University–industry collaboration (UIC) is now commonly accepted and encouraged via federal funding. Previous research on UIC documents its rise and recognizes its role in driving innovation and promoting economic growth. UIC manifests in various forms, ranging from industry sponsorship to fully collaborative research. The form that interests us here is joint research that leads to publications in journals. If a publication includes authors from both university and industry settings, then we consider the paper to be a UIC.

Although we recognize that UIC occurs in many fields, our study focuses on the biomedical and life sciences. This field has grown rapidly and fostered collaboration across disciplines and especially sectors like academia and industry. The biomedical and life sciences are ideal for focused study of collaboration, as the vast majority of biomedical publications in the U.S. are collaborative efforts made by multiple authors: Of the 5,968,019 publications in biomedical and life sciences published between 1980 and 2020 (indexed in OpenAlex), 5,103,096 (about 86%) are multi-authored.

Academic and industry science are governed by distinct institutional logics that shape how knowledge is produced, disseminated, and evaluated. These differences can generate both tensions and synergies when universities and industries collaborate in scientific research. Because UIC is oriented toward dual audiences in academia and industry, such publications may attain broader visibility and impact. Moreover, cross-sector recombination and the exchange of knowledge, tools, and expertise can promote novel combinations that enhance innovative outcomes. Therefore, we hypothesize that research resulting from UIC will achieve wider visibility and be more influential than research conducted solely within one sector.

To test these hypotheses, we pull bibliometric data from two sources: PubMed Knowledge Graph 2.0 and SciSciNet 2.0. We focus on papers published between 1980 and 2020 whose teams are involved in U.S.-based institutions. Based on authors’ affiliations, we distinguish publications co-authored by researchers from both university and industry (UIC: n=121,313), industry only (IC: n=62,130), and university only (UC: n=1,969,904). To operationalize the visibility of a scholarly publication, we employ citation count (the number of times a work has been cited from publication through 2025). To operationalize influence, we rely on Funk and Owen-Smith’s CD index (Disruption Index), which ranges from −1 to 1, where negative values indicate consolidation of existing knowledge and positive values indicate disruption of prior work as measures. Following other works, we dichotomize both outcomes to compare the top 5th percentile of each distribution (e.g., the most highly cited papers) from the rest. We use multivariate logistic regression to assess whether UIC is more highly cited and influential relative to both IC and UC.

We find that UIC publications are more likely to become highly impactful and disruptive, relative to both IC and UC. These effects hold once we control for publication year, journal prestige, open access status, journal topical breadth, number of references, team size, domestic collaboration, and number of institutions. Promoting UIC in team science can thus encourage influential research outcomes.