Measuring the Growth from Better and Better Goods
"The quality growth for durables has been understated by 3 percent per year for the past 15 years."
Much economic growth occurs through growth in quality, as new models of consumer goods replace older, sometimes inferior, models. For 1995, as an example, researchers have estimated that the Bureau of Labor Statistics (BLS) methods of determining economic growth allowed for as much as one percent average quality growth in goods. However, it is often argued that the BLS methods miss much of the growth in goods' quality. The Boskin Commission Report (1996) suggests that the BLS overstates inflation by perhaps one percent a year, with unmeasured growth in the quality of goods the most important component of that overstatement, contributing 0.6 percent per year.
In Measuring the Growth from Better and Better Goods(NBER Working Paper No. 10606), NBER Research Associate Mark Bils estimates that quality growth for durables has been understated by 3 percent per year for the past 15 years. This suggests an actual growth rate of at least 5.8 percent per year, even with computers excluded. These results indicate much faster quality growth for durables, particularly vehicles, than reported by the Boskin Commission.
Bils points to the difficulty of distinguishing quality growth from true price increases for goods such as durables that display frequent model changes. To calculate the consumer price index (CPI), the BLS tracks a large set of prices, with each price specific to a particular product at a particular outlet. At regular sample rotations, the BLS draws a new sample of stores and products within a geographic area to better reflect current consumer spending. But, in addition, a store may stop selling a particular product. The BLS agent then must substitute another model of that brand or of a similar product. These (forced) substitutions occur about once every three years for all non-housing CPI items. They occur much more frequently, nearly once per year, for consumer durables.
Using microdata from the CPI, Bils shows that much of price increases captured in the CPI for durable goods since 1988 reflects not increases in price for a given set of products, but rather shifts to a newer set of product models that have higher prices. He shows that one can arrive at vastly different measures of price inflation and real growth under plausible competing assumptions on how much quality change accompanies these product turnovers.
To judge quality growth, Bils examines how consumer expenditures respond to product substitutions. For automobiles and consumer electronics, he finds that consumer spending clearly moves away from static goods, that is, those with no model changes. This suggests a true rate of inflation that is even lower than that exhibited by these static goods.
For vehicles, the results of this study suggest that quality growth has been understated by as much as 4.4 percent per year. Bils suggests that growth for consumer electronics also has been substantially faster than historically measured, by 2.9 percent or more per year. For the balance of durables, Bils observes that prices for static models respond very little to competing product substitutions, suggesting that price increases accompanying these product substitutions may reflect higher perceived quality as well. This would imply that quality growth is as much as 1.6 percent faster than suggested by BLS measurement methods for these goods.
Product substitutions are more important for consumer durables than for most other consumer goods, so Bils points out that it would not be appropriate to project his findings to non-durables. But the approach could be extended beyond durables by obtaining information for additional goods on how market share responds to product substitutions. Bils suggests that greater availability of scanner data should gradually provide researchers with market information for a broader set of goods and for longer sample periods.
-- Les Picker
The Digest is not copyrighted and may be reproduced freely with appropriate attribution of source.