UC Berkeley

Over the last eight years, China’s government has implemented its social credit system, a project to eliminate societal behavior the state has deemed to be ‘untrustworthy.’ Actions spanning tax evasion, academic plagiarism, and food safety scandals, among a host of others, fall within this system’s purview. The central government reduced these issues into an information asymmetry problem, positing that the appropriate response was to publish information about, shame, and penalize perpetrators. This in turn created market opportunities for the government to partner with both large and small Chinese tech firms who modified their platforms and crafted bespoke technological solutions to advance a state-defined vision of building social trust. This dissertation investigates how Chinese tech firms collaborated with national and local government institutions to implement plans for the country’s social credit system. It explicates how major tech companies differed from smaller tech firms in the roles taken on, strategic compliance, services provided, and benefits accrued when both types of companies partnered with the state. Using qualitative methods, this work draws on Dan Breznitz and Michael Murphree’s framing of “structured uncertainty” to explain how tech companies navigated ill-defined policy goals and limited guidance from the state. Contrary to portrayals of the social credit system as a straightforward, top-down vision the state tasked tech companies with building, this research shows how the firms themselves had a hand in co-producing the system’s infrastructure while balancing their own interests and perceived constraints. The dissertation argues that major tech companies operated as what Chinese legal scholar Shen Kui terms “credit norm entrepreneurs” for having designed platform features that state institutions sought to replicate in rewarding citizens ‘trustworthy’ behavior. It also provides a case study of how three large tech companies operating in the financial technology, ride-hailing, and food delivery sub-sectors developed at least four strategies for negotiating how their data was used in the social credit system. These included resisting orders from government bodies they deemed to be insufficiently authoritative, and experimenting with internal user rating systems to preemptively comply with impending laws. Moreover, rather than indiscriminately handing state institutions user data, companies developed strategies such as sharing aggregated data about ride-hailing passengers’ unpaid debts, and designing narrowly tailored data-sharing channels for state monitoring of food safety incidents. Analysis of how an e-commerce firm and news app worked with courts across China to evaluate, penalize, and socially shame court judgment defaulters additionally demonstrates how state-firm partnerships created business opportunities for firms when giving them government-generated data. Hundreds of smaller tech firms functioned as ‘intermediaries’ in the market for social credit technology. Through the government procurement process, they won contracts to build the hardware and software of province- and city-level databases and scoring platforms. Using structured uncertainty to their advantage, they filled gaps in local governments’ knowledge through designing consultations, training sessions for local officials, and analysis and translation of social credit-related policy documents to educate the public. The combined effect of both types of firms’ tightly scoped participation in building the social credit system reveals how the system’s reliance on over-technological solutionism ultimately made it ineffective at its own purported goals of strengthening social trust. In providing an assessment of the social credit system’s outcomes, this dissertation argues that what began as an ambitious project to reshape citizens’ behavior has ultimately amounted to a form of what political scientist Iza Ding refers to as “performative governance.” The central government’s narrative framing of social credit as an information asymmetry problem in need of technological solutions potentially diverted attention away from how the state and tech industry have eroded other forms of social trust in several of the same fields where tech companies provided rewards. This dynamic arises from “management-oriented governance,” wherein the state defines citizens’ needs without soliciting their input. It also extends Ding’s theorization of performative governance in China as a tool that primarily low-resourced bureaucrats use to assuage, rather than fulfill, the public’s demands. This research shows how higher-resourced Chinese courts as well as national and municipal government bodies partnered with technology firms to present a particular facade of good governance, yielding superficial returns on their efforts. The dissertation concludes with reflections on how features of state-firm co-production of the social credit system may inform similar initiatives to implement technology for governance purposes. Government procurement of technology, the use of government-generated data in major tech firms, and the intermediary role of small tech firms are likely to recur in these projects. Understanding how tech companies and local governments have set boundaries on what they are willing to co-produce with the state can help forecast the effects technological governance will have in China, how these projects might infringe on civil liberties, and how Chinese law may by turns restrict what is implemented, and in other cases be subverted to serve emergent state objectives.