Leader: Watch the devilish details

Some 23 years after its first incarnation, the research assessment exercise has been finally laid to rest. But not before it gave us months of speculation, endless debate, the agony of who was to be entered and who was to be left out, game-playing, wrangles over calculations of the number of staff entered, profiles and, of course, the delightful pockets of excellence. It was a hellishly difficult time, but we were dealing with the devil we knew.

Now it is time to move on, and the RAE's successor, the research excellence framework, is currently being fashioned for implementation in 2014. The academic community wants improvements, but nothing too radically different.

What it has made plain is that the system must be at heart robust and fair, where critical decisions on quality rest with experts in the field rather than on a set of numbers applied blindly. As one vice-chancellor puts it: "Everyone is happy as long as the words 'peer review' are there." But robustness has its price, and there can be no expectation of a significant reduction of the administrative "burden". If universities want the job done properly and comprehensively, they will have to carry the costs.

For the REF, the devil was always going to be in the detail, but what is emerging from the Higher Education Funding Council for England is something that seems to satisfy many people, or at least avoids the worst horrors of the Treasury's original proposals in 2006. It is an evolution of the old assessment setup rather than a radical new metrics system, what some are already dubbing an "RAE with numbers".

Of course, you can't please everyone, and there are concerns from some quarters that the funding council is being bullied into producing something that is just the RAE in a different shade and with the same bureaucratic burden.

There is also the delicate matter of research concentration. While the final RAE identified widespread pockets of excellence, John Denham, the Universities Secretary, has made it clear that cultivating these at the expense of concentrating funding to ensure a world-class system of research is not an option. But how the REF as it stands can deliver this is unclear, with some kind of collaboration between pockets and big players the only likely solution.

The open way that Hefce is consulting, involving academics who are leaders in their field, is admirable. No revision of the RAE has ever had so much time and effort devoted to it. In the autumn, when the proposals are laid out in full, all academics will be able to have their say; and the response of subject associations and institutions will be key to this process.

The funding council must listen and learn, but it has to get this right. The stakes - namely the future of quality-related (QR) research funding - are too high. The simple risk is that the QR budget so vital to universities for investing in new areas of work, infrastructure and the training of new researchers will shift to the research councils. While the dual-support system of funding is by no means perfect, the QR element does allow universities true autonomy over how they spend research income and is crucially a vital protector of blue-skies research. A future where the academics of tomorrow do not have access to the level of QR enjoyed by scholars today would be a very bleak one indeed.

ann.mroz@tsleducation.com.