We transform complex legal documents, tribunal decisions and government records into clear explanations anyone can understand — and publish the data underneath.
Deep analysis of public records and decisions.
Complex information explained simply.
Shining light on systems and institutions.
Empowering people with knowledge and context.
Evidence-based. Non-partisan. Fundamentally public.
Of 52 published decisions reviewed, one resulted in a recorded remedy. But published decisions represent only the matters that produced a public legal record. The investigation is not about wins and losses — it is about what becomes visible, and what disappears.
Very few matters reach the stage where merit is publicly assessed. Four of 37 matters reached a merits hearing. The other 33 ended earlier — on deadlines, forms, screening, or silence. Whether those claims had merit was never publicly tested.
Representation shapes who stays visible. No represented applicant in this record was dismissed for missing a deadline; nearly half of unrepresented ones were. A case can end because the facts fell short — or because a person couldn't present them within the procedural framework. The public record often cannot distinguish the two.
Exits are not verdicts. A matter leaving the record may mean confidential settlement, withdrawal, exhaustion, or a missed step — very different things from a claim tested and rejected. Only a merits hearing answers that question, and the record rarely says which happened.
52 published decisions · 37 matters · joined to subsequent court records
Counted by matter, not by decision. Findings describe the public record, not all outcomes.
When a person enters a legal or administrative system, what outcomes become visible — and what outcomes disappear? Public decisions show the cases that reach a formal conclusion. They reveal far less about the cases that exit earlier: through settlement, withdrawal, procedural dismissal, or outcomes that never generate published reasons.
A matter disappearing from the record does not mean the claim lacked merit. It means the question was never publicly answered.
The public record may reveal where cases end — but not always why. Limited public data makes it difficult to evaluate whether a system is functioning fairly, because the visible record may disproportionately represent the cases that survived every procedural hurdle. Closing that gap is a data problem, and data problems can be worked.
Every figure is extracted from the four corners of a published decision and carries a confidence score. Nothing is imputed — a silent field is recorded as silent, never as a zero.
Rates are calculated by matter — one person's case — not by decision. One matter can produce four decisions, and counting decisions would overweight the most litigated files.
Published decisions are a survivorship sample. Matters that settle or are withdrawn produce no public document, so every rate here is a rate among adjudicated matters. Closing that gap is what the FOI Tracker is for.
Tribunal-only datasets have a known failure mode: a tribunal does not annotate its own published decisions when a court later sets them aside. During pre-publication review, that exact trap surfaced in our own pipeline — so every outcome we publish is now joined to subsequent court records before it ships, and a validation rule blocks any figure that isn't.
When we do publish an error, the correction will appear here: dated, permanent, in full, with what changed downstream. A project that audits institutional record-keeping holds itself to the standard it asks for.