How price indexes affect BLS productivity measures

Monthly Labor Review

Volumn 122, Issue 2, 1999, Pages 35-46

  Eldridge, Lucy P ^a  

^a NONE

Author keywords

[No Author keywords available]

Indexed keywords

EID: 0009375251     PISSN: 00981818     EISSN: None     Source Type: Journal    
DOI: None     Document Type: Article

References (59)

1. The major-sector and industry-level measures are developed independently. Measures of productivity for industries are based on the production of an industry and thus rely heavily on industry producer price indexes. Relatively few industry measures, approximately 14 percent of industries published by BLS, make use of consumer price indexes. In addition, BLS industry productivity statistics do not use input data to measure output and they do not use input data to construct output deflators. For more information, see BLS Handbook of Methods, Bulletin 2490 (Bureau of Labor Statistics, 1997), pp. 89-121.
2. BLS publishes labor productivity measures for six major sectors. Business sector, nonfarm business sector and nonfinancial corporations make use of BEA data; manufacturing, durable goods manufacturing, and nondurable goods manufacturing do not.
3. Allan H. Young and Helen Stone Tice, "An Introduction to National Economic Accounting," Survey of Current Business, March 1985, pp. 59-76.
4. The gross product of general government is the sum of government expenditures on compensation of general government employees and the general government consumption of fixed capital, which measures the services of general government fixed assets. Government expenditures on goods and services purchased from the private sector are not excluded from business sector output. The gross product of private households is the compensation of paid employees of private households. The gross product of nonprofit institutions serving individuals is the compensation paid to employees of these institutions.
5. The rental value of buildings and equipment owned and used by non-profit institutions serving individuals is measured as the sum of consumption of fixed capital, indirect business taxes, and interest paid.
6. It should be noted that this article discusses components of total GDP and business sector GDP, which are derived by aggregating final expenditures on goods and services. The output of individual industries, such as BEA's data on gross product originating (GPO) by industry or the gross output of industries is not discussed. For the latest BEA estimates of gross product by industry and a discussion of the relationship between the estimates of GDP by expenditures and by industry, see BEA's Survey of Current Business, pp. 20-40.
7. For a detailed discussion of BEA's new measures of output, see Steven J. Landefeld and Robert P. Parker, "Preview of the Comprehensive Revision of the National Income and Product Accounts: BEA's New Featured Measures of Output and Prices," Survey of Current Business, July 1995, pp. 31-38.
8. A listing of the product-side components of the national accounts, the methods used to calculate real output, and the source data is published annually by BEA in the Survey of Current Business. The most recent listing appears in the September 1998 issue, pp. 14-35.
9. The real gross product of general government, private households, and owner-occupied housing are included in total GDP as components of expenditures, and removed by BEA to construct business sector output measures for BLS. However, the output of nonprofit institutions enters GDP through a variety of product-side expenditure categories and then is removed by BEA as two income-side components. The first component is an independent aggregate based on compensation of employees of nonprofit institutions, referred to as gross product of nonprofit institutions. This series is constructed as the nonprofits' share of industry employee compensation, as defined by legal form of organization. The compensation of employees of nonprofit institutions is deflated by BEA using a BEA index of compensation per hour for those industries in which nonprofits are concentrated. This deflator is constructed using primarily BLS wages and hours data. The second component is the annual series of rental value of buildings and equipment owned and operated by nonprofit institutions. This component is deflated using an implicit price deflator constructed from the BEA nonresidential building series, which is calculated using a BEA composite cost index. Hence, both of these real components for nonprofits are constructed using input-based deflators. In order to prepare data on the methods used to calculate the various components of real output, this article subtracts the value of these components from the total amount of GDP that is constructed using input-based deflators. This technique is reasonable because the product-side components that reflect nonprofit activities (for example, nonprofit hospitals, clubs and fraternal organizations, religious and welfare activities, and education and research) are also constructed using input-based deflators.
10. "Updated Summary Methodologies," Survey of Current Business, September 1998, pp. 14-35.
11. "Updated Summary Methodologies."
12. Based on 1997 current-dollar data, personal consumption expenditures account for about 68 percent of GDP and 73 percent of business sector output.
13. Other composite indexes refer to BEA deflators that are composites of input data and price indexes. In instances when output data is not available at a lower level of detail, this study groups these components in the category of deflation by other indexes. Approximately 11 percent of all of the "other price indexes" used to construct both GDP and business sector output are composites of input data and price indexes.
14. For military compensation, hours worked are adjusted for changes in employment by rank and length of service. For civilian defense employee compensation, hours worked are adjusted for changes in employment by grades. For State and local government education compensation, hours worked are adjusted for changes in education and experience. For additional details, see "Government Transactions," Methodology Article Series MP-5, November 1998.
15. "Updated Summary Methodologies."
16. As part of the 1998 revision, the sample was updated and item categories were redesigned to allow for the introduction of new items. In 1999, the sample rotation is being redesigned to accelerate the introduction of new items and outlets. For an overview of the 1998 revision, see John S. Greenlees and Charles C. Mason, "Overview of the 1998 revision of the Consumer Price Index, Monthly Labor Review, December 1996, pp. 3-9; see also, Brent R. Moulton, "Bias in the Consumer Price Index, What is the evidence?" The Journal of Economic Perspectives, Fall 1996, pp. 159-77.
17. As part of the 1998 revision, the sample was updated and item categories were redesigned to allow for the introduction of new items. In 1999, the sample rotation is being redesigned to accelerate the introduction of new items and outlets. For an overview of the 1998 revision, see John S. Greenlees and Charles C. Mason, "Overview of the 1998 revision of the Consumer Price Index, Monthly Labor Review, December 1996, pp. 3-9; see also, Brent R. Moulton, "Bias in the Consumer Price Index, What is the evidence?" The Journal of Economic Perspectives, Fall 1996, pp. 159-77.
18. Moulton, "Bias in the Consumer Price Index."
19. The components are presented in the seven major-item groups of the CPI. The expenditure components of the business sector have been regrouped into these CPI groups. In January 1998, the CPI redesigned its major groups to include an eighth major group, education and communications. This new group contains components previously in housing and other goods and services.
20. Toward a More Accurate Measure of the Cost of Living, final report to the Senate Finance Committee from the Advisory Commission to Study the Consumer Price Index, S.Prt. 104-72 (U.S. Senate, Committee on Finance, December 1996.)
21. "Measurement Issues in the Consumer Price Index" (Bureau of Labor Statistics, June 1997). Prepared in response to a letter from Congressman Jim Saxton to Katharine G. Abraham, Commissioner of Labor Statistics.
22. Upper-level substitution bias refers to a measurement error that occurs as the strata indexes are aggregated to form the all-item CPI, using fixed (Computer Expenditure Survey) base-period weights; this can occur if consumers substitute across categories, or strata, of goods and services as relative prices change. Lower-level substitution bias refers to a measurement error that occurs as price data are aggregated to calculate the component, or strata, indexes, using fixed (Point-of-Purchase Survey and Computer Expenditure Survey) base-period weights; this can occur as consumers substitute among items within a specific category, or item strata, as relative prices change.
23. "Measurement Issues in the Consumer Price Index."
24. As part of the 1998 CPI revision, BLS introduced new expenditure weights used for upper-level aggregation. Beginning with January 1998 data, expenditure weights are based on 1993-95 spending patterns in order to reflect more accurately what consumers are buying.
25. The geometric mean index formula requires expenditure and price data for each period and makes use of these data to allow for substitution. This index assumes that customers will substitute to offset relative price changes in order to maintain constant expenditure shares for each item.
26. For a review of studies on upper-level substitution bias, see Ana M. Aizcorbe and Patrick C. Jackman, "The commodity substitution effect in CPI data, 1982-91," Monthly Labor Review, December 1993, pp. 24-33. For a review of studies on lower-level substitution bias, see Brent R. Moulton, "Basic components of the CPI: estimation of price changes," Monthly Labor Review, December 1993, pp. 13-24; and Brent R. Moulton and Karin E. Smedley, "A Comparison of Estimators for Elementary Aggregates of the CPI," paper presented at the Western Economic Association International Conference, San Diego, CA, July 7, 1995.
27. For a review of studies on upper-level substitution bias, see Ana M. Aizcorbe and Patrick C. Jackman, "The commodity substitution effect in CPI data, 1982-91," Monthly Labor Review, December 1993, pp. 24-33. For a review of studies on lower-level substitution bias, see Brent R. Moulton, "Basic components of the CPI: estimation of price changes," Monthly Labor Review, December 1993, pp. 13-24; and Brent R. Moulton and Karin E. Smedley, "A Comparison of Estimators for Elementary Aggregates of the CPI," paper presented at the Western Economic Association International Conference, San Diego, CA, July 7, 1995.
28. For a review of studies on upper-level substitution bias, see Ana M. Aizcorbe and Patrick C. Jackman, "The commodity substitution effect in CPI data, 1982-91," Monthly Labor Review, December 1993, pp. 24-33. For a review of studies on lower-level substitution bias, see Brent R. Moulton, "Basic components of the CPI: estimation of price changes," Monthly Labor Review, December 1993, pp. 13-24; and Brent R. Moulton and Karin E. Smedley, "A Comparison of Estimators for Elementary Aggregates of the CPI," paper presented at the Western Economic Association International Conference, San Diego, CA, July 7, 1995.
29. Moulton and Smedley, "A Comparison of Estimators," July 1995.
30. The seasoning method used to estimate the CPI component indexes is also unbiased if the following assumptions hold: stationarity, small correlation of relative prices over time, and utility functions characterized by constant elasticity of substitution. See Matthew D. Shapiro and David W. Wilcox, "Mismeasurement in the Consumer Price Index: An Evaluation," in Ben S. Bernanke and Julio J. Rotemberg, eds. NBER Macroeconomic Annual 1996 (Cambridge, MA, MIT Press, 1996).
31. Katharine G. Abraham, Commissioner of Labor Statistics, "Research issues related to the geometric mean formula for elementary indexes," attachment C of materials presented by the author in testimony given to the Senate Budget Committee, January 30, 1997.
32. In the January 1996 comprehensive revision of the national income and product accounts, BEA incorporated the adjustments made to the CPI'S for food and housing. BEA incorporated these changes in revised estimates of output beginning with 1978 data.
33. Improvements in CPI methodologies after 1993 are discussed in Paul A. Armknecht, Brent R. Moulton, and Kenneth J. Stewart, "Improvements to the Food at Home, Shelter, and Prescription Drug Indexes in the U.S. Consumer Price Index," BLS Working Paper 263, February 1995.
34. See "Extending the improvements in the CPI sample rotation procedures and improving the procedures for substitute items," CPI Detailed Report: Data for March 1996, April 1996, pp. 4-5.
35. See "The experimental CPI using geometric means," CPI Detailed Report: Data for March 1997, May 1997, pp. 5-18.
36. See "Planned change in the Consumer Price Index formula," CPI Detailed Report: Data for April 1998, June 1998, pp. 6-8.
37. There are 15 categories that will continue to be calculated as they are currently. These categories are: rent of primary residence; owners' equivalent rent of primary residence; housing at school, excluding board; electricity; utility natural gas service; residential water and sewerage maintenance; State and local registration, license, and motor vehicle property tax; telephone services, local charges; cable television; physicians' services; dental services; eyeglasses and eye care; services by other medical professionals; hospital services; and nursing homes and adult daycare.
38. The business sector estimates presented in table 4 are based on the difference between the CPI-U-XG and the CPI-U-XL at a finer level of detail than the seven categories presented. The detailed differences are weighted using business sector shares and aggregated to the seven major-item groups. The seven major-item groups reflect the categories used by BLS prior to January of 1998. Beginning in January 1998, BLS introduced a new structure of eight major-item groups. Paul Liegey, of the BLS Division of Consumer Prices and Price Indexes, provided CPI-U-XG and CPI-U-XL measures for sub-indexes of the CPI for May 1996 and December 1997.
39. This value corresponds with the 0.2 bias estimate discussed in Kenneth Stewart, "The experimental CPI using geometric means," BLS working article, April 1997. The 0.2 level of bias is based on data from the middle of 1996 through December of 1997.
40. Moulton, "Bias in the Consumer Price Index."
41. Marshall B. Reinsdorf, "The Effect of Output Price Differentials on the U.S. Consumer Price Index," in Murray F. Foss, Marilyn Manser, and Allan H. Young, eds., Price Measurements and Their Uses, Studies in Income and Wealth, vol. 57 (Chicago, University of Chicago Press for NBER, 1993).
42. David E. Lebow, John M. Roberts, and David J. Stockton, "Monetary Policy and 'The Price Level'" (Board of Governors of the Federal Reserve System, August 1994).
43. Moulton, "Bias in the Consumer Price Index."
44. In a more recent study, "Price dispersion, seller substitution and the U.S. CPI," BLS Working Paper 252 (Bureau of Labor Statistics, 1994), Reinsdorf shows that, for CPI food indexes, over the period from 1948 to 1963 and from 1967 to 1976, price decline due to new outlets was about 0.1 percent per year. Although this estimate is smaller than the 0.25 percent estimate of his earlier study, there is insufficient evidence to generate estimates of bias for other CPI component indexes.
45. For a detailed discussion of quality change in the CPI, see Brent R. Moulton and Karin E. Moses, "Addressing the Quality Change Issue in the Consumer Price Index," Brookings Papers on Economic Activity, no. 1, 1997.
46. This quality effect is based on an arithmetic method of aggregation. If outliers - which may be a result of sample turnover - are eliminated, the estimate drops to 1.1 percent. Alternatively, a logarithmic method of aggregation indicates a range of quality change of 0.28 to 0.44 percent. Ibid.
47. For all periods through 1997, BEA uses a quality-adjusted index for computers, and beginning in 1995, uses an alternative index for cellular phone service. Beginning with 1998 data, BEA will begin using the new CPI for computers, which uses a hedonic model to adjust prices for changes in quality. BEA uses PPI'S for for-profit and government hospitals, medical laboratories, physicians' services and home health care, beginning with 1993, 1994, 1995, and 1997, respectively.
48. Toward a More Accurate Measure of the Cost of Living.
49. Moulton and Moses, "Addressing the Quality Change Issue."
50. Due to the lack of available data, Moulton and Moses do not attempt to make specific quantity estimates of the potential bias that might result from quality change and new product introduction.
51. Moulton and Moses, "Addressing the Quality Change Issue."
52. Toward a More Accurate Measure of the Cost of Living.
53. In the national accounts, there is a category of other medical professionals for which real trends are calculated using the PPI and the CPI. Detailed data is not available to disaggregate this component and therefore it is not counted as using the CPI. Thus the estimate above should be viewed as a lower bound.
54. See Kenneth Stewart, "The experimental CPI."
55. For an overview of the 1998 CPI revision, see Greenlees and Mason, "Overview of the 1998 revision of the Consumer Price Index."
56. L. Slifman and C. Corrado ("Decomposition of Productivity and Costs," Occasional Staff Studies, oss no. 1, Federal Reserve Board, November 1996) present service industry measures of productivity, displaying negative productivity trends that seem inconsistent with anecdotal evidence of technological advances in service industries. Note that the study primarily discusses data for gross product originating for industries, which are not used to construct business sector GDP.
57. Expenditures on services constitute about 41 percent of final demand in the business sector, and approximately 54 percent of GDP, based on 1997 current dollars. Of the 14 percent of GDP that is constructed using input-based methods, 78 percent is expenditures on services; of the 11 percent of business sector output constructed with input-based methods, 50 percent is expenditures on services.
58. For a detailed discussion of the difficulties in measuring service output, see Mark K. Sherwood, "Difficulties in the Measurement of Service Output," Monthly Labor Review, March 1994, pp. 11-19.
59. For a brief discussion of BLS and BEA efforts to improve the national accounts data, see the article in this issue of the Review by Edwin R. Dean, pp. 24-34.