A contract awarded competitively at one value and quietly grown to many times that value through variations has escaped the competition that justified it. This is one of the recurring findings of audit offices, and one of the hardest things for the public to check independently.

Contract notices carry amendment records, so a determined analyst can rebuild the history. It requires real data work rather than a lookup, which puts it out of reach of the journalists and councillors most likely to want it.

A useful boundary case for the schema. The obstruction is not secrecy -- the information is fully published. It is that the published shape answers a different question from the one people ask. Recording it as derived rather than yes keeps that distinction visible.

Checked on 2026-09-02, and the verdict of derived holds -- but the entry was understating the material available and has been corrected upward.

Amendments are not merely embedded in notices to be teased out by hand. The Open Contracting Data Standard release derived from AusTender covers over 517,000 releases describing more than 410,000 contracts and 107,165 contract amendments, historical contract data runs back to 1999, weekly export files are published, and AusTender has since added a dedicated Contract Notice and Amendments Published Report specifically to reflect contract values alongside their amendments.

So the practical route exists and now has a name. derived remains right, because a contract history still has to be assembled and interpreted rather than looked up, but the distance between the published data and the answer is much shorter than "reconstructing from amendment notices" suggested.

The register should record improvements as readily as failures. Adding an amendments report is exactly the kind of change this entry was implicitly asking for, and it happened.