Attachment F: Experimental CPI for Americans 62 Years of Age and Older
Kenneth J. Stewart and Joseph Pavalone
Introduction
The Consumer Price Index (CPI) of the Bureau of Labor Statistics (BLS) measures the average change in prices over time for a fixed marketbasket of goods and services for two population groups. The CPI for All Urban Consumers (CPI-U) represents the spending habits of about 80 percent of the population of the United States. The CPI for Urban Wage Earners and Clerical Workers (CPI-W) is a subset of the CPI-U population, and represents about 32 percent of the total U.S. population.
In addition to the official CPI's for the CPI-U and CPI-W populations, the CPI calculates an experimental price index for Americans 62 years of age or older. The Older Americans Act of 1987 directed the BLS to develop this experimental index.
A May 1994 Monthly Labor Review article by Nathan Amble and Kenneth J. Stewart, "Experimental Price Index for Elderly Consumers," detailed the methods, sources of data, and limitations of this experimental index series. It provided estimates of the series for all items and major CPI expenditure components from December 1982 through December 1993 and an analysis of price change measured by the experimental series during the 1987-93 period.
BLS plans the periodic release of updated estimates of the Experimental Price Index for the Elderly (CPI-E) in this publication, the Consumer Price Index Detailed Report.1 This article extends the time period of the published estimates of the experimental CPI-E index through December, 1995 (see table 1) and it repeats the description of the methods, sources of data, and limitations of the experimental index contained in the earlier article. Over the 5-year period from December 1990 through December 1995, the experimental price index has risen 15.9 percent. This compares to increases of 14.7 and 14.1 percent for the CPI-U and CPI-W, respectively.
Methodology, sources of data, and limitations.
Although this study indicates a higher overall inflation rate for older Americans compared to the official CPI population groups, any conclusions drawn from it should be used with caution because of the various limitations inherent in the methodology.
Expenditure weights. For each CPI population group, item strata are weighted according to their importance in the spending patterns of the respective population. The definition of the population of older Americans used for the experimental price index was all urban noninstitutionalized consumer units which met one of the following three conditions:
- Unattached individuals who were at least 62 years of age;
- Members of families whose reference person (as defined in the Consumer Expenditure Survey) or spouse was at least 62 years of age; or
- Members of groups of unrelated individuals living together who pool their resources to meet their living expenses, whose reference person was at least 62 years of age.
In the 1982-84 Consumer Expenditure Survey, which is used as the source of expenditure weights in the current CPI, 19 percent of the total sample of eligible urban consumer units (3,135 out of 16,500) met the above definition for older Americans. Because the number of consumer units used for determining weights in the experimental index was relatively small, expenditure weights used in the construction of the experimental price index have a higher sampling error than those used for the larger populations.
For each population group, the base expenditure weight of any component represents the actual expenditure on that component in the base period. The "relative importance" of any component is its base expenditure weight updated for changes in relative prices expressed as a percent of the total updated expenditures for the population. The relative importance data for each of the three population groups for December 1990 and 1995 are shown in table 2.
Areas and outlets priced. The CPI-E is a weighted average of price changes for the same set of item strata and collected from the same sample of urban areas used in calculating the CPI-U and CPI-W.
Retail outlets are selected for pricing in the CPI based on data reported in a separate survey representing all urban households. The experimental index also uses the same retail outlet sample. The outlets thus selected may not be representative of the places of purchase of the older populations.
Items priced. One major limitation of the experimental index is that the categories of items to be priced are selected using expenditure weights calculated from the expenditure surveys for the urban (CPI-U) population. As a result, the specific item classes selected for each stratum may not be representative of the experimental index population. In addition, the specific items selected for pricing within an outlet are selected based upon the distribution of total sales purchased by all customers. This distribution of sales may not be representative of the specific products purchased by the CPI-E population.
Prices collected. A final source of uncertainty about the appropriateness of using the CPI-U prices for the index of the older population concerns the availability of discount prices for older Americans. For example, senior-citizen discount rates are used in the CPI in proportion to their use by the urban population as a whole. To the extent that senior-citizen discounts take the form of a percentage discount from the regular price, this may not be a problem. If, however, the discount is not a fixed percentage of the price, the scarcity of senior-citizen discount prices in the current CPI could lead to error in the experimental index.
Because of the above limitations, any conclusions drawn from these analyses should be treated as tentative.
Relative behavior of price indexes
Table 3 shows the behavior of the CPI for all urban consumers (CPI-U), Urban Wage Earners and Clerical Workers (CPI-W) at the major component levels for the period 1990-95. Over the 5-year period from December 1990 through 1995, the reweighted experimental price index for older Americans (CPI-E) rose 15.9 percent. This compares with increases of 14.7 percent for the CPI-U and 14.1 percent for the CPI-W. The annual average rate of change in the CPI-E was 3.0 percent over this period, compared with 2.8 percent for the CPI-U and 2.7 percent for the CPI-W.
Price change for each major expenditure group varied by population because the distribution of expenditures on the products and services within the major groups varied among the three index populations. For example, the higher rate of increase in four of the seven major groups for the elderly population indicates that they allocate a larger portion of their total expenditures to the components of the groups that are increasing most rapidly. This is especially true within the medical care group. For example, the CPI-E population devoted a substantially larger share of their expenditures to health insurance (see table 2) than did the CPI-U and CPI-W populations, largely because of the availability of employer-provided health care benefits to the latter groups.
A more detailed examination of the indexes shows that the CPI-E had the highest rate of price increase of the three populations for four of the seven major groups. Medical care prices rose in excess of two times the rate of the average for all items in each population group during this 5-year period. Analysis of the relative importance data for the CPI-E, the CPI-U and CPI-W populations indicate that older Americans devote a substantially larger share of their total budgets to the medical care (see table 2). Because of this, and because the medical care component of the CPI showed the largest price increase, medical care accounts for most of the difference between the higher rate of increase in the CPI-E experimental index--as compared to the CPI-U and CPI-W indexes--during the 1990-95 period.
Within the medical care component, the experimental index population has larger out-of-pocket costs than the CPI-U and CPI-W populations as reflected in the difference for medical care services. Medical care services are afforded more weight than medical care commodities, which contains items such as prescription drugs for all three population groups (see table 2). Because the elderly become ill more frequently, with a higher degree of severity,they allocate a larger share of expenditures for services.
The other goods and services component, which also registered a large price increase between 1990-95, showed a slower rate of increase for the CPI-E population than it showed for the CPI-U and CPI-W populations. It rose 28.3, 26.6, and 25.1 percent in the CPI-U, CPI-W, and CPI-E populations, respectively. Within this major group, the CPI-E population allocated a smaller portion of their expenditures to personal and educational expenses, especially college tuition, which increased three times faster than the all-items index during the 1990-95 period.
The CPI and its relationship to Social Security benefits
Adjustments to Social Security benefits are currently based on the percent change in the CPI-W, measured from the average of the third quarter of one year to the third quarter of the succeeding year.
While the Senate Special Committee on Aging specified the population to be covered for this study to be persons 62 years of age and older, this is not likely to be the most appropriate population definition for developing an index for use in indexing Social Security benefits.
First, many Social Security beneficiaries are younger than 62 years of age, and receive benefits because they are surviving spouses and/or minor children of covered workers or because of disability. The spending patterns of this younger group are excluded in the weights for the experimental index for older Americans. Second, a substantial number of persons 62 years of age and older do not receive Social Security benefits, especially those 62-64 years of age. Although these older consumers are included in the population covered by the experimental reweighted index, they presumably should be excluded from an index designed to reflect the experience of Social Security pensioners.
In short, an index designed specifically to measure price change for Social Security beneficiaries--i.e., one that excludes older people not receiving benefits, but includes younger persons receiving survival and disability benefits--might well show price movements that differ significantly from those of the experimental index in this study.
Conclusions
This report analyzed the changes in the prices of three population groups: CPI-U, CPI-W, and the CPI-E, the experimental population of Americans at least 62 years of age, for the period December 1990 through December 1995. Analysis of the relative behavior of the three indexes at the all-items level revealed that the experimental index, which rose 15.9 percent during this period, had a higher rate of increase than the indexes for CPI-U and CPI-W, which rose 14.7 and 14.1 percent, respectively.
The experimental price index, reweighted to incorporate the spending patterns of older consumers, behaved more like the CPI-U than the CPI-W. This was expected, because the CPI-U includes the expenditures of all urban consumers, including those 62 years of age and over. The CPI-W, however, is limited to the spending patterns of wage-earner and clerical families and, therefore, specifically excludes the experience of families whose primary source of income is from retirement pensions.
Finally, the medical care component of the CPI has a substantially larger relative weight in the experimental population compared to the CPI-U or CPI-W. As a result, the medical care component of the CPI-E tends to have a larger effect on the elderly population than it does on the other two populations. However, the experimental price index has limitations as an estimate of the inflation rate experienced by older Americans. Because of the various limitations inherent in the methodology, any conclusions drawn from the data should be interpreted with caution.
Table 1. CPI-E January 1994 through 1995, for all items and for major CPI expenditure components, and December only 1990 through 1993
1994 Jan. Feb. Mar. Apr. May June July Aug. Sep. Oct. Nov. Dec. All items 154.4 155.0 155.6 155.7 155.8 156.3 156.9 157.5 157.7 157.9 158.0 158.0 Food and beverages 148.8 148.0 148.3 148.5 148.7 148.7 149.7 150.4 150.6 150.5 150.8 152.5 Housing 149.4 150.4 150.8 150.5 150.7 151.5 152.0 152.6 152.3 152.3 152.1 151.9 Apparel and upkeep 131.4 132.5 135.5 136.3 135.3 133.1 130.9 130.1 133.5 135.3 134.0 130.3 Transportation 137.7 138.1 138.6 138.9 138.7 139.5 140.3 141.5 141.3 141.4 142.2 142.0 Medical care 220.2 221.7 222.4 223.4 224.1 224.8 226.2 227.1 227.8 229.1 229.8 230.6 Entertainment 157.9 158.5 158.8 159.3 159.4 159.2 159.6 159.4 160.2 160.5 161.1 161.1 Other goods and services 187.3 187.4 187.7 188.9 189.6 190.1 190.4 191.4 192.2 192.8 193.3 193.7 1995 All items 158.9 159.5 160.0 160.5 160.8 161.2 161.5 161.9 162.1 162.5 162.5 162.4 Food and beverages 153.2 153.0 153.0 154.0 154.0 153.7 153.8 154.1 154.8 155.2 155.1 155.6 Housing 153.2 153.9 154.4 154.6 154.7 155.6 156.4 156.8 156.6 156.7 156.5 156.7 Apparel and upkeep 129.1 130.5 133.2 134.2 133.0 129.4 127.9 130.3 133.1 134.8 134.4 130.5 Transportation 142.3 142.4 143.0 144.0 145.2 146.2 145.5 144.4 144.0 144.7 144.7 144.0 Medical care 232.1 233.5 234.1 234.5 234.9 235.3 236.2 237.0 237.5 238.4 239.0 239.4 Entertainment 162.1 162.6 162.9 163.7 164.0 163.4 163.8 164.4 165.6 166.0 167.0 167.0 Other goods and services 194.1 195.3 195.3 195.7 196.4 196.8 197.3 198.8 200.7 201.2 201.7 201.8 Dec. 1990 Dec. 1991 Dec. 1992 Dec. 1993 All items 140.1 144.8 149.2 153.8 Food and beverages 138.7 142.0 143.4 147.7 Housing 136.1 140.8 144.5 148.6 Apparel and upkeep 122.4 126.5 131.2 133.2 Transportation 132.6 130.8 134.6 138.1 Medical care 178.9 194.1 206.9 218.7 Entertainment 141.3 147.5 152.2 157.0 Other goods and services 161.3 172.7 182.3 186.3
Table 2. Comparative analysis of CPI relative importance data of selected expenditure groups, December 1995.
Expenditure Group CPI-U CPI-W CPI-E All items 100.00 100.00 100.00 Food and beverages 17.33 19.26 15.00 Food at home 9.88 11.21 9.66 Food away from home 5.89 6.37 4.23 Alcoholic beverages 1.57 1.68 1.10 Housing 41.35 38.89 46.89 Shelter 28.29 25.98 33.88 Apparel and upkeep 5.52 5.53 3.93 Transportation 16.95 19.02 13.82 Medical care 7.36 6.26 12.14 Medical care commodities 1.28 1.06 2.57 Medical care services 6.08 5.21 9.57 Health Insurance .36 .25 1.09 Entertainment 4.37 4.03 3.35 Other goods and services 7.12 7.01 4.87 College tuition 1.61 1.19 0.59
Table 3. Percentage changes in the CPI-U, CPI-W, and CPI-E by major expenditure roup, December 1990--December 1995.
Expenditure Groups CPI-U CPI-W CPI-E All items: Total 14.7 14.1 15.9 Average annual rate 2.8 2.7 3.0 Food and beverages 12.2 12.1 12.2 Food at home 12.3 12.0 12.1 Food away from home 10.8 10.8 10.9 Alcoholic beverages 18.3 18.2 18.0 Housing 14.7 14.3 15.1 Shelter 17.3 17.1 18.1 Apparel and upkeep 4.2 4.0 6.6 Transportation 9.4 9.6 8.6 Medical care 32.3 32.2 33.8 Medical care commodities 22.2 21.5 23.1 Medical care services 34.6 34.6 37.1 Entertainment 16.0 15.4 18.2 Other goods and services 28.3 26.6 25.1 College tuition 49.6 47.7 43.4
Experimental Consumer Price Index Consumers Aged 62 and Older--UNPUBLISHED DATA U.S. city average All items December 1982=100 SEMIANNUAL 1ST 2ND PERCENT CHANGE YEAR JAN. FEB. MAR. APR. MAY JUNE JULY AUG. SEP. OCT. NOV. DEC. HALF HALF AVG. DEC-DEC AVG-AVG 1982 100.0 1983 100.4 100.5 100.6 101.2 101.7 102.0 102.4 102.7 103.2 103.4 103.5 103.7 101.1 103.2 102.1 3.7 1984 104.4 105.1 105.3 105.7 106.0 106.3 106.7 107.2 107.6 107.8 107.9 108.0 105.5 107.5 106.5 4.1 4.3 1985 108.3 108.8 109.2 109.7 110.1 110.5 110.8 111.1 111.4 111.7 112.1 112.4 109.4 111.6 110.5 4.1 3.8 1986 112.9 112.7 112.3 112.3 112.6 113.1 113.3 113.6 114.1 114.2 114.2 114.4 112.7 114.0 113.3 1.8 2.5 1987 115.2 115.7 116.1 116.7 117.1 117.7 117.9 118.6 119.0 119.3 119.5 119.5 116.4 119.0 117.7 4.5 3.9 1988 120.0 120.3 120.9 121.5 121.9 122.5 123.0 123.6 124.3 124.6 124.8 124.9 121.2 124.2 122.7 4.5 4.2 1989 125.7 125.3 127.1 127.9 128.6 129.0 129.6 129.8 130.0 130.8 131.1 131.4 127.3 130.5 128.9 5.2 5.0 1990 133.0 133.6 134.4 134.6 134.9 135.8 136.6 137.9 138.8 139.6 140.0 140.1 134.4 138.8 136.6 6.6 6.0 1991 141.2 141.6 141.9 142.0 142.4 142.9 143.2 143.6 144.0 144.1 144.5 144.8 142.0 144.0 143.0 3.4 4.7 1992 145.4 146.0 146.7 146.8 147.0 147.3 147.8 148.2 148.4 149.0 149.2 149.2 146.5 148.6 147.6 3.0 3.2 1993 150.1 150.7 151.2 151.7 152.0 152.2 152.4 152.8 152.9 153.4 153.6 153.8 151.3 153.2 152.2 3.1 3.1 1994 154.4 155.0 155.6 155.7 155.8 156.3 156.9 157.5 157.7 157.9 158.0 158.0 155.5 157.7 156.6 2.7 2.9 1995 158.9 159.5 160.0 160.5 160.8 161.2 161.5 161.9 162.1 162.5 162.5 162.4 160.2 162.2 161.2 2.8 2.9 1996 163.5 164.1 164.9 165.4 165.7 165.9 166.5 166.7 167.0 167.6 164.9 Experimental Consumer Price Index Consumers Aged 62 and Older--UNPUBLISHED DATA U.S. city average All items December 1982=100 SEMIANNUAL 1ST 2ND PERCENT CHANGE YEAR JAN. FEB. MAR. APR. MAY JUNE JULY AUG. SEP. OCT. NOV. DEC. HALF HALF AVG. DEC-DEC AVG-AVG 1982 100.0 1983 100.4 100.5 100.6 101.2 101.7 102.0 102.4 102.7 103.2 103.4 103.5 103.7 102.1 1984 104.4 105.1 105.3 105.7 106.0 106.3 106.7 107.2 107.6 107.8 107.9 108.0 106.5 4.1 4.3 1985 108.3 108.8 109.2 109.7 110.1 110.5 110.8 111.1 111.4 111.7 112.1 112.4 110.5 4.1 3.8 1986 112.9 112.7 112.3 112.3 112.6 113.1 113.3 113.6 114.1 114.2 114.2 114.4 113.3 1.8 2.5 1987 115.2 115.7 116.1 116.7 117.1 117.7 117.9 118.6 119.0 119.3 119.5 119.5 117.7 4.5 3.9 1988 120.0 120.3 120.9 121.5 121.9 122.5 123.0 123.6 124.3 124.6 124.8 124.9 122.7 4.5 4.2 1989 125.7 125.3 127.1 127.9 128.6 129.0 129.6 129.8 130.0 130.8 131.1 131.4 128.9 5.2 5.0 1990 133.0 133.6 134.4 134.6 134.9 135.8 136.6 137.9 138.8 139.6 140.0 140.1 136.6 6.6 6.0 1991 141.2 141.6 141.9 142.0 142.4 142.9 143.2 143.6 144.0 144.1 144.5 144.8 143.0 3.4 4.7 1992 145.4 146.0 146.7 146.8 147.0 147.3 147.8 148.2 148.4 149.0 149.2 149.2 147.6 3.0 3.2 1993 150.1 150.7 151.2 151.7 152.0 152.2 152.4 152.8 152.9 153.4