The actual calculation of the CPI is more complicated than our example suggests. The Bureau of Labor Statistics (BLS; http://www.bls.gov/cpi) is the US government agency that is responsible for this calculation, while other countries have similar agencies. The BLS procedure is, in essence, the one we have described: it compares the cost of the same bundle of goods in different years. However, the BLS confronts several difficulties that we have ignored so far.
- Quality changes. Imagine that you now work for the BLS (you took this job after you left the International Monetary Fund [IMF]) and are asked to look at changes in the price of laptop computers. 1 You decide to use the IBM ThinkPad computer. You discover that in 1992 a ThinkPad cost $4,300 on average. Then you find that it is possible to purchase a ThinkPad in 2011 for $899. You calculate the percentage change in the price as ($899 − $4,300)/$4,300 = −$3,401/$4,300 = −0.79 and conclude that the ThinkPad is 79 percent cheaper than two decades previously. You report this to your boss and then go home. But then you start to worry. The 2011 ThinkPad is nothing like the 1992 version. The 1992 computer had 120 MB of memory and weighed over 5.5 pounds. The 2011 ThinkPad has 4 GB of memory and weighs 2 pounds less. It has a vastly bigger hard drive, wireless Internet connection, and a superior display. In short, there were huge quality improvements over this period. A computer with the specifications of the 1992 ThinkPad would be worth much less than $899. By ignoring the improvements in quality, you have understated how much the price of computers has fallen. This problem is particularly acute for computers, but it applies to all sorts of different goods. The new car that you purchase today is very different from a car that your mother or your grandfather might have bought. Cars today come equipped with computerized braking systems, global positioning system (GPS) navigational tools, and numerous other sophisticated engineering features. They are also much more reliable; your grandparents will tell you that cars used to break down all the time, whereas now that is a relatively rare event. It would be a big mistake to say that a 2012 automobile is the same as a 1961 automobile.
- New goods and old goods.The typical basket of goods bought by consumers is changing. In 1970, no one had a mobile phone, an MP3 player, or a plasma television. Similarly, people today are not buying vinyl records, videocassette recorders, or Polaroid cameras. The BLS needs to keep up with every change. As the economy evolves and new goods replace old ones, they must change the basket of goods.
- Changes in purchasing patterns.The bundle purchased by the typical household also changes over time because of changes in the prices of goods and services. The typical household will substitute away from expensive goods to relatively cheaper ones. If the basket of goods is held fixed, the calculation of the CPI will overstate the increase in the cost of living. This effect is most severe if there are two goods that are very close substitutes and the price of one increases significantly relative to another.
Perhaps these seem like minor details in the calculation of the CPI. They are not. A government commission chaired by the economist Michael Boskin provided an extensive report on biases in computing the CPI in 1996. The Boskin Commission concluded the following: “The Commission’s best estimate of the size of the upward bias looking forward is 1.1 percentage points per year. The range of plausible values is 0.8 to 1.6 percentage points per year.” That is, the Boskin Commission concluded that if inflation as measured by the CPI was, say, 3.1 percent, the true inflation rate was only 2 percent. In response to these concerns with measurement, the BLS responded by taking actions to reduce the biases in the measurement of the CPI and deal more effectively with the introduction of new goods. 2
- 851 reads
