GRADUATE EMPLOYABILITY: MEASURING THE PERFORMANCE OF UK HIGHER EDUCATION

In December 1999, the UK government introduced a first wave of university performance indicators. A further wave will be published in 2001, including an indicator of students' employment outcomes. These indicators are intended both to meet the needs of prospective students for better information and to encourage improved performance in the higher education sector. It also seems inevitable that they will be incorporated into the funding formula for higher education.

Such indicators should be designed with great care, according to Jeremy Smith, Abigail McKnight and Robin Naylor, writing in the latest issue of the Economic Journal. In particular, it is vital that evaluation of the performance of higher education institutions adjusts for relevant differences in their characteristics: failure to make such adjustments could lead to very different and potentially misleading institutional rankings.

To date, published performance indicators are based on aggregate university-level data with relatively minor adjustments for differences in university characteristics. This represents a significant improvement on league tables based on raw data that make no allowance for different contexts of universities. Still, much more should be done. These researchers propose an approach to the construction of university performance measures based on the analysis of individual-level data. Potentially, this enables much finer adjustments to take account of different university characteristics.

The researchers' focus is on employment-related performance measures and their analysis shows that the probability of unemployment or inactivity six months after graduation is influenced strongly by the individual's class of degree, by degree subject studied, by prior qualifications and by social class background. For example, a male student with a first class degree is 7.5 percentage points less likely (and a student with a 2:2 is 9 percent more likely) to be unemployed or inactive than a student with a 2:1.

These factors also influence the probability that the individual will be employed in a graduate, rather than in a non-graduate, occupation. A higher degree class raises the unconditional probability of further study but lowers that of employment in a graduate occupation. The type of previous schooling has no significant influence on the probability of unemployment or inactivity, but does affect other outcomes for males: having attended an independent school lowers the probability of further study ( by 2 percentage points) and raises that of employment in a graduate occupation (by 3.5 percentage points).

With respect to university performance, there are large movements in the rankings of individual universities according to whether or not the measure of performance is adjusted to take account of differences. Although the adjusted and unadjusted bases identify similar sets of top and bottom universities, the rank positions of universities in between are not well determined. Hence, it would be inappropriate to have a funding formula that was sensitive to university rank positions.

Finally, there are very big differences by gender in the ranking of universities with respect to the probability of unemployment or inactivity. This underscores the fact that any performance measure should be regarded as only indicative of outcomes for the average student: there is likely to be significant variation in the ranking of universities around the average. Prospective students should be advised not to follow performance rankings slavishly.

Note for Editors: ' Graduate Employability: Policy and Performance in Higher Education in the UK' by Jeremy Smith, Abigail McKnight and Robin Naylor is published in the June 2000 issue of the Economic Journal. The authors are at the University of Warwick, Coventry CV4 7AL.

For Further Information: contact Robin Naylor on 024 76523 529 (email: robin.naylor@warwick.ac.uk); RES Media Consultant Romesh Vaitilingam on 0117-983-9770 or 07768-661095 (email: romesh@compuserve.com); or RES Media Assistant Niall Flynn on 020-7878-2919 (email: nflynn@cepr.org).