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.

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.

 

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':

"The most compassionate approach that balances the risks and benefits of reaching herd immunity, is to allow those who are at minimal risk of death to live their lives normally to build up immunity to the virus through natural infection, while better protecting those who are at highest risk. We call this Focused Protection."

What exactly does the word “allow” mean in this context? Again the passivity is evident, as if humans should just line up in the proper order of virus exposure and submit to nature’s will. How about instead we channel our inner Ayn Rand and stress the role of human agency? Something like: “Herd immunity will come from a combination of exposure to the virus through natural infection and the widespread use of vaccines. Here are some ways to maximize the role of vaccines in that process.”

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.

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.