Social & Behavioral Scoring — The Control-Grid Component

Behavioral scoring is the layer that turns a record into a gate. Once enough of a person’s conduct is logged, it can be compressed into a number — creditworthiness, risk, trustworthiness — and that number can decide what they may access: a loan, a flight, pretrial release, a job. The mythologized version is China’s unified “social credit score,” which never fully existed as one number; the real version, everywhere, is a patchwork of scores that gate real outcomes with little explanation and less appeal. Scored in the Convergence Index.

China

China’s actually-deployed system is the court-enforcement blacklist, not a single citizen score. People judged “untrustworthy” — overwhelmingly court-judgment defaulters — are placed on a list that blocks plane and high-speed-rail tickets, private-school enrollment, and more. The travel bans are not inference from the score mythology — they are two 2018 NDRC opinions, one for civil aircraft and one for trains, that say so in terms (NDRC — Opinions on Appropriately Limiting Specified Seriously Untrustworthy Persons from Taking Civil Aircraft; NDRC — Opinions on Appropriately Restricting Certain Seriously Untrustworthy Persons from Riding Trains); the authoritative primary-source library they sit in is Jeremy Daum’s China Law Translate (China Law Translate). It is real, it gates concrete access, and it operates with administrative discretion and minimal judicial check. The unified-score mythology overstates the integration; the consequences understate nothing.

European Union

The EU runs the most consequential scoring through credit and ESG — and the courts have started to bite. Germany’s SCHUFA assigns credit scores that gate housing and loans; the Court of Justice ruled in C-634/21 that fully automated SCHUFA scoring can constitute a prohibited automated decision under the GDPR (CJEU). In parallel, ESG scoring is being formalized into corporate access-to-capital through the Corporate Sustainability Reporting Directive (CSRD) — a behavioral-scoring layer aimed at firms rather than individuals. Capability present, but with the strongest legal limit of the four.

India

India is scaling consumer scoring fast off the identity-and-payments rail. Credit bureaus — CIBIL, Experian and peers — are expanding coverage to hundreds of millions newly visible through Aadhaar-linked accounts and UPI history, under RBI oversight (RBI). The score follows the rail: as India Stack brings the formerly cash-only into the formal system, it also brings them into the scored system. Broad and growing; the limits are the ordinary (contestable) credit-regulation kind.

United States

The US is the mature private-scoring market and the cautionary tale on algorithmic risk. FICO scores gate credit at population scale, with newer models like Score 10T folding in more behavioral data (FICO); and in the justice system, COMPAS risk scores have been used in pretrial and sentencing decisions — ProPublica’s analysis documented racial disparities in its error rates, the canonical case of an opaque score gating liberty (ProPublica). Privatized and fragmented, but the outcomes it gates — credit, a job, pretrial detention — are as consequential as any state score, with the trade-secret defense often blocking the appeal. That defense has already been tested and upheld: in State v. Loomis the Wisconsin Supreme Court let a COMPAS-informed sentence stand even though the defendant could not examine how the score was produced, because the algorithm was the vendor’s property (Harvard Law Review — State v. Loomis).

United Kingdom

The UK parallel to the Dutch SyRI / toeslagenaffaire: a DWP machine-learning system risk-scoring Universal Credit advance claims, whose own February 2024 fairness analysis found “statistically significant” disparities by age, disability, marital status, and nationality, while the department declined to publish operational detail (Computer Weekly). Unlike the Dutch case — where the District Court of The Hague struck SyRI down outright as an unjustified interference with private life under Article 8 ECHR (District Court of The Hague, ECLI:NL:RBDHA:2020:1878) — it has not been judicially struck. Deployed, contested, not yet checked: sitting between the Netherlands (checked) and Hungary (unchecked) on the rule-of-law-limits axis.

The counter-argument

Scoring is not inherently illegitimate: actuarial credit assessment expanded lending to people redlining had excluded, and a transparent, contestable, narrowly-purposed score can be fairer than a loan officer’s gut. The defect is the recurring one — opacity, scope creep, and no appeal. A score becomes a control-grid component precisely when you cannot see how it was computed, it is used for purposes it was never validated for, and there is no human you can argue your case to. The component scores the gate; transparency and appeal are what the “limit” axis tracks.

Part of the Convergence Index component set. Scored in the interactive index; full cross-country comparison in the convergence table.

Related research

• China body control — the deep primary-source record behind the China summary
• Payment freezing — the off-switch the score triggers · CBDC
• Biometric collection · Digital ID