There’s been a lot of discussion about whether the pandemic lockdowns have been worth it. However, much of the reasoning that we’ve seen has been very motivated and un-nuanced in a way that for us has distorted a lot of the information.
So this is not a thread for taking a position on that. This is a thread for raising individual considerations that are relevant for thinking about the question “Have the pandemic lockdowns, in general, been worth it?”
Every answer to this thread should analyze a single belief that is relevant to whether pandemic lockdowns have been worth it, such as
“No lockdown would have led to 300,000 additional loss of QALYs in the UK”
“GDP $1b lower than the counterfactual in Germany”
“The West will develop a vaccine in 6 months”
“A year in lockdown is worth 70% of a normal year and 85% of a covid year without lockdown”
“The impetus to try remote work has led to changes in workplace norms worth $1 trillion over the next decade”
and provide relevant facts, data and information available about that factor.
Answers in this thread should not attempt to take a position on the overall question. Answers that take a position on the overall question will be deleted. This thread will live up to the virtue of holding off on proposing solutions.
(By ‘lockdown’ we refer to the thing that the US, UK and China have been doing, and what Sweden didn’t. There is naturally a lot of variation between countries, so this cannot have a canonical answer. If your consideration only applies to a small number of countries, that is fine.)
The World Bank now forecasts an that COVID-19 will push 88 million to 115 million people into poverty in 2020. Extreme poverty in this case is defined as living on $1.90 or less per day. This is the first reversal in global poverty in decades taking us back 3 years. The change in trend is very sharp, see figure on page 5.
Many of the new poor are in the urban informal sector, where government redistribution is unlikely to help. The world bank report stresses government interventions in poor countries, but the informal sector is hard to reach (unregistered, less politically powerful, less organized).
Some of these mechanisms are intensified by lockdown policies. The World Bank’s report scrupulously avoid that connection, but I suspect it is important. Disambiguating between rich-country lockdowns and poor country lockdowns is important.
Rich county lockdowns
Much of the effect comes from the contraction of global gdp of 5-8 percent. To the extent that lockdowns increase the GDP reduction, they contributed to the loss. The decisions of the wealthiest countries to lockdown contracted demand for tourism and manufactured goods. South-Asian exporters like India, Bangladesh and Indonesia come up repeatedly in the report. Also the new poor are more urban and formerly worked in tourism and manufacturing. This suggests that western consumer choices contributed to the increase. Probably a minority of total change, roughly.
Poor Country Lockdowns
Like in west, service workers have the least education. They got hit the hardest by lockdowns in counries like India and Ethiopia. Millions of Indian migrant workers had to migrate by foot in one crazy week in India. Documented events along indicate hundreds of deaths. Malnutrition likely the biggest killer. Kids have a lot if QALY’s left.
Total effect?
Back of the envelope calculation. Let’s assume the increase in poverty is 110 Million. It’s unclear when he affect washes out over time, but let’s say that it persists for 5 years. Currently 40% of the extreme poor in SSA and south Asia are 0-14. Over five years, the number of children who will go through the dangerous begining of life will be
110 x 10^6 x .4 x .33 = 15 x 10^6 additional children growing from 0 to 5 in extreme poverty.
A cursory look at OWID’s child mortality and income plots suggests the change in child mortality is about 5%. So assume that an 5% additional counterfactual deaths. Assume 70 QALY’s per child.
5 x 10^6 x .05 x 70 = 54 * 10^6 lost QALYs
Then also assume that lockdowns caused 1⁄4 of the increase in global poverty. Just a guess.
54 x 10^5 / 4 = 13 x 10^6 lost QALYs from lockdowns.
Given my high uncertainties, 1.3 to 130 QALY’s is a 95% range.
This only includes from those people that crossed the magic line at $1.9 / day. Non-extreme poverty also increased.
Building on your statement that many of the affected will be hard to reach with aid payments: some study on just what amount of government redistribution is actually helpful might also be in order. Redistribution may solve the immediately obvious problems of people being suddenly unemployed, but it also slows the economy’s ability to adapt to the significantly changed environment. So there’s undoubtedly a crossover point where it hurts more than it helps in the long run.
Not that I expect most governments would pay any attention at all if somebody did work up a number or a formula, but being able to see which nations come closest to hitting it could be entertaining.
Claim: The decrease in driving during the lock down has significantly increased air quality (https://www.weforum.org/agenda/2020/04/coronavirus-covid19-air-pollution-enviroment-nature-lockdown) , saving many lives. Seeing as we’re dealing with a respiratory disease, the increase in air quality has probably saved even more lives than it would otherwise.
Related claim: the decrease in driving has reduced traffic fatalities, saving many more lives.
Global yearly deaths on roads are ~1.35mil (source).
In the US (well, 23 states) there was a 6% drop in the first 6 months of the year (source—see table halfway down). Naive approximation gives 81k lives saved globally if this turns out to be the yearly average.
Alternative calculation: In the max lockdown month(s) deaths were down ~40-70% (Various European countries, Turkey, UK). Assuming 2 months of severe lockdown this would give 124k lives saved.
Lock downs have led to significant decreases in seasonal influenza in the Southern Hemisphere.
Data from Australia
Data from New Zealand
An important consideration is that the ‘thing that the US, UK and China have been doing, and what Sweden didn’t’, may not refer to anything. There are two meanings of ‘lockdowns have not been worth it’ - ‘allow the natural herd immunity to happen and carry on as normal, accepting the direct health damage while saving the economy’ or ‘we shouldn’t adopt legally mandatory measures to attempt to suppress the virus and instead adopt voluntary measures to attempt to suppress the virus’. The latter of these is the only correct way to interpret ‘thing Sweden did that the other countries didn’t’. The first of these is basically a thought-experiment, not a possible state of affairs, because people won’t carry on as usual. So it can’t be used for cost-benefit comparisons.
In terms of behaviour, there is far more similarity between what the US and Sweden ‘did’ than what the US and China ‘did’. Tyler Cowen has written several articles emphasising exactly this point. What Sweden ‘did’ was an uncoordinated, voluntary attempt at the same policy that China, Germany, the UK and the US attempted with varying levels of seriousness—social distancing to reduce the R effectively below 1, suppressing the epidemic. This thread summarizes the ‘voluntary suppression’ that countries like Sweden ended up with. Tyler Cowen writes an article attempting to ‘right the wrong question’:
So, the question cannot be “should we allow the natural herd immunity to happen and carry on as normal, accepting the direct health damage while protecting the economy”—that is not actually a possible state of affairs given human behaviour. We can ask whether a better overall outcome is achieved with legally required measures to attempt suppression, rather than an uncoordinated attempt at suppression, but since people will not carry on as normal we can’t ask ‘has the economic/knock-on cost of lockdowns been worth the lives saved’ without being very clear that the counterfactual may not be all that different.
The most important considerations have to be,
How long do we expect to have to wait for a vaccine or much more effective treatment? If not long, then any weaker suppression is ‘akin to charging the hill and taking casualties two days before the end of World War I’. If a long time, then we must recognise that in e.g. the US given that a slow grind up to herd immunity through infection will eventually occur.
How does the economic and related damage vary for voluntary vs involuntary suppression? The example of Sweden compared to its neighbours is illustrative here.
How does the total number and spread of infections vary for voluntary vs involuntary suppression? You can’t rerun history for a given country with vs without legally mandated suppression measures.
To what degree do weaker legally mandated measures earlier spare us from stronger legally mandated measures (or greater economic damage from voluntary behaviour change) later?
Edit: Tyler Cowen released another article arguing for a new consideration that I didn’t list—what reference class to place Coronavirus in - ‘external attack on the nation’ or ‘regular cause of death’. Since, for fairly clear rule-utilitarian/deontological reasons, governments should care more about defending their citizens from e.g. wars and terrorist attacks compared to random accidents that kill similar numbers of people. I also think this is a key disagreement between pro/anti-‘lockdown’ positions.
To emphasise this last point, although it falls under ‘questioning the question’, the focus on Lockdowns can be counterproductive when there are vastly more cost-effective measures that could have been attempted by countries like the UK that had very low caseloads through the summer—like funding enforcement and support for isolation and better contact tracing, mask enforcement, and keeping events outdoors. These may fall under some people’s definition of ‘lockdown’ since some of them are legally mandatory social distancing, but their costs and benefits are wildly different from stay-at-home orders. Scepticism of ‘Lockdowns’ must be defined to be more specific.
How long do we expect to have to wait for a vaccine or much more effective treatment?
I can’t think of a better source on this than the Good Judgment project’s COVID-19 recovery dashboard.
How does the economic and related damage vary for voluntary vs involuntary suppression?
This is incredibly complicated and country-specific and dependent on all sorts of factors but maybe this graph from the Financial Times is a good place to start, it tells us how things have gone so far.
How does the total number and spread of infections vary for voluntary vs involuntary suppression?
This is even harder than the previous question. ‘All we can say for sure is “It was possible to get R<1 in Sweden in the spring with less stringent measures’. If you consider that Sweden suffered considerably more death than its comparable neighbours, then you can project that the initial surge in deaths in badly-hit locked down countries like the UK could have been much higher with voluntary measures, but how much higher is difficult to assess. I think that between-country comparisons are almost useless in these situations.
This is also where accounting for coronavirus deaths and debilitations comes into play. ‘Anti-lockdown’ arguments sometimes focus on the fact that even in badly-hit countries, the excess death figures have been in the rough range of +10%, (though with around 11 years of life lost). There are ways of describing this that make it seem ‘not so bad’ or ‘not worth shutting the country down for’, by e.g. comparing it to deaths from the other leading causes of death, like heart disease. This assumes there’s a direct tradeoff where we can ‘carry on as normal’ while accepting those deaths and avoid the economic damage, but there is no such tradeoff to be made. There’s just the choice as to which way you place the additional nudges of law and public messaging on top of a trajectory you’re largely committed to by individual behaviour changes.
And if you do try to make the impossible, hypothetical ‘tradeoff economy and lives’ comparison between ‘normal behaviour no matter what’ and virus suppression, then the number of excess deaths to use for comparison isn’t the number we in fact suffered, but far higher, given the IFR of 0.5-1%, it’s on the order of +100% excess deaths (600,000 in the UK and 2 million in the US).
But again, such a comparison isn’t useful, as it’s not a policy that could be enacted or adopted, in fact it would probably require huge state coercion to force people to return to ‘normal life’.
The basic point that it wouldn’t be worth sacrificing everything to reduce excess deaths by 10% and save a million life-years is true, but that point is turned into a motte-and-bailey, where the motte is that there exists a level of damage at which a particular suppression measure (full lockdowns) is no longer worth it, and the bailey is that in all the situations we are in now most suppression measures are not worth it.
To what degree do weaker legally mandated measures earlier spare us from stronger legally mandated measures (or greater economic damage from voluntary behaviour change) later?
This raises the difficult question of how much to take into account panic over overwhelmed hospitals and rising cases. Tyler Cowen:
A spreadsheet model I made investigates the trade-off between lost life-years from covid-19 deaths versus lost life-years from reduced quality of life being locked down. You can make a copy of it and play with the various parameters.
The spreadsheet uses real data about the mortality risk from covid-19 and the population structure (life expectancy, population pyramid) for the USA.
With the parameters that I chose, lockdowns of 1.25 years destroy about 0.25 life-years per person on net.
https://docs.google.com/spreadsheets/d/1wBcHkt9i_4hGSXRrqurr82j-R_X03nyr_Eds6YLjkzc/
If there was a strictly enforced lockdown for 2-4 weeks, it seems likely that:
The pandemic would dwindle down and end soon afterwards.
Many lives would be saved.
The economy would be able to re-open soon.
There wouldn’t be any sort of large-scale panic.
A potential downside is setting a precedent for too much governmental power. I don’t feel strongly about that but I suspect that the benefit of the government having such power outweighs the cost.
I’m skeptical of this.
Wuhan needed 2 months on lockdown: https://en.wikipedia.org/wiki/COVID-19_pandemic_lockdown_in_Hubei
I’d expect that imposing China-style lockdowns in the West would require significant force and might end up causing a large-scale panic in and of itself.
I’d expect that any lockdown in the West wouldn’t have been effective enough to stamp out 100% of cases, and if you don’t eradicate it then you need ongoing measures or it will just flare up again later, so one strictly enforced lockdown wouldn’t cut it. (Though maybe you could do very rigorous contact tracing and lock down just people who might have been in contact with cases, which could be less costly than full lockdown but probably still need significant enforcement).
The other problem is that a super-strict lockdown tight enough to actually stop the virus in this manner would likely have a higher mortality rate than the virus. COVID spreads like mad, but it’s hit to average life expectancy seems to be pretty small.