Citation stacking is a form of citation manipulation in which two or more journals arrange, explicitly or through informal understanding, to cite each other’s articles at a rate far above what the normal scholarly literature in their field would produce on its own. The goal is to inflate each journal’s citation-based metrics — most consequentially the Journal Impact Factor (JIF) reported in Clarivate’s Journal Citation Reports (JCR) — without a corresponding increase in the actual scholarly value or influence of the work being cited. It is one of several patterns grouped under the broader term citation cartel, and it is distinct from, but related to, simple self-citation, where a journal inflates its own metric by citing itself rather than a partner title.
This guide explains what citation stacking is in operational terms, how it differs from the legitimate citation patterns that occur naturally between journals in the same subfield, the specific mechanics Clarivate and Elsevier use to detect it, and what happens to a journal once a pattern is confirmed.
What citation stacking actually is
The defining feature of citation stacking is reciprocity that is disproportionate to genuine scholarly relevance: Journal A’s authors and editors direct an unusually high share of their citations toward Journal B, and Journal B reciprocates, concentrated in a way that would not occur if authors were simply citing the most relevant prior work in their field. This can happen through several concrete mechanisms:
- Editorial encouragement — reviewers or editors asking authors to add citations to the partner journal as a condition of acceptance, a practice closely related to coercive citation.
- Coordinated special issues — guest editors or overlapping editorial boards across two or more journals commissioning articles that cite each other’s recent output.
- Author-level citation rings — a group of authors publishing across a small set of journals and citing each other’s articles in that same set, sometimes as a paid or brokered arrangement.
What all three have in common is that the citation pattern is concentrated on a narrow set of partner titles rather than distributed across the genuinely relevant literature, and it is timed to land inside the two-year (or, for the 5-Year JIF, five-year) citation window that JCR uses to calculate Impact Factor.
How citation stacking differs from legitimate cross-journal citation patterns
Journals that share a subfield cite each other constantly, and that is normal and expected — it is the ordinary operation of a scholarly literature responding to the same body of prior work. Distinguishing legitimate patterns from stacking is a matter of degree and structure, not the mere existence of reciprocal citation:
- Breadth of the citing/cited relationship. A legitimate subfield relationship involves citations distributed across many journals in that specialty. Stacking is narrow: citation traffic concentrates disproportionately on one or two partner titles rather than the field as a whole.
- Direction and reciprocity. Natural citation is largely one-directional, following the actual flow of foundational-to-derivative research. Stacking shows an unusually symmetric back-and-forth between the same pair of journals.
- Timing. Legitimate citation accrues gradually as the literature engages with a paper over years. Stacking often shows a sharp, coordinated spike concentrated in the specific citation window JCR uses, rather than a gradual accumulation.
- Relevance of the citing context. In genuine scholarship, a citation supports a specific claim or builds on specific prior findings. In stacking, citations are frequently added in ways that are tangential to the argument — a list of references with little connection to the surrounding text, sometimes inserted late in the editorial process.
Editors and citation analysts weigh these factors together rather than relying on any single threshold, because reciprocal citation between two closely related journals is not, by itself, evidence of manipulation.
How Clarivate detects citation stacking for JCR
Clarivate’s Journal Citation Reports separately reports each indexed journal’s journal self-cite figure alongside its Journal Impact Factor, and has used both self-citation rate and reciprocal cross-journal citation concentration as grounds to suppress a journal from a JCR edition — withholding that year’s Impact Factor rather than publishing an inflated one. Clarivate’s own analysis of the 2014 JCR Science Edition found 85% of indexed journals had self-citation rates of 15% or lower, and an earlier 2002 report found 82% of Science Citation Index-edition journals at or below 20%, giving a sense of what a typical distribution looks like against which outliers stand out. Clarivate deliberately does not publish the exact numerical thresholds that trigger a suppression review, on the stated basis that a fixed, published cutoff would simply tell bad actors how far they can go without tripping it.
Suppression for anomalous citation behavior is a distinct action from removing a journal from the Web of Science Core Collection entirely. A journal can be suppressed from a specific JCR edition — losing that year’s Impact Factor listing — while remaining indexed in Web of Science; conversely, full removal from the Core Collection is governed by a separate 28-criteria evaluation process (24 quality criteria plus 4 impact criteria) under Clarivate’s editorial selection process, and citation-manipulation findings can factor into either outcome depending on severity. For the mechanics of full delisting rather than single-edition suppression, see CASRAI’s companion guide, why journals get delisted from Web of Science.
The detection methods Clarivate and independent bibliometricians use to flag stacking include:
- Journal self-citation and cross-citation rate monitoring — tracking what share of a journal’s citations, in a given JCR window, come from itself or from a small set of specific partner journals, against the field-typical distribution.
- Citation network/cluster analysis — graph-based analysis of the citation network can surface unusually dense, isolated clusters of journals that cite each other far more than they cite (or are cited by) the wider field, a signature of a citation cartel rather than an organic subfield relationship. See CASRAI’s guide to citation network analysis for the underlying bibliometric methods (co-citation and bibliographic coupling) that make this kind of clustering visible.
- Timing analysis — a spike in citations to a specific journal concentrated tightly within the JIF’s citation window, rather than distributed across the years following publication, is a pattern flagged for manual review.
- Editorial and authorship overlap checks — shared editorial board members, guest editors, or a recurring small pool of authors publishing across the same narrow set of journals, especially when combined with unusually dense cross-citation.
How Elsevier/Scopus approaches citation manipulation
Elsevier applies a parallel but separately administered process for Scopus. Per Elsevier’s own Scopus content policy, an already-indexed journal can be flagged for re-evaluation by the Content Selection and Advisory Board (CSAB) in two ways: publication-standards or integrity concerns raised by the research community, or Elsevier’s own data models detecting “unexpected patterns” in publication activity, citation graphs, author collaborations, or content — language that explicitly covers citation-stacking-type anomalies, not just plagiarism or peer-review fraud. During a re-evaluation, content flow into Scopus is suspended until CSAB decides whether to continue or discontinue coverage. Content already indexed before a discontinuation decision generally remains part of the scientific record; full retroactive removal is reserved for exceptional, proven cases of severe unethical publication practice. See CASRAI’s companion guide, why and how Scopus delists journals, for the full discontinuation process.
Real, documented cases of coordinated citation manipulation
Citation cartels are not a theoretical risk. Research-integrity investigators have documented citation-boosting schemes operating as an outright commercial service: in a case reported by Chemical & Engineering News in February 2026, investigator Anna Abalkina (Free University of Berlin) identified roughly 20 chemistry papers built around fabricated author identities — fake researchers falsely listed as affiliated with institutions in Senegal, Kenya, Somalia, Namibia, Saudi Arabia, Iraq, and Egypt, partly to exploit article-processing-charge waivers offered to those countries — that were sold as vehicles for purchased citations, reportedly priced at roughly $5–10 per citation. The publishers named in that investigation (Taylor & Francis, Springer Nature, and Elsevier) all confirmed they were investigating at the time of reporting. This is a useful illustration of how far coordinated citation manipulation has moved beyond informal editorial pressure into an organized market, though the specific findings should be treated as still under investigation rather than final as of this guide’s publication.
Citation stacking also frequently co-occurs with other integrity red flags screened for by publishers and integrity offices — see CASRAI’s guide to paper mills and tortured phrases for how citation cartels fit alongside image manipulation, fabricated authorship, and paper-mill output as part of a broader manuscript-screening pattern.
Consequences for a journal found to be citation stacking
- JCR suppression. The journal’s Impact Factor is withheld from that year’s Journal Citation Reports release. This is typically the first and most immediate consequence, and can recur in subsequent years if the pattern continues.
- Removal from the Web of Science Core Collection. In more severe or persistent cases, the journal can be delisted entirely, losing eligibility for any future Impact Factor and triggering Clarivate’s standard multi-year embargo before resubmission is possible.
- Discontinuation from Scopus. A parallel, separately run CSAB re-evaluation can result in the journal being discontinued from Scopus coverage.
- Reputational damage. Even short of formal delisting, a documented suppression or investigation becomes a visible, checkable flag to researchers, tenure and promotion committees, and funders who treat indexing status as a quality signal.
How researchers and editors can avoid false flags
Because legitimate subfield citation patterns can superficially resemble the early signs of stacking, journals and authors who want to avoid an unwarranted integrity review should be able to show that citations to a related journal are distributed rather than concentrated, that citing decisions are made by peer reviewers and authors on scholarly merit rather than editorial instruction, and that any special-issue collaboration between journals is transparently disclosed rather than used to funnel reciprocal citations. Editors evaluating their own journal’s citation profile can compare their self-citation and top-partner-journal citation share against the general JCR distribution benchmarks above as an informal early check, though only Clarivate and Elsevier’s own review processes make the final determination.
Frequently asked questions
Is citation stacking the same as self-citation?
No. Self-citation is a journal citing its own previously published articles. Citation stacking is reciprocal citation concentrated between two or more separate journals. Both are tracked by Clarivate and can each independently trigger a JCR suppression review, and the two often appear together in the same investigation, but they are distinct patterns. See CASRAI’s self-citation rate entry for the narrower concept.
Can normal collaboration between two journals in the same field look like citation stacking?
It can superficially resemble it, which is exactly why detection relies on multiple signals together (concentration, timing, editorial overlap, and relevance of the citing context) rather than any single metric. A legitimate special issue or genuinely overlapping research community is not, by itself, evidence of manipulation.
Does JCR suppression mean a journal is permanently removed from Web of Science?
No. Suppression withholds a specific year’s Journal Citation Reports listing; the journal can remain indexed in the Web of Science Core Collection. Full delisting from the Core Collection is a separate, more severe action governed by Clarivate’s broader 28-criteria evaluation process.
What is a citation cartel?
Citation cartel is the broader umbrella term for any coordinated group — authors, journals, or editorial boards — that inflates citation counts through disproportionate mutual citation. Citation stacking specifically describes the cross-journal version of this pattern. See CASRAI’s citation cartel dictionary entry.
Who actually investigates suspected citation stacking?
For Web of Science-indexed journals, Clarivate’s own editorial and citation-analysis teams review flagged titles as part of the JCR production process. For Scopus, Elsevier’s Content Selection and Advisory Board (CSAB) conducts a parallel re-evaluation. Independent researchers and research-integrity investigators (publishing in venues like Scholarly Kitchen, Retraction Watch, or peer-reviewed bibliometrics journals) also identify and publicize suspected cases, which can prompt a formal review by either indexer.
This guide covers the structural mechanics of citation stacking as documented by Clarivate and Elsevier’s own published policies and independent research-integrity reporting. Exact numerical suppression thresholds are not published by either indexer and should not be treated as fixed rules.
