Impactism
日本語 Join Impactism
Manifesto The Impact Standard Foundations Objections Read For organizations 日本語
Join Impactism
A security camera mounted on a brick wall
Photo by Andrey Strizhkov on Unsplash

Essay

This Is Not a Social Credit Score

June 21, 202611 min read

The fear arrives before we finish the sentence. We say we want a society that honors impact — that makes impact, not money, the measure of a life — and a particular image forms in the listener’s mind, fast and vivid. A number floating beside every face. A score that rises when you please the right people and falls when you don’t. A queue you can be thrown to the back of. A Nosedive: the Black Mirror episode where everyone rates everyone on a five-star scale, and the cumulative rating quietly gates your housing, your flights, your friendships, until a woman drowning in good manners loses everything for the crime of being briefly unpopular (Brooker, Schur & Jones 2016).

We have heard this fear in every room where we have made the argument. It deserves more than reassurance. It deserves to be built at full strength — and then answered structurally, not with a promise that we are the good kind of people who would never. Good intentions are not a safeguard. Architecture is. So we will build the dystopia as real and as frightening as it actually is, and then show you, property by property, why the unit we are proposing cannot become it. Not would not. Cannot — because the failure modes that produce the nightmare are exactly the design choices we have forbidden ourselves.

The fear, at full strength

Start with the cultural image, because it is doing most of the work. Nosedive aired in 2016 and lodged itself in the public imagination because it felt close. Everyone already lived inside something rated — drivers, sellers, hosts, a follower count, a like. The episode simply extended the line: what if there were one number, visible to all, and it governed access to everything? The horror is not the rating. It is the totality and the convertibility — a single score, applied everywhere, that converts directly into whether you get the apartment. That is the picture the word “impact score” summons, and we will not pretend it doesn’t.

Then there is the real-world referent the fear reaches for: China’s Social Credit System. Here we have to be more careful than the fear usually is, because the popular version is largely a myth — and getting it wrong would let us answer a strawman instead of the real thing. The scholarship is clear. Victoria Adelmant of NYU’s Center for Human Rights and Global Justice puts it plainly: “Social Credit is imagined as a comprehensive, nation-wide system in which every action is monitored and a single score is assigned to each individual, much like a Black Mirror episode. This is in fact quite far from reality” (Adelmant 2021). Rogier Creemers of Leiden, who reads the primary documents, describes it not as one score but as “an ecosystem of initiatives at the central and local level, operated by state and private actors,” and notes that the consequential parts run on blacklists, not algorithms: “there is no numerical scoring involved with these blacklists. You are either on or off the list” (Creemers 2018). Jeremy Daum of Yale’s Paul Tsai China Center, who has spent years correcting the record, says the citizen-score story “simply wasn’t happening,” and adds the sharpest line of all: “if you made a single unifying score out of a holistic data set, you’re turning big data into small data” (Daum 2022).

This matters, and not because it makes the system harmless — it doesn’t; the blacklists are real and they bite. It matters because of where the genuine danger sits. The thing people fear — the unified, total, convertible, state-run citizen score — mostly does not exist even in the country that is the byword for it. What exists is messier and, in its real components, still coercive. So the honest steelman is not “China built the Nosedive score.” It is: the moment any society does build a single, total, state-run, convertible score, here is what goes wrong — and the failure modes are well understood, because thinkers mapped them long before the technology arrived.

The deepest of those thinkers is James C. Scott. In Seeing Like a State he gave us the word legibility: the way a central authority, in order to tax and conscript and control, forces the messy diversity of human life into a uniform grid it can read from above — permanent surnames, standard measures, the cadastral map (Scott 1998). Scott’s warning is that the act of making people legible is never neutral. The grid erases the local knowledge it cannot see, and it hands whoever holds the grid an instrument of control that did not exist before the measuring started. A society-wide score is legibility’s purest form: it takes the whole of a person — every loyalty, kindness, dissent, and failure — and renders it as one readable number, sitting in one authority’s hands. Scott’s whole book is the argument that this ends badly even when the people drawing the grid mean well.

And a total score does not even have to be used to do harm. Its mere existence changes behavior, because people manage themselves toward whatever is being watched. We can measure this. After the 2013 Snowden revelations exposed the scale of NSA surveillance, Jonathon Penney documented a roughly 20 percent drop in page views of Wikipedia articles on terrorism-related topics — al Qaeda, car bomb, Taliban — as people quietly avoided looking suspicious to a watcher they could not see (Penney 2016). That is the chilling effect, and it is the subtlest horror of the Nosedive world: not the punishment, but the pre-emptive self-editing of everyone who knows the number is watching. A score you cannot escape becomes a script you cannot deviate from.

Put the pieces together and the fear is not naive. It is a coherent, evidence-backed account of how honoring conduct could curdle into control: make it one number (legibility), make it total (Scott’s grid), make it state-run (one authority holds it), make it convertible (it buys your apartment, like Nosedive), and make it inescapable (the chilling effect does the rest). Western democracies took this seriously enough to legislate against it: the EU’s AI Act, in force since February 2025, names government and private “social scoring” as a prohibited practice — an unacceptable risk, banned outright (European Union 2024). The world has already decided that a total behavioral score is a line you do not cross.

So have we. The question is whether what we are building is on the wrong side of that line. We think the honest answer is no — and the reason is in the structure of the unit, not the purity of our motives.

Why our unit cannot become that score

We want to be exact here, because “trust us, it’s different” is precisely the answer that should never be trusted. The unit of recognition we are proposing has five properties. Each one is not a nicety. Each one is the direct negation of a specific failure mode above. Take any property away and the fear becomes legitimate again — which is why we treat them as non-negotiable, the way an engineer treats a load-bearing wall.

It is peer-conferred, not central. The single most dangerous feature of the Nosedive score and the Scott grid is that one authority holds the measuring instrument. Whoever defines the score defines the person. Our unit has no such center. It is conferred horizontally — by peers, colleagues, neighbours, the people close enough to a contribution to witness it — never issued by a state or a platform from above. There is no central ledger to capture, no single dial to turn, no office that decides what counts. This is the property the surveillance fear is really about, and it is the first thing we refuse. You cannot build Scott’s grid out of a thousand small, overlapping acts of one person seeing another and saying that mattered. There is no grid; there is a web, and no one sits at the middle of it.

It is non-convertible — carried, never cashed. This is the property that does the most work, so we will be blunt about it. The Nosedive horror is convertibility: the rating turns directly into the apartment. The instant standing converts into a spendable balance, two things happen at once. It becomes a price — and a price corrupts the very generosity it tries to reward, which is the argument we make at length in Honor Is Not a Price. And it becomes a penalty — because anything that can be spent up can be docked down, and a convertible score is one policy change away from being a leash. Our unit cannot be spent, cashed, transferred, or deducted. It is carried, the way a reputation for courage or honesty is carried: real, consequential to how people treat you, and yet not a balance anyone administers. You cannot be fined in it. You cannot buy with it. A thing that converts into nothing cannot be weaponized into a punishment.

It decays. There is no permanent dossier. The Scott grid and the social-credit blacklist share a quiet cruelty: they remember. The record is permanent; a person becomes their worst entry, forever. Our unit forgets on purpose. Recognition is a living signal of recent contribution, not an accumulating file — it fades unless it is renewed by new acts, which means standing has to be re-earned, not banked, and a low moment is not a life sentence. Decay is what makes a system forgiving. It is also, not incidentally, what makes it useless as an instrument of control: you cannot hold a permanent dossier over someone if there is no permanent dossier to hold.

It is plural. There is never one global score. The whole nightmare depends on the word one. One number, one authority, one scale that ranks all of humanity from top to bottom. We have written this into our foundations as a hard constraint: never build a single global score, because that simply re-creates money’s single-yardstick tyranny in a new currency. Recognition lives inside many communities — a thousand contexts, each with its own sense of what is worth honoring, none of them commensurable into a grand ranking. The artist honored among artists, the carer honored among carers, the quiet organizer honored in her neighbourhood: these are not points on one leaderboard, because there is no leaderboard. Plurality is not a softening of the system. It is the refusal of the exact thing — the single ranking — that makes Nosedive Nosedive.

It floats above a universal income floor. It is never a leash on survival. This is the property the fear most often misses, and it may be the most important. Nosedive is terrifying because the score controls access to life itself — housing, transport, the means of getting by. A score you must keep up to eat is a score you cannot say no to; that is what makes it a leash. Our proposal is explicit, and explicitly Phase-2: the impact yardstick floats above a floor where survival is already met — the abundance premise where a high income floor means no one goes hungry or unhoused regardless of their standing. When your life does not depend on your recognition, recognition stops being a threat and becomes something else entirely — a game of meaning played by people who are already safe. The floor is the firewall between honor and coercion. It is the difference between do this or you cannot live and do this because it is worth doing.

Read the two lists against each other and the structure is plain. The fear is built from five choices: central, convertible, permanent, singular, survival-gating. The unit is built from their five opposites: peer-conferred, non-convertible, decaying, plural, floor-protected. This is not a coincidence and it is not rhetoric. We derived the properties from the failure modes. The dystopia is the specification, run in reverse.

What we will not claim

We would be doing the very thing we accuse the fear-mongers of — telling you only the flattering half — if we stopped there and declared the danger zero. It is not zero. Honesty requires naming what survives.

Soft social pressure survives. A community that visibly honors some things communicates, by the same act, that it values them — and people respond to what their community values, sometimes by conforming. We would rather a world that pulls people gently toward contribution than one that pulls them toward attention and consumption, which is the pull we have now; but “gentler and better-aimed” is not “absent,” and we will not pretend the nudge disappears just because no one is keeping score. A recognition culture is still a culture, with a culture’s power to make people want to belong.

In-group conformity survives, and can be ugly. Plurality is our safeguard against a single tyrannical ranking, but it buys that safety by accepting many small communities — and small communities can be parochial, cliquish, and cruel to the outsider in their own way. A web of local esteem is not automatically just. It can honor the wrong things, miss the quiet person, reward the performer over the worker. We catalogue these failure modes of honor itself — inflation, error, the slide back toward a price — honestly and at length on The Impact Standard and Objections, because they are the real residual risks, the ones that remain after the architecture has done its job. The architecture defeats the Nosedive score. It does not defeat ordinary human smallness, and we do not claim it does.

And we will say the largest caveat loudest: none of this has been built. This is a proposal. No society has run a standing, decaying, plural, peer-conferred, non-convertible recognition system above a universal floor, and watched what it does over a generation. We are arguing that the design escapes the dystopia by construction — and we believe the argument — but a design is not a result. The thing we are most certain of is the negative claim, and it is the one that matters most for the fear in the room: whatever this becomes, it cannot become the single, total, convertible, state-held score, because we have removed, by hand, every part required to build one.

That is the whole answer. The fear is right that a society which measures the wrong thing, the wrong way, in one place, forever, would be a cage. We have read that warning more carefully than its authors, and built its opposite. A social credit score is one authority, one number, one permanent ledger, convertible into your survival. Recognition, as we mean it, is many witnesses, no ledger, fading on purpose, worth nothing you can spend, floating safely above the floor of a life. The first is a grid you are caught in. The second is a culture that simply refuses to let goodness go unseen.

Make impact, not money, the measure of a life. Confer it, carry it — and never let it become a score you can be caught in.

Sources

← All essays