If you’ve had a conversation with me in the last year, and COVID-19 was brought up at all, you likely heard me say that we are vastly undercounting the number of Covid deaths. I know you’ve heard the stories about people in car crashes being listed as COVID-19 deaths, but they’re no worth taking the time to disprove: you know how to Google. It’s not hard to see right through these stories—they are most often completely fabricated, or a massive exaggeration of some innocuous mistake that isn’t wide spread.
But, it has long been my stance that Covid deaths are vastly underexaggerated (is that a word? It is now). The reason for this is simple. With every death, one question should be asked: would this person had died if there was no COVID-19 pandemic? If the answer to that question is no, the death needs to be counted as a Covid death. I can hear my Qrazy friends yelling at me right now, and my answer is; yes, even if that person didn’t have Covid.
The reason for this is that the Covid fallout is just as important as the virus itself, and this is the case for all types of mass death events. It wouldn’t make sense to argue that people in Japan who died from starvation months after the US dropped atomic bombs did not die from the bombs: if there had been no bombs, there would have been no crop shortage. The same is true with Covid.
So, even if a person did not have Covid, but would have been alive if there were no pandemic, their death should be counted as a Covid death.
Here is an illustration of what I am talking about:
In other professions there may be a different perspective, but as an economist, when I get asked as part of my job (and I have been asked this by state officials) “How can we use Covid to prepare for the economic impact next pandemic,” I tell them what I am telling you now. To properly estimate the impacts of a pandemic we need to know not only the number of people who died from Covid, but also the number of people who died due to Covid.
From Covid is easy to understand. A person caught the virus and was killed by the virus. Due to Covid is less clear, and it’s what this letter is explaining—which will get us to the undercounting of Covid deaths.
The easiest to understand in the graph above is probably type 3 deaths. A healthy person died, so what else could have caused it. Type 1 deaths are the type so many people on The Facebooks kept arguing about. You know the ones: “They had cancer, that can’t be a Covid death,” or “Covid only killed them because they had asthma, it really wasn’t Covid.” But we have to ask the most important question—would the cancer or asthma have killed them at this early stage if there were no pandemic. If not, it should count as a Covid death. But, if someone’s cancer was late stage, and had only a few weeks left to live, that does not need to be counted as a Covid death.
Now, here is where it gets really tricky. Type 2 and 4 deaths. These are people who did not die from Covid, but also would not have died without the pandemic. The unhealthy type 2 could be a scenario where a dialysis patient could not get their dialysis because it was unsafe, or the hospital was on lockdown, or the dialysis wing was temporarily converted to a Covid wing. I’m not sure how often this happened, but the answer is probably not zero. If there were no pandemic, that patient would not have died at that time. Perhaps they would have gotten a kidney transplant if there were no pandemic. Type 2 deaths need to be counted.
Type 4 are the most tricky. Maybe this is someone who was so lonely in isolation they were driven to suicide (more on this later), which is something that would not have happened without the pandemic. Maybe it is a doctor who worked too many hours on too little sleep and suffered a heart attack. Maybe it is someone who was driving recklessly on the empty roads. Like I said, these are the most tricky, but if there were no virus, they would be alive.
From an economic perspective, all 4 types of deaths need to be counted. A dead person is a dead worker, to put it in the dispassionate economic context. That is one more person who cannot make things, buy things, or invest in things. This dead person lowers the productive capacity of their country (what economists refer to as potential GNP). To figure out how to plan for the next pandemic, the impacts cannot be fully realized until the change in productive capacity is understood. Even if someone died due to Covid, rather than from Covid, that change in productive capacity that happened because of the pandemic must be accounted for to plan for the next one.
Which brings me back to undercounting.
To figure out how many deaths there really were from Covid, we need to look at excess deaths—the number of people who died in a given time above and beyond the expected number. The NVSS (National Vital Statistics System) death numbers are out, and they paint an interesting picture.
The earliest numbers on excess deaths look like Covid is seriously undercounted.
As this chart shows, given the US population, we would expect 2.7 to 2.8 million deaths in the US in 2020. Instead there were over 3.3 million—an increase of half a million deaths. Even after taking confidence intervals into account 2020 is a huge outlier. Maybe there would have been high death numbers without a pandemic, but the likelihood of 500,000 extra deaths without the pandemic is approaching zero (asymptotes, for anyone who wants a throwback to your high school math class).
But here is the thing: only 345,323 deaths were officially counted as Covid. That still leaves us with roughly 150,000 unexplained deaths. That’s over 40 percent higher than the official count. If you were trying to plan out the next pandemic and left yourself 40 percent short, you’d be serious trouble when it hit.
So where did those excess deaths come from? Heart disease deaths are about 30,000 higher than normal, unintentional injuries are 20,000 higher, and then there is a smattering of smaller increases in things like strokes and diabetes.
One stat that really jumps out is that suicides, contrary to what Covid skeptics argued, actually went down.
The reason I have been saying Covid deaths were undercounted is because it was not hard to predict a huge number of excess deaths—a lot folks did so, not just me. 500,000 excess deaths is not terribly surprising, but 150,000 excess deaths not from Covid is pretty unexpected. Going back to the chart, the numbers show about 350,000 type 1 and 3 deaths. The most important question that can be asked about the numbers from NVSS is, “how many of the 150,000 additional excess deaths are type 2 and 4 deaths?” My hunch is, quite a lot—probably near 3/4 of them.
The categories that have seen the largest increases—Heart Disease, Stroke, Unintentional Injury, and Diabetes—fit really well into the type 2 and 4 boxes. Perhaps there was some natural increase in these deaths that would have happened without the pandemic, but the likelihood that all of these would have occurred without the pandemic is pretty darn small.
To count the impact of the pandemic, all of these need to be added in. It’s why I’ve been saying we are undercounting.
The NVSS data is only one dataset, and perhaps when we get more data, the complete picture might change. But this is a huge datapoint, and it should lead to a broader discussion about the full impact of the Pandemic.
I had a couple other thoughts, but the NCAA Championship game starts in 9 minutes. The point is, it is important to count all deaths due to the pandemic, not just from it. Otherwise we cannot get quality evaluations of economic impact, or make quality preparations for the next one.
Thanks for reading The Constituent. If you’d like to support the newsletter, here are a few options.
-Thanks,