How the Coronavirus Revealed Authoritarianism’s Fatal Flaw

China’s use of surveillance and censorship makes it harder for Xi Jinping to know what’s going on in his own country.

People wearing masks walk past a portrait of Chinese President Xi Jinping.
Aly Song / Reuters

China is in the grip of a momentous crisis. The novel coronavirus that emerged late last year has already claimed three times more lives than the SARS outbreak in 2003, and it is still spreading. More than 50 million people (more than the combined metro populations of New York, Los Angeles, Chicago, and San Francisco) remain under historically unprecedented lockdown, unable to leave their city—and in many cases, even their apartment. Many countries no longer accept visiting Chinese nationals, or if they do, quarantine them for weeks. Big companies are pulling out of trade shows. Production is suffering. Profound economic consequences are bound to ensue, not just in China but around the world.

How did Xi Jinping—the general secretary of the Communist Party of China, who has been consolidating his power since taking over the post in 2012—let things get to this point?

It might be that he didn’t fully know what was happening in his own country until it was too late.

Xi would be far from the first authoritarian to have been blindsided. Ironically, for all the talk of the technological side of Chinese authoritarianism, China’s use of technology to ratchet up surveillance and censorship may have made things worse, by making it less likely that Xi would even know what was going on in his own country.

Authoritarian blindness is a perennial problem, especially in large countries like China with centralized, top-down administration. Indeed, Xi would not even be the first Chinese ruler to fall victim to the totality of his own power. On August 4, 1958, buoyed by reports pouring in from around the country of record grain, rice, and peanut production, an exuberant Chairman Mao Zedong wondered how to get rid of the excess, and advised people to eat “five meals a day.” Many did, gorging themselves in the new regime canteens and even dumping massive amounts of “leftovers” down gutters and toilets. Export agreements were made to send tons of food abroad in return for machinery or currency. Just months later, perhaps the greatest famine in recorded history began, in which tens of millions would die because, in fact, there was no such surplus. Quite the opposite: The misguided agricultural policies of the Great Leap Forward had caused a collapse in food production. Yet instead of reporting the massive failures, the apparatchiks in various provinces had engaged in competitive exaggeration, reporting ever-increasing surpluses both because they were afraid of reporting bad news and because they wanted to please their superiors.

Mao didn’t know famine was at hand, because he had set up a system that ensured he would hear lies.

Smart rulers have tried to create workarounds to avoid this authoritarian dilemma. Dynastic China, for example, had institutionalized mechanisms to petition the emperor: a right that was theoretically granted to everyone, including the lowest farmers and the poorest city dwellers. This system was intended to check corruption in provinces and uncover problems, but in practice, it was limited in many ways, filtered through courtiers to a single emperor, who could listen to only so many in a day. Many rulers also cultivated their own independent sources of information in far-flung provinces.

Thanks to technology, there is a much more robust option for authoritarians in the 21st century: big-data analytics in a digital public sphere. For a few years, it appeared that China had found a way to be responsive to its citizens without giving them political power. Researchers have shown, for example, that posts on Weibo (China’s Twitter) complaining about problems in governance or corruption weren’t all censored. Many were allowed to stay up, allowing crucial information to trickle up to authorities. For example, viral posts about forced demolitions (a common occurrence in China) or medical mistreatment led to authorities sacking the officials involved, or to victim compensation that would otherwise not have occurred. A corrupt official was even removed from office after outraged netizens on social media pointed out the expensive watches he wore, which were impossible to buy on his government salary.

The public sphere in China during those years wasn’t a free-for-all, to be sure. One couldn’t call for collective action or for deposing the central government. But social media gave citizens a voice and a way to make an impact, and it served as an early-warning system for party leaders. (The only other topic that seemed to be off-limits was the censors themselves—researchers found that they eagerly zapped complaints directed at them.)

This responsive form of authoritarianism didn’t happen just on social media. Beginning in the early 2000s, China held “deliberative polls” in which citizens debated local budgets, important issues, and even reforms that would give them the right to information on government actions. In Zeguo township in Wenling, a municipality of more than 1 million residents, authorities created deliberative bodies wherein they engaged citizens (usually a few hundred, with randomness ensuring they were representative of the population) over a few days by providing information (including detailed accounts of the city’s budget) and hosting discussions to decide on issues of public significance. Authorities sometimes went as far as to pledge, in advance, to abide by the decisions of these bodies. For many years, such experiments flourished all over China and, combined with the digital public sphere, led scholars to wonder whether the “deliberative turn” in the country’s otherwise authoritarian state was not a means of weakening authoritarianism, but of making it more sustainable.

Yet, this deliberative turn was soon reversed.

Since taking power in 2012, Xi has shifted back to traditional one-man rule, concentrating more and more power into his hands. He has deployed an ever-more suffocating system of surveillance, propaganda, and repression, while attempting to create a cult of personality reminiscent of the Mao era, except with apps instead of little red books.

Unlike books, though, apps can spy on people.

One hundred million or so people in China have been, ahem, persuaded to download a party-propaganda app named “Study Xi, Strong Nation,” which makes users watch inculcation videos and take quizzes in a gamified, points-based system. It also allegedly gives the government access to the complete contents of users’ phones. It almost doesn’t matter whether the app contains such backdoor access or not: Reasonable people will act as if it does and be wary in all of their communications. Xi has also expanded China’s system of cameras linked to facial-recognition databases, which may someday be able to identify people everywhere they go. Again, the actual workings of the system are secondary to their chilling effects: For ordinary people, the safe assumption is that if they are in the wrong place at the wrong time, the authorities will know.

An earlier hint that Xi’s China was falling into authoritarian blindness came during the ongoing Hong Kong protests. The demonstrations had started over a minor demand—the withdrawal of an extradition bill of little strategic importance to Beijing. Protest is the traditional way that Hong Kongers, who do not have full voting rights, express discontent. But this time the Beijing insiders miscalculated. They genuinely believed that the real cause for the Hong Kong unrest was the high rents on the densely populated island, and also thought that the people did not support the protesters. Authoritarian blindness had turned an easily solvable problem into a bigger, durable crisis that exacted a much heavier political toll, a pattern that would repeat itself after a mysterious strain of pneumonia emerged in a Wuhan seafood market.

In early December, a strange cluster of patients from a local seafood market, which also sold wildlife for consumption, started showing up in Wuhan hospitals. These initial patients developed a fever and pneumonia that did not seem to be caused by any known viruses. Given the SARS experience of 2003, local doctors were quickly alarmed. With any such novel virus, medical providers are keen to know how it spreads: If the virus is unable to spread from human to human, it’s a tragedy, but a local one, and for only a few people. If it can sustainably spread from human to human, as was the case with SARS, it could turn into a global pandemic, with potentially massive numbers of victims.

Given exponential growth dynamics of infectious diseases, containing an epidemic is straightforward early on, but nearly impossible once a disease spreads among a population. So it’s maximally important to identify and quarantine candidate cases as early as possible, and that means leadership must have access to accurate information.

Before the month of December was out, the hospitals in Wuhan knew that the coronavirus was spreading among humans. Medical workers who had treated the sick but never visited the seafood market were falling ill. On December 30, a group of doctors attempted to alert the public, saying that seven patients were in isolation due to a SARS-like disease. On the same day, an official document admitting both a link to the seafood market and a new disease was leaked online. On December 31, facing swirling rumors, the Wuhan government made its first official announcement, confirming 27 cases but, crucially, denying human-to-human transmission. Teams in hazmat suits were finally sent to close down the seafood market, though without explaining much to the befuddled, scared vendors. On January 1, police said they had punished eight medical workers for “rumors,” including a doctor named Li Wenliang, who was among the initial group of whistleblowers.

While the unsuspecting population of Wuhan, a city of 11 million, went about its business, the local government did not update the number of infected people from January 5 to January 10. But the signs of sustained human-to-human transmission grew. Emergency wards were filling up, not just with people who had been to the seafood market, but with their family members as well. On January 6, Li noticed an infection in the scan of a fellow doctor, but officials at the hospital “ordered him not to disclose any information to the public or the media.” On January 7, another infected person was operated on, spreading the disease to 14 more medical workers.

Things went on in this suspended state for another 10 days, while the virus kept spreading. Incredibly, on January 19, just one day after the death of yet another doctor who had become infected, officials from across the populous Hubei province held a 40,000-family outdoor banquet in Wuhan, its capital, as part of the official celebrations for China’s Lunar New Year.

The dam broke on January 20—just three days before Wuhan would initiate a draconian lockdown that blocked millions of people from leaving. On that day, the respected SARS scientist Zhong Nanshan went on national television, confirming the new virus and human-to-human transmission. That same day, Xi Jinping gave his first public speech about the coronavirus, after he returned from an overseas trip to Myanmar.

Things have dramatically escalated since then. Just one month later, by some estimates, more than 700 million people in China are living under some form of restrictions to their movements, in addition to the severe lockdown in the Hubei province. Domestic social media has erupted in anger at both China’s central leadership and local officials in Hubei province, where the disease began. There are calls for free speech, fury over the death of one of the early medical whistleblowers from the virus, and frustrations with the quarantine.

It’s not clear why Xi let things spin so far out of control. It might be that he brushed aside concerns from his aides until it was too late, but a stronger possibility is that he did not know the crucial details. Hubei authorities may have lied, not just to the public but also upward—to the central government. Just as Mao didn’t know about the massive crop failures, Xi may not have known that a novel coronavirus with sustained human-to-human transmission was brewing into a global pandemic until too late.

It’s nearly impossible to gather direct evidence from such a secretive state, but consider the strong, divergent actions before and after January 20—within one day, Hubei officials went from almost complete cover-up and business as usual to shutting down a whole city.

Another reason to think Xi did not know is that he would have every incentive to act quickly given China’s experience with SARS, during which he was already a member of the Central Committee of the Communist Party. Both SARS and the Wuhan virus (which causes the disease now dubbed COVID-19) are zoonotic coronaviruses, with similar origins and pandemic potential. SARS was contained, though barely, and not before significant economic costs following a failed cover-up. Such an experience should have made it clear that cover-ups are futile when it comes to pandemics, because viruses don’t respect borders. (The Soviet Union learned that radiation doesn’t either, when Sweden alerted the world to the Chernobyl accident.)

It’s hard to imagine that a leader of Xi’s experience would be so lax as to let the disease spread freely for almost two months, only to turn around and shut the whole country down practically overnight.

In many ways, his hand was forced by his own system. Under the conditions of massive surveillance and censorship that have grown under Xi, the central government likely had little to no signals besides official reports to detect, such as online public conversations about the mystery pneumonia. In contrast, during the SARS epidemic, some of the earliest signs were online conversations and rumors in China about a flu outbreak. These were picked up by the Global Outbreak Alert and Response Network, who alerted the World Health Organization, who then started pressuring China to come clean, which finally triggered successful containment efforts.

If people are too afraid to talk, and if punishing people for “rumors” becomes the norm, a doctor punished for spreading news of a disease in one province becomes just another day, rather than an indication of impending crisis. Later, under criticism, Xi would say he gave instructions for fighting the virus as early as January 7, implying that he knew about it all along. But how could he admit the alternative? This is his system.

Contrary to common belief, the killer digital app for authoritarianism isn’t listening in on people through increased surveillance, but listening to them as they express their honest opinions, especially complaints. An Orwellian surveillance-based system would be overwhelming and repressive, as it is now in China, but it would also be similar to losing sensation in parts of one’s body due to nerve injuries. Without the pain to warn the brain, the hand stays on the hot stove, unaware of the damage to the flesh until it’s too late.

During the Ming dynasty, Emperor Zhu Di found out that some petitions to the emperor had not made it to him, because officials were blocking them. He was alarmed and ordered such blocks removed. “Stability depends on superior and inferior communicating; there is none when they do not. From ancient times, many a state has fallen because a ruler did not know the affairs of the people,” he said. Xi would have done well to take note.

Zeynep Tufekci is a contributing writer at The Atlantic and an associate professor at the University of North Carolina. She studies the interaction between digital technology, artificial intelligence, and society.