The international landscape, and the politics, of metascience
“Metascience is political,” according to a new report from the Research on Research Institute. In particular, its vocabulary has been adopted by the American government “in ways its originators may not be able to control”.
Referring to recent developments in the field such as the Trump administration’s push for a new “golden age” of science, and the recent Kyoto vision declaration, the report notes a familiar tension for any field which seeks to identify and resolve problems in the research system:
“Strong claims about systemic failure can help to mobilise attention where established institutions have become complacent, and can provide a language through which otherwise dispersed concerns become politically actionable. Yet the same claims may also contribute to polarisation if they become identified too closely with one political constituency or are presented as establishing a singular diagnosis from which a particular institutional prescription follows.”
Caution about political cooption aside, the report finds a growing international ecosystem, presenting and comparing the work of teams in eighteen different countries who have mapped domestic metascientific activity.
Alongside the report and country-specific case studies, a new metascience atlas plots 476,154 pieces of academic work from the different corners of the landscape, encompassing everything from open access publishing to the philosophy of science.
While noting that the metascience label is used to refer to many strands of work with long histories, it is argued that the new terminology – and the networks and institutions forming around it – represent a genuinely new era of metascience entering mainstream science policy. As authors James Wilsdon, Josie Coburn and Seunghyun Lee put it in an accompanying Research on Research Institute blogpost, “what is genuinely new is the travelling label, the international networks and the appetite for experiment.”
Various factors are found to be behind the recent worldwide growth, including philanthropic backing, better open research information, the rise of AI both as tool and object of study, and even reform efforts based on social justice, reckoning with questions such as “who research serves and who decides how it is judged”.
The report’s authors also sketch out three possible futures for where metascience might go: institutionalisation, marginalisation, or capture. The latter is perhaps the most concerning: “indicators as instruments of control rather than learning, ‘rigour’ as a warrant for defunding inconvenient fields, transparency deployed asymmetrically.”
For the UK metascience unit – a joint UKRI-BIST team which is presented as one of the most mature examples of metascience work currently taking place – the report presents two tests for the future. First, whether it can maintain budget continuity across at least one further spending review, and second, whether it can avoid a “high-profile indicator scandal.”
The unit has said that its key area of focus in 2026–28 will be science systems in the age of AI; it has recently set out further thinking on the development of AI-driven indicators of research quality.