How Many People Are Being Killed by the Lockdown?

From the Washington Examiner, and op-ed, via AEI:

Although lockdowns are preventing some deaths, they are undoubtedly increasing deaths by other causes. This virus is killing people not only by infecting millions but also by inducing a policy response that kills people.

First, …Isolation kills in many ways. We don’t have data that the lockdowns are causing more suicides, but plenty of health experts believe they will. People struggling with drug addiction are finding it harder to get the treatment they need. This could cause relapses and eventually, death.

It’s reasonable to worry about increases in drinking, and thus alcoholism and alcohol-related deaths. (Though, with less driving, we’ll have fewer DWIs.) …

People are also missing out on crucial medical care because of the coronavirus. … Many treatments are simply not happening. The result is that ill people don’t get the healthcare they would be getting.

All these points make sense, although the relevant question is whether the quantitative magnitude is significant. One way to get at this question is to look at excess mortality (which is an estimated figure) and compare against reported covid-19 fatalities. (a statistically reported figure, which is likely subject to undercounting).

Using CDC data accessed a couple days ago, I find that most of the excess mortality is accounted for (in an accounting sense) by covid-19 fatalities.

Figure 1: Excess mortality (blue), and reported covid-19 related fatalities (red), by week. Source: CDC, accessed 5/23/2020, and author’s calculations.

I end the sample at 4/18 because the more recent data are subject to very large revisions (according to a comparison of data ending 5/9 vs. data ending 5/2).

From 3/28 through 4/18, reported covid-19 fatalities accounts for 94.3% of excess fatalities (point estimate).



25 thoughts on “How Many People Are Being Killed by the Lockdown?

  1. pgl

    Is Carney trying to tell us that all of the excess deaths are caused by non-COVID factors? Another AEI quality publication! I guess this is why he had to go to the Washington Examiner to get this published!

  2. Dr. Dysmalist

    So out of the 5.7% “unaccounted for” we know that some of them are stochastic, some are likely due to improperly attributed covid-19 deaths, and some may be due to effects of the lockdowns. It is obvious (to me, at least) that the third reason is a tiny, tiny fraction of total deaths during this pandemic and cannot, therefore, hold up as a rationale for opening up.

  3. dilbert dogbert

    They will = USToday?
    Harder = Politico?
    Increases in drinking = Gurardian?

  4. Jeanne C Scown

    Many of the unaccounted for “excess deaths” may be untested covid-19 victims. It is really hard to trust all the numbers at the moment. We will have to wait until this administration is gone.

  5. Mark

    Automobile accidents result in 1.25 million deaths per year. Decreased driving during the lockdown will save 7.3 times as many deaths by car wreck as the number of suicides it causes. No, I don’t have any data to show this, but who needs data in today’s world?

    1. Barkley Rosser

      In the US automobile deaths were about 38,000 in 2018, several orders of magnitude fewer than what you claim, although it is probably true that the number has been reduced somewhat by the lockdowns.

  6. ltr

    “Although lockdowns are preventing some deaths, they are undoubtedly increasing deaths by other causes. This virus is killing people not only by infecting millions but also by inducing a policy response that kills people.”

    Crazily offensive lying.

  7. Bruce Hall

    Your overall assumptions seem reasonable. The issue, of course, continues to be the accuracy and consistency of reporting. Some sources say Covid-19 deaths are understated while others indicate that deaths from other causes are being attributed to Covid-19. There is no doubt that Covid-19 has caused additional deaths. The exact extent will probably have to remain in a range due to the reporting issues and the probability that death from other diseases are being included as Covid-19 deaths whenever the virus is present.

    USNews had an article dated April 30 (just past the end of your data) that attempts to explain, but not necessarily reconcile, the data.

    Whether or not deaths from other causes are increasing, reports are “anecdotal”:
    At the University of Rochester Medical Center in Rochester, N.Y., emergency room visits dropped by 50 percent, and many of the patients who do come have waited too long to seek treatment. They “are presenting late with strokes and heart attacks,” said Dr. Michael Apostolakos, the system’s chief medical officer. “Or they’re not showing up until they can barely breathe from heart failure.”
    The number of people in the District dying outside hospitals spiked as the novel coronavirus started its sweep through the nation’s capital, raising concerns that people suffering a wide range of critical ailments are not seeking medical attention.
    Some of those people contracted the virus; the city has so far confirmed such deaths at home for three residents. But officials suspect many of the deaths are not related to the virus and may be the result of heart attacks, drug overdoses or other causes.

    Another area that has come under scrutiny has been nursing homes and state policies (NY, MI) that required nursing homes to accept patients without testing or Covid-19 status. Both states reconsidered after significant backlash. Regardless, the issue remains in many states and this can be construed to be a problem of state policies and regulation.

    In April, Dr. Brix stated that if someone dies with the Covid-19 present, it would be counted as a Covid-19 death. Given that such a large percentage of deaths recorded as Covid-19-caused occurred in nursing home patients over the age of 85, it is not unreasonable to suspect that at least a portion of those deaths were actually from other causes.
    So, yes, Covid-19 has caused many deaths that might not have occurred in a “normal” year. But the water is not totally clear on “excess” deaths.

    1. Menzie Chinn Post author

      Bruce Hall: This is the same standard that is used to collect influenza deaths. Hence, if you have a problem with covid-19 stats here, then you have a problem with influenza stats. Yet I have never seen you criticize flu stats. On the other hand, we have ample anecdotal evidence of missed fatalities in homes, uncounted fatalities in nursing homes, and orders from the Florida governor not to report covid-19 fatalities. So which direction do you think the numbers are biased?

    2. pgl

      “Some sources say Covid-19 deaths are understated”.

      Like everyone who has a brain and is not a Trump sycophant. Which of course excludes you on both counts.

    3. pgl

      “Another area that has come under scrutiny has been nursing homes and state policies (NY, MI) that required nursing homes to accept patients without testing or Covid-19 status. Both states reconsidered after significant backlash.”

      You have peddled this BS before, Cuomo has addressed this issue several times. The problem was caused when he followed the advice from the White House. Yes our governor caught on that following Trump’s advice was a really bad idea. BTW assuming you have followed the news – NY has better testing than any other state despite the incompetence on testing ala Trump. But of course you will never admit this reality.

    4. pgl

      Contrast what Bruce’s The Guardian truthfully wrote versus what Bruce wrote:

      The Guardian:

      Parikh said limited testing and a lack of personal protective equipment such as masks hampered efforts to curb the spread of Covid-19 in care homes. Due to limited testing capacity, most state nursing homes are still only able to test residents with symptoms, even though the disease is known to spread asymptomatically. “What I would like to see is the ability to test the entire nursing homes,” Parikh said. “This symptomatic approach is just not cutting it. Many states, including Connecticut, are starting to move in that direction … but I hope it becomes a national effort.”

      Yes The Guardian points to the lack of protective gear and the chaos from the Trump White House on our ability to test. Here is what Brucie boy wrote:

      “Another area that has come under scrutiny has been nursing homes and state policies (NY, MI) that required nursing homes to accept patients without testing or Covid-19 status.”

      No attempt by the nursing home to do what they could do with their limited abilities (thanks to Trump’s incompetence)? That is NOT what his own links said at all.

      OK – we know Bruce is too stupid to read his own damn links. But this blame shifting is based on a rather disgusting lie.
      Come on Bruce – we know you are Trump’s lap dog. But this level of sheer dishonesty?

    5. macroduck

      The author of the OpEd piece, Tim Carney, is not thinking clearly. Here’s how you can tell: The comparison that needs to be made is between loss in the current situation vs potential loss in some alternate situation, not just a counting up of various kinds of loss in the current situation. this comparison of two different states is utterly commonplace in economics, going under the label of “opportunity cost”. You do not learn whether some other choice is better by assessing the results of the current approach – you assess both the current approach and the alternative. Carney made no effort to assess an alternative in which people are allowed (and in many cases compelled) to return to work. An economic argument which implies that some other choice is better than the current one but which ignores opportunity cost is wretchedly inadequate.

      If there were no effort at distancing, hospitals would face even more Covid patients than they do now, so that medical treatment for non-Covid ailments might very well be less available under less stringent distancing, rather than more available, as Carney implies. We have evidence that social distancing efforts improve economic performance over the medium term. during the Spanish Flu epidemic, economic losses were reduced by social distancing efforts. the author is apparently ignorant of that fact.

      Even so, Brucey has teamed up with Carney – quel supris.

      Brucey has his own problem with standard analytical tools. His problem is with the concept of “excess deaths”:

      An octogenarian dies with Covid. That death is counted (usually, but not always) as a Covid-related death. Why? Because an “excess death” is not a death that would never have happened – you see the problem with that, yes? An “excess death” is a death that happened before it otherwise would have. There are mortality data for the 85-year-old set and I assure you, they show that the average life expectancy longer than two weeks. If an 85-year-old contracts Covid and dies soon thereafter, it is an “excess death” because the victim died sooner than is likely in the absence of Covid.

      Brucey’s dismissal of early death for the aged is part of the general Trumpist argument for reopening the economy. If a few people die early, that’s a fair price to pay for an economy that will help Trump get re-elected (thereby avoiding prison), especially since most of the ones who die will be disposable – old, black, sick, urban. As long as the well-to-do politician or pundit lives, others can jolly well die. That is the creed of the right, after all.

      1. baffling

        trump is not worried about those that die of covid. dead people cannot vote. his concern is about those that are still alive. his job is to convince them he is the reason they are still alive.

      2. 2slugbaits

        If a few people die early, that’s a fair price to pay for an economy that will help Trump get re-elected

        I believe that Bruce and Trump share a Maoist vision of a Great Leap Forward. The old and weak must be sacrificed in the service of a “V” shaped recovery.

  8. baffling

    “We don’t have data that the lockdowns are causing more suicides, …”
    at this point, the reader should simply STOP. from here on out, conjecture and opinion.

    i don’t have any data to show that rick stryker sits behind his computer in fishnet stockings and drag queen red lipstick, but plenty of internet trolls tell me his secretive online rendezvous….

  9. Barkley Rosser

    Someone named Jeffrey Tucker at AIER on May 22 published an article claiming lockdowns would increase suicides by increasing unemployment. Thee certainly is such a historic correlation, and we may yet see it, but apparently there has not yet been any noticeable increase in the US suicide rate. The main suicides that have been reported related to it have not been unemployed people but overstressed medical personnel dealing with the virus first hand.

  10. Baffling

    Trump is presiding over the worst economy ever while being engulfed by a virus he is too incompetent to handle. So what does private bone spurs do? Golf.

    1. macroduck

      Come now, you are being too harsh. Imagine a six-foot man (however small his hands may be), swinging a golf club. That action automatically establishes distancing from others on the golf course. Being out of doors further reduces the risk of contagion. Trump is modelling good behavior!

      Now, we know that medical personnel are among the most at risk of contracting Covid. Clearly, they all need to shed their greens (blues, whatever), put on some goofy pants and head for the links. This would not only reduce Covid transmission, but would also stimulate the economy. Think of all the goofy pants sales! Golf will be our salvation.

Comments are closed.