The gap between estimates and official counts was not a matter of degree—it was a chasm.
As China dismantled its zero-COVID architecture in December 2022, the gap between official silence and independent reckoning grew into something harder to ignore. A London-based health analytics firm estimated that roughly 9,000 people were dying each day across the country — a figure the Chinese government's own reporting placed closer to zero. In the space between those two numbers lived a question that has shadowed every great epidemic: who gets to define what counts as a death, and who bears the cost of that definition.
- China's abrupt abandonment of zero-COVID in early December unleashed a viral wave through a largely unshielded population, with Airfinity estimating over 18 million infections and 100,000 deaths in less than a month.
- Official Chinese health authorities reported only 10 COVID deaths since the policy shift — a figure made possible by redefining COVID death to exclude anyone who died with the virus but from underlying conditions.
- Independent projections warn of a catastrophic trajectory: 3.7 million daily infections by January 13th, 25,000 daily deaths by January 23rd, and potentially 1.7 million total deaths by April's end.
- China's chief epidemiologist publicly acknowledged the data gap on December 29th, announcing plans to measure excess mortality — a tacit concession that the official count does not reflect reality.
- The world watches with limited visibility, as dismantled PCR networks and redefined metrics have severed the information channels that once made independent verification possible.
In late December 2022, as a massive COVID wave swept through China following the sudden collapse of its zero-COVID policy, a British health analytics firm offered a count that Beijing would not. Airfinity estimated roughly 9,000 daily deaths — a figure that had nearly doubled in a single week — alongside more than 18 million infections since December 1st. The firm projected the crisis would deepen further, with daily cases peaking at 3.7 million on January 13th and deaths potentially reaching 25,000 per day by late January. If the trajectory held, 1.7 million people could be dead by April.
Against that backdrop, China's official death toll since December 7th stood at exactly 10. The cumulative pandemic total, as of December 28th, was 5,246. The distance between those numbers and Airfinity's estimates was not statistical noise — it was a structural choice. Chinese health authorities had redefined what qualified as a COVID death, counting only those caused directly by respiratory failure from the virus, and excluding anyone who died with COVID alongside other conditions. Simultaneously, the national PCR testing infrastructure had been largely dismantled, removing the diagnostic foundation that case counts depend on.
Airfinity built its estimates from epidemiological models using provincial data gathered before the reporting changes took effect — an attempt to reconstruct a picture the official system had stopped painting. The firm had developed its COVID analysis platform in 2020, giving it a methodological baseline, but acknowledged the difficulty of modeling a situation where the underlying data had been deliberately narrowed.
On December 29th, China's chief epidemiologist Wu Zunyou offered a rare moment of institutional candor. He announced that the Chinese CDC would begin measuring excess mortality — comparing actual deaths against what would have been expected without the epidemic — as a way of identifying what the official count may have missed. It was not a confession, but it was close: an acknowledgment, from within the system, that the numbers being reported did not fully account for what was happening in the country's hospitals and homes.
In late December 2022, as China grappled with a sudden and massive wave of COVID-19 infections, a British health data firm released estimates that painted a starkly different picture from what the Chinese government was reporting. Airfinity, a London-based company specializing in health analytics, calculated that roughly 9,000 people were dying each day from the virus across China—a figure that had nearly doubled in just one week as cases spread through the world's most populous nation.
The surge had begun in November, but accelerated dramatically after Beijing abruptly dismantled its zero-COVID policy in early December. For nearly three years, China had maintained strict testing protocols, lockdowns, and case reporting requirements. When those measures fell away, the virus moved through the population with little resistance. By Airfinity's accounting, somewhere around 100,000 people had already died since December 1st, with more than 18 million infections recorded. The firm projected that infections would peak on January 13th at 3.7 million daily cases, and that deaths would reach a maximum of 25,000 per day by January 23rd. If those projections held, China could see 1.7 million deaths by the end of April.
Yet the official numbers told an almost incomprehensible story. Since December 7th, when the policy shift took effect, Chinese health authorities had reported exactly 10 COVID deaths. As of December 28th, the cumulative official death toll since the pandemic began in 2020 stood at just 5,246. The gap between what Airfinity was estimating and what the government was announcing was not a matter of degree—it was a chasm.
Airfinity based its estimates on epidemiological models built from data collected from Chinese provinces before the recent changes to how cases were counted and reported. The firm had constructed what it described as the world's first dedicated COVID-19 analysis and intelligence platform back in 2020, giving it a foundation for modeling disease spread. But the Chinese government had largely dismantled its national network of PCR testing centers, shifting from prevention to treatment. More significantly, health authorities had redefined what counted as a COVID death: only deaths caused directly by respiratory failure from the virus itself, excluding anyone who died with COVID but from other underlying conditions, even if they tested positive.
This definitional narrowing was not accidental. It reflected a deliberate choice about what the state would count and what it would not. The contrast between the official count and independent estimates raised fundamental questions about what was actually happening in Chinese hospitals and homes, and whether the world had any reliable way of knowing.
China's chief epidemiologist, Wu Zunyou, acknowledged the gap in a briefing on December 29th. He announced that a team from the Chinese Center for Disease Control and Prevention planned to measure deaths differently going forward—specifically, by calculating excess mortality. This approach would compare the number of deaths during the current wave against the number that would have been expected if the epidemic had never occurred. By measuring the difference, Wu suggested, China could identify what might have been undercounted. It was a tacit admission that the official figures did not capture the full scope of what was unfolding, and a signal that at least some officials recognized the need for a more honest accounting.
Notable Quotes
China's chief epidemiologist Wu Zunyou announced plans to measure excess mortality to identify what might have been undercounted in official figures.— Wu Zunyou, chief epidemiologist, Chinese Center for Disease Control and Prevention
The Hearth Conversation Another angle on the story
Why would a government deliberately undercount deaths during a health crisis?
It's not always deliberate in the sense of a conspiracy. When you change the definition of what counts as a COVID death to exclude people with comorbidities, you're making a technical choice that has political consequences. It lets you say the numbers are lower.
But Airfinity's estimate is nine thousand a day. That's not a rounding error.
No. And that's why the epidemiologist Wu Zunyou's announcement about measuring excess mortality matters. He's essentially saying: we know people are noticing the gap. We need a different framework.
Do you think the Chinese government didn't know how many people were actually dying?
They almost certainly knew. Hospitals report to provincial health departments. The question is what gets aggregated upward, what gets reported publicly, and what definition you use to decide what counts.
What happens if the excess mortality numbers come out and they're close to Airfinity's estimates?
Then the government has to reckon with a much larger death toll than it's acknowledged. That changes the narrative about how well zero-COVID actually protected people, and how the transition was managed.
Is Airfinity's model reliable?
They built it on real provincial data from before the policy changed. But modeling is modeling—it's an estimate, not a count. The real answer will come from excess mortality data, if China releases it honestly.