Somewhere in your council's budget papers there is a line item with an unhelpful name. “Cultural development — other.” “Arts and events support.” A figure, a footnote, and nothing else. That line is a door. Behind it sit grant applications, appraisal notes, venue leases, monitoring reports and contracts, most of them public, most of them unread. Arts funding is one of the few areas where local authorities still write things down in detail, because someone, somewhere, has to justify spending public money on a poetry festival. Your job is to go and read it.

This is not a story about fraud, necessarily. It is a story about decisions: who got the money, what they promised, and whether anyone checked. That is usually enough to make a piece your readers will argue about at the bus stop.

Draw the money map before you ask for anything

Public culture money almost never arrives in one clean stream. There is the local authority, the region, the national arts council, lottery distributors, heritage and film bodies, a tourism board, a university with a civic budget, and often a regeneration fund with a cultural strand bolted on. Venues are subsidised in ways that never appear in a grant register at all: peppercorn rents, maintenance contracts, management agreements, staff secondments, free use of a council-owned building.

Before you request a single document, list every body that could plausibly touch your local arts scene and write down the name of each scheme it runs. Annual reports and published funding decisions will get you most of the way in an afternoon. Reporters who skip this step tend to come back with one grant award and no context, then wonder why the story feels thin.

The five record sets worth pulling first

Requests are cheap. Time is not. Start with the documents that carry the most information per page:

1. Grant award registers and decision notices. Who applied, who won, how much, for what, and — crucially — who sat on the panel.
2. Procurement and contract registers. Anything above the local threshold: event management, marketing, consultancy, feasibility studies for a cultural quarter that has not been built.
3. Property and venue records. Leases, management agreements, rent reviews, and the condition surveys nobody wants to talk about.
4. Charity and company filings. Trustees, directors, accounts, and the related-party notes at the back that describe transactions between people who already know each other.
5. Committee papers. Cabinet reports, scrutiny agendas, and the officer recommendations that explain, in careful prose, why the obvious option was rejected.

Ask for the whole file, not the outcome: the application, the assessment, the award letter and the monitoring return. Access rules differ by country and by body, so check what your jurisdiction actually requires and keep a written record of every request and refusal. If a public body is not covered by a freedom of information law, its annual report and its board minutes usually still are.

Read the file like a sceptic

Most funding documents are written to be approved, not scrutinised. Read them with that in mind.

Follow the variance

Compare what was budgeted with what was spent, and what was promised with what was reported. Underspends that get quietly reallocated, targets that change shape between the application and the final report, audience figures rounded to the nearest thousand — these are the small inconsistencies that suggest nobody was checking, or that somebody was managing the paperwork rather than the project.

Read the annexes

The interesting material is rarely in the summary. It is in schedule two, in the conditions attached to the offer, in the line that says funding is “subject to satisfactory progress against milestone three.” Ask what happened to milestone three.

Note who signs what

Names matter more than logos. The officer who recommended the award, the trustee who signed the accounts, the director of the company that received the contract — write them all down. Then look for repeats.

Where the story usually hides

Cross-referencing is where a dull funding table turns into something publishable. Set up a simple spreadsheet with one row per organisation and columns for trustees, registered address, funders, contracts awarded and dates. The patterns surface fast.

• A panel member whose own organisation received a grant in a round they assessed.
• The same accountant, the same solicitor, or the same registered office across three “independent” applicants.
• A council transferring a venue to a trust at no cost, then awarding that trust a grant to run the same building.
• A subcontracting chain where most of the public money ends up two steps down, with no open tender at either step.
• A subsidy justified by access commitments — cheap tickets, community hire, schools work — that the monitoring reports never actually verify.

None of these is automatically wrongdoing. Each is a fair question, and a fair question put in writing is the engine of this kind of reporting.

Give the numbers a scene

A grant register tells you what was promised. The people the money was meant to reach tell you what arrived. Go to the youth theatre that lost its rehearsal space when the building was “rationalised.” Sit in the back of the community choir rehearsal that now charges subs because the grant stopped. Ask the disabled artists' network what happened to the access programme named in the bid.

Ask them for documents too — emails, schedules, invoices. They often have better paperwork than the funder. And attend the meetings: the scrutiny committee, the trust's annual general meeting, the consultation event where the same six people speak. The gap between what is said in the room and what is written in the report is frequently the story.

Ask questions that produce documents, not statements

“Can you comment on the arts budget?” gets you a press release. Instead, be narrow and be written. Ask for the assessment notes for a specific round, the unredacted version of a specific contract, the correspondence between two named officers about a named decision. Ask what the monitoring report for a named year concluded, and who signed it off. If the answer is a refusal, ask for the exemption you are relying on and the name of the person who applied it.

Keep the tone flat. You are not accusing anyone of anything yet; you are assembling a record. Most institutions answer specific questions badly but answer them eventually, and every exchange becomes part of the file.

Before you publish: a short fairness check

Put every significant allegation to the people concerned in writing, with a reasonable deadline and enough detail that they can respond properly. Separate what you can prove from what you suspect, and say so in the copy — an irregularity is not the same as a broken law, and a badly documented decision is not the same as a dishonest one. Publish the key documents alongside the story, or link to where readers can find them, so the piece can be checked rather than merely believed.

Defamation, privacy and data protection rules vary widely, and they bite hardest on stories about named individuals. Have someone with legal training read the draft before it goes out, and if you are unsure whether a document can be published in full, seek advice rather than guessing.

Then keep the spreadsheet. Funding rounds repeat, panels reconvene, the same names reappear with a different letterhead. The second story is always easier than the first, and the third one usually writes itself.

Photo: StockSnap / Pixabay