China has once again explicitly recommitted to its goal of Zero Covid.

Even the WHO has said this policy is unsustainable. China don’t care.

Or rather, Xi doesn’t know and who is going to tell him?

Premiere Li might be trying, but it’s not working.

Xi, whose name keeps coming up headlines during his reign which is never a good sign, is going to see this through as long as China possibly can. How long is that?

My answer last time was almost certainly not so long. The plan wasn’t working.

Yet the market says otherwise? Here’s the six month look at the Shanghai stock market.

For comparison, here’s the S&P:

The Chinese market has been overperforming these past few weeks and is up in absolute terms since mid-April. Presumably that indicates that the market expects less disruption now.

Which makes sense, since (at least officially) cases are down quite a bit.

This is what successful containment would look like if there was successful containment. If, of course, it’s all real.

Here are deaths.

Do you notice anything weird? Deaths peaked at April 30. Cases hit a local peak at April 29 with an absolute high at April 21. After April 30, the decline in deaths is quite rapid, dropping an order of magnitude. But it has not yet been three weeks since cases in China started declining. One could say ‘maybe Chinese death reporting is better because they don’t have unnecessary delays’ but most of the delays in reporting are because it takes time to die.

So these stories don’t make sense together. That’s nothing new for China. What’s interesting is that at the peak we have 0.035/million deaths and about 20/million cases, for a (no-delay) case fatality rate of ~1.75%, with delays that could be ~3%. For comparison, the USA case rate is current 248/million and its deaths are 0.89/million, for a case fatality rate of ~0.3%, or ~0.5% if we go back 21 days for cases – of course the infection fatality rate is far lower since most infectious aren’t recorded here.

We have higher rates of effective vaccination and also higher rates of previous infection, so we should expect that China’s CFR should come in much higher depending on the ratio of how many cases (and deaths) get reported. No doubt a lot of cases were being missed or actively suppressed in China on the way up if there were this many deaths. China has no incentive to report extra deaths.

What is possible is if they were holding back Covid deaths for a while, then dumped them all into the statistics in rapid succession, then went back to either newutral or active suppression. In this story, China was hoping things would go well. When that proved impossible, they cleared the backlog to create a high peak, so they could then show a rapid decline, and then reported however many deaths they decided to report. Presumably other deaths get assigned other causes, as needed.

At the new death rate, China has a CFR similar to America’s. That could in theory be possible if China is reporting all or most of its cases while America misses many of its cases, but the decline is impossible unless cases are going down at the same rate, which they very much weren’t. These two graphs don’t look like two ways of charting the same path through time because at least one of them is lying.

Despite that, both graphs show success, and that success would be essentially impossible to fake in a world where China was steadily losing containment. Lying about deaths can work, but lying about cases only works up to a point.

So somehow, some way, China temporarily has case numbers under control.


On April 25, it looked like Beijing might be headed for lockdown, which would have potential political implications.

Residents braced for the lockdown.

Did the lockdown happen?

If it took weeks of strict lockdown to get a handle on Shanghai and other places, it seems strange that Beijing managed to handle the problem without such widespread measures.

At least in part. The most populous district of Beijing, Chaoyang, has been in lockdown for three weeks. And we have this.

Why are workers in hazmat suits constantly spraying disinfectant everywhere? The answer isn’t because it works.

Officials reassure that Beijing is not going into lockdown (Bloomberg), which comes with this warning to be worried about food supplies from municipal government spokesperson Xu Hejian:

The city’s some 22 million residents don’t need to be nervous about food supplies

What do you think when you see headlines like this?

As Bloomberg put it:

It’s becoming more and more difficult for residents in Beijing to see much difference between their situation and that of Shanghai, where millions of people have been confined to their homes for weeks on end. While most Beijingers are still technically allowed outside, there are few places to go, with schools, gyms and attractions shut, restaurant dining banned indefinitely and some districts cordoned off altogether.

Similar situations apply across China. There were some cases in most places, yet they don’t have major outbreaks.

The puzzle’s solution is that yes, there really are quite a lot of lockdowns going on.

That’s from the German consulate talking to Germans currently in China. I can’t vouch for the number but it seems like a source that wouldn’t say it for no reason.


Whereas in Shanghai, the Chinese CDC kept saying how more testing would take care of things quickly, that kept not being true and the lockdown kept dragging on for weeks. What happened?

There are basically three stories I can come up with here.

Story one is that China is flat out lying and Covid is everywhere, which must be noted for completeness. Once again, I don’t think this is the case because I do not think they would be capable of hiding it under conditions of exponential growth for this long.

Story two is that Shanghai was unusually vulnerable to infection. It had especially cramped and dense conditions, and the necessary countermeasures were much harsher. There certainly could be some effect here, and it makes sense that wherever your biggest problems arise is the place that was most vulnerable to begin with, similar to New York City for the United States. Yet mostly I say the effect sizes here don’t explain things.

Story three is that the lockdown of Shanghai was actually not a very effective means of stopping the spread of Covid. This is my current model. It is easy to equate harshness with effectiveness, but that’s falling into the morality play trap. If Shanghai’s residents are trapped inside their apartment buildings for weeks semi-forced to make broth, that could mean the treatments are working or it could mean everyone in the building is gathered round helping make (reportedly delicious) broth in ways that would give each other Covid if anyone had it.

Thus the ‘silent periods’ that ban group buys in various places keep getting extended indefinitely, out of fear that the act of distribution might be causing infections.

Which in turn means everyone is terrified of it happening to them.

This thread walks through Shanghai’s deserted streets. This CNN post is the story of a reporter escaping via Shanghai airport.

This thread of experiences from May 1 balances ‘how great it is to finally get cheese’ and the subsequent making of homemade pizza with how it feels to know people around you are getting dragged away from their homes on the regular. Where the good news is having cheese.

If anything the amount of production still happening is pleasantly surprising.

Meanwhile China mass disinfects surfaces that have almost zero impact. Although the threat to do this to anyone’s house who tests positive matters, exact incentives left as an exercise to the reader or resident. Details matter.

Things hopefully are finally starting to improve with some businesses slated to be allowed to reopen some 50-odd days into the lockdown. Headlines like ‘as the end nears’ are appearing.


China continues to not mandate vaccinations, which given what else they did mandate is kind of weird (MR)? Yet it makes sense if you assume that the point is to show your superiority and how your system is serious and also your vaccine isn’t very good. Thus Zeynep criticizes the madness because she assumes goals different from those making the decisions. The cruelty and absurdity shows seriousness, and the seriousness is the point. Chinese government are Very Serious People.

Some of Chinese middle class looks to go abroad in wake of restrictions. It would be good if we let them come here. We need to get high skill immigration through the secret congress, somehow avoid it getting held hostage by demands for other policies, but this is not the time or place for that discussion.

This is what actual zero-Covid-whatever-it-takes attitude looks like in practice.

A man convinced he had Covid taped his car shut and stayed inside for ten hours without even opening the window.

The Beijing resident had decorators round that morning who told him they tested positive, the China News Service reported.

At midday on Tuesday he got into his slick white coupe and didn’t move till 10pm, despite never getting tested for the virus.

That’s pretty crazy, but when the cost of Covid cases has been expanded by orders of magnitude due to what might be done in response, it looks a lot less crazy. I know I would much rather be quarantined by myself in a car where I can get contactless food deliveries than in one of the Chinese quarantine facilities, although there are some obvious logistical problems that are going to come up within a few days.

Synthesis and Looking Forward

So what happens now?

We take our next set of data and we keep waiting for the unstoppable force to collide with the immovable object.

The general consensus is that China’s Covid policy, economy and government have sustained major damage. Not everyone agrees – I heard on Odd Lots an opinion that China would recover and that ‘the situation could not get worse’ but if you can’t figure out how the situation could get worse you are not modeling the physical situation or political dynamics in a useful way.

The key questions are whether Covid zero is something China can pull off at some cost, what that cost would be, and what are the political implications for China.

It is clear that these lockdown conditions have disillusioned many Chinese about the supposed vast superiority of their system and its policies. There certainly are advantages. If you can make good decisions, it is better to be capable of doing harsh things. But if you can’t make good decisions, it might be much better to be incapable.

That’s a lesson we in America have learned the hard way. It’s good in theory to have the ability to do useful and potentially necessary things in a coordinated way and have state capacity, but it’s easy for state capacity that does exist to do far more harm than good. The calculation is trickier in China, where unlike in America they did successfully sustain zero Covid for two years. So if the plan works at containing Covid, will it have been Worth It?

My guess is that if you add up the measures taken and costs paid so far China is still ahead in physical terms, and probably also ahead in reputational terms although that is less clear, if this was the last they heard from Covid, versus taking a Western-style approach.

This will not be the last they hear from Covid.

That’s true even if we assume the numbers are real and China is dealing with the situation.

First, how long will it take for China to get back to its previous state? China’s per-million numbers pre-Omicron were something like 0.07. On March 1 they were at 0.22. April 21 was the peak at 21.16, the decline started at April 29 from 20.34. On March 14, 5.45, a 73% decline in 17 days, or a 7.5% decline each day. So assuming that the period from April 29 to now is how well you do under current conditions of hundreds of millions of people in various degrees of lockdown, to get back to a number like 0.1 would take another several months. More likely they will content themselves with numbers somewhere in between.

But there is no ‘content with’ level when it comes to exponential growth. The moment China returns to something resembling normal, I don’t see how Omicron does not emerge once again. I also can’t see China sustaining, essentially forever, what it takes to keep Omicron from rising again – that policy seems outrageously expensive.

It seems like the ‘baseline scenario’ from here is that official cases continue to decline until victory is declared (in some central sense) and after that things start up and we do the whole thing over again.

So in some important sense congratulations, China. You managed to turn around this wave when it seemed no sane response could do so. In another important sense, my condolences to China, because we’re almost certainly going to have to do this again.

The hope would be that over time, China’s techniques will improve in efficiency. They’ll focus on what works and there will be less cruelty and waste, and people will adjust. Yet the disinfection routines continue after two years, and I see no signs that they recognize that their design in Shanghai was deeply flawed.

The fear is that the opposite is the case. There was a reservoir of goodwill and willingness to endure sacrifices, and much of that is now depleted. There was a faith that such events would not happen or would be short and thus one should not hoard food, panic, flee or rise up, and that faith is eroded. There was economic slack in the system, where production could pause and food and other products could be wasted without the consequences being too dire, and that budget is being tested. And so on.

What happened in America was a mix of both effects. We exhausted our willingness to endure stuff, and we also got smarter. Nothing like what a competent civilization would have figured out by now, but we met that civilization halfway. Then the exhaustion got rid of most of the bigger dumb stuff. Which, given what we have to work with, together feels pretty damn good. We muddled through.

China seemed to start off from a better place in many ways, and more able to bring in ‘reinforcements’ as it were to make things stick, but be less adaptable, less able to muddle through and change course. Our data is far from perfect and our discourse far from optimally free, but China’s is far more locked down. We care how it looks, but not like China/Xi cares. Thus, China purses a high-risk, high-reward, all-or-nothing strategy that works until it doesn’t.

Fool me once, though. We keep saying that China can’t possibly get out of this one, then China does, and we say huh, well played. And again, if you pursue a More Dakka martingale strategy, it’s hard to tell if you’re #Winning, about to get mocked on Twitter by Taleb or both. Often it’s both.

We’ll keep watching carefully. If you don’t now give China a substantial chance of ‘getting away with it’ given what they’ve accomplished (where that is defined as not hitting 50 cases per million per day in 2022), which I think I have at something like 30% at this point, either you know something I don’t yet know about the facts on the ground or I’m not sure what’s underpinning that confidence. If you expect China to ‘get away with it,’ I flat out disagree, again unless you have facts not in evidence.

What would be the consequences if China has to go through a long period of large scale lockdowns? It’s very hard to tell, but it feels like the market is underestimating the impact here, in addition to the tail risks. Those tail risks, of things fully breaking down, also still seem real to me.

I’ll check back again in a few weeks.


New Comment
6 comments, sorted by Click to highlight new comments since: Today at 3:33 PM

Cases per million over a country does not automatically follow exponential growth. An overall average figure ignores the likelihood that some regions will be virus-free, and that people in those areas can largely normal lives until the next outbreak comes in from the outside.

I do think China's zero COVID goal is at least sane, even though the cost is high and some of their policy details are pretty terrible from the points of view of both effectiveness and tolerability. I am very glad that elimination is still being attempted by at least one country in the world. We may need to learn from their measures if (or more realistically when) some strain with substantially greater lethality starts spreading.

What’s interesting is that at the peak we have 0.035/million deaths and about 20/million cases, for a (no-delay) case fatality rate of ~1.75%

I think you mean ~0.175%, so deaths are suspiciously low even at the peak.

Also 'March 14' should read 'May 14'.

Two additional conspiracy-ish theories about why China is so persistent with lockdowns:

  1. They know something about long-term effects of Covid we don't (yet) - this seems to be at least partially supported by some of the research results coming out recently
  2. Slowing down exports (both shipping and production) to add momentum to the US inflation problem while simultaneously consuming less energy/metals to keep prices from increasing faster so China can come out of the incoming global economic storm with less damage

A theoretical explanation for the decrease in cases at the same time as the decrease in deaths would be that perhaps most of the early cases were in the elderly/immuno-compromised, while later cases were in young healthy people. I'm not sure there's a good reason to believe that this happened in Shanghai though, when it hasn't had any affect on the cases vs deaths curves elsewhere.

In theory, a place could have a situation where the only testing was happening to confirm already symptomatic patients, so small numbers, but high fatality rate, as all the mild cases gonindetected. Then it moves, over a few weeks, to a mass testing regime, resulting in a lot more tests performed, finding a lot more cases, but the actual peak in cases/population was a few weeks ago.

New to LessWrong?