Closing the EP–Cap gap: a global spectrum of research-integrity governance

<title>Abstract</title> <p> The 2026 whistleblowing wave in China, in which a doctoral candidate ("Classmate Geng") used elementary statistical checks to expose fabricated data in high-impact journals by top talent-title holders ^{<bold>[1]</bold>} , reveals a generic vulnerability of cap-based research evaluation systems. When the expected penalty for misconduct (E[ <italic>P</italic> ]) falls substantially below the cumulative leverage of talent titles (w₁[Cap]), honest research ceases to be an evolutionarily stable strategy (ESS) ^{<bold>[2]</bold>} . Herein, I audit 12 jurisdictions with published sanction records and identify a four-region global spectrum: i) stable (Germany; E[P] ≈ w₁[Cap] <italic>via</italic> criminal referral and clawback); ii) marginally stable (USA, Netherlands, Switzerland, Israel); iii) critical belt (UK, Canada, Australia, Singapore, South Korea, Japan; cap high, (E[ <italic>P</italic> ] lagging); iv) collapsed (China; gap ~1 - 2 orders). Moreover, approximately 30+ further jurisdictions lack public sanction records (ND). The German model provides a portable benchmark: independent investigation, routine criminal referral, and guaranteed fund recovery. Differentiated prescriptions are accordingly derived for each region. The framework generalises: any system where cap weight w₁ exceeds marginal scientific return risks structural convergence to pseudo-scholarship. </p>