The Institutional Boundaries of Academic Retraction: Evidence from the Technical Diffusion of mlncRNA Research

Purpose This study examines whether academic retraction, a key mechanism of knowledge correction, can effectively cross institutional boundaries to inhibit the technical diffusion of flawed research within the patent system. Design/methodology/approach Using a sample of 2,538 retracted papers and 3,543 matched controls in the mlncRNA field, we construct a “paper–patent–patent” citation network and conduct a matched dynamic comparison of patent citation trajectories around retraction events. Findings Retraction significantly suppresses the “adoption stage,” reducing the probability that a paper enters the patent system by approximately 66 %. However, once knowledge is integrated into patents, its “reproduction stage” (patent-to-patent diffusion) remains largely unaffected by the original retraction signal. These findings are consistent with the existence of institutional boundaries that limit the downstream corrective reach of academic retraction. Research limitations First, findings from the mlncRNA field may not generalize to other scientific disciplines. Second, the study focuses on formal citations and may therefore overlook informal knowledge spillovers in industrial R&D. Practical implications The study suggests that to mitigate the risk of flawed science in technology, integrated information-sharing platforms are needed to link academic retractions directly with patent examination and monitoring processes. Originality/value This study provides systematic evidence on the stage-specific efficacy of retractions across the science–technology interface, highlighting the need for integrated cross-institutional governance.