Citation Cartel

A citation cartel is a network of researchers who systematically cite each other's work to inflate their citation metrics, regardless of whether the citations are intellectually warranted. The practice is a direct consequence of the publish-or-perish regime and the academic incentive structure, where citation counts serve as proxies for research impact and are used in hiring, promotion, and funding decisions.\n\nCitation cartels are a form of network manipulation: they exploit the topological properties of citation networks to create artificial clusters of high centrality. The practice is not always explicit or conspiratorial; it can emerge from disciplinary insularity, mutual admiration, or the simple fact that researchers in the same subfield naturally cite each other. What distinguishes a cartel from ordinary citation practice is the systematic inflation of metrics through citation exchanges that would not occur in an unbiased evaluation. The phenomenon is one of many ways in which the systemic incentive structure of science corrupts its epistemic function.\n\n\n

The Topological Signature of Cartels

Citation cartels leave detectable traces in the topology of citation networks. A genuine intellectual cluster is characterized by heterogeneous citations: members cite each other, but they also cite widely outside the cluster. A cartel is characterized by citation homophily: the ratio of within-group citations to between-group citations is anomalously high. This is the network-theoretic equivalent of inbreeding: the cluster's intellectual gene pool shrinks, and its capacity to detect novel ideas degrades.

The topological signature can be quantified. The modularity of a citation network measures the strength of its community structure. A cartel produces a local modularity spike — a subnetwork that is more densely connected internally than would be expected by chance. But modularity alone is not proof of cartel behavior; it is a signal that warrants investigation. The investigation requires examining the content of the citations: are they substantive engagements with the cited work, or are they perfunctory mentions inserted to inflate metrics? This is the difference between a research community and a citation gaming network.

The Epistemic Consequences

The damage of citation cartels is not merely statistical. It is epistemic. When a cartel successfully inflates its metrics, it diverts attention, funding, and talent away from genuinely important work. The cartel's members get hired, promoted, and funded; the cartel's ideas get taught, cited, and institutionalized. The result is a feedback loop in which cartel success produces cartel influence, and cartel influence produces more cartel success. The scientific community becomes an information topology in which centrality is purchased rather than earned.

The most dangerous cartels are not the explicit ones but the emergent ones. A subfield in which everyone attends the same conferences, reads the same journals, and trains in the same labs will naturally produce dense citation clusters without any conscious coordination. The cartel emerges from homophily, not conspiracy. This is the "invisible cartel": a network that behaves like a cartel in its effects but not in its intentions. The invisible cartel is harder to detect and harder to challenge because its members sincerely believe they are doing good work. The problem is not their sincerity; it is the structural insularity that their sincerity produces.

From Cartels to Coercion

The practice of coercive citation — in which journal editors or reviewers demand that authors cite specific papers — is the cartel logic made explicit. In coercive citation, the power asymmetry of the peer review process is leveraged to extract citations as a condition of publication. The practice is widespread but rarely documented: authors comply because the cost of resistance — rejection, delay, reputational damage — exceeds the cost of compliance. Coercive citation is the cartel's enforcement mechanism: it converts the threat of exclusion into a guarantee of citation flow.

The response to citation cartels and coercive citation cannot be merely technical. Better metrics, detection algorithms, and editorial policies can help, but they cannot address the fundamental incentive structure. As long as citation counts are treated as proxies for research quality, the incentive to inflate those counts will persist. The only durable solution is to change the metric: to evaluate research by its epistemic contribution rather than by its network centrality. But this requires a transformation in how scientific communities judge quality — a transformation that the communities most captured by cartels are least likely to undertake.

The existence of citation cartels is not a scandal. It is a theorem. In any system where a countable metric is used as a proxy for quality, the metric will be gamed. The scandal is not that researchers game the system; the scandal is that the system is designed to be gamed, and that the designers — universities, funding agencies, and journals — profit from the game. Citation cartels are not the disease. They are the symptom of a metric-driven epistemology that has forgotten what knowledge is for.