Citation cartels help mathematicians-and their universities-climb the rankings
Gaming of citation metrics and university rankings is drawing comparisons to SEO spam, with reports of “citation cartels,” coerced citations in peer review, and inflated author lists used to boost institutional prestige rather than scientific quality. Commenters link this to Goodhart’s law and broader perverse incentives in academia, arguing that when careers and funding are tied to simple quantitative measures, those measures get systematically abused. Some call for abandoning or radically downplaying citation-based rankings, strengthening replication and integrity checks, and funding alternative publishing models that prioritize reproducible, meaningful work over impact scores.
Citation Cartels & Metrics Gaming
- Many see citation cartels as a predictable outcome of metric‑driven evaluation and university rankings.
- Comparisons are drawn to SEO spam and PageRank gaming: link farms ↔ citation rings.
- Some note this behavior has existed for decades; what’s new is scale and centrality to rankings.
Goodhart’s Law & Misused Metrics
- Repeated reference to Goodhart’s law: once citations become a target, they cease to be a good measure of impact or quality.
- Citations can be high for bad or refuted work, or simply because many people work on a trendy topic.
- Several argue citation counts mainly serve administrators and rankings, not researchers.
Peer Review, Coercive Citations, and Authorship
- Multiple anecdotes of reviewers “suggesting” irrelevant or self‑serving citations, and editors nudging authors to cite their journal.
- Coercive or political authorship (especially in medicine) and inflated author lists are described as common.
- Some defend citation suggestions as often legitimate pointers to relevant prior work; experience varies strongly by field and venue.
Replication and Research Quality
- Strong claims that most papers in many fields cannot be replicated; others call this exaggerated, pointing to domain differences.
- Mathematics is often cited as an exception (results are proof‑checked), though others note nontrivial errors still appear.
- Physics, CS, and engineering are portrayed as more constrained by real‑world validation, but still subject to questionable analysis and poor code.
Discipline Differences & Fragmentation
- Very small specialized “clubs” control niche areas; they review each other’s work and can set local norms, good or bad.
- Some fields (AI/ML) are swamped by volume, leading to under‑citation or rediscovery; others (psychology, some biomedical areas) are seen as particularly fragile.
- Debate over how much pure math actually feeds applied work, and how often theory is disconnected from application.
Rankings, Prestige, and Proposed Fixes
- Rankings based on citations are widely criticized as “obviously gamable” yet still influential.
- Some argue prestige institutions have more to lose and thus cheat in subtler ways; others see blatant gaming in lower‑tier systems with rigid state metrics.
- Suggestions include: abandoning quantitative rankings or taking them very lightly, open preprints, more replication and integrity work, and possibly independent “integrity institutes” — though there is skepticism about incentives and effectiveness.