CONSUMER PRICE INDEX AND COLAS

The U.S. Consumer Price Index (CPI) is a measure of the average change in the prices paid by consumers living in urban areas for a bundle of consumption goods and services. Average price changes can be driven by changes in consumer income, population and demographic changes, and changes in consumer preferences, as well as by the introduction of new product distribution patterns and marketing techniques. The CPI is updated on a regular basis to respond to such changes. In addition, technological and methodological improvements are introduced into the CPI to improve accuracy.

The CPI is not designed to compare prices between areas. For example, one could say that the prices in Chicago increased by 2 percent from one period to another, while they increased by 5 percent in Washington, D.C. But it would be incorrect to say that prices in Washington, D.C., are higher than they are in Chicago, based on the CPI for these cities.

The official CPI is calculated and published by the Bureau of Labor Statistics (BLS), part of the U.S. Department of Labor. CPIs are published for metropolitan areas in the United States (thirty-eight in 2001), as well as for the four census regions and other areas. National indexes, known as the U.S. City Average, are also produced. Not all indexes are published monthly. Indexes are produced for over 200 basic item categories (for example, bananas, prescription drugs, electricity, women's dresses) and eight major groups or item aggregations. The major groups are food and beverages, housing, apparel, transportation, health care, recreation, education and communication, and other goods and services. The CPI includes government-charged user fees, such as water and sewerage charges, and automobile registration fees, along with taxes such as sales and excise taxes, but excludes income and Social Security taxes. The CPI does not include investment items such as stocks, bonds, real estate, and life insurance. For most of the basic items, the BLS chooses a sample of several hundred specific items within selected business outlets (including stores, mail order firms, and the Internet) frequented by consumers, using scientific sampling procedures, to represent the thousands of items available in the marketplace.

Over time, improvements in the CPI have been introduced. The primary and most visible change has been the introduction of a new "market basket" of goods and services upon which to base the indexes. This market basket is a reflection of what consumers (specifically consumer units, also referred to here as households; consumer units share major expenditures, while a group of people in a household may not) buy; data are collected using household expenditure surveys (also known as consumer expenditure surveys). Shares of total expenditures for items in the market basket are produced from the survey data. Shares are produced over all households within an area, and thus the market basket is considered to be "representative" of the purchases by households within an area. These shares are known as "expenditure weights" or "relative importances," and are attached to prices collected by the BLS to produce the CPI. Consumer expenditures from 1999 to 2000 are used (as of 2002) for expenditure weights for items in every CPI index area. To maintain accuracy, the CPI has historically updated the expenditure patterns approximately every ten years. Beginning in 2002, the expenditure weights are to be updated every two years.

The BLS officially produces an index for two population groups; other indexes are also produced, but are referred to as "experimental." The two official indexes are the Consumer Price Index for All Urban Consumers (CPI-U) and the Consumer Price Index for Urban Wage Earners and Clerical Workers (CPI-W). The CPI-U is based on the experience of consumers living in urban areas of the United States only. This population accounted for about 80 percent of the total U.S. noninstitutionalized population in 1978, when the index was introduced. The CPI-W is a subset of the CPI-U population and is based on the experience of wage earners and clerical workers living in urban areas in the United States. To be included in this population, more than half of the household's income must come from a clerical or wage occupation and at least one of the household's earners must have been employed for at least thirty-seven weeks during the previous twelve months. The CPI-W's population represented about 32 percent of the total U.S. population in 1978.

Uses of the CPI

The CPI is the primary source of information concerning trends in consumer prices and inflation in the United States. As a measure of inflation, it is one of the nation's most important economic indicators, and affects nearly all people living in the United States as well as many Americans living abroad. It is also used to create "constant dollar" key economic indicators. Such indicators include, but are not limited to, estimates of income, earnings, productivity, output, and poverty. The index is also widely used for economic analysis and policy formation in the public and private sectors.

The CPI is often employed as a cost-of-living adjustment (COLA), even though it is not a cost-of-living index. It is used to adjust wages and benefit payments to account for the erosion of consumer purchasing power due to price increases faced by consumers. The CPI is the best measure for adjusting wages and payments when the goal is to allow consumers to purchase, at today's prices, a market basket of goods and services equivalent to one that they could have purchased at an earlier time. This is the type of adjustment often called for when the CPI is applied to meet statutory obligations. Such obligations include payments to Social Security beneficiaries and federal and military retirees— all of which are tied to the CPI-W—and those for a number of entitlement programs, including food stamps and school lunches. The impact on the finances of the federal government is quite significant when such adjustments are made. Annual escalation adjustments in the federal income tax brackets, as well as personal exemption and standard deduction amounts, are made using the CPI-U. With regard to the impact on federal taxes, it has been estimated that in the fiscal year 1996, each 1 percent increase in the CPI resulted in a $5.7 billion increase in outlays and a $2.5 billion decline in tax revenues.

Difference between the CPI and COLI

The BLS uses a cost-of-living framework in making practical decisions about constructing the CPI, but the CPI differs in important ways from a complete cost-of-living index (COLI), as noted above. A COLI would measure changes over time in the amount that consumers need to spend to reach a certain level of living ("utility level," in economics terminology). Like the CPI, a COLI would reflect changes in the prices of goods and services, such as food and clothing, that are directly purchased in the marketplace. However, a complete COLI would go beyond this to take into account changes in governmental or environmental factors that affect consumers' well-being. It is very difficult to determine the correct treatment of public goods, such as safety and education, and other broad concerns, such as health, water quality, and crime, that would be in a complete cost-of-living framework.

History of improvements in the CPI

The CPI was begun during World War I. At this time, prices were increasing rapidly, particularly in shipbuilding centers. To deal with the fact that these price increases were eroding the buying power of wages, the federal government decided that an index on which to base cost-ofliving adjustments was necessary. To create such an index, household patterns of expenditures needed to be identified in order to create expenditure weights that would be used to compile the index. These data were collected in ninety-two industrial centers in 1917–1919. The BLS began the periodic collection of prices in 1919, and published indexes for thirty-two cities. Regular publication of the U.S. City Average began in 1921. At this time, indexes were estimated back to 1913. The first indexes were for wage earners and clerical workers.

As would be expected, the buying habits of consumers change substantially over time. In order to account for these changes, several household or consumer expenditure surveys have been conducted. After the 1917–1919 survey, others were conducted in 1934–1936, 1950, 1960–1961, and 1972–1973. In October 1979, continuous consumer expenditure surveys were introduced. A reason for continuous collection of these data is to have data available to update the CPI expenditure weights more frequently.

During these time periods many improvements in pricing and calculation methods were introduced. A Point of Purchase Survey was introduced in 1978. This survey made it possible for the BLS to collect prices directly, based on a new store-specific approach to select items. At this time, the BLS introduced a new index, the more broadly population-based CPI for All Urban Consumers (CPI-U). The CPI-U population included professional and salaried workers, part-time workers, the self-employed, the unemployed, and retired consumers in addition to wage earners and clerical workers.

The last two major revisions in the CPI took place in 1987 and 1998. With the 1987 revision, improvements were made in sampling, data collection, processing, and statistical estimation. Data from the consumer expenditure surveys of 1982–1984 were used for the expenditure weights. The 1980 census was the sample base, and the ongoing Point of Purchase Survey was used to select new item and outlet samples.

With the 1998 revision, another new market basket was introduced, along with methodological changes, a new system for data collection and processing, and a new sample base (the 1990 census). This is the sixth major revision in the CPI's history, and includes changes in the selection and classification of areas, items, and outlets. Several other improvements reflected in the current CPI were not part of a major CPI revision. For example, in 1983, a major change was introduced to separate shelter costs from the investment component of home ownership.

Issues in constructing an appropriate index

The CPI is the aggregate, representative measure of price change as experienced by households. In order to compute the index, one needs to know who buys, where they buy, what they buy, what an item costs (for example, the price of a loaf of bread is $1.00), and what in total is spent for an item (for example, $5.00 is spent for five loaves of bread purchased by the consumer). Statistical samples of household expenditures, prices, and area populations for urban areas provide the inputs for the calculation. Information is collected from a sample of urban households or consumer units to determine their expenditure patterns by using the consumer expenditure survey. Information on prices is collected from a sample of outlets and products based on their likelihood of being patronized and purchased, respectively. The overall CPI is then constructed in two stages by taking a weighted average of household information. At the first stage, prices within CPI item categories are averaged (using appropriate statistical techniques) to form basic item indexes. At the second stage, these indexes are averaged together.

Limitations of the CPI

There are several limitations of the CPI. Two sets of limitations of the CPI are discussed here: application and measure. The CPI may not be applicable to all population groups. Since it is designed to measure the experience of people living in urban areas, it may not accurately reflect the experience of people living in rural areas. Also, the CPI does not separately reflect the inflation faced by subgroups of the population, such as older persons, the poor, various ethnic groups, and males versus females. As noted earlier, the CPI measures only time-to-time changes in relative prices within an area. A higher index would not necessarily mean that prices are higher in one area compared with another area with a lower index value. The CPI cannot be used as a measure of the total change in living costs because social and environmental changes and changes in income taxes, which are beyond the scope of the CPI, are excluded.

Limitations on measure include sampling errors and nonsampling errors. Since the CPI is based on a sample of items, the index will differ from one based on all goods and services purchased in the United States. Nonsampling errors can be the result of problems in data collection, logistical lags in conducting surveys, difficulties in defining basic concepts and their operational implementation, and difficulties in handling the problems of quality change. Such errors can lead to persistent bias in the index.

Critiques, research, and experimental CPIs

There have been many improvements in the CPI over the years, a number of them resulting from research conducted within the BLS. The latest major external review of the CPI followed Senate Finance Committee hearings in the spring of 1995. As a result of the hearings, the committee appointed an advisory commission to study the CPI and make recommendations for methodological improvements. The commission's report, Toward a More Accurate Measure of the Cost of Living, informally referred to as the "Boskin Commission Report," did not identify new issues but did result in focusing greater attention on known limitations of the index.

The CPI program has developed experimental indexes of consumer prices. Only two are described here.

The one most relevant to this volume is referred to as the Experimental Consumer Price Index for Americans 62 Years of Age and Older (CPI-E). The question addressed with this index is how price increases faced by older persons living in urban areas of the United States differ from those faced by all consumers living in urban areas.

The other is the Consumer Price Index Research Series Using Current Methods (CPI-URS). This index incorporates most of the improvements in the CPI-U that have been made since 1978. A question that could be addressed with this experimental index is what would have been the measured rate of inflation since 1978 if the methods currently used in calculating the CPI-U had been in use since 1978. When consistent CPIs are preferred for deflating, this measure gives somewhat different results than the official CPI-U. Such applications include the Bureau of Economic Analysis's use of the CPI-U-RS in its 1999 comprehensive revision of the National Income and Product Accounts; the BLS Office of Productivity's use of the index in its measure of real hourly compensation for its quarterly measure of labor productivity and costs; and the Bureau of the Census's use of the CPI-U-RS for its estimates of historical real income.

The CPI-E

In 1988 the BLS first produced monthly and annual experimental CPIs for Americans age sixty-two and over, with the series going back to December 1982. The index is referred to as the CPI-E. The creation of the index evolved from the 1987 amendments to the Older Americans Act of 1965, in which the BLS was directed to develop an index for older Americans.

The older population, upon which the experimental index is based, is identified as being age sixty-two or older and living in urban areas. Specifically this population includes the following:

1. Unattached individuals who are at least sixty-two years of age
2. Members of families whose reference person (as defined in the consumer expenditure surveys) or spouse is at least sixty-two years of age
3. Members of groups of unrelated individuals living together who share living expenditures, and whose reference person is at least sixty-two years of age.

About 22 percent of urban consumer units (upon which the CPI-U is based) meet the above definition, using data from the consumer expenditure survey data collected in 1993–1995.

Expenditures of some older people are excluded, while those of people younger than age sixty-two are included. For example, older people who are living in institutionalized housing are not included in the sample, nor are older people who are living with adult children. However, expenditures of children or other related individuals living in an older person's household are included.

Over the period 1983–2000, the CPI-E rose by an annual average of 3.5 percent, while the CPI-U averaged 3.3 percent. The CPI-W, the index to which Social Security is tied, increased an average of 3.2 percent.

As noted above, the CPI-W is used to adjust Social Security payments and many retirement payments. However, there are problems with both the experimental index and the CPI-W in terms of applicability to Social Security beneficiaries and to the older population specifically. With regard to the CPI-W, the population upon which this index is based does not include most Social Security beneficiaries, since by definition the population is composed of wage earners and clerical workers. The CPI-E also has several limitations as a potential index for escalating Social Security benefits. First, the group of persons age sixty-two and older is not likely the most appropriate population. Many people younger than sixty-two receive Social Security benefits. About 25 percent of all Social Security beneficiaries are younger people who receive benefits because they are surviving spouses and/or minor children of covered workers, or because of disability. The expenditure experience of this group is not included in the weights for the CPI-E. In addition, a sizable number of people age sixty-two and over do not receive Social Security benefits. In 1988, over 40 percent of the population age sixty-two to sixty-four did not collect Social Security retirement benefits. One might question whether such people's expenditures ought to be included in the calculation of an index for older persons if the goal is to adjust the Social Security benefits.

In addition, the outlets, items, and prices used for calculating the CPI-E are the same as those used for the CPI-U. This means that the way the CPI-E population shops and the prices they pay are not being accounted for separately in the experimental CPI-E. Older people not only may shop at different outlets than people in the CPI-U population, but also may use discounts that are not available to other consumers. Without the development of a series of household surveys for the older population, for example, that obtains detailed descriptions of items purchased by this group and the identification of the outlets where the items were purchased, a full-scale CPI for older persons is not possible within the BLS. To provide such an index would be costly and likely would take years to implement. Much research and development must be done.

Thesia I. Garner Kenneth J. Stewart

See also Consumption and Age; Medicare; Pensions; Social Security; Supplemental Security Income.

BIBLIOGRAPHY

Amble, N., and Stewart, K. J. "Experimental Price Index for Elderly Consumers." Monthly Labor Review 117, no. 5 (May 1994): 11–16.

Boskin, M. J.; Dalberger, E.; Gordon, R. J.; Griliches, Z.; and Jorgenson, D. W. Toward a More Accurate Measure of the Cost of Living. Final Report of the Advisory Commission to Study the Consumer Price Index. S.Prt. 104-72. Washington, D.C.: U.S. Government Printing Office.

Bureau of Labor Statistics. "Frequently Asked Questions." World Wide Web document, April 2001. stats.bls.gov/cpifaq.htm

Bureau of Labor Statistics. "Home Page." World Wide Web document, April 2001. stats.bls.gov/cpihome.htm

Consumer Price Index. Hearings before the Committee on Finance, U.S. Senate. Washington, D.C.: U.S. Government Printing Office, 1995.

Diewert, W. E. "Index Numbers." In The New Palgrave: A Dictionary of Economics. Edited by John Eatwell, Murray Milgate, and Peter Newman. Volume 2. London: Macmillan, 1987. Pages 767–779.

Greenlees, J. S., and Mason, C. C. "Overview of the 1998 Revision of the Consumer Price Index." Monthly Labor Review. 119, no. 12 (December 1996): 3–83.

Handbook of Methods. Bulletin 2285. Washington, D.C.: Bureau of Labor Statistics, 1997. Pages 167–231.

Mason, C. C. "An Analysis of the Rates of Inflation Affecting Older Americans Based on an Experimental Reweighted Consumer Price Index." Report prepared by the Bureau of Labor Statistics for presentation to the Senate Select Subcommittee on Aging, June 1988.

Scholl, K. K. "Cost-of-Living Adjustments (COLAs)." Encyclopedia of Financial Gerontology. Edited by Lois A. Vitt, Jurg K. Siegenthaler, et al. Westport, Conn.: Greenwood Press, 1996, Pages 98-102.

Stewart, K. J., and Reed, S. B. "Consumer Price Index Research Series Using Current Methods, 1978–1999." Monthly Labor Review 122, no. 6 (June 1999): 29–38.

U.S. Senate, Special Committee on Aging. Developing a Consumer Price Index for the Elderly. Washington, D.C.: U.S. Government Printing Office, 1987.