Skip to content
ParamergeParamerge
Stephen Lieberman
Through Paramerge
Contact

PAN Lab example

ProKid (Netherlands)

The colour and the record: a child risk-profiler

A rule-based instrument sorts children under 12 into four colour risk bands from up to twelve years of police records — including children logged only as victims or witnesses — and police controllers decide who to refer to youth care. Modeled on the Dutch police's ProKid. The controllers barely leaned on the colour, so the score was never the operative thing here. The exposure sits upstream, in the record the score is built from: the government's own evaluation found 36% of the flags erroneous or based on irrelevant incidents, and reporting asked whether a wrong entry simply stays in the record to score the next child.

Stylized model of a documented deploymentChild welfare & family services

Open this example in PAN Lab v0.1 to apply pressures and levers and watch what the system does.

What this models

This example runs on the ProKid-class child risk-profiling instrument network: 4 components and 14 pathways between them. Every context in the Lab is a stylized model, never a reconstruction of any actual deployment, and each assumption behind it carries a provenance label.

Evidence base: 4 assumed · 1 published baseline. In the Lab, the shaded evidence band behind each headline readout draws its width from the least-established class below.

  • assumed

    This models the child-risk-profiling pattern documented in the ProKid case file — not a reconstruction of the actual instrument.

  • baseline

    The store→model loop is the strong one here: the colour was computed from up to twelve years of police records, so the record feedback runs at baseline. The pilot found 36% of the flags erroneous or based on irrelevant incidents, and reporting asked whether such entries persist to feed the next score — so the leverage is the memory loop, not the operator override, the inverse of the Allegheny Family Screening Tool (AFST) shape.

  • assumed

    The model→operator pathway is held low: the government evaluation found the colour classification carried little weight in practice — outside Amsterdam the three flagged categories were handled the same way, and case decisions turned on the last incident and family protective factors — so the score anchored the call only weakly.

  • assumed

    Peer pathways are authored on both signs: quality controllers shared and re-checked screening criteria (a modest peer check that was present), while the record-side correction of erroneous entries — the loop reporting asked about — starts closed, this shape's defining absence.

  • assumed

    The documented harm is to children wrongly flagged — including children recorded only as victims or witnesses — and a children's-rights critique of profiling them at all. No ethnic-bias finding specific to this instrument is documented, and none is asserted. The Lab models institutional propagation, not children's outcomes, and estimates no differential harm to served people; that harm is documented in the case file and measured outside any diagram like this one.

What this example does not show

  • The documented harm is to children wrongly flagged — including children recorded only as victims or witnesses — and a children's-rights critique of profiling them at all. The Lab models institutional propagation, not children's outcomes, and estimates no differential harm to served people; no ethnic-bias finding specific to this instrument is documented, and none is asserted. That harm is documented in the case file and measured outside any diagram like this one.

Sources and evidence

What this example rests on, claim by claim. Every entry resolves to the same ledger the Evidence Registry publishes.

  • The Dutch government's own 2011 pilot evaluation of ProKid found that 36% of the tool's red, orange, and yellow child-risk flags (902 of 2,444 over three months across four police regions, rising to 53% in Amsterdam-Amstelland) were system or registration errors or based on irrelevant incidents, and that in none of the four regions was there a well-functioning instrument.

    empirical
    • Government evaluation DSP-groep for the WODC (Abraham, Buysse, Loef & van Dijk), Pilots ProKid Signaleringsinstrument 12- geevalueerd (2011) https://repository.wodc.nl/handle/20.500.12832/1832
    • Investigative Dimitri Tokmetzis (Sargasso), Hoe de politie duizenden risicokinderen produceert (2012) https://sargasso.nl/hoe-de-politie-duizenden-risicokinderen-produceert/

Where this connects

Institutional pressures in this domain

  • Workload surge — Demand outruns staffing; per-case attention shrinks and review becomes triage.
  • Deadline pressure — Statutory or managerial timeliness rules reward fast approval of machine output over slow disagreement.
  • Staff turnover — Experienced skepticism leaves; new staff calibrate their trust on the tool itself.
  • Data & policy drift — The world, the intake process, and the rules change under a system trained on how things used to be — two mechanisms with different remedies: the statistical properties of what the system processes move (concept drift), or the mixture of inputs arriving in deployment differs from the mixture it was trained on (covariate shift).
  • Compliance over substance — Paper controls (sign-offs, checklists) satisfy audits while the behavior they describe erodes.

All of them in context on the Child welfare & family services domain page.

Levers available here and the patterns behind them

Documented case histories