Institutions as Process: Evidence, Expertise, and the Making of HM Treasury’s Green Book

This paper examines how economic techniques become embedded in public policy as institutionalised forms of expertise. Its central empirical object is HM Treasury’s Green Book, a manual that has guided cost–benefit analysis and public sector appraisal in the United Kingdom for over half a century. It is a major, highly persistent, and yet critically under-analysed channel through which welfare economics, alongside public choice theory, financial economics, and operational research, has been translated into everyday policy practice in the UK. The Green Book is widely seen as authoritative: it structures how public decisions are reasoned and justified, and what counts as credible evidence. Yet its status is analytically puzzling. Although it exhibits many characteristics associated with institutions, it fits none of the standard definitions comfortably. This paper therefore asks what makes the Green Book an institution, and what this reveals about how economic knowledge becomes authoritative within the policy process. I argue that the New Institutionalist literature falls short of providing a readily applicable definition of institutions, as it shares an underlying entity view of institutions across its branches. In contrast, building on constructivist accounts of facticity, this paper advances a process view of institutions, conceptualising them as ongoing accomplishments of producing, contesting, and reassembling meaning and authority. This allows me to understand the Green Book not as an institution because it possesses certain properties, but as the outcome of institutionalising practices grounded in institutional facts (Searle) sustained over time. Expertise, in turn, is not simply “used” by policymakers, but is itself constituted through institutional arrangements that define what counts as valid evidence. The Green Book demonstrates the context-dependency of expertise, as Treasury civil servants act as intermediaries between academic economics and departmental policy practice—selectively adopting, adapting, and discarding ideas in ways that shape long-term institutional logics. Empirically, the paper draws on archival research in UK Treasury records from the 1960s to the early 1990s, complemented by elite interviews, to trace how appraisal techniques evolved. The material is presented as vignettes illustrating different facets of institutionalisation. These speak to three dimensions of institutional facticity: perceived social purpose and recognition; constitution of practices and power dynamics; and temporal persistence. The analysis suggests, first, that appraisal techniques gained authority because they were recognised as fulfilling multiple purposes: imposing comparability, disciplining expenditure, and aligning with technocratic ideals. Second, their embedding constituted new evaluative arenas and forms of authority by defining what counted as credible evidence and setting the parameters of policy argument. Third, their durability was sustained by interpretive adaptation in response to political and epistemic challenges while preserving a logic of calculative rationality. By analysing the Green Book as an institutionalised form of evidence and expertise, the paper speaks to core concerns of policy process research: how ideas become embedded in institutions, how expertise is authorised and contested, and how institutionalised knowledge shapes actor behaviour and policy trajectories. It also reveals how the microfoundations of institutional design and public policy instrumentation operate through the everyday practices of civil servants and experts.