I just found a 28-year-old invoice for a refrigerator that illustrates a problem for economists: Measuring price changes is complicated.
Some prices have gone up a lot over the past year. Gasoline is one example, although it is about where it was in 2008. But other things haven’t gone up very much, as that refrigerator invoice in my files demonstrates.
However, human nature is such that people pay much more attention to price increases than to price decreases. And they are particularly oblivious when prices of some category of goods go up much less than general prices. The upshot is that we economists frequently are harangued by people who are convinced that “real inflation” is two or three times that of the government-tabulated Consumer Price Index or that some functionary in the White House dictates CPI numbers for the political benefit of whoever is in the Oval Office at the time.
There is no arguing with those who view all government through conspiracy-theory glasses. But examining refrigerator prices 27 years apart gives a valuable lesson in the complexity of measuring changes in the overall or general price level as the CPI and other indexes like the GDP Deflator attempt to do.
The invoice I found is for an 18-cubic-foot Frigidaire purchased in July 1983 for $677.34, including delivery and sales tax. The purchaser was the previous owner of the house. I was given the invoice, along with sundry owners’ manuals, when we bought the house in 1986. The fridge worked fine, but by March 2010, we decided we needed a new one and purchased a Frigidaire 20-cubic-foot model for $996.77, including delivery and sales tax.
In July 1983, the Consumer Price Index stood at 98.1. For March 2010, it was 217.6. If refrigerators had increased in price exactly in step with the index, a new one would have cost $1,502.44. What we actually paid for the new one is two-thirds of that. Adjusted for inflation, prices of refrigerators have fallen.
However, the average consumer, fuming as a gas pump’s dollar meter flicks ever higher while she fills her tank, is not likely to say, “Well, at least the new fridge was really cheap.”
Yet, appliances are one category that has become markedly cheaper in relative terms for a long time. My mother bought an extremely bare-bones electric clothes dryer for $270 in 1961. That is equivalent to $2,030 in 2011. Now far better models can be purchased for a half or even a third of that amount. Our neighbor bought the first color TV set in the area in 1962, paying $650 for a unit that required at least one service call a year and had terrible picture quality. That would equal $4,900 today.
Some of this reduction in cost is due to many appliances being manufactured in other countries with lower labor costs, but even those produced domestically have not increased in price as much as the general price level.
Such examples lead to one of the knottiest problems in measuring price changes. How do you account for changes in quality? A can of tuna or a cotton T-shirt in 2011 is little different from those items in 1960. But a clothes dryer, a television, a tire or an automobile today is superior in many ways to those items a half-century ago.
In 1964, the filling station owner in our town was amazed that the Lotterman kids had gotten 23,000 miles on the original tires on the Volkswagen they drove to school. At that time, if you got 18,000 miles out of a set of American-made tires on a ’62 Chevy, you were lucky. Now, people routinely get 60,000 or more. Yet, modern tires cost about three times in nominal terms what they did 47 years ago, while the CPI has increased by a factor of seven.
Estimating the value to the consumer of a tire that goes 60,000 miles instead of 15,000 miles or a refrigerator that uses less electricity and has an icemaker isn’t easy. The common method is called “hedonic regression,” a statistical procedure that breaks out individual characteristics of a specific product and consumer willingness to pay for each. The technique also is used in real estate economics to deal with the fact that not all commercial buildings or houses, even of the same square footage, have identical characteristics or are equally desirable.
Most economists think this is the best approach although far from an infallible one. But some critics, mostly from outside the discipline, see this as a way to systematically underestimate inflation.
The final thing to note is that while refrigerators, clothes dryers, TVs and tuna were all common items in the 1960s that remain as widely used today, that is not true for vinyl records, women’s girdles or console stereos. And cellphones, personal computers, arterial stents and myriad other items common today were unheard of only a few decades ago.
Such changes in consumption patterns pose other challenges in measuring general price changes. How do you update the “market basket” or list of goods and services whose prices are measured periodically to construct the index? This involves compiling extensive data on household consumption patterns. In the past, this was done at intervals of several years. But changes can occur quickly. Reportedly, nearly 50 million cellphones were in use by 1996, but they did not enter CPI calculations until 1998.
Constructing price indexes is difficult. Most economists agree that the ones we have here in our country are about as accurate as is possible. But many in the general public will always think the books are cooked.
© 2011 Edward Lotterman
Chanarambie Consulting, Inc.