Put aside the polemic rhetoric and particular policy initiatives. There has never been a wider area of policy consensus in the industry policy position of Australian political parties.

Labor and the conservative parties agree that national prosperity depends on encouraging innovation and increasing investment in research and development. Even the Democrats are firmly on board.

R&D is a sacred cow, and no politician would dare suggest his commitment to its advancement is less than complete.

But in their attempts to outdo each other in their support for the growth of R&D activity, there is a risk that an uncomfortable truth is going unrecognised by policy makers – R&D activity is becoming globally footloose in the same way as manufacturing did in the 1980s and 1990s.

It could be that the rush to build policies around attracting and retaining R&D is both ill conceived and ill fated. It could also distract attention from areas where policy action is most needed and could have the greatest impact in creating a sustainable global competitive advantage in high tech industries, notably developing an entrepreneurial culture. Such a policy emphasis might emerge as the most effective means of cementing and growing the national R&D effort.

This is not an academic issue. It has immediate implications for where scarce industry policy dollars are invested. Seen through this lens, initiatives such as the recent deal with Mitsubishi to have it invest in a new R&D operation in Australia look much riskier.

It is worth considering the recent history of industry policy to understand how R&D came to assume the policy status it has. No one today would seriously suggest that governments should be signing blank cheques to attract or keep generic manufacturing activities. Having experienced the pain of the lose of large parts of the textile, clothing and footwear industries, automotive manufacturers from Leyland to International Harvester to Nissan, and consumer electronics manufacturers, Australians have come to accept that it is pointless thinking we can or should be competing with countries such as China in areas of manufacturing that are particularly cost sensitive.

On the other hand, we continue to believe that there are areas of niche manufacturing where Australia can and should continue to participate. Activities, for example, where the cost of labour is low relative to the cost of intellectual property, where there are performance requirements peculiar to Australia, or where the timeframe to respond to market signals is too short to allow overseas sourcing.

The rise in policy emphasis on R&D ran parallel to the emergence of this more nuanced thinking about manufacturing.

Overseas-headquartered companies emphasised that manufacturing was a costly exercise. New business models saw a rise in the use of contract manufacturing as original equipment manufacturers began to understand that their cost asset was their brand and the technological edge it represented. Nike became the corporate poster child for this thinking – it manufactured nothing, but focused its in house efforts on designing products and finessing its brand image.

As companies moved their manufacturing from Australia, they often offered governments comfort by pointing out that R&D was being retained in Australia. This, we were told, demonstrated that Australia was not being abandoned.

In effect, attracting and retaining R&D has become the centrepiece in industry policy today, replacing attracting and retaining manufacturing.

But the experience of the past two years suggests the R&D story, too, is more complicated than is at first apparent.

The highest relative investors in R&D have been in the ICT industries, where the speed with which innovation moved from labs to market seemed to be accelerating unrelentingly.

Since the "tech sector meltdown" began in April 2000, when the cash spigot was turned off for companies that had been rolling in money, a new imperative has emerged to drive companies and investors alike – cost cutting.

In this environment, a new reality has emerged that policymakers must understand sooner rather than later –R&D in high tech industries is coming to be perceived as a cost centre by the managements of multi-national corporations.

Put aside protestations about R&D being the very heart of their businesses and too important to risk. The same arguments were put in relation to manufacturing a decade ago when manufacturing industries similarly focused on radical cost reduction in the face of global restructuring. Corporations in the ICT industries are in the worst recession for at least 15 years and are completely fixated on quarterly results, as share markets batter them mercilessly and sales stall. They are moving most quickly because they are feeling the most heat, but other industries will trend the same way.

Around the world, R&D labs are being downsized or closed altogether. At the same time, governments are offering incentives for companies to relocate R&D to their borders just as they competed with manufacturing incentives to the same companies a decade ago. R&D activity is moving to places such as Brazil and Singapore who have spotted this trend.

Combined, these phenomena suggest an inescapable conclusion: R&D is becoming globally footloose in precisely the same way as we now understand manufacturing to be.

Some R&D will adhere to specific market where special conditions require particular product characteristics – often, probably, alongside niche manufacturing. But generic R&D for a global market? Anywhere will fit the bill, provided that the necessary engineering talent can be accessed, which is an ever-lengthening list of countries, and where any specific infrastructure requirements can be met, again, an increasing list thanks to the communications revolution. Add to that the prospect of extracting financial incentives from governments and the incentive to move R&D activities to low cost centres becomes irresistible. And, once one companies begins to behave in that way, all its competitors must follow.

So, what industry policy options are there in this environment?

Well, it could be concluded that Australia as a market or source of intellectual property does not have the gravitational mass to hold onto global high tech corporations and has no alternative but to play the investment incentives game. If this is the conclusion, it has to be recognised that there will be an ever-growing number of players to compete with and Australia needs to get into the game as quickly as possible to take advantage of whatever early-ish mover advantage it can.

But high tech industries do not live by R&D alone. Far from it.

It is widely accepted that successful high tech clusters have three foundations: a source of intellectual property; a source of risk finance, and; a source of entrepreneurial management skill.

Australian policy makers have gone to considerable effort to encourage the growth of the venture capital community in the past decade, and the industry has matured quite quickly. Intellectual property is developed not only in big corporate labs, but, historically, has predominantly been created in publicly funded labs in this country.

The element that has been lagging in its rate of development is the creation of a population of entrepreneurs.

In established high tech cluster internationally, the importance of these individuals is well appreciated. Venture capitalists in the US, for example, commonly retain a so-called entrepreneur-in-residence who has a roving brief to oversee the organisation's deal flow, and a licence to nominate a venture they would like to take on themselves.

Practices such as this demonstrate an understanding born of long experience in these economies: that people with the rare talent to be able to identify technologies with corresponding market opportunities, and create businesses out of welding technology to opportunity, are money in the bank.

The Grand Daddy of high tech cluster, Silicon Valley, suggests the presence of entrepreneurs is the cornerstone of enduring regional competitive advantage that goes beyond a single technology. Silicon Valley's first high tech industry was built on electronics components and products. Only with the founding of Shockley Labs to exploit new fangled transistor technology did silicon truly come into play.

But then came personal computer hardware, PC software, multi media software and the Internet revolution. In each case, Silicon Valley was squarely at the heart of the innovation explosion and subsequent wealth creation, even though the core technologies often were developed elsewhere.

Why?

Because in Silicon Valley, generations of entrepreneurs had grown used to constantly seeking the horizon for the Next Big Thing, and generations of venture capitalists had grown rich backing their judgment. Far from seeing companies move away when they reached a certain size, MNCs from around the world flock to buy land at inflated prices to be able to hang up their own Silicon Valley shingle.

Thus, a virtuous cycle has been born.

Building an entrepreneurial culture is not as easy as attracting a large corporate R&D lab, of course. There is no single board of directors to whom a cheque can be offered for a start.

But an entrepreneurial culture has one huge advantage for policy makers over attracting established R&D spenders. By definition, once a culture in established, no one can pick it up and take it away!

Initiatives to encourage such a culture -such as encouraging clustering of activities in technology precincts where entrepreneurs can learn from each other -is not a silver bullet. Imagining that we are just one policy initiative from Nirvana is a mistake that has been made too often before.

Balanced policy is called for. MNCs provide customers, a trained workforce, access to international market and management intelligence and a source of strategic investment, and should be encouraged to be actively engaged in our market. Venture capitalists provide money, but also experience, contacts and an understanding of the pitfalls ahead of a growing company and are necessary partners.

But a balanced stool needs at least three legs. And entrepreneurs are emerging as the most stable support when it matters.