IT has had profound impact on productivity

Information technology (IT) was a key driver behind the productivity gains of U.S. workers in the second half of the 1990s, and it would take a drastic reduction in U.S. business investment in IT to pull down the current strong rate of productivity growth, a U.S. Federal Reserve economist said Tuesday.

The link between a dramatic increase in the productivity of U.S. workers from 1995 through 2000 and the strength of the U.S. economy during that period has been well defined, but economists wanted to know how much of a role IT played in achieving the productivity gain. It turns out it was a lot, according to research conducted by Daniel Sichel and Stephen Oliner, two Federal Reserve economists who presented their findings in a paper titled “The Resurgence of Growth in the Late 1990s: Is Information Technology the Story?”

Sichel, speaking at the Center for Strategic & International Studies, gave a synopsis of the research, saying the goal was to discover the link between developments in computer hardware, software and telecommunications equipment and the extraordinary performance of the U.S. economy over the last five years.

The growth rate of non-farm business labor productivity from 1991 to 1995 amounted to 1.5 per cent per year. That’s respectable for the relatively short period, but it was far outpaced in the second half of the decade when productivity rose to 2.9 per cent per year, an increase of 1.4 percentage points, or nearly double.

“In terms of the long run growth potential of the economy this is a very big deal,” Sichel said.

Sichel and Oliner looked at how the use of IT by businesses, and efficiency gains in the production of computers and semiconductors, combined to achieve this productivity surge. They found that the total contribution of these two “channels” to the improvement in labor productivity was about three quarters of a percentage point, meaning the two factors accounted for roughly 60 per cent of the improvement in labor productivity growth.

The production channel was especially interesting, Sichel said, because while it would seem that computer hardware manufacturers make up such a small part of the U.S.’s multitrillion dollar economy, it turns out that the pace of innovation in the second half of the 1990s was so rapid that it made an important contribution to aggregate labor productivity growth. This was particularly true in the semiconductor sector, the economist said.

“This is the idea of Intel coming out with a new chip every 18 months that sells for the same price as the prior generation, but is much, much more powerful than that prior generation, and the pace at which they were able to do that really appears to quicken in the latter half of the 90s,” he said.

When Sichel and Oliner dug deeper, they found that efficiency in semiconductor production and price declines in the second half of the ’90s were behind the increased use of IT. If the price of semiconductors continues to decline, there are reasons to be quite optimistic about the future path of labor productivity growth, they concluded.

“It may seem funny that this small sector in terms of its share of output could be so important, but if you work the numbers, it really turns out that it’s a very important driver,” Sichel said.

Measuring the contributions made by the use and production of IT to growth in productivity produces a “very upbeat story,” Sichel said. Together they resulted in an increase of 1.8 per cent growth in productivity in the first three quarters of 2000, and that is unlikely to slow given that the amount invested in IT by U.S. business increased by 28 per cent in the first three quarters of 2000. Growth in IT investment would have to virtually cease to pull down productivity, he said.

“Even if there’s quite a substantial slowing in the technology sector … it’s probably no where near as far down as where we were prior to the tech boom in 1995,” he said. “That gives me some reason for being cautiously optimistic that a considerable portion of the gains that we saw in the second half of the 1990s actually are permanent, rather than a cyclic phenomenon that would go away as the economy slows.”