Print“I gather, young man, that you wish to be a Member of Parliament. The first lesson that you must learn is, when I call for statistics about the rate of infant mortality, what I want is proof that fewer babies died when I was Prime Minister than when anyone else was Prime Minister. That is a political statistic.” — Sir Winston Churchill
There are three significant criticisms of the Consumer Price Index (CPI) methodology. The first is that the substitution of items in the basket of goods is not appropriate (the “hamburger versus steak” argument). The second is how to account for changes in the quality of goods sold. The third is the use of rental equivalence to measure housing costs and their contribution to the CPI.
Shadowstats argues that the CPI is underestimated by about 7 percent per year, based on disagreements with the Bureau of Labor Statistics (BLS) methodologies for these three adjustments. If that’s the case, it represents a significant risk to you and your future spending power. We will examine all three criticisms, including the Shadowstats argument and the BLS response, and close with a look at the actual data.
The hamburger versus steak argument can be simply stated: the BLS methodology assumes that if steak gets too expensive, consumers will switch to hamburger, and the inflation rate will reflect that switch by replacing steak with hamburger in the basket of goods. The idea behind the criticism is that the CPI adjusts the basket of goods to reflect relative price levels. In other words, if steak gets too expensive, the CPI calculation substitutes hamburger on the grounds that consumers will switch to hamburger when they can’t afford to buy steak. (Presumably, when hamburger gets too expensive, consumers will switch to pet food, which will then replace hamburger in the index.) According to this argument, the CPI, as calculated, understates inflation, as the changing basket of goods represents a declining standard of living. This argument has been made widely in magazines and on the Internet, most clearly in the Shadowstats primer on inflation.
The BLS is well aware of the criticism and, in the August 2008 Monthly Labor Review, took it on directly: “It must be stated unequivocally that the BLS does not assume that consumers substitute hamburger for steak.” So there! The article explains that any substitution that does occur is within a class—that is, one type of steak for another. Hamburger is in another category—say, “ground beef in Chicago,” not “beefsteak in Chicago.” In fact, the article goes on to say, the change in purchasing is based on relative price changes, so if a more expensive item, like filet mignon, increased by less than flank steak, the basket would actually reweight to the more expensive item.
The BLS provides additional information, with examples of how the strict “basket of identical goods” approach provides an upper limit to the level of price inflation, but it’s probably not the best indicator. So, while the specific example is wrong, the central issue of what constitutes a substitutable item remains. Overall, as the BLS is well aware of the issue, egregious examples like hamburger for steak aren’t likely to exist in general. The effect of substitution is real, however, even though it probably doesn’t affect the CPI estimates to anywhere near the extent estimated by Shadowstats or other hostile analysts.
Recently, technology has given us the tools to examine in real time what prices are doing. The best example of this is MIT’s Billion Prices Project, which we will touch on in part 5 of this series. But first, in part 4, we’ll focus on the remaining significant criticism of the CPI, that of how housing costs are handled.
