Methodology Paper 6: Personal Consumption Expenditures

Acknowledgments

This paper was prepared by Clinton P. McCully, Robert P. Parker, and Helen Stone Tice.Paula C. Young contributed to the section on benchmark estimates. Getty B. Mumford andTeresa A. Williams provided secretarial assistance.

The papers in this series on the methodology of the national income and product accountsare planned and coordinated by Helen Stone Tice. Allan H. Young, Robert P. Parker, and CarolS. Carson provided direction and guidance.

Comments about the paper are invited. Comments, as well as questions about the materialin the paper, should be directed to: National Income and Wealth Division, (BE-54), Bureau ofEconomic Analysis, U.S. Department of Commerce, Washington, DC 20230.

Suggested Citation

U.S. Department of Commerce. Bureau of Economic Analysis. Per-sonal Consumption Expenditures. Methodology Paper Series MP-6.Washington, DC: U.S. Government Printing Office, June 1990.

ii

Contents

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

1 Conceptual basis and framework . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.1 Recording of PCE in the NIPA’s: Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

1.1.1 Purchases of new goods and of services by individuals . . . . . . . . . . . . . . . . . . . . . . . 41.1.2 Purchases of goods and services by nonprofit institutions . . . . . . . . . . . . . . . . . . . 51.1.3 Used goods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51.1.4 Purchases of goods and services abroad by U.S. residents . . . . . . . . . . . . . . . . . . . 51.1.5 Imputed transactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5

1.2 Recording of PCE in the NIPA’s: Special features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61.2.1 Nonprofit institutions serving individuals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61.2.2 Imputed transactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8

Owner-occupied housing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8Rental value of buildings and equipment owned and used by

nonprofit institutions serving individuals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8Farm products consumed on farms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8Compensation in kind . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9Imputed charges for the services of financial intermediaries . . . . . . . . . . . . . 9Services furnished without payment by domestic securities dealers . . . . . 11Services furnished without payment by financial intermediaries except

life insurance carriers and private noninsured pension plans . . . . . . . . 11Expense of handling life insurance and private pension plans . . . . . . . . . . . 12

1.3 Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12Personal consumption expenditures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12

1.4 Classification conventions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

2 Presentation of the estimates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142.1 NIPA tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162.2 Schedule . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16

3 Overview of principal source data and estimating methods . . . . . . . . . . . . . . . . . . . . . . . . . . . 163.1 Current-dollar estimates. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163.2 Constant-dollar estimates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29

Current-dollar estimates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31

1 Annual estimates: Benchmark years . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 311.1 Commodity-flow estimates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31

1.1.1 Commodities purchased by persons . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 311.1.2 Domestic sales . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 311.1.3 Imports . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 321.1.4 Trade margins and related taxes and transportation costs . . . . . . . . . . . . . . . . 321.1.5 Exports . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 331.1.6 Change in trade inventories . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 331.1.7 Allocation of domestic supply among business, government, and persons . . 331.1.8 Intersector sales of used goods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33

1.2 Direct estimates. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 341.3 Sources of the benchmark estimates. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34

2 Annual estimates: Nonbenchmark years . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 342.1 Goods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41

iii

2.1.1 Goods extrapolated by the retail-control method. . . . . . . . . . . . . . . . . . . . . . . . . . . . 412.1.2 Other goods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41

2.2 Services . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48

3 Quarterly and monthly estimates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 483.1 Historical estimates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 483.2 Current estimates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55

4 Special procedures used for selected PCE categories . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 554.1 New autos . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55

4.1.1 Data sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 554.1.2 Unit sales . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 564.1.3 Average expenditure. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56

4.2 Net transactions in used autos . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 564.2.1 Total unit stock and components of change . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 564.2.2 Residual estimates of change in unit stock held by persons . . . . . . . . . . . . . . . . 574.2.3 Valuation in wholesale prices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 584.2.4 Quarterly and monthly estimates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58

4.3 Net transactions in used trucks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 584.4 Gasoline . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59

4.4.1 Quantities purchased . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59Autos . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59Trucks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59Airplanes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60Motorcycles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60Boats . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60

4.4.2 Average retail price . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 604.4.3 Quarterly and monthly estimates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60

4.5 Rental value of owner- and tenant-occupied nonfarm dwellings . . . . . . . . . . . . . . . . . 604.5.1 Number of permanent-site units . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 614.5.2 Rent on tenant-occupied permanent site dwellings . . . . . . . . . . . . . . . . . . . . . . . . 614.5.3 Space rent on owner-occupied permanent site dwellings . . . . . . . . . . . . . . . . . . . 614.5.4 Rent on owner- and tenant-occupied mobile homes . . . . . . . . . . . . . . . . . . . . . . . . . 614.5.5 Quarterly and monthly estimates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62

4.6 Brokerage charges and investment counseling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 624.6.1 Annual estimates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62

Securities commissions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62Commodities commissions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63Investment counseling fees . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63Services furnished without payment by domestic securities dealers . . . . . 63Sales charges on purchases of investment company securities . . . . . . . . . . . 64

4.6.2 Quarterly and monthly estimates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 644.7 Services furnished without payment by financial intermediaries except life

insurance carriers and private noninsured pension plans . . . . . . . . . . . . . . . . . . . . . . . . . 644.8 Expense of handling life insurance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65

4.8.1 Legal reserve companies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 654.8.2 Fraternal benefit societies and mutual savings banks . . . . . . . . . . . . . . . . . . . . . 654.8.3 Private noninsured pension plans . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66

4.9 Religious activities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66

Constant-Dollar estimates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67

1 Source data used for deflation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 671.1 Price indexes from other sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 671.2 Indexes prepared by BEA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67

1.2.1 Computers and peripheral equipment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 671.2.2 Food produced and consumed on farms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 671.2.3 Tolls . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 671.2.4 Airline transportation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 671.2.5 Foreign travel by U.S. residents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67

iv

1.2.6 Hospitals. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 721.2.7 Nursing homes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 721.2.8 Expense of handling life insurance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 721.2.9 Other . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72

2 Source data for quantity indicators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 722.1 Direct base-year valuation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 722.2 Quantity extrapolation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73

2.2.1 Rental value of farm housing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 732.2.2 Services furnished without payment by financial intermediaries except

life insurance carriers and private noninsured pension plans . . . . . . . . . . . . 732.2.3 Stock brokerage charges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 732.2.4 Other insurance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 732.2.5 Parimutuel net receipts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73

Sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75

Appendix—Selected NIPA Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83

Tables

1. Summary National Income and Product Accounts, 1988 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22. Content of PCE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43. Consolidation of Appropriation Accounts of Nonprofit Institutions and Individuals . . . . . 74. Imputed Transactions in PCE, 1988 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95. Production and Saving-Investment Accounts for Owner-Occupied Housing . . . . . . . . . . . . 116. Income Statement and Production Account for Depository Institutions . . . . . . . . . . . . . . . . . 137. Income Statement and Production Account for the Life Insurance and Pension Business of

Life Insurance Carriers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148. Location of PCE in the NIPA Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159. Summary of Source Data and Estimating Methods Used in Preparing Current-Dollar

Estimates of PCE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1810. Steps in the Commodity-Flow Method Used for PCE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3211. Source Data and Estimating Methods Used in Preparing Benchmark Estimates of

Current-Dollar PCE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3512. Source Data and Estimating Methods Used in Preparing Nonbenchmark Annual

Estimates of Current-Dollar PCE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4213. Source Data and Estimating Methods Used in Preparing Quarterly Estimates of

Current-Dollar PCE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4914. Retention Rates and Percent of Original Value by Age of Auto . . . . . . . . . . . . . . . . . . . . . . . . . 5715. Constant-Dollar PCE: Sources of Estimates. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68

v

Introduction

This paper presents the conceptual basis and frame-work of personal consumption expenditures (PCE) inthe national income and product accounts (NIPA’s), de-scribes the presentation of the estimates, and describesthe sources and methods used to prepare annual, quar-terly, and monthly estimates of PCE. This introductiondiscusses the conceptual basis and framework. It alsodefines the PCE measures that appear in the NIPAtables and indicates the tables in which they can befound. Subsequent sections describe in detail thesources and methods used to prepare the estimates ofPCE.

1 Conceptual Basis andFramework

As described in “An Introduction to National Eco-nomic Accounting” in the March 1985 SURVEY OF

CURRENT BUSINESS, the national economic accounts canbe viewed as aggregations of accounts belonging tothe individual transactors in the economy.1 The ba-sic approach used in national economic accounting isto group economic transactors into sectors and to setup uniform types of accounts that show the broad cat-egories of economic transactions in which the sectorsengage. Four sectors are commonly distinguished:Business, household, government, and foreign.2 Foreach sector, a set of three accounts is created: A produc-tion account, which records the production attributableto that sector; an appropriation account, which recordsthe sources of the sector’s income, its current out-lays, and its saving; and a saving-investment account,which records the sector’s gross saving, net increase inassets, and net increase in liabilities. Taken together,these sector accounts constitute a double-entry systemin which an outlay recorded in one account for one sec-tor is also recorded as a receipt in another account,either for the same sector or for another sector.

The NIPA’s are a combination of the sector accountsjust described, a combination designed to display thevalue and composition of national output and the dis-tribution of incomes generated in its production. Insummary form, the NIPA’s consist of five accounts:(1) The national income and product (NIP) account,

1. This article, by Allan H. Young and Helen Stone Tice, is reprinted asMethodology Paper Series MP-1 (Washington, DC: U.S. Government PrintingOffice, March 1985).

2. In MP-1, the term “households” is used; in this paper, the NIPA terms“persons” or “personal sector” are used.

which is a consolidation of the four sector productionaccounts and the business appropriation account; (2)the personal income and outlay account, which is anexpansion of the household appropriation account; (3)the government receipts and expenditures account,which is the government appropriation account; (4) theforeign transactions account, which is a consolidationof the foreign appropriation account and the foreignsaving-investment account; and (5) the gross savingand investment account, which is a consolidation ofthe saving-investment accounts of the business, house-hold, and government sectors. The five accounts of theNIPA’s, with entries for 1988, are shown in table 1.PCE appears in the NIP account, where it is by far thelargest component of final demand, and in the personalincome and outlay account, where it is the dominantoutlay.

1.1 Recording of PCE in the NIPA’s:Overview

PCE measures the goods and services purchased by thepersonal sector. The personal sector in the NIPA’s con-sists of persons resident in the United States. Personsare defined as individuals and the nonprofit institu-tions serving them. Persons resident in the UnitedStates are those who are physically located in theUnited States and who have resided, or expect to re-side, in this country for 1 year or more. Employees ofU.S. businesses sent abroad on assignments of 1 yearor less are included, as are U.S. Government civilianand military personnel stationed abroad, regardless ofthe duration of their assignments.

The content of PCE is highlighted in table 2. Thistable shows the kinds of transactions included. Mostof PCE consists of purchases of new goods and of serv-ices by individuals from business. In addition, PCEincludes purchases of new goods and of services bynonprofit institutions, net purchases of used goods byindividuals and nonprofit institutions, and purchasesabroad of goods and services by U.S. residents trav-eling or working in foreign countries. Finally, PCEincludes imputed purchases that keep PCE invariantto changes in the way that certain activities are carriedout—for example, whether housing is rented or ownedor whether employees are paid in cash or in kind.

1

4 Personal Consumption Expenditures

Table 2.—Content of PCE

Category of expenditure Comments

1. Purchases of new goods and of services by individuals frombusiness and government and purchases of the services of paidhousehold workers. 1

Purchases for business use are not included.Purchases from other individuals are not included.Purchases of dwellings are not included.Finance charges are not included.Payments for public goods, such as services provided by govern-

ment-owned schools and hospitals, are not included.Health and casualty insurance premiums in excess of benefits

paid are included.

2. Purchases of goods and services by nonprofit institutions frombusiness, individuals, and government.

Purchases of goods and services consist of the current operatingexpenses, including capital consumption. Purchases of struc-tures and equipment are not included.

3. Net purchases of used goods by individuals and nonprofit institu-tions from business and from government. 1

Transactions between persons are not reflected in PCE becausethey cancel in the aggregation of the personal sector.

4. Purchases of goods and services abroad by U.S. residents. These transactions are included in PCE for foreign travel andother, net, and in PCE. They are not included in the various de-tailed PCE components.

5. Purchases imputed to keep PCE invariant to whether:Housing and institutional structures and equipment are rented

or owned.Employees are paid in cash or in kind.Farm products are sold or consumed on farms.Saving, lending, and borrowing are direct or are intermediated.Financial service charges are explicit or implicit.

PCE components entirely imputed are the space rent of owner-oc-cupied housing, farm products consumed on farms, wages andsalaries paid in kind, services furnished without payment by fi-nancial intermediaries except life insurance carriers and privatenoninsured pension plans, and expense of handling life insur-ance.

Other imputed purchases are the imputed rental value of buildingsand equipment owned and used by nonprofit institutions (in-cluded in their current operating expenditures), the imputedvalue of employer-paid life insurance, the imputed value of em-ployer-paid health insurance, and the imputed value of the serv-ices furnished without payment by domestic securities dealers.

1. The source data underlying the estimates of many of the detailed PCEcomponents include expenditures in the United States by foreign visitorsand personal remittances in kind to foreigners. These transactions are ex-cluded in the aggregate from total PCE by means of negative entries in

PCE for foreign travel and other, net.

1.1.1 Purchases of new goods and of services byindividuals

The first category of table 2 constitutes most of PCE. Itconsists of purchases by individuals of new goods and ofservices from business and government and purchasesof the services of paid household workers. Examples ofgoods and services purchased from business are autos,major appliances, food, clothing, medical care, utilities,and rental housing.3 Purchases from government areincluded in PCE if the goods and services are similar tothose sold by private business. For these purchases, anoffsetting entry is made in the government purchasescomponent of final demand for the government sale—

3. PCE also includes the imputed services of owner-occupied housing; thisis taken up in the discussion of category 5 in table 2.

a negative purchase—or in the surplus of governmententerprises on the income side of the NIP account.

PCE does not include the business share of autosand other goods and services that are purchased for acombination of business and personal use.

PCE does not include purchases from individuals,other than the services of paid household workers.

PCE does not include purchases of dwellings—thatis, structures and mobile homes—nor does it in-clude landlords’ purchases of durable goods; theseare all included instead in private residential fixedinvestment.

PCE does not include finance charges. These areincluded in interest paid by consumers to business, acomponent of personal outlays.

PCE does not include payments to government forgoods and services that either are administrative or

Introduction 5

regulatory in nature or are provided privately pri-marily by nonprofit institutions; these payments areclassified as personal nontax payments, a componentof personal taxes and nontaxes. Thus, consumer postalcharges and payments to State universities for roomand board are included in PCE, but consumer pay-ments for medical care to county hospitals and fortuition to State universities are included in personalnontax payments.4

PCE includes outlays for insurance. Purchases ofhealth insurance and of casualty insurance from stockand mutual companies are measured in PCE by pre-miums paid to insurance carriers less the benefitsand dividends paid by them.5 This definition in-cludes in PCE the service of being insured. It avoidsdouble-counting the reimbursed expenditures, whichare accounted for separately as automobile repair, pro-fessional services, and the like. Purchases of lifeinsurance are measured in PCE by the operating ex-penses of life insurance carriers, as discussed in thenext section.

1.1.2 Purchases of goods and services bynonprofit institutions

The second category in table 2 is purchases of goodsand services by nonprofit institutions. Nonprofit in-stitutions serving individuals include trade unionsand professional associations, clubs and fraternal or-ganizations, educational institutions, foundations foreducation and research, and religious and welfareorganizations. The expenditures of nonprofit in-stitutions are not distributed among the individualcommodities in PCE. Instead, certain components ofPCE—primarily medical care, private education andresearch, and religious and welfare activities—includea measure of the services of nonprofit institutionsserving individuals equal to their current operatingexpenses.6 For example, airline tickets purchased bya private university are included in PCE for privateeducation and research, not in PCE for airline trans-portation. The accounting for nonprofit institutionsserving individuals is discussed in the next section.

1.1.3 Used goods

The third category in table 2 is used goods. PCEincludes the net purchases—purchases less sales—ofused goods from other sectors; transactions betweenpersons are not included because they cancel in the ag-gregation of personal sector transactions in used goods.

4. For a more complete discussion, see U.S. Department of Commerce,Bureau of Economic Analysis, Government Transactions, Methodology PaperSeries MP-5 (Washington, DC: U.S. Government Printing Office, November1988), p. 5.

5. Expenditures for health insurance provided by nonprofit organizations,such as Blue Cross-Blue Shield, are measured in PCE by the current operatingexpenses of the carriers. This treatment of nonprofit institutions is explainedin the next section.

6. These expenses are net of receipts from meals, rooms, andentertainment, which are accounted for separately.

Net purchases include costs incurred in the transac-tions, such as dealers’ margins. These entries definePCE so that it correctly measures total purchases byU.S. persons, whenever the goods were produced.

Including transactions in used goods makes PCE amore useful construct for the analysis of consumer be-havior than it would be were it limited to the newlyproduced goods and services that constitute GNP. Netpurchases of used goods also are included in other finaldemand components of the NIP account—investment,net exports, and government purchases. Except forthe dealers’ margins on the sale of used goods, theseentries cancel in the derivation of GNP as the sum ofthe final demand components so that GNP consists ofonly newly produced goods and services.

1.1.4 Purchases of goods and services abroad byU.S. residents

The fourth category in table 2 is purchases of goodsand services abroad by U.S. residents. When travel-ing abroad, U.S. residents pay foreigners for passengerfares and travel services and also purchase goodsand other services. Conceptually, these outlays arecomponents of PCE, but they are not part of U.S.production.

The entries for foreign travel and expendituresabroad by U.S. residents define PCE so that it cor-rectly measures the total purchases of U.S. persons,wherever the goods and services were produced. Thesepurchases are included in total in the PCE cate-gory, “foreign travel and other, net.”7 They are notdistributed among the individual commodities in PCE.

As in the case of used goods, including purchasesabroad by U.S. residents makes PCE a more usefulconstruct for the analysis of consumer behavior. Off-setting entries are made in imports in the NIP accountin order to cancel these PCE entries in the derivationof GNP as the sum of the final demand components.

1.1.5 Imputed transactions

The fifth category in table 2 is imputations. GNP con-sists primarily of those goods and services reflected inthe transactions of the market economy. The limitedexceptions to the market transaction principle in theNIPA’s are imputations that serve the purpose of keep-ing GNP, PCE, and other NIPA aggregates invariantto whether: (1) Housing and institutional structures

7. This PCE category also contains two negative entries that remove con-sumption by foreigners from the PCE estimates. (1) When traveling in theUnited States, foreigners make expenditures indistinguishable from thosemade by U.S. residents in the source data underlying many of the estimatesof detailed PCE components. These outlays are removed in the aggregatefrom PCE and are shown instead as exports in the NIP account. (2) Expendi-tures on goods subsequently remitted to foreignerswithout quid pro quo (suchas CARE packages) are removed in the aggregate from PCE and are showninstead as exports in the NIP account. These goods are treated as transferpayments; they are shown as a separate component of personal outlays inthe personal income and outlay account and as a component of payments toforeigners in the foreign transactions account.


and equipment are rented or owned; (2) employees arepaid in cash or in kind; (3) farm products are soldor consumed on the farm; (4) saving, lending, andborrowing are direct or are intermediated; and (5) in-termediated financial transactions involve an explicitor implicit service charge. The accounting for imputedtransactions is discussed in the next section.

The customary presentations of PCE are not alongthe lines of table 2. Instead commodities from thecategories in table 2 are grouped in two ways—by typeof product and by type of expenditure. A description ofthese groupings and the numbers of the NIPA tablesin which PCE entries may be found are provided in thesection called “Presentation of the Estimates.”

1.2 Recording of PCE in the NIPA’s:Special features

1.2.1 Nonprofit institutions serving individuals

Nonprofit institutions serving individuals receiveproperty income (that is, interest, rent, and dividends),receive payments for goods and services that they pro-vide, pay compensation of employees and incur otheroperating expenses, make and receive transfer pay-ments (that is, donations and grants), and purchaseand rent buildings and equipment. These transactionsare consolidated with those of individuals to form theNIPA personal income and outlay account.

The consolidation is illustrated in the three T-accounts in table 3, which includes hypotheticalnumbers as an aid to the reader. The first two ac-counts in table 3 are the appropriation accounts ofnonprofit institutions and of individuals; they both in-clude transactions between individuals and nonprofitinstitutions as well as those with other sectors.

• The appropriation account of nonprofit institu-tions shows as uses current operating expenses,transfer payments to individuals, and saving. Thecurrent operating expenses are wages, salaries,and other labor income paid; purchases frombusiness; mortgage interest paid; and capital con-sumption allowances.8 It shows as sources salesof goods and services to individuals, transfer pay-ments received, and property income received.Sales to individuals include both receipts for pri-mary services (such as education) and receiptsfor other goods and services (such as universitydining hall receipts). Transfer payments includeboth donations from individuals and business andpayments from government in support of research.

• The appropriation account of individuals showsas uses purchases from business and from non-profit institutions; wages, salaries, and other labor

8. The term “capital consumption allowances” is used to mean “capitalconsumption allowances with capital consumption adjustment,” the NIPAterminology for depreciation plus accidental damages valued at currentreplacement cost.

income paid; interest paid; rent paid; transfer pay-ments to nonprofit institutions; and saving. Itshows as sources wages, salaries, and other la-bor income received; property income received;and transfer payments received from nonprofitinstitutions, from business, and from government.

The third account is the personal income and out-lay account; entries in this account are obtained byconsolidating the first two accounts. Sales of serv-ices by nonprofit institutions to individuals cancel thepurchases of these services by individuals. Trans-fer payments to persons consists only of paymentsby business and government; transfer payments thatoccur between individuals and the nonprofit institu-tions serving them cancel in the consolidation. Wages,salaries, and other labor income is included both inwages, salaries, and other labor income and in PCE,however. These purchases of factor services by indi-viduals and nonprofit institutions do not cancel in theconsolidation because they measure both the value ofpersonal sector production purchased by the personalsector and the factor income received by persons frompersonal sector production.

PCE is shown in two parts: Current expenditures ofnonprofit institutions and expenditures by individuals.Each part consists of wages, salaries, and other laborincome paid; rent paid; and other purchases. In eachpart, rent paid includes imputed rental charges, whichare discussed in the next section.

The expenditures of nonprofit institutions are notdistributed among the individual commodities in PCE.Instead, certain components of PCE—primarily medi-cal care, private education and research, and religiousand welfare activities—include the services of non-profit institutions serving individuals. Thus, PCE foreach of these components shows the total expendi-tures of persons, whether the services are provided byfor-profit or by nonprofit organizations.

The expenditures of individuals for wages, salariesand other labor income paid represents payments tohousehold workers; these payments are shown as aseparate category of PCE. Rent paid is also shownas a separate category of PCE. The remainder oftheir expenditures are distributed by commodity; theseexpenditures include purchases of clothing, meals,rooms, and entertainment from nonprofit institutions.

In the estimates of PCE, the services of nonprofit in-stitutions are measured net of their receipts from salesof clothing, meals, rooms, and entertainment to indi-viduals, which are accounted for separately in otherPCE categories of expenditures by individuals. Forexample, PCE for private education and research in-cludes the educational services provided by nonprofituniversities, while PCE for housing includes paymentsby individuals to nonprofit universities for dormitorycharges.


1.2.2 Imputed transactions

The imputed transactions included in the NIPA’s area limited set of exceptions to the principle that GNPconsists of market transactions in goods and services.In order to keep PCE invariant to how certain activi-ties are carried out, imputations are made to place amarket value on certain transactions that do not occurin the market economy, and that, therefore, are notobservable in its records. In this process, some markettransactions are restructured to provide a represen-tation of the activity that is more appropriate for theNIPA’s. Both a measure of production and the incomesassociated with that production are imputed. The im-putations described here are those for which purchasesof the imputed production are included in PCE.

The effect of these imputed transactions on the NIPand personal income and outlay accounts is shown intable 4, which shows these two accounts with entriesfor 1988. The entries for the imputed transactionsare shown in boldface both in table 4 and in tables 5,6, and 7, which illustrate three of the more compleximputations.

Owner-occupied housing.—The imputation for owner-occupied housing creates a business that purchaseshousing and subsequently sells housing services topersons. The business has a production account thatshows the receipts from these sales and the chargesagainst these receipts and a saving-investment ac-count that shows investment in housing and thecharges for capital consumption on the housing stock.This treatment of housing is illustrated in table 5,which shows the production and saving-investmentaccounts for owner-occupied housing.

In the production account, purchases of housingservices by persons are imputed; these services aremeasured as space rent, which is estimated by anal-ogy with rental property.9 Gross product equals spacerent less consumption of purchased materials andservices—that is, expenditures that would be treatedas expenses by owners of rental property.

Charges against gross product equals the sum ofnet interest paid (mortgage interest), indirect busi-ness taxes (property taxes), capital consumption al-lowances, subsidies (a negative entry), and the owner-occupants’ income. The last, like corporate profits, iscalculated as a residual. It is included in rental incomeof persons, in the case of nonfarm property, and in farmproprietors’ income, in the case of farm property.

When this production account is consolidated withthe rest of the NIP account, the imputed gross prod-uct originating in owner-occupied housing is includedin GNP. As shown in table 4, the space rent is en-tered in PCE on the right side of the account, andthe consumption of purchased materials and servicescancels in the consolidation. The incomes, taxes, subsi-dies, and capital consumption allowances are entered

9. Space rent excludes charges for utilities, major appliances, furniture,and furnishings.

in the same categories of charges on the left side ofthe NIP account. Similarly, in the personal incomeand outlay account, space rent is included in PCE andin personal outlays; proprietors’ income, rental incomeof persons, and net interest are included in personalincome. In addition, purchases of newly constructedhousing, shown in the saving-investment account intable 5, are business fixed investment and are includedin gross private domestic investment on the right sideof the NIP account.

In the absence of the imputation, the measure ofhousing services would not be invariant to changesin the ownership of the housing stock. The expensesof homeownership, such as expenditures for materialsand supplies used in maintenance, would enter finaldemand (as PCE) rather than intermediate purchases;and purchases of owner-occupied housing would bepart of PCE rather than of investment outlays. Prop-erty taxes and mortgage interest would be personalrather than business payments and therefore not in-cluded in charges against GNP; rental income onowner-occupied nonfarm housing, the owner-occupiedrental component of farm proprietors’ income, and cap-ital consumption allowances for owner-occupied homeswould not enter the national economic accounts at all.

Rental value of buildings and equipment owned andused by nonprofit institutions serving individuals.—Because nonprofit institutions are considered per-sons in the NIPA’s, an imputed rental charge for thebuildings and equipment that they own and use isincluded in PCE as well as in the appropriate cate-gories of factor and nonfactor charges against GNP.Purchases of buildings and equipment by nonprofitinstitutions for their own use are included in invest-ment outlays rather than in PCE. The rental chargecovers net interest (mortgage interest), indirect busi-ness taxes (property taxes), and capital consumptionallowances. Unlike that for owner-occupied housing,the imputed rental charge does not include any im-puted income. (The treatment of nonprofit institutionswas taken up earlier.)

In the absence of this imputation, this componentof current expenditures of nonprofit institutions wouldconsist only of net purchases of buildings and equip-ment. PCE would not be invariant to changes in theownership of such assets. As with owner-occupiedhousing, without the imputation, mortgage interestand property taxes would not enter GNP, because theywould be personal rather than business outlays.

Farm products consumed on farms.—PCE includesestimates of the value of the food and fuel that are bothproduced and consumed on farms, and farm propri-etors’ income includes the margin on that production.Consumption of purchased materials and services can-cels in the consolidation of the aggregate businesssector.

In the absence of the imputation, these commoditieswould be treated as if consumed during production,

Introduction 9

Table 4.—Imputed Transactions in PCE, 1988

[Billions of dollars]

National Income and Product (NIP) Account

Uses Sources

Compensation of employees ............................................ 2,907.6 Personal consumption expenditures ............................... 3,235.1Compensation paid in kind 1 ...................................... 153.5 Space rent ................................................................... 348.0Other ............................................................................. 2,754.1 Nonprofit buildings and equipment 4 ....................... 24.7

Proprietors’ income 2 ......................................................... 327.8 Farm products consumed on farms ........................ .8Owner-occupied farm housing .................................. 3.4 Compensation paid in kind 5 ..................................... 143.5Farm products consumed on farms ......................... .4 Services furnished without payment by domesticOther ............................................................................. 324.0 securities dealers ................................................... 3.3

Rental income of persons 3 .............................................. 15.7 Services furnished without payment by financialOwner-occupied nonfarm housing ........................... –15.0 intermediaries except life insurance carriersOther ............................................................................. 30.7 and private noninsured pension plans ................ 91.1

Corporate profits 2 ............................................................. 328.6 Expense of handling life insurance 6 ....................... 45.4Services furnished without payment by domestic Other ............................................................................ 2,578.3

securities dealers .................................................... 3.3 Gross private domestic investment ................................. 750.3Other ............................................................................. 325.3 Residential ................................................................... 232.4

Net interest ....................................................................... 392.9 Owner-occupied housing ...................................... 193.6Owner-occupied housing ........................................... 174.6 Other ......................................................................... 38.8Nonprofit buildings and equipment .......................... 6.2 Nonresidential .............................................................. 517.8Imputed interest paid to persons .............................. 233.1 Nonprofit buildings and equipment ..................... 20.0

Depository institutions ........................................... 91.1 Other ......................................................................... 497.9Life insurance carriers and private noninsured Net exports of goods and services 7 ............................... –73.7

pension plans ....................................................... 141.9 Government purchases of goods and services 7 ............ 968.9Other ............................................................................. –21.0

Business transfer payments ............................................. 29.0Indirect business tax and nontax liability ......................... 393.5

Owner-occupied housing ........................................... 57.5Nonprofit buildings and equipment .......................... 2.7Other ............................................................................. 333.3

Less: Subsidies less current surplus of governmententerprises .................................................................... 18.5Owner-occupied housing ........................................... .2Other ............................................................................. 18.3

Capital consumption allowances 3 .................................... 513.6Owner-occupied housing ........................................... 79.6Nonprofit buildings and equipment .......................... 15.9Other ............................................................................. 418.1

Statistical discrepancy ...................................................... –9.6

Charges against gross national product .......................... 4,880.6 Gross national product .................................................... 4,880.6

See footnotes at end of table.

and GNP and farm proprietors’ income would not in-clude their value. PCE would not include the valueof this production and thus would not be invariant towhether farmers choose to consume their own productsor to sell them and use the revenue to purchase farmproducts on the market.

Compensation in kind.—Both PCE and compensationof employees include estimates of compensation paidin kind. Three categories of wages and salaries in kindare identified in PCE: The value of food furnished tomilitary personnel and to other employees, the valueof standard clothing issued to military personnel, andthe value of employees’ lodging. In addition, twofringe benefits, employer-paid health insurance pre-miums and employer-paid life insurance premiums

are included in PCE and in the other labor incomecomponent of compensation.

In the absence of the imputation for compensationpaid in kind, these expenditures would be employers’intermediate purchases rather than a factor cost—compensation of employees—and thus not included inGNP. PCE would not include these expenditures, norwould personal income include their value.

Imputed charges for the services of financial in-termediaries.—The measures of PCE for the servicesof financial intermediaries include imputations forservice charges that the intermediaries do not collectexplicitly. These institutions are viewed as making ex-plicit charges for some of the services that they renderand making implicit charges for other services by pay-

Introduction 11

Table 5.—Production and Saving-Investment Accounts for Owner-Occupied Housing

Production Account

Uses Sources

Farm proprietors’ income 1 Space rent 3

Rental income of persons 2

Net interest Less: Consumption of purchased materials and servicesIndirect business tax and nontax liabilityCapital consumption allowances 2

Less: Subsidies

Charges against gross product Gross product

Saving-Investment Account

Uses Sources

Net purchases of owner-occupied housing units SavingNet acquisition of financial assets Capital consumption allowances 2

Less: Net increase in liabilities

Gross investment Gross saving

1. With inventory valuation and capital consumption adjustments.2. With capital consumption adjustment.3. Excludes charges for utilities, major appliances, furniture, and furnish-

ings.

ing depositors or policy holders less interest, dividends,and other property income than they earn on thesefunds. The NIPA’s impute an estimate of this implicitcharge and include it in PCE. The NIPA’s also imputean income payment and include it in net interest paid.

The financial services imputations treat a financialinstitution as an intermediary, which channels sav-ing into productive uses by giving savers the kind ofassets they want (for example, bank deposits or lifeinsurance policies) and giving business investors thekind of liabilities they want (for example, loans tocarry inventories). Although the property income paidby business borrowers can be viewed as income at-tributable to the depositors or insurance beneficiaries,the intermediary typically does not pass on all thisincome to those who hold its liabilities. Three im-puted financial service charges are shown in table 4:(1) Services furnished without payment by domesticsecurities dealers; (2) services furnished without pay-ment by financial intermediaries except life insurancecarriers and private noninsured pension plans—thatis, by depository institutions; and (3) expense of han-

dling life insurance. These are discussed more fully inthe following paragraphs.

Services furnished without payment by domesticsecurities dealers.—The PCE category brokeragecharges and investment counseling includes an im-plicit charge equal in amount to the price spread onthe transactions carried out by securities dealers withpersons. Dealers who make markets in securities donot charge commissions; instead, as income they re-tain the revenue resulting from acquiring securities ata price lower than the price at which the securitiessubsequently are sold to their customers.

Omitting this imputation, and the parallel onefor transactions on behalf of business, would reducebusiness income because the trading gains would beclassified as capital gains income, which is omitted inthe NIPA’s.

Services furnished without payment by finan-cial intermediaries except life insurance carriersand private noninsured pension plans.—Depositoryinstitutions—that is, commercial banks, mutual sav-ings banks, savings and loan associations, credit


unions, and regulated investment companies (includ-ing money market funds)—are considered to pay animputed interest that the depositor uses to purchasethose checking, bookkeeping, and investment servicesfor which he does not pay an explicit service charge.

The nature of the interest imputation calcula-tions for depository institutions is illustrated withT-accounts in table 6, which shows the income state-ment and the production account for depository institu-tions.10 In this example, net property income receivedby the institution consists entirely of monetary inter-est. The first panel of table 6 shows the T-account formof the income statement; the second, the income state-ment converted to a production account that includesthe imputed transactions.

For depository institutions, the imputed servicecharge—or the equivalent imputed interest—consistsof monetary interest received less monetary interestpaid; it is entered on both sides of the productionaccount. The sum of monetary and imputed servicecharges is equivalent to the operating expenses—thatis, the charges against gross product (including profits)plus consumption of purchased materials and services.When this production account is consolidated with therest of the NIP account, imputed interest is entered asan element of both PCE and net interest paid.11

In the NIP account shown in table 4, imputed inter-est paid to persons is included both in net interest paidand in PCE, as services furnished without payment byfinancial intermediaries except life insurance carriersand private noninsured pension plans. When theseentries are carried over to the personal income andoutlay account, personal interest income is higher bythe amount of the imputation, but personal saving isunaffected because PCE is increased by an equivalentservice charge.

In the absence of the imputation, PCE would includeonly service charges actually paid by persons, andpersonal income would include only property incomedirectly paid to persons.12

Expense of handling life insurance and private pen-sion plans.—Life insurance carriers provide servicesthat combine elements of both insurance and saving;private pension plans provide services to savers. Theseinstitutions earn property income on insurance andpension reserves, which have been contributed directlyby—or for the benefit of—policy holders or pensionbeneficiaries and which are to be paid out to the benefi-ciaries as annuity or lump-sum distributions of incomein the future.

10. This example covers stock companies only and ignores the treatment ofmutual institutions.

11. Depository institutions also pay imputed interest to government, for-eigners, and business. Government purchases and exports include servicesfurnished without payment by depository institutions. Imputed interest paidto business and the imputed service chargespaid by business cancel in the con-solidation of the production accounts of financial and nonfinancial business.

12. In addition, in the absence of the imputation, gross product originatingin depository institutions would be negative were property income receiptssubstantially in excess of property income payments.

In the NIPA’s, life insurance premiums and bene-fits and pension plan contributions and benefits aredisregarded; instead, the institutions are regarded ascharging policyholders a fee equal to the institutions’operating expenses for the package of services pro-vided. These imputed fees, which include profits inthe case of stock companies, appear as “expense ofhandling life insurance” in PCE.

The treatment of this imputation is shown in table 7,which consists of the income statement and productionT-accounts for the life insurance and pension businessof stock life insurance carriers. Because the invest-ment vehicles of these institutions are more diversethan those of depository institutions, table 7 includesthe detail on property income omitted in table 6. Inthis example, net imputed interest paid is equal to netmonetary interest and dividends received. When thisproduction account is consolidated with the rest of theNIP account, the imputed fee for the costs of handlinglife insurance, which equals the operating expensesof life insurance companies, is added to PCE; the fac-tor and nonfactor charges of which it is composed areadded to the income side of the account.

The net effect of these transactions is to consolidatethe saving of life insurance carriers and pension planswith that of households and institutions. Underwrit-ing income (premiums less benefits) is treated as atransfer payment within the personal sector, and theproperty income and operating expenses of the insti-tutions are treated as personal income and outlays.The gross product originating consists of the expenseof handling life insurance less purchases of materialsand business services.

Although personal saving is unaffected by the im-puted interest paid by depository institutions, it isaffected by the imputed interest paid by life insur-ance carriers. That paid by life insurance carriers andprivate noninsured pension plans, which is added topersonal income, exceeds the imputed fee, which isadded to PCE, by an amount equal to the excess ofthe net additions to policy reserves over underwritingincome (premiums less benefits). In effect, the imputa-tion performs a timing change, because it records theproperty income that has been withheld to the accountof policy holders and beneficiaries as if it were actuallydisbursed to them in the current period.

In the absence of the insurance imputation, PCEfor life insurance would be measured by underwrit-ing income. Investment returns and increases in lifeinsurance and pension reserves would be included inbusiness income and saving rather than in personalincome and saving.

1.3 Definitions

Personal consumption expenditures.—Goods andservices purchased by individuals; the operating ex-penses of nonprofit institutions serving individuals;and the value of food, fuel, clothing, rent of dwellings,

Introduction 13

Table 6.—Income Statement and Production Account for Depository Institutions

Income Account

Uses Sources

Compensation of employees Service charges (monetary)Consumption of purchased materials and services Monetary interest receivedMonetary interest paidCapital consumption allowancesProfits

Outlays and profits Receipts

Production Account

Uses Sources

Compensation of employees Total service chargesNet interest paid Monetary

Monetary interest paid Services furnished without payment by financialImputed interest paid intermediaries except life insurance carriers and

Monetary interest received private noninsured pension plansLess: Monetary interest paid Monetary interest received

Less: Monetary interest received Less: Monetary interest paidCapital consumption allowancesProfits Less: Consumption of purchased materials and services

Charges against gross product Gross product

Note.—Monetary interest paid and monetary interest received are the onlytypes of ‘‘property income’’ included in the imputed interest calculation inthis example. This example covers stock institutions only.

and financial services received in kind by individu-als. Net purchases of used goods are also included.Purchases of residential structures by individuals andof buildings and equipment used by nonprofit institu-tions serving individuals are classified as gross privatedomestic investment.

1.4 Classification conventions

The following conventions are used to classify eachcommodity as a good, either durable or nondurable, oras a service.

In general, goods are commodities that can be stored,or inventoried. Durable goods have an average life ofat least 3 years; all other goods are considered non-durables. Services are commodities that cannot bestored and that are consumed at the place and time ofpurchase.

If commodities have both a good and service com-ponent, the classification generally is based on therelative importance of the two components.

• The following are the major commodities of thistype that are classified as goods: Restaurant

meals, expenditures abroad by U.S. residents,and replacement parts whose installation cost isminimal.

• The following are the major commodities of thistype that are classified as services: Airline trans-portation and hospital charges, which include food;natural gas and electricity; current operating ex-penses of nonprofit institutions; foreign travel byU.S. residents; expenditures in the United Statesby foreigners; and repair services, which includethe cost of parts, except for replacement partswhose installation cost is minimal.

Other important conventions in the classificationsystem include the following.

• All clothing and shoes are classified as nondurablegoods, regardless of their average life.

• Margins on used goods are included in goods.

• Repair services for which separate data are notavailable are included in the goods category of thecommodity being repaired.


Table 7.—Income Statement and Production Account for the Life Insurance and Pension Business ofLife Insurance Carriers

Income Statement

Uses Sources

Death and annuity benefits paid Premiums and annuity considerationsNet additions to reserves Property income receivedCompensation of employees Monetary interest receivedConsumption of purchased materials and services Dividends receivedMonetary interest paidCapital consumption allowancesProfits

Outlays and profits Receipts

Production Account

Uses Sources

Compensation of employees Expense of handling life insuranceNet interest paid Premiums and annuity considerations

Monetary interest paid Property income receivedImputed interest paid Less: Net additions to reserves

Property income received Less: Death and annuity benefitsMonetary interest received Less: Monetary interest paid to policyholdersDividends received Less: Consumption of purchased services

Less: Monetary interest paid to policyholdersLess: Monetary interest received

Capital consumption allowancesProfitsLess: Dividends received

Charges against gross product Gross product originating

Note.—This example covers only the life insurance and pension businessof stock life insurance carriers. The real estate and accident and health in-surance business of these carriers is excluded.

2 Presentation of the Estimates

In the NIPA tables, PCE is presented in one of twogroupings of commodities—type of product or type ofexpenditure.

• PCE by type of product, the classification mostfrequently used, is based on the 3-way classifica-tion of commodities just mentioned. It consists ofthree broad categories, with additional detail un-der each: Durable goods, nondurable goods, andservices. The most detailed presentation of PCEby type of product is shown annually in NIPA table2.5, which is reproduced in appendix A. Less detailis shown quarterly and monthly.

• PCE by type of expenditure is a functional classifi-cation. Each category may include both goods and

services. There are 12 broad functional categories,with additional detail under each: Food and to-bacco; clothing, accessories, and jewelry; personalcare; housing; household operation; medical care;personal business; transportation; recreation; pri-vate education and research; religious and welfareactivities; and foreign travel and other, net. PCEby type of expenditure is presented only on an an-nual basis in NIPA table 2.4, which is reproducedin appendix A.

2.1 NIPA tables

The estimates of PCE are published in the NIPA tables,which appear in the SURVEY (and reference volumescited therein). Table 8 indicates the location, by NIPAtable number, of the various annual, quarterly, and


monthly estimates of PCE in current dollars, and,where applicable, of estimates in constant dollars andof the associated price indexes.13 Annual estimatescover 1929 to the present; quarterly estimates in cur-rent dollars generally cover the beginning of 1946 tothe present; corresponding estimates in constant dol-lars generally cover the first quarter of 1947 to thepresent; and monthly estimates cover January 1959 tothe present.

Most presentations of the PCE estimates in the NIPAtables are for the three major types of product detail:Durable goods, nondurable goods, and services. An-nual and seasonally adjusted quarterly estimates forthese types in current dollars and in constant dollarsare shown in NIPA tables 1.1 and 1.2, respectively.The associated fixed-weighted price indexes are shownin table 7.1; implicit price deflators in NIPA table 7.4;and chain price indexes, shown only as the percent-age change from preceding period, in NIPA table 8.1.Percent changes from preceding period in annual andquarterly seasonally adjusted current and constantdollars, fixed-weighted price indexes, and implicit pricedeflators also are shown in NIPA table 8.1. Quarterlycurrent-dollar totals, not seasonally adjusted, andseasonally adjusted monthly current-dollar estimatesare shown in NIPA tables 9.1 and 2.8, respectively.Monthly seasonally adjusted constant-dollar estimatesand the associated implicit price deflators are shown inNIPA tables 2.9 and 7.11, respectively. (As indicatedin table 8, the monthly current- and constant-dollarestimates also appear in the “Current Business Statis-tics” section of the SURVEY.) Quarterly constant-dollartotals and monthly totals, not seasonally adjusted, arenot prepared.

PCE estimates at the more detailed type of prod-uct level also are shown in the NIPA tables. Annualand seasonally adjusted quarterly estimates of PCEin current dollars and in constant dollars are shownin NIPA tables 2.2 and 2.3, respectively. The associ-ated fixed-weighted price indexes are shown in table7.9. Quarterly current-dollar totals, not seasonallyadjusted, are shown in NIPA table 9.2.

The most detailed estimates of PCE are shown annu-ally in current and in constant dollars in NIPA tables2.4 and 2.5, respectively. In NIPA table 2.4, commodi-ties are grouped by type of expenditure, a functionalclassification. In NIPA table 2.5, the commoditiesare grouped by major type—durable goods, nondurablegoods, and services. Annual fixed-weighted price in-dexes corresponding to the detail in NIPA table 2.5 areshown in NIPA table 7.10. Annual and seasonally ad-justed quarterly fixed-weighted price indexes for threespecial expenditure categories—food, energy, and allother PCE—are shown in NIPA table 7.1.

Annual and quarterly estimates of the PCE compo-nents of auto and truck output are in NIPA tables 1.17

13. Table 8 does not include references to NIPA table 2.1, which shows PCEas part of personal outlays and of the disposition of personal income.

and 1.19 in current dollars and in NIPA tables 1.18and 1.20 in constant dollars.

Annual estimates of the PCE components that alsoare part of farm and of housing sector output are inNIPA tables 1.21 and 1.23 in current dollars and inNIPA tables 1.22 and 1.24 in constant dollars.

Annual per capita PCE estimates in current and inconstant dollars are shown in NIPA table 8.2.

Annual estimates of the PCE components classifiedas imputations are shown in NIPA table 8.9.

2.2 Schedule

The initial monthly seasonally adjusted estimates ofPCE are prepared about 4 weeks after the end ofthe month, revised in each of the following monthsuntil the preparation of the final estimate of the corre-sponding quarterly PCE estimate. Advance quarterlyseasonally adjusted estimates of PCE are prepared inthe first month after the end of the quarter, revised amonth later (preliminary estimate), and revised againthe following month (final estimate). No further revi-sions are made in the quarterly or monthly estimatesuntil the annual revisions, which usually occur eachJuly and cover the 3 most recent years. Following thethird annual revision, no further revisions are made inthe estimates until the comprehensive revisions (oftenreferred to as “benchmark revisions”), which usuallyoccur every 5 years.

3 Overview of Principal SourceData and Estimating Methods

3.1 Current-dollar estimates

The current-dollar estimates of PCE are based on sta-tistical reports, primarily from the Census Bureau, butalso from other government agencies; on governmentadministrative and regulatory agency reports; and onreports from private organizations, hereafter denotedtrade sources.

• The Census Bureau statistical reports cover thefollowing: Sales, inventories, and cost of pur-chased goods for manufacturing and trade; serviceindustry receipts and expenses; and residentialrental payments.

• The statistical reports of other government agen-cies cover the following: Cash receipts by farmersfor agricultural products, sales of electricity andfuel oil, international trade in services, receiptsand expenses for higher education, and prices paidby consumers.

• Government administrative and regulatory agencyreports cover the following: Federal and Stateand local government revenues and purchases,merchandise trade, sales of gasoline, wages and

Introduction 17

salaries, revenues from transportation services,brokerage commissions, and bank service charges.

• Reports from trade sources cover the following:Sales to persons of motor vehicles, tires, drugs,natural gas, gasoline and oil, water and sanitaryservices, telephone and telegraph service, tran-sit service, brokerage and investment counseling,broadwoven fabric, yarn, and knit fabric; expensesfor life insurance; premiums and benefits for non-life insurance; receipts and expenses for hospitals;and expenses of religion.

Most of these source data are “value data,” that is,they embody both the quantity and the price dimen-sions that are required for current-dollar estimates.Other source data provide information on physicalquantities and related prices, and the price andquantity data are combined to obtain value estimates.

The source data used for PCE are complete onlyfor “benchmark” years, that is, years in which BEA’sbenchmark input-output accounts are used to estab-lish the level of PCE and of its components duringa comprehensive revision. In nonbenchmark years,as well as for quarterly and monthly estimates, mostof the estimates are prepared by interpolation andextrapolation, using indicator series.

There are significant differences among the method-ologies for the current-dollar benchmark annual, non-benchmark annual, and quarterly and monthly PCEestimates. These differences are shown in table 9,which provides an overview of the source data and esti-mating methods used. Table 9 also identifies the PCEcomponents for which there have been major changesin source data, estimating methods, or both during theperiod covered by the estimates.

For benchmark years, the current-dollar PCE es-timates are based on the final demand componentsof BEA’s benchmark input-output (I-O) accounts, ad-justed to reflect the definitional, classificational, andstatistical changes incorporated into the NIPA’s sincecompletion of the I-O accounts. These accounts pro-vide information on the flow of commodities—goods,services, and structures—from each producing indus-try to other industries and to final users—that is,persons, governments, foreigners, and business on cap-ital account. These accounts are prepared for yearsin which the Census Bureau conducts the quinquen-nial economic censuses, which provide extensive dataon agriculture, transportation, manufactures, whole-sale trade, retail trade, and service, construction, andmineral industries; and the quinquennial censuses ofgovernments, which provides extensive data on Stateand local governments.

In the benchmark I-O accounts, the estimatingmethods used for purchases by persons depend on thetype of source data available. For a relatively smallnumber of commodities, purchases by persons are es-timated directly. For most commodities, purchasesby persons are estimated using methodologies thatare variations of the commodity-flow method, which

links domestic commodity sales, domestic commoditysupply, and commodity purchases by industries andby final users. These detailed I-O commodity esti-mates are grouped into the PCE categories used for theNIPA’s.

At present, the most recent benchmark I-O accountsreflected in the NIPA’s are for 1977. Other benchmarkyears based on I-O accounts are 1958, 1963, 1967, and1972. Before the use of I-O benchmarks, a less system-atic commodity-flow procedure was used for 1929–39(when the census of manufactures was biennial), 1947–48, 1951, 1954, and 1956. See [161, pp. 103–22] and[163, pp. 74–82] for a description of this procedure.

For nonbenchmark years, the current-dollar annualestimates of PCE are not prepared at the commoditylevel but are prepared at the level of the PCE cate-gories. For most PCE categories, the estimates areprepared by interpolating between and extrapolatingfrom the benchmark estimates using less comprehen-sive indicator series; this estimating method assumesthat some relationship underlying the benchmark esti-mates remains unchanged. For example, the use of allreceipts of domestic legal service establishments as theindicator for the PCE legal services category assumesthat the share of total domestic purchases attributableto persons remains unchanged. For most of the PCEcategories for which benchmark estimates are basedon quinquennial census data, the indicator series arederived from annual Census Bureau surveys of retailstore sales and service industry receipts. For otherPCE categories, a variety of related indicator series areused. In a few cases, the source data and estimatingmethods used are the same for both the benchmarkand the nonbenchmark estimates.

Current-dollar quarterly and monthly estimates ofmost PCE categories are prepared using indicator se-ries to interpolate between and extrapolate from theannual estimates. Among the more important monthlyindicator series are retail store sales, unit sales of au-tomobiles and trucks, wages and salaries, securitiestransactions, quantities of gasoline purchases, changesin the housing stock, and utility usage; where these se-ries provide quantity measures, monthly price indexesare used to obtain value indicators.

For some PCE categories, such as net foreign traveland several types of insurance, only quarterly sourcedata are available; for still others, only annual sourcedata are available. Where monthly, quarterly, or bothmonthly and quarterly source data are not available,monthly estimates are prepared by judgmental inter-polation and extrapolation of constant-dollar annualor quarterly series; the constant-dollar monthly seriesare then converted to current dollars using monthlyprice indexes.

Introduction 29

ACLI American Council of Life Insurance.AGA American Gas Association [6, 7, 8].AHA American Hospital Association [9, 10].AKC American Kennel Club [11].APTA American Public Transit Association [14, 15].ARTS Annual Retail Trade Survey (Census Bureau) [128].ASM Annual Survey of Manufactures (Census Bureau)

[129].AT&T American Telephone and Telegraph Company [16,

17].ATA Air Transport Association of America [1, 2].BLS Bureau of Labor Statistics, U.S. Department of Labor

[197].BPA Balance of Payments Accounts (Bureau of Economic

Analysis) [158].CA Census of Agriculture (Census Bureau) [113].CES Center for Education Statistics (U.S. Department of

Education) [164, 165, 166, 167, 168].CM Census of Manufactures (Census Bureau) [117].CMI Census of Mineral Industries (Census Bureau) [118].COG Census of Governments (Census Bureau) [114].CPB Corporation for Public Broadcasting [30, 31].CPI Consumer Price Index (BLS) [201].CPS Current Population Survey (Census Bureau) [137].CRT Census of Retail Trade (Census Bureau) [119].CSI Census of Service Industries (Census Bureau) [120].CWT Census of Wholesale Trade (Census Bureau) [124].EE Employment and Earnings (BLS) [202].EEI Edison Electric Institute [33].EIA Energy Information Administration, U.S. Department

of Energy [169, 170, 172-177, 180, 182].EW Employment and Wages (BLS) [203].FAA Federal Aviation Administration [210, 211, 212].FCC Federal Communications Commission [39-42].FDIC Federal Deposit Insurance Corporation [43, 44].FEC Federal Election Commission [222, 223, 224].FHLBB Federal Home Loan Bank Board [46].FHWA Federal Highway Administration, U.S. Department of

Transportation [213].

FRB Federal Reserve Board [24].FRBNY Federal Reserve Bank of New York.GF Governmental Finances (Census Bureau) [138].HCFA Health Care Financing Administration, U.S. Depart-

ment of Health and Human Services [189, 190,191].

HIAA Health Insurance Association of America [50].HR House of Representatives [226].ICC Interstate Commerce Commission [57, 58, 60, 61,

62].IRS Internal Revenue Service, U.S. Department of the

Treasury [205, 206].MRTS Monthly Retail Trade Survey (Census Bureau) [127,

131].NAB Newspaper Advertising Bureau [96].NADA National Automobile Dealers Association [76].NASD National Association of Securities Dealers.NASRC National Association of State Racing Commissioners

[74].NCHS National Center for Health Statistics, U.S. Department

of Health and Human Services [192].NFL National Football League [86].NSA Not seasonally adjusted.NYSE New York Stock Exchange.OTSAI Office of Transportation Systems Analysis and Infor-

mation [215].PPI Producer Price Index (BLS) [204].QS Quarterly Summary of Federal, State, and Local Tax

Revenue (Census Bureau) [141].SAS Service Annual Survey (Census Bureau) [132].SEC Securities and Exchange Commission [229].SSA Social Security Administration, U.S. Department of

Health and Human Services [193].TIUS Truck Inventory and Use Survey (Census Bureau)

[123].TSC Transportation Systems Center [216, 217, 218].USDA U.S. Department of Agriculture [108].USPS U.S. Postal Service [228].

3.2 Constant-dollar estimates

Constant-dollar estimates are prepared using one ofthree methods. For most PCE categories, the methodused is deflation, in which the constant-dollar es-timates are obtained by dividing the most detailedcurrent-dollar categories by appropriate price indexes,primarily Bureau of Labor Statistics (BLS) consumerprice indexes. The other methods of preparingconstant-dollar estimates both use quantity indica-

tors; these methods are quantity extrapolation anddirect base-year valuation. For quantity extrapolation,constant-dollar estimates are obtained by extrapolat-ing the current-dollar estimates from the base yearby quantity indicators. For direct valuation, constant-dollar estimates are obtained by multiplying base-yearprices times quantity data for the current period. Allconstant-dollar estimates are expressed in terms of asingle base period, currently 1982. This period usuallyis changed at the time of a comprehensive revision.

Current-Dollar Estimates

15. The description in the following pages focuses on the source data thatare important determinants of the benchmark PCE estimates. It ignores theunique strength of the I-O accounts for benchmarking. Briefly, the compilationof the I-O accounts requires that all inputs and outputs for each industry beaccounted for fully. This requirement means that each data source is cross-checked indirectly by every other data source. If there are major imbalances in

This section provides more detailed informationon the source data and estimating methods used toprepare the current-dollar PCE estimates for bothbenchmark and nonbenchmark years and the quar-terly and monthly estimates of PCE. This informationfor benchmark years is summarized in table 11; fornonbenchmark years in table 12; and for quarters andmonths in table 13. In the section on special pro-cedures, more detailed descriptions are provided forselected categories.

1 Annual Estimates: BenchmarkYears

PCE estimates for benchmark years are the estimatesfrom BEA’s benchmark I-O accounts, adjusted forthe latest definitional, classificational, and statisticalchanges incorporated into the NIPA’s. These accountsshow domestic output of each commodity and its dis-position either as consumption by industries or as partof one or more components of final demand, whichsum to gross national product (GNP). PCE and theother components of final demand are presented inthe I-O accounts as the sum of purchases of detailedcommodities—that is, goods, services, or structures.In the presentation of PCE in the NIPA’s, these com-modities are grouped into the detailed type of productor type of expenditure categories shown in the NIPAtables.14 Two methods are used in preparing the bench-mark estimates of PCE: The commodity-flow methodand direct estimation.

1.1 Commodity-flow estimates

For most commodities purchased by persons, the es-timates in the I-O accounts are calculated using thecommodity-flow method, which is outlined in table 10.This method first converts domestic sales, which is thevalue of sales of commodities produced by domesticfirms at producers’ prices, to domestic supply, which isthe value of sales to domestic purchasers at purchasers’prices and, therefore, includes imports and excludesexports. Then it allocates domestic supply among

14. A complete listing of the commodities underlying each PCE productcategory is available from BEA. For additional information, write to BEA(BE-51), U.S. Department of Commerce, Washington, D.C. 20230.

domestic purchasers—that is, persons, business, andgovernment.15

1.1.1 Commodities purchased by persons

In step 1 of the commodity-flow method, commoditiespurchased by persons are identified by BEA on thebasis of the nature of the product from the producttitles contained either in the detailed tabulations in thequinquennial economic censuses or in the StandardIndustrial Classification Manual [221].

1.1.2 Domestic sales

In step 2, estimates of total domestic sales—that is,shipments, revenues, or receipts—are prepared foreach of the commodities identified in step 1. These esti-mates reflect sales of commodities produced by privatefirms, regardless of their principal activity, and by gov-ernment agencies; the value of the sales includes excisetaxes and tips. For example, shipments of jewelry areobtained by combining shipments by manufacturerswith the value of lapidary work performed by whole-sale and retail jewelers; receipts for auto repair includerepair work performed by gasoline service stations andautomobile dealers, as well as by auto repair shops;and revenues for electricity include electricity providedby both private and government utilities.

In the 1977 I-O accounts, shipments of goods are pri-marily from the detailed product statistics of the 1977Census of Manufactures [117]; sales, receipts, or ex-penses for services are primarily from the 1977 Censusof Service Industries [120] and the 1970 and 1980 Cen-suses of Housing [115]; the estimates of excise taxesare those prepared for BEA’s government accounts, asdescribed in MP-5 [152, pages 24 and 25]; tips are es-timated by BEA by applying appropriate tipping ratesto sales or receipts. For goods produced in nonmanu-facturing industries, shipments data are from the U.S.Department of Agriculture (USDA) [108] and from the1977 Census of Agriculture [113], the 1977 Census of

inputs or outputs, source data and proceduresmust be reviewed and estimatesreworked. Small imbalances between the supply and use of a commodity aredistributed mathematically across the outputs (inputs) purchased by businessand, sometimes, the components of final demand.

31


Table 10.—Steps in the Commodity-Flow Method Used for PCE

1. Identify commodities purchased by persons.

2. Estimate domestic sales—shipments, sales, and/or receipts of these commodities at producers’ prices.

3. Add imports.

4. Add trade margins and related taxes and transportation costs.

5. Subtract exports.

6. Subtract change in trade inventories (mainly wholesale and retail).

7. Allocate domestic supply among business, government, and persons.

8. Adjust for intersector sales of used goods to derive purchases by persons.

Mineral Industries [118], the 1977 Census of Whole-sale Trade [124], the 1977 Census of Service Industries[120] and the 1977 Census of Retail Trade [119].16 Re-ceipts of government enterprises and sales by othergovernment agencies are from the 1977 Census of Gov-ernments, Governmental Finances, and Federal budgetdocuments as described in MP-5 [152, pages 33, 49, 91,and 99].

1.1.3 Imports

In step 3, imports are added because in the NIPA’s,they are included in the final demand of each typeof purchaser and subtracted as a separate entry indetermining GNP.

The source for merchandise imports is the CensusBureau monthly merchandise trade series [145]. Forimports of services, the source is BEA’s balance of pay-ments accounts, which rely heavily on quarterly andannual BEA surveys, as described in MP-3 [151, pages22, 23, and 37] and in the June 1989 SURVEY OF CURRENT

BUSINESS [158].

1.1.4 Trade margins and related taxes andtransportation costs

In step 4, distribution costs—consisting of trademargins, sales and excise taxes collected by thetrade establishments, and transportation costs—areadded because transactions measured so far in thecommodity-flow method mostly have been stated atproducers’ prices while the components of final demandare valued at purchasers’ prices.

16. The economic census data used to prepare the 1977 I-O accounts wereadjusted to account for undercoverage due to the exclusion of certain smallfirms from those censuses and for misreporting on the tax return data used bythe Census Bureau to conduct the censuses. The adjustment for misreportingis described in the June 1984 SURVEY OF CURRENT BUSINESS [153].

In the 1977 I-O accounts, trade margins for bothwholesalers and retailers are estimated for eachcommodity in two stages.

• First, margins for each establishment indus-try (that is, the unit that is classified isthe establishment) are estimated by multiplyingestablishment-industry margin rates by the cor-responding establishment-industry sales. Themargin rates are based on establishment-industrydata on sales, sales tax rates, and cost of goodssold from the Census Bureau 1977 annual sur-veys of merchant wholesalers and retailers [128,130]; wholesale and retail trade sales by estab-lishment industry and by commodity are from the1977 Censuses of Wholesale and Retail Trade [119,124].

• Then, margins for each of the commodities soldin each industry are estimated. This is ac-complished for each industry by multiplying themargin rate for the establishment industry inwhich the commodity is primary by sales of thecommodity in the given industry and then adjust-ing the sum over all commodities to the marginof the given industry. Margins also are added toaccount for sales of the commodity by other in-dustries, such as government-owned liquor storesand serviceindustry establishments. When morecommodity detail is needed, wholesale margins areestimated for the appropriate commodities basedon the composition of manufacturing shipments towholesalers; detailed retail margins are estimatedbased on the composition of products purchased bypersons.

Sales taxes at both the wholesale and the retaillevels are estimated for each commodity using thesame source data and procedures as used for the trademargins.

Current-Dollar Estimates 33

Transportation costs are derived primarily from1977 revenue data by mode of transportation: Forrailroads and motor freight, from the Interstate Com-merce Commission (ICC) [62] and the 1977 Censusof Transportation: Nonregulated Carriers and PublicWarehousing [12]; for pipelines, from the Departmentof Energy (DOE) [184]; for airlines, from the Depart-ment of Transportation (DOT) [216]; and for water,from the Department of Defense (DOD) [111]. Thecosts are allocated by commodity using data from thesesources, from the ICC [59, 61], and from the 1977Census of Transportation: Commodity TransportationSurvey from the Census Bureau [121].

1.1.5 Exports

In step 5, exports, which include distribution costs, aresubtracted because they are recorded in the NIPA’s asa separate final demand component.

The source for merchandise exports is the CensusBureau monthly merchandise trade series [144]. Forexports of services, the source is BEA’s balance of pay-ments accounts, which rely heavily on quarterly andannual BEA surveys, as described in MP-3 [151, pages22, 23, and 37] and in the June 1989 SURVEY OF CURRENT

BUSINESS [158].

1.1.6 Change in trade inventories

In step 6, changes in inventories held by distribu-tors are subtracted, because not all goods shipped byproducers or imported in a given period are final pur-chases in the same period. In the 1977 I-O accounts,the inventory data used for this step are primarilyfrom inventories reported by establishments in the1977 Census of Wholesale Trade [124] and in the 1977Annual Retail Trade [128].

Although sales of establishments are collected bytype of commodity, inventories are collected only forthe establishment as a whole. The commodity compo-sition of inventories for each establishment industry,therefore, is estimated by multiplying sales of eachcommodity by the inventory-sales ratio of the estab-lishment industry in which the product is primary. Thesum of the resulting commodity inventories is thenadjusted to equal the corresponding establishment-industry total. Total inventories for each commodityare obtained by summing the commodity detail acrossall industries.

1.1.7 Allocation of domestic supply amongbusiness, government, and persons

In step 7, purchases by government and businessare subtracted from domestic supply at purchasers’prices—the result of steps 2 through 6—to obtain pur-chases by persons. However, this residual procedure isnot used where additional source data are available onpurchases by persons. There are two variations on the

standard commodity-flow procedure: (1) For a few com-modities, source data are available to allocate domesticsupply at purchasers’ prices among persons, business,and government. (2) For some commodities, sourcedata are available to estimate purchases by personsdirectly.

For commodities where the residual procedure isused to obtain purchases by persons, estimates of com-modity purchases by governments and governmententerprises are described in MP-5 [152]. Estimates ofbusiness purchases are derived in part from CensusBureau data on purchased materials and services, butbecause such data are not available for all business,most business purchases must be estimated usingother data and, where necessary, judgment in placeof data. In these cases, inputs to an industry maybe estimated from trade source data, administrativerecords, or the cost structure of a similar industry forwhich data are available. For example, the expensesof business associations and professional membershiporganizations are estimated from the following di-verse sources: Census Bureau data on payroll andemployment, BEA estimates of capital consumptionallowances, and trade source data and administrativerecords for the remaining inputs. In other cases, sourcedata can provide the means for allocating the output ofa commodity among purchasers. Thus, trade sourcesprovide the basis for allocating the output of carpet-ing among industry inputs, business investment, andPCE.

In the first variation on the standard commodity-flowprocedure, independent estimates are prepared of thedistribution of domestic supply at purchasers’ pricesamong persons, government, and business. This pro-cedure is used for new autos, new trucks, and gasoline.In the second variation on the standard commodity-flow procedure, purchases by persons are estimateddirectly, either at producers’ prices (coal, broadwovenfabric, yarn, and knit fabric) or at purchasers’ prices(natural gas, electricity, telephone and telegraph, pre-scription drugs, motor oil, fuel oil, liquified petroleumgas, and kerosene).

1.1.8 Intersector sales of used goods

In step 8, net transactions in used goods between per-sons and other sectors are added. Margins on sales ofused goods are included in earlier steps because theyare based on the same source data as the trade marginsfor new goods described in step 4. Intersector sales(net transactions) are net sales of used goods to personsby business, foreigners, and government. The largestof these transactions are those in used motor vehicles;the sources used for the motor vehicle estimates are de-scribed in the section on special procedures. For othergoods, net exports are from the Census merchandisetrade series; business sales, primarily sales of usedrental equipment, are derived from data on receipts ofequipment rental establishments reported in the 1977


Census of Service Industries [120]; and governmentsales of used equipment are estimated as described inMP-5 [152, pages 93–96].

1.2 Direct estimates

The commodity-flow method is not used for PCE cat-egories where, by definition, the commodities arepurchased only by persons. These categories arethe following: Food furnished employees (includingmilitary); food produced and consumed on farms; stan-dard clothing issued military personnel; the rentalvalue of owner- and tenant-occupied dwellings; ex-pense of handling life insurance; personal remittancesin kind to foreigners; expenditures abroad by U.S.residents; expenditures in the United States by for-eigners; domestic service other than cleaning services;health insurance; and reimbursement for governmentemployee use of own motor vehicles.

Most of these categories are discussed in the in-troduction. The last category, reimbursement forgovernment employee use of own motor vehicles, is anegative entry in PCE for used autos that offsets a pos-itive entry in government purchases. Estimates for allthese categories are prepared directly using a varietyof source data.

1.3 Sources of the benchmark estimates

Table 11 identifies the principal source data and esti-mating methods for key elements that are important inthe derivation of the PCE estimates. This informationis presented for each PCE category shown in NIPA ta-ble 2.5; additional commodity detail is presented wheredifferent methodologies are used for the commoditieswithin the category. Where special procedures areneeded either because of the nature of the commodi-ties (for example, rent) or the availability of moredetailed source data (for example, new autos) table 11provides only a summary description. More detail ispresented at the end of this section for the followingPCE categories: New autos, net transactions in usedmotor vehicles, gasoline, rental value of owner- andtenant-occupied nonfarm dwellings, brokerage chargesand investment counseling, services furnished with-out payment by financial intermediaries except lifeinsurance carriers and private noninsured pensionplans, expense of handling life insurance, and religiousactivities.

Estimates of most PCE categories are prepared us-ing the standard commodity-flow method describedabove; these categories are identified by the entry“residual estimate” in the “purchases by persons” col-umn of table 11. The “domestic supply” columnidentifies the primary source of domestic commoditysales, which is one of the key elements in preparing theestimates of supply and which differs among the cate-gories. This column also identifies secondary sourcesof sales, such as those by governments, for categories

where these sales are significant. Generally, no infor-mation is provided about the other key elements in thestandard commodity-flow method—exports, change ininventories, distribution costs, and imports—becausetheir sources, which are described in the text, are thesame for all PCE categories. Where this is not the case,the other sources and methods are described in table11.

Table 11 also indicates the categories that areprepared using variations on the standard commodity-flow method. For these categories, the “purchases bypersons” column describes the additional source dataand the methods used. In addition, there are a fewcategories, such as firewood, for which purchases bybusiness and government are assumed to be statisti-cally insignificant. These are identified by the entry“all purchases assumed by persons.”

Finally, table 11 indicates the categories that areestimated directly because, by definition, they areunique to PCE, such as food furnished civilian em-ployees. These categories are identified by the entry“all purchases defined to persons” in the “purchasesby persons” column. The entry “n.a.” (not applica-ble) in the “domestic supply” column indicates that thecommodity-flow method is not used for this categoryand that the category is estimated directly. In thesecases, the “purchases by persons” column identifies theprincipal source data used for the estimate.

For most of the PCE categories and commodities,“purchases by persons” are purchases by individualsonly. As explained in the introduction, the expend-itures of nonprofit institutions are not distributedamong the individual commodities in PCE. A measureof their services equal to their current operating ex-penses is included in those PCE categories—other thanlife insurance—for which the domestic supply columnhas the entry “expenses.”

2 Annual Estimates:Nonbenchmark Years

In general, estimates of PCE categories for nonbench-mark years are prepared using an indicator series tointerpolate between and extrapolate from the bench-mark estimates, because most of the latter dependheavily on data available only from quinquennial cen-suses. The most important census data that are notavailable for other years are the following: (1) De-tailed commodity shipments, revenues, and receipts,which are used to identify commodities purchasedby persons and which provide measures of outputof these commodities; (2) sales of wholesalers andretailers, by merchandise line, which provide the in-formation necessary to convert trade inventories andtrade margins from an establishment-industry basisto a commodity basis; and (3) materials purchased bymineral industries, construction, and manufacturingestablishments, which provide measures of business


purchases used in the residual calculation of purchasesby persons.

In addition, certain data that are important in de-riving the PCE estimates are not available for otheryears because they are derived in preparing the bench-mark I-O accounts. For example, the commodity detailon purchases by State and local governments entersinto the estimates of purchases by business and gov-ernment used in the residual calculation of purchasesby persons, and State and local government sales de-tail enters into the calculation of domestic commoditysales.

The source data and estimating methods used to pre-pare the 1986–88 estimates released in July 1989 areshown in table 12. (Major changes over the years in thesource data used are indicated in the footnotes to table9.) Where necessary, the table distinguishes amongthe source data and estimating procedures used in thefirst, second, and third July revisions. Finally, the ta-ble indicates the few categories for which the indicatorseries is not based on annual source data, but is thesum of the monthly or quarterly estimates. The infor-mation in table 12 is presented in the same detail as intable 11 except for computers, American Eagles, andvideo cassette rentals, three commodities that were notpurchased by persons in 1977.

2.1 Goods

2.1.1 Goods extrapolated by the retail-controlmethod

The retail-control method provides, first, the indica-tor series used in interpolating and extrapolating thetotal of most goods and, second, provides the “con-trol” total to which the categories included in thisgroup must sum. The goods to which it is appliedare called the “retail-control group,” consisting of allgoods except motor vehicles, computers and peripheralequipment, gasoline and oil, food furnished employees,food produced and consumed on farms, standard cloth-ing issued military personnel, tobacco products, andnet foreign remittances.

The indicator series for the retail-control total is de-rived primarily from retail store sales as reported bythe Census Bureau, excluding sales of building ma-terial stores, hardware stores, garden supply stores,mobile home dealers, gasoline service stations, and mo-tor vehicle dealers. (These outlets are omitted becausemost of their sales either relate to homeownershipitems that are not included in PCE or consist of prod-ucts that can more reliably be estimated from othersource data.) To the remaining retail store sales areadded (1) sales of military commissaries and exchangesand (2) sales of goods other than gasoline and oil bygasoline service stations. The latter sales are calcu-lated as total sales at gasoline service stations lesssales of gasoline and oil and less receipts for repairservices. (Sales of gasoline and oil are estimated using

quantities and average price data; receipts for repairservices are estimated as a fixed proportion of salesother than for gasoline and oil, based on merchandiseline sales from the most recent census of retail trade.)The indicator series is used to extrapolate total PCEfor the retail-control group.

Estimates of the underlying commodity detail arederived primarily from data on merchandise line salesfrom the most recent census of retail trade and annualdata on retail sales by kind of business. First, indepen-dent estimates of purchases by persons of tobacco andof computers and peripheral equipment are removed.Then, the commodity composition of the remainder isdetermined using a two-stage process. In the firststage, the benchmark levels of the component com-modities are extrapolated by estimates of retail salesof those commodities; these estimates are preparedby disaggregating annual data on retail sales for eachkind of business, using commodity proportions fromthe census of retail trade data on merchandise linesales, and then aggregating each commodity acrosskinds of business. In the second stage, estimatedsales of the extrapolated commodities are adjusted pro-portionately to agree with the previously determinedretail-control group total. (The retail-control group in-cludes landlord purchases of durable equipment, whichare included in the residential investment componentof GNP, and these purchases are also estimated.)

For 1948–58, the data used for the retail control in-dicator series were a combination of Census Bureauretail sales of specialty stores [131] and Federal Re-serve Board data on retail sales of department stores,by department [23]. This indicator was also used tointerpolate between the commodity flow estimates of1929–39 and 1947.

The retail-control method assumes that the follow-ing three ratios remain constant: (1) The ratio ofmerchandise line sales of each commodity to total salesfor each kind of business within the control group; (2)the ratio of total purchases by persons at retail-controlgroup stores to total purchases by persons; and (3) theratio of purchases by persons at these stores to totalsales of these stores.

2.1.2 Other goods

Table 12 also indicates the methods used for nonbench-mark year estimates of goods that are not extrapolatedby the retail-control method. For new trucks, an ab-breviated commodity-flow method is used. For othermotor vehicles and for gasoline and oil, the proce-dures are discussed in more detail at the end of thissection. For food furnished military employees, foodproduced and consumed on farms, standard clothingissued military personnel, and net foreign remittances,the method is the same as that used in the benchmarkestimates. For tobacco products and for computersand peripheral equipment, independent source dataare used.


Note.—Number in parentheses after title is line number in NIPA table 2.4. Number in brackets refers to the item number in the listing of sources.

ACLI American Council of Life Insurance.AGA American Gas Association [6, 7, 8].AHA American Hospital Association [9, 10].AKC American Kennel Club [11].APTA American Public Transit Association [14, 15].ARTS Annual Retail Trade Survey (Census Bureau) [128].ASM Annual Survey of Manufactures (Census Bureau)

[129].BLS Bureau of Labor Statistics, U.S. Department of Labor

[197].BPA Balance of Payments Accounts (Bureau of Economic

Analysis) [158].CES Center for Education Statistics (U.S. Department of

Education) [164, 165, 166, 167, 168].CPB Corporation for Public Broadcasting [30, 31].CPI Consumer Price Index (BLS) [201].CPS Current Population Survey (Census Bureau) [137].EE Employment and Earnings (BLS) [202].EEI Edison Electric Institute [33].EIA Energy Information Administration, U.S. Department

of Energy [169, 170, 172-177, 180, 182].EW Employment and Wages (BLS) [203].FAA Federal Aviation Administration [210, 211, 212].FCC Federal Communications Commission [39-42].FDIC Federal Deposit Insurance Corporation [43, 44].FEC Federal Election Commission [222, 223, 224].

FHWA Federal Highway Administration, U.S. Department ofTransportation [213].

FRBNY Federal Reserve Bank of New York.GF Governmental Finances (Census Bureau) [138].HCFA Health Care Financing Administration, U.S. Depart-

ment of Health and Human Services [189, 190,191].

HIAA Health Insurance Association of America [50].ICC Interstate Commerce Commission [57, 58, 60, 61,

62].IRS Internal Revenue Service, U.S. Department of the

Treasury [205, 206].MRTS Monthly Retail Trade Survey (Census Bureau) [127,

131].NAB Newspaper Advertising Bureau [96].NADA National Automobile Dealers Association [76].NASD National Association of Securities Dealers.NASRC National Association of State Racing Commissioners

[74].NFL National Football League [86].NYSE New York Stock Exchange.PPI Producer Price Index (BLS) [204].SAS Service Annual Survey (Census Bureau) [132].SEC Securities and Exchange Commission [229].TSC Transportation Systems Center [216, 217, 218].USDA U.S. Department of Agriculture [108].USPS U.S. Postal Service [228].

2.2 Services

A variety of methods is used to construct indicatorseries for the PCE service categories, as indicated intable 12. The methodologies used for the rental valueof owner- and tenant-occupied dwellings, brokeragecharges and investment counseling, services furnishedwithout payment by financial intermediaries exceptlife insurance carriers and private noninsured pen-sion plans, the expense of handling life insurance, andreligion are described in greater detail at the end ofthis section. The preferred approach for the remain-der is to use receipts—or, for nonprofit institutions,expenses—as the indicator.

For services that use industry receipts as the indica-tor series—for example, legal services—it is assumedthat the proportion of these receipts accounted forby persons in the benchmark year remains constant.For services for which annual data on receipts arenot available, several approaches are used. In somecases—for example, spectator sports—an indicatorseries for receipts is derived from the number of ad-missions and the CPI for admissions, again on theassumption that the proportion of industry receiptsaccounted for by persons remains constant. In othercases—for example, employment agency fees—the in-dicator series is wages and salaries for the industry.This approach assumes that both the ratio of wagesand salaries to receipts and the proportion of industryreceipts accounted for by persons remain constant.

Finally, for services provided mainly by nonprofitinstitutions—for example, social welfare—the indica-tor series also is wages and salaries for the industry.This approach assumes that the ratio of wages andsalaries to total current operating expenditures and

the proportion of these expenditures accounted for bypersons remain constant.

3 Quarterly and MonthlyEstimates

The source data and estimating methods used inpreparing the quarterly and monthly estimates ofcurrent-dollar PCE released in July 1989 are shownin table 13, using the same categories and detail asthose shown in table 12. Table 13 identifies categoriesfor which different sources and methods are used for“historical” estimates—that is, those prepared for theyears covered by annual or benchmark revisions—andfor “current” estimates—that is, those prepared for theyears that have not yet been subject to an annual revi-sion. (Major changes over time in the source data usedare identified in the footnotes to table 9.) Table 13 alsoprovides information about the seasonal adjustment ofthe PCE estimates. Except as noted, indicator seriesare seasonally adjusted using the Census X-11 sea-sonal adjustment program. For categories for whicha judgmental trend is used for quarterly or monthlyestimates, the trend represents a seasonally adjustedseries.

3.1 Historical estimates

Historical estimates are prepared from annual inter-polations that generally use the same indicator seriesor judgmental trend as that used for the currentestimates, unless additional source data become avail-able. Where source data are quarterly, the historical


17. Domestic autos consist of units assembled in the United States, Canada,and Mexico; imported autos are those assembled elsewhere.

monthly series is derived from the quarterly series byinterpolation using a judgmental trend.

Historical quarterly not seasonally adjusted esti-mates are prepared by interpolation using quarterlyindicator series that are either quarterly source dataor the sum of monthly source data. For categoriesin which a judgmental trend is used for the season-ally adjusted indicator series, the unadjusted quar-terly estimates are prepared using the constant-dollartrend reflated by not seasonally adjusted price in-dexes. (Historical monthly unadjusted estimates arenot prepared.)

3.2 Current estimatesCurrent estimates for most categories are extrapo-lations of the most recent July revision estimates.In most cases, these extrapolations are prepared us-ing indicator series; judgmental trends are used toestimate the remaining categories, as well as thoseindicator series for periods for which the source dataare not yet available. For each judgmental extrapola-tion, the estimate is prepared by first extrapolating theconstant-dollar series and then reflating it to currentdollars with the price index used as the deflator for thecategory, usually a component of the CPI.

Although the sources and methods used to preparethe monthly estimates of current-dollar PCE are notshown separately in table 13, information about theirpreparation can be determined from the quarterlydescriptions:

• For PCE categories for which 3 months of sourcedata or indicator series are available for theadvance quarterly estimate, the initial monthlyestimate is prepared using that information.

• For PCE categories for which 1 or 2 months ofsource data or indicator series are available for theadvance estimate, a judgmental trend is used forthe initial monthly estimate and any other missingmonth.

• For PCE categories for which quarterly source dataor indicator series are used, monthly estimatesare prepared by interpolating between and ex-trapolating from the quarterly estimates using ajudgmental trend as the indicator series.

• For PCE categories for which a judgmental trend isused to prepare the quarterly estimates, the trendis estimated monthly and is used to prepare boththe quarterly and monthly estimates.

Current quarterly and monthly not seasonallyadjusted estimates are not prepared.

4 Special Procedures Used forSelected PCE Categories

The source data and estimating methods for the PCEcategories discussed in this section were too complex

to be summarized adequately in the tabular presen-tation of tables 11, 12, and 13. These descriptionscover the benchmark annual, nonbenchmark annual,and quarterly and monthly estimates.

4.1 New autos

Estimates of PCE for new autos are derived by valuingunit sales purchased by persons by average expendi-ture, which is the average retail list price adjusted fordiscounts, sales taxes, and transportation costs. Sepa-rate estimates are made for expenditures on domesticand on imported autos.17 The proportion purchasedby persons (consumer share) is derived using informa-tion on new auto registrations. The source data arefrom trade sources, except for the data on discounts,sales taxes, and transportation costs. Annual esti-mates for all years and quarterly estimates are the sumof monthly estimates. Where source data are availableonly annually, relationships determined from these an-nual data are held constant for the monthly estimatesextrapolated from the most recent year. For quarterlyestimates, 3 months of unit sales and list prices of bothdomestic and foreign autos, 3 months of price adjust-ments, and 1 month of registrations data are availablefor the advance estimate. Projections of missing sourcedata are based on judgmental trends.

4.1.1 Data sources

Data on monthly unit sales of domestic autos, bymodel, are from the Motor Vehicle Manufacturers As-sociation of America (MVMA) [68]; monthly unit salesof imported autos are from Ward’s Automotive Reports(Ward’s) [234]. Data on the proportion of each modelmanufactured or imported with specified options areavailable for model year to date at approximately 4-month intervals from Ward’s. Data on the proportionof new autos by type of owner—that is, persons, busi-ness, or government—are from monthly R.L. Polk &Company data on registrations of new autos [100].The MVMA and Ward’s data have been used since1951 and 1956, respectively. Estimates for 1946–50are based on R. L. Polk data on new registrations;those for 1942–1945 on rationing data from the Of-fice of Price Administration (OPA) [162]; and those for1929–41 on data on dealers’ sales from the AutomobileManufacturers’ Association [19], the predecessor of theMVMA.

Manufacturers’ wholesale and retail list prices foreach style and option of each model are from the Au-tomobile Invoice Service (AIS) [18]. Average dealerdiscounts, or premiums, and transportation costs arefrom the Bureau of Labor Statistics (BLS) [197], whichcollects this information monthly as part of its reg-ular pricing program for the Consumer Price Index(CPI). Sales taxes are estimated using a tax rate for


franchised car dealers from the Census Bureau 1977Annual Retail Trade Survey [128] and taxes from theCensus Bureau’s quarterly survey of State and localtax revenue [141]. Estimates from 1946–66 are basedon price data from the BLS composite price for se-lected models [199], adjusted beginning in 1954 forallowances and discounts. Prices from 1942–45 areestimated in accordance with OPA regulations, in-creasing 1 percent per month from the December 1941price and an additional 1 percent per month from May1944 to December 1944. The December 1944 priceis held constant through the first quarter of 1945.Price data from 1929–41 are from the AutomobileManufacturer’s Association.

4.1.2 Unit sales

Unit sales of domestic autos to persons are derived asthe product of total unit sales and the proportion ofnew auto registrations recorded in the names of indi-viduals from the Polk data. The Polk data are adjustedby BEA for “mixed-use” autos—that is, autos used bothfor business and personal use. This adjustment isbased on Census data [134] on business mileage drivenby household members, updated annually to reflectchanges in the number of self-employed persons perhousehold from the Census Bureau current populationsurvey (CPS) [136]. This procedure applies to esti-mates from 1946 forward. The PCE share in earlieryears is based on survey data from the OPA (1942–45)[227] and the Bureau of Public Roads [159], prede-cessor to the Federal Highway Administration. Thisshare, after a small deduction for Federal, State, andlocal government purchases, is 70 percent from 1929–41 and 50 percent from 1942–45. The allocation forthese years applies to new autos and to gross marginson used autos.

Sales to persons of imported autos are assumed tobe 96 percent of total sales of imported autos less fleetpurchases, which are estimated from annual data fromAutomotive Fleet Fact Book [28].

4.1.3 Average expenditure

Average list prices are estimated by combining re-tail list prices with unit sales of each model. Theseaverage list prices are adjusted to an average expend-iture basis using BLS data, classified by size class andmanufacturer, on discounts and transportation costs;sales taxes are added based on the tax rate from thebenchmark estimates.

Further adjustments are made for fleet purchasesby businesses and governments, which are assumed tobe made at wholesale prices. These fleet purchase ad-justments are based on unit sales from monthly Polkdata and from annual data on fleet purchases by busi-nesses and government from the Automotive Fleet FactBook. (Expenditures by business and government are

used to estimate the new auto component of producers’durable equipment and government purchases.)

4.2 Net transactions in used autos

In PCE, purchases of used goods consist of the dealers’margin on goods purchased by persons and net trans-actions (purchases less sales) between persons andother sectors valued in wholesale prices. For used au-tos the estimate of net transactions represents severaltypes of transactions. Most purchases are from newor used auto dealers, but some are directly from othertypes of businesses and from government and foreign-ers. Most sales are also to new and used auto dealers,either as trade-ins or as direct sales. Individuals alsosell autos that are at the end of their useful lives toscrap dealers, and in some instances they sell directlyto other businesses and to government and foreigners.(Transactions between persons are intrasectoral andcancel in aggregation.) The estimate of net transac-tions in used autos is not developed by explicitly takinginto account each type of transaction listed above butby estimating the change in unit stocks of used au-tos held by persons and then valuing that change atwholesale prices.

The procedure is described in three parts; in eachpart, autos are grouped by age—determined by year oforiginal sale—and the calculations are carried out byage group. First, the total unit stock of autos in useis estimated and the year-to-year change in the stockseparated into its various components—purchases ofnew autos, the change in dealers’ inventories of usedautos, exports and imports of used autos, and scrappedunits.

Second, unit stocks held by persons are calculatedas a residual by estimating unit stocks held by gov-ernment and by business and then deducting theseestimates from total unit stocks. In establishing unitstocks held by persons, the business portion of “mixed-use” autos held by persons is assigned to business.Then the change in unit stocks held by persons is di-vided into purchases of new autos, scrapped units, andnet unit transactions, except scrapped units.

Third, the change in unit stocks held by persons isvalued at wholesale prices, with a lower value assignedto scrapped units than to other autos.

4.2.1 Total unit stock and components of change

The total unit stock of autos in use is based on salesof new autos for the current and previous years, as-sumed retention rates for autos less than four yearsold, and R.L. Polk Company registration data on thenumber of autos in use by model year [234]. The as-sumed retention rates are shown in table 14. The R.L.Polk Company data are for autos in use on July 1 ofeach year. This information is converted to an end ofcalendar year basis.


Table 14.—Retention Rates and Percent of Original Value by Age of Auto

Age

Assumed retention rates[Percent] Percent of original value

in constant dollarsTotal Government Business

< 1 year ............................ 100.0 100.0 100.0 100.0

1–2 years .......................... 99.8 99.0 70.0 68.4

2–3 years .......................... 99.6 96.0 65.0 51.4

3–4 years .......................... 99.0 90.0 60.0 36.8

4–5 years .......................... n.a. 80.0 55.0 26.4

5–6 years .......................... n.a. 70.0 50.0 17.6

6–7 years .......................... n.a. 60.0 45.0 11.4

7–8 years .......................... n.a. 50.0 45.0 7.6

8–9 years .......................... n.a. 50.0 45.0 3.6

9–10 years ........................ n.a. 0 0 1.7

10–11 years ...................... n.a. 0 0 0.8

>11 years .......................... n.a. 0 0 0.4

n.a. Not applicable.

The procedure for estimating purchases of new do-mestic and imported autos is described above. Usedauto inventories held by franchised domestic autodealers are from the Motor Vehicle Manufacturers As-sociation (MVMA) [68]. The number of used autos heldby franchised foreign auto dealers is estimated judg-mentally, and those held by independent used autodealers are estimated from the value of yearend inven-tories from the annual retail trade survey [128], andan average value based on a five-term moving averageof monthly auction prices for autos less than 8 yearsold from Automotive News [32]. Annual data on ex-ports and imports of used autos are from the CensusBureau’s U.S. Exports [144] and U.S. General Importsand Imports for Consumption [145].

Given estimates of the other components of thechange in the stock of autos in use, scrappage isestimated as a residual. In estimating the compo-nents of change by age of auto, it is assumed that thechange in dealers’ inventories and exports and importsare confined to newer models and scrappage to oldermodels.

4.2.2 Residual estimates of change in unit stockheld by persons

The unit stock of autos held by government and theunit stock of autos held by business (excluding the

business portion of “mixed-use” autos) are based onnew auto purchases for the current and previous yearsand assumed retention rates. The estimates of pur-chases of new autos by government and business arederived by allocating total sales of new autos amonggovernment, business and household purchases withnew auto registration data from R.L. Polk Company.The assumed retention rates are shown in table 14.The calculation of the unit stock held by business isreviewed against information on the average length oftime autos are retained in business fleets according tothe National Association of Fleet Administrators andon the total number of autos in fleets of four or more[28]. As necessary, the results of the calculation areadjusted so that the change in the unit stock is in linewith the change in the total number of autos in fleetsof four or more.

The unit stock of autos held by persons is derivedby subtracting the unit stocks held by government andbusiness (excluding “mixed-use”) from the total unitstock. The business portion of “mixed-use” autos isthen separated from the unit stock held by personsand included in the business stock. The separation isbased on information from the Census Bureau on theshare of total mileage by age of auto that householdsreport as business usage [126]. The effect of splittingthe stock held by persons between business and per-sonal use on the basis of mileage shares is to measure


the business use of personal autos on a “full-car equiv-alent” basis and thereby to measure the unit stocksheld by business and persons on a “full-car equivalent”basis.

The change in the unit stock of autos held by persons(excluding “mixed-use”) is separated into purchases ofnew autos, scrapped units, and net unit transactions,except scrapped units. Purchases of new autos by per-sons is known and scrapped units are calculated byage of auto as a proportion of the total scrappage pre-viously established. The proportion is assumed to beequal to the ratio of the unit stock held by personsto the total unit stock. Net unit transactions, exceptscrap, are calculated as the residual.

4.2.3 Valuation in wholesale prices

Net transactions, except scrap, grouped by age, arevalued in wholesale prices. Scrapped units by age arevalued at 8 percent of the wholesale price.

The estimate of the average wholesale value for eachage group of used autos is calculated by (1) stating theoriginal average retail value for each group in constantdollars, (2) adjusting that value for depreciation, and(3) restating the depreciated value in current prices.The original average retail value is the average ex-penditure calculated in estimating purchases of newautos. These values are adjusted for price change withthe consumer price index (CPI) for new autos [201].The depreciated values are calculated according to theschedule shown in table 14. This schedule was devel-oped from information on the wholesale value of usedautos published by the National Automobile DealersAssociation [77] for the years 1963 to 1972. Beginningwith August 1987, the restatement to current prices iscarried out by multiplying the depreciated values bythe CPI for used autos [201]. Before then, the CPI forused autos, unlike that for new autos, did not includean adjustment for quality change. The restatementwas carried out using a price index obtained by ad-justing the CPI for used cars for quality based on thequality adjustment made for new autos in the CPI fornew autos.18

4.2.4 Quarterly and monthly estimates

Quarterly and monthly estimates of net transactionsare calculated using a procedure that parallels aspectsof the annual procedure but in which calculations interms of values replace those in units. Five elementsof the annual unit calculations, expressed in values,are interpolated and extrapolated. They are (1) changein dealers’ inventories of used autos, (2) net exportsof used autos, (3) scrappage, (4) net transactions inused autos by business, and (5) net transactions inused autos by government. The sum of the first three

18. The adjustment for quality change in the used auto price index wasintroduced in July 1987 and covers the period beginning January 1984. It willbe carried back to earlier periods in the next comprehensive revision of GNP.

provides an estimate of the change in the total stockof used autos in use valued in wholesale prices. Theother two provide the basis for a residual calculationof net transactions in used autos by persons given theestimate of the change in the total stock of used autos.

The value of the change in dealers’ inventories ofused autos is estimated from dealers’ unit stocksvalued at average wholesale prices. Unit stocks offranchised domestic auto dealers are reported monthlyby the Motor Vehicle Manufacturers’ Association [68],and stocks of franchised foreign auto dealers and usedauto dealers are estimated judgmentally. Unit stocksare valued with the average wholesale price of usedautos published by Automotive News [32]. The val-ues of net exports and of scrappage are held constantfrom the previous annual estimates. The values of nettransactions by business and by government are in-terpolated and extrapolated by the products of theirunit purchases of new autos and the average wholesaleprice of used autos. Their unit purchases are thoseestimated in the procedure for new autos; the averagewholesale price is that from Automotive News.

4.3 Net transactions in used trucks

In PCE, purchases of used goods consist of the deal-ers’ margin on goods purchased by persons and nettransactions (purchases less sales) between personsand other sectors valued in wholesale prices. For usedtrucks, the benchmark estimate of net transactionsby persons is obtained as the difference between pur-chases of used trucks by persons from dealers and salesof used trucks by persons to dealers. It is assumedthat persons have no transactions in used trucks withbusinesses (other than dealers), governments, or for-eigners and that trucks owned by persons are not soldto scrap dealers at the end of their useful lives.

Purchases of used trucks by persons from dealersis estimated as (1) dealer sales less (2) exports less(3) dealers’ margins, (4) split into purchases by per-sons and by business. Dealer sales are from the 1977Census of Retail Trade merchandise line data [119],adjusted to include sales taxes using data from the1977 ARTS [128]; exports are from Census Bureaumerchandise trade data [144]; dealers’ margin is fromthe 1977 ARTS [128]; and the split between purchasesby persons and by business is based on informationon truck usage in the 1977 Truck Inventory and UseSurvey (TIUS) conducted by the Census Bureau [123].

Sales of used trucks by persons to dealers is esti-mated as (1) dealer sales less (2) dealers’ margin plus(3) the change in dealers’ inventories of used trucks, (4)split into sales by persons and by business. The changein dealer inventories is based on inventory data fromthe 1977 ARTS as described in the section on currentdollar benchmark estimates; the split between sales bypersons and by business is based on data from TIUS[123].


For nonbenchmark years, annual estimates of nettransactions by persons are interpolated betweenbenchmarks and extrapolated from the last bench-mark, using the number of new truck registrations tobusiness five years earlier from the R.L. Polk Company[100], and the CPI for used autos [201] as indicatorseries. Quarterly and monthly estimates are interpo-lations and extrapolations using as an indicator seriesthe monthly estimates of new trucks purchased bypersons.

4.4 Gasoline

Estimates of PCE for gasoline are prepared as theproduct of the quantity purchased and an averageretail price.

4.4.1 Quantities purchased

Except for the first annual revision, separate estimatesof quantities purchased by persons are prepared forgasoline used in each of the following types of vehi-cles: Autos, trucks, airplanes, motorcycles, and boats.Before 1946, estimates are only for auto gasoline.

Quantities are expressed in gallons, except as oth-erwise noted. For autos, consumer purchases areobtained as a residual after accounting for other uses.For trucks and airplanes, purchases by persons are es-timated directly. For motorcycles and boats, purchasesby persons are estimated judgmentally as a portion oftotal purchases for these vehicles.

For the first annual revision, a single estimate isprepared of the total quantity purchased for use in alltypes of vehicles.

Autos.—Purchases by persons are equal to total gaso-line purchased for use in autos less purchases bygovernment and business. Business purchases are thesum of purchases for autos held and used exclusivelyby business and the business portion of purchases bypersons for “mixed-use” autos.

Total gasoline purchased for use in autos is fromHighway Statistics [213, 214] or, before 1940, theMinerals Yearbook [196].19

Purchases for exclusively business autos and for gov-ernment autos are each estimated as the product of anauto stock estimate and estimated fuel usage per auto.The stock estimates are prepared by BEA, as describedin the section on net transactions in used autos. Inboth cases, fuel usage per auto is calculated as milesdriven per vehicle divided by average miles per gallon(MPG).

For exclusively business autos, the 1977 benchmarkestimate of miles driven per auto is based on 1977

19. For the 1968 to 1982 estimates, BEA adjusted the FHWA data to reducequantity purchased for autos and increase quantities purchased for trucks,based on an Energy Information Administration (EIA) survey [181]. Detailedinformation about the derivation of this adjustment, which now is reflected inthe FHWA data, as well as in the procedures described here, is contained inBEA Working Paper 5 [149].

data from a survey of fleet operators conducted by theBrookhaven National Laboratory [103]. Miles drivenper auto is assumed constant at the 1977 level forall years beginning in 1972, except for 1974, when itis assumed to have declined 5 percent in response tothe gasoline shortage. For 1946 to 1971, miles drivenper auto are from the National Association of FleetAdministrators (NAFA) [70].

MPG for 1977, 1982, and 1983 are from the NAFA.For other years from 1972 to 1981, MPG is interpolatedand extrapolated by an estimate of MPG calculatedusing MPG of new autos sold by model year, froman Environmental Protection Agency study [219], andweights derived from the age distribution of the BEAstock of exclusively business autos. After 1983, MPG isextrapolated by MPG for all autos from FHWA. From1946 to 1971, MPG is assumed to equal FHWA’s MPGfor all autos.

For government autos, the 1977 benchmark estimateof miles driven per auto is calculated as a weightedaverage of three groups—civilian Federal Governmentautos, State and local governments, and other Stateand local government autos. For civilian Federal Gov-ernment autos, miles driven per auto is based on datafrom the General Services Administration (GSA) [225];for State and local government police civilian autos,estimates are based on the Brookhaven survey. Milesdriven per auto for government autos is assumed con-stant at the 1977 level for all years beginning in 1972,except for 1974, when it is assumed to have been re-duced 5 percent in response to the gasoline shortage.From 1946 to 1971, miles driven per government autois assumed to be 77 percent of miles driven per auto forexclusively business autos. Average MPG for govern-ment autos for all years from 1972 to 1983 is assumedto be the same as that for civilian Federal Governmentautos from the GSA report. After 1983, it is extrapo-lated by MPG for all autos from FHWA. From 1946 to1971, MPG is assumed equal to the FHWA measurefor all autos.

The business portion of purchases by persons is esti-mated as a proportion of total auto gasoline purchasedby persons. The proportion is based on Census data[134] on the share of total mileage by age of autoreported by persons as business usage. Total autogasoline purchased by individuals is calculated as totalauto gasoline less that for exclusively business autosand government autos.

For the period 1929 to 1941, total auto gasolinepurchases are allocated 70 percent to PCE, the samepercentage used in the allocation of new autos andgross margin on used autos. For the period of gasolinerationing during World War II, the percentages usedare 65 percent for 1942, 55 for 1943 and 1944, and 60percent for 1945.

Trucks.—Beginning with 1972, estimates of purchasesof truck fuel by persons are derived using BEA stocksof consumer trucks (including vans) and fuel usage pertruck, which is assumed to be 20 percent more than


average usage of consumer autos, based on a 1983 EIAsurvey [181].

From 1946 to 1971, estimates are based on unitstocks and average mileage of trucks for personal use,and MPG assumed to be equal to the FHWA measurefor passenger autos. Benchmark estimates of stocksand average mileage are from the Truck Inventory andUse Survey [123]; for non-benchmark years, unit stocksand average mileage are judgmentally interpolatedand extrapolated.

Airplanes.—Benchmark estimates of purchases of avi-ation fuel by persons are based on EIA data [171] onbarrels used in general aviation and a PCE share basedon Federal Aviation Administration data [211] on theproportion of hours flown accounted for by persons. Fornonbenchmark years beginning in 1946, quantities arefrom FHWA data on gallons of gasoline consumed foraviation, with the share purchased by persons inter-polated between benchmark years and held constantfrom the latest benchmark year.

Motorcycles.—Total gasoline for motorcycles is fromthe FHWA; 95 percent is assumed to be purchased bypersons.

Boats.—Total gasoline purchased for boats is from theFHWA; 50 percent is assumed to be purchased bypersons.

4.4.2 Average retail price

The 1977 benchmark estimate of the average retailprice per gallon of nonaviation gasoline is from an EIAreport [178]. The same price is used for all types of ve-hicles except airplanes. For aviation fuel, the averageretail price is from EIA data [176], adjusted to includetaxes.

After 1977, a single average retail price is used forall vehicles. For 1978 to 1981, this price is the averageprice of nonaviation gasoline, from the same sourceused for the benchmark estimate. Beginning in 1982, aweighted average of price per gallon of leaded gasolineand of unleaded regular and premium are used. Pricesby type and grade are published by EIA [175]. Weightsare based on data from the same source on quanti-ties of leaded and unleaded gasoline supplied and dataon the quantities of regular and premium grades ofunleaded gasoline from information on supplier salespublished in the Petroleum Marketing Annual [179].For the first annual revision, supplier sales are notavailable, and the distribution of unleaded gasoline bygrade is judgmentally trended.

For 1976, the average price is from the “Survey ofGallonage Sales of Gasoline” conducted by the Cen-sus Bureau [143]. For the 1972 benchmark estimate,the average price is the unit value of gasoline salesat gasoline service stations from the 1972 Census ofRetail Trade [119]. For the 1967 benchmark estimate,for other annual estimates from 1958 to 1966, and for

interpolation between the 1967, 1972, and 1976 esti-mates, a weighted average of prices per gallon of leadedregular and premium and of unleaded gasoline fromBLS [198] is used. Weights are based on quantitiesfrom the Ethyl Corporation [37]. Before 1958, the over-all average price, adjusted to include taxes, is from theAmerican Petroleum Institute [13].


Quarterly and monthly estimates of purchases by per-sons are based, when available, on the same productsupplied data from EIA used for the first annual revi-sion estimate and on the same average price data andquantity weights from EIA used for annual estimatesbeginning in 1982. For the advance quarterly estimate,the total product supplied is based on monthly EIAdata [180] for the first month of the quarter and weeklyEIA data [183] for the other two months. Averageretail prices and the distribution of gasoline suppliedbetween leaded and unleaded are from EIA for the firsttwo months of the quarter [175]. For the third month ofthe quarter, prices by type and grade are extrapolatedwith consumer price index components [201], and thedistribution of gasoline by type is based on weekly EIAdata.

4.5 Rental value of owner- andtenant-occupied nonfarm dwellings

As explained in the introductory section, PCE includesboth monetary rents paid by tenants and an imputedrental value for owner-occupied dwellings. PCE forrent on tenant-occupied nonfarm dwellings is meas-ured as net contract rent plus tenants’ unreimbursedexpenditures for major replacements, maintenance,and repairs. Net contract rent is defined as the rentthat tenants contract for or agree to pay—includingany charges for major appliances and furnishings, util-ities, or services—less the charges for utilities. PCEfor owner-occupied nonfarm dwellings is defined as“space rent,” which is an imputed value of the rental ofthe dwelling alone, exclusive of any charges for majorappliances and furnishings, utilities, or services.

Separate estimates are prepared for owner-occupiedpermanent site dwellings, for owner-occupied mobilehomes, for tenant-occupied permanent site dwellings,and for tenant-occupied mobile homes. For each type ofdwelling, the estimate of PCE is essentially the prod-uct of the number of occupied units and an appropriaterent per unit. Benchmark estimates are based primar-ily on data from the decennial census of housing (COH)[115] and the survey of residential finance (SRF) [116],which is conducted in conjunction with the COH. Esti-mates for other years are interpolations between andextrapolations from the benchmark estimates; they arebased primarily on data from the biennial Americanhousing survey (AHS) [112], the current populationsurvey (CPS) [133], and the CPI. (The AHS is available


annually from 1973 through 1981 and biennially sincethat time; the CPS is available annually and, in recentyears, quarterly.) Quarterly and monthly estimatesare based primarily on data from the CPS and CPI andfrom BEA estimates of the stock of mobile homes.

4.5.1 Number of permanent-site units

The benchmark for the number of tenant- and owner-occupied permanent site dwellings is from the COH,adjusted by BEA to reflect the stock as of July 1 andto include certain vacant units. For tenant-occupieddwellings, two types of vacant units are treated asif they were occupied for one-half of the year: Unitsthat have been rented and are awaiting occupancy,and seasonal and migratory units. For owner-occupieddwellings, two types of vacant units are treated as ifthey were occupied for the entire year: Units that havebeen sold and are awaiting occupancy, and units heldoff the market for occasional use. The COH providesseparate estimates for the seasonal and occasional useunits and a combined total for units sold or rentedawaiting occupancy; BEA allocates this total betweenowner- and tenant-occupied units using the propor-tions from the occupied housing stock. Before 1960,COH data are supplemented by intercensal bench-marks based on the 1956 National Housing Inventory[140]; Census Bureau surveys done in conjunction withthe Monthly Report on the Labor Force [136] for 1944,1945, and 1947; and data from the Census Bureau RealProperty Inventory of 1934 [136].

Similar information from the AHS for selected yearsand from the CPS is used to interpolate betweenand extrapolate from the census data. Before 1948,the number of units is interpolated between and ex-trapolated from benchmarks using Current PopulationSurvey data on the number of households [136] (1948–72) and by data on new housing units adjusted toallow for demolitions and conversions (1929–47). Newhousing units are based on data on starts [139], anddemolitions and conversions are from a reconciliationof cumulative starts and the net change in the stockbetween successive benchmark estimates.

4.5.2 Rent on tenant-occupied permanent sitedwellings

For tenant-occupied permanent site dwellings, PCE isthe sum of (1) net contract rent per unit multiplied bythe number of units and (2) tenants’ unreimbursed ex-penditures for major replacements, maintenance, andrepairs.

Net contract rent equals contract rent less paymentsfor utilities—electricity, gas, water, and heating fuel.The benchmark estimate for net contract rent per unitis derived from unpublished data on mean contractrent per unit from the COH; similar information fromthe AHS is used to interpolate between census yearsand to extrapolate from the latest census benchmark.Estimates for years when the AHS is not available

are based on changes in the CPI for residential rent,adjusted judgmentally for changes in the quality of thehousing stock. Payments for utilities are estimated asthe product of average expenditures per room and thenumber of rooms. Average expenditures per room areextrapolated from the 1976 estimate by changes in theCPI for fuel and other utilities; until 1976, averageexpenditures were obtained from Apartment BuildingIncome and Expense Analysis [53], published by theInstitute of Real Estate Management. The averagenumber of rooms per tenant unit and the number oftenant units in two or more unit buildings are from theCOH.

Tenants’ unreimbursed expenditures for major re-placements, maintenance, and repairs—available fromthe survey of residential alterations and repairs (SO-RAR) [125] from the inception of the survey through1963—have been extrapolated from the 1963 SORARestimate by data on maintenance and repair expend-itures from SORAR, from 1964–82, and on data fromthe BLS consumer expenditures survey [200] from1983 forward. (Expenditures from 1973 forward arefor tenant-occupied properties; those for 1964–72 arefor owner-occupied unit properties.)

4.5.3 Space rent on owner-occupied permanentsite dwellings

For owner-occupied permanent site dwellings, PCE isspace rent per unit times the number of units. Bench-mark estimates of space rent per unit are derivedusing (1) the estimated tenant-occupied net contractrent per unit and (2) SRF information on the mar-ket values of owner- and tenant-occupied units and onthe rent-to-market-value ratios of the latter by valueclass. An estimate of owner-occupied net contract rentis calculated by assuming that it is the equivalent ofnet contract rent of tenant-occupied units in the samevalue class.

Estimates of owner-occupied net contract rent foryears between censuses are interpolated and, until1984, extrapolated from the latest census using thechange in rents for tenant-occupied units. Beginningin 1984, the estimates are extrapolated using the CPIfor homeowners’ equivalent rent adjusted judgmen-tally for changes in the quality of the housing stock.Owner-occupied net contract rent, the product of therent per unit and the number of units, is convertedto space rent by subtracting an amount attributableto the use of major appliances and furnishings. Theamount is set equal to the capital consumption al-lowance with capital consumption adjustment on theseappliances and furnishings estimated by BEA as partof its estimates of capital stock [150].

4.5.4 Rent on owner- and tenant-occupied mobilehomes

For mobile homes, the benchmark for the number oftenant- and owner-occupied units is the COH. Inter-


polation between these benchmarks and extrapolationfrom the latest census is based on a perpetual inven-tory calculation, using data on new units from theNational Conference of States on Building Codes andStandards [81].

The average rent per unit for tenant-occupied unitsis also derived from COH. The rent on owner-occupiedmobile homes is assumed to be slightly higher, basedon the average number of rooms in each type of unit.(In 1970 the differential was eleven percent; currently,it is about 7.5 percent.) Interpolations between bench-marks and extrapolation from the latest census yearare based on the CPI for residential rent. Beginningwith 1984, contract rent for owner-occupied units isextrapolated with the CPI for homeowners’ equivalentrent. Contract rent for mobile homes is calculatedas the number of units times average rent. PCE forowner-occupied and for tenant-occupied mobile homesare contract rents less payments for utilities. Bench-mark data on utility payments are based on a 1966survey by the Department of Housing and Urban De-velopment [194]. These payments are extrapolatedbackward and forward, using as indicator the productof the number of units and the CPI for gas and electricutilities.


Quarterly and monthly current-dollar estimates areprepared by reflating constant-dollar PCE for eachtype of dwelling using the CPI for residentialrent (tenant-occupied) or owner’s equivalent rent(owner-occupied).

Monthly constant-dollar estimates are based on thenumber of units for each type of dwelling adjustedfor changes in the quality of new units. Estimatesof the stock of owner-occupied and of tenant-occupieddwellings are derived by interpolating between and ex-trapolating from annual estimates from the AHS orCPS, whichever is available. (In recent years theseCPS data also have been available quarterly, usuallyin time for the final quarterly estimate.) The totalstock of mobile homes is estimated by interpolatingbetween and extrapolating from annual estimates, us-ing as an indicator monthly data on shipments ofmobile homes [81]. The distribution of the mobilehome stock between owner- and tenant-occupied unitsis based on recent trends. Stocks of permanent sitedwellings are estimated as the difference between thetotal stocks of owner- and tenant-occupied dwellingsand the estimates of mobile homes. The units es-timates are adjusted for changes in the quality ofnew units based on historical relationships betweenaverage rental values and the respective CPI’s.

For the advance quarterly estimate the estimates ofunits are extrapolated judgmentally, and the CPI’s areavailable for all 3 months. For the preliminary esti-mate, 1 month of data on manufacturers’ shipmentsof mobile homes becomes available. For the final es-

timate, the quarterly CPS data on units and anothermonth of shipments of mobile homes become available.

4.6 Brokerage charges and investmentcounseling

This PCE category consists of (1) commissions paidby persons on securities transactions, (2) commis-sions paid by persons on commodities transactions,(3) investment counseling fees paid by persons, (4)services furnished without payment by domestic secu-rities dealers, and (5) sales charges paid by persons onpurchases of investment company securities.

4.6.1 Annual estimates

Benchmark and nonbenchmark annual estimates forall five categories are based primarily on consoli-dated income statements for securities broker-dealers.For all estimates except the first July revision, theseincome statements are from the Financial and Oper-ational Combined Uniform Single (FOCUS) Report,which all security broker-dealers are required to filewith the Securities and Exchange Commission (SEC).SEC tabulations of broker-dealer income statementsare published in its Annual Report [229]. For the firstJuly revision, the identical revenue items obtainedfrom consolidated income statements of New YorkStock Exchange (NYSE) member firms [93], which con-stitute a major portion of all broker-dealers, are usedto extrapolate the detailed SEC data. Except as noted,all data used in the estimates are from these sources.

Broker-dealer revenues for each of the five cate-gories, from the FOCUS or NYSE reports, are adjustedto exclude interest and revenues derived only frombusiness and government (such as underwriting). Theremainder is allocated between the public—that is,persons and institutions—and other broker-dealersusing independent information from a variety ofsources.

Securities commissions.—For the benchmark esti-mate of securities commissions, revenues from theFOCUS reports are first adjusted to exclude floorbrokerage—that is, commissions paid by broker-dealers to other broker-dealers—based on data oncommissions paid from the FOCUS reports. The re-mainder, commissions charged the public, is allocatedbetween persons and institutions using data on the dis-tribution of the value of stock transactions accountedfor by each and data on their relative commission rates.

• The distribution of the value of stock transactionsis based on annual data on the total value of stocktransactions from the SEC [230] and the NYSE[94] for registered exchanges, and from the Na-tional Association of Securities Dealers (NASD)[72] for over-the-counter markets. The value ofpublic trading is estimated by excluding membertrading for their own account, based on the 1976


and 1980 versions of the NYSE Public Transac-tions Study (PTS) [95]. The percentage of the valueof public trading accounted for by persons also isbased on PTS data.

• The ratio of individual to institutional commissionrates is based on the SEC Survey of CommissionRates [231], with commission rate levels adjustedso that commissions paid by individuals and byinstitutions sum to the SEC’s total commissions.

For nonbenchmark years, the percentage of the valueof stock transactions accounted for by members is in-terpolated and extrapolated by members’ percentageof total shares traded on the NYSE [93]. For publictransactions on the NYSE, the percentage of the valueof trading accounted for by persons is extrapolated byan estimate of the percentage of publicly traded sharesaccounted for by individuals. This estimate is based onNYSE data on the distribution of trading by order-sizeclass [93] and on unpublished data for 1975 to 1981,from the SEC’s commission rate survey, on the dis-tribution of publicly traded shares, by order-size classbetween individuals and institutions. For public trans-actions on markets other than the NYSE, the shareof the value of public transactions accounted for bypersons is held constant at the level determined fromPTS. The ratio of individual to institutional commis-sion rates is based on the last SEC commission ratesurvey.

Commodities commissions.—Benchmark estimatesof commissions paid by individuals on commoditiestransactions are based on commodities revenues of se-curity and commodity broker-dealers. Commoditiesrevenues of security broker-dealers are from the FO-CUS reports. Commodities revenues of commoditybroker-dealers are derived as the product of BLS wagesand salaries for the industry [203] and the ratio of rev-enues to wages and salaries of security broker-dealers[229]. For nonbenchmark years, commodities revenuesare extrapolated by security broker-dealer revenues.

Allocation to persons is based on a judgmental esti-mate for 1977 of their share of trading; the share isheld constant for nonbenchmark years.

Investment counseling fees.—Investment counsel-ing services are provided by security broker-dealers,commercial banks, and independent investment coun-selors. Allocation to persons is based on the marketvalue of holdings of equity securities not accounted forby institutional investors, from the 1979 SEC AnnualReport [229].

Benchmark estimates of total fees charged at secu-rity broker-dealers are based on FOCUS report data.Benchmark estimates of fees charged at commercialbanks are based on Federal Deposit Insurance Corpo-ration data [44] on income from fiduciary activities,allocated to personal agency accounts on the basisof data from FRBNY Functional Cost Analysis [47].Benchmark estimates of fees of independent invest-

ment counselors are based on the wages and salariesof services allied with the exchange of securities orcommodities and the ratio of revenues to wages andsalaries of security broker-dealers, adjusted to excludeservices other than investment counseling [229].

For nonbenchmark years, investment counselingfees at broker-dealers and commercial banks are ex-trapolated by FDIC data on the fiduciary income ofcommercial banks [44]. Fees at independent coun-selors are extrapolated with wages and salaries fromBLS [203].

Services furnished without payment by domestic se-curities dealers.—As explained in the introduction,the PCE category brokerage charges and investmentcounseling includes an implicit charge equal in amountto the price spread on the transactions carried outby securities dealers with persons. Dealers whomake markets in securities do not charge commissions;instead, they retain as compensation the income re-sulting from acquiring securities at a price lower thanthe price at which the securities subsequently are soldto their customers.

Services furnished without payment by domesticsecurities dealers on transactions involving personsare estimated in three parts: Income from over-the-counter market making, income from trading debtsecurities, and income from trading other securities.All data on services furnished without payment bydomestic securities dealers come from the FOCUSreport.

The income of over-the-counter market makers, whomaintain inventories of stocks and provide market liq-uidity, is allocated to persons on the basis of data fromthe Mutual Fund Fact Book.

Income from trading debt securities reported in theFOCUS report first is adjusted to exclude interestearnings on the basis of interest income reported in theSEC The Securities Industry in 1980 [232] and infor-mation from 10-K reports and from annual reports toshareholders of large publicly traded broker-dealers;the proportion of income accounted for by interest hasbeen held constant since 1980. The adjusted incomefrom trading debt securities is allocated among Fed-eral, State and local, and corporate securities on thebasis of broker-dealer holdings of these securities fromthe FOCUS report.

• Income from trading Federal securities is ad-justed to include that received by commercialbanks, based on New York Federal Reserve Bank(FRBNY) data [48], and allocated to persons onthe basis of marketable Federal securities held byindividuals, based on Treasury Department data[209].

• Income from trading State and local debt securitiesis allocated to persons in the same proportion asFederal securities.

• Income from trading corporate debt securities isallocated to persons on the basis of the percentage


of corporate and foreign bonds held by householdsfrom the Federal Reserve Board’s flow of fundsdata [25].

Income from trading other securities reported inthe FOCUS report is allocated to persons on the ba-sis of data on individuals’ share of the value of stocktransactions used in allocating securities commissions.

In nonbenchmark years, services furnished withoutpayment by domestic securities dealers for State andlocal, corporate debt, and other securities on behalfof persons are extrapolated from benchmark levels bydata on income from trading these securities.

Sales charges on purchases of investment companysecurities.—Benchmark estimates of sales charges onpurchases of investment company securities are allo-cated to persons on the basis of the share of marketvalue accounted for by noninstitutional and fiduciarymutual fund holdings, using data from the MutualFund Fact Book [63]. For nonbenchmark years, theshare is held constant.


Quarterly and monthly estimates of brokerage chargesand investment counseling are prepared in three steps:(1) Commissions on securities transactions, (2) incomefrom the sale of investment company securities, and(3) all other commissions and fees, including serv-ices furnished without payment by domestic securitiesdealers.

Commissions paid by persons on securities trans-actions are estimated as the product of total sharestraded and a cents-per-share commission rate.

• Shares traded by persons are based on data onthe volume of securities transactions from theNYSE [94], the National Association of Securi-ties Dealers (NASD) [73], and the Securities andExchange Commission (SEC) [230]. Estimatesof shares traded by persons are prepared sepa-rately for the NYSE and other markets, whichinclude other registered exchanges and the over-the-counter market. For the NYSE, publiclytraded shares are estimated as total shares tradedless shares accounted for by members; the share ofpersons is derived from (1) information on sharestraded by order-size class and from (2) data fromthe last SEC commission rate survey [231] on theshares of persons and institutions in each sizeclass. For other markets, shares traded by personsare assumed to be a constant proportion of totalshares traded, based on PTS.

• Estimates of the commission rate, expressed ascents per share, are derived from the latest Julyrevision estimate as an implicit rate, which is cal-culated from the estimates of shares traded bypersons and commission revenue accounted for bypersons on securities transactions. During the

year, the estimated commission rate may be ad-justed when there are large changes in the numberof shares per order accounted for by individualsto reflect the fact that larger orders generally costless per share than smaller ones. For the NYSE,the number of orders placed by persons is esti-mated in the same way as the number of sharestraded by persons. Shares per order for persons onother markets are estimated as a constant propor-tion of NYSE shares per order, based on the lastPTS.

Income from the sale of investment company se-curities is extrapolated by the sales of the shares ofopen-end investment companies reported by the In-vestment Company Institute and published by theFederal Reserve Board [24].

All other commissions and fees, including serv-ices furnished without payment by domestic securitiesdealers, are extrapolated judgmentally.

For the advance quarterly estimate, 3 months ofsource data are available except for SEC data on trans-actions on exchanges other than the NYSE, for which1 month of data is available. For the preliminary andfinal estimates, additional detail on shares traded byorder size class from the NYSE becomes available. Forthe initial monthly estimate, NYSE and NASD dataare available.

4.7 Services furnished without paymentby financial intermediaries exceptlife insurance carriers and privatenoninsured pension plans

As explained in the introduction, purchases of serv-ices include an imputation for the value of servicesfurnished to persons without payment by depositoryinstitutions—that is, financial intermediaries exceptlife insurance carriers and private noninsured pensionplans. Estimates of this service charge are preparedas part of the estimates of imputed interest and asso-ciated service charges. The service charge is equal toimputed interest received by persons from commercialbanks and savings and loan associations and all theimputed interest paid by mutual savings banks, creditunions, and investment companies.

For commercial banks, total imputed interest paidequals the sum of monetary interest received on de-posits plus other property income and imputed interestreceived from Federal Reserve Banks less monetaryinterest paid on deposits. For mutual savings banks,savings and loan associations, and credit unions, im-puted interest equals monetary interest received ondeposits plus other property income less monetaryinterest paid on deposits less profits before tax ofmutual institutions. For investment companies, im-puted interest equals monetary interest received plusdividends received less the interest share of cash dis-tributions less dividends paid other than capital gains


distributions. The personal share of commercial bankand savings and loan association imputed interest isequal to the share of deposits of these institutions heldby persons; imputed interest paid by mutual savingsbanks, credit unions, and investment companies is allallocated to persons.

Estimates for all years are derived from data oninterest paid and received from the FDIC [44], the Fed-eral Home Loan Bank Board [45], and the NationalCredit Union Administration [85]. Profits of mutualinstitutions and interest and dividend income of in-vestment companies are from tabulations of corporateincome tax returns or other sources as described in MP-2 [147, pages 10, 19, 38–39]. The share of commercialbank deposits attributable to persons is based on datafrom the FDIC [44] and the Federal Reserve Board’sdemand deposit ownership survey [24]. The share ofsavings and loan association deposits attributable topersons is based on a 1968 study by the Federal HomeLoan Bank Board on ownership of deposits [46].

Quarterly and monthly estimates are preparedjudgmentally.

4.8 Expense of handling life insurance

This PCE category consists of current expenses of pri-vate businesses that underwrite life insurance andadminister pension plans. As explained in the intro-duction, current expenses are used to measure PCEfor this category because it allows the savings at-tributable to life insurance and private pension plansto be treated as personal saving rather than as busi-ness saving. Current expenses are defined to consist ofall expenses related to life insurance and pension ac-tivities, including investment expenses, but to excludeexpenses related to real estate activities. Current ex-penses are defined to include imputed interest receivedfrom commercial banks and the profits of stock lifeinsurance companies.

Life insurance is underwritten by three types ofbusinesses: Legal reserve companies, including bothstock and mutual companies; fraternal benefit soci-eties; and mutual savings banks. All types of policiesare included: Whole life, term, endowment and re-tirement income, group life, industrial life, and creditlife insurance. Private pension plans include all typesof annuity arrangements administered by life insur-ance companies and pension plans administered byother institutions (that is, private noninsured pensionplans), so long as the plan is subject to the provisionsof the Employees Retirement Security Act of 1974,which covers virtually all types of plans provided byemployers.

4.8.1 Legal reserve companies

For domestic legal reserve companies, the bench-mark and nonbenchmark annual estimates, except forthe first July revision estimates, are based on tab-

ulations prepared by the American Council of LifeInsurance (ACLI) from annual statements filed bythe companies with State insurance departments [3].The following items related to the life insurance andpension business in these tabulations are consideredcurrent expenses: Commissions paid on premiums andannuity considerations; general insurance expenses;investment expenses; insurance taxes, licenses, andfees; and other miscellaneous expenses. Commis-sions paid on premiums and annuity considerations,which measure only commissions on direct insur-ance business, are adjusted to a measure of total netcommissions paid by including commissions paid onreinsurance assumed and by subtracting commissionsreceived on reinsurance ceded.

Because the annual statements of domestic com-panies consolidate their activities world-wide, theexpenses of their operations in foreign countries mustbe subtracted. In addition, the expenses of foreign lifeinsurance companies operating in the United Statesmust be added. The expenses of domestic companiesabroad are estimated using the relationship betweendomestic premium receipts and total premium receiptsfrom the ACLI’s Life Insurance Fact Book (LIFB) [4].The expenses of foreign companies in the United Statesare estimated as the product of current expenses ofU.S. companies chargeable to U.S. residents and theratio, based on data from the LIFB, of premiums paidby U.S. residents to Canadian companies operatingin the United States to premiums paid by U.S. resi-dents to U.S. companies. For nonbenchmark years, thenet of these geographic adjustments is extrapolated bycurrent expenses of domestic legal reserve companies.

The last steps in estimating current expenses are (1)to add imputed interest paid by commercial banks tolife insurance companies and (2) to add the profits ofstock life insurance companies. The estimates of im-puted interest paid by commercial banks are describedin the previous section; as described in [147], the esti-mates of profits are based on tabulations of corporatetax returns from IRS [206] for the third July revisionor on data from ACLI for the second July revision.

4.8.2 Fraternal benefit societies and mutualsavings banks

For fraternal benefit societies and mutual savingsbanks, data on current expenses are not available.Consequently, PCE is estimated as premiums less ben-efits and dividends paid to members and beneficiaries.For the benefit societies, estimates are based on datafrom the National Fraternal Congress of America [87].For mutual savings banks, estimates are based on datafrom the LIFB [4]. Both of these sources are availableannually, except for the most recent year.


4.8.3 Private noninsured pension plans

For private noninsured pension plans, the estimateof expenses for the last benchmark year (1977) arebased on IRS tabulations of reports for pension plans[205]. (Previous benchmark and annual estimateswere based on SEC data on noninsured plans [230],since discontinued.) Annual estimates except for themost recent year are prepared separately for commis-sions and for all other expenses. Commissions areestimated as the product of the value of stock transac-tions by these plans and an institutional commissionrate. Stock transactions is the sum of purchases andsales from the IRS report, extrapolated by the prod-uct of pension fund assets, from LIFB, and the NYSEturnover rate [93]. The institutional commission rateis that derived in preparing the PCE estimate for bro-kerage and investment counseling described above. Allother expenses of private pension plans are extrap-olated by the wages and salaries of private pension,health and welfare funds from BLS [203].

For the most recent year, expenses for both life in-surance and private noninsured pension plans areextrapolated by BLS tabulations of the wages andsalaries of life insurance carriers [203].

For the extrapolation of current quarterly andmonthly estimates, and the interpolation of histori-cal estimates, wages and salaries of insurance carriersare estimated based on BLS employment and earnings

[202]. For the advance quarterly estimate, 3 monthsof BLS data are available.

4.9 Religious activities

Religious activities, like those of other nonprofit in-stitutions, are measured in PCE by their currentoperating expenses, as explained in the introduction.For all years except the first July revision estimates,the annual estimates are the sum of a number of sepa-rate calculations for each denomination. Expenses arederived by multiplying per capita contributions to con-gregational expenses, taken from the National Councilof Churches (NCC) Church Financial Statistics [83],by membership estimates based on data from the samesource, from the Yearbook of American and CanadianChurches [84], and from the Official Catholic Direc-tory [64]. These expense estimates are reduced by thevalue of construction of religious buildings, based onCensus Bureau data [154].

For the first July revision estimate, expenses areextrapolated by the product of Census Bureau data onthe U.S. adult resident population [135] and the mostrecent BEA estimate of per capita disposable personalincome.

Quarterly and monthly estimates are interpolationsand extrapolations of the annual estimates, using asindicator wages and salaries derived from monthlyBLS data on employees and earnings of employees ofreligious institutions [202].

Constant-Dollar Estimates

The constant-dollar estimates of PCE are preparedat a detailed level using one of three methods. Themethod used for most of the PCE categories is defla-tion, that is, constant-dollar estimates are obtained bydividing current-dollar estimates, at the most detailedcategory level, by appropriate price indexes with thebase period—at present, the year 1982—equal to 100.The other methods, direct base-year valuation andquantity extrapolation, are similar in that they bothuse quantity indicators. In direct base-year valuation,constant-dollar estimates are obtained by multiplyingbase-period prices by actual quantity data for the otherperiods. In quantity extrapolation, constant-dollar es-timates are obtained by extrapolating the base-periodcurrent-dollar estimates in both directions from thebase period (1982) by quantity indicators. For eachPCE category shown in NIPA table 2.5, table 15 indi-cates which of the three methods is used and providesinformation about the source data for the quantity in-dicators and price indexes. Additional information onthese sources is provided later in this section.

Table 15 also identifies which estimates are cal-culated monthly, quarterly, or annually and whichprice indexes or quantity indicators are seasonallyadjusted. For virtually all PCE categories, constant-dollar estimates are prepared on a monthly basis; thequarterly and annual estimates are the sum of themonthly estimates. For categories for which monthlysource data are not available, the constant-dollar es-timates are prepared on a quarterly basis, if possible:Annual estimates are summed from the quarterly es-timates, monthly historical estimates are derived byinterpolation, and current monthly estimates are de-rived judgmentally. For PCE categories for which onlyannual data are available, the constant-dollar esti-mates are prepared on an annual basis. Quarterly andmonthly historical estimates are derived by interpola-tion, and quarterly and monthly current estimates arederived judgmentally.

1 Source Data Used for Deflation

1.1 Price indexes from other sources

The price indexes used to deflate the detailed PCE cat-egories are almost all CPI’s. CPI’s, published monthlyby BLS in the CPI Detailed Report [201], are based onactual transaction prices for the items in a fixed, orconstant, “market basket” of consumer goods and serv-

ices purchased by the population of urban consumers.Price indexes from other sources include the BLS Pro-ducer Price Index (PPI) for apparel [204]; the HealthCare Financing Administration input price indexes fornonprofit hospitals (quarterly and annually) and fornursing homes (annually) [188]; a trade source inputprice index for higher education (annually) [101]; anda composite index of CPI’s for industrialized countriesfrom the International Monetary Fund [56].

1.2 Indexes prepared by BEA

Indexes are prepared by BEA for several PCEcategories.

1.2.1 Computers and peripheral equipment

This category is deflated with a quarterly price indexfor personal computers [146, 148].

1.2.2 Food produced and consumed on farms

This category is deflated with a composite index pre-pared from monthly U.S. Department of Agriculture(USDA) prices received by farmers [109].

1.2.3 Tolls

This category is deflated with an index prepared annu-ally from information on toll rates and weights basedon historical toll revenue traffic information, bothfrom the International Bridge, Tunnel, and TurnpikeAssociation [54, 55].

1.2.4 Airline transportation

This category is deflated with an index prepared fromthe CPI for airlines and Department of Transportation[216, 217] and trade association data [2] on revenueper passenger mile.

1.2.5 Foreign travel by U.S. residents

This category is deflated with a composite index of an-nual CPI’s for selected countries adjusted for exchangerates and combined with travel expenditure weights.

Finally, some categories are deflated with compos-ite indexes of input prices prepared from price andcost indexes. The latter are primarily indexes of av-erage earnings, by industry: Monthly estimates are

67


Table 15.—Constant-Dollar PCE: Sources of Estimates—Continued

Linein

NIPAtable2.5

Category Price index for deflation, quantity extrapolator ofbase-year value, or directly valued quantity

1982 weight(percent)

86 Higher education ............................................................ Composite index of input prices from trade source (A);monthly, BEA composite index of input prices.

.588

87 Elementary and secondary schools ............................... BEA composite index of input prices ............................ .52988 Other private education and research ........................... BEA composite index of input prices* .......................... .47389 Religious and welfare activities ...................................... BEA composite index of input prices ............................ 2.16490 Net foreign travel:

Foreign travel by U.S. residents ................................. BEA composite index of CPI’s for selected foreigncountries (A).

.796

Less: Expenditures in the U.S. by foreigners ............. BEA composite index of U.S. CPI components ........... .666

n.e.c. Not elsewhere classified.

1. Current-dollar estimates of these categories are available annually only.

Note.—Except for series marked with an asterisk, all source data are sea-

sonally adjusted. Unless otherwise noted, all source data are availablemonthly. Source data marked ‘‘(A)’’ or ‘‘(Q)’’ are available only annually orquarterly, respectively; the corresponding monthly estimates for these se-ries, unless otherwise noted, are derived by interpolation or, for current peri-ods, judgmentally.

from BLS Employment and Earnings [202], and an-nual estimates are from BLS Employment and Wages[203]. The weights used to combine these price andcost indexes are derived from the most recent censusof service industries, except for the life insurance cat-egory, for which the weights are derived from the mostrecent benchmark I-O accounts.

1.2.6 Hospitals

This category—both nonprofit and for-profit hospitals—is deflated monthly with a composite index preparedfrom average earnings and the CPI for all items.

1.2.7 Nursing homes

This category is deflated monthly with a compositeindex prepared from average earnings, an index offringe benefits per hospital employee from the Ameri-can Hospital Association [9], and the PPI for industrialcommodities less fuel and related products and power.

1.2.8 Expense of handling life insurance

This category is deflated annually with a compositeindex prepared from average earnings, the PPI forelectric power, and 4 CPI’s.

1.2.9 Other

The categories trade union expenses, professional as-sociation expenses, clubs and fraternal organizationsexcept insurance, higher education (monthly), elemen-tary and secondary schools, other private education and

research, and religious and welfare activities are de-flated with composite indexes prepared from averageearnings and the PPI for industrial commodities lessfuel and related products and power.

2 Source Data for QuantityIndicators

2.1 Direct base-year valuation

The direct base-year valuation method is used for boththe dealers’ margin and net transactions componentsof the estimates of net purchases of used autos.

The constant-dollar margin is estimated as unitsales of used autos multiplied by the base-period aver-age margin per used auto sold. Unit sales of franchisedcar dealers are from the Motor Vehicle ManufacturersAssociation [68]; unit sales of used car dealers are es-timated using Census Bureau retail sales of used cardealers [128] and an average value based on the aver-age auction price of used cars reported in AutomotiveNews [32].

For the annual estimates, net transactions in usedautos is the constant-dollar value of the net change inthe stock of used autos held by consumers, estimatedas a residual after accounting for changes in used au-tos held by business and government, and accountingfor exports, imports, and changes in dealer invento-ries. The net change is derived using unit sales fromtrade sources and BEA constant-dollar stocks with de-preciated original value in base-year prices. For thequarterly estimates, net transactions is measured in

Constant-Dollar Estimates 73

current dollars and deflated using the CPI for usedcars. (Further details on the derivation of the esti-mates of net transactions in used cars can be found atthe end of the current-dollar section.)

2.2 Quantity extrapolation

The quantity extrapolation method is used for the es-timates of rental value of farm dwellings, services fur-nished without payment by financial intermediariesexcept life insurance carriers and private noninsuredpension plans, stock brokerage charges, insuranceother than life insurance, and parimutuel net receipts.The source data used for the quantity indicators forthese categories are described below.

2.2.1 Rental value of farm housing

The indicator series for this category is the constant-dollar net stock of farm housing, based on USDAdata [108]. This series is estimated annually by BEAby adding to the previous year’s stock in base-periodprices net investment in farm housing—that is, capitalexpenditures net of depreciation—deflated by the farmdwelling construction cost index.

2.2.2 Services furnished without payment byfinancial intermediaries except lifeinsurance carriers and private noninsuredpension plans

The indicator series for this category is employee hoursof relevant financial institutions. This series is de-rived annually by BEA from Bureau of Labor Statistics(BLS) tabulations of average weekly hours worked andof employment and wages reported by employers cov-ered by State unemployment insurance programs [202,203].

2.2.3 Stock brokerage charges

The indicator series for brokerage charges is the num-ber of orders placed by individuals, based on data from

the New York Stock Exchange [94], the Securities andExchange Commission (SEC) [230], and the NationalAssociation of Securities Dealers [73]. The use of thesesource data in estimating orders is presented in thecurrent-dollar section.

2.2.4 Other insurance

The indicator series for the insurance categories otherthan life insurance is either deflated premiums or de-flated benefits. Current-dollar PCE is measured aspremiums earned less losses incurred and dividends topolicyholders for commercial insurers, and as operat-ing expenses for nonprofit insurers. (The source datafor these estimates are described in the current-dollarsection.)

For household personal property and auto insur-ance, the Consumer Price Index (CPI) for tenants’and auto insurance, respectively, are used to deflatepremiums.

For health insurance, benefits are deflated in threeparts: Medical and hospitalization insurance, incomeloss insurance, and private workers’ compensation.For medical and hospitalization insurance, a weightedaverage of the CPI for hospital and other medical serv-ices and the CPI for physicians’ services is used. Forincome loss insurance, losses are deflated by the CPIfor all items for urban workers. For private workers’compensation, a weighted average of the deflators formedical and hospitalization insurance and for incomeloss insurance is used annually; monthly the CPI forall items is used.

2.2.5 Parimutuel net receipts

The indicator series for parimutuel net receipts iswinnings deflated by the CPI for all items. Winnings—that is, the amount wagered less the amount retainedby tracks and State governments—is based on datafrom the National Association of State Racing Commis-sioners [74] that are used to estimate current-dollarPCE.

SourcesThis is a list of publicly available information used in preparing the estimates of personal consumption

expenditures. Whenever possible, a specific portion of a larger publication is cited—a chapter, a series, or atable number/title. In some cases, the information used is more detailed than that available in the listed source,which is the publication most accessible to the public.

1. Air Transport Association of America. “Air-line Passenger Traffic.” Washington, DC: ATA,monthly.

2. . “Monthly Discount and YieldReport.” Washington, DC, monthly, unpublished.

3. American Council of Life Insurance. “AnnualStatement of the Condition and Affairs of theU.S. Legal Reserve Life Insurance Companies.”Washington, DC: ACLI, annually.

4. . Life Insurance Fact Book.Washington, DC: ACLI, annually.

5. American Express Company. Annual Report. NewYork, NY: American Express, annually.

6. American Gas Association. Gas Facts. Arlington,VA: AGA, annually.

7. . “Monthly Gas Utility StatisticalReport.” Arlington, VA: AGA, monthly.

8. . “Quarterly Report of Gas IndustryOperations.” Arlington, VA: AGA, quarterly.

9. American Hospital Association. HospitalStatistics. Chicago, IL: AHA, annually.

10. . “National Hospital Panel SurveyReport.” Chicago, IL: AHA, monthly.

11. American Kennel Club. Pure-Bred Dogs—American Kennel Gazette. New York, NY: AKC,monthly.

12. American League of Professional Baseball Clubs.American League Red Book. New York, NY: AL,annually.

13. American Petroleum Institute. AmericanPetroleum News. Washington, DC: API, monthly.

14. American Public Transit Association. Transit FactBook. Washington, DC: APTA, annually.

15. . “Transit Ridership Report.”Washington, DC: APTA, quarterly.

16. American Telephone and Telegraph Company.Personal consumption expenditures for telephoneservice. New York, NY, annually, unpublished.(Discontinued.)

17. . Total domestic telephone revenue.New Brunswick, NJ, 1978, unpublished.

18. Automobile Invoice Service. New Car Cost Guide.San Jose, CA: H.M. Gousha, annually.

19. Automobile Manufacturers Association. (NowMotor Vehicle Manufacturers Association.) “Au-

tomobile Facts and Figures.” Detroit, MI: AMA,annually.

20. Best, A.M., Company, Inc. Aggregates and Aver-ages: Property-Casualty. Oldwick, NJ: A.M. Best,annually.

21. . “Property-Casualty Quarterly By-Line Report.” Oldwick, NJ: A.M. Best, quarterly.

22. Billboard Communications, Inc. AmusementBusiness. New York, NY: BCI, weekly.

23. Board of Governors of the Federal Reserve Sys-tem. Department Store Trade, United States.Washington, DC: Board of Governors, monthly.(Discontinued.)

24. . Federal Reserve Bulletin. Washing-ton, DC: Board of Governors, monthly.

25. . Flow of Funds Accounts. (Sta-tistical Release Z.1.) Washington, DC: Board ofGovernors, quarterly.

26. . Sales Finance Companies. Wash-ington, DC: Board of Governors, monthly.(Discontinued.)

27. . Survey of Consumer Finances.Washington, DC: Board of Governors, annually.(Discontinued, but now published occasionally in[24].)

28. Bobit Publications. Automotive Fleet Fact Book.Redondo Beach, CA: Bobit Publications, annually.

29. Cigar Association of America, Inc. “MonthlyStatistical Bulletin.” Washington, DC: CAA,monthly.

30. Corporation for Public Broadcasting. “Expendi-tures of Public Radio Stations.” Washington, DC,annually, unpublished.

31. . “Expenditures of Public Tele-vision Licensees.” Washington, DC, annually,unpublished.

32. Crain Communications, Inc. Automotive News.Detroit, MI: CCI, weekly.

33. Edison Electric Institute. Statistical Yearbook.Washington, DC: EEI, annually.

34. Election Research Center. America Votes.Washington, DC: ERC, biennially.

35. Electronic Industries Association. “Color Tele-vision Activity.” Washington, DC: EIA,monthly.

75


36. . “Monochrome Television Activity.”Washington, DC: EIA, monthly.

37. Ethyl Corporation. “Yearly Report of GasolineSales.” Houston: Ethyl Corp., annually.

38. Fairfield Research, Inc. Home Video: State of theIndustry Report. Lincoln, NE, annually.

39. Federal Communications Commission. “CableTelevision Industry Revenues.” Washington, DC,annually. (Discontinued after 1981 report.)

40. . “Interexchange Carriers, Total Op-erating Revenue.” Washington, DC, annually,unpublished.

41. . “Report of Telephone Company Rev-enues and Expenses.” Washington, DC, monthly,unpublished.

42. . Statistics of Communications Com-mon Carriers. Washington, DC: U.S. GovernmentPrinting Office, annually.

43. Federal Deposit Insurance Corporation. AnnualReport. Washington, DC: FDIC, annually.

44. . Statistics on Banking. Washington,DC: FDIC, annually.

45. Federal Home Loan Bank Board. Combined Fi-nancial Statements. Washington, DC: FHLBB,annually.

46. . “Mid-1968 Characteristics of Asso-ciation Accounts.” Federal Home Loan Bank BoardJournal. Washington, DC: FHLBB, February1969.

47. Federal Reserve Bank of New York. FunctionalCost Analysis. New York, NY: FRBNY, annually.

48. . “Gross Dealer Transactions in U.S.Government and Federal Agency Securities.” NewYork, NY, monthly, unpublished.

49. Greyhound Lines, Inc. Intercity passengerrevenue. Des Moines, IA, monthly, unpublished.

50. Health Insurance Association of America. SourceBook of Health Insurance Data. Washington, DC:HIAA, annually.

51. Hunter Publishing Co., Inc. National PetroleumNews Factbook Issue. Des Plaines, IL: HPC,annually.

52. IMS America Ltd. National Prescription Audit.Ambler, PA: IMS, annually.

53. Institute of Real Estate Management. ApartmentBuilding Income and Expense Analysis. Chicago,IL: IREM, annually.

54. International Bridge, Tunnel & Turnpike Associ-ation. Toll Rates Survey: United States & CanadaBridges and Tunnels. Washington, DC: IBTTA,biennially.

55. . Toll Rates Survey: United StatesRoads. Washington, DC: IBTTA, biennially.

56. International Monetary Fund. InternationalFinancial Statistics. Washington, DC: IMF,monthly.

57. Interstate Commerce Commission. Annual re-ports of Class I line-haul railroads. Washington,DC, annually, unpublished.

58. . Annual reports of Class I motor car-riers of passengers. Washington, DC, annually,unpublished.

59. . Freight Commodity Statistics:Class I Railroads. Washington, DC: U.S.Government Printing Office, annually.

60. . “Large Class I Household GoodsCarriers Selected Earnings Data.” Washington,DC: ICC, quarterly.

61. . Transport Statistics in the UnitedStates: Motor Carriers Part 2. Washington, DC:ICC, annually.

62. . Transport Statistics in the UnitedStates: Railroad Companies and Motor Car-riers Subject to the Interstate Commerce Act.Washington, DC: ICC, annually.

63. Investment Company Institute. Mutual Fund FactBook. Washington, DC: ICI, annually.

64. Kennedy, P.J. & Sons. The Official CatholicDirectory. Wilmette, IL: PJK, annually.

65. Laventhol & Horwath. “National Trends of Busi-ness in the Lodging Industry.” Philadelphia, PA:L&H, monthly.

66. McGraw-Hill Publishing Company, Inc. “MonthlyPrices: Handy and Harman Gold Price.” MetalsWeek. New York, NY: McGraw-Hill, weekly.

67. Moody’s Investors Service. Moody’s Public UtilityNews Reports. New York, NY: MIS, twice-weekly.

68. Motor Vehicle Manufacturers Association of theU.S., Inc. “Passenger Car Sales in the U.S.Reported by U.S. Manufacturers.” Detroit, MI:MVMA, monthly.

69. . “U.S. Retail Sales of Trucks byGVWR and Manufacturer.” Detroit, MI: MVMA,monthly.

70. National Association of Fleet Administrators.“Fleet Operating Expense Survey.” NAFA FleetExecutive. Iselin, NJ: NAFA, annually.

71. National Association of Professional BaseballLeagues. The Orange Book. St. Petersburg, FL:NAPBL, annually.

72. National Association of Securities Dealers, Inc.Fact Book. Washington, DC: NASD, annually.

73. . “NASDAQ Daily Closing Indexand Volume Report.” Washington, DC: NASD,monthly, unpublished.

74. National Association of State Racing Commission-ers. Pari-Mutuel Racing. Lexington, KY: NASRC,annually.

75. National Association of Water Companies. Finan-cial Summary for Investor-Owned Water Utilities.Washington, DC: NAWC, annually.

Sources 77

76. National Automobile Dealers Association. NADAData. McLean, VA: NADA, annually.

77. . NADA Official Used Car Guide.McLean, VA: NADA, monthly.

78. . Operating Averages for the Automo-bile Retailing Industry. Washington, DC: NADA,quarterly. (Discontinued.)

79. National Cable Television Association. “Ca-ble Television Developments.” Washington, DC:NCTA, several times a year.

80. National Catholic Education Association. U.S.Catholic Elementary and Secondary Schools: En-rollment and Staffing. Washington, DC: NCEA,annually.

81. National Conference of States on Building Codesand Standards, Inc. “Mobile Home ProgramStatistics.” Herndon, VA: NCSBCS, monthly.

82. National Cotton Council of America. CottonCounts Its Customers. Memphis, TN: NCCA,annually.

83. National Council of Churches of Christ in the USA.Church Financial Statistics. New York, NY: NCC,annually.

84. . Yearbook of American andCanadian Churches. New York, NY: NCC,annually.

85. National Credit Union Administration. YearendStatistics for Federally Insured Credit Unions.Washington, DC: NCUA, annually.

86. National Football League. Net gate receipts forprofessional football. New York, NY, annually,unpublished.

87. National Fraternal Congress of America. Statis-tics of Fraternal Benefit Societies. Chicago, IL:NFCA, annually.

88. National League of Professional Baseball Clubs.National League Green Book. New York, NY: NL,annually.

89. National Railroad Passenger Corporation. “Pas-senger Accounting System—Systems EarningsSummary.” Washington, DC, monthly, unpub-lished.

90. Nevada State Gaming Control Board. “GamingRevenue Analysis.” Carson City, NV: NSGCB,monthly.

91. . Nevada Gaming Abstract. CarsonCity, NV: NSGCB, annually.

92. New Jersey Casino Control Commission. Pressrelease on New Jersey casino industry ‘win’.Trenton, NJ: NJCCC, monthly.

93. New York Stock Exchange. Fact Book. New York,NY: NYSE, annually.

94. . “Marketing Research Report.” NewYork, NY: NYSE, monthly.

95. . Public Transactions Study. NewYork, NY: NYSE, 1977 and 1981.

96. Newspaper Advertising Bureau. News release onnewspaper advertising expenditures. New York,NY: NAB, monthly.

97. Newspaper Enterprise Association, Inc. “Associa-tions and Societies.” World Almanac and Book ofFacts. New York, NY: NEA, annually.

98. Nielsen Media Research. “Cable Penetration Es-timates.” Nielsen Station Index News. New York,NY: NMR, four times a year.

99. PennWell Publishing Company. The Oil and GasJournal. Tulsa, OK: PWPC, weekly.

100. Polk, R.L. & Company. “Firm Name Registrationsvs. Total Industry.” Cincinnati, OH: R.L. Polk,monthly.

101. Research Associates of Washington. Higher Edu-cation Prices and Price Indexes. Washington, DC:RAOW, annually.

102. Rubber Manufacturers Association. Rubber Facts.New York, NY: RMA, annually.

103. Shonka, D.B. Characteristics of Automotive Fleetsin the United States, 1966–77. Oak Ridge, TN:Oak Ridge National Laboratory, 1978.

104. Smith Travel Research. “Monthly LodgingOutlook.” Hendersonville, TN: STR, monthly.

105. Spectator Company. Spectator Insurance Year-book. Philadelphia, PA: Spectator Company,annually. (Discontinued.)

106. Textile Economics Bureau, Inc. “Textile FibersEnd-Use Survey.” New York, NY: TEB, monthly.

107. The Foundation Center. Foundation Grants Index.New York, NY: FC, annually.

108. U.S. Department of Agriculture. Economic In-dicators of the Farm Sector: National FinancialSummary. Washington, DC: U.S. GovernmentPrinting Office, annually.

109. . Agricultural Prices. Washington,DC: USDA, monthly.

110. . Agricultural Prices: AnnualSummary. Washington, DC: USDA, annually.

111. U.S. Department of the Army. Corps of Engineers.Waterborne Commerce of the United States. NewOrleans, LA: Corps of Engineers, annually.

112. U.S. Department of Commerce. Bureau of the Cen-sus. American Housing Survey. Washington, DC:U.S. Government Printing Office, biennially.

113. . Census of Agriculture. Wash-ington, DC: U.S. Government Printing Office,quinquennially.

114. . Census of Governments. Wash-ington, DC: U.S. Government Printing Office,quinquennially.

115. . Census of Housing. Washington,DC: U.S. Government Printing Office, decennially.

116. . Census of Housing: Survey ofResidential Finance. Washington, DC: U.S.Government Printing Office, decennially.


117. . Census of Manufactures. Wash-ington, DC: U.S. Government Printing Office,quinquennially.

118. . Census of Mineral Industries.Washington, DC: U.S. Government Printing Office,quinquennially.

119. . Census of Retail Trade. Wash-ington, DC: U.S. Government Printing Office,quinquennially.

120. . Census of Service Industries. Wash-ington, DC: U.S. Government Printing Office,quinquennially.

121. . Census of Transportation: Com-modity Transportation Survey. Washington, DC:U.S. Government Printing Office, quinquennially.

122. . Census of Transportation: Nonreg-ulated Carriers and Public Warehousing. Wash-ington, DC: U.S. Government Printing Office,quinquennially.

123. . Census of Transportation: TruckInventory and Use Survey. Washington, DC: U.S.Government Printing Office, quinquennially.

124. . Census of Wholesale Trade. Wash-ington, DC: U.S. Government Printing Office,quinquennially.

125. . Construction Reports: Survey ofResidential Alterations and Repairs. (Series C-50). Washington, DC: U.S. Government Printing,annually.

126. . County Business Patterns. Wash-ington, DC: U.S. Government Printing Office,annually.

127. . Current Business Reports: AdvanceMonthly Retail Sales. (Series BR.) Washington,DC: U.S. Government Printing Office, monthly.

128. . Current Business Reports: AnnualRetail Trade. (Series BR-13.) Washington, DC:U.S. Government Printing Office, annually.

129. . Current Business Reports: AnnualSurvey of Manufactures. Washington, DC: U.S.Government Printing Office, annually.

130. . Current Business Reports: AnnualWholesale Trade. (Series BW-13.) Washington,DC: U.S. Government Printing Office, annually.

131. . Current Business Reports: MonthlyRetail Trade. (Series BR.) Washington, DC: U.S.Government Printing Office, monthly.

132. . Current Business Reports: ServiceAnnual Survey. (Series BS.) Washington, DC: U.S.Government Printing Office, annually.

133. . Current Housing Reports: HousingVacancies and Homeownership. (Series H-111.)Washington, DC: U.S. Government Printing Office,monthly.

134. . Current Population Reports: Con-sumer Buying Indicators. Washington, DC: U.S.Government Printing Office, July 1971.

135. . Current Population Reports: Esti-mates of the Population by Age, Race, and Sex.(Series P-25.) Washington, DC: U.S. GovernmentPrinting Office, annually.

136. . Current Population Reports: House-hold and Family Characteristics. (Series P-20.)Washington, DC: U.S. Government Printing Office,annually.

137. . Current Population Reports: SchoolEnrollment—Social and Economic Characteristicsof Students. (Series P-20.) Washington, DC: U.S.Government Printing Office, annually.

138. . Governmental Finances. (SeriesGF.) Washington, DC: U.S. Government PrintingOffice, annually.

139. . Housing Construction Statistics,1889 to 1964. Washington, DC: U.S. GovernmentPrinting Office, 1966.

140. . 1956 National Housing Inventory.Washington, DC: U.S. Government Printing Office,1958.

141. . Quarterly Summary of Federal,State, and Local Tax Revenue. Washington, DC:U.S. Government Printing Office, quarterly.

142. . “Selected Spectator Sports.” Statis-tical Abstract of the United States. Washington,DC: U.S. Government Printing Office, annually.

143. . “Survey of Gallonage Sales ofGasoline.” Washington, DC: BOC, monthlyunpublished. (Discontinued.)

144. . U.S. Exports. (Series FT 410.)Washington, DC: U.S. Government Printing Office,monthly.

145. . U.S. General Imports and Importsfor Consumption. (Series FT 135.) Washington,DC: U.S. Government Printing Office, monthly.

146. U.S. Department of Commerce. Bureau of Eco-nomic Analysis. “BEA’s Measurement of Com-puter Output.” SURVEY OF CURRENT BUSINESS.Washington, DC: U.S. Government Printing Office,July 1989.

147. . Corporate Profits: Profits BeforeTax, Profits Tax Liability, and Dividends. (BEAMethodology Papers Series MP-2.) Washington,DC: U.S. Government Printing Office, May 1985.

148. . “Deflators for Purchases of Com-puters in GNP: Revised and Extended Estimates,1983–88.” SURVEY OF CURRENT BUSINESS. Wash-ington, DC: U.S. Government Printing Office,November 1988.

149. . Dollar Measures of Energy Pro-duction and Consumption in the United States,1972–82. (BEA Working Papers Series WP-5.)Washington, DC: BEA, 1987.

150. . Fixed Reproducible Tangible Wealthin the United States. Washington, DC: U.S.Government Printing Office, June 1987.

Sources 79

151. . Foreign Transactions. (BEAMethodology Papers Series MP-3) Washington,DC: U.S. Government Printing Office, May 1987.

152. . Government Transactions. (BEAMethodology Papers Series MP-5.) Washington,DC: U.S. Government Printing Office, November1988.

153. . “Improved Adjustment for Mis-reporting of Tax Return Information Used toEstimate National Income and Product Accounts,1977.” SURVEY OF CURRENT BUSINESS. Washington,DC: U.S. Government Printing Office, June 1984.

154. . “Table 5.4—Purchases of Structuresby Type.” SURVEY OF CURRENT BUSINESS. Washing-ton, DC: U.S. Government Printing Office, Julyissues.

155. . “Table 2.1—Personal Income andIts Disposition.” SURVEY OF CURRENT BUSINESS.Washington, DC: U.S. Government Printing Office,monthly.

156. . “The Input-Output Structure of theU.S. Economy.” SURVEY OF CURRENT BUSINESS. Wash-ington, DC: U.S. Government Printing Office, May1984.

157. . “The U.S. National Income andProduct Accounts: Revised Estimates.” SURVEY

OF CURRENT BUSINESS. Washington, DC: U.S.Government Printing Office, July 1986.

158. . “U.S. International Transactions.”Quarterly in March, June, September, and De-cember issues of the SURVEY OF CURRENT BUSINESS.Washington, DC: U.S. Government Printing Office,monthly.

159. U.S. Department of Commerce. Bureau ofPublic Roads. Public Roads. Washington,DC: U.S. Government Printing Office, monthly.(Discontinued.)

160. U.S. Department of Commerce. National Oceanicand Atmospheric Administration. “Monthly De-gree Day Data Summary.” Washington, DC,monthly, unpublished.

161. U.S. Department of Commerce. Office of Busi-ness Economics. National Income, 1954 Edition:A Supplement to the SURVEY OF CURRENT BUSINESS.Washington, DC: U.S. Government Printing Office,1954.

162. . “Price Ceilings and Wartime Con-trol Over the Economy.” SURVEY OF CURRENT

BUSINESS 22 (June 1942): 19–26.163. . U.S. Income and Output: A Sup-

plement to the SURVEY OF CURRENT BUSINESS. Wash-ington, DC: U.S. Government Printing Office,1958.

164. U.S. Department of Education. Center for Edu-cation Statistics. Digest of Education Statistics.

Washington, DC: U.S. Government Printing Of-fice, annually. (Formerly National Center forEducation Statistics.)

165. . “Fall Enrollment in Colleges andUniversities.” Washington, DC: CES, annually.

166. . “Financial Statistics of Institutionsof Higher Education.” Washington, DC: CES,annually.

167. U.S. Department of Education. National Centerfor Education Statistics. Private Schools in Ameri-can Education. Washington, DC: U.S. GovernmentPrinting Office, 1981.

168. . Statistics of Public Elementaryand Secondary Day Schools, 1977–78 School Year.Washington, DC: U.S. Government Printing Office.

169. U.S. Department of Energy. Energy Informa-tion Administration. Annual Report to Congress.Washington, DC: U.S. Government Printing Office,annually.

170. . Electric Power Monthly. Wash-ington, DC: U.S. Government Printing Office,monthly.

171. . Energy Data Reports: Crude Petro-leum, Petroleum Products, and Natural Gas Liq-uids, 1977. Washington, DC: U.S. GovernmentPrinting Office, 1978.

172. . Energy Data Reports: Deliveries ofFuel Oils and Kerosene. Washington, DC: U.S.Government Printing Office, annually.

173. . Energy Data Reports: Sales ofFuel Oil and Kerosene. Washington, DC: U.S.Government Printing Office, annually.

174. . Energy Data Reports: Sales ofLiquified Petroleum Gas and Ethane in 1977.Washington, DC: U.S. Government Printing Office,1978.

175. . Monthly Energy Review. Wash-ington, DC: U.S. Government Printing Office,monthly.

176. . Monthly Petroleum Product PriceReport. Washington, DC: U.S. GovernmentPrinting Office, monthly.

177. . Natural Gas Annual. Washington,DC: U.S. Government Printing Office, annually.

178. . Petroleum Market Shares: Re-port on Retail Sales of Gasoline. Washington,DC: U.S. Government Printing Office, monthly.(Discontinued.)

179. . Petroleum Marketing Annual.Washington, DC: U.S. Government Printing Office,annually.

180. . Petroleum Supply Monthly. Wash-ington, DC: U.S. Government Printing Office,monthly.

181. . Residential Energy ConsumptionSurvey: Consumption Patterns of Household Ve-


hicles, 1983. Washington, DC: U.S. GovernmentPrinting Office, 1984.

182. . Statistics of Privately Owned Utili-ties. Washington, DC: U.S. Government PrintingOffice, annually.

183. . Weekly Petroleum Status Report.Washington, DC: U.S. Government Printing Office,weekly.

184. U.S. Department of Energy. Federal Energy Regu-latory Commission. Annual reports of oil pipelinecompanies (Form P). Washington, DC, annually,unpublished.

185. U.S. Department of Health, Education, and Wel-fare. Social Security Administration. “PrivateHealth Insurance: Health Care Services, Enroll-ment, and Finances.” Social Security Bulletin.Washington, DC: U.S. Government Printing Office,annually. (Article discontinued.)

186. . “Social Welfare Expenditures.” So-cial Security Bulletin. Washington, DC: U.S.Government Printing Office, annually. (Articlediscontinued.)

187. . “Workmen’s Compensation Pay-ments and Costs.” Social Security Bulletin.Washington, DC: U.S. Government Printing Office,annually. (Article discontinued.)

188. U.S. Department of Health and Human Services.Health Care Financing Administration. Hospitaland nursing home input price indexes. Baltimore,MD, quarterly, unpublished.

189. . “National Health Expenditures.”Health Care Financing Review. Baltimore, MD:HCFA, annually.

190. . “Prepaid Health Plans.” Baltimore,MD, annually, unpublished.

191. . “Total Private Health Insurance.”Baltimore, MD, annually, unpublished.

192. U.S. Department of Health and Human Services.National Center for Health Statistics. MonthlyVital Statistics Report: Births, Marriages, Divorcesand Deaths. Washington, DC: U.S. GovernmentPrinting Office, monthly.

193. U.S. Department of Health and Human Services.Social Security Administration. “Income Replace-ment During Sickness.” Social Security Bulletin.Washington, DC: U.S. Government Printing Office,May 1981.

194. U.S. Department of Housing and Urban Devel-opment. Survey of Utility Payments by MobileHome Owners. Washington, DC: U.S. GovernmentPrinting Office, 1966.

195. U.S. Department of the Interior. Bureau of Mines.Mineral Industry Surveys. Washington, DC: U.S.Government Printing Office, monthly.

196. . Minerals Yearbook. Washington,DC: U.S. Government Printing Office, annually.

197. U.S. Department of Labor. Bureau of LaborStatistics. “Average Price of Car Components.”Washington, DC, monthly, unpublished.

198. . “Average Retail Price for Gasolineand Motor Oil.” Washington, DC: BLS, monthly,unpublished. (Discontinued.)

199. . “Average Retail Prices of Se-lected Popular-Sized and Compact Automo-biles.” Washington, DC, monthly, unpublished.(Discontinued.)

200. . Consumer Expenditures Sur-vey: Interview Survey. Washington, DC: U.S.Government Printing Office, biennially.

201. . CPI Detailed Report. Washington,DC: U.S. Government Printing Office, monthly.

202. . Employment and Earnings. Wash-ington, DC: U.S. Government Printing Office,monthly.

203. . Employment and Wages. Wash-ington, DC: U.S. Government Printing Office,annually.

204. . Producer Prices and Price Indexes.Washington, DC: U.S. Government Printing Office,monthly.

205. U.S. Department of the Treasury. Internal Rev-enue Service. “Employee Pension Benefit Plans,1977.” SOI Bulletin. Washington, DC: U.S.Government Printing Office, Spring 1982.

206. . Statistics of Income: Corpora-tion Income Tax Returns. Washington, DC: U.S.Government Printing Office, annually.

207. . Statistics of Income: Partner-ship Returns. Washington, DC: U.S. GovernmentPrinting Office, annually. (Discontinued.)

208. . Statistics of Income: SoleProprietorship Returns. Washington, DC:U.S. Government Printing Office, annually.(Discontinued.)

209. U.S. Department of the Treasury. Treasury Bul-letin. Washington, DC: U.S. Government PrintingOffice, quarterly.

210. U.S. Department of Transportation. Federal Avi-ation Administration. 1980 Survey of AirportServices. Washington, DC: FAA, 1980.

211. . FAA Statistical Handbook ofAviation. Washington, DC: FAA, annually.

212. . General Aviation Activity andAvionics Survey. Washington, DC: FAA, annually.

213. U.S. Department of Transportation. FederalHighway Administration. Highway Statistics.Washington, DC: U.S. Government Printing Of-fice, annually. (Formerly U.S. Department ofCommerce, Bureau of Public Roads.)

214. . Highway Statistics: Summary to1985. Washington, DC: U.S. Government PrintingOffice, 1987.

Sources 81

215. U.S. Department of Transportation. Office ofTransportation Systems Analysis and Informa-tion. Taxicab Operating Characteristics. Spring-field, VA: National Technical Information Service,1977.

216. U.S. Department of Transportation. Transporta-tion Systems Center. Air Carrier Financial Statis-tics. Cambridge, MA: TSC, quarterly. (Formerlypublished by Civil Aeronautics Board.)

217. . Air Carrier Traffic Statistics. Cam-bridge, MA: TSC, monthly. (Formerly published byCivil Aeronautics Board.)

218. . National Transportation Statistics.Cambridge, MA: TSC, annually.

219. U.S. Environmental Protection Agency. Light DutyAutomotive Fuel Economy Trends Through 1984.Washington, DC: U.S. Government Printing Office,1984.

220. U.S. Executive Office of the President. Officeof Management and Budget. Budget of theUnited States Government. Washington, DC: U.S.Government Printing Office, annually.

221. . Standard Industrial Classifica-tion Manual. Washington, DC: U.S. GovernmentPrinting Office, 1972.

222. U.S. Federal Election Commission. FEC Index ofIndependent Expenditures. Washington, DC: FEC,quadrennially.

223. . Final Report, U.S. Senate and HouseCampaigns. Washington, DC: FEC, biennially.

224. . Reports on Financial Activity.Washington, DC: FEC, annually.

225. U.S. General Services Administration. FederalMotor Vehicle Fleet Report. Washington, DC: U.S.Government Printing Office, annually.

226. U.S. House of Representatives. Committee onBanking and Currency. Subcommittee on Do-mestic Finance. Tax-Exempt Foundations: TheirImpact on Our Economy. Washington, DC: U.S.Government Printing Office, 1972.

227. U.S. Office of Price Administration. “Eligibil-ity Requirement Survey.” Washington, DC: OPA,1942–45, unpublished.

228. U.S. Postal Service. Annual Report of the Post-master General. Washington, DC: USPS, annually.

229. U.S. Securities and Exchange Commission. An-nual Report. Washington, DC: U.S. GovernmentPrinting Office, annually.

230. . Monthly Statistical Review. Wash-ington, DC: U.S. Government Printing Office,monthly. (Discontinued.)

231. . “Survey of Commission Rates.”Washington, DC, monthly, unpublished. (Discon-tinued.)

232. . The Securities Industry in 1980.Washington, DC: SEC, 1981.

233. Variety, Inc. Variety. New York, NY: Variety,weekly.

234. Ward’s Communications, Inc. Ward’s AutomotiveReports. Detroit, MI: Ward’s, weekly.

235. Western Union Corporation. Annual Report.Upper Saddle River, NJ: Western Union, annually.

Appendix–Selected NIPA Tables

This appendix reproduces the four NIPA tables that present PCE by detailedtype of product and type of expenditure: (1) Table 2.2.–Personal ConsumptionExpenditures by Major Type of Product, (2) Table 2.3.–Personal Consump-tion Expenditures by Major Type of Product in Constant Dollars, (3) Table2.5.–Personal Consumption Expenditures by Type of Product in ConstantDollars, and (4) Table 2.4.–Personal Consumption Expenditures by Type ofExpenditure. The 1988 estimates in these tables are from the July 1989SURVEY.

83

Methodology Paper 6: Personal Consumption Expenditures

Documents

Transcript of Methodology Paper 6: Personal Consumption Expenditures