China’s digital paradox
Beijing’s ambitions in digital technology are increasingly shaping debates internationally, says Vincent Brussee. But only facts provide a solid foundation for good policy – not myths.
Recent European proposals for the regulation of artificial intelligence (AI) included one to prohibit “AI-based social scoring for general purposes done by public authorities.” In making the latter point, the European Commission appeared inspired by Western depictions of China’s Social Credit System (SoCS) as the germ of an AI-driven Orwellian dystopia. But it is useful to remember that China’s SoCS, at least in its current iteration, differs significantly from what many foreign observers make it out to be. It is fragmented and almost entirely driven by humans and low-tech digitalization, not AI.
China’s SoCS stands less for a technological dystopia than for the paradox that is the country’s information and communication technologies sector. On the one hand, Xi Jinping has declared the digitalization of China’s bureaucracy and surveillance apparatus a priority, many government initiatives come flanked by buzzwords such as “AI” or “Internet+,” and some of the world’s most formidable technology companies call China home. On the other hand, fragmentation is a persistent and debilitating feature of the tech landscape.
In the same way, the SoCS remains a key part of Xi’s vision for data-driven governance and is meant to break down information barriers, while remaining encumbered by innumerable “data islands” six years after its construction formally began. Similarly, digital initiatives for grassroots governance still require intensive manual labor for inputting and managing information. Even the ubiquitous Health Code system for Covid-19 prevention – a contact tracing system that uses cellular data to track citizens’ exposure risk – was initially marred by regional differences and fragmentation.
An objective narrative is crucial
It is crucial for European policymakers and observers alike to be aware of this paradox. If they want to invoke Beijing’s use and governance of tech as a benchmark for EU policy, they cannot afford to misrepresent developments in China. Unfortunately, there is a tendency to lump all Beijing’s tech and surveillance initiatives together. While they are right to be concerned about surveillance in China, focusing attention on myths like machine-driven social credit scoring obscures much more invasive surveillance initiatives – Project Sharp Eyes, for example, aims to establish full video surveillance coverage of key public spaces in Chinese cities.
Europeans should also free themselves of the assumption that digitalization initiatives spring into integrated action immediately after Beijing’s top leadership declares them a priority. Different digital initiatives follow their own trajectories and should be viewed as separate puzzle pieces. Some may indeed involve strict surveillance for solely political purposes, but others may target genuine societal issues. Reducing Beijing’s use of technology to nothing but automated surveillance and control diminishes our understanding of what it is up to.
And where it is falling short. Digitalization initiatives regularly fail to live up to their buzzwords. The proverb “the mountains are high, and the emperor is far away” aptly captures one of the reasons for this. China has a huge and complex bureaucracy, with many parts that are used to doing things their own way. Although many digitalization plans date back to the 90s, government data management often did not stretch beyond filing cabinets, fax machines and red stamps until well into the 2010s.
China's extensive bureaucracy is slowing down digitalization
With 90 million Communist Party (CCP) members, some 800 million urbanites living in developed megalopolises and 600 million mostly rural people making do on less than RMB 1000 per month, authorities at every level have to contend with an immense variety of interests – not least within the bureaucracy itself. Although Xi has taken decisive steps to centralize control, getting all cadres in China to move roughly in the same direction remains a mammoth task.
This challenge is made all the greater by the inner workings of the bureaucracy. Hard-wired performance assessments give local cadres many incentives to distort data. They get rewarded for presenting a rosy picture to outsiders – including foreign observers – while the real picture is more gloomy. For example, cities that serve as models for the application of social credit claim to have collected billions of data pieces, but most of it is of low quality or irrelevant.
Many government departments lack the funds to tackle ambitious digitalization plans and mainly employ low-skilled staff who rely on hand-scribbled notes and messaging app WeChat to pass around information. Government cadres also recognize that digitalization does not always serve their interests. Local governments in the past used information asymmetries to strengthen their own position in dealings with Beijing. But digitalization strengthens central authorities by allowing access to source data – and even enables surveillance of local bodies.
There are three factors that determine the pace at which Chinese authorities adopt digital technologies. The first one is political imperative: the greater the threat to regime stability, the more likely all actors will be aligned and funds available to realize quick adoption. Then comes scope: the more factors that need to be tracked and integrated by an initiative, the more difficult actual implementation becomes. The last one is scale: the more regions and departments become involved, the greater the risk of fragmentation becomes.
These factors explain why a complex, cross-sectoral and inter-regional initiative like SoCS remains fragmented and digitized to a low degree. They also explain why one-dimensional and more regional political surveillance initiatives have had fewer problems getting off the ground. Building the technical architecture is not the main problem. Designing them so to suit everyone’s policy interests and getting all cadres to use them consistently is the bigger challenge. The smaller and more coherent the user group, the more manageable the problems.
This challenge may not sound alien to European bureaucrats and IT specialists. But an acute form of “Sinophrenia” – the belief that China is at the same time about to collapse and take over the world – prevents observers from treating Beijing’s tech developments with the same diligence as we would at home. Instead, the default assumption has become that they are all part of a CCP masterplan to cling on to power. When myths about the “Social Credit superscore” started to surface, few stopped to ask why Beijing would want such a thing in the first place. In fact, such a single basket for all societal ills would be so obscure and vague in meaning it becomes utterly worthless – a concern that Chinese policymakers are clearly aware of.
Only once we treat these developments with the right level of diligence, can we start to take away the real lessons and avoid founding policy responses on myths.