As the national system is still under construction and there is no consistent explanation from the Chinese government, there are currently many different views of how the social credit system will work and exactly what it will involve.
We can assume that it will build upon already existing and widespread listing systems. These listings are based on a “rap sheet”, a type of behavior certificate. State authorities will blacklist individuals and companies that have broken existing laws, rules, and local regulations. Conversely, people and businesses who behave particularly well within the system will be put on red lists. Databases with these lists are already publicly available in China today.
For an insight into the current blacklists and social scoring system, consult the National Credit Information Sharing Platform and the National Enterprise Credit Information Publicity System. On this second site, you can look up your own company’s rating by searching for the company’s name or Unified Social Credit Identifier (an 18-digit number) (the site is available in Chinese only). To find out how to look for your company and more, you can browse the Practical Guide to China’s Corporate Social credit system, provided by the German Außenhandelskammer (chamber of foreign trade).
Many media reports assume that these listing practices are closely related to a points system. Points-based scoring systems are already being tested in various projects. For example, since 2015, there has been the Sesame Credit System from Ant Financial, an affiliate of the Alibaba group. Chinese citizens can participate in this system on a voluntary basis. Its first purpose is to assess creditworthiness and calculate a personal credit score based on a points system.
When collecting data, which is also linked to the B2B platform Alibaba.com and the company’s own online auction house Taobao, information about place of residence and work, family status, payment behavior with bills and credit card use is also saved. Political expression and purchasing habits on social media are also recorded.
While the Alibaba group is often compared with Amazon, the Chinese equivalent to Facebook is called Tencent. This company set up its own credit system called Tencent Credit years ago, which is the same as Alibaba’s Sesame credit system. It populates its databases primarily using WeChat, originally a chatting service (like WhatsApp) that was expanded with further services (e.g., the mobile payment system WeChat Pay, comparable to Apple and Google Pay).
A points-based system is also the basis used by the Chinese coastal city of Rongcheng in its social credit system tested since 2014. 670,000 residents already have a social score in their everyday lives, which must be given for loan applications or for potential promotions at work.
Given this and other pilot projects, many experts assume that planned national social credit system roll-out will also be points-based, although this has not yet been officially confirmed by Chinese authorities. The concept would be that all actors begin with an initial score (1000 points). Good behavior raises the score (up to a maximum of 1300 points). Bad behavior will lose points (to a minimum of 600 points). Exemplary actors can hope to be rewarded. People who don’t behave as the State would want them to can, in the worst cases, be blacklisted and recorded on the official “heavily distrusted entities list”.
The social credit systems in place in China so far are based on predefined evaluation criteria that will very likely also play a role in the national system.
Important evaluation factors for individuals are currently credit scores and purchasing habits (online and offline). Activities and rule-breaking, both on social media and in everyday and social life, will be reflected in a person’s score. People who protect public property, campaign for family values or look after their parents or sick family members can expect positive effects on their score. Criminal records and behavior in public transport (e.g., riding without a ticket, smoking in trains, etc.) are also considered in China’s social scoring. In addition, eating habits and evaluations from superiors and landlords will impact the rating.
Special evaluation criteria for businesses include product quality, environmental protection (following/breaking rules on emissions), pricing, licensing, and use of data and data transfers. The level of tax paid and adherence to payment deadlines are also included in the evaluations. Company reviews by social media users are also considered. Even the behavior of employees can affect the rating of the whole company. If there are several branches of a company in China, the evaluation of one of its locations can influence the rest of the company.
However, it remains unclear how these evaluation criteria will be weighted in a national scoring system and whether every infraction will automatically have a negative impact. According to experts on China, such as Jeremy Daum, it is more likely that a serious, relevant, and criminal incident must occur first (fraud, theft, tax evasion), and only then will the negative profile be expanded with more bad behavior.
It also remains unclear for now as to when ratings and blacklists would be updated under the national system. Currently, a blacklist entry for a serious misdemeanor may be kept on record for longer than 5 years, or it could be deleted after 6 months at the earliest. The process can be accelerated if a solution to the underlying problem is quickly sought. Businesses can counter a poor ranking by signing a Credit Rescue Commitment Letter and providing suitable evidence in their defense, which could allow the negative evaluation to be canceled. Furthermore, legal and administrative aid is available.
According to media reports, the technology behind China’s social credit system will involve database networks, digital image, and sound recordings, big data and data mining analyses, and artificial intelligence. The technical analyses will be able to pull data from pre-existing projects and systems, as well as new databases, such as the national digital/electronic surveillance system, Skynet, which should also be able to provide further data. It currently consists of almost 600 million surveillance cameras that, among other abilities, work with facial recognition and, as of recently, partial gait recognition.