Printer Friendly

Annual revision of the U.S. national income and product accounts.

In this issue of the Survey of Current Business, the Bureau of Economic Analysis (BEA) presents revised estimates of the national income and product accounts (NIPA'S) for 1990-92 and the first two quarters of 1993. As is usual in annual NIPA revisions, source data that are more complete, more detailed, and otherwise more appropriate than previously available information have been incorporated into the estimates, and seasonal factors have been updated. In addition, several methodological changes have been made.

The first section of this article discusses the impact of the revisions on several measures of economic activity, the second section provides a summary of the revisions and the major source data underlying them, and the third section describes the changes in methodology and summarizes the source data and methods used to prepare the NIPA estimates. Appendix A to this article shows, in current dollars, the revised annual estimates and the revisions for the five summary accounts of the NIPA'S. A set of most of the NIPA tables follows this article; the missing tables are scheduled to appear in the September 1993 Survey (see page 52). An index to the complete set of NIPA tables begins on page 122.

Impact of the Revisions

The revised estimates show that the U.S. economy was stronger than indicated by the previously published estimates. From the fourth quarter of 1989 to the second quarter of 1993, the growth rate (average annual rate of change) for real gross domestic product (GDP) was revised up 0.5 percentage point to 1.4 percent (table 1). The stronger growth was mainly due to upward revisions to personal consumption expenditures (PCE) for goods, to nonresidential producers' durable equipment (PDE), and to State and local government purchases.

[TABULAR DATA OMITTED]

Growth in real GDP is stronger for each of the 3 full years that were revised. The increase for 1990 was revised from 0.8 percent to 1.2 percent, the decrease for 1991 was revised from 1.2 percent to 0.8 percent, and the increase for 1992 was revised from 2.1 percent to 2.6 percent. On a fourth-quarter-to-fourth-quarter basis, the 0.5-percent decrease during 1990 was revised to a 0.2-percent increase; the increase during 1991 was revised from 0.1 percent to 0.3 percent; and the increase during 1992 was revised from 3-1 percent to 3.9 percent.

The revised estimates show slightly more inflation than previously indicated. From the fourth quarter of 1989 to the second quarter of 1993, the average annual rates of increase in the price indexes fixed weights) for both GDP and gross domestic purchases were revised up slightly: 0.1 percentage point to 3.8 percent for the GDP price index and 0.2 percentage point to 3.8 percent for the gross domestic purchases price index (table 2). Among major components, none of the revisions to the average annual rates of change exceeded 0.4 percentage point.

[TABULAR DATA OMITTED]

Business cycle. - As in the previously published estimates, real GDP reaches a cyclical peak in the second quarter of 1990 and a cyclical trough in the first quarter of 1991. However, in the revised estimates, the contraction is less severe than in the previously published estimates. The total decrease in real GDP from the second quarter of 1990 to the first quarter of 1991 is now 1.6 percent (not at an annual rate); in the previously published estimates, the decrease was 2.2 percent. An upward revision to gross private domestic investment contributed most to the lessened severity of the contraction.

In the revised estimates, the current expansion is stronger than in the previously published estimates. The increase in real GDP from the first quarter of 1991 to the second quarter of 1993 is now 2.4 percent (average annual rate); in the previously published estimates, the increase was 2.0 percent. As a consequence of the stronger growth in the revised estimates, the preceding peak in real GDP (that is, the level of GDP in the second quarter of 1990) is surpassed in the second quarter of 1992, one quarter earlier than in the previously published estimates.

Gross saving and investment. - Gross saving as a percentage of gross national product (GNP) is somewhat higher on the revised basis, but a downtrend is still apparent: It was unrevised at 13.0 percent for 1990 and was revised up 0.4 percentage point to 12.8 percent for 1991 and 0.4 percentage point to 11.9 percent for 1992.

The level of gross saving is revised up substantially for 1991 and 1992, reducing the previously published decline of 6.9 percent from 1989 to 1992 to a decline of 3.2 percent. For 1991, the revision was more than accounted for by corporate net cash flow - undistributed corporate profits with inventory valuation adjustment (IVA) and with capital consumption adjustment (CCAdj) and corporate consumption of fixed capital) - which was revised up substantially mainly because of revisions to undistributed profits. For 1992, the upward revision to gross saving was mainly accounted for by an upward revision to personal saving and a downward revision to the government deficit.

Personal saving was revised up substantially for 1992; as a result, the personal saving rate (personal saving as a percentage of disposable personal income) was revised up 0.5 percentage point to 5.3 percent. The revision to personal saving stemmed mainly from a substantial upward revision to wage and salary disbursements, which in part reflected a large upward revision to the previous estimate of accelerated bonus payments (that is, the payment in the fourth quarter of 1992 of yearend bonuses that typically would have been paid in the first quarter of 1993). Because this revision affected wage and salary disbursements and not wage and salary accruals, the "wage accruals less disbursements" component of gross saving was revised down by the same amount. (See the section "Changes in methodology.")

The government deficit (NIPA basis) for 1992 is somewhat smaller than previously estimated. The downward revision was more than accounted for by the Federal Government deficit, which, at $276.3 billion, is $21.7 billion smaller, mostly reflecting higher receipts. The State and local government surplus, at $7.2 billion, is $8.3 billion smaller, mostly reflecting higher expenditures.

Gross investment - which is measured as the sum of gross private domestic investment, net foreign investment, and the statistical discrepancy - is somewhat higher on the revised basis. Within gross investment, gross private domestic investment was revised up for all 3 years, and net foreign investment was revised down for all 3 years.

Summary of the Revisions

The incorporation of newly available source data and of changes in methodology leads to revisions to current-dollar estimates and to the prices and quantities used to prepare constant-dollar estimates. In turn, these revisions lead to revisions to constant-dollar estimates. This section describes the revisions to the annual current-dollar, price, and constant-dollar estimates and then briefly describes the revisions to the quarterly estimates.

Annual current-dollar estimates

Table 3 summarizes the current-dollar revisions to major NIPA components. It provides a guide to the revisions by identifying the subcomponent series for which revisions were $2.0 billion or more and by listing the major source data that underlie the revised estimates. (For a list of the principal source data and estimating methods used in preparing the current-dollar estimates, see table 7.) It should be noted that newly available source data lead to a revision in the level of an estimate not only for the year into which they are directly incorporated, but usually in the levels for subsequent years as well. The discussion in this section follows the sequence of entries in table 3.

[TABULAR DATA OMITTED]

GDP. - The level of current-dollar GDP was revised up for all 3 years: $23.9 billion, or 0.4 percent, for 1990; $45.4 billion, or 0.8 percent, for 1991; and $87.8 billion, or 1.5 percent, for 1992. These revisions were large by historical standards. For example, the revision for 1992, the "most recent year" in this annual NIPA revision, is 1.0 percentage point larger than the average revision for the "most recent year" in the annual NIPA revisions since 1969.

In this annual revision, the largest dollar revisions among the major components of GDP were to personal consumption expenditures for goods and for services, to nonresidential producers' durable equipment, and to State and local government purchases.

Personal consumption expenditures (PCE) for goods. - PCE goods was revised up for all 3 years: $8.6 billion for 1990, $18.1 billion for 1991, and $27.1 billion for 1992. Upward revisions to "goods other than motor vehicles and parts" accounted for most of the revisions for all 3 years; these revisions resulted from the incorporation of revised Census Bureau retail sales data. Within this category, the largest upward revisions were to furniture and household equipment, food, clothing and shoes, and "other durable goods." Gasoline and oil was revised down for all 3 years, reflecting revised gasoline consumption data by type of vehicle for 1990 from the Federal Highway Administration (FHWA), newly available FHWA consumption data for 1991, and Energy Information Administration data on gasoline supplied to retailers for 1992.

PCE for services. - PCE services was revised up for all 3 years: $4.1 billion for 1990, $0.6 billion for 1991, and $16.9 billion for 1992. Revisions to recreation services accounted for most of the upward revision for 1990; revisions to personal business services more than accounted for the upward revisions for 1991 and 1992. The upward revision to recreation services for 1990 was accounted for primarily by revised Census Bureau data on annual service receipts. The upward revisions to personal business services for 1991 and 1992 reflected the incorporation of newly available information from a variety of regular sources for two components - "services furnished without payment by financial intermediaries except life insurance carriers and private noninsured pension plans" and "expense of handling life insurance."

Among the other types of PCE services, household operation services was revised up for all 3 years; the revisions were primarily to telephone and telegraph services. Within telephone and telegraph services, the revisions largely reflected a change in BEA'S assumptions about the consumer share of purchases of cellular phone services; previously, the consumer share of these purchases had been assumed to be zero. In addition, newly available source data for cellular phone services revenue for 1990-92 were incorporated.

Transportation services was revised down for 1990 and 1991 and up for 1992. Revisions to motor vehicle repair, greasing, washing, parking, storage, rental, leasing, and related services more than accounted for the 1991 and 1992 revisions. These revisions primarily reflected the incorporation of revised Census Bureau receipts data for 1991, new receipts data for 1992, and new Bureau of Labor Statistics (BLS) data on automobile rental and leasing for 1991.

Medical care services was revised up for 1990 and down for 1991 and 1992. Within medical care services, health insurance was revised up for all 3 years, reflecting revised Health Care Financing Administration (HCFA) data for 1990, new HCFA data for 1991, and revised BEA projections for 1992. Physicians' services was revised down for all 3 years, reflecting the incorporation of revised Census Bureau annual receipts data for 1990 and 1991 and new receipts data for 1992. Personal care services was revised down for all 3 years, also reflecting the incorporation of newly available Census Bureau receipts data.

Net foreign travel was revised up for 1990 and down for 1991 and 1992, reflecting revisions to the BEA balance of payments accounts (BPA'S) and an improved procedure for estimating passenger fare payments by U.S. residents; the improved procedure incorporates BPA estimates of passenger fare exports and imports and U.S. Department of Transportation data on international passenger revenue.

Nonresidential structures. - Nonresidential structures was revised up for all 3 years: $0.5 billion for 1990, $2.5 billion for 1991, and $4.2 billion for 1992. Revisions to utilities, mainly to electric light and power, more than accounted for the upward revision for 1991. The upward revision to electric light and power reflected newly available data from a variety of regular sources. For 1992, nonresidential buildings was revised up, and "other nonfarm structures" was revised down. These revisions reflected revised Census Bureau data on the value of new construction put in place. Nonresidential producers' durable equipment (PDE). - Nonresidential PDE was revised up for all 3 years: $8.6 billion for 1990, $12.4 billion for 1991, and $13.0 billion for 1992. The upward revisions were widespread among the components of PDE. For 1990 and 1991, the revisions primarily reflected the introduction of commodity shipments data from the 1990 and 1991 Annual Surveys of Manufactures. For 1992, the revisions reflected newly available industry shipments data from the Census Bureau monthly survey of manufacturers and from the Census Bureau Current Industrial Report covering civilian aircraft. For all 3 years, the revisions also reflected a BEA adjustment to account for a downward bias in the Census Bureau shipments data that resulted from an incomplete incorporation of new businesses (see the section "Changes in methodology").

Residential fixed investment. - Residential fixed investment was revised down slightly for 1990 and 1991 and up $5.9 billion for 1992. For 1992, the revision was largely accounted for by an upward revision to "improvements," reflecting the incorporation of information from BLS and Census Bureau surveys.

Change in business inventories. - The change in business inventories was revised up for all 3 years: $0.6 billion for 1990, $1.6 billion for 1991, and $2.9 billion for 1992. The change in farm inventories was unrevised for 1990 and 1991 and was revised up $2.8 billion for 1992; the revision for 1992 reflected the incorporation of new estimates from the U.S. Department of Agriculture (USDA).

The change in nonfarm inventories was revised up for all 3 years: $0.5 billion for 1990, $1.7 billion for 1991, and $0.1 billion for 1992. For 1990, upward revisions to the IVA more than offset downward revisions to the change in book value; both revisions largely reflected the incorporation of data from a variety of regular sources. For 1991, the upward revision was primarily due to revisions to the IVA. For 1992, upward revisions to the IVA more than offset downward revisions to the change in book value; the latter reflected downward revisions to retail trade automotive dealers and to industries other than manufacturing and trade that offset an upward revision to merchant wholesale trade.

Net exports of goods and services. - Net exports of goods and services was revised down $2.5 billion for 1990, up $2.2 billion for 1991, and up $0.8 billion for 1992. The downward revision for 1990 was due to upward revisions to imports of merchandise and of services. The upward revisions for 1991 and 1992 were due to upward revisions to exports of merchandise; for 1992, the revision to merchandise exports was mostly offset by an upward revision to imports of services. Except for the revisions to merchandise exports for 1991 and 1992, the revisions to exports and imports mainly reflected revisions to the BPA'S. The revisions to merchandise exports mainly reflected revisions to the territorial adjustment and were based on revised data from Puerto Rico and the U.S. Virgin Islands.(1) (For more information about revisions to the BPA'S, see the section "Changes in methodology.")

Government purchases. - Government purchases was revised up for all 3 years: $4.2 billion for 1990, $8.8 billion for 1991, and $16.9 billion for 1992. The revisions were primarily to State and local government purchases, which accounted for most of the revision for 1990 and more than accounted for the revisions for 1991 and 1992, and reflected revised fiscal year 1991 and new fiscal year 1992 data from Census Bureau surveys of government finances. Within State and local government purchases, the revisions for 1991 and 1992 were largely to structures and to "goods and services other than compensation of employees."

Net receipts of factor income. - Net receipts of factor income from the rest of the world, which is excluded from GDP but included in GNP, was revised up $1.0 billion for 1990, down $3.3 billion for 1991, and down $3.9 billion for 1992. Both receipts and payments of factor income were revised up for all 3 years. Receipts of factor income was revised up $8.0 billion for 1990, $2.6 billion for 1991, and $0.4 billion for 1992; payments of factor income was revised up $7.0 billion for 1990, $5.9 billion for 1991, and $4.3 billion for 1992. The revisions reflected the incorporation of the results of two 1989 benchmark surveys: BEA'S benchmark survey of U.S. direct investment abroad and the U.S. Treasury Department's benchmark survey of foreign portfolio investment in the United States. (See the section "Changes in methodology.")

GNP. - The level of GNP was revised up for all 3 years: $24.9 billion, or $0.4 percent, for 1990; $42.2 billion, or 0.7 percent, for 1991; and $83.9 billion, or 1.4 percent, for 1992. The revision to GNP for 1990 was roughly the same as that to GDP, and it was somewhat smaller than that to GDP for 1991 and 1992. The relatively smaller revisions to GNP for 1991 and 1992 reflected the downward revisions to net receipts of factor income.

Gross national income (GNI) and the statistical discrepancy. - GNI measures the costs incurred and the profits earned in the production of GNP; GNP measures final expenditures - the sum of consumer spending, investment, net exports, and government purchases - plus net receipts of factor income from the rest of the world. In concept, GNP is equal to GNI. In practice, they differ because they are estimated using less than perfectly consistent source data. The difference between GNP and GNI is the statistical discrepancy.

The level of GNI was revised up for all 3 years: $22.6 billion, or 0.4 percent, for 1990; $54.4 billion, or 1.0 percent, for 1991; and $94.4 billion, or 1.6 percent, for 1992. For 1990, the revision to GNI was slightly less than that to GNP, and the statistical discrepancy was revised up from $5.4 billion to $7.8 billion. For 1991 and 1992, the revisions to GNI were considerably larger than those tO GNP, and the statistical discrepancy was revised down from $21.9 to $9.6 billion for 1991 and from $34.1 billion to $23.6 billion for 1992. As a percentage of GNP, the revised statistical discrepancy is 0.1 percent for 1990, 0.2 percent for 1991, and 0.4 percent for 1992.

Within GNI, corporate profits with IVA and CCAdj was revised up considerably for all 3 years. Net interest, compensation of employees, and proprietors' income with IVA and CCAdj were revised up considerably for 1991 and 1992. For 1992, the upward revisions were partly offset by a downward revision to rental income of persons with CCAdj.

Compensation of employees. - Compensation of employees was revised up for all 3 years: $6.4 billion for 1990, $11.6 billion for 1991, and $56.8 billion for 1992. Both wages and salaries and supplements to wages and salaries were revised up for all 3 years. The substantial upward revision to wages and salaries for 1992 resulted from the incorporation of BLS tabulations of unemployment insurance wage and salary data. For 1992, these data were adjusted from a "when-paid" basis to an accrual basis for inclusion in national income by excluding the larger-than-usual amount of bonuses paid in 1992. (For more information, see the section "Changes in methodology.")

For all 3 years, the revisions to supplements were mostly accounted for by other labor income. For 1990, the revision to other labor income was to group health insurance; it reflected revised HCFA data. For 1991, the revision was to group health insurance, reflecting revised HCFA data, and to pension and profit-sharing plans, reflecting newly available Internal Revenue Service (IRS) tabulations of corporate tax return data. For 1992, the revision reflected the revised levels for 1991 and revised BEA projections for 1992. In addition, employer contributions for social insurance was revised up for 1992, reflecting the incorporation of newly available regular source data.

Proprietors' income with IVA and CCAdj. - Proprietors' income with IVA and CCAdj was revised down $3.6 billion for 1990, up $8.4 billion for 1991, and up $9.8 billion for 1992. For 1990 and 1991, the revisions were mostly accounted for by the nonfarm component of proprietors' income; for 1992, the farm and nonfarm components contributed about equally to the revision. For farm proprietors' income, the revision for 1992 reflected the incorporation of new estimates from USDA.

For nonfarm proprietors' income, the revision for 1990 primarily reflected the incorporation of newly available data from the IRS Taxpayer Compliance Measurement Program, which BEA uses for the tax return misreporting adjustment. This adjustment modifies tax return data for underreported incomes, for overreported expenses, and for nonreported incomes. The revisions for 1991 and 1992 primarily reflected newly available IRS tabulations of 1991 tax return data for sole proprietorships and partnerships. The CCAdj for nonfarm proprietors' income was revised down for 1991 and 1992 (see the discussion of consumption of fixed capital).

Rental income of persons with CCAdj. - Rental income of persons with CCAdj was revised down for all 3 years: $1.9 billion for 1990, $2.4 billion for 1991, and $13.6 billion for 1992. Rental income of persons (without CCAdj) accounted for all of the revisions for 1990 and 1991 and for $11.1 billion of the revision for 1992. For all 3 years, the revisions to rental income of persons were primarily due to revised mortgage interest payments - the largest expense item in calculating rental income. Specifically, the revised estimates incorporate an improvement in the procedure introduced in last year's annual revision for estimating the effective rate of interest on mortgage debt (see the section "Changes in methodology"). The CCAdj for rental income of persons was unrevised for 1990 and 1991 and was revised down $2.5 billion for 1992 (see the discussion of consumption of fixed capital).

Corporate profits with IVA and CCAdj. - Corporate profits with IVA and CCAdj was revised up for all 3 years: $18.9 billion for 1990, $23.2 billion for 1991, and $13.4 billion for 1992. Revisions to domestic profits primarily reflected the incorporation of revised IRS tabulations of corporate tax return data for 1990 and newly available tabulations for 1991. Revisions to rest-of-the-world profits reflected the incorporation of revisions to the BPA'S.

Corporate profits before tax was revised up for all 3 years. Domestic profits of financial corporations, mainly commercial banks and savings and loan associations, accounted for most of the revisions. For 1990, the revisions reflected the revised IRS tax return tabulations. For 1991, the revisions resulted from the replacement of profits derived from regulatory agency reports with profits derived from the tax return tabulations. For 1992, the revision largely reflected the revised levels for 1991.

The IVA for corporate profits, mainly for manufacturing, was revised up for all 3 years, reflecting revised price indexes and new information on inventory accounting methods from Census Bureau annual surveys.

The CCAdj for corporate profits was revised up for 1990 and down for 1991 and 1992 (see the discussion of consumption of fixed capital).

Net interest. - Net interest was revised up for all 3 years: $3.0 billion for 1990, $13.3 billion for 1991, and $26.8 billion for 1992. For each year, a downward revision to net monetary interest was more than offset by an upward revision to net imputed interest. For 1990, monetary interest paid to and received by domestic business were both revised up, reflecting revised IRS tabulations of corporate tax return data. Rest-of-the-world monetary interest paid and received were both revised up (received more than paid), reflecting revisions to the BPA'S. The revision to net imputed interest was to interest paid both by life insurance carriers, reflecting new source data (see the section "Changes in methodology"), and by private noninsured pension plans, reflecting revised Federal Reserve Board flow-of-funds accounts assets data and BEA estimates of effective interest rates. For 1991, monetary interest paid to and received by domestic business were both revised up, reflecting new IRS tabulations of tax return data for corporations and for nonfarm sole proprietorships and partnerships. The revision to net imputed interest was mostly to interest paid by pension plans, reflecting the revised flow-of-funds data and the estimated interest rates noted earlier. For 1992, the revisions were similar to, albeit larger than, those for 1991.

Consumption of fixed capital. - Consumption of fixed capital - that is, economic depreciation and accidental damage - was revised little for 1990 and 1991 and was revised up $4.5 billion for 1992. The revisions reflected revised BEA estimates of fixed investment and prices and, for 1992, revised trade source data used to estimate the damage caused by Hurricane Andrew.

Capital consumption allowances - that is, tax-return-based depreciation for corporations and nonfarm proprietorships, historical-cost depreciation using consistent service lives for farm proprietorships and rental income of persons, and accidental damage - was revised up $5.4 billion for 1990, down $8.2 billion for 1991, and down $13.6 billion for 1992. The upward revision for 1990 reflected revised IRS tabulations of corporate tax return data. The downward revisions for 1991 and 1992 reflected newly available IRS tabulations of 1991 tax return data for corporations and for nonfarm sole proprietorships and partnerships.

The CCAdj - which is derived as the difference between capital consumption allowances and the economic measure of depreciation - was revised up $5.5 billion for 1990, down $8.2 billion for 1991, and down $18.0 billion for 1992.

Nonfactor incomes. - Nonfactor incomes - indirect business tax and nontax liability, business transfer payments, and subsidies less current surplus of government enterprises - was revised little for 1990, up $0.4 billion for 1991, and down $3.3 billion for 1992. The revision for 1992 was largely due to business transfer payments.

National income. - National income was revised up for all 3 years: $22.7 billion for 1990, $54.1 billion for 1991, and $93.2 billion for 1992. These revisions reflected the aforementioned revisions to compensation of employees, proprietors' income, rental income of persons, corporate profits, and net interest.

Personal income and its disposition. - Personal income was revised up for all 3 years: $9.6 billion for 1990, $22.6 billion for 1991, and $86.8 billion for 1992. These revisions partly reflected the previously described revisions to the components of national income that are included in personal income - wages and salaries, other labor income, proprietors' income, and rental income of persons. They also reflect revisions to personal dividend income, personal interest income, and transfer payments to persons, all of which are in personal income but not in national income. Personal dividend income was revised up $4.1 billion for 1990, down $9.1 billion for 1991, and up $1.1 billion for 1992; these revisions reflected the incorporation of revised IRS tabulations of corporate tax return data for 1990 and newly available tabulations for 1991, as well as the incorporation of data from publicly available corporate financial statements for 1992. Personal interest income was revised up for all 3 years, largely reflecting the aforementioned revisions to net interest. Transfer payments was revised up for 1990 and down for 1991 and 1992. The revision for 1992 was more than accounted for by newly available HCFA data on State and local government payments for medicaid.

The smaller revisions to personal income than to national income for the 3 years were traceable to the components of corporate profits that are not in personal income. For 1992, the difference in the size of the revisions to the two income measures was much smaller than for the 2 preceding years because of the revisions to wages and salaries: Wages and salaries in national income was revised up $36.5 billion for 1992; wages and salaries in personal income was revised up $55.0 billion. (See the section "Changes in methodology.")

Personal tax and nontax payments was revised up for all 3 years: $2.0 billion for 1990, $1.7 billion for 1991, and $17.5 billion for 1992. The upward revision for 1992 was due mostly to payments to the Federal Government and reflected newly available data from the U.S. Treasury Department. As noted in the April 1993 Survey, there was an unusually large difference between the fiscal year 1992 Collections of personal taxes and the corresponding NIPA series. The difference was due to two factors: First, the impact of a change to withholding tables that was implemented in March 1992; and second, as noted above, a substantial understatement in the previously published NIPA wage and salary series used to estimate tax payments.(2)

Reflecting the revisions to personal income and to personal tax and nontax payments, disposable personal income (DPI) was revised up for an 3 years: $7.6 billion for 1990, $20.9 billion for 1991, and $69.4 billion for 1992.

Personal outlays - PCE, interest paid by persons, and personal transfer payments to rest of the world (net) - was revised up for all 3 years: $13.3 billion for 1990, $19.1 billion for 1991, and $43.4 billion for 1992. The revisions for all 3 years were largely to PCE.

Personal saving - the difference between DPI and personal outlays - was revised down $5.6 billion for 1990, up $1.9 billion for 1991, and up $26.1 billion for 1992.

Annual price estimates

Revisions to fixed-weighted price indexes result from the incorporation of newly available source data and of methodological changes. Source data that affect these indexes consist not only of price indexes, which are used for deflation, but also of current-dollar estimates and quantity data, which are used for preparing constant-dollar estimates by quantity extrapolation or direct valuation (see the section "Updated summary methodologies").

Newly available price index information includes revised price indexes for domestic, exported, and imported computers, for foreign travel, for single-family houses, and for defense goods and services. Newly available current-dollar estimates affected the price estimates for two types of PCE services - "services furnished without payment by financial intermediaries except life insurance carriers and private noninsured pension plans" and health insurance for petroleum and natural gas exploration, and for Federal Government and State and local government employee compensation.

In addition, methodological changes were made in the deflation of two GDP components. For residential structures, BEA'S Multifamily price index, which had been introduced in the 1991 comprehensive revision of the NIPA'S, was extended through 1992. For net exports, changes were made in the deflation of several service components - air passenger fares, royalties and license fees, "other private services," and the territorial adjustments. (See the section "Changes in methodology.")

The level of the GDP fixed-weighted price index was unrevised at 113.6 for 1990 and was revised up 0.1 index point to 118.2 for 1991 and up 0.5 index point to 122.1 for 1992. Reflecting these revisions in level, the annual percent increase in the index was unrevised at 4.6 percent for 1990 and was revised up 0.1 percentage point to 4.1 percent for 1991 and up 0.4 percentage point to 3.3 percent for 1992. (Revisions to price changes for gross domestic purchases were about the same as those for GDP.) Revisions to the price changes for components of GDP were generally small; the largest revisions were to PCE services for 1992 and to State and local government purchases for 1992 (table 4).

[TABULAR DATA OMITTED]

The increase in prices of PCE services for 1992 was revised up 0.9 percentage point to 5.0 percent. The revision was to services furnished without payment by financial intermediaries, to health insurance, and to foreign travel by U.S. residents. The prices of services furnished without payment by financial intermediaries and of health insurance are "implicit" prices because the constant-dollar estimates of these components are prepared by quantity extrapolation. The revisions to both components were due primarily to upward revisions to the current-dollar estimates. For services furnished without payment by financial intermediaries, the quantity extrapolator was revised down slightly; for health insurance, the quantity extrapolator was revised up slightly. The revision to foreign travel reflected a newly available composite index of consumer prices for foreign countries.

The increase in prices of State and local government purchases for 1992 was revised up 0.6 percentage point to 2-5 percent. The revision was mainly due to the implicit price for compensation of employees. For this component, the current-dollar estimate was revised up more than the quantity extrapolator.

Annual constant-dollar estimate

In general, revisions to real GDP reflect four factors: (1) Revisions to current-dollar components of GDP whose constant-dollar estimates are prepared by deflation, (2) revisions to the prices used in deflation, (3) revisions to the quantities used to estimate components of real GDP by extrapolation or direct valuation, and (4) revisions caused by shifts in the composition of current-dollar GDP. The following tabulation provides information on the effects of these factors on the revisions to the level of and percent change in real GDP, which are shown in the last line. Revisions to current-dollar GDP are shown in the first line; revisions to the GDP price index fixed weights) are shown in the second line; and "other" revisions, a residual, is shown in the third line.

In this presentation, the first line, "revisions to current-dollar GDP," includes current-dollar revisions for those components of real GDP that are estimated by extrapolation or direct valuation; the second line, "revisions to the GDP price index," includes revisions to the "implicit" prices that result from the relationship between the current-dollar estimates and the quantities used to prepare the constant-dollar estimates for those components. Thus, for those components, current-dollar revisions are offset by revisions to the "implicit" prices if there are no revisions to the constant-dollar estimates.

The level of real GDP was revised up for each year: 0.4 percent for 1990, 0.8 percent for 1991, and 1.3 percent for 1992. Reflecting these revisions in level, the annual percent change in real GDP was also revised up for each year: 0.4 percentage point to 1.2 percent for 1990, 0.5 percentage point to -0.7 percent for 1991, and 0.5 percentage point to 2.6 percent for 1992 (table 4). For 1990 and 1991, the current-dollar revisions fully accounted for the constant-dollar revisions; for 1991, the price and "other" revisions had small offsetting effects. For 1992, the current-dollar revision was more than the constant-dollar revision, and "prices" were revised up. The pattern of revisions for 1992 largely reflected revisions to the PCE component "services furnished by financial intermediaries except life insurance carriers and noninsured pension funds" - for which the constant-dollar estimate is prepared using quantity extrapolation. In current dollars, this component was revised up for 1992, contributing 0.2 percentage point to the 0.7-percentage point upward revision to current-dollar GDP. Because there was virtually no revision to the quantity extrapolator, there was no revision to the constant-dollar estimate of this component.

For 1990, the upward revision to the increase in real GDP was more than accounted for by PCE (both goods and services) and by nonresidential fixed investment. For 1991, the upward revision to the decrease in real GDP reflected smaller decreases in PCE goods and in nonresidential fixed investment and larger increases in State and local government purchases and in net exports of goods and services. For 1992, the upward revision to the increase in real GDP was accounted for by larger increases in PCE goods, residential fixed investment, and State and local government purchases and by a smafler decrease in net exports.

Revisions to components of real GDP. - The annual percent change in PCE was revised up for all 3 years: 0.3 percentage point to 1.5 percent for 1990, 0.2 percentage point to -0.4 percent for 1991, and 0.3 percentage point to 2.6 percent for 1992. For 1990, the upward revisions were to durable goods, nondurable goods, and services. For 1991, the smaller decrease was mainly due to an upward revision to durable goods, which was more than accounted for by furniture and household equipment; a smaller increase in services was partly offsetting. For 1992, the larger increase was mostly due to goods.

The change in nonresidential fixed investment was revised up 1.6 percentage points to an increase of 1.2 percent for 1990, up 1.1 percentage points to a decrease of 5.9 percent for 1991, and was revised little for 1992. The revisions for 1990 and 1991 largely reflected upward revisions tO PDE. For 1990, information processing, transportation equipment, and "other equipment" were revised up. For 1991, the smaller decrease mainly reflected revisions to industrial equipment.

The change in residential investment was revised little for 1990, down 0.3 percentage point to -12.9 percent for 1991, and up 3.1 percentage points to 16.3 percent for 1992. The upward revision for 1992 largely reflected revisions to the "improvements" component.

The change in exports of goods and services was revised little for 1990 and 1992 and up 0.6 percentage point to 6.4 percent for 1991. The upward revision for 1991 largely reflected revisions to goods.

The change in imports of goods and services was revised up 0.6 percentage point to 3.6 percent for 1990, down 0.4 percentage point to -0.5 percent for 1991, and down 0.9 percentage point to 8.7 percent for 1992. The revisions for all 3 years reflected current-dollar and price revisions to both goods and services. In goods, the price revisions were mainly to capital goods. In services, the price revisions were mainly to travel and passenger fares (see the section Changes in methodology").

The change in government purchases was revised up for all 3 years: 0.3 percentage point to 3.1 percent for 1990, 0.3 percentage point to 1.5 percent for 1991, and 0.2 percentage point to -0.1 percent for 1992. The revisions for all 3 years largely reflected revisions to State and local government purchases of structures and of "goods and services other than employee compensation." For 1991, a downward revision to the change in Federal Government purchases, mainly to services other than compensation of employees," partly offset the upward revision to State and local government purchases.

Quarterly estimates

Quarterly estimates are affected by annual NIPA revisions in three major ways: (1) Adjustments to reflect revisions to the annual estimates, (2) incorporation of new and revised source data (including the updating of seasonal factors that are used to indicate quarterly patterns), and (3) changes in the methodology used to prepare the quarterly estimates.

In general, the quarter-to-quarter patterns of changes in BEA'S measures of real output and prices are not markedly different on the revised basis (table 5), but output was clearly stronger in the quarters of 1990 and 1992. For real GDP, the revisions to the 14 quarterly percent changes (at annual rates) averaged 0.5 percentage point (without regard to sign). Only one of the 14 quarters had a downward revision in change (the second quarter of 1991), and two had no revision in change (the fourth quarter of 1991 and the third quarter of 1992). The two largest revisions in change were upward, and both were for 1992: The second quarter was revised up 1.3 percentage points to 2.8 percent, mainly reflecting PCE, and the fourth quarter was revised up 1.0 percentage point to 5.7 percent, mainly reflecting residential fixed investrnent.

[TABULAR DATA OMITTED]

For gross domestic purchases prices, the revisions to the 14 quarterly percent changes (at annual rates) averaged 0.2 percentage point (without regard to sign). For only one quarter was the revision larger than 0.5 percentage point: For the first quarter of 1992, the increase was revised up 1.0 percentage point to 3.9 percent. Sources of this revision were widespread.

The quarterly pattern of changes in disposable personal income (DPI) and related measures was revised substantiauy beginning with the fourth quarter of 1992. These revisions mainly reflected the incorporation of a larger-than-expected acceleration of bonus payments (see the section "Changes in methodology"). Primarily because of these revisions, the percent change (at an annual rate) in real DPI was revised up from 4.3 percent to 10.6 percent for the fourth quarter of 1992, down from 2.7 percent to -7.8 percent for the first quarter of 1993, and up from 1.3 percent to 5.9 percent for the second quarter of 1993.

Methodology

The revised NIPA estimates incorporate several changes in methodology - either in the source data or in the methods used to prepare the estimates. This section of the article describes these changes and updates previously published tables showing methodologies for current- and constant-dollar GDP.

Changes in methodology

Producers' durable equipment (PDE). - In the previously published PDE estimates for 1988 and 1989, BEA adjusted shipments data from the 1988 and 1989 Annual Surveys of Manufactures (ASM'S) to account for a downward bias that resulted from an incomplete incorporation of new businesses. The adjustments were based on information from the 1987 Census of Manufactures that implied shortfalls in the level of total manufactures' shipments of 0.7 percent for 1988 and 1.4 percent for 1989. In the previously pubfished estimates, BEA had discontinued the adjustment for estimates after 1989 because the Census Bureau improved its mailing list for subsequent ASM'S. However, because new Census Bureau studies indicated that the improvements reduced but did not eliminate the bias, BEA has reintroduced the adjustment for the estimates from 1990 forward. These adjustments implied shortfalls in the level of total manufactures' shipments of 1.2 percent for 1990, 1.6 percent for 1991, and 2.0 percent for 1992. For the components of manufactures' shipments used to estimate PDE, the adjustment added $3.9 billion for 1990, $5.1 billon for 1991, and $5.1 billion for 1992.

Price index for multifamily housing. - In the comprehensive revision of the NIPA'S that was released in December 1991, a hedonic price index for multifamily housing was used to prepare constant-dollar estimates of this component of residential fixed investment for 1977-89.(3) Until this annual revision, the price index for 1990 forward was extrapolated by the Census Bureau price deflator for single-family houses under construction. For the revised estimates, the multifamily price index was updated through 1992, based on regression equations for 1990, 1991, and 1992. The multifamily index is calculated only for annual estimates; for quarterly estimates, the Census Bureau single-family deflator is used as both the interpolator and extrapolator.

U.S. balance of payments accounts revisions. - The BPA'S provide the basis for the foreign transactions entries in the NIPA'S. (Differences between the BPA'S and the NIPA'S are identified in NIPA table 4.5 in this issue of the Survey.) In this year's annual BPA revision, BEA incorporated newly available data from regular sources and introduced a number of major improvements that involve new source data and new estimating methodologies. (For a description of the annual BPA revision, see "U.S. International Transactions, Revised Estimates for 1983-92" in the June 1993 Survey.)

The effects of the BPA methodological changes on the components of GDP were small. The change that provides improved estimates of foreign air carriers' expenses incurred in the United States had the largest effect; it reduced exports of other transportation services by $1.9 billion for 1992. This reduction in exports of services was partly offset by the effect of another methodological change, the inclusion of late reporters in the estimation of net "other private service receipts by U.S. parents from their foreign affiliates"; this change increased exports of "other private services" by $1.1 billion for 1992.

Changes in BPA methodology that affected receipts of factor income included the incorporation of the 1989 benchmark survey of U.S. direct investment abroad, which added $1.3 billion for 1992, and the incorporation of new source data on foreign commercial paper and certificates of deposit, which added $1.5 billion. Changes in methodology that affected payments of factor income included the incorporation of new data sources for interest payments on U.S. corporate bonds and U.S. Treasury securities, which added $3.9 billion for 1992; the new data sources included results from the U.S. Treasury Department's benchmark survey of foreign portfolio investment in the United States for 1989.

Because this annual NIPA revision - as usual - covered only 3 years, the BPA revisions were brought into the NIPA'S at the best level beginning with the estimates for 1990; the estimates for earlier years were not revised. As a result, there are discontinuities in the NIPA estimates. The extent of these discontinuities between 1989 and 1990 is quantified in table 6. For net exports of goods and services, the discontinuities are small; the change in GDP from 1989 to 1990 is understated by $0.3 billion. For net receipts of factor income, the discontinuities are also small; the change in GNP from 1989 to 1990 is overstated by $0.2 billion. However, the discontinuities in receipts and in payments of factor income, mainly in interest flows, are more substantial: The change in NIPA receipts of factor income from 1989 to 1990 is overstated by $4.2 billion, and the change in NIPA payments of factor income is overstated by $3.7 billion. (Revisions to the Nipa's for years prior to 1990 will be made in the next comprehensive NIPA revision.)

[TABULAR DATA OMITTED]

Deflation of net exports. - Several changes in the prices used for the deflation of net exports of services are introduced with this annual revision. The prices used to deflate exports and imports of air passenger fares are affected by two modifications in the way BEA uses BLS price indexes for these fares. Rapid changes in air passenger fares in recent years have made the BLS indexes, which are based on fares for a single week of the quarter specifically, the first week of the third month of the quarter), less representative of the average quarterly fare; as a result, beginning with 1990, this revision introduces a two-quarter moving average of the BLS indexes. In addition, the BLS indexes are now seasonally adjusted.

Also beginning with 1990, the implicit price deflator for final sales to domestic purchasers is used to deflate exports and imports of royalties and license fees, exports and imports of other private services, and BEA'S territorial adjustments to exports and imports of services. Previously, these components were deflated using the GDP implicit price deflator.

Beginning with the first quarter of 1993, the BEA composite index of foreign country consumer price indexes, adjusted for changes in exchange rates and weighted by country of destination, is used to deflate the "final" current quarterly estimates of the imports of travel services. Previously, this index was available only for annual revisions, and the current quarterly estimates were deflated with an index of foreign exchange rates. (For the "advance" and "preliminary" quarterly estimates of GDP, judgmental extrapolation will be used.)

Wage and salary accruals. - The methodology for estimating annual wage and salary accruals (a component of national income) was changed for 1992 to reflect the large amount of bonuses paid in 1992. For previous years, the revised estimates for wage and salary accruals and the revised estimates of wage and salary disbursements (a component of personal income) reflected the incorporation of BLS tabulations of wages and salaries paid to employees covered by State unemployment insurance (UI). These tabulations were used for both wage and salary series because BEA assumed either that (1) all wages earned during the year are paid during the same year or that (2) the amount of wages earned during a year and paid the next year is the same from year to year.

In early 1993, reports indicated that bonus payments earned by many employees in the securities industry in 1992 that typically would have been paid in early 1993 had instead been paid in late 1992. Because of the assumption that most of the bonuses earned in 1991 had been paid in early 1992, the acceleration created an imbalance between the accruals and disbursements series for 1992 and potentiauy for 1993. BEA took this fragmentary information into account in preparing the initial estimates for the fourth quarter of 1992; at that time, BEA estimated that $1.5 billion in bonus payments had been accelerated and increased fourth-quarter wage and salary disbursements by $6 billion (the $1.5 billion at an annual rate). This adjustment was necessary because the source data used to prepare the current quarterly wage and salary estimates - the employment, average weekly hours, and average hourly earnings from the BLS establishment survey - do not cover these types of bonus payments.

In June 1993, BEA received fourth-quarter 1992 UI tabulations, which cover all bonus payments. These UI data showed a substantially larger-than-expected increase in wage and salary disbursements. In addition, collections of personal income taxes showed an unusually strong increase at the end of 1992. To determine the cause of the large fourth-quarter increase in the UI data, BEA analyzed the quarterly tabulations for 1982-92 and concluded that about $20.0 billion of the fourth-quarter 1992 increase - $18.5 billion more than originally estimated - reflected an acceleration of bonus payments from early 1993. Consequently, the revised 1992 estimate of wage and salary accruals is based on wages and salaries from the UI tabulations that are adjusted to exclude the $20 billion in accelerated bonus payments.

Until June 1994, BEA Win receive no regular source data to indicate whether the acceleration of bonus payments that occurred in December 1992 will occur again in December 1993. For the 1993 estimates of wage and salary disbursements, BEA is assuming that the acceleration of bonus payments in 1992 was unusual. Next June, when the UI tabulations for the fourth quarter of 1993 become available, BEA will reassess this assumption.

Mortgage interest. - Annual and quarterly estimates of the mortgage interest component of net interest and of the deduction for mortgage interest used in deriving rental income of persons are estimated through the first quarter of 1993 using the product of mortgage debt outstanding, which is based on Federal Reserve Board data, and a BEA estimate of the effective rate of interest on mortgage debt. (These data are not available for the advance" and "preliminary" quarterly estimates.) The result is then used to extrapolate a benchmark estimate of mortgage interest paid based on data from the Census Bureau decennial survey of residential finance.

In last year's annual NIPA revision, a new procedure was introduced for estimating die effective rate of interest.(4) This procedure for the first time accounted for prepayments of all types of residential mortgages, but it was based on data that reflected only the prepayment pattern of mortgages insured by the Federal Housing Administration (FHA). In this year's annual revision, the FHA data are replaced with prepayments data from an industry source that are representative of all types of mortgages.

Imputed interest paid by life insurance carriers. - The revised estimates of net interest reflect a change in the source data used for estimating imputed interest paid by life insurance carriers, which is defined as their property income - that is, net monetary interest received plus dividends less monetary interest paid to policyholders. In the revised estimates for 1990 and 1991, monetary interest received by life insurance carriers - the largest component of their property income - is based on interest receipts from tabulations of income statements for life insurance carriers prepared by the American Council of Life Insurance. Previously, these estimates had been based on IRS tabulations of corporate tax return data on interest receipts for corporations classified as life insurance carriers. This change came about because in last year's and this year's annual revisions, the IRS tabulations of interest receipts have been unreasonably low, probably due to industry classification problems in the data.(5) (The estimates for the latest full year - in this case, 1992 - continue to be based on Federal Reserve Board flow-of-funds accounts assets data and corresponding interest rates estimated by BEA.)

Updated summary methodologies

Table 7 identifies the principal source data and estimating methods used to prepare the currentdohar estimates of the product- and income-side components of GDP, and table 8 identifies the principal source data and estimating methods used to prepare the constant-dollar estimates of the product-side components. These tables have been updated to reflect the methodological changes introduced in this annual NIPA revision.

Current-dollar estimates of GDP.-THE components in table 7 are as shown in the national income and product account (see table A, account 1), starting on the product side and proceeding to the income side. The subcomponents in table 7, with their 1992 dollar values, are grouped according to the methodology used to prepare them.

The column in table 7 for annual estimates covers the several annual estimates in the estimating cycle; the major differences in methodology as the estimates move through the three annual revisions to a benchmark revision are few enough to condense into the table. For example, for most goods in PCE (the first item on the product side), the table indicates one methodology for benchmark years and another for all other years.

The column for the quarterly estimates is a condensation in two respects. First, it refers to the advance estimate for the current quarter - that is, the estimate prepared in the first month following the end of the quarter. That one estimate, rather than all three of the current quarterly estimates, is described because more attention focuses on the first look" at the quarter. Second, even for the advance estimate, the column does not detail how many months of source data are available nor whether the data are subject to revision by the source agency.

Table 7 lists source data representing a variety of different economic measures - wages and salaries, premiums, expenses, interest rates, mortgage debt, tax collections, unit sales, housing stock, employment, and average price, to name a few. For most components, the source data are "value data"; that is, they embody both the quantity and price dimensions that are required for current-dollar estimates. In these cases, the methodology indicated in table 7 is the adjustment of the value data to derive estimates consistent with NIPA definitions and coverage.

When value data are not used in preparing an estimate, the table indicates the combination of data with separate quantity and price dimensions that is used to derive the required value estimate (as well as any major adjustments needed to derive estimates consistent with NIPA definitions and coverage). On the product side, a "physical quantity times price" method is used for several components. For example, the estimate for new autos is prepared as unit sales times average list price. An "employment times earnings times hours" method and variations of a "stock of assets/liabilities times an interest rate" method also are used for several components.

Some of the source data shown in table 7 for the annual estimates are used to interpolate and extrapolate the levels established by source data that are viewed as final, and all of the source data shown for the advance quarterly estimates are used to extrapolate the level of the preceding quarter. In addition to using indicator series, as is the case when specific source data are listed in the table, extrapolation and interpolation may be based on trends, as is the case when "judgmental trend" is listed in the table.(6)

Constant-dollar estimates of GDP. - Table 8 shows which of three methods is used to prepare constant-dollar estimates and indicates the source data with which it is implemented.7 The method used for most of GDP is deflation. In deflation, constant-dollar estimates are obtained by dividing the most detailed current-dollar components by appropriate price indexes, with the base period - at present, the year 1987 - equal to 100.

The other two methods, quantity extrapolation and direct base-year valuation, are similar to each other in that they both use quantity data. For quantity extrapolation, constant-dollar estimates are obtained by extrapolating the base-year current-dollar estimates in both directions from the base period (1987) by the quantity indicators. For direct base-year valuation, constant-dollar estimates are obtained by multiplying base-year prices by actual quantity data for each period.

The subcomponents in table 8 are the same as those shown in table 7, except where more detail is needed to highlight differences in methodology for constant-dollar estimates. For table 8, the distinction between annual and quarterly methodology is far less important than it is for the current-dollar methodology, and major differences between the annual and quarterly source data are noted within the individual entries.

Tables 7 and 8 and Appendix A follow.

[TABULAR DATA OMITTED]

(1.) The territorial adjustment converts merchandise trade from a BPA basis, which includes Puerto Rico and U.S. territories as part of the United States, to a NIPA basis, which includes Puerto Rico and U.S. territories as part of the "rest of the world." Similar adjustments are made for services and for receipts and payments of factor income.

(2.) See Karl Galbraith, "Federal Budget Estimates, Fiscal Year 1994," Survey 73 (April 1993): 46-52.

(3.) For a summary description of the index, see "The Comprehensive Revision of the U.S. National Income and Product Accounts: A Review of Revisions and Major Statistical Changes," Survey 71 (December 1991): 40. For additional detail, see Frank de Leeuw, "A Price Index for New Multifamily Housing," Survey 73 (February 1993): 33-42.

(4.) See "Annual Revision of the U.S. National Income and Product Accounts," Survey 72 (July 1992): 22-23.

(5.) In the IRS data, some life insurance carriers are classified in the casualty insurance industry and others are classified in industries not related to insurance because the IRS information is based on the filing of consolidated tax returns.

(6.) Extrapolation is a method of extending estimates from one period forward (or backward) in time to other periods. In simple terms, extrapolation applies a percent change - either the percent change in the indicator series or the percent change in the trend - to the level of the preceding (following) period. Interpolation is a method of filling in estimates between two periods. Interpolation applies a more complex mathematical formula - there are several in use - to preserve the pattern of the indicator series consistent with the level of the source data viewed as final.

(7.) With few exceptions, BEA does not prepare constant-dollar estimates of income measures because, unlike product measures, price indexes cannot be associated with them. Three exceptions are disposable personal income and, as presented in the "Business Cycle Indicators" section of the Survey, personal income and corporate profits, In these cases, the estimates are adjusted for price change by reference to the prices of the goods and services on which the income is spent. BEA derives constant-dollar net national product and national income by preparing constant-dollar estimates of consumption of fixed capital and of the nonfactor incomes and then subtracting these estimates from constant-dollar GNP.
COPYRIGHT 1993 U.S. Government Printing Office
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1993 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:1990 to 2nd quarter 1993
Publication:Survey of Current Business
Date:Aug 1, 1993
Words:9911
Previous Article:The business situation.
Next Article:The business situation.
Topics:

Terms of use | Copyright © 2016 Farlex, Inc. | Feedback | For webmasters