Regulatory impact assessment (RIA) has emerged on the political agenda of the member states of the European Union (EU). RIA is the cornerstone of programmes for ‘better regulation’ and ‘good regulatory governance’. This article argues that the diffusion of RIA has not produced convergence. The lack of convergence is explained by the different political contexts – measured in terms of types of bureaucracy, government's capacity to handle distributional conflict, policy processes, and actors' preferences. These findings stand in contrast to the explanation for the adoption of regulatory policy tools based on the need to protect efficient markets. EU governments adopt RIA for a variety of reasons, and consequently the outcomes of the diffusion process differ markedly. Further, the evidence shows that ideas can travel more easily than the content of policy tools – a result at odds with the notion that changes in policy ideas are the most fundamental type of policy change. However, when ideas travel lightly, they may well stimulate the introduction of some instruments that are similarly labelled ‘impact assessment’, but RIA practice may exist only on paper, and in some cases basic assessments of administrative burdens are disguised under the RIA label. The diffusion of a common RIA ‘bottle’ has not produced similar ‘wines’.
Stored only when both papers are in the corpus — 346 citations exist upstream.