Official figures exaggerate the employment prospects of graduates in some of higher education's most popular subjects by failing to distinguish between different types of job, according to university league tables published this week.
An analysis of graduate destinations for The Times Good University Guide reveals sharp contrasts between courses. In almost half of the subjects surveyed, at least one graduate in five took work that did not require high-level qualifications.
Previous surveys, including the established measure of graduate employment, the First Destinations Survey, have shown more than 90 per cent of graduates in all but a handful of subjects working or studying within six months of completing a degree. But the figures mask wide variations when non-graduate occupations are excluded.
Degrees leading to the professions and other elite occupations maintain their strong employment record when graduates in low-level work are excluded. Dentistry, medicine and veterinary medicine all have near-100 per cent graduate employment and virtually no graduates in lower-grade jobs.
Subjects with a clear vocational path such as nursing also perform strongly.
But in many subjects, up to a quarter of graduates work in jobs deemed to be non-graduate. American studies has the highest proportion (34 per cent), with psychology, sociology, art and design, linguistics and media studies among those exceeding 25 per cent.
In this year's tables, jobs have been assigned as "traditional graduate" and "non-graduate", with a category of "graduate track" to cover skilled clerical and technical occupations that have become accepted destinations for graduates. The categories are based on research by Abigail McKnight, Toyota research fellow at the Centre for Analysis of Social Exclusion at the London School of Economics, conducted while she was at Warwick University's Institute for Employment Research.
The graduate category comprises traditional graduate jobs - doctors, lawyers, engineers, teachers and technical occupations - and occupations that have more recently been regarded as graduate jobs, including high-level sales and skilled clerical.
Graduate-track jobs require a high level of education but not necessarily a degree, and include "entry route" occupations and new areas of graduate employment. Examples include low-level management, such as purchasing and catering managers. The category also covers some skilled manual work.
Non-graduate occupations do not require high-level qualifications, and include low-level clerical jobs and manual jobs.
The researchers used typical entry requirements and the average level of qualifications held by employees to assign each occupation to a category.
Graduate and graduate-track occupations have higher average earnings than those classified as non-graduate. In an IER survey of those who left university in 1995, average annual earnings of graduates in graduate occupations three and a half years after graduation was about 5 per cent higher than for graduates in graduate-track occupations and about 35 per cent higher than for graduates in non-graduate occupations.
The research by Ms McKnight and her former colleagues at Warwick was cited by the Higher Education Funding Council for England in its Indicators of Employment report of April 2001. But Hefce disregarded type of employment because it considers that the available information does not give a full subject-based analysis or detailed data by institution.
The effect on institutional performance can be seen from the destinations data published on pages iv-v. Stripping out the non-graduate jobs has only a limited effect on institutions with large medical, dental and veterinary schools. But the impact on mainstream institutions can be dramatic. The University of Derby's employment rate plunges from 91 to 58.5 per cent when it is judged only for graduate and graduate-track employment.
The First Destinations Survey has limitations because it is a snapshot taken within six months of graduation. But the IER concluded that employment six months after graduation is related not only to employment prospects over three years, but also to the quality of job ultimately obtained.
• Subscribers, click here for the full 2003 University league tables