Clarivate publishes Journal Impact Factors (JIF) annually in late June in its Journal Citation Reports. It’s a mathematical calculation: The number of citations to works published in a journal in the previous two years, divided by the total number of citable items published in those previous two years.
In a 15 May 2025 blog post Nandita Quaderi, Senior Vice President & Editor-in-Chief, Web of Science, announced the cessation of including citations to retracted content in the numerator of the mathematical calculation that determines impact factor. Putting retractions into historical perspective, she identified the first retraction as happening in 1756 (it concerned Benjamin Wilson’s article about electricity criticizing Benjamin Franklin).
She also acknowledged that reasons for retractions run the gamut from honest mistakes to intentional manipulation. Regardless of why articles are retracted, they "play a crucial role in maintaining the integrity of the scholarly record." Quaderi goes on to assert, rightfully, that Clarivate tries very hard to ensure that the journals indexed in Web of Science Core Collection, which is the source for JCR, are trustworthy. "Our commitment to research integrity is not limited to our robust editorial selection and introducing new features in the Web of Science; it extends to the metrics, indicators and visualizations we include in the JCR and the insights we publish in our Institute for Scientific Information (ISI) reports."
Over the past several years, Clarivate has been working on making the statistics more valuable. Elaine Lasda commented on this in her Metrics Mashup column in the September/October 2021 issue of Online Searcher ("The Revamped Journal Citation Reports: Are More Metrics Better Metrics?", available through EBSCO, Factiva, LexisNexis, and ProQuest databases but not, ironically, included in Web of Science).
This has involved extending JIF to all journals in the Web of Science Collection, not just ones deemed impactful fo the sciences and social sciences and introduced new unified rankings. When considering the impact of retracted articles on the quality measures Clarivate employs, it’s important to recognize what a very small percentage of total content those articles represent—only 0.04%. Add to that the lengthy amount of time it takes publishers to issue a retraction (one recent retraction was of a paper published 50 years ago), and the problem of including them in calculating JIF is probably minimal. However, that is changing, with more retractions ( and shortened time span to retraction, propelling Clarivate toward this policy change.
Retraction Watch weighs in
In its commentary on Clarivate’s decision, Retraction Watch welcomed the new approach. It wrote that the overall retraction rate, not the Web of Science rate, was now up to 0.2%. This continues to threaten trust in scholarly research. Clarivate’s JIF, which Retraction Watch describes as a proxy for the importance of faculty research, although the measurement is not without controversy. Counting retractions in JIF to inform impact, Clarivate decided, distorts the metric.
Quaderi told Retraction Watch that Clarivate would stop counting citations once the paper is retracted, but would keep those that occurred before. She added that the company will continue to use Retraction Watch to flag retracted papers in indexed journals. It’s done that for the past 3 years. She added that the change would not impact a researcher’s h-index, as retracted papers will continue to be included in that calculation.
What it means for librarians
The library community is very aware of retractions. The rise of paper mills and predatory journals has long been a concern. It’s now been joined by worries about the role generative AI plays in scholarly literature. Recent retractions have occurred when AI-written papers are uncovered. This latest move by Clarivate will be carefully scrutinized by librarians to assess how it will affect their faculty’s publishing activities. That it doesn’t affect researchers’ h-index is reassuring. The decision to eliminate retracted articles from the numerator but not the denominator is interesting but how that plays out in the actual calculation remains a mystery since we haven’t seen the latest JCR.
What is clear is that librarians should watch this closely so they can effectively advise faculty and graduate students to be wary of citing retracted articles in their papers.