You claim lockdowns are almost certainly net negative but plenty of research papers on the cost-benefit analysis of lockdowns come to the opposite conclusion. Lockdowns saved plenty of lives, prevented long Covid and stopped hospitals from being overwhelmed at the costs of ~5% of GDP and decreased mental health. I could see that for certain segments of the population such as young people that lockdowns are a net negative, though a large decision like lockdowns should likely be decided based off the benefits to all of society and for older people lockdowns were definitely a huge net positive.
Cost-benefit analyses of lockdowns are hard because on one side you have a bunch of reasonably well-defined positives: preventing deaths, and on the other you have a long tail of weird second-order long-term effects like children learning loss, calcification of social relationships because of masks, politicians getting used to exerting this sort of power over the daily lives of normal people, small business closures and all the second-order effects from that, increased videogame addiction, the dramatic increase of Tinder usage instead of face-to-face flirting, increased depression and suicides, that one study that showed babies born during the pandemic had lost 2 standard deviations of cognitive development compared to babies born earlier (babies are likely more resilient than we think and this loss will be temporary, but still). We’ve had pandemics and old people dying before, so the overall impacts can be better estimated than the impacts of the lockdowns and measures we’ve taken in the past 2 years, I think people have drastically underestimated the very long-term and far-reaching effects of these restrictions.
that one study that showed babies born during the pandemic had lost 2 standard deviations of cognitive development compared to babies born earlier
Whoa, that effect size is huge. Too big for me to believe it without more evidence. Seems more likely to be a confounding factor. The discussion section of that paper is pretty good, listing a bunch of hypothesis of what the reason could be, but not finding any obviously good ones.
One thing that stands out to me:
One aspect also not investigated here is the impact of mask-wearing by the study staff during child visits and assessments [53]. The inability of infants to see full facial expressions may have eliminated non-verbal cues, muffled instructions, or otherwise altered the understanding of the test questions and instructions.
This seems like it could be a big confounder. (Though it only makes sense if it has a differentially larger effect on younger children, since the cognitive loss supposedly applies to babies born during the pandemic rather than babies tested during the pandemic.)
babies are likely more resilient than we think and this loss will be temporary
What makes you think so? My prior is that ‘babies are more resilient than we think’ is a fashionable idea because the opposite would be tantamount to blaming parents, especially poor ones, and that’s unfashionable. I’m interested in learning more about the topic.
Here’s a study where they try to predict adult IQ from infant IQ, the correlation is something like 0.32, so about 10% of the variance of adult IQ is explainable by infant IQ, meaning that low-IQ babies can in fact end up with large IQs as adults, which would either indicate that infant cognitive development measures are pretty bad for predicting adult IQ or that it doesn’t matter that much what you do to a baby in the first years of life. There’s also the existence of periods of historical deprivation where babies have been subjected to much worse conditions than during lockdowns, and they didn’t all end up as village idiots like you’d expect from a 2 standard deviation drop. But still, even if most of the effect is going to disappear, something like a 5 point drop in IQ would be a catastrophe, let alone a 30 point one.
Please note that “infant IQ” is a very non-standard use; the article says “ability to process information” instead.
Before three years, things are so difficult to measure that even mental retardation is not diagnosed at that age. I am not surprised by the correlation being low; but from my perspective it simply means “if you try to measure a baby’s mental life, you will get a lot of noise”.
infant cognitive development measures are pretty bad for predicting adult IQ
They are probably also pretty bad for predicting IQ at the age of five years.
(I don’t have evidence for this that I could link; it’s just my understanding of how things work.)
One of the reasons is that in people 3 and more years old, we can distinguish between someone being generally smart/slow vs someone having a specific talent/disorder (e.g. “a genius kid with dyslexia”); but a baby does not have a sufficiently wide range of specialized activities to diagnose this.
Yeah, it’s overconfident to claim that lockdowns are “almost certainly net negative”. This stuff is complicated.
But it’s also not certain that lockdowns were “definitely a huge net positive” for older people. For example, for my 90 year old grandmother the life-saving benefits are much larger than for younger people. But the costs of a couple years in lockdown has also been huge for her. She’s been persistently depressed, and her health has deteriorated a lot. Presumably from not moving around much any more. She’s felt really bad about life since the pandemic started.
Especially given that her statistical risk of dying per year is something like 50% pre-covid, it’s not obvious whether this is a good trade-off. It all comes down to details about just how big the mental health costs are and the specific number for mortality reduction from covid.
Yeah, I wanted to make much the same point. My grandma died in 2015, but I’ve talked about it with my family members and we’re all pretty certain she would have chosen to die of COVID rather than go into lockdown. In the last years of her life (especially after being widowed in 2008), she was very realistic about the fact that she didn’t have much longer to live, and her whole life revolved around her community — she lived alone, but she was an important community fixture in her small town, and had been for decades. Not socializing would have removed her only source of meaning, and there’s no way she would have thought that it was worth locking down to buy herself two additional years of life in isolation.
Many cost-benefit analyses (especially early in the pandemic) compared modelled, not actual, benefits of lockdowns—note that epidemiological models have repeatedly proven overly pessimistic—usually with estimated short-term monetary costs of lockdowns. Generally these cost-benefit analyses ignored non-monetary costs, even though they’re huge. They also tended to calculate benefits by comparing lockdowns vs doing nothing, ignoring options like supporting the elderly to isolate themselves while letting everyone else go about their lives.
We now know that lockdowns had minimal-to-zero benefits in terms of preventing short-term excess deaths: https://www.nber.org/papers/w28930#fromrss (If you don’t have patience to read the paper, just consider Sweden’s average-for-Europe performance despite its decision not to lockdown, or the minimal difference between the outcomes of US red states and blue states.)
So given that the actual factually-demonstrated benefits are statistically indistinguishable from zero (so realistically some small positive), and the costs are immense, I find it hard to claim that a reasonable cost-benefit analysis would support lockdowns.
Do you have links to any scientifically rigorous cost-benefit analyses that actually take into account all Covid costs, including long-term impacts to health, education and life expectancy? And that also use factual inputs when estimating benefits rather than epidemiological models of dubious accuracy? If you can post links to such high-quality cost-benefit analyses, I will consider updating my views. If not, I hope you will update yours.
Compared to other Scandinavian countries, the Swedish death rate is very high: Sweden has 1459 covid deaths per million. For Norway that number is 238, for Finland it is 280 and for Denmark it is 559.
The Danish population density is more than 5 times that of the other Scandinavian countries, so perhaps Scandinavia without Denmark is a better reference class. I’m not claiming that Scandinavia (with or without Denmark) is the correct reference class, I’m just saying that Europe is not the only possible reference class and other reference classes paints a completely different picture.
Population density is entirely the wrong metric to look at here. You could fudge the denmark “population density” count by just including Greenland, and including the empty swathes of land in the nordic countries has the same effect.
Well said. Eyeballing the chart shows no correlation between population density and covid death rate. Looking strictly at Europe, CSPI has a good chart (halfway down the long detailed post) showing minimal correlation. My take on population density is that the pro-lockdown media in March and April 2020 were predicting bodies piled high in the streets and the total collapse of healthcare systems without lockdowns. Because back then populations hadn’t normalised extreme losses of freedom so you needed to predict Armageddon in order to sell the lockdowns. Sweden, and later the US red states, proved that Armageddon was factually false. So the journalists needed some rationalisations for why Sweden hadn’t turned into a hell-hole and population density was the first thing they thought of.
If you still want to control for population density, compare Stockholm against European cities likeParis with similar population density. I’m pretty sure I read a good report finding that Stockholm was average again, but I seem to have lost the link.
Fair point. The relevant metric might be something like average population density around an average person (or density might not be that relevant at all). Looking at this map of population density, my conclusion is that every country is different, so using Scadinavia without Denmark as a reference class is probably unfair cherry-picking. Still, Sweden has more than 2.6 times as many covid deaths per million as any other Scandinavian country.
My prior for this is that the population-level differences are probably almost entirely centered around care-homes, and questions related to care-homes (or, more generally, very old people). Since my knowledge on comparisons between care-homes in scandinavian countries is close to zero, I cannot really provide any insight here. But something as banal as “did people who are sick stop going to work in care-homes” will probably bias the results far more than population density.
Edit: regarding population density: I don’t think there’s no effect, I just don’t know what the effect is or if it is even monotone. Almost all the people I know who got covid either (a) live in a village and don’t trust the government or (b) are highly-connected even for big-city standards. So it could be that there is a sweet spot local-minimum where people in cities feel super at risk because of all the strangers around them and isolate, but the effects of the additional density don’t meaningfully lead to more covid.
Maybe that’s more from lower vaccination rates, than lockdowns—but it still undermines the argument that´s based on no significant red/blue state differences.
Dumbledore’s Army was speaking about excess deaths. The article to which you link is about COVID mortality.
In addition, it says “Either way, the per capita death toll in blue America and red America was similar by the final weeks of 2020.” That suggests that the lockdowns didn’t cause a difference in COVID-19 mortality but the vaccines actually did. Sweden is also an example of no lockdown but decent vaccine uptake.
In one thread someone wrote about one rationalist dying to sepsis because they locked down to the extent that they didn’t go to the hospital. I have one friend who said that three acquaintances of hers committed suicide during the lockdowns.
Looking at excess deaths makes a lot more sense because the additional deaths caused by the lockdown matter.
Yeah, I agree that excess death data is preferable when available. For some reason Dumbledore’s Army’s original link isn’t working for me (“page not found”). So I haven’t yet seen state by state excess mortality data. But if it actually doesn’t find any difference between the red/blue states that would undermine the argument from the NYT article above.
Looking at Our World in Data’s limited cumulative excess mortality data Sweden has 2-8X higher excess mortality during the pandemic compared to other Scandinavian countries (with similar vaccination rates). That undermines any simple arguments based on Sweden (other than ones with weaker conclusions—such as that avoiding lockdowns don’t 10X+ net covid mortality).
Ofc there could be some other explanation for excess Swedish mortality. But an argument against lockdowns based on Sweden as a datapoint would need a pretty solid explanation of this.
Fixed the link, thanks for letting me know it wasn’t working. If For whatever reason that link doesn’t work, it’s called The Impact of the COVID-19 Pandemic and Policy Responses on Excess Mortality andthe authors are Agrawal, Cantor, Sood and Whaley.
Looking at Our World in Data’s limited cumulative excess mortality data Sweden has 2-8X higher excess mortality during the pandemic compared to other Scandinavian countries
“The cumulative difference between the reported number of deaths since 1 January 2020 and the projected number of deaths for the same period based on previous years.”
Sweden 883 Finland 411 Denmark 154 Norway 110 Iceland 92
Ah, that data isn’t cumulative. It is just looking at current excess mortality. A lot of Sweden’s excess mortality happened early on (I believe, while the other Scandinavian countries were locking down more). So the cumulative number is higher, but not the current number.
I think it’s entirely from different vaccination rates. Check the stats for 2020 only ie before vaccines were widely available. Because there were indeed no significant red-blue state differences then. The paper I linked above includes all US states in its analysis as well as 43 other countries.
Edited to add: try going through the US states data in this report, written March 2021 ie before the Delta wave with differential vaccine rates.
There has been some evidence both ways, but I think that the best argumentation has been that lockdowns were net-negative by a lot. Dumbledore’s Army linked this paper, which is not the same as the one that originally convinced me, but it’s also good and I don’t have the other link on hand.
In particular, analyses in favor of lockdowns tended to make two errors:
They compare a lockdown to a hypothetical situation in which there is no mitigation at all, rather than allowing for private, voluntary mitigation.
They assume that infections which do not occur during the lockdown never occur at all, rather than simply occurring later. Since we never got close to actually eradicating COVID, deaths during lockdown were mostly delayed rather than prevented. (The exception would be the cases where we delayed long enough to get good treatments or vaccines, or where we avoided medical system collapse, but even taking these into account the benefit is still smaller than assumed.)
Most importantly: relying on voluntary isolation allows people to choose their level of risk based on their vulnerability, their risk tolerance, and their ability to bear the costs. This is not perfect (you wind up with some people who would prefer to self-isolate but can’t do so for economic reasons), but it’s still miles better than forcing the same policy onto everyone.
From earlier this year: Douglas Allen of Simon Fraser University reviews 80 studies, concludes that lockdowns have costs which are 3.6 − 282x more than their benefits. The 3.6 is based on unrealistic extreme assumptions intended to steel-man the case for lockdown. He concludes: “It is possible that lockdown will go down as one of the greatest peacetime policy failures in Canada’s history.”
I repeat my request for you to post links to any high-quality cost-benefit analyses which come to the opposite conclusion.
You claim lockdowns are almost certainly net negative but plenty of research papers on the cost-benefit analysis of lockdowns come to the opposite conclusion. Lockdowns saved plenty of lives, prevented long Covid and stopped hospitals from being overwhelmed at the costs of ~5% of GDP and decreased mental health. I could see that for certain segments of the population such as young people that lockdowns are a net negative, though a large decision like lockdowns should likely be decided based off the benefits to all of society and for older people lockdowns were definitely a huge net positive.
Cost-benefit analyses of lockdowns are hard because on one side you have a bunch of reasonably well-defined positives: preventing deaths, and on the other you have a long tail of weird second-order long-term effects like children learning loss, calcification of social relationships because of masks, politicians getting used to exerting this sort of power over the daily lives of normal people, small business closures and all the second-order effects from that, increased videogame addiction, the dramatic increase of Tinder usage instead of face-to-face flirting, increased depression and suicides, that one study that showed babies born during the pandemic had lost 2 standard deviations of cognitive development compared to babies born earlier (babies are likely more resilient than we think and this loss will be temporary, but still). We’ve had pandemics and old people dying before, so the overall impacts can be better estimated than the impacts of the lockdowns and measures we’ve taken in the past 2 years, I think people have drastically underestimated the very long-term and far-reaching effects of these restrictions.
Whoa, that effect size is huge. Too big for me to believe it without more evidence. Seems more likely to be a confounding factor. The discussion section of that paper is pretty good, listing a bunch of hypothesis of what the reason could be, but not finding any obviously good ones.
One thing that stands out to me:
This seems like it could be a big confounder. (Though it only makes sense if it has a differentially larger effect on younger children, since the cognitive loss supposedly applies to babies born during the pandemic rather than babies tested during the pandemic.)
What makes you think so? My prior is that ‘babies are more resilient than we think’ is a fashionable idea because the opposite would be tantamount to blaming parents, especially poor ones, and that’s unfashionable. I’m interested in learning more about the topic.
Here’s a study where they try to predict adult IQ from infant IQ, the correlation is something like 0.32, so about 10% of the variance of adult IQ is explainable by infant IQ, meaning that low-IQ babies can in fact end up with large IQs as adults, which would either indicate that infant cognitive development measures are pretty bad for predicting adult IQ or that it doesn’t matter that much what you do to a baby in the first years of life. There’s also the existence of periods of historical deprivation where babies have been subjected to much worse conditions than during lockdowns, and they didn’t all end up as village idiots like you’d expect from a 2 standard deviation drop. But still, even if most of the effect is going to disappear, something like a 5 point drop in IQ would be a catastrophe, let alone a 30 point one.
Please note that “infant IQ” is a very non-standard use; the article says “ability to process information” instead.
Before three years, things are so difficult to measure that even mental retardation is not diagnosed at that age. I am not surprised by the correlation being low; but from my perspective it simply means “if you try to measure a baby’s mental life, you will get a lot of noise”.
They are probably also pretty bad for predicting IQ at the age of five years.
(I don’t have evidence for this that I could link; it’s just my understanding of how things work.)
One of the reasons is that in people 3 and more years old, we can distinguish between someone being generally smart/slow vs someone having a specific talent/disorder (e.g. “a genius kid with dyslexia”); but a baby does not have a sufficiently wide range of specialized activities to diagnose this.
Yeah, it’s overconfident to claim that lockdowns are “almost certainly net negative”. This stuff is complicated.
But it’s also not certain that lockdowns were “definitely a huge net positive” for older people. For example, for my 90 year old grandmother the life-saving benefits are much larger than for younger people. But the costs of a couple years in lockdown has also been huge for her. She’s been persistently depressed, and her health has deteriorated a lot. Presumably from not moving around much any more. She’s felt really bad about life since the pandemic started.
Especially given that her statistical risk of dying per year is something like 50% pre-covid, it’s not obvious whether this is a good trade-off. It all comes down to details about just how big the mental health costs are and the specific number for mortality reduction from covid.
Yeah, I wanted to make much the same point. My grandma died in 2015, but I’ve talked about it with my family members and we’re all pretty certain she would have chosen to die of COVID rather than go into lockdown. In the last years of her life (especially after being widowed in 2008), she was very realistic about the fact that she didn’t have much longer to live, and her whole life revolved around her community — she lived alone, but she was an important community fixture in her small town, and had been for decades. Not socializing would have removed her only source of meaning, and there’s no way she would have thought that it was worth locking down to buy herself two additional years of life in isolation.
Many cost-benefit analyses (especially early in the pandemic) compared modelled, not actual, benefits of lockdowns—note that epidemiological models have repeatedly proven overly pessimistic—usually with estimated short-term monetary costs of lockdowns. Generally these cost-benefit analyses ignored non-monetary costs, even though they’re huge. They also tended to calculate benefits by comparing lockdowns vs doing nothing, ignoring options like supporting the elderly to isolate themselves while letting everyone else go about their lives.
We now know that lockdowns had minimal-to-zero benefits in terms of preventing short-term excess deaths: https://www.nber.org/papers/w28930#fromrss (If you don’t have patience to read the paper, just consider Sweden’s average-for-Europe performance despite its decision not to lockdown, or the minimal difference between the outcomes of US red states and blue states.)
So given that the actual factually-demonstrated benefits are statistically indistinguishable from zero (so realistically some small positive), and the costs are immense, I find it hard to claim that a reasonable cost-benefit analysis would support lockdowns.
Do you have links to any scientifically rigorous cost-benefit analyses that actually take into account all Covid costs, including long-term impacts to health, education and life expectancy? And that also use factual inputs when estimating benefits rather than epidemiological models of dubious accuracy? If you can post links to such high-quality cost-benefit analyses, I will consider updating my views. If not, I hope you will update yours.
Compared to other Scandinavian countries, the Swedish death rate is very high: Sweden has 1459 covid deaths per million. For Norway that number is 238, for Finland it is 280 and for Denmark it is 559.
The Danish population density is more than 5 times that of the other Scandinavian countries, so perhaps Scandinavia without Denmark is a better reference class. I’m not claiming that Scandinavia (with or without Denmark) is the correct reference class, I’m just saying that Europe is not the only possible reference class and other reference classes paints a completely different picture.
Covid numbers are from https://www.worldometers.info/coronavirus/
Population density is entirely the wrong metric to look at here. You could fudge the denmark “population density” count by just including Greenland, and including the empty swathes of land in the nordic countries has the same effect.
Well said. Eyeballing the chart shows no correlation between population density and covid death rate. Looking strictly at Europe, CSPI has a good chart (halfway down the long detailed post) showing minimal correlation. My take on population density is that the pro-lockdown media in March and April 2020 were predicting bodies piled high in the streets and the total collapse of healthcare systems without lockdowns. Because back then populations hadn’t normalised extreme losses of freedom so you needed to predict Armageddon in order to sell the lockdowns. Sweden, and later the US red states, proved that Armageddon was factually false. So the journalists needed some rationalisations for why Sweden hadn’t turned into a hell-hole and population density was the first thing they thought of.
If you still want to control for population density, compare Stockholm against European cities likeParis with similar population density. I’m pretty sure I read a good report finding that Stockholm was average again, but I seem to have lost the link.
Fair point. The relevant metric might be something like average population density around an average person (or density might not be that relevant at all). Looking at this map of population density, my conclusion is that every country is different, so using Scadinavia without Denmark as a reference class is probably unfair cherry-picking. Still, Sweden has more than 2.6 times as many covid deaths per million as any other Scandinavian country.
My prior for this is that the population-level differences are probably almost entirely centered around care-homes, and questions related to care-homes (or, more generally, very old people). Since my knowledge on comparisons between care-homes in scandinavian countries is close to zero, I cannot really provide any insight here. But something as banal as “did people who are sick stop going to work in care-homes” will probably bias the results far more than population density.
Edit: regarding population density: I don’t think there’s no effect, I just don’t know what the effect is or if it is even monotone. Almost all the people I know who got covid either (a) live in a village and don’t trust the government or (b) are highly-connected even for big-city standards. So it could be that there is a sweet spot local-minimum where people in cities feel super at risk because of all the strangers around them and isolate, but the effects of the additional density don’t meaningfully lead to more covid.
Hm. I wonder if there’s really a ” minimal difference between the outcomes of US red states and blue states”. From the graph here it looks like red states had ~40% higher mortality per capita: https://www.nytimes.com/2021/11/08/briefing/covid-death-toll-red-america.html
Maybe that’s more from lower vaccination rates, than lockdowns—but it still undermines the argument that´s based on no significant red/blue state differences.
Dumbledore’s Army was speaking about excess deaths. The article to which you link is about COVID mortality.
In addition, it says “Either way, the per capita death toll in blue America and red America was similar by the final weeks of 2020.” That suggests that the lockdowns didn’t cause a difference in COVID-19 mortality but the vaccines actually did. Sweden is also an example of no lockdown but decent vaccine uptake.
In one thread someone wrote about one rationalist dying to sepsis because they locked down to the extent that they didn’t go to the hospital. I have one friend who said that three acquaintances of hers committed suicide during the lockdowns.
Looking at excess deaths makes a lot more sense because the additional deaths caused by the lockdown matter.
Yeah, I agree that excess death data is preferable when available. For some reason Dumbledore’s Army’s original link isn’t working for me (“page not found”). So I haven’t yet seen state by state excess mortality data. But if it actually doesn’t find any difference between the red/blue states that would undermine the argument from the NYT article above.
Looking at Our World in Data’s limited cumulative excess mortality data Sweden has 2-8X higher excess mortality during the pandemic compared to other Scandinavian countries (with similar vaccination rates). That undermines any simple arguments based on Sweden (other than ones with weaker conclusions—such as that avoiding lockdowns don’t 10X+ net covid mortality).
Ofc there could be some other explanation for excess Swedish mortality. But an argument against lockdowns based on Sweden as a datapoint would need a pretty solid explanation of this.
Fixed the link, thanks for letting me know it wasn’t working. If For whatever reason that link doesn’t work, it’s called The Impact of the COVID-19 Pandemic and Policy Responses on Excess Mortality and the authors are Agrawal, Cantor, Sood and Whaley.
No. https://ourworldindata.org/excess-mortality-covid is the page for excess mortality.
Finland has +19 pp absolute change and +165% relative change while Sweden only has +8 pp and +125%. Norway even has +24 pp and +355%.
Sweden has lower excess mortality than Finland and Norway.
Yeah, precisely that page. Scroll down to the graph:
”Excess mortality: Cumulative number of deaths from all causes compared to projection based on
previous years, per million people, Dec 19, 2021
“The cumulative difference between the reported number of deaths since 1 January 2020 and the projected number of deaths for the same period based on previous years.”
Sweden 883
Finland 411
Denmark 154
Norway 110
Iceland 92
Proportions are similar if you check out the economist’s data below: https://ourworldindata.org/grapher/excess-deaths-cumulative-per-100k-economist?country=OWID_WRL~CHN~IND~USA~IDN~BRA
Where are you getting your numbers?
I’m referencing the numbers on https://ourworldindata.org/grapher/excess-mortality-p-scores-projected-baseline?tab=table&country=MEX~RUS~ZAF
Sweden has a higher population than the other countries listed so total numbers are not comparable. That alone doesn’t explain all the difference.
It’s unclear to me why https://ourworldindata.org/grapher/excess-mortality-p-scores-projected-baseline?tab=map&country=MEX~RUS~ZAF and https://ourworldindata.org/grapher/cumulative-excess-deaths-per-million-covid come to such different conclusions.
“Sweden has a higher population than the other countries listed so total numbers are not comparable. That alone doesn’t explain all the difference.”
The numbers I’m citing above are population normalized. They are total excess deaths per million (and per 100k in the economist link).
”It’s unclear to me why https://ourworldindata.org/grapher/excess-mortality-p-scores-projected-baseline?tab=map&country=MEX~RUS~ZAF and https://ourworldindata.org/grapher/cumulative-excess-deaths-per-million-covid come to such different conclusions.”
Ah, that data isn’t cumulative. It is just looking at current excess mortality. A lot of Sweden’s excess mortality happened early on (I believe, while the other Scandinavian countries were locking down more). So the cumulative number is higher, but not the current number.
I think it’s entirely from different vaccination rates. Check the stats for 2020 only ie before vaccines were widely available. Because there were indeed no significant red-blue state differences then. The paper I linked above includes all US states in its analysis as well as 43 other countries.
Edited to add: try going through the US states data in this report, written March 2021 ie before the Delta wave with differential vaccine rates.
Funnily enough, I have yet to read a single not-completely-ridiculous cost-benefit analysis that goes this way. We must live in different bubbles.
There has been some evidence both ways, but I think that the best argumentation has been that lockdowns were net-negative by a lot. Dumbledore’s Army linked this paper, which is not the same as the one that originally convinced me, but it’s also good and I don’t have the other link on hand.
In particular, analyses in favor of lockdowns tended to make two errors:
They compare a lockdown to a hypothetical situation in which there is no mitigation at all, rather than allowing for private, voluntary mitigation.
They assume that infections which do not occur during the lockdown never occur at all, rather than simply occurring later. Since we never got close to actually eradicating COVID, deaths during lockdown were mostly delayed rather than prevented. (The exception would be the cases where we delayed long enough to get good treatments or vaccines, or where we avoided medical system collapse, but even taking these into account the benefit is still smaller than assumed.)
Most importantly: relying on voluntary isolation allows people to choose their level of risk based on their vulnerability, their risk tolerance, and their ability to bear the costs. This is not perfect (you wind up with some people who would prefer to self-isolate but can’t do so for economic reasons), but it’s still miles better than forcing the same policy onto everyone.
From earlier this year: Douglas Allen of Simon Fraser University reviews 80 studies, concludes that lockdowns have costs which are 3.6 − 282x more than their benefits. The 3.6 is based on unrealistic extreme assumptions intended to steel-man the case for lockdown. He concludes: “It is possible that lockdown will go down as one of the greatest peacetime policy failures in Canada’s history.”
I repeat my request for you to post links to any high-quality cost-benefit analyses which come to the opposite conclusion.