Universities spent about ?4,000 for each researcher they submitted to the research excellence framework, a report has revealed.

The estimate of institutions¡¯ own total spend on the REF exceeds ?230 million, of which ?55 million went on preparing impact statements and ?19 million for panellists¡¯ time.

The cost to the four UK funding councils is estimated at ?14 million, according to the report by policy advisory group Technopolis, putting the total cost of REF 2014 at an estimated ?246 million. This is almost four times that of an adjusted estimate of the cost of the previous research assessment exercise in 2008.

It is considerably less than a?previous estimate of ?1 billion?for the REF made by academic Robert Bowman, director of the Centre for Nanostructured Media at Queen¡¯s University Belfast. But a previous assessment by Rand Europe correctly?calculated the cost to universities of articulating impact at ?55 million.

ÖйúAƬ

The review ¨C?REF Accountability Review: Costs, Benefits and Burden ¨C?looked at the institutional costs of the REF for a ¡°carefully constructed¡± sample of 20 universities, which was informed by a survey of REF managers and a series of interviews with pro vice-chancellors for research.

It found that universities spent about ?212 million collectively on the submission process, giving a cost per researcher of ?4,000.

ÖйúAƬ

But the report cautions that the ?212 million figure may include ¡°double counting¡± as universities had difficulty distinguishing between REF-related costs and those of ¡°business as usual¡±, as well as separating the costs of impact submissions from the rest of the REF.

To better compare the cost of the REF with that of the RAE, the report authors adjusted the RAE cost estimate to ?66 million in 2014 prices.

David Sweeney, director for research, education and knowledge exchange?at the ÖйúAƬ Funding Council for England, said that the ¡°substantial increase¡± in overall expense reflected the cost of improvements in the REF compared with the RAE to account for individual researchers¡¯ circumstances and ¡°making a success of impact¡±.

¡°We knew [these elements] would increase the costs, but we thought it was worth it,¡± he told?Times ÖйúAƬ.

The report adds that the total cost of the REF is less than 1 per cent of the ?27 billion of public money spent on research in the UK over a six-year period.

¡°In the context of dual support, the total cost amounts to roughly 2.4 per cent of the ?10.2 billion in research funds expected to be distributed by the UK¡¯s funding bodies in the six years 2015-16 to 2020-21,¡± it adds.

ÖйúAƬ

A recent report on the cost of quality assurance by consultancy KPMG found that universities?spend about ?1 billion a year, or almost 8 per cent of teaching budgets, to ensure that academic standards are upheld.

The Technopolis report also found that some institutions ran mock REFs to make sure that they made the strongest possible submission, which drove up costs. But the ¡°main cost driver¡± was the research output element of the assessment, as universities took time to review and negotiate the selection of staff and publications.

ÖйúAƬ

Universities found that a ¡°disproportionately costly element¡± of the REF was the treatment required for staff with individual circumstances, although institutions welcomed the changes to accommodate such staff.

Derek Sayer, a professor of cultural history and well-known critic of the REF, said that the 2.4 per cent cost as a proportion of quality related research funding was ¡°misleading¡±.

This was because some of the money went into funding streams for charity support, business research and national research libraries, which is not allocated on the basis of REF. Taking this into account a more correct figure would be ¡°close to 4 per cent¡±, he said.

Kieron Flanagan, senior lecturer in science and technology policy at the University of Manchester, said that there were some places where the cost of the REF was ¡°more nuts than others¡±.

¡°Northern Ireland subscribes to the REF to move funding between two institutions. So there is a real question about what the point is,¡± he said.

Meanwhile, Richard Black, pro-director (research and enterprise) at Soas, University of London, said that he did not find the Technopolis analysis helpful because it assumed that without the REF, universities would not need to plan for research. ¡°Much of the planning and monitoring that is done for the REF is valuable in its own right,¡± he added.

ÖйúAƬ

?holly.else@tesglobal.com