Standards · How we work

Data Changelog

Our sourcing methodology promises that a correction is applied at the data source and reflected on every page at once, according to the published corrections process. This page is the honest record of that promise kept: every dated entry below is a real factual correction made since the site launched in June 2026, for example a June 2026 State Pension figure that was live here at £240.30 per week and corrected to the Department for Work & Pensions' own £241.30 per week rate (a 4.8% triple-lock uprate), what was wrong, and what it became. We don't log routine content additions here, only corrections to something that was previously inaccurate, mislabeled, or unverifiable.

  1. Separated tax-year rates from the database compile stamp

    Homepage, statistics, and CiteThis surfaces had been able to read as if the current tax-year State Pension / NI rates and the ETL compile stamp shared one clock. Disclosed the three populations on methodology (current tax-year rates, adjacent prior-year comparison rates, and the _metadata.data_vintage compile stamp) so Dataset temporalCoverage and side-by-side totals cannot pretend they share one year.

  2. Removed a misattributed statistic

    The "finding lost pensions" guide claimed the median UK worker holds 11 jobs over their working life, attributed to the Office for National Statistics. That figure traces to a US Bureau of Labor Statistics study of American workers, not UK ONS data, independent industry analysis puts the real UK figure at roughly 3-3.5 jobs per working lifetime. Removed the misattribution and the unverified claim next to it.

  3. Removed a hidden overstated byline

    A screen-reader-only span on research pages used an outdated organisation label and the verb "Reviewed", independent of the page's own visible byline, which correctly says "Compiled by" (PlainPension has no named individual reviewer). Fixed the hidden text to match the honest visible text.

  4. Refreshed the lost-pensions figure and reframed advice-adjacent language

    The "£30 billion forgotten-pot problem" statistic (used across 3 pages) was superseded, the current Pensions Policy Institute research (Briefing Note 138, "Lost Pensions 2024", October 2024) puts the figure at £31.1 billion across 3.3 million pots, with an average pot size of roughly £9,400 (not £30bn / 4.8 million pots / ~£6,000, which traced to an older report). Corrected all 3 occurrences with the specific report cited. Separately, reframed 3 sentences that read as personal-finance sequencing advice ("almost always the right route") as conditional information pointing to Pension Wise or an FCA-regulated adviser, the underlying calculations were correct, only the framing needed to change.

  5. Corrected the 2026/27 State Pension rate

    The full New State Pension for 2026/27 was published on this site as £240.30/week (£12,495.60/year, 4.4% triple-lock uprate). Verified against gov.uk and corrected to £241.30/week (£12,547.60/year, 4.8% uprate); the basic State Pension figure was corrected from £184.15 to £184.90/week. Fixed at the data source and rebuilt so every page and calculator result reflects the corrected rate.

Found something that looks wrong?

Use the typed corrections pathway or the contact page with the page URL and the figure you believe is off. Real, confirmed corrections are added to this page.

Download the UK pension statistics CSV cited on this page: uk-pension-statistics.csv (CC0).