Improved management at Wal-Mart probably played a bigger role in the US's productivity "miracle" of the late 1990s than all the expensive investment in high-speed computers and fibre-optic cable by businesses, according to a challenging new analysis of US economic performance published this week.

McKinsey Global Institute, the research arm of the management consulting group, says the US did indeed experience a sharp improvement in its underlying economic performance between 1995 and 2000, but says the change can be accounted for by growth in just a handful of business sectors and was not principally the result of the surge in investment in information technology experienced over the same period.

Belief in an IT-driven transformation of US economic performance has come close to being accepted as economic orthodoxy, although the scale of the improvement has been revised down in recent government figures.

Many leading tech firms had claimed the sharp improvement in US productivity - from 1.4 per cent a year between 1972 and 1995 to 2.5 per cent between 1995 and 2000 - was the result of a surge in IT investment. Between 1995 and 2000, IT investment increased 20 per cent a year, nearly double the rate from 1987 to 1995. In the four years from 1995, US businesses doubled their IT capabilities.

But the McKinsey report says in most sectors the large increases in IT investment did not produce any improvement in productivity.

"There was a big jump in capital spending on IT and a big jump in productivity in the economy as a whole at the end of the '90s," says Mr Bill Lewis, director of the McKinsey Global Institute. "But the actual correlation between the two is very weak."

The report is the latest in a number of analyses of the productivity changes in recent years. Some suggest the change was widespread, based on the application of new technology, others have been more sceptical, saying the improvement was confined to a small number of IT-producing sectors and was, in any case, largely cyclical.

"There's no doubt technology investment played a role, especially in the IT sectors themselves, but the gains for much of the rest of the economy were much less clear," says Mr Barry Bosworth, an economist with the Brookings Institution, which acted as an independent adviser to McKinsey.

McKinsey's researchers used official government data to determine which sectors were the principal drivers of productivity growth. They then did a detailed analyses of what was behind the productivity growth by looking at companies in each sector.

Nearly all the post-1995 jump in productivity was in just six sectors - retail, wholesale, securities, telecoms, semiconductors and industrial machinery and equipment (mainly computers) - representing about 31 per cent of the non-farm private sector economy.

The other 53 sectors, representing 69 per cent of the economy, were mixed, but overall contributed just 0.3 per cent productivity growth. Yet these 53 sectors accounted for 62 per cent of the acceleration in IT spending. Many of them actually experienced productivity deceleration.

In the six sectors that produced almost all the net productivity improvement, a number of factors contributed to the improvement - of which IT was just one.

But the report's contention that IT played a limited role in the overall economy was brought into question by its admission that, in all the sectors where productivity gains were recorded, it was a contributing factor.

Thus, in retailing, it was the introduction of electronic data interchange, bar codes and RF gun screening; in wholesale trade, better inventory management as a result of faster data processing; computer manufacture benefited from the increase in chips; securities trading was transformed by the advent of the internet; and introduction of digital cellular equipment led to a leap in telecoms productivity.