San Francisco

It could well be the greatest paradox of the post-industrial age.

The creation and application of knowledge is increasingly the battleground on which commercial competition is fought.

Yet we still do not know exactly how lightbulbs turning on in people's heads are systematically turned into commercial applications.

We know that humanity has demonstrated an innate ability to do this, and to do it repeatedly, but history is replete with examples of great breakthroughs that have occurred more by fluke than design.

But the world is entering an age where luck, more than ever, is not enough. If staying in business means continuing to apply knowledge more effectively, companies need to be able to rely on an institutional framework that allows them to be confident that the process that takes an idea to a product can be repeated.

Yet this is also an era of rapid structural economic change, which some students of innovation fear might break up the very economic entities which enabled the breakthrough flukes to occur. see articles: The frictionless economy and The view from the R&D coalface

Economists, not surprisingly, have different opinions about the type of industrial organisation that creates the best environment for these breakthroughs.

Classical economic theory holds that monopolies suppress innovation. Indeed, according to Nobel Laureate, Stanford University Economics Professor, Kenneth Arrow, the effects of monopolies on national innovation are a more compelling reason for competition policy than their effects on prices.

Arrow says the loss caused by inefficiency in monopolistic markets has been calculated to be only about one percent.

"If you take the classical apparatus which tells you that monopolies are inefficient, then say, okay, what is the loss to the monopoly, that is, using the same theoretical base, these losses turn out to be not very big," Arrow says.

"The effect on things like innovation are far more serious," he says.

"The trouble is, if you're already a monopolist, say due to a previous innovation, you've got two conflicting tendencies.

"On one hand, you want to protect your monopoly power through a new innovation. On the other hand, you are competing with yourself, so to speak, if the new innovation replaces the old one.

"If it's a small enough innovation, you may prefer not to introduce it, because you'll be undermining your previous monopoly."

But the problem also extends to the structure of industries overall. Arrow says history suggests that more large, transformational innovations come from small businesses.

He argues it is difficult for large businesses to break free in their thinking from what they have done in the past to take genuinely new approaches. Also, it is too tempting for them to concentrate their resources in a "big push" for what appears to be the most promising lines of research, based on their past experience. "This leaves room for others to follow up the lines that the big firms don't take," he says.

Arrow concedes that big business has done "some pretty innovative things", but cites IBM as an example of both the strengths and limitations of big corporate innovation.

"The development of electronic computers by IBM was pretty impressive," he says.

"They had a whole way of life built on punch card machines and they broke away from it. Admittedly, they didn't start the idea, but they picked it up and made it practical, with the development of the big mainframes.

"They were a little slow on the PC. They saw mainframes as everything, didn't have the imagination to develop personal computers.

"I won't say big firms haven't introduced some new ideas alright, but the disproportionate amount comes from small firms, precisely because of this," Arrow says.

If one large company dominates an industry, a small firm attempting to enter the market on the basis of an innovation is vulnerable to being driven out of the market.

US courts are being asked if just such a situation has arisen in the Microsoft case. The US Department of Justice case is against the company that most benefited from IBM's tardiness in getting into the PC market. Microsoft is accused of leveraging its domination of the desktop software market into domination of the new, internet browser market. In the process, Microsoft is accused of trying to unfairly drive out Netscape, which created the market with its Navigator browser. Microsoft allegedly used a series of tactics, including predatory pricing, forcing business partners to favour its browser, and bundling its own browser product into its operating system.

Microsoft founder and chairman Bill Gates, a man who must sometimes wonder how he has attracted so much opprobrium, regularly points to the $US3 billion a year that his company spends on R&D, to defend Microsoft from charges that it is not innovative. But Arrow dismisses the bulk of this as simply the price of incremental product development needed for a company such as Microsoft to stay in business.

Other leading economists, however, take a different view. They are not convinced that large businesses, even monopolists, have a stifling effect on national innovation. In fact, they argue that the presence of monopolists has been an essential element in the innovation process of at least some industries.

In his recent book, The Productive Edge, Professor Richard Lester, director of the Massachusetts Institute of Technology Industrial Performance Center, argues that the existence of monopolies has allowed countries such as the US to fudge a lack of understanding about how innovation actually happens.

Lester quotes work by the late Princeton University Professor Donald Stokes, who produced a matrix of R&D. At one end of the R&D spectrum is work such as that by Neils Bohr into atomic structure. This work was designed to advance basic knowledge but had no applied use in mind.

At the other end of the spectrum is the work of Thomas Edison, who Lester says went as far as to ban his researchers from doing any work that sought to advance theoretical understanding, so determined was he to focus only on the application of existing knowledge.

But in the middle, was the work of Louis Pasteur, who wanted to address real world problems, such as treating disease and preventing milk spoiling, but had to make advances in basic knowledge to do so.

Lester says this middle category of research has been little understood, but is absolutely crucial in an effective national innovation system. The difficulty with it is that it is difficult to see how the research should be paid for and what incentives should be created to encourage it.

Pure research aimed at advancing knowledge, with no specific application in mind, is properly the responsibility of the nation as a whole, and should be publicly funded through institutions such as universities or research centres. Work to commercialise knowledge should be paid for by the businesses or individuals that expect to profit from their work. But what of work that could take many years and might never result in a commercial application, or, if it does, takes so long to result in a product that no private financier would be able to recoup the investment before others are able to imitate or recreate their work?

"Historically, this sort of work has been funded partly by industry and partly by government with a variety of ad hoc approaches of varying effectiveness," Lester says. Most of these relied on the sheer size of the research budgets available to public or private enterprises that enjoyed monopoly or near monopoly power. He cites AT&T's famous Bell Labs, IBM's Yorktown laboratories and Du Pont's central research labs, the place where nylon was invented. "But many of these great labs have been dismantled in recent years, and those that have escaped this fate have mostly been repositioned to do shorter term, product-specific research."

Lester says Microsoft is an exception that proves the rule. He says the software giant announced plans to create a long-term research centre in 1995, swimming against the corporate tide.

Lester concedes that there are strong arguments that large corporations lack the flexibility to recognise transformational technological breakthroughs at times, sometimes even when they are occurring within their own laboratories, and gives IBM's failure to move into personal computers as an example. Nonetheless, he argues that the trend for large companies, such as Intel, to have their own venture capital funds to encourage new applications for their own products does not represent an alternative to the giant corporate labs as sources of Pasteur-type innovation.

Lester says start-up companies based on innovative technologies are very heavily concentrated at the Edison end of the R&D cycle, and doubts they can adequately substitute for the giant labs of the past. They are too dependent on their ability to add to shareholder value in the short-term, so that the funds can realise capital gains by selling their shares. See articles: An Audience with Jeff Sohl and How Angels are Giving Flight to the High Tech Boom.

Lester proposes a scheme devised by Stanford University economist Paul Romer as a potential funding solution. Put simply, Romer suggests industries which see an opportunity for technological or scientific advances that individual companies could not or would not fund should be able to approach the government with a proposal to establish an industry research board. The government, after accepting such a proposal, would levy a sales tax on the industry, the proceeds being allocated by companies to research projects they assess as most valuable. All businesses in the industry would have equal access to the results of the research.

Romer's solution seems elegant. But it would take a courageous politician to champion it in an age where any suggestion of government involvement in business is politically anathema, and where economic specialisation and the release of individual entrepreneurial zeal are lauded as the drivers of the latest regional economic miracles in places such as Silicon Valley. More likely, perhaps, the solution to developing new innovative institutions will be in the hands of business people themselves. See accompanying article: The view from the R&D Coalface.

Further Reading

The Productive Edge: How U.S Industries Are Pointing the Way to a New Era of Economic Growth. Richard K Lester.

Bell Labs is Dead, Long Live Bell Labs. Robert Buderi. Technology Review, September-October 1998.