Value for money is always important, but particularly so in times of increasing financial strain across the UK in general and the Higher Education sector in particular.  This report presents the results of a value for money assessment using the public domain outputs of REF 2021, the UK Research Excellence Framework.  The public domain data includes both output measures and also information on the income that generated this research, thus the basis for a value-for-money assessment.

The analysis was carried out for the Computer Science and Informatics area (SP11 of REF).  This is an area that is dependent on financial input but where there are few major equipment expenses which might unduly skew the results; so, it is likely to give a good indication of trends across REF areas, especially in science and technology.

REF has multi-faceted outcome measures, so several forms of summary metric were used and value for money calculated for each.  The results of all follow a roughly similar pattern.  If ranked by raw outcome measure the top institutions were well known leaders (Imperial, Oxford, etc.), with post-1992 institutions at the tail of the league table  However, when a value for money measure is used the situation is fully or largely reversed, with post-1992 institutions on the whole offering the best value for money in terms of REF measures of research output per pound of investment.  The value for money differences between top and bottom of the tables are quite large, between ten and thirty to one depending on the precise measure.  These results are in accord with an earlier HEFCE value-for-money study in the late 1990s and analysis of NSERC funding in Canada in 2013.

As with any data analysis, the results should be treated with caution taking into account other contextual factors, but the general lesson is clear.  While the ‘obvious customers’ do produce the greatest volume of high-quality research, it appears that funding directed towards these high-achieving institutions is beyond the point of diminishing returns. Channelling more funding towards traditionally less research-intensive parts of the sector could achieve greater quality research for the country as a whole.

Furthermore, a better understanding of the factors that determine how these lower-ranked institutions achieve such good results with limited income may help improve research outcomes across the whole sector.