AIS

The UK exercises in research quality assessment since 1986 have had ill-defined objectives. For the first exercises the results were simply «to inform funding», seeking value for money in an evaluative, regulatory state. That aim remained, almost word for word, until the Research Assessment Framework (REF) in 2014, when there was an additional articulated intent «to change behaviour». There had already been much changed behaviour over the years, perhaps unintended, apparently unexpected, as staff adapted their responses to changing «rules of the game» and the meaning they attributed to these. This article outlines the changing mechanics of successive exercises – the process means to ill-defined policy ends, and analyses the impact of design features which have affected the staff, and distorted institutional strategies of both policy development and control of delivery. The cumulative effect is to imbalance the system to favour a small élite, leading to isomorphism and funding concentrated to an extent that risks loss of diversity and stifling of challenges to established ideas, failing to recognise a variety of excellences.