Nobel laurates and the role of the industry in the emergence of new scientific breakthroughs

Since the 1980s, many companies recognized for their major scientific breakthroughs (e.g., IBM, AT&T, etc.), cut their investments in fundamental research activities. In parallel, academics from public research organizations (PRO) and universities engaged more extensively with the industry through research collaborations. The conditions, determinants, and effects of academic engagement have been deeply analyzed. But, the extent to which major scientific breakthroughs of the last century have emerged either from (1) academics and researchers with no interaction with the industry or (2) from scientists interacting with the industry-either as engaged academics belonging to PRO or universities or as corporate scientistsare yet to be more systematically documented. To fill this gap, we explored the extent to which scientists from the quasi-complete cohort of Nobel laureates in Physics, Medicine, and Chemistry were interacting with the industry before their breakthrough discoveries. We designed a unique dataset of their ties with the industry based on affiliations review of 84,423 academic papers and applicant review of 5,207 patent families. First, we showed that one-fifth of the studied cohort of laureates was interacting with the industry before their breakthrough discovery. More importantly, this share is still increasing, mainly through academic engagement, while the share of awarded corporate scientists has remained stable since 1970. Second, we were able to analyze the effects of those interactions with the industry on the post-discovery period by comparing interacting and noninteracting with industry laureates' follow-on research works. While some scientific discoveries were partly made possible thanks to Nobel laureates' industrial partners, those laureates' follow-on knowledge works were not bound to their initial sets of partners. They experienced similar knowledge diffusion-to-industry rates than other laureates but higher academic production rates and diffusion-to-academia rates. Finally, we claim that the extent to which scientific new knowledge still emerges in relation to industrial contexts in modern science has been underevaluated and opens rooms for further research.