The EP–Cap Instability: Why Cap-Based Evaluation Systems Converge to Pseudo-Scholarship — Implications for Global Research Integrity and Journal Quality

<title>Abstract</title> <p> Cap-based research evaluation systems now dominate global science funding, yet their parameter vulnerability remains unformalised. Here we show, by extending a companion framework (Zhang, submitted to <italic>Science</italic> PF ¹¹ ), that coupled ( <italic>p,a)</italic> saddle–node dynamics permit endogenous decay of expected penalty even under frozen caps, <italic>via</italic> audit fatigue and sanction euphemisation. A twelve-jurisdiction scan anchored to Retraction Watch and PubPeer finds only Germany satisfies the stability condition E [P] &gt; w₁[Cap] + w₂[G]; the discounted present-value gap in foundational knowledge dwarfs short-term cap inflation. Beyond evaluation policy, the EP-Cap gap offers a parsimonious diagnostic of why leading journals in cap-heavy systems disproportionately absorb pseudo-scholarship—linking national incentive architecture to global knowledge-network integrity. A dissipative-structure analogy reveals epistemic-engine inefficiency and abnormal specific heat. Crucially, cap leverage converts almost exclusively through top-[ <italic>IF</italic> ] venues, making leading journals themselves the absorbers of the pseudo-phase — linking the EP–Cap gap to journal-level scientific quality. Three levers — clawback + criminal referral, raw-data escrow, cap–audit coupling — close the gap without abolishing caps. </p>