In Tibet, Officials’ Pursuit of Zero-COVID Sent Tens of Thousands into Mass ‘Isolation’ Facilities

When COVID hit Tibet in early August, after more than 900 days without a case, the response by Tibetans on social media was generally supportive of the local government and its efforts to contain the spread of the infection. Social media postings by Tibetans within Tibet applauded the heroic dedication of the white-suited anti-epidemic workers known in Chinese as dabai (“Big Whites”), sympathized when videos showed one of them collapsing from exhaustion, and joined the nation in praying for the epidemic to end soon. In Shigatse, Tibet’s second-largest city, four local businesses donated 5 million renminbi (U.S.$740,000) to the COVID management effort, while a group of Tibetan monasteries contributed 3 million renminbi ($444,000) in cash and materials to show their support, as they usually do at times of national crisis. Some posts complained about the difficulty in getting vegetables in Lhasa or expressed irritation at the extraordinarily long queues that people had to join in order to get their required tests, but there were few if any signs of serious dissent. There was no indication on Tibetan social media of mockery or criticism of the official videos showing dabai riding on horseback to test nomads in remote grassland camps for a virus that most likely the dabai were bringing with them. Neither were there complaints about the firetrucks and crews spraying streets and hotel rooms with disinfectant, even though the WHO rates the risk of surface transmission of COVID as insignificant and the Tibet government has warned that these sprays are toxic. The only serious signs of tension at that time came from angry tourists stuck in four-to-five-mile traffic jams as they tried to flee back home to inland China from a disease which they or other tourists had probably brought with them to Tibet during their summer holidays.

By the first week of September, that mood had changed. Dozens of short but critical videos were posted by Tibetans on Douyin, China’s version of TikTok, followed soon after by posts, mainly written in Chinese, on Weibo, the Chinese equivalent of Twitter. The videos and posts came mainly from the Tibetan capital, Lhasa. They described the anxieties and anger caused by the Lhasa government’s form of “silent management” (静默管理, jingmo guanli), the euphemism used in China for lockdowns. One of the earliest such videos is not dated, but was shared on Twitter in late August. In it, an older Tibetan man fights to hold back his tears as he begs officials to find some way for people to isolate in their homes instead of being taken into quarantine:

People in our courtyard are being taken away . . . there will be no one left in the courtyard. I felt happy when there were people and we could talk to one another, but now people in the courtyard are being taken away. I feel sad. What to do? Don’t you people think about [a solution] for this? Otherwise, no one will be left. Don’t you people think about it? Please!

Many other Tibetans followed by posting similar videos on Douyin describing their concerns about the lockdown and its effects.

The wave of online testimonies from Tibet resembled the explosions of anger that emerged online from other locked-down cities in China as officials sought to implement Xi Jinping’s policy goal of zero-COVID transmission, or “societal zeroing” (社会面清零, shehui mian qingling), meaning the elimination of the virus from the community. The most prominent of these online waves came from Shanghai between March and August this year and from Ghulja (known in Chinese as Yining) in Xinjiang in early September; these were social media tsunamis that spread so rapidly and so widely that it took the authorities several days to contain and neutralize them. The videos from Shanghai and Ghulja were gut-wrenching—people dying in their beds without medical attention, suicides attributed to prolonged social isolation, accounts of children and others dying from lack of food and medical treatment. In Ghulja, after the local government apologized at some length for its failures, the wave of public complaints seemed to decrease, either because of improvement in the situation or because of greater repression by the state, or both.

But the lockdown videos from Lhasa were not primarily about fears of imminent starvation, suicides (though there are now reports of these in Lhasa too), or widespread deaths, as in Ghulja. Neither are the Tibetan videos focused on outrage at China, its leaders, or the zero-COVID policy itself. In fact, in some Tibetan areas, and even in parts of the Tibetan capital, officials are said to be handling the epidemic restrictions well and their efforts are popularly appreciated, something the team of researchers whom I work with have been able to verify by studying social media reports and a number of unsolicited first-hand accounts sent to us by residents of the region.

The general attitude in Lhasa since early September, however, has been marked by concern and discontent. Its focus has been primarily on the mass transfer by city officials of thousands of citizens to isolation camps, and on the ways in which officials have carried out those transfers. But what has made the government pay attention to those concerns is that, on or just after September 15, messages from Tibet attracted the attention of sympathizers throughout China and went viral on Sinophone social media.

* * *

When China first began to deal with COVID during the mass outbreak in Wuhan in early 2020, the state’s flagship response, which it widely promoted and lauded, was the rapid construction of 16 field hospitals known in Chinese as fangcang (方舱) or “square cabin” hospitals. Usually referred to in China’s English-language media as “makeshift hospitals,” these were mostly erected in sports stadiums, exhibition halls, warehouses, gymnasiums, and similar spaces, in which people with mild-to-moderate COVID symptoms were isolated, monitored, and treated for their illness. Scores of studies have since been published by Chinese scientists and scholars describing these hospitals and their design as a world-leading innovation in healthcare and epidemic management. Media reports and medical studies enthuse about how the layout of fangcang (“three areas and two channels”) reduces the risk of cross-contamination; gushing articles in China’s media, especially those aimed at Westerners, note that the word fangcang sounds like the Chinese term for Noah’s Ark; multiple studies detail the ways in which on-site doctors, health workers, and psychiatrists helped fangcang patients form “community activities [including] eating together, watching television, dancing, reading, and celebrating birthdays.” One academic study measured the distance between beds in a fangcang—between 1 and 1.5 meters—and approvingly declared it to be optimal for recovery from illness because of patients “who had been strangers to each other being forced to socialize in intimate spatial conditions until they got to know each other better.” Internationally, the makeshift hospitals became part of China’s pandemic export drive: the government produced translations of the fangcang floorplans and management manuals in multiple foreign languages and sent experts abroad to advise on their construction. Chinese scholars in turn recommended that “other countries facing the COVID-19 pandemic should consider using Fangcang shelter hospitals as part of their public health response.” According to these scholars, Italy, Iran, Serbia, the U.S., and the U.K. have since followed China’s advice and imported the fangcang model.

Even before COVID arrived in the Tibet Autonomous Region (TAR) in early August, the local government had begun constructing fangcang hospitals. By September 7, there were 25 in the TAR, according to Chinese state media, and nine of them, with some 17,500 beds, were in Lhasa. Official publicity shows sites that resemble the original makeshift hospitals in Wuhan, with lavish supplies, extensive medical supervision, and hundreds of beds laid out in rows within a single open space.

Unofficial reports on social media, however, indicate that beside those nine fangcang in Lhasa there may be as many as 10 other such sites in or near the Tibetan capital, which have been little covered by the official press. Videos posted by people in these fangcang show few doctors, health workers, or medical equipment, let alone dance groups or psychiatrists. There are indeed some fangcang hospitals in Lhasa which are well equipped and which treat those who are moderately or seriously ill; it is these that have been widely praised in the Chinese press, and probably for good reason. But the unofficial online testimonies make it clear that most of the Lhasa fangcang are of a very different type and have a very different purpose. Their primary function is to quarantine COVID patients who have mild if any symptoms. This is because, under Chinese regulations issued in April 2020, following a principle known as the “Four Early’s” (“early detection, early reporting, early isolation, and early treatment”), all those who have tested positive for COVID but are not dangerously ill have to be quarantined in “centralized isolation points” (集中隔离点, jizhong geli dian); they have to remain there usually for 14 days, provided that they then receive negative test results on two successive days, at least 24 hours apart. That was what “societal zeroing” requires: removing COVID cases from the community by not, in principle, allowing people with mild cases of COVID to self-isolate at home. The majority of the Lhasa fangcang thus are not hospitals—they are the “centralized isolation points” or quarantine camps that house patients with mild or asymptomatic cases.

As officials in Lhasa strive to fulfill the zero-COVID goal, they have shipped thousands of Tibetans, as well as Han Chinese, Chinese Muslims, and other members of China’s minority groups (and at least one foreigner) to these isolation fangcang.

As officials in Lhasa strive to fulfill the zero-COVID goal, they have shipped thousands of Tibetans, as well as Han Chinese, Chinese Muslims, and other members of China’s minority groups (and at least one foreigner) to these isolation fangcang. Conditions in them vary widely. Official videos show people in Lhasa dancing in private rooms with food delivered to their doors in hotels that have been requisitioned to serve as centralized isolation points, and a number of private reports describe smaller isolation fangcang—usually ones with separate rooms for occupants—with ample staff, space, and provisions.

But the unofficial online videos show, in most cases, vast arenas, exhibition halls, storage spaces, office courtyards, or multi-story car parks, where hundreds of people share a single space with little effective separation between beds, if they have beds at all. Some are housed in unfinished apartment blocks; these are probably the better ones, because the occupants get separate rooms, but photographs show that the walls and floors in many of these are raw, dusty cement. Others are in larger camps, where videos and photographs show bathrooms flooded with water or sewage through which one has to wade to reach the toilet. Among the occupants are elderly people, infants, and the infirm. Multiple videos and Weibo messages describe long queues to enter the camps, difficulties in getting food once in them, and endless bureaucratic obstacles and delays in getting the test results without which one cannot be released.

One Tibetan student from a college in Lhasa, in a diary circulated anonymously on WeChat in mid-September, described his arrival at an isolation fangcang. He and a group of neighbors were driven to a converted school in Toelung Dechen on the western outskirts of Lhasa and confined with 6,000 others. According to the diary, they waited 24 hours before being given something to eat apart from milk. On subsequent days, the dabai served one meal a day. “When they brought the food,” the student wrote, “we just grabbed it without letting go and ate it immediately, because otherwise we wouldn’t get enough—if you don’t fight for the food you’ll be left hungry, there isn’t any other way. It makes no difference whether you are young or old—once the food comes, then they start fighting.” After several days, the fangcang occupants worked out a system whereby each group of eight people designated one person to collect food for that group; this solved the problem of the fighting, but by the time the food arrived it was always cold.

Other such stories are rife: An occupant in a camp of 4,000 people described eating steamed buns collected off the floor; others complained of unhygienic conditions in a fangcang kitchen; several note that diapers were distributed to women in a fangcang because no sanitary materials were available for them. By early September, in private conversation, though rarely online, Tibetans had started to refer to the isolation camps by the Tibetan word phag tshang, which sounds similar to fangcang but means pigsty.

* * *

The conditions in the Lhasa fangcang are not unique in China and may not be the worst: Earlier this year, Manya Koetse, editor-in-chief of the trend-monitoring site What’s on Weibo, surveyed Chinese social media accounts which described similar or worse conditions in isolation fangcang in the cities of Xi’an and Shanghai. In April this year, The Wall Street Journal published an account of a 4,000-person fangcang in Shanghai where there were no bathing facilities and the lights were never dimmed; more recently, videos have circulated showing a fangcang in Guixi, Jiangxi province, where occupants are confined to warehouses without bedding, let alone beds. A set of photographs posted on social media shows people in beds laid out in an open-air parking lot in Lanzhou, the capital of Gansu province; another video shows a row of people sleeping on the floor beneath a row of urinals, apparently in a fangcang in Xinjiang.

@fangshimin via Twitter

A screenshot of a tweet by @fangshimin, showing a video taken inside a fangcang in Lhasa, Tibet, posted September 15, 2022.

What is striking about the zero-COVID policy and its reality in Lhasa is thus not the conditions in the isolation fangcang there, but the number of these fangcang and of their occupants relative to the size of the city’s population. We know that on September 27, according to official figures, 42,937 people were held under “centralized isolation and observation” (集中隔离观察, jizhong geli guancha)—meaning that they were in fangcang of some kind—throughout the TAR. By that date, a total of 97,213 others had been in a TAR fangcang since the outbreak began in early August but had been released, probably after each spending at least two weeks in confinement. That’s a total of around 140,000 people who had gone through or were still in the TAR fangcang system by September 27. Since 48 percent of the COVID cases in the TAR up to that date were reported from Lhasa, we can guess that around half of the TAR fangcang population were residents of the capital. That suggests that between early August and late September, some 70,000 people in Lhasa—around 15 percent of the 480,000 residents of the Lhasa metropolitan area—had been or were still in a fangcang.

That figure does not include the unknown numbers of people in Lhasa who were living until now in the remaining traditional courtyards for which Lhasa is famous, clustered around the 7th-century Jokhang Temple in the old part of the city. After the outbreak in August, epidemic workers decided that those courtyards represented an unacceptable risk of cross-infection—a potential obstacle to the goal of “societal zeroing”—because their residents shared toilets or water sources. So those residents have been moved too, irrespective of their test results, to fangcang or to alternative accommodations, as a preventive measure.

The scale of COVID disruption and internment in Lhasa is, therefore, exceptional. The equivalent in a city the size of Shanghai would be around 3 to 4 million people. In Wuhan, by the time the 16 original fangcang in the city were closed in March 2020, they had housed a total of only 12,000 people, about 0.1 percent of the city’s population. The mass internment of 50,000 people in isolation fangcang in Xi’an in January this year was the largest known use of such facilities to date, but it represented only 0.6 percent of the city’s population—and they were distributed among 443 different fangcang, meaning that there were on average only about 100 people at each site. The Wuhan fangcang housed a total of 750 each on average over the course of a month; according to the diary of the Tibetan college student in Lhasa, the largest known Lhasa fangcang held 10 times that number six weeks ago. While some of the Lhasa facilities (the better ones) hold only about 400 people, on average it seems that there are 2,000 people at each site at any one time.

Evidently, the Lhasa government has been unable to manage these sites effectively. Official planning may have faltered because it focused largely on the fangcang hospitals, where medical treatment is the priority, rather than on the many isolation fangcang, where the vast majority of people with COVID in Lhasa are held: 93 percent of COVID cases in Lhasa, according to the government’s daily reports on the epidemic, have no symptoms and do not require treatment, and they are sent not to fangcang hospitals but to isolation fangcang. Once there, these people do not require treatment. But they do need food, water, sanitary conditions, and a transparent, consistent way to obtain release, which the Lhasa government has struggled to provide.

But patients with COVID, with or without symptoms, are only a small part of the problem the Lhasa government faced in executing its fangcang strategy. 140,150 people had been or were still in the TAR’s isolation fangcang from the outbreak of the epidemic as of September 27. Only 18,470 of these, however, can have been COVID patients, because that is the total number of COVID cases reported in the TAR by that date. Therefore just over 120,000 people—87 percent—in the TAR fangcang system must not have tested positive for COVID. Local officials quarantined them because they were close contacts of COVID cases, and the zero-COVID policy requires all close contacts to be “centrally isolated.” According to the regulations, they are to remain in a centralized isolation point for 14 days, and are then released if they have tested negative on days 1, 4, 7, and 14 of their confinement.

The crisis in Lhasa thus resulted firstly from its officials taking seriously orders from Beijing to quarantine all such people in centralized isolation points. Secondly, Lhasa officials seem to have defined the close-contact category in very broad terms. In so doing, they were following instructions from China’s top man in the TAR, Party Secretary Wang Junzheng:

Effectively take hold of community prevention and control on the frontline following the requirements: “encircle, scoop up [positive cases from the community], and extinguish [transmission],” focus on compounds, multi-rental buildings, urban villages, and other key areas, . . . and avoid [any situation where a place seems] sealed off but in reality is not tightly controlled.

In some cases, it appears officials sent entire communities to isolation fangcang because of a handful of positive results within their compound or neighborhood. In early October, for example, a resident in Lhasa sent us an audio recording of a discussion among members of a 500-strong community in north-central Lhasa. In the discussion, which was held on WeChat as an exchange of voice messages, the residents responded to news that they were likely to be transferred en masse to an isolation fangcang. While we don’t yet know what transpired afterwards, or how many positive cases had been found in that community, it is clear that the residents were not eager to comply, and that, as almost all of them pointed out, it made no sense in terms of epidemic management to transplant the entire community to a quarantine camp.

The city has thus ended up transferring an unusually high proportion of its population into the fangcang system. China publishes each day the total number of people throughout the country who are “under medical observation” (医学观察, yixue guancha), of whom an unknown proportion are in isolation fangcang. On September 27, the government reported that 145,548 people were under medical observation throughout China, meaning that they were held in isolation fangcang or under home isolation; on that same day, as we have seen, 42,937 close contacts were held in isolation fangcang in the TAR. This means that at least 30 percent of the close contacts held in isolation fangcang throughout China on that day were in Tibet, even though only 12 percent of China’s COVID cases were occurring there. Since 91 percent of the COVID cases in the TAR during the previous two weeks were reported from Lhasa, it is probable that the vast majority of those held in Tibet fangcang were in fact in Lhasa.

* * *

Isolation fangcang pose particular difficulties for the officials who run them. One is that many of those listed as asymptomatic COVID cases suspect, as they argue in videos or online posts, that their test results were wrong. Such claims are common wherever epidemic restrictions are in place, but in the case of Lhasa they appear to have some foundation, since at least one senior official from the city government has said that “individual test results are not accurate enough,” hinting at a high rate of false positives. In addition, the 93 percent rate of asymptomatic COVID cases in Lhasa is unusually high compared to the average rate in China, which is around 70-80 percent, leading to further doubts about the reliability of the COVID tests.

As a result, the isolation fangcang in Lhasa contain a few very mild COVID cases, some asymptomatic cases, a number of people who believe they are false positives, and a very large number of close contacts whose tests were negative. Although the state in Tibet has mobilized 120,000 people to serve as COVID volunteers and join the many officials who work as dabai, working desperately long hours alongside thousands of low-level cadres and local officials (at least two have died so far from overwork, according to government reports), it seems that the Lhasa bureaucrats interned far too many people, far too fast, and in camps that are far too large, and they have been unable to provide sufficient services to what at times has been over a tenth of the city’s population.

That failure, however, goes beyond issues of service provision. For obvious reasons, camp occupants with negative results and those with positive results—yin and yang, as people refer to them on Chinese social media—are supposed to be housed separately. But China’s fangcang appear to have been designed primarily to hold sick patients, not close contacts, and, at least in some large camps, it seems that there is not always separation between those with positive and those with negative results. As one Lhasa resident wrote in a private text message this week, “The most discussed issue now is that there are loads of people getting sent to fangcangs while having a negative [test] result. Negatives and positives are quarantined together. No one understands why.” Numerous Weibo posts express similar concern about the likely spread of COVID within the isolation fangcang, while other posts and videos are by occupants who have tested negative but know of no reason why they have been brought to a camp and fear cross-infection there. One such case is outlined in a phone call, a transcript of which is excerpted below, recorded and circulated privately in Lhasa in late September. The conversation is between a local official and a Tibetan man who had been sent to a fangcang with his wife and children. The family had arrived with negative test results, but had contracted COVID in the camp:

Father: I can’t get anyone [of the officials] to talk about this at all. It’s only because I can’t get there right now that I haven’t gone to the central government [in Beijing] to report what’s happened to us, otherwise I swear a hundred percent I would.

Official: I understand.

Father: During this epidemic, I don’t understand whether community [leaders] have money to waste or some kind of corruption, but we are clean [have negative test results] and we’ve been put with people who’ve tested positive. . . It drives me crazy, it makes me so angry that those of us who’ve tested negative are mixed with those who’re positive.

Official: It was the mayor’s decision.

Father: I begged them not to, because one of my children is only just one year old, the other one is six, seven years old. I was talking to my wife last night and said to her that it would be fine to be here if we were positive, we wouldn’t complain, because at least we would have a room to stay in. So we were hoping that things would get better. But this morning they told us that we can’t stay here, because the younger child, the one who’s under one year old, is not allowed to stay in this camp. I swear by the Three Jewels you can take us as a family wherever you want. We are ready to go together. But if you are planning to separate us as a family, I will stay and die here. If I die here, people will know. And my daughter’s COVID is getting more serious now, she was crying all morning.

Official: Please wait. I will tell them to find a way.

Father: I just told you, if I could, I would go to the center [Beijing] to appeal. There is no hope for the TAR, I think the whole TAR has already gone rotten, I swear!

Official: I understand you.

Father: If we were positive, I would thank the government for their kindness, and I would be so sorry that I’d made difficulties for the government, because here the food and other things are okay. But even though I tested negative the government forced me to come here, to this, whatever you call it, pigsty. . .

The father goes on to repeat that he will kill himself rather than abandon his family or allow it to be split up. His current situation is not known. It may not be coincidental that although the TAR government announced on September 4 that the entire region had “achieved or basically achieved ‘zero-COVID in society’,” it is still under lockdown six weeks later. Over the two weeks following that announcement, the case rate in Lhasa indeed fell from 289 to 33 cases a day, but then increased sharply to a daily average of 132 for the next 10 days, perhaps a sign that cross-infection might have been happening in the camps.

* * *

Popular doubts about why so many people are being sent to the isolation fangcang in Lhasa are heightened by the mechanics of the transfer process. The Tibetan student’s diary describes his experience after he and his housemate had positive test results, but with no symptoms:

On Monday evening at 6:00 we got a notice saying we needed to gather outside, and we immediately went there. We didn’t know what the notice was about, so we hadn’t brought any food or clothes with us. First they registered each household, and then there were buses and ambulances to transfer patients [to isolation fangcang and hospitals]. We went there too, waited in a line for six hours, and got our names registered. At 1:00 a.m., I got a seat on a bus. . .

I fell asleep on the bus and when I woke up saw that we were at a crossroads, where there was a line of buses [waiting]. I took this chance to go outside the bus to have a pee, and looked at the time. I saw it was 4:30 a.m. I asked my friend if we had already reached Nyemo county, since the road there [from Lhasa] is good. But he said we had been on the bus without moving for two hours.

At that time, everyone was talking, making noise, saying that we needed to report upward to the higher-level authorities. Both men and women were saying this. Then the landlady [who was also on the bus] borrowed an old man’s phone and left a lot of messages in the [Neighborhood] Committee’s [WeChat] group. In particular, she asked, why are we treated like this? She said a lot about the fact that we had had nothing to eat or drink. . .

We were given one piece of bread each and some milk. Many people were saying, if the bus is not going anywhere, let’s get out and walk back home. It was the early hours of the morning then, and there were all these buses on the road, going right and left. At around dawn, we arrived at a place with a multi-storied building, which from a distance looked like a school. Once there, we were assigned to our rooms.

The student’s account of the transfer to the camp is typical of several that my research team has seen on social media or that have been relayed to us directly. It seems that the transfers are almost always done at night, that the registration process, let alone the drive, often takes many hours, if not all night, that no one seems to know where they are being taken, and that food or drink are hard to come by. A glimpse of the transfer process is conveyed by a video that shows long queues as people wait to board the buses; another shows passengers crowded on a bus, each with a plastic refuse bag with some belongings; a third shows a bus convoy on its way through the city at the dead of night, escorted by two police cars; a more disturbing one shows dabai forcing two women into a minibus. The nocturnal transfers have become so standard that a group of anonymous Tibetans has circulated mock posters on social media advertising a non-existent film called Lha sa’I nam gung gi spyi ‘khor, “The Lhasa Night Bus.” They described the film as “a thriller in Chinese and Tibetan” and gave it an 8.6 audience rating.

* * *

Initially, the testimonies about fangcang conditions in Lhasa were shot on videos by Tibetans speaking in their own language. These videos were then posted on Douyin, a platform that is easy to use even for those who cannot write, and perhaps harder for automated censoring software to track. The videos were directed at fellow Tibetans within China. Many of the posters were ordinary people, not intellectuals, including a number of Tibetan migrant manual workers or truck-drivers within Tibet who, stranded by the lockdown far from their homes, had no way to find work or generate income. Soon others joined them, posting comments and testimonies about the situation in Lhasa as text messages on Weibo, often calling on people in China to take note of the worsening conditions in Lhasa. These were written or spoken mainly in Chinese. One was a Tibetan online influencer with a major following, certainly putting her career at risk by speaking out, while others were ethnic Chinese studying in Tibet or perhaps working in the private sector. On September 15, this strategy of mobilizing Chinese online attention in order to amplify calls for help succeeded: A number of Chinese-language posts and threads produced by these individuals went viral on Chinese social media. A Weibo thread with the hashtag #TibetEpidemic (#西藏疫情#) received 210 million hits within four days; since there are only 7 million Tibetans in China, most readers must have been Chinese. Numerous expressions of concern about the situation in Lhasa were posted by Chinese netizens on these and other threads, calling on the government to pay attention to conditions there.

The initial response by the authorities was the usual one: Numerous posters reported that their texts or videos had been deleted or that they had decided to remove them. Alexander Boyd, a Senior Editor at China Digital Times, noted that on September 15, hundreds of posts on the #Tibet Epidemic Prevention and Control (#西藏疫情防控#) thread, many of them highly critical eye-witness accounts from Lhasa,disappeared—they had been removed from search results. “Big V” posters—public figures seen as friendly to the government—started to appear on the thread, flooding it with positive material, and hits on these threads dropped from 36 million on September 15 to around 1 million on the 19th. Most posts on the #TibetEpidemic thread also disappeared on that day. A group of concerned people in Tibet and China, however, were trying to keep such threads online. “People are working on this one now,” one of these people wrote to us from Lhasa the next day, referring to another thread about the Lhasa epidemic. “They keep getting deleted but people just keep posting. It keeps getting taken down from the trending hashtags section. But then people keep working on it and it appears again. Then it disappears again.” Their efforts failed, however: The daily hits on that thread reached 115 million on the day they sent that message but were close to zero four days later.

On September 17, however, two days after “harmonization,” the online mobilization of the Chinese public by Tibetans and their Chinese supporters produced a remarkable result: The Lhasa government issued a rare apology, and not just as a printed statement. The executive deputy mayor of Lhasa, Dramdul (referred to as Zhandui in Chinese), was shown on television promising “to strive to improve the service guarantee level of fangcang, isolation points, and other services.” “I would like to express my deep apologies to the general public!” he proclaimed. “We know that there are still a lot of gaps between our work and the expectations and requirements of the masses. . . We must face up to problems, improve shortcomings, and further optimize prevention and control measures.” He referred specifically to “public opinion and demand channels,” which he promised would be unblocked. He then, remarkably, stood up and bowed briefly to the camera.

Dramdul is of relatively low rank, and neither his superior, the Lhasa Party secretary, nor the seniormost figure in Tibet, regional Party Secretary Wang Junzheng, who is ethnic Chinese, appeared at the press conference or apologized for the chaos in Lhasa. But buried in a newspaper report that day about Party Secretary Wang’s inspection of a new lighting system at the local airport was an apparently random sentence attributed to him: “We will further improve the setting up of designated hospitals, fangcang hospitals, and isolation points, so as to ensure that we are prepared, available, and able to respond at critical times.” China’s top man in Tibet had not apologized, but it was a rare moment, nonetheless: He had signaled his recognition of public discontent.

The following day, September 18, the deputy director of the Lhasa Fangcang Hospital Working Group Office, an ethnic Chinese official called Tu Xin, was rolled out to give his own televised press conference. Again, it was described as a response to “the issues that the masses are keenly concerned about, namely the operation and management of the fangcang hospitals in Lhasa.” Lhasa, Tu announced, will “actively improve the supply of quilts, blankets, and other warm items to make patients feel warm.” He promised to arrange for the delivery of milk, instant noodles, ham sausages, mineral water, and other foods in a timely manner, and especially to arrange for halal restaurants to provide food for Muslims, “so that patients can eat hot meals and eat well.” The camp staff henceforth would “do a good job in sterilizing the fangcang environment and in cleaning up garbage to provide good surroundings and living conditions,” Tu assured the public, in what was clearly a form of urgent crisis management. Even so, Tu couldn’t resist reverting to standard Chinese messaging about fangcang: His team will also now organize all fangcang in Lhasa, he said, “to carry out ‘sing, dance, move, and get well’ activities and will set up a team to carry out psychological counseling to improve public health.” The Tibetan satirists on social media in due course released another mock film poster, this time advertising a film called Sgrom tshang gi sgor gzhes, “The Fangcang Circle Dance.” It showed a photograph of happy Tibetans dancing in front of the Intercontinental Hotel in the eastern outskirts of Lhasa, where a real fangcang (next to a long-standing police detention camp) has been set up.

In other ways, however, the state has reverted to the norm. The Weibo threads remain largely “harmonized” for some time and in late September the top post on one of them declared that “individual netizens who publish and repost false rumors about the epidemic on the Internet without verification cause bad social influence.” On September 16, an exile Tibetan media outlet in India reported that seven Tibetans from Nagchu, north of Lhasa, had been fined 3,000 renminbi ($410) each and put in administrative detention for at least five days for sharing videos online about conditions in the fangcang. Three people—one from Lhasa, one from Shigatse, and one from Meldrogungkar (Mozhuogongka), a county just outside Lhasa—were detained for “posting epidemic-related rumors on Weibo and Douyin in order to confuse the public, incite antagonism, and encourage people to gather together to make trouble,” according to a notice issued by the Lhasa police on September 18. The detainees are likely to have spent up to 15 days in detention, given the law cited in the case. Two days later, the Lhasa police announced that 786 people had been punished under the same law for epidemic-related offences, including “going out without permission” and “spreading rumors.” By October 10, that number had risen to 1,081.

* * *

By September 30, the number of people held in “centralized isolation points” in the TAR had decreased to 28,000, half the number two weeks earlier; since then, the government has stopped releasing figures about close contacts in the fangcang. On October 13, the daily number of reported COVID cases in Lhasa dropped to a single figure for the first time since lockdowns in the city started some 66 days earlier. The government announced that all but 516 of the 90,000 tourists trapped in the city by the lockdown since August had been allowed to leave for their homes in inland China. Of the 366 residential compounds or communities in the city that had been categorized as high- or medium-risk two months before, only six were listed as still at risk, all of them at a medium level. Although much of the city remained under lockdown, in most areas one person from each household was now allowed to leave their home for an hour a day, albeit only within the boundaries of their community or block. Later, residents in several areas, such as the university, were given permission to go outside for two hours a day on every second day, and on October 21, the TAR government announced that “the overall and orderly restoration” of normal life—a gradual end to restrictions—would start soon.

So far, however, the lockdown remains in place in much of Lhasa, night buses are still ferrying people to the fangcang, and there are no indications of any camps ceasing operation. Meanwhile, reports from the city suggest increasing signs of strain. Besides six lockdown-related suicides documented by exile-linked organizations in India and the U.S., three of which took place in isolation fangcang or because of family members having been taken to a fangcang, rumors of other suicides in the city are rife. Videos of two failed attempts have appeared online and the Lhasa government has placed increasing emphasis on the need to “do a good job of psychological adjustment and emotional counseling for quarantined persons.” “If you make the common people unable to survive,” as one online poster put it, “it ends with mass jumping from buildings.” Although there are no official reports of deaths from COVID in Tibet, exile-linked media based in the U.S. have attributed at least three fatalities in Tibet to lack of medical attention for COVID during the lockdown, and our research confirmed through personal connections at least three other cases of deaths of people who reportedly could not get prompt treatment for non-COVID conditions. The pain of those deaths was exacerbated for their families by the fact that, along with prayer ceremonies involving monks or others, sky burials—the principal form of funeral among Tibetans, considered exceptionally beneficial for the deceased because the entire body is donated for the benefit of other creatures—are not permitted under lockdown rules, which say officials should “encourage” citizens to hold “simple funerals.”

The cyber environment in Lhasa reflects these stresses. Video testimonies by Tibetans describing the difficulties they face under lockdown have started to appear again, and criticisms of the local government appear frequently on Weibo, written mostly by Chinese residents in Lhasa rather than Tibetans, perhaps because Tibetans anticipate much more serious repercussions if they speak out than do their Chinese co-nationals. Many of these complaints focus not on restrictions or conditions in the city, but on the fear of confinement in a fangcang. As one Lhasa resident wrote to us last week, “I’m not particularly concerned about getting infected—I’m more concerned about being sent to those camps.” That discontent has spread, he added, so that “people who do not have a red QR health code on their phones [indicating that they are COVID-positive] now refuse to go there [to the fangcang].” Anyone actually resisting an isolation order is unlikely to have succeeded, but on October 26, hundreds of Chinese migrant workers in Lhasa were successful in one demand: They staged a day-long protest calling on officials to let them return to their homes in lowland China. The following day, they were assured that they could leave the city if they could show the required number of successive negative test results, and thousands appear to have left soon after. But for those who remain, the lockdown has not ended. On average, five new COVID cases have been reported in Lhasa each day since mid-October—zero-COVID has not been achieved. The Lhasa government no longer announces the number of areas in the city that remain locked down, but in many places, residents are still limited to an hour a day outside their homes but within their neighborhood.

* * *

The story of the fangcang camps, and of the Lhasa government’s handling of the epidemic, has revealed some of the limits to the ability of a Chinese-style administration to manage a community in crisis. That failure has been blamed by some Chinese social media commentators on the fact that Tibet has for years been the recipient of “paired assistance,” meaning lavish funding from other Chinese provinces, and has been largely run by ethnic Chinese cadres on temporary postings. This, these critics say, has led Tibetan officials to become spoiled by Chinese largesse and competence, so that the local officials have not acquired the ability to run such camps on their own. This argument, reminiscent of colonial attitudes, is weak. No Tibetan has been allowed to hold the topmost position in the local government since China annexed Tibet in the 1950s, and thousands of ethnic Chinese cadres are still running much of the administration in Tibet; hundreds more Chinese were imported from provinces throughout the People’s Republic of China in August to handle the epidemic, in many cases specifically to set up and run the fangcang system. The failure of governance in Lhasa is thus not a question of Tibetan competence or of governmental inefficiency; in fact, it is more likely to be the opposite. Lhasa officials did exactly what the Chinese leadership ordered them to do: According to the zero-COVID policy, they were to remove all traces of the virus from society. Therefore, they dutifully moved much of society into isolation camps.

According to the zero-COVID policy, they were to remove all traces of the virus from society. Therefore, they dutifully moved much of society into isolation camps.
The consequences of the Lhasa government’s unwavering implementation of orders from above tell us much about how policy works, and doesn’t work, in the deeply hierarchical, command-based system of contemporary Chinese politics. It also tells us about policy in China’s colonial peripheries: Within the overall top-down system, officials in highly sensitive areas like Tibet or Xinjiang, especially if they themselves are members of a minority group, have to do much more, or believe that they have to do much more, than their majority Chinese counterparts to demonstrate compliance with the demands of their superiors. Officials in Tibet thus may not have much space in which to exercise discretion about which rulings from Beijing to overlook or modify in practice, an option which is essential to efficient running of the system.

What thus seems to have happened in Lhasa is that officials there applied the state’s definition of close contacts in a literal way, emptying out entire communities in response to minor outbreaks in order to meet the requirements of the zero-COVID policy. They also focused their attention and resources on equipping and running fangcang hospitals, which have housed the 1,500 confirmed, symptomatic COVID cases in Tibet, rather than the isolation fangcang, which have housed about a hundred times that number. That management decision is probably not due to incompetence or malice, but to the persuasive power of the Party’s narrative about its excellence in epidemic management, which, through incessant praise of makeshift hospitals in lavishly illustrated articles and glossy videos, insisted that fangcang hospitals are the key to success in defeating COVID. If so, both the officials and the citizens of Lhasa have been victims of the Party’s own self-praising propaganda.

Not all officials in Tibet rigidly imposed the fangcang policy: We know from direct sources of local Tibetan officials who quietly took the risk of allowing a patient and his or her immediate contacts to isolate at home, just as the elderly Tibetan man had begged officials to do in the earliest known video testimony. The September 15 viral wave likely led other officials in Tibet to think twice before sending people off to a fangcang. But we also know of over 100 officials, many of them Tibetans, who have been fired or penalized during the COVID crisis for “inadequate implementation of pandemic prevention and control work” and related lapses. The pressure on officials in Tibet, as in Xinjiang and now Inner Mongolia too, is to over-demonstrate their compliance with orders from Beijing—notwithstanding that these are technically “autonomous regions”—even when those orders are dangerously inflexible and arguably ill-conceived.

This feature of political culture in Tibet or other autonomous regions in China is not news for those who live in them, since policy implementation has almost always been more rigid there than in the Chinese heartlands. For many among the Tibetan and the wider Chinese publics, however, there have been other lessons learned from the lockdown months in Lhasa, besides the pain, grief, and discomfort that they, along with innumerable others elsewhere in China, have experienced: the inability or reluctance of the Lhasa administration to moderate the orders imposed upon it; the evident risks built into China’s system of inflexible, top-down policy commands; and the effectiveness of multiethnic, cross-regional mobilization on social media in pushing local officials to apologize for failure.

For China as a whole, however, the events in Lhasa have a more worrying significance. According to images and videos posted on the Twitter account @Songpinganq, scores of far larger isolation fangcang are currently under construction in many if not all of China’s towns and cities. At least one of these fangcang sites is reportedly designed to hold up to 30,000 people. These videos are as yet unverified, but they strongly suggest that China has quietly dumped the much-praised fangcang model it exported around the world, which for some unexplained reason placed hundreds of people in a single open space, ideal for spreading infection. Instead, the new fangcang appear to be composed of single-cell isolation units, one for each occupant or family; some videos suggest that the occupants’ only contact with the outside world will be via a hatch in the door. Whether this new type of fangcang is yet in operation is unknown, but it seems likely that more and more cities in China will soon be moving people to such sites. By October 30, China’s official figure for current “close contacts under medical observation,” which includes all those in the fangcang system, had reached over half a million—569,567 to be precise—nearly four times the figure one month earlier. As officials strive to achieve the goal of zero-COVID, the experience of fangcang overreach in Lhasa looks set to be replayed across the country in the weeks and months to come.