Assessing China’s “National Model” Social Credit System
Researchers study a city-level social credit system known as a “national model” that tracks the widest scope of behavior of any such system in China
INSIGHTS
A 2019 “national model” social credit system scores people in a city of 1 million using 389 rules — 124 reward “good” behavior and 265 punish “bad” behavior.
66% of offenses (and 83% of the most severe offenses) already fall under established laws and regulations. Others expand local government authority into moral and social domains beyond the law.
The system also allows local authorities to act more flexibly than the law, adapting the scoring system to shifting priorities like zero COVID policies or other campaigns.
The system rewards and surveils government employees more intensely than ordinary citizens, reflecting the state’s effort to discipline both the public and its own agents.
The system’s scoring gives urban residents and government staff greater access to rewards, while rural residents face more penalties and fewer reward opportunities.
Source Publication: Chuncheng Liu and Akos Rona-Tas (2024). Trusting by Numbers: An Analysis of a Chinese Social Credit System Governance Infrastructure. Critical Sociology.
According to official documents, China’s social credit systems are designed to augment social governance by promoting order, encouraging “civilized” behavior, and expanding government oversight where legal tools fall short. Rather than a single unified system, China’s social credit systems are comprised of multiple experiments run by different agencies. These can be categorized into four types: (1) blacklists for serious legal violations; (2) a government-run financial credit report system; (3) commercial credit systems from tech firms that use algorithms to predict creditworthiness; and (4) city government-run systems that collect data on behavior and combine rewards and punishments. This study focuses on the fourth type. Out of about 330 cities in China, 19 have such local systems. The researchers examine one, anonymized as “Meritown” social credit system (SCS), which is considered a national model and tracks the widest scope of behavior of any such system in China. How does the system score people? Who is targeted by the system’s scoring rules? How does the system impact social groups differently?
The data. Meritown is a county-level city in northeast China with a population of one million. The primary data source for the analysis is the 2019 Regulation of Meritown Society Members’ Social Credit and Evaluation, which defines the rules and scoring metric of the system. The authors also analyzed 45 additional Meritown policy documents referencing credit (xinyong) and supplemented this with fieldwork conducted between 2019 and 2022, including participatory observation in three communities and over 100 interviews with residents and officials. They use the Ministry of Justice of the People’s Republic of China Law and Regulation database and the China Law and Regulation Database to determine whether a Meritown SCS indicator was already legally sanctioned.
Points, punishments, and perks. The Meritown SCS is run by a local government office under the dual supervision of the local CCP secretary and the mayor. The system gives every adult in Meritown a social credit score tied to their national ID, starting at 1,000 points (level A). It draws data from agencies like traffic police and banks and relies on village and residential committee staff to report everyday behaviors, such as trash disposal. The system scores people using 389 rules — 124 reward good behavior, 265 punish bad — and classifies them into eight levels (AAA to D). Perks like heating bill discounts go to top scorers (AAA), while those rated D face police monitoring. Scores change over time and can be appealed online. Researchers grouped punishable actions into six types: political, economic, disorderly conduct, moral and ethical, defying authority, and unspecified punishment. The researchers grouped positive behaviors into five areas: donations, volunteer work, governmental work, social honor, and reporting problems.
Excerpt from the 389-item Meritown SCS scoring measure
Seventy-one “severe” offenses instantly lower one’s letter classification (e.g., a drunk driving conviction automatically reclassifies an AAA person to a B). While political behavior is not the most common type of penalty overall, it makes up the largest share (25%) of these severe offenses, followed by economic (20%), unspecified (20%), defying authority (15%), disorderly conduct (10%), and moral or ethical issues (10%). Notably, 67% of the political offenses involve petitioning local government — an activity legally allowed but often discouraged by local officials. Unlike the legal system, which looks at intent and context, very different actions are treated as equal. For example, 300 hours of volunteer work (+50) can cancel out abusing a family member (–50).
Punished and rewarded behaviors by population type
When policy shifts, so can the points. The Meritown SCS largely reinforces existing legal and regulatory frameworks — 83% of severe behaviors triggering letter reclassification and 66% of those affecting scores already fall under established legal sanctions. However, the system also allows local authorities to act more flexibly than the law, adapting the scoring system to shifting policy priorities. For instance, the number of volunteer-related rewards increased when the city promoted a civic engagement campaign, and during the COVID pandemic new rules were rapidly added to penalize behaviors like not wearing a mask (–10) and reward participation in disease control efforts (+100).
Officials get more perks but extra scrutiny. The Meritown SCS primarily targets the general population, with 81% of reward and 67% of penalty indicators applying to all people. However, government and military personnel face significantly more detailed and moralized scrutiny, along with more opportunities for rewards, followed by business owners and professionals. For example, public school teachers are penalized for moonlighting as private tutors (–20), and company heads can lose points if their firms go bankrupt (–50). Government staff are evaluated with greater granularity, and even minor behaviors, like playing cards at work or showing a bad attitude toward clients, are penalized. Public servants are also required to earn bonus points annually (since 2020, exclusively through volunteering) to maintain promotion eligibility.
Government workers are eligible for more reward behaviors to reach credit levels like AAA, while ordinary citizens mostly depend on low-value rewards like 2 points per 10 hours of volunteering. Rural residents face 28% of penalties (vs. 12% for urban residents), but get only 6% of reward opportunities (vs. 48% for urban residents).
Disciplining society and the state. In their analysis, the researchers challenge simplistic portrayals of China’s social credit systems as uniform tools of authoritarian control. They nevertheless conclude Meritown’s SCS reflects the discretionary judgments and priorities of local officials, often through vaguely defined indicators that allow for strategic ambiguity and administrative flexibility. This ambiguity, while enabling adaptable enforcement, raises concerns about legality, legitimacy, and bias. The system’s scoring method tends to disadvantage already marginalized groups, particularly rural residents. At the same time, government employees experience both privileged opportunities for credit accumulation and heightened internal surveillance, reflecting the dual role of the state in disciplining society and itself.