Do computers really boost productivity?

Paul Mason offers a 10-point briefing on the issues confronting strategists and policy makers as they get to grips with the new...

Paul Mason offers a 10-point briefing on the issues confronting strategists and policy makers as they get to grips with the new economy's biggest unanswered question...

Behind all the hype about the new economy there is only one real question that matters: can the Internet bring a rise in productivity and, if so, will it last? Here are ten questions to help you follow the debate.

  • What is happening? For decades economists accepted the "Solow Paradox" - that the impact of computers is seen everywhere except in the productivity figures. But a series of studies in the USA show that since 1995 there has been a discernible IT effect on productivity. Productivity surged after 1995 - that was noteworthy because it usually tails off at the end of a period of growth. (See top graph, right)

  • What is the impact of IT on US productivity? Estimates vary but the most conservative puts the IT contribution to productivity growth at 50%. The boldest, from economist Karl Whelan, says 73%.

  • Why does it matter? The Bank of England has to decide whether economic growth is sustainable without provoking inflation. If it assumes no IT effect on productivity - and none has been proved to exist here - then it will put the brakes on, through interest rate hikes, sooner rather than later.

    If, however, there is a real Internet effect on productivity, policy makers can hit the accelerator, not the brakes. The International Monetary Fund recently said that getting the policy wrong could provoke a slump. "In a situation where future productivity growth is highly uncertain, the costs of maintaining an inappropriate policy rise rapidly with time."

  • How does IT show up in the statistics? Economists meas-ure two kinds of productivity: that which results from "capital deepening" and what they call "total factor productivity". Put simply, the first is the result of deploying more capital per worker; the second includes all the organisational changes that accompany new technology.

  • What is the main IT factor in productivity growth? Falling hardware prices. Since 1995 there have been "sharp declines in the prices of computer processing, data storage and retrieval, and communications" says the US Department of Commerce. In turn this allowed a rapid upturn in IT investment.

    According to the US government, the relationship between capital stock of IT hardware and hours worked increased 16% a year in the first half of the 1990s but shot forward to 34% a year between 1996 and 1999. Software grew by around 11% in both periods. The rate of increase for other kinds of capital was 0.5% throughout the whole of the 1990s.

  • Can the trend continue? Experts are divided. A recent paper by Charles Fine of MIT and Daniel Raff of the Wharton School looked at the IT effect in the US auto industry. They concluded that the biggest savings were to be made in the "order to delivery cycle" - for example by copying the Dell Direct model of sales - rather than in the production process itself. "The predominant effect thus seems likely to wring inefficiency out of the value chain through what is basically a one-shot improvement in forecasting, communication and co-ordination.

    "The automobile sector is large, and this could well have a discernible effect on productivity growth while it is happening: but it does not seem likely to be ongoing."

    Another Department of Commerce-commissioned report, however, predicted that productivity growth would continue as long as real hardware price declines of 20% a year continued.

  • Will the productivity effect come to the UK? One factor inhibiting this is the smaller size of the UK's information and communications industry. Falling prices in the ICT sector had a direct impact on US inflation: without productivity gains in these industries, inflation from 1995-8 would have been 2.3%, not 1.8% as it turned out. Likewise, the IT industry alone accounted for one-third of real GDP growth in the USA (1995-9).

  • What do UK experts think? DeAnne Julius, a member of the Bank of England Monetary Policy Committee, told a conference in August: "A convincing case can be made that developments in the US are a leading indicator of events here.

    "The UK, like the US, has seen a sharp increase in investment spending on ICT in the latter part of the 1990s, and productivity growth has recently begun to rise."

  • How will we know it's happening? We won't, until 2002. The UK government is only now putting in place methods to measure the economic impact of IT investment, and initial figures will not be out until late 2001.

  • What does it mean for my business? If you are investing heavily in IT don't think that will automatically translate into measurable productivity growth. The US statistics found that IT investment without accompanying "re-engineering" of the business process actually depressed productivity.

    And the IT effect is still only measurable in the goods-producing industries. According to the DoC: "Between 1990 and 1997, despite heavy investments in IT and a three-decade build up of the real net IT capital stock, IT using service industriesÉrecorded declining productivity."

    Basically, all the evidence suggests that only when IT is combined with decentralised organisation does it show any appreciable effect on productivity - but watch this space.

  • Read more on Data centre hardware