The goal of the consumer price Index is to measure changes in the cost of living. In other words, the consumer price Index tries to gauge how much incomes must rise to maintain a constant standard of living. The consumer price index, however, is not a perfect measure of the cost of living. Three problems with the index are widely acknowledged but difficult to solve.
The first problem is called substitution bias. When prices change from one year to the next, they all do not change proportionately: Some prices rise more than others. Consumers respond to these differing price changes by buying less of the goods whose prices have risen by large amounts and by buying more of the goods whose prices have even fallen. That is, consumers substitute towards goods that have become relatively less expensive. If a price index is computed assuming a fixed basket of goods, it ignores the possibility of consumer substitution and therefore overstates the increase in the cost of living from one year to the next.
Let us consider a simple example. Imagine that in the base year apples are cheaper than pears, and so consumers buy more apples than pears. When the Bureau of Labour Statistics constructs the basket of goods, it will include more apples than pears. Suppose that next year pears are cheaper than apples. Consumer will naturally respond to the price changes by buying more pears and fewer apples. Yet when computing the consumer price Index the BLS uses a fixed basket, which in essence assumes that consumers continue buying the now expensive apples in the same quantities as before. For this reason the index will measure a much larger increase in the cost of living than consumers actually experience.
The second problem with the consumer price index is introduction of new goods. When a new good is introduced consumers have more variety from which to choose and this in turn reduces the cost of maintaining the same level of economic well-being To see why, consider a hypothetical situation: Suppose you could choose between a Rs1000 gift certificate at a large store that offered a wide array of goods and a Rs500 gift certificate at a small store with the same prices but with a more limited selection. Which would you prefer? Most people would pick the store with greater variety. In essence the increased set of possible choices makes each rupee more valuable. The same is true with the evolution of economy over time: As new goods are introduced consumers have more choices, and each rupee is worth more. Yet because price index is based on a fixed basket of goods and services, it does not reflect the increase in the value of the dollar that arises from the introduction of new goods.
Again, let’s consider an example. When VCRs were introduced consumers were able to watch their favourite movies at home. Although not a perfect substitute for a first run movie on a large screen, an old movie in the comfort of your family room was a new option that increased set of opportunities. For any given number of dollars the introduction of the VCR made people better off; conversely to achieve the same level of economic well-being required a smaller number of dollars. A perfect cost–of–living index would have reflected the introduction of the VCR with a decrease in the cost of living. The consumer price index however, did not decrease in response to the introduction of the VCR. Eventually the Bureau of Labour Statistics did revise the basket of goods to include VCRs, and subsequently the index reflected changes in VCR prices. But the reduction in the cost of living associated with the initial introduction of the VCR never showed up in the index.
The third problem with the consumer price index is unmeasured quality change. If the quality of a good deteriorates from one year to the next, the value of a dollar falls, if the price of the good stays the same, because you are getting a lesser good for the same amount of money. The Bureau of Labour Statistics does its best to account for quality change. When the quality of a good in the basket changes – for example, when a car model has more horsepower or gets better gas mileage from one year to the next – the Bureau adjusts the price of the good to account for the quality change. It is, in essence trying to compute the price of a basket of goods of constant quality. Despite these efforts changes in quality remain a problem because quality is so hard to measure.
The Bureau of Labour Statistics adopted several technical changes to improve the CPI, and many economists believe the bias is now only about half as large as it once was. The issue is important because many government programs use the consumer price index to adjust for changes in the overall level of prices, Recipients of Social Security for instance get annual increases in benefits that are tied to the consumers price index.