What a Records Register Actually Tells You

Every public institution keeps some version of a records register — a list of what document types it produces and, in principle, which of those are open to request. Reporters tend to treat this as background paperwork, something to glance at once and move past. In our experience it deserves closer reading than that, because it quietly defines the boundary of what a story can be checked against.

An institutional briefing starts here. Before we describe who reports to whom, we work out which of an agency's internal decisions leave a paper trail that's actually accessible, and which live only in meeting minutes that may or may not be published. That distinction shapes how a reporter should plan their requests, weeks before any single record is asked for.

Reporting lines are not always where they look

Organisational charts published by an agency are a starting point, not a conclusion. In practice, a body described as independent may still route its most consequential decisions through a coordinating ministry, and a register entry that looks routine can be the only public trace of that dependency. Briefings we write spend real time on this — tracing where a decision is actually made, not just where the chart says it sits.

• Which records are published on a fixed schedule versus available only on request
• Which decisions require sign-off from a body outside the one you're investigating
• Where named points of contact exist for public information requests

A briefing, not a verdict

We're careful to keep an institutional briefing descriptive. It tells you how a body is structured and what it's obliged to disclose; it doesn't tell you what to conclude from that structure. That judgment belongs to the reporter, closer to the story than any briefing we could write.

The value, most reporters tell us afterward, is less in any single fact and more in having a map before the first request goes out — knowing which office to ask, and roughly how long a reasonable answer should take.