The Reporting Question

A topic is a subject: our playground. A question is something you can find out: why was the slide removed last week? A real question points at people who know — whoever removed it — and at things you can check, like the notice the city posted.

Test a question two ways. Could somebody actually know the answer? Could you reach that somebody, or read it somewhere? If both are yes, you can report it.

A reporting question passes three tests. What exactly happened — a specific event or change, not a mood. Who says so — people or records that can attest to it. How do they know — the way those sources came by the fact. "Why did the district drop its lunch vendor in March?" passes all three: an event (the switch), sources (board members, the vendor, the minutes), and a knowable basis (they were in the room; the paper trail exists).

Once the question is set, the reporting plan writes itself: every task is a step toward the answer. Read the March minutes. Call the food-services director. Call the vendor. A reporter with a question knows when to stop — when it is answered. A reporter with a topic never does.

A working question names an actor, an action, and a time and place — which makes it checkable and, just as important, refutable. "The city is wasting money" cannot be wrong in any specific way, so it cannot be right in one either. "Overtime in the water department doubled in two years while staffing stayed flat" can be confirmed, corrected, or killed by a payroll ledger. Precision is not pedantry; it is what lets the evidence answer.

Editors think in questions because questions can be costed: this one needs two interviews and one records request; that one needs six months. Reporters also set the minimum story before pitching — the piece you can still file if the biggest claim does not hold — so the question is worth asking even when the answer disappoints.

"Tell me about dogs" is too big. "Why does Rex bark at the mail carrier?" is a question you can answer. Watch Rex. Ask the carrier.

Small questions get answered. Big wonderings just float. When you want to find something out, shrink it until someone could answer it.

The reporting question does the work a research question does in study design: it operationalizes. Narrowing "homelessness" to "why did the county's overflow shelter close in January despite a funded contract?" fixes the population, the period, and the outcome, and thereby fixes which evidence bears on it — the contract, the county correspondence, the provider's filings. Vague questions are unfalsifiable, and unfalsifiable stories are unreportable.

Investigative practice formalizes this as the hypothesis memo: state the question, the minimum and maximum story, the sources that could confirm or kill it, and the first three steps. The memo forces the disconfirming source into the plan from day one — the person whose account, if true, ends the story.

Epistemically, the question is the newsroom's unit of inquiry, and its form controls what inference the finished story can license. Tuchman's classic observation — that objectivity operates in newsrooms as a strategic ritual — cuts here: a vague question invites a quote-balanced piece in which nothing specific is at stake, while a sharply posed question forces the evidence to adjudicate. Question design is an epistemic commitment, not housekeeping.

Compare pre-registration in the sciences: fixing the question before collection disciplines the temptation to report whatever pattern the notebook happens to hold. The reporter's analogue is writing the hypothesis memo before the first interview, then recording how the question changed and why — an audit trail for the story's own reasoning.