Mr Paul Strassmann is playing his favourite role once more: the computer expert who claims companies are wasting money on computers. The IT industry's leading iconoclast has been rehearsing the part during more than 40 years of rigorous research into companies and their use of IT. However, as the latest of his more than 200 reports, books and articles on the subject nears publication, his message has been refined into its ultimate, devastating form.

Mr Strassmann's central thesis is that no relationship can be demonstrated between the amount a company spends on computer systems and its profitability. And, he believes, none ever will be.

He quotes for emphasis Mr Robert Solow, the Nobel Prize-winning economist who taught him statistics at Massachusetts Institute of Technology: "You can see computers everywhere - except in economic statistics."

A second argument is no less contentious: that the IT industry - customers and vendors alike - is on the cusp of an important and final disjunction.

"Every seven years, we have torn up what has gone before and started again," he says. "There have been eight cycles of `build and scrap' since 1946. The first cost $100 million (€118 million), equal to 7 per cent of business investment at the time. The last cost $2 trillion or 47 per cent. The next would have cost $5 trillion but we have run out of money: we have come to the end of history as we know it."

These conclusions have a significance beyond the computer industry. The US, which spends more on IT than any other nation, is on the brink of its first recession in a decade, dragging with it much of the developed world.

It has sustained growth during the past 10 years, many think, by investing in information systems. The pundits argue that this has delivered competitive gains, accelerated transactions and increased customer satisfaction. Many other nations, envious of the performance of the US economy in the past decade, have set out to achieve the same miracle. But is this belief in IT as the powerhouse of improved productivity misplaced? Has the US share of that $2 trillion been squandered on a false promise? Few are better placed to answer these questions than Mr Strassmann, now 72 and as active in retirement as when he was chief information officer for the US Department of Defence.

His career has seen him managing the information interests of large companies, including Kraft and Xerox. It culminated in his appointment to a newly created post of director of defence information at the Department of Defence, where he was responsible for a $35billion cost-reduction programme. He oversaw the department's IT strategy at the time of the Gulf war and was rewarded with the Defence Medal for Distinguished Public Service, its highest civilian award.

It is, however, the quality and quantity of the data underpinning his chief contentions that set Mr Strassmann apart. He has been collecting financial statistics on the world's principal companies with all the enthusiasm of a trainspotter since 1954, soon after leaving Cooper Union college in New YorkCity with an engineering degree.

Today, the numbers - revenue, profits, assets, productivity, head count and so on - fill appendix after appendix in his published works. He is contemptuous of the profusion of studies from consultancies and the computer industry "proving" the value of investment in computer systems.

"I give little credence to the gloss. I say `show me the numbers'," he declares with a grin.

Not for Mr Strassmann the superficial "150 leading companies" that contribute to most surveys. His latest work, on European competitiveness, analyses in depth more than 3,000 companies. The question of how investment in computers may be linked to profitability is as intriguing as it is difficult to answer, in spite of intensive efforts during the past 50 years. Mr Strassmann believes his numbers indicate unequivocally that there is no connection between investment in computers and profitability. He has looked at the data in every possible way.

Plot, for example, investment in IT per employee against return on equity for several hundred companies in the US and Europe. If greater IT investment led to increased profits, you would expect to be able to plot a straight(ish) line climbing from left to right. But Mr Strassmann's chart shows only a random scatter of points, suggesting that IT spending has no bearing on profits.

Mr Strassmann asserts: "Ingesting profuse quantities of the best and latest IT may be largely irrelevant and possibly damaging if the corporation already suffers from excess assets (too much fat) and a huge administrative cost (clogged arteries), and chases every marketing fad (schizophrenia)."

IT spending, Mr Strassmann concludes, has an important but chiefly catalytic role. This is not to argue that IT is worthless, only that it is not an end in itself. Its role, he says, is to promote the factors such as competitive advantage, strategic positioning, and management style and quality that do most to determine profitability.

A company that spends wisely - even if sparsely - on IT will see its performance enhanced. A company that spends indiscriminately on IT will see its performance diminished, because IT will merely amplify its poor business practices.

For evidence, he cites the Profit Impact of Marketing Strategies (PIMS) programme, based in Cambridge, Massachusetts, which concluded, using data from 3,000 client companies, that market environment, competitive situation and position in the market made a 65 per cent contribution to profitability; operating effectiveness contributed only 15 per cent.

The fact remains that over the past 15 years or so, US companies have devoted a dramatically increasing proportion of their investment to IT.

"I find the similarity between IT spending and the arms race the most plausible explanation for what has been corporate practice over the past decade," says Mr Strassmann.

He says IT budget reviews in many companies have followed the military pattern, where requests for added funding are based on little more than the need to match the capabilities of competitors.

"In this respect, the job of the IT proponents is made easier since vendors and an advertising-supported press will make sure the accomplishments of the leading innovators in each industry are well described and promoted."

This IT arms race is no more about to end than the salesforce is about to pack away its Powerpoint presentations. However, Mr Strassmann believes that companies will no longer be able to afford to escalate their IT spending according to the past pattern, where each wave of investment has been bigger than the last.

Because of this constraint, spending will remain stagnant. This will delay the next chapter in the story - the move from IT based on companies' own systems to IT based on central servers and the network.

This is a frightening scenario for vendors that are used to rapid revenue growth. The answer, Mr Strassmann argues, is to escape the inefficiencies inherent in today's customer-based IT - "a do-it-yourself cottage mode of production and medieval guild mentality".

The way to do this, he says, is for hardware and software vendors to fund the next round of IT development by supplying IT services on a fee basis over the Internet. In simple terms, IT will be supplied through a conduit in the wall, like electricity, water or gas.

He believes that the first green shoots of this transformation in the industry can be seen in the explosion of application service providers (ASPs), companies that rent software over the Internet for a monthly fee.

Many experts regard Mr Strassmann's predictions about ASPs with almost as much suspicion as they view his ideas of productivity. Companies, they say, will not trust ASPs with sensitive data, or to run complex systems.

Mr Strassmann, however, rejects these objections. "The records of most company-embedded IT shops are now sufficiently blemished that management will certainly be inclined to listen to the enticing promises of the increasingly revenue-hungry IT vendor."

So, not content with years of rubbishing vendors' claims about the value of their wares, he is now predicting the demise of the corporate IT department.

A Czech by birth, Mr Strassmann was a resistance fighter during the Second World War. It must have been useful preparation for taking on the battalions of the IT industry.

The Economics of Information, to be published by Butler Direct (www.butlergroup.com)