Here are the dates of foundation of the precursors of the major British government funding agencies for science and ÖйúAƬ: Medical Research Council, 1913; Science and Engineering Research Council, 1916; University Grants Committee, 1919. Here are someUnited States dates of foundation: National Academy of Sciences, 1863 (American Civil War, 1861-65); National Research Council, 1916; Office of Scientific Research and Development (parent of National Institutes of Health), 1941.

Britain and the US grew rich during the 18th, 19th and early 20th centuries under laissez faire. The funding of science was never regarded as a responsibility of government. Congress even tried to refuse the Smithsonian bequest in 1829, only accepting it in 1846. As late as 1940, the American federal and state governments were only spending $81 million a year on research and development, mostly for agriculture ($38 million) and defence ($26 million). But US economic or scientific growth was not hindered. In 1940, the private sector was spending $265 million on R&D, of which $31 million was pure science funded by the universities or the private research foundations, and some $23 million was basic science funded within industry.

It was war that persuaded the Anglo-Saxon governments to fund science, to create agencies to develop poison gas, atom bombs and tanks. But, once created, the science agencies agitated for their retention into peace time. George Ellery Hale, the chairman of the NRC, wrote: "during the war, the plans I had in mind looked forward to peace." To achieve their aim, scientists embraced the argument of "public good". Private funders would, it was claimed, not fund pure science because they could never capture the majority of the benefits for themselves. The taxpayer must, therefore, fund it.

Empirically, this argument still fails. Companies now fund ever more basic science. Of 911 large US companies surveyed for the years 1966-77, Griliches found that those that did basic science outperformed the ones that did not. Of 16 major US oil and chemical companies surveyed for 1960-76, Mansfield showed that the more basic science they funded, the greater their profits.

ÖйúAƬ

Why, therefore, does the "public good" argument fail? Certainly, most of the benefit of basic science does indeed escape the funder. Odagiri and Murakimi, surveying the ten major Japanese pharmaceutical companies ($13 billion joint sales in 1981), found that each individual company could only expect an annual return on its investment in R&D of 19 per cent -- its competitors could collectively extract a 33 per cent annual return on each company's research. Yet an annual return of 19 per cent is good enough to encourage anyone to invest.

Moreover, it takes a researcher to capture research. As Rosenberg has shown, the main value of researchers in industry is not the research they actually do, but the research they capture. The time they spend in libraries is more important than that spent at the bench. Since the best researchers cannot be retained unless they are free to research, industry pays them to do that. The more research they do, the better they are as scientists, the better they are at capture -- and the higher the profits.

ÖйúAƬ

David Edgerton, the historian of science, has argued that even if governments need not fund science for economic reasons, there remains a social problem. If all science is industrial, how can society monitor it? Would, for example, the tobacco companies have funded Richard Doll's discovery that cigarettes cause cancer?

But not all science is industrial under laissez faire -- witness the rise of the private, civic universities. The science historian Roy Macleod has shown that Britain's academic science base doubled every 12 to 20 years during the 19th century. There were 60 university science lecturers in 1850, 400 in 1900 and 1,600 scientists who enjoyed "congenial employment" (academics and others with scientific freedom) in 1914.

Consider the rise of great philanthropic foundations such as the Howard Hughes Medical Institute ($332 million spent in 1991) or the great charities (American Heart Association, $105 million spent last year). Britain's medical foundations and charities now outspend the government's MRC.

Laissez faire trains. Nineteenth-century Britain seethed with private technical and scientific education. No less than 700 mechanics' institutes for technicians were established between 1820 and 1840 alone. Their economics were those of today's US medical students. Individuals financed themselves through loans, which they repaid through the higher salaries their skills commanded. Further, those loans fostered greater social mobility than do our grant-funded, overly middle-class universities. Moreover, repeated reports (Board of Education, 1909; Balfour Committee, 1929; Malcolm Committee, 1929) confirmed that industry never lacked for trained scientists or technicians.

ÖйúAƬ

But the free market ensured that scientists earned proper salaries, unlike today's whose earnings are undercut by a state-funded over-supply, and by fixed, government-set salaries across ÖйúAƬ and research.

Government funding of civil R&D damages because it merely displaces private funding. Countries whose governments now tax them least, and fund R&D least -- Japan or Switzerland, for example -- enjoy the highest R&D budgets, and the highest gross domestic products per capita. Today's high taxes in Britain (taxes were less than 10 per cent GDP in 1913) have empowered the politicians only to displace the philanthropists and industrialists who created the pure science, academic plurality and technological leadership of the industrial revolution.

Terence Kealey is a lecturer in the department of clinical biochemistry, University of Cambridge.