“Don’t Panic, Don’t Spread Rumours” reads a People's Daily propaganda poster.

Amid the ongoing efforts to tackle the coronavirus, this is one of the key messages being relayed to Chinese citizens through government propaganda. Yet the outbreaks of panic and misinformation that Beijing fears are partly down to its own practices. For, within a system lacking free expression and unrestricted information flows, people inevitably resort to alternative and often unverified sources of data.

In recent years, Xi Jinping’s administration has exacerbated this systemic problem by tightening censorship and shrinking an already limited space for information exchange. The result is that any potentially problematic statements are dismissed as “rumours” (just as certain Western political actors now refer to any facts they don’t like as “fake news”). Indeed, it was precisely this charge that was laid upon Dr. Li Wenliang – one of the whistle-blowers of the outbreak in Wuhan, who tragically died from the virus himself.

Where the Chinese government has often been quick to clamp down on perceived rumours, it has been slow to counter those statements through robust facts of its own. Even during this recent crisis, the information released by official organs like the Centre for Disease Control has been scarce and inaccessible. Yet such reluctance by official organs to take initiative should be unsurprising given the harsh climate that has emerged in China’s increasingly centralised political structure. (The sacked mayor of Wuhan, Zhou Xianyang, even alluded to this issue in explaining his own failure to act.)

Amid such bureaucratic inaction, it has been non-governmental actors who have taken the initiative to inform the Chinese population on health issues such as the COVID-19 coronavirus. Perhaps the most notable of these is an online health advice platform called Doctor Dingxiang (丁香医生). Started in 2014 by former physician Chu Yang, it is essentially a fact-checking service for everyday medical issues. (Online misinformation relating to health problems has long been an issue on the Chinese internet.)

Since the start of this epidemic, Doctor Dingxiang has published over a hundred articles clarifying rumours circulated online regarding prevention and treatment of the coronavirus. But by far its most valuable asset has been an online database reporting comprehensive case numbers and fatalities, presented through dynamic maps and easy-to-grasp visual charts. For the past month or more, this has become the main way that most ordinary Chinese people keep themselves informed about the outbreak.

Screenshot of an interactive map of coronavirus cases across China, provided by Doctor Dingxiang.

To date, the database has been viewed almost two and half billion times which, I have no doubt, is greater than the views accumulated by any official online data provider. Doctor Dingxiang has thus emerged as one of the most respected sources of information regarding the coronavirus, a lifeline both for people across China and indirectly (by informing global media reports) the entire world.

Yet, despite this success, resources like Doctor Dingxiang have had their work complicated by erratic official statistics and reporting methodologies. Arousing particular confusion has been the government’s approach to counting confirmed infections, revised twice in only a week.

Firstly, on February 13, the authorities decided to include cases confirmed via CT scans (rather than nucleic acid test kits), causing a sudden 45% spike in the overall count. Then, just seven days later, the methodology reverted to only include test-confirmed cases. To make matters more perplexing, the Hubei authorities also announced a ban on subtracting already-confirmed cases, ordering any such cases to be re-added to the total.

A graph of new coronavirus cases shows a sudden spike following a change in official counting methods (Doctor Dingxiang).

Beijing has stated its changing testing capacity as the reason for this latest revision. But even if that explanation is to be believed, to have so many revisions in such a short space of time is strange. It has sparked mistrust in official data domestically and revived existing concerns about Chinese statistics internationally. Further fuelling doubts is the fact that the changes in methodology have occurred at the same time as high-level personnel changes in Hubei. Many will assume that this is no coincidence, and that there is a political motive – not a medical or statistical reason – for the data reporting changes.

For medical professionals, businesses and anyone trying to get an accurate picture of the current outbreak situation in China, these shifting official stats are puzzling and worrying. But, fortunately, there are steps we can take to try and understand what is going on, beyond the headline numbers:

1. Realise this is nothing new

Firstly, we must acknowledge that these doubts around the coronavirus case statistics are not a one-off issue, but part of a broader, pre-existing phenomenon of dubious data in China. While there is little concrete evidence of deliberate data manipulation by statistical authorities – and little chance of acquiring such evidence in China’s closed-off environment for public information – there are two reasons to always treat Beijing’s official numbers with caution.

One is a matter of logic – in autocratic systems like China’s (and perhaps even in liberal democracies), official statistics are not just numbers but political statements. This makes sensitive data such as coronavirus cases a likely target for manipulation at both national and sub-national levels by actors motivated more by political survival than statistical realities. Indeed, academics like Jeremy Wallace have found strong signs of GDP data being doctored in China at politically sensitive times.

Another reason for caution is that even senior leaders in Beijing have themselves been known to express doubts over China’s official statistics. Most notably, Premier Li Keqiang was quoted in a diplomatic cable revealed by Wikileaks as saying that GDP figures are “man-made” and thus unreliable. Instead, he has looked at alternative data sources to understand the true picture of the domestic economy, including electricity consumption, rail cargo volumes, and loans paid out.

2. Triangulate data where possible

Secondly, we can learn from this strategy of triangulation adopted by Premier Li as well as scholars like Kyle Jaros in his recent book, China’s Urban Champions (Princeton, 2019). As Jaros writes, “In a context where statistical, textual, and interview data all must be treated carefully, it becomes even more important to use diverse sources and mixed methods. Instead of assuming that any one source or method is fully dependable, I gather information from a range of textual, interview and statistical materials and use different types of evidence to develop and test [my] claims.”

In the case of the present coronavirus outbreak, this triangulation could mean merging official data with unofficial numbers from news reports, social media and platforms like Doctor Dingxiang. One valuable metric to look at right now is the number of regional authorities lowering their alert status for the outbreak (now at six, including the relatively badly-affected Guangdong Province. We could also consider associated data like railway passenger numbers, flight traffic and trade cargoes, although these numbers may not change significantly until after the outbreak has been officially declared over.

3. Seek signals not accurate numbers

Thirdly, in the absence of many reliable statistical alternatives to the official coronavirus figures, we can look to non-statistical sources of information. In other words, we can seek signals or actions from key actors, rather than accurate numbers or statements. As China media aggregator Bill Bishop recently said in his Sinocism newsletter, “Look at what they do, not at what they say.”

Possible actions to look out for include a lifting of community lockdowns, an end to face mask enforcement, a revised date for the postponed parliamentary sessions, and the all-clear for schools and universities to resume classes. Eventually, when the battle has been deemed well and truly won, this will likely be signalled through an official visit to Wuhan by paramount leader Xi Jinping. Thus, while the official numbers from China currently appear promising, concrete insights will ultimately come not just through statistical analysis but also by observing carefully-choreographed propaganda activities.

Premier Li Keqiang in Wuhan on January 27 a visit notable not so much for the substance of his statements, but as a signal of concern and intent by the top leadership (CGTN).

Final Thoughts

It is now a month since Wuhan went under lockdown on the eve of a rather sombre Spring Festival holiday here in China. My observations online and off suggest that more and more people are starting to head out of their homes and back into public spaces. This may suggest that restrictions on movement have been relaxed in places, and that people have confidence in the increasingly positive news flows. It may also be that people are now normalised to the outbreak and simply eager to get outside after weeks of being mostly home bound.

In any case, the concern now for China is that people and organisations grow complacent and abandon new habits of personal hygiene and prevention, delaying total containment. At the time of writing, that end goal is still far from certain, as numbers continue to fluctuate, and new hotspots are reported including inside prisons. It is also critical for China that lessons being learned during this outbreak are taken forward, whether to fight a near-term return of the same pathogen, or a future novel virus.

Yet, surely of greater significance now is the spread of the COVID-19 infection outside of China. Recent days have seen a worrying surge of cases from South Korea to Iran and even Italy. And as these countries’ governments try to respond, their task has likely been made more difficult by a perceived lack of reliable statistical data on cases and trends from within China.

It is also likely that international medical practitioners currently lack sufficient technical information about the viral outbreak on the front lines, partly because of China’s closed-off environment for information exchange, and partly owing to language barriers. In order for other parts of the world to really understand the coronavirus, it will probably (unfortunately) be necessary for the situation to deteriorate there too so that local personnel can observe and understand the nature of the viral outbreak for themselves.

The good news for many Western nations is that health systems have been preparing for precisely that eventuality. In most cases, they have been doing so with more advanced existing public health infrastructure and expertise than was present in Hubei. The challenge now for the WHO and national health authorities is to leverage those advantages in stabilising the public mood, and sensitising people on basic hygiene, prevention and treatment. It is imperative that such information reaches the public above the noise of the press and social media, so as to avoid incidents like the recent attack on a bus of evacuees in Ukraine.

The toughest task of all will be in less developed countries, particularly within Africa. While volunteering in Sierra Leone during the early days of the Ebola outbreak, I saw first-hand how dangerous rumours can be during the initial spread of a new disease. In places where there is limited electricity, let alone access to reliable information online, the dangers of distorted facts are even greater than in China. Key to overcoming such challenges will be the continued sharing of high-quality data and insights throughout the global community.