Ask For The Annex, Not The Record

I have filed 3,114 Freedom of Information Act requests and litigated nineteen. People ask me what the trick is, and expect an answer about persistence, or lawyers, or knowing somebody.

The trick is grammar.

The eleven-year request

PX-1330 took eleven years and four months. Most of that was my fault.

I asked, in 2015, for "the Sector 41 anomaly log." The agency said, correctly, that it held no record by that name. I appealed. It said the same thing. I appealed again. It said the same thing more slowly.

In 2021 I telephoned a retired disclosure officer named Cordelia Strand, who died last month and whose obituary this newspaper ran on 29 July, and read her my request.

She listened, and then said: "That is not a record. That is an annex to a record. Ask for the annex."

The next request took four years instead of six and produced 211 pages.

Why this is not a technicality

Every large organisation has a filing vocabulary, and it is not the vocabulary the public uses.

You want to know what happened at a meeting. The organisation does not hold "what happened at the meeting." It holds a readout, or a note for file, or a record of decision, or an action log — and if you ask for the wrong one, the honest answer to your request is that it does not exist.

Most people hear "we hold no such record" and conclude they are being stonewalled. Nine times in ten they are being told the truth by somebody who is not permitted to help them ask a better question.

That is a design failure and not a conspiracy, and the difference matters, because you cannot fix a conspiracy with a form.

What I do now

I ask for the retention schedule first. Every agency has one, most publish them, almost nobody reads them, and it is a complete list of the kinds of thing the organisation admits to keeping, with the words it uses for them.

Then I write my request in their nouns.

It is boring. It works. It is the single most useful thing I know and it took me six years to learn it because I spent those six years assuming the problem was that somebody was hiding something.

The part I still get wrong

I still write requests that are too wide, because a wide request feels braver.

A wide request produces a fee estimate, a fee estimate produces a negotiation, and a negotiation produces a delay measured in years. Every reporter I know does this and every disclosure officer I have ever met has told me so, kindly, more than once.

Strand's advice on that, the last time I spoke to her: "Ask for the smallest thing that would prove you right. If you get it, ask for the next smallest thing."

She then told me to ask for the withheld pages of PX-1330 separately, because they would have been indexed separately.

That appeal is pending. She was almost certainly right.

The Reading Room appears fortnightly on the opinion pages.