The Journal Impact Factor is a perfectly respectable statistic about journals. The trouble starts when it is asked to stand in for judgements about people. A hiring panel that ranks candidates by the impact factors of the journals they have published in is not using a crude version of a good method. It is using a measure built for one purpose to answer a different question, and it gets the wrong answer often enough to matter.
Our position is simple. The JIF has a legitimate place in understanding journals: a librarian comparing titles within one field, or an editor tracking how a journal's citation profile changes over time, can use it sensibly alongside other evidence. It has no legitimate place as a score for individual articles or individual researchers. That is not a fringe view. It is the stated position of the research-assessment reform movement and, in plainer language than many people realise, of the company that publishes the number.
We have explained elsewhere what the Impact Factor measures and why journal rankings move. This piece makes the argument for drawing a firmer line around how it is used.
An average that describes almost no one
The JIF is an average, and averages behave badly when the underlying data are lopsided. Citations to papers in a journal are heavily skewed: a minority of highly cited articles pulls the average upwards, while a typical paper in the same journal is cited less often than the headline figure suggests.
In 2016 a group of authors including Vincent Larivière, Véronique Kiermer and Stephen Curry published a simple method for generating the citation distributions that sit behind impact factors, and applied it to journals across the JIF spectrum. Their conclusion was blunt: distributions are highly skewed, they overlap extensively between journals with very different impact factors, and the citation performance of an individual paper cannot be inferred from the JIF of the journal that published it.
That finding should settle the matter for anyone using journal metrics to evaluate a CV. A paper in a lower-JIF specialist journal can easily outperform many papers in a higher-JIF title. Assigning each paper the score of its venue rewards authors for the success of other people's articles, and penalises them for venues that suit their audience rather than their promotion committee.
The numbers do not travel between fields
Citation practices vary sharply by discipline. Fields differ in how many references authors include, how quickly work is cited, how much of the literature appears in journals rather than books, and how large the research community is. The San Francisco Declaration on Research Assessment (DORA) lists the field-specific properties of the JIF among its core deficiencies, along with the mixing of article types such as reviews and primary research.
Clarivate itself now publishes field-normalised indicators, such as the Journal Citation Indicator, alongside the JIF in the Journal Citation Reports. In our view that is a tacit acknowledgement that raw citation rates do not compare cleanly across disciplines. Yet institutional processes still place a mathematician, a cell biologist and a historian on the same scale and treat the raw numbers as comparable. That is not a subtle misuse, and it is one that every major statement on responsible metrics warns against.
Even the publisher says so
When Clarivate released the 2026 Journal Citation Reports in June, it described the journal-level indicators in the JCR as designed to support journal evaluation, not to assess the performance of individual researchers or articles. Institutions that fold the JIF into individual promotion or funding decisions are therefore not following the provider's guidance; they are going beyond it.
This matters because defenders of metric-heavy evaluation sometimes present criticism of the JIF as an ideological objection to measurement. It is not. The objection is to using a well-defined journal statistic outside its domain.
Incentives invite gaming
A metric that carries career consequences will attract manipulation. DORA warned in 2012 that impact factors can be gamed through editorial policy. The evidence since then has borne this out in a more organised form.
Clarivate suppresses JCR metrics for journals with anomalous citation patterns, including excessive self-citation and citation stacking, where two or more journals cite each other at unusually high rates. Retraction Watch reported that 20 journals had their impact factors suppressed in the 2025 release, four for citation stacking and 16 for self-citation, compared with 17 in 2024 and four in 2023. The same release, it reported, excluded citations to retracted papers from impact factor calculations for the first time.
Suppression shows that the system has defences. It also shows that the incentive to inflate a journal's number is strong enough that some editors and publishers act on it. The more weight institutions place on the JIF, the stronger that incentive becomes. Reducing the misuse of the metric is part of protecting its integrity.
The reform consensus is already written down
The case for responsible use has been made repeatedly, by different communities, in documents that institutions can sign and implement.
- DORA (2012). Developed by journal editors and publishers at the American Society for Cell Biology meeting in San Francisco in December 2012, its first recommendation is not to use journal-based metrics such as the JIF as a surrogate for the quality of individual articles, or to assess individual scientists in hiring, promotion or funding decisions. It also asks publishers to reduce their emphasis on the JIF as a promotional tool, or to present it alongside other metrics.
- The Leiden Manifesto (2015). Published in Nature in April 2015 by Diana Hicks, Paul Wouters, Ludo Waltman, Sarah de Rijcke and Ismael Rafols, its ten principles include that quantitative evaluation should support qualitative, expert assessment; that assessment should account for variation by field; that individual researchers should be judged on a qualitative assessment of their portfolio; and that evaluators should avoid misplaced concreteness and false precision.
- CoARA (2022). The Agreement on Reforming Research Assessment, published in July 2022, commits signatories to abandon inappropriate uses of journal- and publication-based metrics in research assessment, in particular inappropriate uses of the JIF and the h-index, and to avoid using rankings of research organisations in assessing research. Unlike DORA, it asks signatories to publish an action plan and report on progress.
None of these documents calls for abolishing metrics. All of them place judgement first and indicators in a supporting role.
What responsible use looks like
For publishers, it means presenting the JIF with its year and subject category, alongside other indicators, and never implying that it certifies the quality of an individual paper. It means offering article-level information, as DORA recommends, and treating any approach that might inflate citations as an integrity risk rather than a growth strategy.
For institutions and funders, it means asking candidates to describe their most significant contributions and reading the work, with metrics used, if at all, as context that a panel can explain and challenge. Signing DORA or CoARA is the easy part; changing the templates, scoring rubrics and committee habits is where reform succeeds or fails.
For researchers, it means choosing journals for their audience, editorial standards and policies, and resisting the temptation to let a single number make that decision for them.
Conclusion
The Impact Factor is not the villain of research assessment. It is a narrow journal statistic that has been pressed into service far beyond its design, and its own publisher now says as much. The responsible course is not to pretend it does not exist, but to confine it to questions it can answer. Journals can be compared within fields with care. Papers and people deserve to be judged on what they actually contributed.
