Public anger and attendant online censorship are running high after a May 22 gas explosion at Liushenyu coal mine in Shanxi province killed at least 82 miners and injured 128 others. While the cause of the accident remains under investigation by a team assembled by China’s State Council, Xinhua reported earlier this week that “ concealed mining tunnels, falsified drawings and outsourced and unregistered miners, who had not been provided with required life-saving location trackers, were contributing factors to the deadly incident.” It was China’s deadliest mining accident in nearly 17 years, after a 2009 gas explosion at a mine in Heilongjiang province that killed 108 people.
In addition to expressions of sympathy for the victims and and concern about mine safety in general, many initial online reactions were critical of what appeared to be a cover-up of the true death toll—at least, until higher authorities in Beijing issued a directive demanding the timely and accurate release of information, at which point the reported death toll leaped from eight to 82. Also fueling public anger were reports that although 247 miners were underground at the time of the blast, only 124 were listed on the company’s public-notice board, meaning that 123 others were unregistered and unauthorized to enter the mine: the fact that their information was missing from the system further complicated search and rescue efforts. Several online articles expressed incredulity that no documentation could be found for those 123 workers, with one noting that in 2026—with all our cell phones, cameras, and sophisticated information technology— it is inconceivable that a mine could simply “lose track” of 123 people working underground, unless the elision was a deliberate attempt to skirt the law.
CDT Chinese editors have archived nine articles about the mine disaster, at least three of which have been deleted from WeChat. Two of the undeleted articles were from large news outlets Phoenix News (“ Behind the ‘Hidden Coalfaces’: An Investigation into the Mine Disaster in Qinyuan, Shanxi Province”) and China Newsweek (“ Exposing a Mine’s Hidden Facets: Fake Rectifications, Fake Shutdowns, Fake Data, Fake Blueprints, and Fake Reports.”) The Phoenix News article, published two days after the explosion, is a longform investigation into the many longstanding safety violations at the Liushenyu mine, including the existence of illegal, unregistered coalfaces (暗面, àn miàn): secret tunnels bricked up and concealed from inspectors, where workers labored without the required trackers, meaning rescuers had no way of locating them after the explosion. Workers describe a culture of routine safety violations and corner-cutting resulting in inadequate ventilation, substandard safety training, and outdated or poorly maintained equipment. The company is characterized as unusually stingy even by mining industry standards, with frequent wage arrears and workers forced to buy their own boots, bedding, and basic supplies.
A WeChat article from blogger Wei Chunliang, “ The Shanxi Mine Disaster: The Deeper You Dig, the More Shocking It Is,” discusses some of the many serious and long-running safety violations at the mine, and marvels at the fact that they were allowed to go on for so long. A brief excerpt:
This coal mine is like a house with gaping holes on all four sides. If you wanted to plug those gaps, you’d hardly even know where to begin.
The mine’s safety management and oversight had completely collapsed. Even during the rescue operation, they were still lying and covering up, providing the rescue teams with useless, doctored blueprints. This prompted an outburst from the on-site operations commander: "Seriously? At a critical moment like this, you’re still trying to cover things up?" [ Chinese]
One of the deleted articles archived by CDT is “ The Moment Instructions Came Down From Beijing, the Death Toll in the Shanxi Coal Mine Disaster Rose Tenfold,” from WeChat account Li Yuchen. Li includes a timeline illustrating how, as soon as higher authorities in Beijing issued instructions to be more forthcoming in releasing information, the reported death toll in the mine disaster quickly went from eight to 82, suggesting that information on casualties had been suppressed. Li mentions that the mine had been fined twice for safety violations, but was allowed to continue operating, and suggests that tax revenue from the mine was too important to allow it to close down. He also notes that Ren Tiezhu, the chairman of the mining group and a two-time delegate to the Shanxi Provincial People’s Congress, has now been detained by the authorities. A portion of Li’s deleted article is translated below:
Looking back at May 23, the day after the mining disaster, the clearest record is Xinhua News Agency’s own timeline of the rescue effort.
Before dawn, there were four dead. At 7:19 a.m., eight dead. At 10:23 a.m., CCTV broadcasts the directive [calling for the public release of information]. At 12:46 p.m., over 50 are dead. At 1:06 p.m., 82 dead. At 2:00 p.m., the final report shows 82 dead, and nine still missing. Some outlets follow up with a figure of 90 deaths.
So this is what an official notification of a mining disaster looks like. There are still nine miners missing underground at Liushenyu. As for the 82 miners whose bodies were recovered, their numbers gradually appeared in the official reporting—but only in the hours after the central government directive was handed down.
So 247 people went down into the mine. That number, 247, was in compliance with procedure. Eight deaths were reported, then 82, then 90, all procedurally compliant. Past mine-safety violations resulted in fines, in compliance with procedure, but no shutdowns were ordered, because that wouldn’t be in line with procedure. Downplaying the initial death toll, then revising the figures upward? Just following procedure. Ren Tiezhu serving as a People’s Congress delegate, his mine being twice-fined but allowed to continue operating, and Ren being detained only after the tragedy occurred? Check, check, check—everything compliant with procedure.
Everyone behaved in accordance with procedure. Procedure wasn’t the problem—it was those 90 people trapped underground. Numbers like that are never released all at once: you report one figure first, wait for approval from the higher-ups, then report the next figure. That’s the procedure.
While search and rescue operations were still underway at Liushenyu mine, CCTV News reported that the State Council had issued a directive calling for “all-out efforts to locate and rescue trapped personnel, provide medical treatment to the injured, properly handle the aftermath of the accident, release information in a timely and accurate manner, ascertain the cause of the accident as soon as possible, and identify and punish those responsible in accordance with laws and regulations."
Note that one particular phrase: “Release information in a timely and accurate manner.”
The standard template for such a directive usually concerns the handling of search and rescue, medical treatment, aftermath and accountability. This time, the top decision-makers made a point of adding specific language about the public release of information.
There is only reason for such an addition: the higher-ups had already anticipated that the information would not be released ‘in a timely or accurate manner.’” [ Chinese]
Another of the censored pieces, a short article from Xu Peng’s WeChat account History Rhymes, focuses on the workers and families affected. Like many other commentators, Xu wonders whether the names of the 82 miners who died in the disaster will ever be made public. On a similar theme was a persuasive article (also later deleted) from former journalist Huang Zhijie, who argues that the best way to prevent cover-ups of mining and other industrial accidents is to release the lists of victim names. Huang describes past disasters in which that was the case, and discusses international norms regarding publishing lists of victims, although he acknowledges that in China, sometimes the pressure not to publish victims’ names comes from the surviving family members. A portion of Huang’s article is translated below:
When I was working as a journalist, we would frequently be assigned to report from the scene of major accidents, and the most frustrating and challenging aspect of that coverage was how to obtain and confirm the list of casualties. I used to think that if only there existed the societal consensus and institutional requirements for casualty lists to be made public, then many avoidable incidents and accidents could be averted, many unnecessary conflicts could be dispelled, and the work of journalists would be less arduous and less dangerous.
So how can we prevent the cover-ups and chronic underreporting of death tolls? The answer is to publish the names of those who have died. Once such a list is made public, it makes it harder to maintain an information blackout, and reduces the likelihood of the death toll being low-balled.
This is also a matter of public rights. When major disasters and accidents occur, the government has an obligation to investigate, and the people have a right to know the truth, the core of which is an accurate accounting of those who died. The public’s right to know is inextricably bound up with our ability to make informed choices about our own safety.
Internationally, it has become standard practice to publish the names of those killed in major accidents. Here in China, this has also been done on multiple occasions in the past. For example, a list of fatalities was published after the 2008 Wenchuan earthquake; the first batch alone numbered 19,065 victims. For the 2011 Wenzhou high-speed rail collision, the published list included 39 fatalities; for the 2015 Shanghai Bund stampede, 35 fatalities; and for the 2015 Tianjin port explosion, 165 fatalities. For a time, it seemed a consensus had formed: publishing the names of those who died in major accidents was the mark of a civilised society.
In recent years, however, the handling of most major accidents [in China] hasn’t included published lists of fatalities. Even in cases where it’s been officially confirmed that the death toll was underreported, the names of the victims have never been made public.
[…] Is the loss of human life merely a statistic, or does it represent a collection of specific, vibrant individuals with myriad human connections and social interactions? The answer depends on whether or not the list of victims is made public. Publishing such lists demonstrates respect for humanity, allows the deceased to rest in peace, offers solace to the living, helps clarify who was responsible, and may even prevent similar tragedies from recurring. [ Chinese]
