April 12, 2002

Productivity Forecasts

The broad universe of commentators and economists is still somewhat divided on what the medium-term future holds. Writers like Julie Kosterlitz of the _National Journal_ talk of how growth in the late 1990s was unsustainably fast because of the NASDAQ bubble. Nobel Prize-winning economist Joe Stiglitz talks of how there must be "real restructuring" because of the "capital overhang" produced by excessive investment that has created "overcapacity," and how the "real restructuring takes time."

By contrast, the smaller slice of economic forecasters have reached near consensus that the next decade of productivity growth is much more likely to see a continuation of recent trends than a renewed productivity slowdown. In their near-consensus analysis, the information technology revolution has driven rapid productivity growth because of the interaction of three factors. First, technological progress in and price declines of information technology capital goods have been extremely rapid. Second, the share of spending on adding to the information technology capital stock took an upward leap in the 1990s, and a higher spending share coupled with lower prices accelerated the pace of growth of the real information technology capital stock. Third, the share of total income attributable to the returns on the existing information technology capital stock jumped as well, which meant that the faster pace of growth of the information technology capital stock produced even faster growth in labor productivity. The effect of the ongoing technological revolution is proportional to the strength of these three factors multiplied together. And growth accelerated because all three of these factors took an upward leap in the 1990s.

The belief that the future will see rapid productivity growth is built on this quantitative foundation. The speed of growth can decline, in this near-consensus analysis, only if there is (a) a medium-term decline in the pace of the information technology revolution, (b) a decline in the share of total expenditure devoted to information technology goods, or (c) a decline in the share of total income attributable to the higher productivity made possible by information technology. The near-consensus analysis sees no sign of a decline in any of these three facts. And this implies, in standard economic growth models, that the future will be like the recent fast-growth past.

Moreover, there are four jokers in the deck: four things that could well make the future even brighter than the past since 1995...

The course of U.S. productivity growth during the recent recession is yet another piece of evidence that the productivity growth future looks bright. Usually recessions see falling productivity: firms hoard labor they cannot use productively now so as to have a pool of trained workers when demand picks up, and the pressure to find ways to economize on resources to expand capacity is absent in times of slack demand. But the 2001 recession saw productivity in the American economy rise. Consider the fourth quarter of 2001, in which hours worked in the U.S. economy fell at an annual rate of 3.3 percent. But in that quarter production rose at an annual rate of 1.7 percent, and productivity rose at an annual rate of 5.0 percent.

If you are a politician in office, such growth lets you score political points by saying that there was never a recession at all (never mind that the U.S. economy has shed so many jobs over the past year). But to an economist interested in forecasting more than the next three month's statistics, it suggests something much more interesting--that the process of automation and investment in information technology that fueled such rapid real wage and production growth in the late nineties is still going on under the surface even in the depth of the recession, even when businesses are under no immediate pressure to boost their capacity. This suggests that the underlying computer-driven productivity growth trend is rapid indeed.

There is, however, one significant puzzle confronting those who (like me) believe in optimistic scenarios about future productivity growth. If the information technology revolution is such a transformative technology, where is the accleration of productivity growth in western Europe? To this question, there are four answers: The first is that the premise is wrong, and that the U.S. is experiencing a short cyclically-related boost to output rather than a long-run upward shift in the rate of productivity growth. The second is that western Europe's productivity miracle is coming, but that processes of economic diffusion take time, and that one should not expect diffusion immediately to regions more remote in space and social distance from Silicon Valley. The third--Alan Greenspan's favorite--is that businesses will be motivate to move to more efficient computer-intensive modes of organization only when they see cost savings from doing so, that in Europe businesses by and large see neither the possibility increasing profits through cost saving nor the possibility of increasing profits through expanding production, and hence thorough-going neoliberal reform is necessary before Europe's economies can reap large aggregate benefits. The fourth is that it is not rigid labor markets and slack demand per se that account for the slow spread of the information technology revolution throughout Europe, but instead a failure of governments to cut their red tape. On the periphery of western Europe from Ireland to Finland, after all, the information technology revolution is doing just fine.

I do not have the answer to this puzzle. But it must be answered and answered correctly, if western Europe is to have as bright a productivity growth future as America seems to have right now.

Posted by DeLong at April 12, 2002 03:52 PM | TrackBack

Post a comment