The relationship between scientific research and business is vital to Britain’s economic future. But using top research universities to promote economic development requires a subtlety and understanding that is not evident in the government’s science strategy. Indeed, odd though it sounds, by focusing on ideas as the means of creating wealth, we are making a big mistake. People create wealth, and we need to get the right people into the right places.

The link between scientific research and economic growth is well established and, in academic terms, Britain is well placed—much better than France, for instance, where Jacques Chirac has just announced the formation of an agency for industrial innovation. Britain has eight of the world’s top 50 universities, with scientific research a particular strength. Although it has only 1 per cent of the world’s population, it produces 5 per cent of the world’s scientific research, and achieves 12 per cent of all citations in scientific papers—more citations per pound of GDP than any other country.

But Britain’s spending on R&D at 1.9 per cent of GDP is well below the OECD average—comparing particularly badly with France, Germany and the US. The government’s goal is to raise it by 2014 to 2.5 per cent of GDP, requiring an extra £5bn a year to be spent by both public and private sectors.

In making the case for this kind of investment, the treasury’s science and innovation report observed last July: “Studies show that R&D delivers benefits by allowing an economy to do two things: understand and appreciate the value of others’ findings and results; and make new discoveries.”

But which of these “two things” should we concentrate on? It is clear that the government believes that the answer lies in making new discoveries in the hope of “commercialising” them. This is why the Higher Education Innovation Fund (HEIF) is to receive £90m in 2005-06 to stimulate enterprise from research, largely through university technology-transfer offices.

The government—or the relevant departments—believes in what has become known as the linear model of commercialising scientific research. This envisages the university, which owns a discovery, copyrighting, patenting or otherwise claiming it as intellectual property, then either putting it into a company (a “spin-out”), licensing it to industry or building a consultancy round it. Hence the role of technology-transfer offices in licensing, issuing patents and registering intellectual property rights.

The problem is that most of the academic research that has been done in this area—including some of the government’s own—suggests that this approach is misconceived. Wealth is created not by the exploitation of intellectual property but through the actions of entrepreneurs. Businesses do not engage with universities primarily in order to license intellectual property, but to recruit people. From the university’s point of view, licensing income is a trivial component of research income. Indeed, most US universities lose money on licensing.

Barry Bozeman, the leading US expert in technology transfer and public policy, observed in 2000 that there was “an emerging consensus that university and federal laboratory technology transfer have only modest potential for creating new businesses.” Moreover, a study commissioned by the British consulate general in San Francisco reported not only that “technology-transfer offices do not cover their own costs” but also that “direct donations from businesses typically far exceeded licensing revenues.”

That same British government study from San Francisco also hinted at what might be the most productive way of bolstering the link between scientific research and business. “All interviewees agreed that the most effective form of technology transfer was the migration of highly skilled people from universities to business. The technical knowhow that researchers carry with them can be significantly more valuable to businesses than the legal right to commercialise inventions.”

The Lambert review of business-university collaboration, commissioned by Gordon Brown, noted the view of business and universities that too many spin-outs were being created and that technology-transfer offices were not making money. Lambert made many other useful observations. But his report did not challenge the prevailing view that a successful strategy was all about putting enough money into R&D so that brilliant ideas could be developed, licensed and commercialised.

An alternative approach should be based on using universities to attract, and retain entrepreneurial and technical talent. This is the “people-centric” approach, and it is where the government should be concentrating its efforts.

To begin with, we need to find out where businesses get their ideas for new R&D-based products. Fortunately, Wesley Cohen, Richard Nelson and John Walsh asked just this question in 2002 as part of a Carnegie Mellon survey on industrial R&D. They also reviewed the contributions of university and government research labs—”public research”—to industrial innovation. Their survey showed that customers were the best source of ideas for business projects (named by 90 per cent of respondents), with the next best source being manufacturing operations (74 per cent). Public research comes second bottom on the list, named by just 32 per cent of respondents. So, the best business ideas do not come from the lab but from business and the market. Where the scientists come in is identifying the technology capable of meeting customer need.

The government’s idea-centric approach relies on ideas that emerge from the laboratory, where PhD students are studying whatever happens to be their chosen area of interest. The idea emerges whenever the academic discovers it—typically after three years of research—regardless of the market’s readiness, or business’s willingness to commercialise it. By contrast, the people-centric approach involves looking to business, customers and the manufacturing process for inspiration. Doctoral students look things up, drawing on any and all of the world’s scientific research. And the research is only tapped into when needed, when the market is ready.

It is odd, on the face of it, that the government should have built its policy for commercialising research on a technology-transfer model that, according to the published literature, only works for the biotech and pharmaceutical sectors. The explanation may lie in the fact that pharmaceuticals are a great British success story—we have the second largest drugs industry in the world behind the US—and it is one area in which we spend disproportionate amounts on R&D.

There is nothing wrong with the idea of trying to grow hi-tech business “clusters” out of universities, much as Sun Microsystems emerged from Palo Alto, California (with the eponymous acronym standing for Stanford University Network). And there is no reason why the so-called “silicon fen” of 900 tech companies that has emerged around Cambridge should not achieve something similar. But the key is to concentrate on entrepreneurs.

In Cambridge, we have the outstanding example of Hermann Hauser, an Austrian who came to Cambridge to do a PhD in physics because of the reputation of the Cavendish lab, and went on to found the computer company Acorn, which begat ARM Holdings. Hermann and these companies have sprouted a further 100 companies.

The idea that research universities should attract and educate the next generation of entrepreneurs is backed up by plenty of research—and by my own experience. From 2001-04, I was a full-time industrial visitor in the Cambridge University computer laboratory, and this experience convinced me that we could do much more to attract, retain and develop entrepreneurs. One way would be to offer scholarships to a small number of graduate students on grounds of demonstrable entrepreneurial ability rather than mere academic accomplishment. Another would be to incorporate more training in business management into computer science and other suitable degree subjects. Yet another—and perhaps the easiest to introduce—would be to offer bursaries to encourage would-be entrepreneurs to stay on locally after graduation, rather than having to get a business plan funded.

What entrepreneurs and business leaders want above all is to be able to hire the right people. In the world of computer science, technology professionals are the hottest properties, and places like Cambridge have lots of them. If you want these people to grow the kind of local technology cluster that might really boost the economy, and not drift off and get jobs elsewhere, you need to help them find relevant, often local, jobs. This is particularly important for doctoral students.

There are various ways of doing this, of which perhaps the most powerful is the creation of “industrial supporters’ clubs,” whereby companies pay a subscription in return for the right to recruit from each year’s crop of students. In Cambridge, this worked best when we invited every corporate member to deliver a three-minute pitch to a hall full of computer science students. We also invited them to exhibit at a jobs fair. Apart from the other benefits, this makes substantial amounts of money for the university department concerned.

The government has got one important thing right: research universities can indeed grow technology clusters and boost economic growth. But other than that, the treasury and DTI have lost their way. The way to build technology clusters is not to rely on the licensing or patenting of brilliant ideas and discoveries that emanate from the laboratory, but to enhance the flow of people—entrepreneurs, technology professionals and PhDs—from university to business and vice versa.