Martin Feldstein writes The U.S. Underestimates Growth [1] in The Wall Street Journal:
Government statisticians are supposed to measure price inflation and real growth. Which means that, with millions of new and rapidly changing products and services, they are supposed to assess whether $1,000 spent on the goods and services available today provides more “value” or “satisfaction” to American consumers than $1,000 spent a year ago. Even more difficult, they are tasked with estimating exactly how much it costs now to buy the same quantity of “value” or “satisfaction” that $1,000 could buy a year ago.
These tasks are virtually impossible, and the problem begins at the beginning—when an army of shoppers go around the country at the government’s behest to sample the prices of different goods and services. Does a restaurant meal with a higher price tag than a year ago reflect a higher cost for buying the same food and service, or does the higher price reflect better food and better service? Or what combination of the two? Or consider the higher price of a day of hospital care. How much of that higher price reflects improved diagnosis and more effective treatment? And what about valuing all the improved electronic forms of communication and entertainment that fill the daily lives of most people?
In short, there is no way to know how much of each measured price increase reflects quality improvements and how much is a pure price increase. Yet the answers that come out of this process are reflected in the consumer-price index and in the government’s measures of real growth.