How a rort gets documented, what a grade means, and what happens when we get something wrong. Published so it can be cited, and versioned so a correction to it is visible too.
A rort is a documented mechanism by which something is taken from the public. Not an allegation, not a suspicion, not a person.
The subject is always the device: a reclassification lever, a deemed-refusal drift, a voluntary planning agreement, a tax structure, a sunset clause that keeps moving, a threshold quietly lowered. The people named are the ones the record shows operating it.
Not every rort is denominated in dollars. Of the fifteen cases on this station, twelve carry no money figure at all. A date an obligation took force, a vote that rezoned a suburb, a certificate never issued: each is a finding of exactly the same kind as a sum, and this station presents them with the same weight. Ranking by money would be a claim about this archive that is not true of it.
No case is published without at least one primary source. A primary source is the document itself: legislation, a court record, a filing, a regulator's own publication, an annual report, an official transcript. Reporting about a document is not the document.
The evidence grade and the source tiers below are identical to those used by change4planet.org, the other station in this fleet. A grade must mean the same thing across every site or it means nothing at all.
What is not shared is that registry's Transferability grade. It asks how portable an intervention is to another place, and a rort is not an intervention.
As of version 1.0, every case on this station carries the status ongoing. That is honest: none has closed.
The vocabulary above exists so that when one does close, is reformed, or widens, the record can say so rather than going quiet. A watch station that can only report that something is still happening cannot report the thing it most wants to: that the pressure worked.
On 16 August 2026 this station audited every citation it carries: 646 unique addresses across 748 references.
It found 568 deep links returning 404 because a sentence's full stop had been captured into the address; 23 sources genuinely gone; nine citations naming articles that do not exist, attributed to real publishers with headlines they never printed; and 25 published facts that were false or imprecise, of which 15 have been corrected.
Seven of the nine fabricated citations were in one section — the profiles of people, rather than the investigations built on filings and regulator reports. Sourcing discipline held wherever primary documents existed and collapsed where the subject was a person. That finding is why this document leads with the primary source rule, and why it exists at all.
The station argues for accountability. It had never published how it works. That was the gap.