From the June 2000 SURVEY OF CURRENT BUSINESS
By Sherlene K.S. Lum, Brian C. Moyer, and Robert E. Yuskavage
IN this article, the Bureau of Economic Analysis (BEA) presents new estimates of gross product, or gross product originating (GPO), by industry for 1998 and revised estimates for 1947-97.(1)The estimates reflect the results of the recent comprehensive revision of the national income and product accounts (NIPA's), and they incorporate newly available source data and methodological changes for GPO by industry. (See the box "Gross Product Originating: Definition and Relationship to Gross Domestic Product.")
As part of this comprehensive revision of the GPO estimates, the double-deflation method for computing real GPO was extended to all industries, starting with the estimates for 1987 .(2) The extension of the double-deflation method required the development of time series of gross output and intermediate inputs for all industries, also starting with 1987. The availability of these integrated estimates of gross output, intermediate inputs, and GPO opens new possibilities for economic and industrial analysis. (See the box "GPO Estimates as a Set of Accounts.")
Using the double-deflation method improves the real GPO estimates in two ways. First, the GPO estimates can now incorporate more complete information on industry output, inputs, and prices, so the preparation of the estimates requires fewer assumptions about the relationships among industry outputs and inputs. Second, the source data on prices used to prepare the estimates of real GPO are now more consistent with the source data on prices used to prepare the estimates of real gross domestic product (GDP). As a result, the growth rates of real GPO are now more consistent with those of real GDP, thus increasing the reliability of the computed measures of industry contributions to real GDP growth.
|
Gross Product Originating: Definition and Relationship to Gross Domestic Product Gross product, or gross product originating (GPO), by industry is the contribution of each private industry and government to the Nation's output, or gross domestic product (GDP). An industry's GPO, or its "value added," is equal to its gross output (which consists of sales or receipts and other operating income, commodity taxes, and inventory change) minus its intermediate inputs (which consist of energy, raw materials, semifinished goods, and services that are purchased from domestic industries or from foreign sources).For the national income and product accounts
(NIPA's), GDP is measured as the sum of the expenditure components
and is benchmarked to the input-output accounts. Gross domestic
income (GDI) measures output as the sum of the costs incurred and
the incomes earned in the production of GDP. In concept, GDP and GDI
should be equal; in practice, they differ because their components
are estimated using largely independent and less-than-perfect source
data. The difference between GDP and GDI is the "statistical
discrepancy," which is recorded in the NIPA's as an
"income" component that reconciles GDI with GDP. BEA views
GDP as the more reliable measure of output, because the source data
underlying the estimates of expenditures are considered to be more
accurate
.(1)
Current-dollar GPO by industry is measured as the sum
of distributions by industry of the components of GDI that are
attributable to labor and property in the United States.
Consequently, the sum of the current-dollar GPO estimates also
differs from current-dollar GDP by the statistical discrepancy. In
the GPO estimates, the statistical discrepancy is included in the
GPO of private industries because of BEA's view that most of the
measurement problems with the components of GDI affect the GPO of
private industries rather than the GPO of general government or
government enterprises. |
Improvements from the comprehensive NIPA revision that were incorporated into this GPO revision include the recognition of business purchases and government expenditures for software, including own-account production of software, as investment and the incorporation of improved NIPA estimates of the real value of banking services into the gross output and GPO estimates for depository institutions.(3) Other major improvements that were incorporated include more recent data on industry intermediate inputs from the 1992 benchmark input-output (I-O) accounts and the 1996 annual I-O accounts and improved industry distributions of indirect business taxes.
Two other improvements are related to the conversion of NIPA income components from a company basis to an establishment basis: The conversion of corporate net interest from a company basis to an establishment basis to parallel the treatment of other corporate property-type income and the incorporation of a newly available 1992 employment matrix for the estimation of corporate property-type income on an establishment basis.
The major results from this GPO revision include the following.
|
A major element of this comprehensive revision of the gross product originating (GPO) by industry estimates is the development of an integrated set of estimates of gross output, intermediate inputs, and GPO for all industries, beginning with 1987. As a result, the GPO estimates are now a consistent set of industry production accounts that are more closely integrated with BEA's national income and product accounts and input-output accounts, thus opening new possibilities for analysis of structural change, industrial performance, and output per employee. In addition, the development of the new estimates of gross output enables the extension of the double-deflation method for computing real GPO to all industries, thus improving the reliability of the real GPO estimates and their consistency with estimates of real GDP. Gross output, intermediate inputs, and GPO are now generally derived from a consistent framework based on an industry production account that depicts how an industry uses both intermediate inputs and value-added inputs (GPO) to produce its gross output. The industry production account is a key component in international guidelines for the development of integrated economic accounts . Gross output and GPO are both important measures of
industry output. Many economists prefer to use gross output in
studies of industry production and output per employee because it
reflects the use of both primary and secondary
inputs
.(1)
Until recently, the lack
of gross output measures for many services industries limited the
analytical possibilities for the nonmanufacturing group. These gaps
in available gross output measures were largely attributable to the
lack of detailed source data for current-dollar gross output and for
price indexes, but they were also partly due to conceptual problems
in defining the output of some services industries--such as
depository institutions, which includes banking, and business
services, which includes computer services. As a result, some
analysts used GPO as an output measure because of GPO's
comprehensive coverage and widespread availability from the national
economic accounts of most countries. The availability of an
integrated set of industry production accounts that includes
estimates of gross output, intermediate inputs, and GPO opens new
possibilities for analysis of structural change, industry
performance, and output per employee. For example, analysts can
examine the substitution of purchased services for labor input
(outsourcing), the changes in cost-price markup patterns over time
and across industries, and the changes in output per employee for
nonmanufacturing industries and manufacturing industries. Several of
the nonmanufacturing industries for which estimates of gross output
and intermediate inputs are available for the first time--including
depository institutions and business services--have invested
significantly in information technology in recent years, and these
new estimates may help in assessing the impact of these investments
on output. Some of the problems that have limited the
availability of gross output estimates and the use of the
double-deflation method remain. For example, problems persist in
defining the gross output of many services industries, and
limitations remain in the methodology that uses cost indexes as
proxies for output-price indexes. Nevertheless, the real GPO by
industry estimates have been significantly improved, because fewer
assumptions are required about the relationship between gross output
and intermediate inputs. More importantly, a framework has been
established for incorporating the new producer price indexes (PPI's)
for services industries that began in 1995 and any additional PPI's
from the Bureau of Labor Statistics. These improvements in source
data will continue to improve the quality of the gross output and
real GPO estimates and will further enhance their analytical value
and their consistency with the estimates of GDP
. |
This article is presented in four parts. The first part discusses the relative performance of industries in terms of real growth rates, contributions to real GDP growth, shares of current-dollar GDP, and the composition of current-dollar GPO. The second part discusses the estimates of GPO prices and unit costs. The third part describes the changes in methodology and presentation, and the fourth part describes the methodology used to prepare the GPO estimates. The detailed GPO estimates for 1987-98 are presented in tables 1-14 at the end of the article. (See the box "Data Availability" on page 34.)
Industry Growth, Contributions, Shares, and Composition
The relative performance of particular industries or industry groups can be assessed by examining their real GPO growth rates, their contributions to real GDP growth, their shares of GDP, and the composition of their current-dollar GPO. Comparison of the growth rate of an industry's real GPO with that of real GDP indicates whether an industry is growing above or below the pace of U.S. economic growth. An industry's contribution to real GDP growth indicates the extent to which the industry is affecting the growth of real GDP. Changes in an industry's share of current-dollar GDP indicate whether the industry's use of the economy's resources is increasing or decreasing. Changes in the composition of an industry's current-dollar GPO indicate whether the labor and capital shares for the industry are changing.
Real GPO growth rates
Over the period 1977-98, the gross product for all industry groups increased. The increases ranged from 5.5 percent for wholesale trade to 1.2 percent for government; real GDP increased at an average annual rate of 3.1 percent (table A, chart 1). Within the transportation and public utilities industry group, the communications industry grew 5.7 percent. In 1977-87, the growth rates ranged from 4.9 percent for wholesale trade to 1.1 percent for construction. In 1987-98, wholesale trade again grew the fastest at 5.9 percent.
The revisions to the growth rates of real GPO for 1977-97 were small. The growth rate for private industries was revised up 0.3 percentage point, the same as the 0.3-percentage point upward revision to real GDP growth (table B). By industry group, the real GPO growth rate for transportation and public utilities was revised up the most (0.6 percentage point), and that for agriculture, forestry, and fishing was revised down the most (0.2 percentage point).
In 1992-98, real GDP increased at an average annual rate of 3.6 percent; private industries increased 4.2 percent, and government increased 0.5 percent. Among the industry groups, manufacturing increased 4.9 percent, as durable-goods manufacturing increased 8.1 percent and non-durable-goods manufacturing increased 0.8 percent. Excluding manufacturing, the largest increases were in transportation and public utilities (7.2 percent in communications and 5.1 percent in transportation), and the smallest increase was in agriculture, forestry, and fishing (1.5 percent).
At the industry level of detail, the changes in real GPO varied widely in 1992-98. Real GPO increased at an average annual rate of 5 percent or more in 18 industries. The increases were especially large in electronic and other electric equipment (20.3 percent) and industrial machinery and equipment (13.1 percent) in durable-goods manufacturing, partly reflecting the rapid growth in high-tech products, and in nondepository institutions (20.0 percent) and security and commodity brokers (17.9 percent) in FIRE.(4) Real GPO decreased in nine industries. The largest decreases were in instruments and related products (4.1 percent) and "other transportation equipment" (1.7 percent) in durable-goods manufacturing.
Beginning with this GPO revision, estimates of gross output and intermediate inputs by industry for 1987-98 are available for all industries and industry groups (table C). These expanded estimates present a more comprehensive picture of an industry's output. The growth rate of real gross output for an industry can be decomposed into the real growth rates of its primary factors of production (value-added inputs--labor and capital) and its secondary factors of production (intermediate inputs--including energy, raw materials, semifinished goods, and services--that are purchased from domestic industries and from foreign sources). For example, in 1987-98, real gross output for services grew at an average annual rate of 4.7 percent as a result of a 3.4-percent growth in real value-added from labor and capital inputs (real GPO) and a 6.8-percent growth in real intermediate inputs.(5) This decomposition demonstrates that most of the growth in real gross output for services resulted from the growth in real intermediate inputs rather than from the growth in real value-added inputs.
Contributions to real GDP growth
The contribution of an industry to real GDP growth indicates the extent to which the industry's growth affects the growth of real GDP. This contribution depends on both the industry's growth rate and its relative size.(6) In 1977-98, services and FIRE were the two largest contributors, accounting for 0.6 percentage point and 0.5 percentage point, respectively, of the 3.1-percent growth in real GDP (table D).
In 1987-98, services accounted for 0.6 percentage point and FIRE for 0.5 percentage point of the 3.0-percent growth in real GDP. In 1992-98, durable-goods manufacturing and services were the largest contributors; each accounted for 0.7 percentage point of the 3.6-percent growth in real GDP.
Shares of current-dollar GDP
An industry's share of current-dollar GDP is a good indicator of its relative size in the economy. (It is a better measure than an industry's share of real GDP, because shares of real GDP depend on the choice of the reference year.) Shares of current-dollar GDP can also be used to examine long-term trends in relative size, because these shares do not become distorted for years that are far from the reference year.
In 1992-98, the share of GDP accounted for by private services-producing industries increased from 61.2 percent to 64.7 percent, while the share accounted for by private goods-producing industries decreased from 24.0 percent to 23.3 percent and the share accounted for by government decreased from 14.2 percent to 12.6 percent (table E).(7) In the private services-producing industries, all industry groups increased their shares of GDP. The largest increases were in services, from 19.3 percent to 21.0 percent, and in FIRE, from 18.1 percent to 19.1 percent. The decrease in the share of private goods-producing industries was the net effect of a decrease in the share for manufacturing from 17.1 percent to 16.4 percent and of small decreases in the shares for mining and agriculture, forestry, and fishing that were partly offset by an increase in the share for construction from 3.7 percent to 4.3 percent. State and local government enterprises maintained its share, but the shares of all the other components of government declined.
In 1947-98, the share of private services-producing industries increased from 44.8 percent to 64.7 percent, while the share of private goods-producing industries fell from 41.9 percent to 23.3 percent (chart 2). The shares of all the private goods-producing industry groups except construction decreased. Among the private services-producing industry groups, services and FIRE more than accounted for the increase in share of GDP.
About two-thirds of the shift in GDP shares from private goods-producing industries to private services-producing industries over the 1947-98 period occurred in 1977-98. In 1947-77, the share of private services-producing industries increased only 6.9 percentage points, from 44.8 percent to 51.7 percent, and the share of private goods-producing industries decreased 9.1 percentage points, from 41.9 percent to 32.8 percent. In 1977-98, the share of private services-producing industries increased more rapidly than in 1947-77 --up 13.0 percentage points from 51.7 percent to 64.7 percent; this increase was mainly accounted for by services and FIRE. The share of private goods-producing industries decreased 9.5 percentage points from 32.8 percent to 23.3 percent.
The revisions to GPO had little effect on the industry-group shares of GDP. The largest downward revisions were to manufacturing, 0.2 per-centage point for 1987 and 0.4 percentage point for 1997, and to FIRE, 0.2 percentage point for 1987 and 0.5 percentage point for 1997. These downward revisions were offset by upward revisions to the statistical discrepancy, 0.4 percentage point (from a negative value to a positive value) for 1987 and 0.6 percentage point (from a more negative to a less negative value) for 1997, and to transportation and public utilities, 0.3 percentage point for 1997.
The revisions to the shares of GDP reflect the revisions to current-dollar GDP, which in turn, largely reflect the NIPA revisions to the components of gross domestic income (GDI) and to the industry distributions of these components. GDP was revised up for all years: The revisions were generally small before 1987, but starting with 1987, the revisions to current-dollar GDP increased substantially--from $50.2 billion (1.1 percent) for 1987 to $74.5 billion (1.2 percent) for 1992 and to $189.9 billion (2.3 percent) for 1997 (table F).
For all years, a major contributor to the upward revision to current-dollar GDP was the NIPA definitional change that recognized software as investment. The new treatment of software mainly affected the share of business services, which was revised up 0.3 percentage point for 1997, from 4.5 percent to 4.8 percent.
Composition of GPO
The changes over time in an industry's share of labor and capital reflect differences in the growth rates of the
components of current-dollar GPO.(8) For the total economy, the share of GDP that was accounted for by
labor decreased from 58.1 percent in 1987 to 57.7 percent in 1992 and to 57.3 percent in 1998, while the share
of capital dropped from 34.0 percent to 33.5 percent and then rose to 35.5 percent (table 4). The share of
indirect business tax and nontax liability increased from 7.8 percent to 8.1 percent and then dropped to 7.7
percent.(9)
Although the shares of labor and capital for the total economy did not change greatly in 1987-98, the shifts were substantial for manufacturing. The labor share of GPO for manufacturing decreased from 68.6 percent in 1987 to 63.9 percent in 1998, despite a decrease of less than 1.0 percent in full-time equivalent employment, while the capital share increased from 27.9 percent to 32.4 percent.(10) The decrease in labor share and the increase in capital share can be largely attributed to durable-goods manufacturing, which accounts for nearly three-fifths of manufacturing; the labor share decreased 6.2 percentage points, and the capital share increased 5.9 percentage points. Most of this shift occurred in 1992-98.
The shifts in the labor and capital shares in mining, in agriculture, forestry, and fishing, and in wholesale trade were also large. The biggest shift was in agriculture, forestry, and fishing, where the labor share increased from 27.4 percent to 37.0 percent, and the capital share decreased from 67.3 percent to 57.3 percent. In mining, the labor share decreased from 36.9 percent to 34.3 percent, and the capital share increased from 50.0 percent to 54.5 percent. In wholesale trade, the labor share decreased from 57.7 percent to 54.6 percent, and the capital share increased from 19.4 percent to 24.7 percent.
GPO Prices and Unit Costs
This section presents the estimates of GPO prices and unit costs by industry. First, it presents the GPO price indexes, including a discussion of contributions to GDP price change and a discussion of the relationship among GPO, gross output, and intermediate inputs price indexes. Second, it defines and presents the estimates of unit costs by industry and discusses their relationship to the GPO prices.
GPO prices
The GPO price index for an industry represents the price of its primary factors of production. For all industries and industry groups, the GPO price indexes are computed using a Fisher chain-type price-index-number formula. Because the GPO price indexes are computed using a Fisher formula, an industry's price index can be used in combination with its quantity index to separate changes in current-dollar GPO into price changes and quantity changes. For example, the 1987-98 average annual growth of 4.4 percent in current-dollar manufacturing GPO can be viewed as the product of a 1.4-percent growth in the manufacturing GPO price index (table G) and a 3.0-percent growth in the manufacturing GPO quantity index (table C)--that is, 1.044=1.014×1.030.
In 1987-98, the average annual change in the price of GPO for private services-producing industries (2.8 percent) was greater than, and the change in the price of GPO for private goods-producing industries (1.5 percent) was less than, the change in the GDP price index (2.6 percent). The slower growth in the index for private goods-producing industries partly reflected the rapidly declining prices of high-tech goods--including computers, digital telephone-switching equipment, and semiconductors.
In 1992-98, the GPO price index for private industries increased 1.7 percent, slightly less than the 1.9-percent increase in the GDP price index. The slower growth in the prices for private industries is partly attributable to a decline in manufacturing prices, as a decrease in durable-goods prices more than offset an increase in nondurable-goods prices. In addition, the price indexes for three other private industry groups that are involved in the distribution of goods to consumers either increased less than the GDP price index or decreased: Transportation and public utilities (1.3 percent), wholesale trade (-0.1 percent), and retail trade (0.2 percent). FIRE (2.9 percent) and services (3.5 percent) were among the industry groups for which the GPO price index increased more than the GDP price index.
Contributions to change.--GPO prices can be used to assess an industry's contribution to the change in GDP prices. This contribution depends on the industry's size relative to GDP and on the growth rate of its GPO price index.(11) In 1987-98, the largest contributors to the change in the GDP price index (2.6 percent) were services (0.8 percentage point) and FIRE (0.6 percentage point), both of which were large and rapidly growing industry groups (table H). Manufacturing contributed 0.3 percentage point, all of which was accounted for by nondurable-goods manufacturing.
Gross output prices and intermediate inputs prices.--Because price indexes for gross output and intermediate
inputs are now available for all industries and industry groups for 1987-98, the relationships among percent
changes in the chain-type price indexes for GPO, gross output, and intermediate inputs can be analyzed in
a similar manner as the relationships for the chain-type quantity indexes. For example, in 1987-98, the gross
output price index for wholesale trade grew at an average annual rate of 0.9 percent as a result of a 0.5-percent
growth rate in the price index for capital and labor inputs (GPO) and a 1.6-percent growth rate in the price
index for intermediate inputs.(12)
Unit costs
The GPO chain-type price index for an industry represents the price of its primary factors of production. Therefore, an industry's GPO price index can be used in combination with its current-dollar GPO components to assess each component's contribution to total industry labor and capital costs.(13) The GPO measures of unit costs are computed by dividing current-dollar GPO and its components by real (chained-dollar) GPO.(14) The resulting quotients are the GPO chain-type price index and the part of the price index that is associated with each component.(15) GPO unit-cost measures by private industry group are presented in table 14.(16) If the percent change in the unit cost for a component is greater than the percent change in the GPO price index, the relative importance of that component in the industry cost structure has increased.
The cost per unit of GPO for private industries increased 2.4 percent in 1987-98. Unit labor costs (compensation of employees per unit of GPO) increased 2.3 percent (table I). Unit costs for indirect business tax and nontax liability increased 2.1 percent, and unit costs for property-type income increased 2.7 percent. The larger increase in the unit costs for property-type income indicates that capital costs became a larger part of GPO unit costs during the period--that is, the return to capital per unit of gross product increased.
In 1987-98, unit labor costs declined in mining, durable-goods manufacturing, and wholesale trade, and they increased in all the other private industry groups. In agriculture, forestry, and fishing, in FIRE, and in services, the increases in unit labor costs were larger than the increases in total unit costs.
In 1992-98, the larger declines or smaller increases in unit labor costs relative to total unit costs were in industry groups engaged in the production and distribution of goods to consumers. In transportation and public utilities, unit labor costs increased 0.5 percent, less than the 1.3-percent increase in total unit costs. In wholesale trade, unit labor costs declined 1.0 percent, while total unit costs declined only 0.1 percent. In retail trade, unit labor costs declined 0.6 percent, while total unit costs increased 0.2 percent.
Changes in Methodology and Presentation
This section presents changes in methodology and presentation. First, it discusses changes to methodology resulting from changes in NIPA sources, GPO methodology, and GPO source data. Second, it discusses changes in the presentation of the GPO estimates.
NIPA sources
Most of the definitional and statistical changes that were incorporated in the recent comprehensive revision
of the NIPA's had little or no effect on the GPO estimates. Two notable exceptions were the recognition of
business purchases and government expenditures for software as investment and the redefinition of the value
of imputed services of regulated investment companies.(17)
Beginning with estimates for 1959, the NIPA's now recognize business and government expenditures, including own-account production of software, as fixed investment. This change, which resulted in an upward revision to GDP, primarily affected property-type income. In the GPO estimates, the reclassification of software also resulted in a decrease in intermediate inputs, and the recognition of own-account software resulted in an increase in gross output. The industries that were most affected by the new treatment of software were wholesale trade, retail trade, FIRE (depository institutions, nondepository institutions, and insurance carriers), and services (business services, including computer and data processing services).
In addition, beginning with the NIPA estimates for 1959, imputed services of regulated investment companies (mutual funds) were redefined to equal operating expenses rather than net property income received. This redefinition affected both GDP and the net interest component of GDI, increasing both in some years and decreasing both in some years. This revision affected the GPO property-type income of several industries, primarily the regulated investment companies component of holding and other investment companies.
The GPO revision also incorporated statistical changes that were made in the comprehensive NIPA revision.(18) Two changes--the incorporation of improved estimates of the real value of unpriced banking services and the incorporation of geometric-mean-type consumer price indexes--affected real GPO through the prices of both gross output and intermediate inputs. The NIPA price index for banking services was used to develop the gross output estimates for depository institutions and the intermediate inputs price indexes for banking services. The geometric-mean-type consumer price indexes were used in the computation of the estimates of real gross output for many of the private services-producing industries, particularly for retail trade.
The GPO components that were most affected by the comprehensive NIPA revision were property-type income and the statistical discrepancy, which is included in the aggregate for private industries. Revisions to compensation of employees and to indirect business taxes and nontax liability were generally small for all industry groups. For 1987, the revisions to the current-dollar GPO estimates were largely to transportation and public utilities, wholesale trade, and the statistical discrepancy (table F). For 1992, they were largely to manufacturing and to services, and in 1997, they were largely to transportation and public utilities, to services, and to the statistical discrepancy.
GPO methodology
Extension of the double-deflation method.--In the previously published GPO estimates, the double-deflation method was used to compute real GPO by industry for 51 of the 63 industries for which it is the preferred method.(19) In the double-deflation method, estimates of gross output and intermediate inputs are used in the calculation of real GPO. Generally, double deflation is the conceptually preferred method because it requires fewer assumptions about the relationships among gross output and intermediate inputs.
In this GPO revision, the double-deflation method has been extended to the remaining 12 industries, starting with the estimates for 1987.(20) The methodology used to compute the gross output estimates is generally the same as that used for the previously published estimates. In 1987-98, real gross output for the 12 industries increased at an average annual rate of 5.3 percent, compared with the 3.4-percent increase for "all industries" (table 10), and real GPO for the 12 industries increased 4.4 percent.
New gross output estimates.--The extension of the double-deflation method required the development of new gross output estimates for 12 industries. Several aspects of the methodology deserve special attention.
|
The summary estimates of gross product by industry that are presented in this article and more detailed estimates for 1947-98 are available on BEA's Web site; go to www.bea.doc.gov and click on "Industry and Wealth data." These estimates are also available online to subscribers to STAT-USA's Economic Bulletin Board and Internet services (call 202-482-1986, or go to www.stat-usa.gov ). In addition, the estimates will be available on the following diskettes for $20.00 each in late June. Gross Product by Industry, 1947-98--product number NDN-0256 Gross Output by Detailed Industry, 1977-98--product number NDN-0257 Manufacturing Industry Shipments, 1977-98--pro-duct ID number NDN-0258 Manufacturing Product Shipments, 1977-96--pro-duct ID number NDN-0259 To order, call the BEA Order Desk at 1-800-704- 0415 (outside the United States, call 202-606-9666) |
Improved gross output estimates.--Several types of improvements were made to the previously published gross output estimates, generally beginning in 1987. For agricultural services, forestry, and fishing, for trucking and warehousing, and for telephone and telegraph, the use of additional component detail improved the gross output estimates. For electric, gas, and sanitary services, an improved indicator series and a price index from the Department of Energy were introduced to reflect changes in the structure of the gas utility industry. For security and commodity brokers, the improved NIPA adjustment for excluding interest from gains on trading accounts was introduced, and for insurance carriers, corrections were made to the previous method for computing the current-dollar indicator series for the property and casualty insurance component.
Treatment of corporate net interest.--For the estimates of current-dollar GPO, NIPA income components
often need to be converted from a company basis to an establishment basis. As part of this GPO revision, the
estimates of corporate net interest, starting with 1987, were converted from a company basis to an
establishment basis; previously, corporate profits before tax and corporate capital consumption allowances
were converted from a company basis to an establishment basis, but the estimates of corporate net interest
were not.(22)
In general, this change did not significantly affect the current-dollar GPO for most industries. The industries that were affected the most were nondepository institutions, security and commodity brokers, insurance carriers, and holding and investment offices. Although the revisions to net interest were relatively large for these industries, the adjustments from a company basis to an establishment basis are consistent in sign and magnitude with the adjustments for corporate profits before tax in which interest received is recorded as a receipt.
Improved industry distributions of indirect business tax and nontax liability.--Starting with 1977, the industry distributions of selected components of indirect business tax and nontax liability were updated.(23) Industry distributions of State and local general sales and gross receipts taxes were updated to incorporate data from an expanded sample of States and the District of Columbia. The industry and State distributions of other selected State and local taxes were updated with Census of Governments data for States, supplemented as needed with other State data. Selected taxes and payments include severance taxes, alcohol excise taxes, occupation and business license fees, other selective taxes, documentary and stock transfer taxes, public utility excise taxes, and motor vehicle license fees. Benchmark distributions were developed for 1977, 1982, 1987, 1992, and 1997.
GPO source data
The GPO estimates also incorporated new source data. The 1992 benchmark I-O accounts provided updated benchmarks for deriving the gross output estimates. Industry compositions of intermediate inputs were incorporated from the 1992 benchmark I-O accounts and the 1996 annual I-O accounts. The 1992 Census Bureau employment matrix provided an updated relationship among companies and establishments that was used to convert profits before tax, corporate net interest, and corporate capital consumption allowances from a company basis to an establishment basis.
The extension of the double-deflation method required the use of additional source data. For example, for depository and nondepository institutions, data from the Federal regulatory agencies, federally sponsored credit agencies, and BLS were used to develop gross output estimates. For business services, social services, membership organizations, and other professional services, extensive use was made of data from the Census Bureau's services annual survey. The gross output estimates of Federal Government enterprises and State and local government enterprises incorporated data from the U.S. Postal Service and the Census Bureau.
In addition, estimates from the comprehensive NIPA revision, new and revised estimates from Census Bureau annual surveys, and newly available producer price indexes from BLS were incorporated into the GPO estimates. Data from the Census Bureau on manufacturers' shipments, inventories, and orders were used to prepare the estimates of industry shipments for 1997-98.
Changes in presentation
Two major changes in presentation were made for this GPO revision. First, the reference year for chain-type measures was updated from 1992 to 1996; this change did not directly affect the percent changes in the price or quantity indexes or in chained dollars, because these changes are measured with chain-type indexes. Second, the presentation of the estimates of gross output and intermediate inputs was changed to reflect the extension of the double-deflation method to all industries.
Methodology
This section discusses the methodology, including the source data and estimating procedures, that is used to prepare the estimates of current-dollar and real GPO. Previously published tables that summarize the methodology have been updated to incorporate the major changes introduced in this GPO revision.
Current-dollar GPO estimates
The current-dollar GPO estimates are prepared as the sum of the distributions by industry of the components of GDI (see the box "Gross Product Originating: Definition and Relationship to Gross Domestic Product").(24) This section describes the methodology for distributing the current-dollar estimates of these components.
For most components of GDI, the estimates are based on source data that provide industry distributions on either a company basis or an establishment basis. Only the estimates with distributions based on establishment data can be used directly to calculate industry GPO. The industry distributions of the components that are estimated from Internal Revenue Service (IRS) tabulations of business tax returns, which are on a company basis, must generally be converted to an establishment basis; this conversion is particularly necessary for companies whose establishments are classified in several industries. For the components for which the source data do not provide industry distributions, BEA has developed establishment-based industry distributions from related sources (table J).
To convert company-based data to an establishment basis, Census Bureau matrixes of the employment of establishments of corporations are used. These matrixes present employment of establishments cross-classified by the company-industry classifications assigned by the IRS in preparing tabulations of corporate tax returns and by the establishment-industry classifications assigned by the Census Bureau in the economic censuses. These matrixes are used to adjust the data on corporate profits before tax, corporate net interest, and corporate capital consumption allowances from a company basis to an establishment basis. For integrated petroleum companies, the results of applying these matrixes are supplemented by tabulations of net income and depreciation of energy companies on an establishment basis from the Department of Energy financial reporting system. These matrixes are also adjusted to reflect publicly available information about large mergers, acquisitions, or changes in company diversification that have occurred since 1992, the year covered by the latest matrix.
Real GPO estimates
For this GPO revision, the 1987-98 GPO chain-type quantity indexes for all industries and industry groups except private households and general government are calculated using the double-deflation method.(25) The GPO chain-type quantity indexes for 1977-87 are calculated using one of three alternative methods--double deflation, extrapolation, or direct deflation--depending on the availability and reliability of source data.(26) In the double-deflation method, separate estimates of gross output and intermediate inputs are combined in a Fisher chain-type quantity-index-number formula to generate chain-type quantity indexes of GPO.(27) The double-deflation method is the preferred method for computing GPO chain-type quantity indexes because it requires fewer assumptions about the relationships among gross output and intermediate inputs. The real, or chained (1996) dollar, GPO estimates for each industry and industry group are derived as the product of the GPO chain-type quantity index (divided by 100) and the corresponding 1996 current-dollar GPO value.
Table K provides a summary description of the principal source data used to prepare the detailed estimates of current-dollar gross output and gross-output prices for 1987-98. For 1977-87, this table is applicable only for those industries for which the double-deflation method was used to compute the GPO chain-type quantity indexes.
Except for farms and nonfarm-housing services, current-dollar intermediate inputs are derived by subtracting current-dollar GPO from current-dollar gross output.(28) Detailed current-dollar estimates of intermediate inputs are derived in four steps: (1) The input compositions for 1977, 1982, 1987, 1992, and 1996 are derived from BEA's benchmark and annual I-O accounts; (2) the input compositions for 1978-81, 1983-86, 1988-91, and 1993-95 are estimated by interpolating the detailed compositions from 1977, 1982, 1987, 1992, and 1996; (3) the imported and domestically produced shares of each detailed input for 1977-96 are estimated; and (4) the input compositions for 1997-98 are estimated, primarily based on the 1996 composition.
Detailed prices are prepared separately for domestically produced intermediate inputs and for imported intermediate inputs. Prices for domestically produced intermediate inputs are largely based on the detailed prices associated with the gross output estimates, as shown in table K. For example, the same quality-adjusted semiconductor price index is used for selected semiconductor products in the deflation of both gross output and domestically produced intermediate inputs. The prices of imported intermediate inputs are developed from a variety of sources, primarily from BLS import price series, and are the same as those used for the NIPA estimates of imports.
Tables 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, and 14 follow.
|
Robert E. Yuskavage, Chief of the GDP by Industry Branch of the Industry Economics Division (IED), supervised the preparation of the estimates. Sumiye O. Okubo, Associate Director for Industry Accounts, and Ann M. Lawson, Chief of the Industry Economics Division, provided overall guidance. Felicia V. Candela, Thea C. Graham, Peter J. Lee, Tameka R. Lee, Sherlene K.S. Lum, Kimberly A. Mourey, Brian C. Moyer, and Robert A. Sylvester prepared the estimates. Other IED staff, including Mark A. Planting, assisted in the use of the input-output accounts. Staff members from the National Income and Wealth Division and the Government Division contributed to the development of the estimates. Staff members from the Regional Economic Analysis Division participated with IED staff in a joint project to reengineer the estimation of indirect business taxes and nontax liability by industry. Special thanks to Alan C. Lorish, Jr., Chief of the Computer Systems and Services Division, and to members of his staff--particularly Stephen P. Holliday and Douglas J. Klear--for their assistance in reengineering the data-processing application used for this GPO revision. |
2. Double-deflation is the preferred method for computing industry real GPO because it requires fewer assumptions about the relationships among gross output and intermediate inputs. It is not the preferred method for the three industries--private households, Federal general government, and State and local general government--for which gross output and GPO are equivalent .
3. See Eugene P. Seskin, "Improved Estimates of the National Income and Product Accounts for 1959-98: Results of the Comprehensive Revision," Survey 79 (December 1999): 15-39.
4. Chain-type quantity indexes of GPO are shown in table 5, and estimates of real (chained-dollar) GPO are shown in table 6. For information about the computation of the real GPO estimates, see the box "Computation of the Chain-Type Quantity Indexes for Double-Deflated Industries" in Robert E. Yuskavage, "Improved Estimates of Gross Product by Industry, 1959-94," Survey 76 (August 1996): 142.
5. The contributions to growth of these components are a function of their relative size and of changes in their relative size over time as a result of factors such as increased outsourcing .
6. See the box "Using Chained-Dollar Estimates for Computing Contributions to Economic Growth: A Cautionary Note" in Lum and Moyer, "Gross Product by Industry," 24-25.
7. Private goods-producing industries consist of agriculture, forestry, and fishing; mining; construction; and manufacturing. Private services-producing industries consist of transportation and public utilities; wholesale trade; retail trade; finance, insurance, and real estate; and services.
8. The labor share of production is approximated using compensation of employees--which consists of wage and salary accruals, employer contributions for social insurance, and other labor income. The capital share of production is approximated using property-type income--which consists of corporate profits and proprietors' income with inventory valuation adjustment, rental income of persons, net interest, private capital consumption allowances (CCA's), business transfer payments, the current surplus of government enterprises less subsidies and government consumption of fixed capital. Proprietors' income is included in property-type income; however, an unknown portion of proprietors' income represents a labor share of production. Indirect business tax and nontax liability (primarily sales, property, and excise taxes) is not included in property-type income, because it is the part of the pretax return to capital that accrues to government rather than to business. Private CCA's are used to calculate GPO because detailed industry estimates of private consumption of fixed capital are not available. The GPO by industry estimates are not affected by this difference in the valuation of depreciation because the estimates of private CCA's are deducted in computing corporate profits before tax and proprietors' income and are derived from the same source data.
9. The shares do not add to 100.0, because of the statistical discrepancy, which was 0.1 percent in 1987, 0.7 percent in 1992, and -0.5 percent in 1998 .
10. For some analytical purposes, such as multifactor productivity analysis, the labor and capital shares of gross output are more appropriate than the labor and capital shares of GPO. For most industries, particularly for manufacturing, the labor and capital shares of GPO are larger than the labor and capital shares of gross output, because gross output also includes intermediate inputs. For example, labor's share of manufacturing gross output was 22.9 percent in 1998, whereas labor's share of manufacturing GPO was 63.9 percent .
11. For a description of the calculation of these contributions, see the reference in footnote 6. For price calculations, the procedure described in the reference was modified to replace the chain-type quantity index with the chain-type price index.
12. The contributions to growth of these components are a function of their relative size and of changes in their relative size over time as a result of factors such as increased outsourcing.
13. See the section "Composition of GPO."
14. Current-dollar cost per unit of real GPO equals the GPO price index divided by 100 .
15. GPO unit-cost measures are not derived from separate price indexes for labor and capital; instead, these measures are derived by proportionally distributing an industry's GPO price index to the components of its current-dollar GPO .
16. These unit-cost measures differ from the unit-labor-cost and unit-nonlabor-cost series published by the Bureau of Labor Statistics (BLS). See Department of Labor, Bureau of Labor Statistics, BLS Handbook of Methods, Bulletin 2490 (Washington, DC: U.S. Government Printing Office, April 1997).
17. Brent R. Moulton, Robert P. Parker, and Eugene P. Seskin, "A Preview of the 1999 Comprehensive Revision of the National Income and Product Accounts: Definitional and Classificational Changes," Survey 79 (August 1999): 7-20.
18. Brent R. Moulton and Eugene P. Seskin, "A Preview of the 1999 Comprehensive Revision of the National Income and Product Accounts: Statistical changes," Survey 79 (October 1999): 6-17.
20. The 12 industries are water transportation, transportation services, depository institutions, nondepository institutions, other real estate, holding and other investment offices, business services, social services, membership organizations, other (professional) services, Federal Government enterprises, and State and local government enterprises.
21. For this industry, the year-to-year changes in GPO largely reflect changes in the property-type income. In many years, these changes have been very large, making the estimation of real GPO extremely difficult. As a result, for the GPO estimates, an indirect method is used to estimate gross output in which current-dollar intermediate inputs are estimated by extrapolating the intermediate inputs for this industry from the 1992 benchmark I-O accounts by BEA wages and salaries for this industry. A preliminary study of the sources of these large changes indicates classification problems in the underlying source data .
22. As a result of this change in methodology, separate estimates of corporate net interest and of noncorporate net interest are now available as detailed components of GPO .
23. Statistical revisions to control totals for selected components of indirect business taxes for State and local government that were introduced in the comprehensive revision of the NIPA's slightly affected the industry distribution of these taxes.
24. Private capital consumption allowances (CCA's) are used to calculate GPO because detailed industry estimates of private consumption of fixed capital are not available. The GPO by industry estimates are not affected by this difference in the valuation of depreciation because the estimates of private CCA's are deducted in computing corporate profits before tax and proprietors' income and are derived from the same source data.
26. For a description of these alternative methods, see Yuskavage, "Improved Estimates," 143-149.
28. For farms and nonfarm housing services, complete and consistent gross output and intermediate inputs series are available from the NIPA's .
