Editorial Standards

Why this page exists

A publication that will not describe its own method is asking to be taken on trust. This page sets out how NEWS ANALYSIS decides what to cover, how we source it, and what we do when we are wrong, so that a reader can judge our work against a stated standard rather than an impression.

Choosing what to cover

We are an international general news publication with ten desks, and we publish a small fraction of what happens. A story earns a place when we can add something a headline cannot: the mechanism behind an announcement, the number that was left out, the consequence that follows, or the context that makes a claim plausible or doubtful.

We do not cover a story simply because it is being covered elsewhere. Volume of attention is not, by itself, evidence of importance. We also try to resist the opposite error of contrarianism for its own sake. The test is whether we can say something specific and supportable.

Sourcing

Our default is the primary source. If a story rests on a regulatory filing, a court document, a peer reviewed paper, a transcript, a product specification or a set of published statistics, we read that document before writing. Reporting a summary of a study as though it were the study is one of the most common ways an inaccurate claim enters circulation, and we try hard not to be part of that chain.

Where we rely on another publication for a fact we cannot independently verify, we say so and we link to them. We do not present another outlet’s reporting as our own, and we do not launder a single unverified claim into apparent consensus by citing several outlets that all trace back to it.

We attribute. Claims made by a company about its own product, by a government about its own policy, or by an organisation about its own performance are reported as claims, with the claimant named. The reader should never have to guess whether a sentence is an established fact or a party’s assertion.

Numbers

Figures get specific treatment because they are unusually easy to get wrong and unusually persuasive when wrong. We check that a number in our article matches the number in the source, that the units and the time period are the same, and that a percentage change has a stated baseline. Where a figure is an estimate, a forecast or a range, we describe it that way rather than presenting a midpoint as a fact.

We are careful with comparisons across countries and currencies, because definitions often differ even when labels match. If two figures are not measured the same way, we would rather say so than imply a precision that is not there.

Fairness

When our reporting is critical of an identifiable organisation or person, we represent their position accurately, in their own terms, and we say plainly when they have declined to comment or have not responded. Absence of a response is reported as absence of a response, never as agreement or as evasion.

We distinguish between a disputed question and a settled one. Presenting a well established finding as an open controversy misleads readers, and so does presenting a genuinely open question as resolved. Where informed people disagree, we describe the disagreement and what would settle it.

Language

We write plainly. We avoid the vocabulary of promotion, and we try not to use intensifiers in place of evidence. If something is significant, the specifics should show it. We prefer the concrete number to the adjective.

Images

Images on this site are illustrative unless captioned otherwise. We do not present an illustration as documentary evidence of an event, and we do not use imagery to imply a claim the article does not support.

Independence

Editorial decisions are made on editorial grounds. Advertising and commercial relationships do not determine what we cover or how we cover it, and the people who sell advertising do not review articles before publication. Our Ownership and Funding page explains how this publication is financed and who controls it.

Correcting the record

We will make mistakes. When we do, we fix the article and disclose the change, because an error quietly removed is an error a reader cannot account for. Our Corrections page sets out how we handle everything from a typographical slip to a substantive error that changes the meaning of a story.

We would rather publish a correction than defend an indefensible sentence. A publication’s credibility comes from how it behaves when it is wrong, not from a claim never to be.