Did the lockdown work in Wuhan?

The Wuhan lockdown may not have worked; so why are so many people convinced that it did?

by George J. Dance

When the Chinese communist government locked down the city of Wuhan on January 23, effectively putting most of the population under house arrest, it was rightly called "a radical experiment in authoritarian politics."(1) Public health experts, as The Guardian later noted, were "sceptical": "Beijing’s decision was a vast experiment, epidemiologists warned, that might not work despite its huge human and economic cost. Quarantine had never been tried on such an enormous scale in the modern world."(2) 

That scepticism persisted as late as March 12. Reporting that day on Italy's national lockdown, The Economist quoted just one health care expert – "Gabriel Leung, an epidemiologist from Hong Kong University who was part of a World Health Organisation team that examined China’s efforts to contain its epidemic" – who told the journal that "nobody knows yet what combination of controls works best against covid-19. 'Do you need to do everything that the Chinese have done to control it?' he says, or is it enough to copy only certain elements. 'That', says Mr Leung, 'is really the big question.'"(3)  

On March 12, the World Health Organization (WHO) declared COVID-19 a pandemic, and google daily searches for "coronavirus" spiked to over 20 million. Many of those millions found, read, and shared a March 10 article on Medium by San Francisco business writer Tomas Pueyo, "Coronavirus: Why You Must Act Now" (working title, "Coronavirus: Act Today or People Will Die"). "On the face of it," sociologist Warren Pearce noted, Pueyo's article "appeared an unpromising addition to attention-sapping digital platforms: described as a 27-minute read, it ran to over 6000 words and contained 23 charts. What’s more Pueyo made no claims to special expertise or relevant credentials."(4) However, to a world scrambling for answers, Pueyo offered a message of complete certainty. Pearce notes that "the article received a stunning 40 million views in the first nine days since publication." 

Pueyo's message was grim: The coronavirus was coming to get us. It would "spread at an exponential speed,"(5) doubling every few days until most of us were infected. All of us were in danger: "If you get it you'll die" (as he later declared to a television audience). The "health care system will be overwhelmed." Countries that did not eliminate the virus would witness "a fatality rate between ~3%-5%" – at least a million dead Canadians, 2 million dead Britons, 10 million dead Americans. 

However, Pueyo also had a message of hope. Our leaders could save us from all of this, but only if they did as Pueyo told them, when he told them to: ""As a politician, community leader or business leader," he told his online readers, "you have the power and the responsibility to prevent this." "The only way to prevent this is social distancing today. Not tomorrow. Today. That means keeping as many people home as possible, starting now." "These measures require closing companies, shops, mass transit, schools, enforcing lockdowns…"  

Although Pueyo was calling for an almost inconceivable disruption of society, he confidently predicted that it would come to pass "in 2-4 weeks." But how, when scientists admittedly did not know what measures worked best, could he have such certainty? Like many others, he had followed the pandemic online, crunched the numbers, and reached his own conclusions. For him, the data from Wuhan proved unquestionably that Wuhan-style lockdowns worked; he presented that proof, dramatically, in what he called "one of the most important charts" in his article, Chart 7 or the "Wuhan graph": 

This chart, prepared by the Johns Hopkins Institute based on data from the Chinese Centers for Disease Control (and annotated in red by Pueyo), graphs the trajectory of the Wuhan pandemic from two different perspectives. The orange bars on the right are the diagnosed cases, sorted by date of diagnosis. (Pueyo calls those the "official cases".) The grey bars to their left are the same cases, but this time sorted by the date each patient first developed symptoms: the symptomatic cases (Pueyo calls those the "true cases"). "The Chinese CDC found these by asking patients during the diagnostic when their symptoms started." Pueyo explains the significance of the latter set:

Up until Jan 23rd, when Wuhan closes, you can look at the grey graph: it’s growing exponentially. True cases were exploding. As soon as Wuhan shuts down, cases slow down. On Jan 24th, when another 15 cities shut down, the number of true cases (again, grey) grinds to a halt. Two days later, the maximum number of true cases was reached, and it has gone down ever since. 

Indeed, that is exactly what the chart shows: the growth in new symptomatic cases began slowing down as soon as Wuhan was locked down, and by the next day had begun to decline (ultimately to reach zero by March 19). For Pueyo the moral was clear: "If you want to be safe, do it Wuhan-style." 

Of the 40 million readers of Pueyo's article over the next week, untold millions became true believers, sharing it and proselytizing for it online. "I’ve had an overwhelming amount of positive feedback," he told his alma mater in April. "I’ve lost count of the number of friends who have said, 'I received this from four different people.' I pushed it out through my networks and they received feedback from several different governments who acted the day they received the articles."(6) By March 19, the day China declared no new cases, the whole world (or at least the online portion of it, except a few hopeless dullards) had become convinced that lockdowns worked. 

The sheer amount of Pueyo's online converts and their proselytizing, plus that of their online converts in turn (not to mention a simultaneous social media campaign originating in China), quickly translated into solid majorities in almost every democratic country demanding that their governments lock them down for their own safety. Now governments that held back were the ones being accused of performing a dangerous experiment. By the end of March, over 100 countries, and billions of people, were under full or partial lockdown. In the United States, 42 states, with more than 90% of the population, were under stay-at-home order orders.  Self-fulfilling prophecy or not, Pueyo's prediction had come true: The world was doing it Wuhan-style.

Eventually the number of new COVID cases did stop growing and begin to fall, everywhere, which was seen by Pueyo's true believers as confirmation of their truth. In no cases, though, did a state that locked down experience the Wuhan-style outcome of an instant drop in new cases, nor their fall to zero in two months. Some researchers were finding no correlation between lockdown dates and the eventual fall in cases, leading to a renewed debate on lockdowns' effectiveness.  

For instance, the next chart, which Washington Post columnist Philip Bump presented in a May column as "obvious evidence" that "lockdowns saved lives,"(7) shows the time lag between seven states' lockdown orders and the first drop in new cases ranging from one week in Pennsylvania to six in New Jersey.   

Part of that time lag can be explained by the fact that these counts are of "official" or diagnosed cases, and the onset of symptoms would have happened earlier; as in Pueyo's Chart 7, where diagnosed cases (the orange bars) continued to increase for days after the lockdown. But Bump points out that the time from infection to becoming an official case would necessarily be even longer than the tune from becoming symptomatic:

 

Remember that the effects of a stay-at-home order on new cases and on deaths isn’t immediate. The virus can take two weeks before symptoms manifest.... We’ve highlighted the two-week window after the implementation of the stay-at-home order in each place and, because this is a seven-day average of new cases, the third week as well, since its average includes part of that two-week window.

Bump argues that "in New York alone, there’s clearly a correlation between the order and the decline in cases," based on the decline there happening almost precisely three weeks after the order.

The WHO estimates an "incubation period [of] from 1-14 days with median estimates of 5-6 days between infection and the onset of clinical symptoms" for COVID-19.(8) Bump seems to be giving only the maximum time in order to exaggerate its length. In addition, his use of a 7-day average (when a 3-day average would have worked just as well) seems designed only to add a third week so that New York would fit. Still, his point that there is a delay due to the incubation period is correct. Since it takes time for infections to turn into cases, it will take the same time for a fall in new infections to manifest as a fall in new cases.  

Given that, though, how can one account for Pueyo's Chart 7, which shows an immediate slowdown in symptomatic cases beginning the very first day of the lockdown? 

"If you go back to the Wuhan graph," Pueyo reminds us later in his paper, "you will remember that as soon as there was a lockdown, cases went down. That’s because people didn’t interact with each other, and the virus didn’t spread." However, given the incubation period, the symptomatic cases on January 24  were caused by "spreading" – infection – that had already occurred before January 22, mostly before January 17, and some as early as January 9. A fall in symptomatic cases on January 24, then, can only have been due to a corresponding fall in infections by January 22. Whatever might have made infections fall by January 22, it could not have been a lockdown that had yet to occur.  

I can see just two possible explanations. The first is that something else happened, before January 22, to slow the spread and reduce infections. That something else is unlikely to have been voluntary social distancing, given the Chinese government's suppression of information on the disease pre-lockdown; and in fact, the only study to examine the question concluded that "voluntary self-isolation driven by individuals' perceived risk of becoming infected kick[ed] in only towards the peak of the epidemic and ha[d] little or no impact on flattening the epidemic curve."(9) A better candidate would be the massive depopulation that occurred at the time. Wuhan's mayor estimates that 5 million residents had left the city by January 23.(10) The sudden disappearance of that many people – almost half the population of 11 million – is bound to have led to a marked decrease in social interaction and therefore in virus spread, though just how much of a decrease needs further study. 

The second possible explanation is that the data used to construct Pueyo's Chart 7 is bogus. That possibility cannot be simply dismissed: China's communist regime was accused of concealing information during the SARS coronavirus outbreak in 2003, and some researchers are convinced it is doing the same thing this time. At the same time, it seems reasonable to accept the official data barring any actual proof of falsification, since the alternative would mean having no data.  

Even accepted at face value, though, the data presented on the Wuhan graph shows only that the lockdown there was followed by an almost immediate drop in cases. It does not show, much less prove, that the lockdown caused that drop. Rather, given the biological facts, it shows that the lockdown could not have done so.

Sources (accessed October 9, 2020):

1. James Hamblin, ''A Historic Quarantine'', The Atlantic, January 24, 2020. 
2. Emma Graham-Harrison and Lily Kuo, "China's coronavirus lockdown strategy: brutal but effective," The Guardian, March 19, 2020. 
3. "Should other countries copy Italy’s nationwide lockdown?", The Economist, International Edition, March 12, 2020. 
4. Warren Pearce, "What does Covid-19 mean for expertise? The case of Tomas Pueyo," International Science Council. 
5. Italicized quotations are from: Tomas Pueyo, "Coronavirus: Why You Must Act Now," Medium, March 10, 2020. 
6. "Tomás Pueyo, MBA ’10 on His Viral Post, 'Coronavirus: Why You Must Act Now'," Stanford Graduate School of Business, April 1, 2020.
7. Philip Bump, "Tucker Carlson claims there’s ‘no evidence’ stay-at-home orders saved lives. He’s wrong.", Washington Post, May 22, 2020. 
8. Centers for Disease Control and Prevention, "Clinical Questions about COVID-19: Questions and Answers," Coronavirus Disease 2019 (COVID-19), August 4, 2020. Web, Sep. 13, 2020. 
9. Alexander Chudik, M. Hashem Pesaran and Alessandro Rebucci, "Voluntary and Mandatory Social Distancing: Evidence on COVID-19 Exposure Rates from Chinese Provinces and Selected Countries," Federal Reserve Bank of Dallas, April 2020. 
10. Erica Kinetz, "Where did they go? Millions fled Wuhan, China, before coronavirus lockdown," Associated Press, February 9, 2020.