Universities will be encouraged to revert to traditional recruitment policies by government-backed performance indicators to be introduced from next year, economics experts warned this week.
Researchers at Warwick University have concluded that the employability of graduates - the key performance indicator ministers want funding chiefs to introduce - is most closely linked to the class of degree obtained and the subject studied. Which university a graduate attended has only a small effect on employability.
The findings will come as a blow to the government, which has told the Higher Education Funding Council for England that it must produce performance indicators relating to employment outcomes by next year for individual universities and higher education as a whole.
In this time, the funding council can only produce figures based on the first destination survey of graduates, which will fail to take into account all the characteristics that can influence graduate employment rates.
"If the government places a heavy reward on placing graduates in good jobs, it will give universities an incentive to take from the social classes that are successful," said Robin Naylor, who carried out the study with Jeremy Smith and Abigail McKnight of the University of Warwick.
Such a scheme would be in conflict with the existing initiative to reward institutions that recruit students from under-represented groups.
The team examined the employment records of 70,000 students who graduated from the old universities in 1993. The findings will be presented at a conference on Monday. The team hopes to repeat the exercise for 1998 graduates from all universities.
Dr Naylor's study looked at the effects of the university attended, the subject studied, the degree class, the sex of the graduate, and whether they were full-time or part-time, a mature student, married or single. It also examined social class, type of schooling and entry qualifications.
"Of all the things that determine labour market success, degree class was the most important. A man with a first-class degree is 10 per cent less likely to be unemployed than a man with a 2:1: a man with a 2:2 is 10 per cent more likely to be unemployed than a man with a 2:1. For women the effect was less pronounced but still there," he said.
"The subject studied also makes a big difference. For example, people reading medicine or law are the least likely to be unemployed on graduation," he said.
HEFCE is working on performance indicators for employability for the coming years. It intends to take into account university location, subject studied and sex of the graduate. It will also consider the findings of Dr Naylor's work.
"We have to take into account all sorts of characteristics," said John Thompson, who works on performance indicators for the HEFCE.
Another issue is whether graduate employment should be sorted into different types. Some senior figures believe that some types of job are more relevant to the outcomes of higher education than others: others argue against creating tiers of "graduate" and "non-graduate" jobs.
Dr Naylor said that although social class, like institution attended, had a relatively minor impact on employability, "it does have a big effect on the salary. High social classes get better-paid jobs."