Gross product by industry, 1947-96.

Lum, Sherlene K.S. ; Yuskavage, Robert E.

This article presents new estimates of gross product, or gross product originating (GPO), by industry for 1995-96 and revised estimates of current-dollar GPO for 1947-94 and of real (chained-dollar) GPO for 1977-94.(1) The new and revised estimates incorporate the final results of the comprehensive revision of the national income and product accounts (NIPA'S) released in May 1997; the estimates for 1993-96 also incorporate the results of the annual NIPA revision released in July 1997 and newly available source data.(2) In addition, two new tables have been added to present the estimates of gross output and intermediate inputs.

For the first time since 1988, GPO estimates are available for the most recent complete year; this is the latest step in a continuous GPO improvement program. The improvements that were introduced into the GPO estimates last year included the improved chain-type measures of real GPO and a quality-adjusted BEA price index for selected semiconductor products.(3) Future improvement efforts will focus on integrating the GPO estimates with the benchmark input-output (I-O) accounts and with other BEA industry estimates.(4)

The first part of this article discusses the relative performance of industries for 1993-96 in terms of real growth rates, industry shares of current-dollar gross domestic product (GDP), and the composition of current-dollar GPO. The second part discusses the revisions to the estimates, and the third part describes methodology. The fourth part briefly describes the revisions to the historical estimates, and the fifth part discusses the estimates of gross output and intermediate inputs. Tables following the text present the new and revised estimates for 1993-96 for detailed industries, including the new tables for gross output and intermediate inputs, and new and revised estimates for 1947-96 for industry groups.

Industry Growth, Shares, and Composition

Comparisons of real gross product growth rates and of shares of GDP across industries show the relative performance of particular industries or industry groups. For example, comparisons can be made of the relative growth rates of real gross product among industries and of their contributions to the growth rate of the economy as a whole. A comparison of the share of current-dollar GDP accounted for by the gross product of an industry over time indicates whether that industry's claim on the economy's resources is increasing or decreasing. The composition of an industry's current-dollar GPO indicates whether the labor and capital shares for that industry are changing over time.

Real growth rates

Real GDP increased at an average annual rate of 2.6 percent in 1992-96; private industries increased 3.2 percent, and government showed minimal growth (table 1). The real gross product of all private industry groups except agriculture, forestry, and fishing increased; the increases ranged from 8.2 percent in durable goods manufacturing to 2.3 percent in finance, insurance, and real estate (FIRE).

[TABULAR DATA NOT REPRODUCIBLE IN ASCII]

By detailed industry, 19 industries recorded average annual increases in real gross product of 5 percent or more, and 3 industries in particular stand out: Electronic and other electric equipment (21.9 percent) and industrial machinery and equipment (14.4 percent) in durable goods manufacturing and security and commodity brokers (16.8 percent) in FIRE.(5) Real growth declined for 10 industries; the largest decreases were also in durable goods manufacturing: Instruments and related products, down 8.3 percent, and "other transportation equipment." down 5.8 percent. Instruments and related products has shown annual declines over the whole period.

In 1995, real GDP slowed to a 2.0-percent increase from a 3.5-percent increase in 1994. The growth in private industry GPO slowed to a 2.7-percent increase from a 4.5-percent increase. All the private industry groups except FIRE and services grew at a slower rate in 1995 than in 1994. In 1995, the fastest growing industry groups were durable goods manufacturing (8.9 percent), nondurable goods manufacturing (4.1 percent), and mining (5.8 percent). Agriculture, forestry, and fishing declined 6.5 percent.

In 1995, among the detailed industries, four of the five fastest growing industries were in manufacturing, and the other was in FIRE. Among durable goods industries, industrial machinery and equipment increased 25.1 percent, and electronic and other electric equipment increased 20.4 percent. Among nondurable goods industries, food and kindred products increased 13.5 percent, and petroleum and coal products increased 16.4 percent. Insurance carriers in FIRE grew 15.1 percent. Two industries had decreases of over 10 percent: Paper and allied products, down 13.0 percent, and farms, down 12.7 percent.

In 1996, real GDP growth accelerated to 2.8 percent from 2.0 percent. The acceleration was mainly accounted for by agriculture, forestry, and fishing; construction; transportation and public utilities; wholesale trade; and retail trade. All the industry groups except mining, nondurable goods manufacturing, and government increased. Wholesale trade grew the fastest (7.8 percent), followed by durable goods manufacturing (7-4 percent); agriculture, forestry, and fishing grew the slowest (0-3 percent). Although durable goods had the second fastest growth in 1996, it grew slower than in 1995- Although agriculture, forestry, and fishing increased only slightly in 1996, it had fallen sharply in 1995.

By detailed industry, two of the four fastest growing industries were in durable goods manufacturing: Electronic and other electric equipment increased 23.8 percent, and industrial machinery and equipment increased 13.1 percent. These two industries were also two of the four fastest growing industries in 1995. The two other fast-growing industries in 1996 were transportation by air (18.4 percent) and metal mining (14.8 percent). Two industries declined more than 10 percent: Oil and gas extraction, down 10.6 percent, and instruments and related products, down 10.1 percent.

Data Availability

This article presents the summary estimates of gross product by industry. These estimates and more detailed estimates for 1947-96 are available on the Internet on BEA'S home page at [is less than] http://wwwbea.doc.gov [is greater than]. They are also available online to subscribers to STAT-USA'S Economic Bulletin Board (EBB) (call 202-482-1986, or go to [is less than] http://wwwstat-usa.gov [is greater than]).

In addition, the following estimates win be available from BEA on diskettes:

* Gross Product by Industry: 1947-96, product number NDN-0174, price $20.00.

* Gross Output by Detailed Industry: 1977-96, product number NDN-0175, price $20.00.

* Manufacturing Industry Shipments: 1977-96, product number NDN-0176, price $20.00.

* Manufacturing Product Shipments: 1977-95, product number NDN-0177, price $20.00.

To order using Visa or MasterCard, call the BEA Order Desk at 1-800-704-0415 (outside the United States, call 292-606-9666). To order by mail, send ;& check payable to "Bureau of Economic Analysis, BE-53" to BEA Order Desk, Bureau of Economic Analysis, BE-53, U.S. Department of Commerce, Washington, DC 20230.

Contributions to real GDP growth. Growth rates alone do not indicate the extent to which industries contribute to the growth of real GDP; the contribution also depends on the industry's relative size. In 1992-96, durable goods manufacturing was the largest contributor, o.8 percentage point, to the 2.6-percent growth in real GDP; services was the next largest, 0.5 percentage point (table 2).(6) For 1995, durable goods manufacturing contributed 0.9 percentage point to the growth in real GDP, and services contributed 0.7 percentage point. For 1996, durable goods manufacturing and services each contributed 0.7 percentage point.

Table 2.-- Contributions to Percent Change in Real Gross Domestic Product

 Average
 1993 1994 1995 1996 1992-96(1)

Percent change:
 Gross domestic product 2.3 3.5 2.0 2.8 2.6

Percentage points:
 Private industries 2.3 3.8 2.4 2.5 2.8

 Agriculture, forestry, and -.2 .3 -.1 .0 .0
 and fishing
 Mining .1 .1 .1 -.1 .0

 Construction .1 .2 .1 .2 .1

 Manufacturing .6 1.4 1.2 .7 1.0
 Durable goods .6 1.0 .9 .7 .8
 Nondurable goods .0 .5 .3 .0 .2

 Transportation and public .4 .5 .1 .2 .3
 utilities
 Transportation .1 .2 .0 .1 .1
 Communications .2 .1 .0 .0 .1
 Electric, gas, and
 sanitary services .1 .2 .1 .1 .1

 Wholesale trade .2 .5 .1 .5 .3

 Retail trade .4 .5 .3 .4 .4

 Finance, insurance, and real
 estate .4 .4 .5 .4 .4
 Services .4 .5 .7 .7 .5

 Statistical discrepancy(2) .1 -.6 -.6 -.4 -.4

 Government .0 .0 .0 .1 .0

 Not allocated by industry(3) .0 -.4 -.4, .2 -.2

1. Annual rate. 2. Equals GDP measured as the sum of expenditures less gross domestic income. 3. Equals GDP less the so" discrepancy and the sum of GPO of the detailed industries.

NOTE. -- For information on the calculation of the contributions to parent change, see footnote 5 in text

Shares of current-dollar GDP by industry

Shares in current-dollars are a better indicator of an industry's relative size in the economy in any one period than shares in real dollars. Industry shares in real dollars, whether measured in chained dollars or in constant dollars, are dependent on the choice of the base period and therefore are not good indicators of relative size.

The share of current-dollar GDP that was accounted for by private goods-producing industries increased from 24.0 percent in 1992 to 24.6 percent in i996, and the share accounted for by private services-producing industries increased from 61.3 percent to 63.1 percent (table 3).(7) The increase for private services-producing industries was mostly accounted for by "services" and by FIRE; the share of services rose 1.0 percentage point, and that of FIRE rose 0.6 percentage point. In the FIRE group, the shares of security and commodity brokers (0-4 percentage point) and insurance carriers (0-5 percentage point) increased the most.

[TABULAR DATA NOT REPRODUCIBLE IN ASCII]

The share of current-dollar GDP that was accounted for by government fell from 14.0 percent to 13.0 percent; the decline was concentrated in Federal general government (table 7)(8)

Composition of GPO

Current-dollar GPO is measured as the sum of costs incurred and incomes earned in production in each industry; it is equal to gross domestic income, whose components can be grouped into categories that approximate shares of labor and capital. Differences over time and among industry groups in shares of labor and capital can thus be observed using these approximations. The labor share of production can be 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 (property-type income) can be approximated using the remaining components of GPO except indirect business tax and nontax liability, which is excluded because it can be viewed as a part of the pre-tax return to capital that accrues to government rather than to business.(9)

For the total economy, the share of GDP that was accounted for by compensation of employees decreased slightly, from 58.4 percent in 1992 to 58.o percent in 1996, while the share of property-type income increased from 32.8 percent to 34.9 percent (table 4). The labor and capital shares Of GPO, and the degree of change in these shares, varied among industry groups. The labor share of manufacturing GPO declined 5.0 percentage points over the period despite increases in full-time equivalent employment and compensation per full-time equivalent employee.(10)

The decline in labor's share of manufacturing GPO continued a trend that started in 1980. After reaching a postwar peak of 74.6 percent in 1980, labor's share of manufacturing GPO declined nearly 12 percentage points to 63.0 percent in 1996, the lowest share since 1950 (chart 1). By contrast, labor's share of GPO for all industries (including government) declined much less -- from 60.0 percent in 1980 to 57.6 percent in 1996 -- and this decline is more than accounted for by the decline in manufacturing.(11) For nonmanufacturing industries (not shown on the chart), labor's share of GPO increased slightly, from 49.6 percent in 198o to 50.9 percent in 1996.

The shifts in the labor and capital shares in mining and in agriculture, forestry, and fishing were also relatively large (table 8). In mining, the labor share decreased from 35.2 percent in 1992 to 29.7 percent in 1996, and the capital share increased correspondingly. In agriculture, forestry and fishing the labor share increased from 27.3 percent to 30.5 percent, and the capital share decreased correspondingly.

Revisions to the GPO Estimates

Table 5 presents revisions to current-dollar GPO and to real GPO growth rates by industry group for 1993 and 1994.

Current-dollar estimates

The revisions to current-dollar GPO largely reflect the effects of the annual and comprehensive NIPA revisions on the components of gross domestic income and, to a lesser extent, on the industry distributions of these components. Relatively large revisions to several of the income components mostly offset one another. Mining was revised up $5.6 billion for 1993, reflecting revisions to corporate profits before tax, and it was revised up $4.8 billion for 1994, reflecting revisions to corporate capital consumption allowances. A large upward revision of $12.9 billion to nondurable goods manufacturing for 1994 was primarily due to a revision to corporate profits before tax in the chemicals and allied products industry. In FIRE, a large downward revision of $20.4 billion to holding and other investment offices for 1994 was somewhat offset by an upward revision of $9.0 billion to security and commodity brokers.

Real growth rates

The revisions to real GPO growth rates primarily reflect the revisions to current-dollar GPO, but they also reflect the incorporation of new and revised source data for gross output and prices and the incorporation of revised data on the composition of gross output from the 1992 benchmark input-output (I-O) accounts. By industry, the revisions to real GPO growth rates for both 1993 and 1994 were generally small. However, the growth rate for mining GPO for 1993 was revised up 6.2 percentage points, and the growth rate for agriculture, forestry, and fishing GPO for 1994 was revised up 4.4 percentage points. The upward revision to mining was in off and gas extraction and reflected a revision to corporate profits. The upward revision to the industry group agriculture, forestry, and fishing was in "agricultural services, forestry and fishing" and reflected new source data for gross output. In transportation and public utilities for 1994, a downward revision of 11.2 percentage points to radio and television broadcasting was largely offset by an upward revision of 3.8 percentage points to electric, gas, and sanitary services. The downward revision to radio and television broadcasting was due to an upward revision to intermediate inputs, while the upward revision to electric, gas, and sanitary services was due to a downward revision to intermediate inputs.

Methodology

This part of the article describes changes in source data and estimating methods that affect the GPO estimates for each year, and it discusses the GPO methodology that was required to prepare estimates for 1996 on a more timely basis.(12)

NIPA SOURCES

The primary change in methodology that was incorporated from the NIPA annual revision was the use of new prices for deflation. The GPO estimates for real gross output and real intermediate inputs incorporate the revisions to BEA'S quality-adjusted prices for semiconductors and computers and the introduction of a new quality-adjusted price index for telephone switching equipment. These changes raised real gross output and real GPO in the electronic and electric equipment industry for 1994-96. In addition, real gross output for health services was affected by the incorporation of the Bureau of Labor Statistics (BLS) producer price index for skilled and intermediate care facilities into the deflation of for-profit nursing home services.

GPO SOURCES

The new and revised GPO estimates also reflect the use of revised composition of gross output from the 1992 benchmark I-O accounts, by the introduction of a revised concordance for matching price indexes with manufacturing products that affects both gross output and intermediate inputs, and by the use of new techniques for computing chain-type price indexes for selected manufacturing products.

In the previously published estimates, preliminary gross output estimates from the I-O accounts were used to set the level of gross output for the double-deflated industries.(13) Revised gross output levels are now used as weights in developing industry indicator series for extrapolating the previously published estimates of gross output. In general, these revisions did not significantly affect the levels of gross output.

For the computation of chain-type price indexes for selected manufacturing products, a new concordance jointly developed by BLS and BEA that matches BLS price indexes and Census Bureau product-class codes was introduced. In addition, improved aggregation techniques were introduced for developing composite price indexes from detailed BEA price indexes for computers and semiconductors at the product-class level.

As mentioned previously, new sources and methods were used to prepare the GPO estimates for 1996 when data from regular sources -- primarily annual Census Bureau surveys of manufacturing, trade, and services -- were not available. For manufacturing, annual totals of Census Bureau monthly industry shipments data were used to extrapolate 1995 annual survey of manufacturers (ASM) shipments data, because ASM data for 1996 were not yet available.(14)

Data from new sources or preliminary data were also used for some nonmanufacturing industries. In services, preliminary data from the Census Bureau's services annual survey (SAS) were used to extrapolate final 1995 SAS data. In transportation and public utilities, partial-year data or proxies for the regular series were used to extrapolate 1995 levels for a number of industries. In retail trade and wholesale trade, margin rates by detailed kind-of-business from the Census Bureau's annual retail trade survey and annual trade survey (wholesale) were held constant from 1995.

Historical GPO Revisions

The release of the current-dollar GPO estimates for 1947-96 (table 11) and chained (1992) dollar GPO estimates for 1977-96 (table 12) completes the comprehensive GPO revision. The chained-dollar GPO estimates have been revised to reflect the current-dollar GPO revisions. The revisions were generally small except for relatively large upward revisions to current-dollar GPO for general government for 1947-58. These revisions reflect the new NIPA treatment of government investment, which was incorporated in the revised GPO estimates for 1959-94 that were released in August 1996.

Gross Output and Intermediate Inputs by Industry

In addition to the estimates of GPO by industry, this article presents estimates of gross output and intermediate inputs by industry. Gross output measures each industry's total output, including the intermediate products used -- raw materials, semifinished goods, energy, and services purchased from other industries or imported -- and the value added generated in production. Gross output by industry is shown in table 13, and intermediate inputs by industry are shown in table 14; current-dollar estimates are presented in billions of dollars, and real estimates are shown as chain-type quantity indexes.(15)

Current-dollar gross output, which is roughly equivalent to an industry's sales or receipts, is often used by industry analysts as a measure of an industry's size relative to that of other industries. Current-dollar GPO, or value added, is the contribution to output by factors of production, as measured by compensation of employees, profits, and other property-type income. GPO is a measure of the industry's contribution to GDP because, like GDP, it is an unduplicated measure of total output. Thus, GPO is a better measure than gross output of the industry's contribution to the economy's output. For manufacturing, current-dollar GPO as a share of current-dollar gross output was nearly unchanged, increasing from 35.9 percent in 1992 to 36.0 percent in 1996 (table 6). Current-dollar intermediate inputs as a share of current-dollar gross output correspondingly declined slightly, from 64.1 percent in 1992 to 64.0 percent in 1996.

Quantity indexes for real gross output, real intermediate inputs, and real GPO are computed from detailed data on sales, purchases, and prices using the formula for the chain-type quantity index. Real measures adjust for the effects of price change that are included in current-dollar measures. In addition to their role in computing real GPO, estimates of real gross output and real intermediate inputs are used in studies of industry productivity. In 1992-96 real gross output for manufacturing grew 4.6 percent, real intermediate inputs grew 4.0 percent, and real GPO grew 5.6 percent.

Tables 6 through 14 follow.

[TABULAR DATA NOT REPRODUCIBLE IN ASCII]

(1.) For the previously published estimates of gross product by industry for 1959-94, see Robert E. Yuskavage, "Improved Estimates of Gross Product by Industry, 1959-94:SURVEY OF CURRENT BUSINESS 76 (August 1996): 133-155. The previously published GPO estimates for 1947-58 appeared in Robert E. Yuskavage, "Gross Product by Industry, 1988-91," SURVEY 73 (November 1993): 33-44.

(2.) For more information, see Robert P. Parker, "Completion of the Comprehensive Revision of the National Income and Product ACCOUNTS, 1929-96," SURVEY 77 (May 1997): 6-9; and Robert P. Parker and Eugene P. Seskin, "Annual Revision of the National Income and Product Accounts:" SURVEY 77 (August 1997): 6-35.

(3.) For a description of these improvements, see Yuskavage, "Improved Estimates," 133-155.

(4.) The 1992 benchmark I-O accounts are presented in Ann M. Lawson, "Benchmark Input-Output Accounts for the U.S. Economy, 1992" in this issue. A comparison of the GPO estimates with those in the I-O accounts is presented in "Note on Alternative Measures of Gross Product by Industry" this issue.

(5.) Annual and average annual growth rates for detailed industries are computed from the chain-type quantity indexes that are shown in table 9. Chained (1992) dollar GPO estimates for detailed industries and industry groups are shown in table 10.

(6.) For a description of the calculation of these contributions, see "Note on Computing Alternative Chained Dollar Indexes and Contributions to Growth" in J. Steven Landefeld and Robert P. Parker, "BEA'S Chain Indexes, Time Series, and Measures of Long-Term Economic Growth," SURVEY 77 (May 1997): 63.

(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 statistical discrepancy as a share of current-dollar GDP fell from 0.7 percent to -0.8 percent.

(9.) Property-type income is the sum of corporate profits, proprietors' income, rental income of persons, net interest, capital consumption allowances, business transfer payments, and the current surplus of government enterprises less subsidies. Proprietors' income is included in property-type income as a capital share of production; however, an unknown portion of proprietors' income represents the labor share of production.

(10.) For some analytical purposes, the labor and capital shares of gross output are more appropriate than the labor and capital shares of GPO. For most industries and for manufacturing in particular, 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 22.7 percent in 1996, whereas labor's share of manufacturing GPO was 63.0 percent.

(11.) The labor share of GPO for all industries differs slightly from the labor share of GDP, because GDP includes the statistical discrepancy.

(12.) For a detailed description of the GPO Methodology, see Yuskavage, "Improved Estimates:" 143-149.

(13.) In the double-deflation method, separate estimates of gross output and of intermediate inputs enter into the calculation of real GPO.

(14.) ASM shipments data are available on a four-digit Standard Industrial Classification (sic) basis, whereas monthly shipments data are generally available only on a three-digit sic basis. Product-class shipments, which are used as weights to develop deflators at the four-digit industry level, also were not available for 1996, so the product composition of industry shipments was held constant from 1995.

(15.) Gross output and intermediate input estimates are prepared only for those industries for which the double-deflation method is used to estimate real GPO. For a list of these industries, see Yuskavage, "Improved Estimates," 145. For the other industries, source data are not adequate for preparing gross output estimates.

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, often referred to as its "value added," is equal to its gross output (sales or receipts and other operating income, commodity taxes, and inventory change) minus its intermediate inputs (consumption of goods and services purchased from other industries or imported).

For the national income and product accounts (NIPA'S), GDP is measured as the sum of expenditure components. Gross domestic income (GDI) is measured as the sum of costs incurred and incomes earned in the production of GDP. In concept, GDP and GDI should be the same; in practice, they differ because their components are estimated using largely independent and less-than-perfect source data. 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) The difference between GDP and GDI is called the "statistical discrepancy"; it is recorded in the NIPA'S as an "income?" component that reconciles GDI with GDP.

Current-dollar GPO by industry is measured as the sum of distributions by industry of the components of GDI. Consequently, the sum of the current-dollar GPO estimates also differs from current-dollar GDP by the statistical discrepancy. In presenting 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.

Real GDP in the NIPA'S is also measured as the sum of the expenditure components. Real GPO estimates for most industries are derived using separate estimates of gross output and intermediate inputs.(2) The sum of the real GPO estimates differs from real GDP by the real statistical discrepancy, which is shown as part of private-industry GPO, and by the category entitled not allocated by industry," which is the difference between real GDP and the sum of real GPO for the detailed industries and of the statistical discrepancy. The value of the category "not allocated by industry" reflects the lack of additivity of detailed real GPO estimates that results from the formula used to calculate real output and from differences in the source data (both current dollars and prices) used to estimate industry GPO and the expenditures measure of real GDP. As with the current-dollar measures, BEA views the source data used to estimate the components of real GDP to be more reliable. In addition, the amount of detailed expenditures data available to calculate real GDP is greater than that for the gross output and intermediate inputs available to calculate real GPO. For some industries, no source data are available to measure gross output, and the resulting real GPO estimates are prepared using less reliable methodologies.

(1.) For additional information on the accuracy of the two measures, see the box "Statistical Discrepancy" in Parker and Seskin, "Annual Revision," 19.

(2.) 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 Yuskavage, "Improved Estimates," 142.