John Kao is an innovation guru described as "Mr Creativity" by this newspaper a decade ago. Now he is concerned about America losing its global lead and becoming "the fat, complacent Detroit of nations". In his new book, "Innovation Nation", he points to warning signs, such as America's under investment in physical infrastructure, its slow start on broadband, its pitiful public schools and its frostiness toward immigrants since September 11th 2001—even though immigrants provided much of America's creativity. What America needs, he reckons, is a big push by federal government to promote innovation, akin to the Apollo space project that put a man on the moon

As head of California's Stanford Research Institute, Mr Carlson knows the strengths of Silicon Valley from first-hand experience. And yet here he is insisting that America's information technology, services and medical-devices industries are about to be lost. "I predict that millions of jobs will be destroyed in our country, like in the 1980s when American firms refused to adopt total-quality management techniques while the Japanese surged ahead." The only way out, he insists, is "to learn the tools of innovation" and forge entirely new, knowledge-based industries in energy technology, biotechnology and other science-based sectors.

It is natural to be sceptical of such dour arguments and calls for government action. After all, the United States still leads in innovation. Whether it is by traditional measures, like spending on research and the number of patents registered, or less tangible but more important ones, like the number of entrepreneurial start-ups, levels of venture-capital funding or the payback from new inventions, America is invariably at the top of the league. Indeed, the Council on Competitiveness recently concluded in a report that, by and large, the outlook is bright for America.

Yet the same council's innovation task force also gave warning that other countries are making heavy investments that threaten to erode America's position. It would like a big push in four areas: improving science, engineering and maths education; welcoming skilled immigrants; beefing up government spending on basic research; and offering tax incentives to spur "US-based innovation."

These are mostly sensible recommendations because they focus on those framework conditions and bits of infrastructure that the market would not provide on its own. Where such prescriptions tend to go awry is when they argue for specific subsidies or tax breaks for favoured industries (like supporting only "US-based" innovation in today's world of global creative net-works). After all, the Schumpeterian forces of creative destruction must be allowed to work their magic.

Resilience in the face of those disruptive forces gave Silicon Valley the edge over its nearest high-tech rival, Boston's Route 128 technology corridor. Both clusters were riding high until the personal computer and distributed-computing changed the market. Firms went through wrenching change, but those in northern California, like Hewlett-Packard and Xerox, emerged stronger than those near Boston, like Digital Equipment and Wang—which no longer exist. As Berkeley's Anna Lee Saxenian has shown, Silicon Valley's champions were nimble and networked but those on Route 128 were brittle, top-down bureaucracies.