This paper uses data from the submissions and ratings from the UK Research Assessment Exercise 2001 (data on staff selection, income, research groups, methodologies, audiences, and areas on expertise) to reflect on the shifts in public understandings of researcher accountability over the past few decades at the level of institutional (rather than individual) accountability. In particular, it looks at what has been described as a decline of professional, ethical and communicative modes of accountability in favour of more technical, managerial and bureaucratic ones. This was accompanied by a conceptual change, from accountability as responsibility and communicative reason to accountability as hierarchical answerability (with corresponding shifts in values, concepts of public good, and hierarchies of knowledge). In recent years, attempts have been made to develop generally acceptable hybrids between professional accountability (i.e., accountability that resides within education and research communities and relies on conceptions of intrinsic excellence and on professional ethics – Winch, 2001) and bureaucratic accountability (i.e. financial concepts of accountability and auditability, transposed to the context of professionals being held to account by third parties such as government and administrative structures or corporations – Biesta, 2004). The UK RAE may be described as one such attempt. However, rather than producing accountability regimes with wider legitimacy, it – and other initiatives along similar lines - has been criticised by commentators for having aggravated the “erosion of professional power from state governments and the corporate community” (Webb, 2005, p.191). Despite the RAE intending to recognise and draw together the standards of excellence characteristic to the education research communities and the measures of performance favoured by the state administration, arguably in the end the emphasis fell firmly on the latter. The data analysed in the paper suggest that, while effective at screening out poor quality research (due to the peer review processes at its core), the RAE also endangered pockets of expertise and emerging research cultures, through the cutting of funds for lower-rated institutions. This had to do less with the quality of the peer reviewing, and more with the managerial logic of the exercise, which presupposed: commensurability of research quality across sub-fields, types of institutions, and types of research cultures and communities; meaningful aggregation of quality at institution level; the value of competition and selectivity in creating quasi-markets of state-funded research; and the theoretical possibility of comprehensive assessments of submissions (as opposed, for example, to peer-reviewing a representative sample of publications, or to skimming through large submissions to gain a flavor of the overall “performance”). Many of these are deep-cutting issues that are unlikely to be addressed by simply replacing the RAE with a more quantitative assessment exercise, such as currently proposed in the UK.

The paper draws on work commissioned by the British Educational Research Association and aimed at gathering information about the distribution of educational research expertise across different types of institutions throughout the UK, post-RAE 2001. The strategies for data gathering and analysis were multifold, ranging from survey of education departments and overview of newspaper articles, to analysis of the authorship of academic journal articles after 2001, and analysis of the RAE 2001 submissions.

The paper argues that, in a post-RAE context (given plans to replace the current RAE with a more metrics-centred exercise), neither the reinforcement of targets, indicators, standards and techniques of managerial accountability, or the closure of academia to external scrutiny, is likely to be the way forward. Rather, what is needed is a restoration of the discursive, democratic and ethical dimensions of the relationship between research, the public, and policy.