Setting the Record Straight on Social Credit

An exchange

In a recently posted animated experiment, Glenn and I touched on China’s Social Credit System. After we discussed the explainability of AI algorithms, we had the following (unplanned) exchange:

GL: So it strikes me now that something is at stake beyond the immediate effects of applying the algorithm. If we submit, habitually, to being governed by processes that we don't understand ... Cultivating the habit of “Oh, well, it works. I don't know how it works, but it works” is a step into something where you reduce the humanity, you kind of regiment ... We become regimented and almost subordinate...

NP: Right. Do you know about the social credit system in China?

GL: Not details, not details.

NP: Yeah, me neither. But the basic idea [is] it's this all-encompassing system: everything that you do affects the score that you have, and then the score in turn affects what's available to you. And it's, like, everything together. You might not be able to travel to a different city if your score is too low, and the way I understand it, you usually don't really know what affected the score. There's not a whole lot of that explainability coming back to the citizen.

Well, it turns out that, while what I described as the “basic idea” is an accurate representation of much of the Western reporting on the Social Credit System, it is not an accurate representation of how the Social Credit System actually works or how it’s experienced by ordinary people in China.

A TGS listener, astrophysicist Gregory Green, wrote to us and provided a great explanation of what this system is and isn’t. You can find our email exchange below.

As you might expect, I feel a bit of a sting from being wrong in public. But more than that, I’m excited about what we’re building here at Substack: there’s space for experimentation, for expanding the range of the show, and for the TGS community to help us get better. We have the freedom to make mistakes and a mechanism that enables us to learn from them.

Please do keep these emails coming.

---

This post is free and available to the public. To receive early access to TGS episodes, an ad-free podcast feed, Q&As, and other exclusive content and benefits, click below.

---

Dear Prof. Loury,

Before I launch into a correction of your recent discussion of China's Social Credit System, I just want to say that I enjoy the discussions you publish.

In a recent discussion you posted on YouTube, your interlocutor described China's Social Credit System as a score that is computed for every citizen, based on everything they do, and which determines what they can and cannot do.

This has nothing to do with reality. No such system exists in China.

While there is something called the “Social Credit System” in China, it's completely different from the meme that has spread in the West.

The real Social Credit System is actually primarily a regulatory framework for businesses, and it has very little to do with individual people. It's not even a single system, but rather an umbrella term for a whole number of data-sharing systems between different regulatory agencies. The basic idea is that companies that fail to comply with regulations from one government agency are then subject to increased scrutiny from other government agencies. Let's say you run a restaurant and you fail a food safety inspection. The government agency that handles occupational safety might then also decide to take a closer look at your operations. The flipside is that companies that have a good compliance track record are then subject to less scrutiny. That's the idea behind the system, at least: using data to target regulatory enforcement and burdens where needed.

Individuals do not have a “Social Credit Score” and in fact rarely interact with the system at all. The main way that a private citizen could come into contact with the system would be by defaulting on a legal judgment made by a court. For example, if you are ordered by a court to pay a debt, but you refuse to do so, you may then be barred from buying high-speed rail tickets. Again, there is no score involved, and unlike what your interlocutor claimed, it will in fact be obvious to you exactly how you ran afoul of the system.

I don't know how the myth of the Chinese Social Credit score got started, but it spread because it plays right into the image most people in the West have of China (helped along by some news outlets that uncritically propagated the myth). Now it has become “common knowledge” that there's an all-encompassing score that follows everyone around in China. The funny thing is that if you ask people in China about their Social Credit score, they will have no idea what you're talking about and will probably think that you're asking about their financial credit score (which actually exists, after all).

In any case, I just thought that I would write to correct this issue, because the “Social Credit score” is a widespread myth, and I think it gives people in the West a very distorted view of China and helps to poison attitudes towards the country. One can rightly be very critical of many things in the country, but one's view should be grounded in reality.

Regards,

Gregory Green

Astrophysicist

Heidelberg, Germany

P.S.: If you want to see a good discussion about the myth and reality of the Social Credit System, take a look at this discussion organized by the Asia Society. Even the first few minutes are highly enlightening:

---

Hi Gregory,

I'm Glenn's interlocutor in the AI video. Thank you for your correction. I do have to admit (as I think I did in the conversation) that my understanding of the Social Credit score is very superficial.

I wanted to ask you for a follow-up. Consider this quote from a 2019 Wired article:

One city, Rongcheng, gives all residents 1,000 points to start. Authorities make deductions for bad behaviour like traffic violations, and add points for good behaviour such as donating to charity. One regulation Ohlberg recently read specifically addresses stealing electricity. Of course, you'll have to get caught first or be reported by someone else. While facial recognition is infamously used to spot jaywalkers, in some cities it's not so automated, Ohlberg notes.

Private projects, such as Sesame Credit, hoover up all sorts of data on its 400 million customers, from how much time they spend playing video games (that's bad) to whether they're a parent (that's good). That can be shared with other companies. One infamous example is Sesame Credit linking up with the Baihe dating site, so would be partners can judge each other on their looks as well as their social credit score; that system is opt-in.

So far, taking part in both the private and government versions is technically voluntary; in the future, the official social credit system will be mandatory. That said, there's plenty of pressure to take part now. “There are incentives for participating, and disincentives for not participating,” Hoffman notes.

Do you think it's fair to say that, while there is no all-encompassing system as I described it in the exchange with Glenn, there are signs of such a system starting to emerge, perhaps in a less centralized fashion than “the myth of the Social Credit score” would have us believe?

There's also this line from a 2015 BBC article:

That system isn't in place yet. For now, the government is watching how eight Chinese companies issue their own “social credit” scores under state-approved pilot projects.

Do you think this too is a misrepresentation, and these private projects that include versions of a Social Credit score (for the lack of a better term) are just that, private projects, and not a part of the state's research into the viability of such approaches?

Thank you,

Nikita

---

---

Hi Nikita,

This is an article about the Rongcheng system, written by an academic who studies the Social Credit System. A lot of the mythology about the Social Credit System seems to come from the city of Rongcheng's system. Just for context, Rongcheng is a small city, as far as cities in China go (about 750,000 people). It is one of the few places in the country that does actually have some sort of point scoring system. However, as the article I linked to explains, the system is very low-stakes: both the rewards and punishments are fairly insignificant. Most people could ignore it and go about their lives as normal.

Sesame Credit (also called Zhima Credit) is something altogether different, and it has basically nothing to do with the Social Credit System. Sesame Credit is essentially a financial credit score, developed by the tech company Alibaba to facilitate loans and payments in its online marketplace. Your Sesame Credit is determined primarily by your transactions in the Alibaba marketplace. There was some consideration of linking Sesame Credit to the Social Credit System a few years ago, but the government eventually rejected the idea.

Do you think this too is a misrepresentation, and these private projects that include versions of a social credit score (for the lack of a better term) are just that, private projects, and not a part of the state's research into the viability of such approaches?

The BBC article from 2015 is talking about the possibility of linking privately developed financial credit scores (like Sesame Credit) to the Social Credit System. As I understand it, the government eventually rejected the idea of linking these private systems to the Social Credit System.

Do you think it's fair to say that, while there is no all-encompassing system as I described it in the exchange with Glenn, there are signs of such a system starting to emerge, perhaps in a less centralized fashion than "the myth of the Social Credit score" would have us believe?

I do not yet see any signs of an all-encompassing system of social scoring emerging. Based on my interactions with people in China, the Social Credit System plays no role in their lives. A lot of people in the West think that Chinese people are living in some sort of Black Mirror episode, but as far as Social Credit goes, that isn't the case. I actually think that fears by people in the West about the possibility of increased surveillance and use of personal data in their own countries are what drive the myths about Social Credit in China. People are fearful that their own societies will turn into an imagined version of China. Most people in the West are not familiar with China at all, but they vaguely know that it has a lot of government surveillance, so it's easy to turn it into a bogeyman.

If you want to see what Chinese people themselves think of the Social Credit System, you can take a look at these street interviews from Shanghai: 

Perhaps the most surprising thing (from a Western perspective) is that most Chinese people do not even know what the Social Credit System is and assume they're being asked about financial credit scores. When the interviewees are asked about the Western perception of the Social Credit System, they have a range of reactions (from saying it would be good to punish people who misbehave in public in certain ways to expressing concern about the possibility of the government interfering too much in people's lives).

Regards,

Gregory

---

Thank you so much for this, Gregory.

I'm kind of glad that I've said the wrong thing—because now the TGS audience, like me, will be able to read your response and become better informed on this topic.

Nikita

---

Hi Nikita,

Thanks to you as well for posting my response. I wrote in the first place because I know that Glenn is fair enough to make corrections. The topic of the Social Credit System has caused a lot of confusion in the West, so it's good to give a bit of a reality check.

Regards,

Gregory

---