the level of economic activity, the number of suicides due to isolation, the amount of medical treatments postponed or foregone, etc. We should neither favor nor oppose shutdowns in that event.
But it’s unrealistic to suppose that shutdowns don’t affect the level of close human contact. Few believe this, which is why most people either favor or oppose shutdowns.
In fact, the most sophisticated critics of shutdowns oppose them in part because they share the common sense understanding that shutdowns reduce the number of infections (and, by extension, deaths). They argue that the best response to a pandemic is to reach herd immunity quickly — in other words, to have a pretty large percentage of the population become infected.
As I understand it, this was part of Sweden’s rationale for not shutting down its economy. Swedish authorities expected a heightened number of infections as a result of their policy. They regarded this as a feature, or at least not a bug.
The expectation of Swedish authorities was fulfilled, and this brings us to the evidence.
By now, there is reasonably reliable data from enough jurisdictions to assess the effect of shutdowns on coronavirus deaths by comparing outcomes. But differing conclusions can be reached, depending on the comparisons one chooses to make.
If one compares outcomes in Sweden to outcomes in New York, which had a stringent shutdown, one might well conclude that shutdowns haven’t reduced deaths from the coronavirus. The evidence is that in Sweden there have been 575 coronavirus deaths per one million people, compared to more than 1,600 in New York. (Unless otherwise indicated, all numbers cited here are from Worldometer.)
But why compare Sweden to New York state? The distance from Stockholm to New York City is about 6,000 miles. The demographics are different, as are the population densities. Sweden has no city like New York City.
In addition, by the time New York shut down, there had already been a high degree of community spread. Shutdowns don’t help those who have already been infected. Nor do they prevent spread in nursing homes if the virus has already arrived.
Why not compare Sweden to its neighbors? In Norway, where a shutdown was implemented, there have been 49 deaths per one million people. Only 49, compared to 575 in Sweden.
I’m sure there are demographic and population density differences between these two bordering nations. But I’d be shocked if they can come close to explaining the dramatically different results between the two. It’s also worth noting that, before their policies diverged, the coronavirus health outcomes were not very different.
Denmark also imposed a shutdown, but lifted it earlier than Norway did. The evidence is that, in Denmark, 107 people per one million have died. Again, that’s vastly fewer deaths per capita than Sweden has experienced, though significantly more than Norway, with its longer shutdown, has.
Turning to the U.S., South Dakota did not shut down. Thus, it provides a potentially useful point of comparison to other states.
If we compare South Dakota to nearby Wisconsin, which did have a shutdown, we see almost identical per capita death rates. But Wisconsin differs from South Dakota in terms of demographics and population density. There is no Milwaukee in South Dakota. Thus, the fact that the per capita death numbers in the two states are similar suggests to me that Wisconsin’s shutdown prevented loss of life due from the Wuhan virus.
Minnesota, with its shutdown, hasn’t fare so well. It has had 321 deaths per one million, compared to 182 in South Dakota and 186 in Wisconsin. However, as Scott has often pointed out, something like 75 percent of Minnesota’s coronavirus victims were residents of long term care facilities. A shutdown won’t prevent the virus from spreading among those confined there.
In South Dakota, residents of long term facilities reportedly make up only 43.5 percent of the state’s coronavirus victims. Thus, it seems that Minnesota has had fewer deaths per capita than South Dakota among those not trapped in nursing homes. This, despite demographics and population densities that would tend to make South Dakota a less deadly place than Minnesota in a normal pandemic.
Therefore, whatever else one might say about Minnesota’s shutdown, I believe that for residents not stuck in nursing homes, it probably resulted in fewer people dying than would have died with no shutdown.
More generally, for the reasons set forth above, I believe that shutdowns have reduced significantly the number of deaths from the Wuhan coronavirus in Europe and America. As noted, however, this doesn’t mean they have been beneficial, on balance. Nor, even if they have been, is this an argument for shutdowns of a particular level of severity or a particular duration.
