Daily Chart - China’s data reveal a puzzling link between covid-19 cases and political events | Graphic detail

/chinas-data-reveal-a-puzzling-link-betw

  • Daily Chart - China’s data reveal a puzzling link between covid-19 cases and political events | Graphic detail | The Economist
    https://www.economist.com/graphic-detail/2020/04/07/chinas-data-reveal-a-puzzling-link-between-covid-19-cases-and-political-e

    Erratic infection numbers raise questions about the accuracy of the country’s statistics

    EVER SINCE the new coronavirus started to spread beyond China’s borders, the country’s official tally of infections has served as a grim benchmark for the outbreaks that followed. On March 26th the count in China was surpassed by that in America, now the centre of the pandemic. Since then China’s total, now close to 83,000, has also been overtaken by those of Italy, Spain, Germany and France.

    But there is growing suspicion that China’s official statistics on the covid-19 pandemic cannot be trusted. On March 24th China’s prime minister, Li Keqiang, came close to admitting that the numbers had been miscounted when he warned officials that “there must be no concealing or under-reporting.” Classified reports to Congress from American intelligence agencies have concluded that the numbers of both cases and deaths from the disease in China are much higher than the official government figures would suggest.

    Such scepticism may be deserved. An analysis by The Economist of data reported by China’s National Health Commission reveals two peculiar features. First, the data are volatile. Across the nine Chinese provinces with serious outbreaks, we identified 15 episodes in which new cases of covid-19 jumped by more than 20% in a single day, before quickly returning to earlier levels. Although such spikes can occur in any dataset—because of erratic record-keeping, for example—we found that other countries and regions with covid-19 outbreaks, of a similar size to these provinces, have experienced fewer. Second, when spikes occur, they are often accompanied by important decisions by government officials. Of the 15 such episodes observed in the data, two-thirds appeared to occur within a day of the sacking of a provincial official or other significant political event.

    Take Hubei, the province hit hardest by covid-19. On February 9th the region reported a 27% increase in new infections. On the next two days, new cases declined by 20% and 22%, respectively. Then on February 12th they surged by a whopping 742% to almost 14,000, before immediately falling back sharply. Chinese authorities say the spike was caused by revisions to the government’s methodology for counting cases. But these changes were introduced nearly a week earlier and were reversed seven days after the spike (see chart). An alternative explanation for the surge in new cases on February 12th was another event, announced the next day: the sacking of the party chiefs of both Hubei and its capital city, Wuhan.

    Other spikes in new covid-19 cases have also coincided with changes in personnel or policy announcements. On January 27th officials in Zhejiang province held a press conference detailing the opening of 335 clinics and a 1,000-bed hospital to accommodate a surge of patients. The next day, new cases nearly tripled to 123, before declining sharply in the next few days. On February 20th authorities in Shandong province sacked the chief of the provincial justice department. That same day new covid-19 cases at a local prison jumped from two to 200, and then immediately returned to two the next day.

    Although most of these episodes occurred independently of one another, we identified one day when cases jumped in several places. On February 3rd every Chinese province with a sizeable outbreak of covid-19—at least 50 new infections per day—suffered a big increase in new cases (the mean was 35%). This was the only day during the epidemic on which this happened. One possible explanation came two weeks later, when it was revealed that on the same day President Xi Jinping, in a speech to the Standing Committee of the Politburo, had called on authorities battling the virus to “face up to existing problems” and “release authoritative information in a timely manner”.

    Do these data prove that China manipulated its covid-19 data, or that the country’s official tally of cases and deaths is lower than it should be? No. But the unusual spikes in new cases, and the curious way their timing matches political developments, are bound to raise questions about their accuracy.