Why China Can’t Export Its Model of Surveillance

Over the past two decades, Chinese leaders have built a high-tech surveillance system of seemingly extraordinary sophistication. Facial recognition software, Internet monitoring, and ubiquitous video cameras give the impression that the ruling Chinese Communist Party (CCP) has finally accomplished the dictator’s dream of building a surveillance state like the one imagined in George Orwell’s 1984. 

A high-tech surveillance network now blankets the entire country, and the potency of this system was on full display in November 2022, when nationwide protests against China’s COVID lockdown shocked the party. Although the protesters were careful to conceal their faces with masks and hats, the police used mobile-phone location data to track them down. Mass arrests followed.

Beijing’s surveillance state is not only a technological feat. It also relies on a highly labor-intensive organization. Over the past eight decades, the CCP has constructed a vast network of millions of informers and spies whose often unpaid work has been critical to the regime’s survival. It is these men and women, more than cameras or artificial intelligence, that have allowed Beijing to suppress dissent. Without a network of this size, the system could not function. This means that, despite the party’s best efforts, the Chinese security apparatus is impossible to export. 

The CCP’s state security system has worked well for China. But as the country faces unprecedented economic headwinds, the apparatus will become subject to new pressures and strains. The party-state may find it harder not only to maintain its technological stranglehold but also to rely on the involvement of civilian informants who act as the lifeblood of its surveillance regime.

STRONG, BUT NOT TOO STRONG

China has two main domestic security agencies. There is the Ministry of State Security, which is responsible for external espionage and domestic counterintelligence. It does not spy on Chinese citizens, except when they are suspected of having foreign connections. The Political Security Protection unit in the Ministry of Public Security is in charge of domestic surveillance. The MPS includes specialized units, as well as frontline beat cops. There is a clear division of labor between the MSS and the MPS, and they recruit informants separately. Beijing does not publish up-to-date information on the MSS, although it did disclose a decade ago that the total number of uniformed policemen was around two million. Today it is likely to be moderately higher because of a bigger domestic security budget.

Building a surveillance apparatus is a complex task for an autocratic regime. Generations of Chinese leaders have struck a delicate balance between making the secret police powerful enough to do its job, but not so powerful that it threatens the regime itself. Although Chinese leaders have succeeded in this task, the resulting security apparatus has not come cheap. In 2022, Beijing spent 1.44 trillion yuan (around $202 billion) on domestic security, a category of expenses covering the regular police, the MSS, the People’s Armed Police, courts, prosecutors, and prisons—which is roughly equivalent to its total defense spending. This figure will likely increase as Beijing expands, upgrades, and maintains its facial recognition programs, Skynet and Sharp Eyes. Neither is cheap: when Sharp Eyes was rolled out in 2016, Beijing spent 300 billion yuan on hardware and installation alone.

To avoid creating a rival to its own power, the CCP distributes surveillance tasks to different units in the security forces and other state-affiliated actors. This organizational arrangement has two distinct advantages. It prevents the formation of a powerful secret police that can control the upward flow of information and become a threat to the party. And it enables the party to benefit from the involvement of state-owned enterprises, universities, and other entities that channel information to the government, without increasing the size of the secret police. This model requires close coordination, so the CCP maintains a political-legal committee that has overall responsibility for domestic security at all levels of the state. The party further controls the surveillance state by placing CCP officials in every security unit and imposing a five-year limit on the tenure of the top security chiefs, including MSS and MPS ministers. In the former Soviet Union and East Germany, no such limits existed; unsurprisingly, secret police chiefs in those regimes amassed enormous power. 

IT’S NOT CHEAP

Funding constraints have long limited the CCP’s ability to maintain a large domestic security force, let alone an elite, well-paid, and well-equipped secret police network. A comparison with the East German Stasi is instructive. Before the fall of the Berlin Wall, in 1989, the Stasi was the largest secret police force in the world in relative terms, with one officer for every 165 East German citizens. Additionally, the Stasi had 189,000 informants (roughly 1.1 percent of the population).The number of China’s political security protection officers responsible for domestic surveillance is classified. But data I collected from a small number of localities, which was likely disclosed by mistake, suggest there are only 60,000 to 100,000 of them, meaning that there is at most one officer for every 14,000 or so Chinese citizens. 

As a consequence, the CCP must rely on its organizational presence in state-affiliated social and economic institutions, as well as local communities, to recruit a large number of informants. These citizens can spy on their colleagues or neighbors, and because their participation is secured by coercion or enticement, it does not cost much to maintain them. Data disclosed by 30 local governments show that between 0.73 percent and 1.1 percent of China’s population—perhaps as many as 15 million people—serve as informants. On top of that number, the regular police and secret police maintain separate networks of remunerated spies and unpaid informers whose exact numbers are classified. The number of informants in communities and workplaces varies according to the needs and discretion of local authorities. In some universities, for example, every classroom has an informant who provides a biweekly or monthly report to a handler who is usually a party functionary. The information or intelligence generated by informants includes updates on the activities of members of banned cults and underground religious groups and on public reaction to major government policy or political events.

Analysis of the information generated by informants from a small sample of jurisdictions shows that only about 40 percent of the informants are active. Nevertheless, the awareness that fellow students, co-workers, and neighbors could be spies is likely to deter ordinary people from taking part in activities or speech that could get them into trouble.

THE OLD METHODS

Chinese surveillance systems, which were established long before the advent of advanced surveillance technologies, began as labor-intensive systems—and still are. The MSS and MPS’s most prevalent tactic remains what the party calls “controlling battlefield positions,” or monitoring critical public venues (airports, railway stations, and hotels) and social institutions (in particular, universities and Tibetan Buddhist monasteries) for activities that pose threats to public safety and regime security. Typically, controlling battlefield positions involves frequent in-person inspection of compliance by the frontline police to ensure that the staff of these public venues are keeping records and reporting their customers’ identities and transactions. For example, the cyberpolice frequently inspect Internet cafes to confirm that the proprietors are recording their customers’ ID cards. Local data suggest that about 40 percent of the “special intelligence personnel” recruited by the police are assigned battlefield-position duties. Technology, such as digitized information system and video cameras, plays a complementary but largely secondary role in this.

Another effective surveillance tactic is intimidating and monitoring people the CCP classifies as “key individuals,” or KIs. These are members of banned cults and religious groups, petitioners, protesters, people with mental illnesses, and drug users. Police conduct “door knock” operations to verify KIs’ physical whereabouts, to intimidate them, and to warn them against taking part in undesirable activities. Another method is to form a team of five people—usually a beat cop, a neighborhood committee official, a representative of the targeted individual’s employer, a family member of the target, and an informant in physical proximity to the target—to keep a close eye on potential troublemakers, especially repeat petitioners who may stage protests or travel to Beijing to embarrass local authorities.

Chinese surveillance systems began as labor-intensive systems—and still are. 

These two tactics are highly labor-intensive, which is also true of China’s two mass surveillance programs. One is operated by the police and the other by local authorities. Data from dozens of local jurisdictions suggest that the police program monitors somewhere between 3.4 million and 5.0 million individuals, mostly ex-convicts and criminal suspects, and that the other keeps tabs on somewhere between 3.9 million and 7.7 million people. Although a breakdown of this targeted population is not available, it appears that many of those under surveillance are viewed as potential protesters. In particular, they appear to be former People’s Liberation Army soldiers, repeat petitioners, members of ethnic minorities, or adherents of groups that the CCP considers cults.

Advanced technology has significantly improved the state’s ability to track blacklisted individuals. The MPS maintains a digitized national database of KIs to which local police authorities have access, meaning that these people are tracked both physically and electronically. When a KI travels, a system that captures mobile phone information can trigger automatic alerts. The online activities of individuals labeled “Internet KIs” are closely monitored by China’s cyberpolice, again thanks to advanced but undisclosed technical means.

BEHIND THE CAMERAS

Fancy tech tools notwithstanding, it is the organizational capabilities of China’s Leninist party-state that allow its surveillance to function with unrivaled potency. This model, then, is not exportable to less organized autocracies with a shallow reach into their societies and economies. Such countries can import Chinese hardware. But they cannot import Beijing’s system, meaning that they are unlikely to develop equivalent surveillance capabilities.

China’s surveillance state may have helped prevent the emergence of organized opposition to the CCP in the post-Tiananmen era, but political stability was also a product of the country’s generally high levels of economic growth. The Chinese surveillance state remains untested in a less benign economic environment. But such an environment is emerging: the real estate bubble has burst, putting pressure on local government budgets that formerly relied on receipts from land sales to fund their operations, and roughly one in five young people are now unemployed. These economic problems will make it harder for Beijing to handle the spiraling costs of maintaining and upgrading its high-tech surveillance equipment. This may be a particular problem for the Skynet and Sharp Eyes projects, which are funded by debt-ridden local governments and are therefore likely to experience mounting challenges in the lean years ahead.

The impact of a protracted economic malaise is likely to create three problems for China’s labor-intensive surveillance apparatus. First, unemployment, falling incomes, and waning opportunities will fuel social discontent, increasing the burden on the security services as more people engage in activities that the party deems threatening. Second, the rising discontent could make it more difficult to recruit new informants. Third, if incomes are stagnating, informants may demand greater compensation, raising the cost of gathering intelligence. 

A faltering surveillance state will present Chinese leaders with several options. Beijing may decide to treat its citizens better in the hope that doing so will limit protests. Alternatively, the country’s leaders may demand less of the security apparatus, which currently targets many people who pose little or no real threat to the party’s rule. The likeliest outcome, however, is that when soft repression through surveillance fails, the party will begin to use hard repression. The surveillance state, then, may be replaced by something much worse.