How can governance change research content? Linking science policy studies to the sociology of science

Although science studies routinely assume that governance can change research content, we know little about how it does that. Science policy actors, commercial interests and civil society actors utilise a variety of governance instruments and processes to change the directions or to improve the quality of research. Both can be achieved only by making researchers select different goals and approaches, that is, change the content of their work. A recent review returned very few studies that investigated the impact of governance on research content and even fewer that convincingly established causality (Gläser and Laudel 2016). Empirical studies of competitive project funding focused on the validity and reliability of selection procedures and largely shunned the investigation of pre-emptive adaptation of project content or effects of selective funding on the dynamics of research fields (ibid: 122–125). The governance of emerging fields has been studied with the fields’ growth as the only dependent variable approximating content (ibid: 126–129). Most studies of the effects of performance-based funding limit themselves to changes in publication behaviour, whose causal ascription to performance-based funding can charitably be described as tenuous (ibid: 129–134; Gläser 2017). Finally, studies of academy–industry links have produced some interesting results on the impact of such links on the diffusion of knowledge (Evans 2010a; b) and on the increased likelihood of findings that are consistent with an industrial funder’s interests (Krimsky 2013) and on the impact of such links on the diffusion of knowledge (Evans 2010a; b). However, the empirical evidence of many studies is contradictory, which has led to a call for more in-depth studies of academy–industry links (Krimsky 2013). Most of these findings do not lend themselves to theoretical generalisation because they draw on specific cases in particular fields. More importantly, theoretical progress cannot be achieved with the current implicit division of labour between subfields of science studies. Science policy