Welcome back to the Ethical Reckoner. Today, I wanted to write about China’s Social Credit System (SCS), a topic that I find fascinating from both a China studies and computer science perspective, and the similar systems that are on the rise in America and Europe. Based on my (limited) crowd research, most people either don’t know much about the SCS or think it’s a Black Mirror-esque tool of dystopian, autocratic control (a perception aided by sensationalist headlines full of misinformation).
In reality, the system is not a Nosedive-like panopticon that rates your every move. Rather, it’s a mish-mosh of small(ish)-scale public and private initiatives. The Chinese government likes to use a pilot model of development where the central government sets out guidance, then tasks corporations or local governments to implement them, creating a system that fosters competition and innovation while also preserving the Chinese Communist Party (CCP)’s centralised control. The SCS started out as a consumer credit initiative in the late 1990s intended to stimulate the economy and was run by the People’s Bank of China (PBOC). Essentially, it was the FICO of China. In 2014, though, amidst a narrative of societal moral crisis similar to the discourse around economic crisis, it evolved to go from beyond the promotion of fiscal trustworthiness (征信/zhēngxìn) to general honesty and integrity (诚信/chéngxìn). These words are both captured in the term 信用 (xìnyòng), which is what is translated as “credit” in the Chinese phrase interpreted as “social credit system” (社会信用体系/shèhuì xìnyòng tǐxì) (Knight 2020).**
On the private side, the PBOC gave eight firms permission to establish pilot systems that could lead to official licenses, but in 2017 the PBOC declined to certify any of them because “none of the eight pilot programs were meeting the PBOC’s expectations with respect to fairness in rating, institutional independence, and data privacy and security” (Dai 2018). Some of these have continued independent of the government, most notably Ant Financial’s “Sesame Credit.” These are often presented as being the SCS, but they’re independent, private systems. They essentially function like company loyalty programs, but instead of getting a coupon for every $100 you spend, having a good score means you can get easier access to credit from Ant Financial Group, deposit-free housing, and access to the fast security lane at airports (Creemiers 2018). The latter benefit is a public-private collaboration similar to the TSA-Clear collaboration in the US.
On the public side, “pilot model social credit cities” started being announced in 2015. After official assessments in 2017, 12 model cities were announced in 2018, with 16 cities being added in 2019 (Knight 2020). These systems vary hugely, with only a handful using a points-based system. One of those, in Rongcheng, has published a long list of behaviours that can increase or an individual or organisation’s score, making it relatively transparent. The part of the system dedicated to encouraging honesty in taxi drivers through Uber-like star ratings has decreased passenger complaints by 72% and seen the return of lost property worth over 2 million RMB (Knight 2020).
The part of the SCS that has perhaps garnered the most attention from Western media is the blacklist-redlist system. The blacklist began as a list of people who were not complying with court orders and can bar people from (among other things) entering the civil service and purchasing luxury goods; once they carry out their legal obligations, they are removed from the list, and can appeal being put on the list (Creemiers 2018). In 2018, it expanded to include anti-social behaviours like tax evasion and rail ticket scalping. According to an official report, the blacklist had stopped individuals from purchasing air and rail tickets 23 million times by the end of 2018—and encouraged 3.5 million people to pay the taxes or debts that landed them on the blacklist in the first place. This is the national-level blacklist; it uses data from national banks and courts. Individual cities have their own lists, but data sharing is limited. Recently, focus has shifted from the “stick” of the blacklist to include a “carrot” redlist approach; the new “CreditEase+” program would provide benefits when applying for loans or government services and discounts on travel for “good” individuals, which will be integrated into municipal systems (Knight 2020). These civil rewards mimic the rewards offered by private systems like Sesame Credit.
So, the SCS as it stands is not a single unified system. Rather, it is a collection of disjoint public and private initiatives guided by the government’s overarching goal to improve social trustworthiness, a carrot-and-stick model that has wide support among the Chinese public—over 80% of respondents in one study reported “somewhat approving or strongly approving” SCS systems (Kostka 2019).
With all the fragmented systems, the SCSs (as there is no single SCS) are far more innocuous than the scaremongering Western media would have you believe. It’s important to note that China generally places a higher emphasis on social stability than individual liberty and that this approach is generally supported by the people—hence the SCSs’ high approval rating (Ess 2020). The SCSs rely on publicly available (e.g. court rulings) and voluntarily provided (e.g. taxi ride ratings) data to build social trust—using a particular Chinese definition of trust—and encourage public morality.
Where the SCSs get worrying is when more advanced technology comes in. For example, Nanjing is planning to include jaywalking in the list of antisocial behaviours, with people cited five or more times subject to penalty. Xiangyang, meanwhile, is using facial recognition to identify jaywalkers. Combine the two, and you have the beginnings of an integrated technocratic system that could could involuntarily measure and punish behaviour that previously would go unnoticed. This is concerning from a Western individual liberty perspective, as it’s a violation of individual privacy, and from a Chinese perspective. After the announcement of the jaywalking inclusion, Weibo users started calling for specific standards for what can be included in SCSs to prevent them from being abused. Chinese citizens do value privacy—they’re just generally more permissive of government use of individual data for national security reasons. If plans to centralise the system go through, though, it could be quite dangerous for individual freedom. Suddenly, every innocuous behaviour can be tracked and judged.
Now, why should you, dear reader, care about this? (Sadly, my fascination with it is probably not enough reason.) It’s clear that this system won’t be encroaching on your life anytime soon (since none of my readers are currently in China). I’m not about to tell you that American or European governments are setting up a similar system. What I will say, though, is that this form of “reputation state” is already forming in a much more insidious fashion.
Consider the workplace fitness challenge, where companies offer incentives to employees who take a certain number of steps, or lose a certain amount of weight, or achieve some other “wellness” metric. Often, this relies on data voluntarily provided by employees, either manually or collected from their devices. This is basically the same mechanism used in the Chinese SCSs—encouraging certain behaviours through carrots (perhaps a free Fitbit for the winning team) and sticks (the shame of not participating or failing to hit the goal). This is what Steffan Mau calls “omnimetrics”—the measurement of everything. Think about it. Your phone tells you how many steps you took yesterday, how well you slept, how many likes your Instagram post got—and we use these to judge ourselves and each other. Assigning moral value to numbers measuring our everyday actions is the exact purpose of the SCSs.
This is a key component of surveillance capitalism, a term coined by Shoshana Zuboff. Surveillance capitalism is “a parasitic economic logic in which the production of goods and services is subordinated to a new global architecture of behavioural modification” (Zuboff 2019). In other words, it’s the drive for consumer data by private companies so that they can sell you the things they want you to buy in more and more effective ways. Google, Facebook, Microsoft—all are collecting vast troves of data that they use to get you to click on ads, share posts, and visit the sites they want you to.
This risks developing a Western SCS, one much more insidious than China’s. For example, Acxiom, a provider of personalised information profiles, has data on over half the adult populations of the US and Germany, with over 1500 data points per person, on average. Their data profiles are used by companies to serve more relevant ads and deals. Acxiom classifies every individual into one of seventy socioeconomic clusters, such as “top professionals,” “metro strivers,” and “children first.” These classifications are based on publicly available data and also shockingly personal data, like accessing information on medication or chronic diseases. As the system becomes more and more detailed, they will invisibly contribute to what offers and deals you can get: “Anyone relegated to Acxiom’s lowest consumer category (known in the industry as ‘waste’) will struggle to get a good mobile phone deal, a half-decent health policy or a holiday loan, nor will they find it easy to rid themselves of the stigma afterwards” (Mau 2019, 165). Doesn’t this sound a bit like a blacklist? Tell me which is worse: a system that determines what you can and can’t do based on clearly laid out lists of social and anti-social behaviours, or one that determines what you can do based on data that it’s scraped from your on- and offline activities?
This should make you feel uncomfortable. If you think that you’ll surely be in one of the top classifications or have the best score, well, so did the Capitol rioters who were shocked they were put on the no-fly list after attacking the centre of American democracy. Chinese citizens understand that losses of individual liberty due to data collection will impact them, but they are willing to make the trade-off; Americans (especially white Americans) often support centralised data collection initiatives assuming they will only apply to the racial or ideological “other,” and then are shocked when their own privacy is invaded. Privacy invasion regardless of status is inherent to the “reputation state,” and there are already undertows in the official and unofficial SCSs that can trap you in a “waste” world as a second-class citizen***, struggling to rise to the surface but kept down by the force of your data—which is forever. Your jaywalk at 11pm on a Friday night when no one was around will never be forgotten, nor will the time you missed the office yoga class. As the status drive is digitised and cemented through ever-more-detailed quantification measures, if we’re not careful, we could end up blindly sliding further into an SCS-like system where everything is measured, quantified, and—officially or unofficially—used to judge our character and decide what we do and don’t deserve.
A quick update on the last newsletter:
Amazon booted Parler from their hosting service, while Apple and Google removed it from the App Store, in what initially looked like a PR-motivated skirmish but may actually be the preliminary salvo against the echo palace. This week, Apple and Google both banned Wimkin, a smaller platform claiming “100% uncensored social media.” However, Parler is still hosted by Epik—the same company that hosts Gab, which was dumped by GoDaddy after the man behind the 2018 Pittsburgh synagogue shooting used it to post anti-semitic content. The only entities that Epik seems to deny service to are literal Nazis. Parler also has retained the service of DDoS-Guard, a Russian company providing cybersecurity services. So, while some American companies may be striking against the echo palace, other companies are happy to take over providing electricity and heat; the palace will survive, but perhaps slightly deeper into the woods than it was before. Whether this will keep people from wandering across it or foster even more toxic dialogue, only time will tell.
*Yes, the “West” is a flawed and nebulous concept, but here I’m using it to refer generally to the US and EU/UK.
**It may translate closer to something like “societal trustworthiness system,” but that’s not as catchy.
***Thanks to Dolly Bai for these insights.
Thumbnail generated by DALLE-3 via ChatGPT with the request “Generate an abstract painting that's an abstract representation of "the reputation state" in a palette of yellows and oranges”.
Lazy bibliography:
Creemers, Rogier. China’s Social Credit System: An Evolving Practice of Control. 2018. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3175792
Dai, Xin. Towards a Reputation State: The Social Credit System Project of China. 2018.
Ess, Charles. Digital Media Ethics. 2020.
Knight, Adam. Technologies of Risk and Discipline in China’s Social Credit System. 2020.
Kostka, Genia. China’s social credit systems and public opinion: Explaining high levels of approval. 2019. https://journals.sagepub.com/doi/full/10.1177/1461444819826402
Mau, Steffan. The Metric Society: On the Quantification of the Social. 2019.
Zuboff, Shoshana. The Age of Surveillance Capitalism. 2019.
this is super nice!
Clear, concise and food for thought. I liked that you included an update on the previous Ethical Reckoner. Well done, my dear.