Physical scientists may not yet be an endangered species, but a report published this week warns that careful nurturing will be needed to ward off their eventual extinction. Between 1995 and 2000, although the total number of British graduates rose by 12%, enrolment in physics and engineering courses fell by 7% and enrolment in chemistry dropped a whopping 16%. Gareth Roberts, president of Wolfson College in Oxford, launched SET for Success, his review of the supply of highly skilled scientists and engineers in the United Kingdom, on Monday, saying that the review committee had "identified a number of serious problems in the supply of people". These problems, added Roberts, are "not equally spread, being concentrated in the fields of maths, engineering, chemistry, and physics". Life science is not off the hook, however: Fields such as molecular genetics are "becoming increasingly dependent" on highly numerate people, he pointed out.

Welcoming the findings of the Roberts report, which was commissioned last year by the Treasury, Minister for Science and Innovation Lord Sainsbury promised that "we in government will consider the recommendations as part of this year's spending review". He confirmed that the stakes are high. The shortfall of scientists is "almost universal", with countries such as Germany and Japan also seeing precipitous drops in the numbers of students in the physical sciences. "The country that gets this right", pointed out Sainsbury, has a real opportunity to strengthen its position in high tech industry.

The report calls for action throughout the educational supply chain: Fully one-third of Roberts's 37 recommendations relate to schools. For example, physical science teachers are rapidly heading towards retirement--more than a quarter are age 55 or over. And with a median class size of 30 in England, pupil-to-staff ratios are already "far too high", Roberts says--particularly when it comes to practical work.

Among the report's more novel suggestions, and one aimed squarely at redressing this teaching shortfall, is a recommendation whose implementation Roberts describes as "very urgent": Undergraduate and postgraduate students should be recruited to work as teaching assistants in schools. "In many ways", he admits, the idea is not new, since "many students are already assisting in schools on a voluntary basis" through programmes such as Researchers in Residence. What is new is that Roberts says they should be paid for their work in schools, because it is a better use of their expertise than "filling supermarket shelves". Roberts sees the scheme as having benefits all around, by "assisting hard-pressed teachers", helping students improve their transferable skills, and exposing children to enthusiastic young scientists.

Pay at all levels also comes in for scrutiny, and the report recommends that teachers, PhD students, postdocs, and academics all get raises. However, salaries should not necessarily be created equal: The report says that head teachers and governing bodies should "use all the pay flexibility at their disposal" to recruit and retain teachers in shortage fields. Meanwhile, Roberts himself asserts that "universities must differentially increase the salaries of scientists and engineers, particularly those engaged in research of international quality".

"It is vital that PhD stipends keep pace with graduate salary expectations", said Roberts, and the report recommends that Research Council maintenance grants be raised, over time, to the equivalent of the average, tax-free, graduate starting salary. This currently stands at just over £12,000, in stark contrast to the £7500 minimum stipend for Engineering and Physical Sciences Research Council students. In addition, Roberts wants the Treasury to hand extra cash to the Research Councils so that they can extend stipends from their current 3 years to 3½ years, which is now the average amount of time it takes to complete a PhD in the UK. As well as allowing students more breathing space he "would wish students to use the additional time to improve their transferable skills, and also gain some work experience". The report's recommendation is that, for both PhD students and postdocs, at least 2 weeks of dedicated transferable skills training should be provided each year.

As chair of the Research Careers Initiative, which was set up to monitor progress in the good management of contract research staff, Roberts holds the plight of postdocs close to his heart. Starting salaries for this group "should be raised in the near term to over £20,000 [per annum]", he asserts in the report, as opposed to the less than £18,000 new research staff can currently expect. Moreover, Roberts would like to see more support for young researchers seeking to get a leg up to a permanent academic post. Praising the Royal Society's University Research Fellowships, which offer up to 10 years' support for a small number of promising young researchers, Roberts would like to see large increases in the numbers of this kind of grant.

Consequently, his report asks the government to provide funding for 200 new 5-year fellowships each year, to be administered by the Research Councils. Outlining his vision for these fellowships, which he said should be "focussed on shortage areas", Roberts suggested that fellows should spend the first 3 years concentrating on producing high quality research, acquiring in years 4 and 5 the professional skills they need to become lecturers. However, his hope is that universities would be quick to spot talent and would offer permanent employment to these high fliers "long before the 5-year period is up."

It's not just the public sector that comes in for criticism on pay. Responding during the report's consultation period, many science-based businesses lamented the difficulty they have in recruiting high quality scientists. But "other sectors from which there is strong, and growing, demand for the skills and knowledge of science and engineering graduates tend to offer more generous pay and more attractive careers structures", points out the report. "For too long in the physical science sector", said Roberts, "there has been an apprenticeship culture that we don't see in other sectors". Although salaries have begun to pick up in recent years in response to the shortage of scientists, the message to companies engaged in R&D is "must try harder".

Sadly, as Peter Cotgreave, director of the pressure group Save British Science, points out, "the government cannot tell companies to pay their R&D staff more". They have been pushing them in the right direction, however, he says, by creating tax credits to encourage companies to up the amount they spend on R&D, which currently averages 1.8% of gross domestic product. The upshot of the extra investment "might be that people who do R&D might get paid more", he hopes.

While broadly welcoming the report, particularly its recommendations on pay, Cotgreave is disappointed that "it doesn't put any figures on how much it's going to cost". This, he says, will make it very difficult to judge whether the Treasury truly stumps up all the cash that is needed. Stump up they must, however, or the United Kingdom is "seriously in danger" of no longer having a world-class science base. Although it may be possible to "hide the problem" for some time longer in the universities, school teaching "is already at an absolute crisis point", he says. Some 66% of the people that teach physics are not trained in physics, he states, and he is skeptical of the report's recommendation for more continuing professional development to fill the teaching void. In a school that is already short of teachers, he wonders, how do you find a teacher a half day off for training?

So is it all doom and gloom in British science? Not entirely. Sainsbury is convinced that the United Kingdom is seeing "the beginnings of a brain gain." And Roberts says that in terms of both quality and quantity of graduates, we're doing okay. Overall numbers show that only France, New Zealand, and Finland produce more science and engineering graduates per head of population than Britain. The problem--as Roberts sees it--lies in the balance of physical to biological scientists. While the past 5 years have seen marked drops in the numbers of the former, the number of biological sciences graduates has increased over the same period by 49%, no doubt fueled by excitement over advances in genetics. But as Roberts so rightly points out, these hot fields can only prosper with the support of able physicists, chemists, and engineers.