stay home
Did the original lockdown of January 2020 work in China? It was the original example of how extreme restrictions could be effective in 'controlling' the new virus, praised and commended by the WHO in its visit the following month. Even if there was uncertainty at the time about whether Western nations could or should emulate the authoritarian state, there was no doubt among all the main players that China's intervention had been successful and brought the virus under control. It was this apparent success that inspired governments around the world to take the plunge and follow China's lead, beginning with Italy in February and March, and it likely lay behind the constant sense over the following years that restrictions should be working, even when they plainly were not.

So did China's original extreme lockdown work? My colleague Dr. Noah Carl is among those who think it did. Noah accepts the well-attested evidence that lockdowns in the rest of the world largely didn't work. But he argues that it did in Wuhan. Initially he implies the difference was due to the strictness of China's lockdown ("To begin with, China's lockdown was far stricter"), but his main position, stated in his conclusion, is that lockdowns with border controls when imposed early enough can stop the virus in its tracks. Thus: "When prevalence was relatively low, countries had a shot at containing Covid, so long as there were strict border controls in place."

To defend this point of view he points to Australia and New Zealand, which seem to have kept the virus at bay until 2022 by following this method.

The main problem to my mind with the claim that China's lockdown worked by stopping the virus in its tracks is that none of the rest of East Asia had a large wave in 2020, despite nowhere else imposing a lockdown like in Wuhan, and some places like Japan not imposing one at all. The rest of China did not have a lockdown like the one imposed in Wuhan, or initially any restrictions at all - that was supposedly part of the success, that the extreme Wuhan measures 'contained' the virus and protected the rest of the country, hence Italy initially imposing lockdown in a local area and then just in the north of the country.

But we now know that the idea the virus was contained in Wuhan is nonsense. The virus was not contained in Wuhan, it went all round the world and also round the country as five million people are reported to have left the city ahead of the quarantine. Whatever prevented a large outbreak occurring at that time in China and its neighbouring countries, it was not the measures imposed in Wuhan.

In the first wave it was frequently stated that countries had successfully controlled the virus whenever 'cases' went down or they had only a small outbreak. This happened with Wuhan's lockdown, and also with South Korea's contact tracing, which was credited with keeping the virus at bay, setting the stage for the later global obsession with test and trace, while Germany and Eastern Europe's mild first wave was attributed to their quick, sharp lockdowns. We were all supposed to learn lessons from these countries: from China, that the harder the better; from Eastern Europe that the quicker the better; and from South Korea, that slick contact tracing was a viable alternative if you could pull it off. Of course, all these lessons were demolished as 2020 went on, as new waves of Covid kept coming irrespective of how fast, hard and often lockdown was imposed and how intensively contacts were traced. The fact that Japan didn't do any of these things but still had low death rates in 2020 was largely ignored or put down to a peculiar Japanese X-factor. But Japan wasn't exceptional in the region: none of the countries in East Asia had large outbreaks in the rest of 2020.

excess mortality uk asia

Putting this down to whatever each country did, when Japan did none of them while Western countries found none of them worked, makes no sense, and was usually based at some level on the silly idea that East Asians are just better at this kind of thing than Westerners. The more logical conclusion is that East Asia just wasn't very susceptible to the initial strains, for whatever reason (save for the isolated outbreaks in Wuhan in January and Daegu, South Korea, in February). After all, once Delta and especially Omicron came along, these countries saw substantial waves, though their responses hadn't changed.

As has often been noted, the first wave between January and April 2020 was an oddly patchy affair. It snowballed into large, deadly outbreaks in only a selection of locations: Wuhan, Daegu, Lombardy, Iran, north-eastern U.S., parts of Western Europe, and so on, but much of the world was spared, including areas like Eastern Europe that would later be hit hard. Numerous studies have shown that this pattern had no discernible relationship with the measures imposed. To take one example, New York implemented a strong lockdown but experienced an extraordinary number of deaths (many of which are suspected to be related to faulty treatment protocols); South Dakota, on the other hand, imposed almost no restrictions and saw no excess deaths that spring.

Some scientists have suggested this early patchiness may be due to the virus adapting to its human hosts - that it is a variant-based phenomenon, in other words. If so, this would explain why very strict border controls such as those used in New Zealand and Australia can be successful, for a time, in keeping the virus at bay, as they keep out the new variants that produce the new waves.

The question we're looking at here, though, is not whether strict border controls can keep out the disease - I think the evidence suggests they can, for a time. It's whether lockdown and other internal restrictions can "stop the virus in its tracks" and bring an ongoing outbreak under control. Noah suggests that lockdown can work, provided it is accompanied by strict border controls. But given the evidence that lockdowns don't work, the question is why we should suppose that in this combination of lockdowns and strict border controls, the lockdowns are pulling any weight, rather than just being there as an unnecessary extra.

Lockdowns were all the rage in the autumn and winter of 2020-21, but the winter waves seen across the world at that time were as strong as any flu waves in recent decades, with no sign of lockdowns preventing them. Consider London, where new infections were clearly rising at the end of November despite the city being in lockdown until December 2nd and then under strict 'Tier 2' restrictions.

ONS infection survey London
ONS infection survey for London in 2020
On the other hand, the few places which did not impose such strong restrictions, such as Florida, had no worse outcomes than those which did, such as California, which imposed a stay-at-home order between November 21st and December 21st, the impact of which in the chart below is not discernible.

new deaths covid-19 2021
Of course, if lockdowns, like strict border controls, did work it could only ever be a short-term solution, one that would never be justified on any sensible cost-benefit analysis. That is the argument against border controls - even if Australia and New Zealand can shut their borders tight and keep new variants (and thus new waves) out for a period of time, that doesn't mean it's worth it, not least because the virus will still be there when you reopen, plus the gain, such as it is, is simply not worth the pain.

The argument against lockdowns, on the other hand, is much simpler, because they do not even succeed on their own terms, as disease control. Many people struggle to accept this conclusion because it runs contrary to the 'commonsense' idea that keeping people apart will reduce spread. There are many reasons that lockdowns do not do this, including that large numbers of people do still go to work and share physical space, not least in hospitals, care homes, doctor's surgeries, offices and shops, as well as in their own homes. But whatever the reason they fail, the evidence is clear that they don't make a significant impact on outcomes - and China gives us no reason to think they do.