Universities are producing about 25 per cent more PhDs in science and engineering than the United States economy needs, according to a study by the Rand Corporation and the Institute for ÖйúAƬ Research at Stanford University.

This is far higher than anyone thought, because it compares numbers of specialists being produced with jobs for which they are qualified. It does not look simply at whether they have work, however humble.

The overproduction of doctorates appears to be highest in computer sciences, say the two economists who conducted the research, Charles A. Goldman at Rand, and William F. Massy at Stanford. Twice the number of computer science PhDs are being produced as are needed.

The surplus of doctoral computer science degrees being awarded over the number of those who acquire desirable jobs in the field is 50.5 per cent.

ÖйúAƬ

In other areas of science and engineering, the job prospects are better, but the surpluses of supply over demand are still large. In physics, chemistry and mathematics, for example, the surplus is 31.5 per cent; in chemical engineering 26 per cent; in mechanical engineering 44 per cent; and in the geological sciences 23 per cent. Psychology has the lowest surplus: 4 per cent.

The question is what should be done about all the unemployed or underemployed PhD holders. Many in the universities argue for more federal government funding for research to mop up the excess capacity.

ÖйúAƬ

But, after experimenting with mathematical models, the authors decided that would be the wrong answer. More money for research would, indeed, provide jobs for professors. But it would also produce even more PhDs.

What needs to happen is for the universities to cut back on the production of PhDs. The study, funded by the Alfred P. Sloan Foundation with federal money, included interviews with heads of department about how they decided on the numbers of graduate students to recruit. Department heads said they calculated how many new teaching assistants they would need to staff their undergraduate classes and how many new researchers for professors' research programmes.

Then they admitted PhD students. "When there are ever-increasing resources to be put into the system in the form of research money or ever-expanding undergraduate enrolment, then the system stays in balance and the numbers of opportunities are sufficient to absorb new PhDs," said Dr Goldman.

"But if resources are failing to grow, there will be an oversupply of PhDs. That is what is happening at the moment. Resources have levelled off."

ÖйúAƬ

A task force set up by the American Chemical Society and headed by David Lavallee, provost of City College of New York, reported finding "an annual oversupply of chemistry PhDs in the workforce of between 250 and 400".

The unemployment rate for chemists is put only at 2.5 per cent. Many of those who appear as employed in the figures are thought to be working in temporary postdoctoral appointments as teaching assistants, usually for two years at a time and on relatively low pay.

There is evidence that the tight job market has already affected demand. The American Institute of Physics found that the number of first-year physics graduate students had dropped by 6 per cent for the second year running. James Madison University in Harrisonburg, Virginia, has abolished its physics degree. All ten physics professors have been fired.

The outlook is not much better for mathematicians. In a survey, the American Mathematical Society found an "alarming" 12 per cent of those receiving PhDs between July 1, 1990, and June 30, 1991, to be unemployed and looking for work. That was double the rate for the year before.

ÖйúAƬ

But the National Academy of Sciences has a slightly different point of view. In a survey published earlier this year, Reshaping the Graduate Education of Scientists and Engineers, it acknowledged "indications of employment difficulties". More people were seeking jobs in academia than there were available positions, it said.