Recently, evidence has been accumulating that the innovative power of research is stagnating, despite the ever-growing size of the scientific community. In addition, a social divide between unquestioning belief and irrational skepticism, as evidenced during the Covid pandemic, reinforces the impression of a science in crisis.

A central starting point of this work is the idea that excessive division of labor and specialization could be a fundamental cause of this crisis. Despite the importance that thinkers such as Adam Smith and Emile Durkheim ascribed to the division of labor on in explaining scientific progress, modern economic findings on coordination costs have rarely been applied in the context of scientific institutions. If addressed at all, the topics are explored in highly specialized studies on topics such as cognitive diversity or bibliometric specialization. Given the prevalence of economic thought within science studies, this is surprising and may itself be evidence of a lack of coordination among increasingly isolated branches of science. Thus, a thorough discussion of division of labor and specialization as factors influencing scientific inquiry has been largely missing until now.

This work takes a first step toward addressing this gap by introducing a new dataset that documents the development of professorial denominations at twenty globally highly ranked universities over the last century. The data reveals a continuous increase in specialization, as well as thematic path dependencies, especially in European institutions. The results of a non-parametric conditional efficiency framework indicate a functional relationship between the degree of division of labor and specialization with the productivity of the universities studied. Findings across university clusters point to a beneficial effect of the US department system on productivity when compared to its European professorial chair model counterpart. Part of the observed crisis could thus be mitigated by monitoring coordination costs to avoid overspecialization and promote organizational forms that enable enhanced collaboration within research institutions. This work further contributes to the literature by proposing an alternative to the vague concept of interdisciplinarity, offering an institutional-based and more granular perspective that allows for monitoring and measurement.