Every news outlet makes mistakes. The difference between a trustworthy publication and an unreliable one is not the error rate. It is what happens after the error: whether it is acknowledged, fixed visibly, and recorded where readers can find it.
A news corrections policy is the set of rules a publication follows when it gets something wrong. It answers three questions: who decides a correction is needed, how prominently the fix is displayed, and whether the original mistake stays visible or quietly disappears. Readers who want to judge a source can learn more from that policy than from any slogan about integrity on the About page.
This explainer explains how corrections work, why the practice signals quality rather than weakness, and how to read a corrections page the way an analyst reads a data release — including the revisions. Anyone who follows economic statistics already understands the logic: a first estimate is a draft, and the honest institutions are the ones that publish the revision. The same holds for journalism, as our coverage of how GDP figures get calculated and revised shows. Readers following this should also see How GDP Figures Get Calculated — and Revised.
What is a news corrections policy?
A news corrections policy is a publication's standing procedure for handling factual errors. It typically defines what counts as a correction, who has authority to issue one, where it appears, and how long the record is kept. Good policies also distinguish between a correction, which fixes a fact, and a clarification, which fixes context or wording without changing the substance.
The strongest policies share a few features. They are written down, not left to individual editors' judgment. They name a channel through which readers can report errors. They place the correction in a visible location — at the top of the corrected article, or in a dedicated corrections section — rather than in a footnote. And they never silently rewrite a published story, because a silent edit destroys the one thing a correction is meant to protect: the reader's ability to see what was claimed and what changed.
One terminological aside, because the word misleads search engines. In criminal justice, "corrections" refers to the punishment, treatment and supervision of convicted offenders — a system of prisons, parole and probation, as Wikipedia's article on corrections explains. A newspaper's corrections page has nothing to do with that system. The shared word is a coincidence of English, not a shared practice.
Why do mistakes happen even at careful outlets?
News is produced under time pressure, by people, from sources that are themselves imperfect. A statistic gets transcribed wrong. A name is misspelled. A claim from an official turns out to be false, but the outlet reported it accurately — which is a different kind of problem, and a good corrections policy handles it differently too.
Economic journalism is a useful case because the underlying data keeps changing. Jobless claims, inflation readings and growth estimates all arrive as preliminary figures that get revised. A story can be perfectly accurate on publication day and misleading a month later. That is why serious financial desks distinguish between an error, which they correct, and a subsequent data revision, which they usually note in an update. Readers can see the same discipline in our piece on how weekly jobless claims numbers get made.
The incentive structure matters here. A newsroom that punishes every error harshly teaches its writers to hide mistakes. A newsroom that treats a self-reported error as a professional act gets told about errors early, while they are still cheap to fix. The policy is really a management tool: it sets what staff do when something goes wrong at 11 p.m. before deadline.
What should a correction look like?
A well-made correction has four parts.
- The error, stated plainly. What the story said, and what is wrong with it. Vague phrasing — "we have updated this article" — tells the reader nothing.
- The correct information. Not just a deletion. A correction that removes a wrong number without supplying the right one leaves the reader worse off than before.
- The date and location. When the fix was made, and whether it appears at the top of the article, at the relevant paragraph, or in a corrections log.
- The right prominence. Small typos can be fixed with a brief note. A wrong claim that was central to the story deserves a correction at the top, and if the whole premise fails, a retraction.
The escalation ladder matters. A correction fixes a discrete fact. A clarification fixes framing. An editor's note flags a substantive problem with context. A retraction withdraws the piece entirely, reserved for stories whose core claim collapses. Outlets that use one blunt instrument for all four cases are telling you something about their standards.
How can readers use corrections to judge a source?
The analysis: treat a corrections page the way an auditor treats a revision history. The presence of corrections is neutral or positive — it means the outlet is checking its work. What you are evaluating is the pattern.
Ask four questions. First, are corrections dated and specific, or generic? Second, do they appear on the original article, where readers of that article will actually see them, or only in a back corner of the site? Third, does the outlet correct errors in its own original reporting, or only trivial typos — the easy confessions? Fourth, does the tone treat the reader as a partner or as an inconvenience?
An outlet with no visible corrections is not necessarily flawless. It may simply be unwilling to admit error, and there is no way to tell the difference from the outside. That asymmetry is the core argument for the practice: a visible corrections record is a signal that cannot easily be faked, while the absence of one proves nothing. The reader's rational move is to trust the publication that shows its work.
What would change this reading? Evidence that a corrections page exists mainly for show — corrections that are undated, buried, or never applied to consequential stories. If the record shows that pattern, the signal flips.
What this means for the economics reader
Statistical agencies and newsrooms face the same problem: publish a first version and be judged on a draft, or wait for certainty and be useless. The agencies publish, then revise, and the credible ones print the revision history alongside the current figure. Our explainers on how inflation is measured and how the unemployment and participation rates differ rest on that same understanding of data as a living record.
A news corrections policy is the journalistic equivalent. It says: here is what we claimed, here is what we now know, here is the difference. Readers who learn to check for it gain a durable, low-cost test of reliability — one that works before they read a single story.
The practical takeaway: before trusting any outlet, find its corrections page. If it exists, read a few entries. If it is specific, dated and prominent, the outlet probably deserves the benefit of the doubt. If you cannot find one at all, discount accordingly — and remember that the absence of admitted mistakes is not evidence that mistakes are absent.
