#283 2010 · Retraction Watch (Ivan Oransky & Adam Marcus) · Academic publishing / scientific integrity
Two journalists made scientific fraud visible just by writing down where it already was
the problem
A failure record scattered across thousands of independent institutions, each recording it privately and separately, functions the same as no record at all — no single party can see the pattern, and no individual institution has an incentive to publicize its own failures
background
When a scientific paper was retracted, the correction lived only inside the individual journal that published it — thousands of separate journals, each handling retractions differently, most with no obligation to state why a paper was pulled or to make that information easy to find later. A researcher whose paper was quietly retracted for misconduct in one journal faced essentially no aggregated consequence, because no institution — not the researcher's employer, not other journals, not funding bodies — had a single place to check a track record. The information existed, technically, but scattered and unindexed information that requires searching thousands of separate sources one at a time is functionally the same as no information.
In 2010, health journalists Ivan Oransky and Adam Marcus started Retraction Watch, initially a blog reporting on individual retractions, and later, with MacArthur Foundation funding starting in 2015, built it into a single searchable database aggregating retraction records across journals and tagging each one with its actual cause — fraud, honest error, image manipulation, plagiarism — rather than leaving the vague, often euphemistic language many journals used to describe why a paper was pulled.
what everyone would do
Lobby journals and institutions for stricter retraction policies and mandatory disclosure requirements — the standard reform-minded response to a transparency gap. It fails because the information was already technically public; the problem was never that journals refused to disclose retractions, it was that thousands of separate disclosures scattered across independent outlets, with no standard categorization, were functionally invisible to anyone trying to check a researcher's record.
what they saw
Oransky and Marcus saw that scattered, unindexed public information behaves exactly like no information at all — nobody could reasonably search thousands of individual journals one at a time to check a track record, so the accountability the disclosure requirement was supposed to create simply never materialized. The fix wasn't demanding new disclosure, it was aggregating and standardizing what already existed into one place anyone could actually check.
the move
Rather than lobby for new retraction policy or try to compel journals to disclose more, Oransky and Marcus simply aggregated and standardized information that individual journals were already, technically, making public — converting a scattered, effectively invisible record into a single, searchable, comparable dataset anyone (a hiring committee, a funder, a fellow researcher) could check in seconds.
why it works
Building a single searchable database that pulls retractions from thousands of separate journals and tags each with its actual cause converts information that required an impractical, journal-by-journal search into something checkable in seconds. Because no single institution had an incentive to publicize its own failures, and no individual journal's retraction record meant anything without the others to compare against, the value only existed once the records were combined — a hiring committee or funder could not previously see a pattern across a researcher's history, and now could immediately. That newly visible pattern created real accountability pressure on its own, without any new disclosure rule, which is part of why documented retraction volume rose sharply after the database existed to make retractions actually matter.
the payoff
The Retraction Watch Database now contains more than 35,000 tracked retractions with roughly 3,500 added annually, the site draws approximately 150,000 monthly readers, and documented retraction volume has risen sharply since the project's founding — from an estimated 45 retractions a month industry-wide in 2010 to nearly 300 a month by 2022 — a trend Oransky and Marcus attribute partly to increased scrutiny the aggregated, public record itself created.
where it breaks
The approach depends on the underlying records genuinely being public, even if scattered — it cannot surface information institutions are actively concealing rather than merely failing to aggregate, since there's nothing to index if the disclosure never happened in the first place. It also requires sustained effort to keep the aggregation current and accurately categorized across thousands of ongoing sources, which is real, continuous work rather than a one-time build; a database that goes stale or miscategorizes causes loses the trust that makes it useful. And it only creates accountability pressure if someone with actual power over outcomes — hiring committees, funders, other researchers — is willing to check and act on what the aggregated record shows, since a searchable database nobody consults produces visibility without consequence.
what came after
Retraction Watch won the Council of Science Editors' Award for Meritorious Achievement in 2025 and now operates with institutional funding from Crossref, functioning as the standard reference database publishers, universities, and funding agencies check when evaluating a researcher's publication record, and is widely credited in academic-integrity literature with materially increasing accountability pressure on journals and authors simply by making an already-public record actually usable.
references
- [1]Nature — Retraction Watch featureNature, 2022media.nature.com
- [2]Science (AAAS) — Scientific retractions may become easier to spot as Retraction Watch finds new partnerScience / AAAS, 2023science.org