CONSUMER PRICE INDEX (CPI)

The Consumer Price Index (CPI), sometimes called the cost-of-living index, measures the average change in prices that typical American wage earners pay for basic goods and services, such as food, clothing, shelter, transportation, and medical care. It is expressed as a percentage of the cost of the same goods and services in a base period. For example, using the years 1982 to 1984 as a base period with a value of 100, the CPI for April 1995 was 151.9, meaning that prices had increased by an average of 51.9 percent over time. The CPI is often used to measure inflation, so it is closely monitored by government policymakers and by individuals whose wages vary with the purchasing power of money. The practice of indexing wages to the CPI is known as a cost-of-living adjustment (COLA). The term "cost of living" is often applied to the numerical result of the CPI. Loosely defined, it refers to the average cost to an individual of purchasing the various goods and services needed to maintain a reasonable living standard.

The U.S. Bureau of Labor Statistics (BLS) began calculating the CPI in 1917, and over the years it has become an important economic statistic. The CPI is calculated monthly and is usually reported within the first two weeks of the following month. In order to calculate the CPI, the BLS surveys about 24,000 households to find out where families shop regularly and what types of goods and services they purchase. It then contacts about 21,000 retail businesses in 85 major metropolitan areas to obtain prices for 90,000 items. All of this information is combined in the CPI, which represents the average price of a "market basket" of goods and services. The BLS tries to incorporate any new developments in the market by changing 20 percent of the retail outlets and items in its survey every year on a rotating basis.

A separate CPI is calculated for different income levels, geographical areas, and types of goods and services. For example, the CPI-U is calculated for all urban households, which includes about 80 percent of the U.S. population. In contrast, the CPI-W measures average price increases for the 32 percent of Americans who derive their primary income as wage earners or clerical workers. The BLS also publishes a CPI for each of seven major categories of items: food and beverages, housing, apparel, transportation, medical care, entertainment, and other goods and services. In addition, it compiles individual indexes for 200 different items and combined indexes for 120 smaller categories of items. Separate CPI measurements are also released for four major geographical regions of the United States—Northeast, North Central, South, and West—as well as 29 large metropolitan areas.

The CPI influences the American economy in several ways. A high annual percentage increase in the CPI reflects a high rate of inflation. The Federal Reserve Board, which controls the nation's money supply, often reacts to such increases by raising interest rates. This makes it more expensive for individuals and businesses to borrow money, which usually slows spending, encourages saving, and helps to curb inflation in the economy. The CPI also determines the percentage of annual increase or decrease in income for many Americans. For example, COLA formulas based on the CPI are built into many employment contracts. The federal government also uses the CPI to adjust Social Security and disability benefits, to determine the income level at which people become eligible for assistance, and to establish tax brackets. In addition, the CPI is often used to compare prices for certain goods within a set of years, and to calculate constant dollar values for two points in time.

Some economists believe that the CPI has historically overstated actual increases in the cost of living by I percent or more annually. They generally attribute the discrepancy to some combination of the following four factors: improvements in the quality of goods; the introduction of new goods; substitution by consumers of different goods or retail outlets; and the difficulty of measuring the prices consumers actually pay for goods. These measurement problems can be easily understood with the help of some examples. As Daniel Mitchell explained in Challenge, substitution by consumers means that when the price of apples rises in comparison to the price of bananas, consumers will not necessarily continue to purchase the base-period basket of goods. Instead, they might choose to buy more bananas and fewer apples, which would affect the CPI. Improvements in quality can affect the CPI in a similar way. A personal computer purchased in 1998 would have higher processing speed, larger memory, and more features than one purchased just a few years earlier. Thus it is not effective to track the price of computers over time without adjusting for the change in quality.

In 1996 the Boskin Commission produced a report criticizing the BLS and summarizing various problems with the CPI. Such criticisms were taken seriously because the CPI has such a significant effect on government revenues as well as the income of many Americans. Since that time, a series of modifications have been made to correct some of the technical deficiencies in the CPI. These changes have combined to reduce the index by .74 percentage points. The last change, which took place in January 1999, was intended to reflect how consumers shifted to lower-priced substitutes when prices rose on the products they purchased previously.