Chinese authorities have placed an estimated 760 million people into lockdown as part of an epic campaign to contain the spread of covid-19, the disease caused by the novel coronavirus. As of Sunday, there were over 77,000 confirmed cases and more than 2,500 deaths in China, mostly in Hubei province. Wuhan, the provincial capital and the epicenter of the outbreak, has been hard hit.

Why did China’s CDC system, once touted as among the world’s best disease control programs, fail to help contain the virus early on? And what has the crisis exposed about China’s system of governance? Here’s what you need to know.

China built a system to prevent another SARS crisis

In the aftermath of the 2003 SARS crisis, China invested heavily to improve its system for infectious disease control and prevention. These measures included new laboratories and a nationwide Infectious Diseases Reporting System, as well as new laws on infectious diseases control and public health emergencies. The reporting system is extensive, covering all of China’s more than 2,800 county-level jurisdictions.

This sentinel system for infectious diseases helped China tackle various outbreaks — including H1N1avian flu and malaria. Successive China CDC directors have taken great pride in this system. In a March 2019 interview, Gao Fu, the China CDC director general, said he was “very confident that the SARS incident will not recur. This is due to our country’s well-built infectious disease surveillance network; we can block the virus when it appears.”

The system worked, according to local authorities

The Wuhan Health Commission (WHC) began to release information on its website on the atypical pneumonia cases on Dec. 31, 2019.

But local authorities didn’t tell the full story

The China CDC official line, however, suggests a different timeline. According to Feng Zijian, deputy director general of the China CDC, the direct reporting system was “not activated that expeditiously.” In fact, the award to Zhang for reporting on Dec. 29 reminds us that the pre-Dec. 29 cases were not reported, let alone filed into the disease reporting system in real time.

Two separate sources reveal that Gao himself was the real sentinel of the coronavirus outbreak. In the evening of Dec. 30, Gao Fu noticed from scanning group-chats that the WHC had just issued two internal notices on atypical pneumonia cases. Alarmed that such information had not been submitted to the national reporting system, he called the Wuhan CDC head and learned that the number of cases was well above the threshold for reporting. Troubled by what he heard — and didn’t hear — Gao immediately alerted the National Health Commission (NHC) leadership. The following day, Dec. 31, the NHC dispatched a national team of experts to Wuhan to investigate.

Local authorities also silenced whistleblowers

As the national team was on its way to Wuhan, the WHC issued its first public statement about the atypical pneumonia outbreak, reassuring the public that the health administrations and hospitals were managing the situation well. Of the 27 cases, “seven were critical, but the rest were stable and controllable, including two that … are expected to be discharged.” In fact, the latest retrospective study by China CDC reveals there were already 104 cases, including 15 deaths, in December.

In contrast, on Jan. 2, health authorities in Singapore and other countries began to screen passengers from Wuhan.

The case of Li Wenliang has captured global attention. Early on Dec. 31, the Chinese doctor was reprimanded by WHC and the Wuhan Central Hospital he worked at for spreading false rumors about SARS-like cases on Dec. 30. Police later forced him to sign a document promising not to spread “false rumors” again. Other doctors were also admonished for “irresponsible behavior that caused social panic and disrupted Wuhan’s development and stability.”

Systemic fissures contributed to further delays

Local officials, including Wuhan’s mayor, blamed their inadequate public disclosure on the need to secure approval from above. But the truth is more complicated. We now know that on Dec. 30, a joint Hubei-Wuhan CDC investigation team concluded that there were no clusters of cases but there were nonetheless a family of several members that became infected.

Had China CDC experts seen this report — or engaged with the infectious disease doctors at the major Wuhan hospitals — they would likely have recognized earlier that the virus was spreading from human to human. Three crucial weeks would elapse before a new national experts team, including Zhong Nanshan and Gao, finally concluded that the coronavirus was highly contagious.

The infectious diseases sentinel system only works if the hospitals and local health administrations actively engage with it and contribute to the information. In Wuhan, the system failed, monumentally. The failure has laid bare the inherent tensions of a reporting system that is also beholden to the political imperatives of provincial and municipal Communist Party bosses.

For now, President Xi Jinping has replaced the top leaders of Hubei and Wuhan. China remains in the midst of an unprecedented and enormously costly effort to contain covid-19. While the Chinese leadership can lay some of the blame for the crisis on local missteps, a more effective public health emergency response system will depend on encouraging information flows and realigning institutional interests.

Dali L. Yang is the William C. Reavis Professor of Political Science at the University of Chicago. His research has emphasized governance and regulation in China.

Read more:

Updated COVID-19 (Coronavirus) statistics

Data update dates: World Health Organisation: 25 Feb | Hubei: 25 Feb | China: 25 Feb

COVID-19 (Coronavirus) is the number 1 issue facing investors at the moment. Given issues with data from China, we have put together these charts (updating throughout the day) to highlight the data from outside of China. Often the final data point will only include countries which have reported that day and so will change throughout the day. 

NOTE: China has re-classified statistics at least three times. There are also numerous revisions to prior numbers. We have made some adjustments to the charts below to normalise these statistics where possible, but treat China and Hubei data with scepticism. We now use both suspected and confirmed cases in Chinese ratios. 

COVID-19 cases caught outside of China

Whilst at first most cases of COVID-19 outside China were people who had flown from China to another country, we now seeing transmission of the virus outside of China taking off:

Source of new Covid-19 cases

Number of new Covid-19 cases


Total Covid-19 cases outside China


Given that a single cruise ship made up the bulk of cases outside China in early February, it is still useful to look at cases with and without that ship.

Total Covid-19 cases caught outside of China

New Covid-19 cases caught outside China each day

The average incubation period of COVID-19 probably less than a week (but could be as much as 24 days), and then an additional 3-4 days before diagnosis. So, you would expect measures like quarantines and travel restrictions to take around 10 days before showing up in statistics.

Time to doubling

This is an examination of how long it takes for cases or deaths to double.

Days taken for Covid-19 cases outside China to double to double

Number of days for Covid-19 cases and deaths in China to double


Winter is here

If we limit cases to only those caught in a particular country, exclude China, and then split countries into:

  • Winter countries: Northern Hemisphere Countries currently in winter (including Vietnam as the domestic transmission cases are in the north)
  • Summer/Equatorial countries: Southern Hemisphere countries currently in summer or Countries near the equator where temperatures are relatively high all year

Covid-19 cases by season

Covid-19 cases by season

Note: Countries near China are more likely to have contact with Chinese citizens and these countries are in winter which probably distorts this data.

New and total COVID-19 case numbers in Hubei, the rest of China

Our analysis (and the analysis of many others) suggests reporting of COVID-19 cases in Hubei province were under-reported.

Then, on the 7th of February, China changed its definition of how it is reporting new cases to exclude patients who test positive for the virus but have no symptoms will no longer be regarded as confirmed. This means up to 80% of cases might no longer be reported. On the 13th of February Hubei reclassified how it classifies cases. On 20th February Hubei reclassified again. All changes affect the quality of the data. Confirmed + suspected cases in China are our key measure.

We are tracking data from Hubei and the rest of China separately. We are sceptical of the China data, but there is some information in the series.

Total number of Covid-19 cases in China

On 7 Feb China made some adjustments to how they report data. Below we have made an estimate of what the case count might look like if China did not make this adjustment:

Extrapolated and suspected Covid-19 cases in Hubei province

Extrapolated, suspected and confirmed Covid-19 cases in China province


New confirmed Covid-19 cases in China

Daily change in confirmed & suspected Covid-19 cases

New and total Coronavirus death toll in Hubei, the rest of China, and the rest of the world

Total Covid-19 death toll in China


New daily Covid-19 deaths in China


Total Covid-19 deaths outside China & Iran

COVID-19 Mortality Rate using lag periods

The mortality rate is where we can see distinct differences in data. Dividing the number of deaths by the number of cases during the early stages of an outbreak is very misleading. People who were diagnosed today with the disease are still alive, but they still might die from the disease in the coming days.

A better way is to compare the current deaths to the number of cases from “x” days ago. We still don’t know how many days we should be looking back. The stats so far suggest that the median days from the first symptom to death is 14. But with a broad range from 6 to 41. And, we don’t know how long on average after the first symptom a person would take to become a case.

The below charts show the death rate if the right period to look back is 4, 8 or 12 days. Using data without Hubei, a mortality rate of somewhere between 0.5% and 3% is likely.

In recent days, data from Iran has skewed the results. There is likely a significantly larger outbreak in Iran than what is being reported. We have started showing our mortality rates for the rest of the world excluding Iran.

For more on what this means, see our article on understanding COVID-19 statistics


China Covid-19 mortality rate using different lag periods


Hubei province Covid-19 mortality rate using different lag periods


Worldwide Covid-19 mortality rate

More Analysis

See our latest investment view and here for our latest podcast. Keep in mind that the economic impact is not particularly related to the number of deaths, more important is the disruption to business which already looks to be significant.

Data sources

This is a list of some of the main data sources we use:  Probably the best one  Faster than worldometers for Chinese data, but slower on rest of the world data Official source for Chinese Data. Explains data adjustments. Official source for Hubei Data. Usually comes out a few hours before the China data.  Doesn’t always explain adjustments. The slowest to update, but the most authoritative in our view. More consistent with definitions than other sources. The prettiest pictures, but one of the slower sites to update. I don’t find the charts that useful. has a daily youtube wrap-up has a daily youtube wrap-up

Updated COVID-19 statistics and analysis