How Fast Is The U.S. Growing?

A lot of economic observers think that U.S. productivity growth is finally on a roll. Not only did it hit a 20-year high in 1992 but it has now risen by at least 1.5% for four straight years--the first such string of gains since the 1960s.

What's puzzling is that this impressive record has yet to raise the living standards of many Americans. The Census Bureau reports that real median family income declined steadily from 1990 through 1993, for example, and real median earnings of full-time year-round workers fell in 1993, as well.

A new study by economist Mark J. Lasky of DRI/McGraw-Hill seeks to resolve this seeming contradiction. It suggests that a big part of the answer lies in the statistical method used to calculate gross domestic product, which tends to exaggerate recent hikes in investment, productivity, and economic growth.

The government uses a fixed-weight method to calculate GDP. That is, it uses the expenditures in a particular base year (currently 1987) to measure the real prices of individual components of GDP and thus their contribution to overall output. This method has many advantages, but one big disadvantage: It gives too large a weight to pieces of output whose prices lag behind overall inflation after the base year.

The problem is especially troublesome when, as in recent years, investment has been heavily concentrated in computers, whose prices have been falling rapidly. Under the fixed-weight method, the purchase of a $1,000 computer today that has the same computing power as a $10,000 workstation in 1987 would be counted as a $10,000 expenditure. Yet its actual impact on overall economic activity would be similar to $1,000 spent on other items, whose prices may have risen. To lessen such distortions, the government issues a comprehensive revision of GDP data every five years, using a more recent base year.

The next such revision, which will shift to 1992 as the base year, will occur in the summer of 1996. But economists have already begun to calculate what it may reveal. Dean Baker of the Economic Policy Institute says it will show that investment in the current expansion has been substantially overstated.

DRI's Lasky estimates that growth in 1992 and 1993 may be reduced by an average 0.5% a year. And annual productivity gains during the expansion will be cut by about 0.45% (chart). These downward revisions help explain why income and employment growth have been so sluggish relative to output.

Although productivity growth so far in the 1990s will be lowered, the good news is that the economy has more breathing room for noninflationary growth than recent measures of GDP suggest. And the long-heralded productivity breakthrough may still lie ahead.

Before it's here, it's on the Bloomberg Terminal.