An open-access article published on 21 July 2026 in Scientometrics presents findings that should unsettle anyone who still judges universities’ contribution to innovation by counting patents, academic inventors, and spin-offs. Entitled From green thesis to green invention, the study follows two decades of doctoral research conducted at German universities in Physics, Electrical Engineering, and Computer Science. The authors linked 44,720 doctoral theses with their authors’ publications, patent records, and the scientific literature cited by those patents, asking a question that conventional innovation rankings largely avoid: how much university knowledge reaches technology without ever appearing in a university-owned patent?
Among PhD graduates with publications included in the analysis, around 10% of those from Physics and 20% of those from Electrical Engineering and Computer Science appear as inventors on patents. Yet the publications of 36% of the former and 51% of the latter are cited in patent documents. The indirect connection between science and invention therefore appears roughly four times more often than direct participation as an inventor. In green technologies, the contrast is even more revealing: only 0.8% to 5.8% of PhD graduates appear as inventors on a green patent, while 4.5% to 17.6% have publications cited by green patents. https://link.springer.com/article/10.1007/s11192-026-05736-5
These findings also cast doubt on the standard interpretation of the so-called European paradox, according to which Europe produces excellent science but fails to convert it into innovation and economic value. The study does not prove that the paradox is a statistical fiction, particularly because it is limited to Germany and to two broad scientific fields. It does, however, suggest that part of Europe’s supposed innovation failure may be manufactured by the indicators used to diagnose it. If universities contribute chiefly by producing knowledge that others quietly convert into inventions, then evaluating them through patents and spin-offs is not merely incomplete; it is intellectually fraudulent. It erases the very process by which science becomes technology and then congratulates the indicators for finding so little.
PS - One bibliographic omission in the aforementioned paper deserves explicit mention. Its reference list contains no citation of a closely related study based on 4,460 Swedish innovations, which I discussed in an earlier post. That study found that most innovations were never patented, while patents captured barely 15% of the overall innovation signal, leaving the remaining 85% statistically invisible to conventional indicators. It is the same post that opens by recalling European Commission officials’ warning that Europe suffers from an “unhealthy obsession with patents.” https://pachecotorgal.com/2025/08/31/the-invisible-85-how-patents-create-a-false-economy-of-innovation-value/
