Covid tradeoffs

The argument by some conservatives that eschewing lockdowns would have yielded enormous economic benefits without any appreciable increase in covid deaths – i.e., that there were no tradeoffs here – is the kind of argument liberals often make. Normally, it’s liberals who say the policies they favor will yield only benefits. Normally, it’s liberals who deny there are tradeoffs.

Whether it’s liberals or conservatives who make these claims, I’m always suspicions of them. But suspicions aren’t evidence, much less effective rebuttal. It’s conceivable that even in the midst of a virus that spreads rapidly through human contact, a jurisdiction could maintain the high level of person-to-person contact necessary to avoid sharp increases in unemployment without incurring excess deaths as a result.

However, the evidence persuades me that this was not the case during this pandemic.

I always start with Scandinavia because, as John and I agreed at the time, the divergence in pandemic policies between the neighboring countries of Sweden and Norway provided something like a controlled experiment. Would Sweden, which did not lock down, fare no worse in terms of covid deaths than Norway, which did?

The answer turned out to be a resounding no. Deaths per capita in Sweden are reported at ten times the amount of deaths per capita in Norway.

Denmark rejected Sweden’s approach, but lifted its lockdown earlier than Norway did. Not surprisingly, Denmark’s reported covid deaths per capita are significantly higher than Norway’s, but far lower than in Sweden’s.

State-to-state comparisons in the U.S. are less probative than in Scandinavia because of the prevalence of travel from neighboring state to neighboring state. For example, a big biker rally permitted in South Dakota will likely result in spreading the virus to neighboring states. However, the state-to-state comparisons I have focused on point to the same conclusion as the Scandinavian comparisons – lockdowns helped reduce deaths from covid.

Let’s start with the neighboring states of Minnesota and South Dakota. Other things being equal, I would expect more covid deaths per capita in Minnesota than in South Dakota. Minnesota is more densely populated than South Dakota and has a larger African-American population per capita. It’s my understanding that Blacks have been hit harder by the virus than Whites.

Yet, per capita deaths in South Dakota (2,277 per one million) are much higher than they are in Minnesota (1,331). South Dakota was one of America’s most lenient states with regard to lockdowns. Minnesota was one of the most stringent.

John compares Minnesota and Wisconsin, along with South Dakota and North Dakota. Regarding the Dakotas, John says that North Dakota locked down, while South Dakota did not, yet deaths per capita are almost the same in the two states.

But North Dakota was one of only a handful of states (South Dakota was another) that never issued a stay-at-home order. And throughout much of the pandemic, it apparently kept bars and restaurants open at 75 percent capacity. This helps explain why North Dakota’s per capita death rate is so high, much higher than Minnesota’s.

What about Wisconsin and Minnesota, where the per capita death rates are similar? Minnesota locked down stringently. Wisconsin also locked down until mid-May when a court overturned the state-wide order requiring this.

However, some jurisdictions in Wisconsin (Madison and Dane counties) promptly imposed their own lockdown orders. And the governor issued new restrictions in the Fall, when the pandemic was heating up again.

Thus, there was less divergence overall than meets the eye between the lockdown policies of Minnesota and Wisconsin.

But there was divergence. Why don’t the per capita death numbers diverge?

The answer lies, I believe, in nursing homes. By way of a possible explanation of the relatively low death rate in Minnesota compared to the Dakotas, John points out that Minnesota had a lower percentage of its population in nursing homes than did the Dakotas.

But what really matters is what happened in the nursing homes – how many people died in them. A jurisdiction in which officials fail to deal with the covid’s spread in nursing homes is going to experience higher death numbers than would otherwise be the case. Lockdowns won’t help much, if at all, because the virus is being spread inside the facilities.

Minnesota’s handling of nursing homes was disastrous. Scott has shown this in his long-running series on coronavirus in one state.

As of early March, 4,070 people had died in Minnesota nursing homes. Total deaths from the virus were around 6,700. So nursing home deaths in Minnesota were 61 percent of covid deaths in the state.

In Wisconsin, the picture is very different. As of the same date in March (when my source stopped reporting nursing home deaths, probably because of the high vaccination rates in these facilities), 1,976 people had died in Wisconsin nursing homes during the pandemic. Total Wisconsin deaths were around 7,500. Nursing home deaths, therefore, were only about 27 percent of covid deaths. (Some nursing home deaths during the pandemic were likely primarily to non-covid factors, but this is true in both Minnesota and Wisconsin.)

It looks to me, then, like nursing home deaths are what drove Minnesota’s overall numbers to the level of Wisconsin’s. We wouldn’t expect lockdowns to have much impact on the spread of the virus in nursing homes. They are intended primarily to limit the spread elsewhere.

If we focus on the situation outside of nursing homes, where one might expect lockdowns to prevent deaths, the Minnesota-Wisconsin comparison indicates that they did just that. It looks like Minnesota’s lockdowns prevented a non-trivial number of people not in nursing homes from dying of covid, though not without generating significant costs.

In South Dakota the percentage of deaths in nursing home as a percentage of covid deaths is lower than in North Dakota, although the difference isn’t as stark as the one between Wisconsin and Minnesota. In South Dakota the number is around 45 percent. In North Dakota, it’s around 60 percent. (These percentages are derived from the same source materials cited above for Wisconsin and Minnesota, as applied to the Dakotas.)

Absent this disparity at nursing homes, North Dakota’s overall covid death numbers would look better in comparison to South Dakota’s.

Finally, a note of caution. All of the numbers cited on both sides of this argument may be subject to some doubt. The counts are inherently imperfect and it’s possible that some have been manipulated.

That’s why I’m so focused on Scandinavia. The disparities there, especially between Norway and Sweden, are too enormous to be the result of error or manipulation.

These disparities point unambiguously to the conclusion that lockdowns — whatever we traded off by imposing them so widely, and it was plenty — have prevented deaths from covid. The evidence from states we have examined here reinforces this view, I believe.

Responses