The economics pilot, and how to read its chart.
We ran the same reading as our medicine census on retracted economics papers. Today the pilot is one chart. This page says what the pilot is, how to read the chart and what it can’t tell you.
One question, asked outside medicine.
After a paper is retracted, do later papers still cite it as support? The census asks this of medicine. The pilot asks it of economics.
- Which papers.Papers Retraction Watch tags as business or economics, retracted 1990 to 2023, with a DOI. Paperwork retractions, which find no fault in the work, are set aside.
- Which citations.Open-access papers in PubMed Central that cite one of them in a year after the retraction. Those citing papers are mostly biomedical, health and environmental journals.
- How it was read.Three AI models from different families labelled every citing passage, with the same labels as the medicine census. Before the run, their labels were checked against blind reference labels on a separate set, as the pilot’s plan required.
- How big.231 citations to 130 retracted papers, against 51,626 citations to 12,328 papers in the medicine census.
How to read it.
38.1% of citations to retracted economics papers still cite them as support.
Close to medicine (40.7%). 231 citations to 130 retracted economics papers, all made after the retraction. Only 1.3% mention the retraction.
Cite the retracted paper as support, after the retraction
Why it matters. Retracted research keeps being cited as support outside medicine too, about as often.
A pilot. The citing papers come from free full text in PubMed Central, mostly biomedical, health and environmental journals, so this shows how economics retractions are cited there, not in economics journals.
A pilot. Citing papers from PubMed Central full text.
Source: SaneCite economics pilot: 231 citations to 130 papers Retraction Watch tags as business or economics, retracted 1990–2023. Three AI models labelled each citation; pending human review. The data. Data as of 29 Oct 2026.
Source: SaneCite economics pilot, 231 citations, 29 Oct 2026, pending human review · sanecite.com/charts
Show the numbers
| Field | Citations judged | Cited as support | Share (95% CI, %) | Retracted papers |
|---|---|---|---|---|
| Economics (pilot) | 231 | 88 | 38.1% (32.1–44.5) | 130 |
| Medicine (Phase 2 census) | 51,626 | 21,037 | 40.7% (39.9–41.5) | 12,328 |
Economics: papers Retraction Watch tags as business or economics, retracted 1990–2023, cited after the retraction in PubMed Central full text. Three AI models labelled each citation. The economics interval is a Wilson interval; the medicine interval comes from re-running the numbers over whole retracted papers. 3 of the 231 economics citations mention the retraction.
- 01
Two bars, one measure
Each bar is the share of citations, made after the retraction, that cite the retracted paper as support. The top bar is medicine (40.7%); the cyan bar is the economics pilot (38.1%). Both are drawn to the same scale.
- 02
What “cited as support” means
The citing sentence rests on a finding, number, method or conclusion of the retracted paper. If the retracted data were made up, that sentence would lose its support.
- 03
It counts citations
Economics is 88 of 231 citations, not papers. One retracted paper can be cited many times.
- 04
Open “Show the numbers”
It gives the counts and a 95% range for each bar: 32.1–44.5% for economics, 39.9–41.5% for medicine. The economics range is much wider because the pilot is much smaller.
The subtitle adds one more figure: only 1.3% of the economics citations (3 of 231) mention the retraction. The “Pilot · pending human review” tag means AI models did the reading and people have not yet checked it.
A pilot, read in biomedical journals.
- Not economics journals.The citing papers come from PubMed Central, so this shows how economics retractions are cited there, not in economics journals.
- Not checked by people.These are AI readings, pending human review, like the medicine census.
- Small.It is a pilot, much smaller than the medicine census. Read it as a first look beyond medicine.
- Check it yourself.The chart’s data · the pilot’s results · code and data. The chart page has the image and a caption, free to share with credit to SaneCite (sanecite.com).
Short answers.
- Is this a study of economics journals?No. The retracted papers are economics papers, but the papers citing them come from PubMed Central, mostly biomedical, health and environmental journals.
- Did people check these readings?Not yet. Three AI models labelled every citation, and the results are pending human review.
- Can the chart be shared?Yes. It is free to share with credit to SaneCite (sanecite.com). The chart page has the image, a caption and the data.