Gross domestic product by industry for 1999-2001

THE Bureau of Economic Analysis (BEA) has recently released new estimates of gross domestic product (GDP) by industry for 2001 and revised estimates for 1999-2000.1 The estimates incorporate the results of this year's annual revision of the national income and product accounts (NIPA's) and newly available

1. For the previously published estimates, see Sherlene K.S. Lum and Brian C. Moyer, "Gross Domestic Product by Industry for 1998-2000," SURVEY OF CURRENT BusiNEss 81 (November 2001): 17-33.

2. See Eugene P. Seskin and Stephanie H. McCulla, "Annual Revision of the National Income and Product Accounts," SuRvEY 82 (August 2002): 7-14 components of gross domestic income (see the box below).

After growing at a rapid pace in the last half of the 1990s, real GDP slowed in 2000, declined in the first three quarters of 2001, and increased in the last quarter of 2001. Relative to past cycles, the slowdown in real GDP was mild, though real growth rates varied widely among industries. Declines, particularly in some manufacturing industries, were steep. Growth rates in other industries-mainly in services but also in manufacturing-remained strong in 2001, partly reflecting decreases in unit labor costs and unit capital costs through increases in productivity.

Highlights in 2001 include the following:

* Real GDP increased 0.3 percent. Real GDP in private services-producing industries increased 1.7 percent, while real GDP in private goods-producing industries decreased 4.2 percent.3

* Among the private goods-producing industries, manufacturing was the hardest hit by the 2001 recession. Its real GDP decreased 6.0 percent. Durable-goods manufacturing decreased 5.2 percent, and nondurable-goods manufacturing decreased 7.1 percent.

* The performance of the information technology (IT)-related industries was mixed. Real GDP in the industrial machinery and equipment industry and in the instruments and related products industry decreased sharply. In contrast, real GDP in the communications industry group and in the electronic and other electric equipment industry increased strongly; in the face of falling prices in both industries and falling demand in the electronic and other electric equipment industry, these industries were able to lower their unit labor costs and unit capital costs through increases in productivity, and thereby increase their real GDP (value-added) growth.

* Real GDP growth in the finance, insurance, and real estate (FIRE) industry group was strong. Doubledigit growth was reported for nondepository institutions (including credit-card companies, motor vehicle finance leasing companies, and mortgage banking firms) and for security and commodity brokers.

* In several of the private services-producing industry groups, real GDP growth slowed but remained relatively strong. Growth in retail trade slowed from 7.5 percent in 2000 to 4.6 percent in 2001, and growth in FIRE slowed from 6.2 percent to 2.8 percent. Growth in the communications industry group was 12.3 percent in both 2000 and 2001. * Real GDP for services-mainly business and personal services-grew only 0.9 percent.

* Declines in real GDP in the transportation-by-air industry and in the hotels and other lodging places industry lowered overall real GDP growth by 0.1 percentage point, partly reflecting sharp reductions in business air travel and tourism-related services after the terrorist attacks on September 11th.

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 growth, industry shares of current-dollar GDP, and the components of income for current-dollar GDP by industry. The second part discusses the prices of GDP by industry, contributions to price change, and unit costs. The third part discusses the revisions to the estimates. The fourth part describes the changes to the methodology and presentation. The detailed GDP-by-industry estimates for 1998-2001 are presented in tables 1-14 at the end of the article (see also the box "Data Availability" on page 27).

Measures of Industry Performance

The relative performance of particular industries or industry groups can be assessed by examining their real growth rates, their contributions to real GDP growth, their shares of current-dollar GDP, and their components of income for current-dollar GDP by industry.

Real GDP-by-industry growth rates

From 1995 to 2000, a period of rapid economic growth, real GDP increased at an average annual rate of 4.0 percent. Private industries increased 4.6 percent, and government increased 1.4 percent; in private industries, goods-producing industries increased 4.1 percent, and services-producing industries increased 5.3 percent (table A and table 5A).4

In 2001, real GDP growth slowed to 0.3 percent from 3.8 percent in 2000. Growth in both the private goods-producing industries and the private services-- producing industries was adversely affected by the recession. Real growth in the goods-producing industries dropped from 3.6 percent to -4.2 percent, and real

4. Real GDP by industry is computed using the double-deflation method in which separate estimates of real gross output and intermediate inputs combine in a Fisher chain-type quantity-index-number formula. For more information, 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.

growth in the services-producing industries dropped from 5.4 percent to 1.7 percent. Among the goodsproducing industries, growth in manufacturing dropped the most, from 4.7 percent to -6.0 percent.

The -6.0-percent growth in manufacturing in 2001 resulted from negative growth in both durable-goods manufacturing (-5.2 percent) and nondurable-goods manufacturing (-7.1 percent). At the detailed industry level, growth in 19 of the 21 manufacturing industries was negative. Growth decreased sharply in fabricated metal products (-10.2 percent), industrial machinery and equipment (-10.7 percent), textile mill products (-10.7 percent), and petroleum and coal products (-17.9 percent).

Real growth in the private services-producing industries slowed but was still a positive 1.7 percent in 2001. This growth was largely driven by growth in retail trade (4.6 percent) and in FIRE (2.8 percent). The growth in FIRE primarily reflected growth in nondepository institutions (16.4 percent) and security and commodity brokers (13.1 percent). In contrast, growth declined in the transportation-by-air industry (-7.7 percent) and in hotels and other lodging places (-4.1 percent), partly reflecting sharp reductions in business air travel and tourism-related services after September 11th.

The performance of the IT-related industries in 2001 was mixed. In durable-goods manufacturing, real growth in industrial machinery and equipment (including computers) decreased 10.7 percent, and real growth in instruments and related products decreased 9.6 percent. In contrast, real growth in electronic and other electric equipment (including semiconductors, digital telephone switching equipment, and local area network (LAN) equipment) increased 7.5 percent, and real growth in the communications industry group increased 12.3 percent.

The growth in real GDP in the electronic and other electric equipment industry was positive despite , sharp decrease in industry output. In current-dollars, gross output (primarily sales) decreased 22.4 percent in 2001 (table 8), and intermediate inputs (purchased energy, materials, and services) decreased 29.1 percent (table 9). By absorbing much of the decrease in output through decreased intermediate inputs, the industry's current-dollar value added-that is, its current-dollar GDP by industry-decreased only 11.7 percent (table 1). In addition, the prices of gross output fell more than those of intermediate inputs, so industry value-- added prices fell 17.8 percent (table 7A). As a result of increases in productivity, the industry was able to lower its unit labor costs 15.7 percent and its unit capital costs 24.8 percent, and thereby raise its real value-- added.

Real growth in the communications industry group was strong in 2001 despite falling prices for output. Gross output prices (primarily reflecting prices for cellular telephone services) fell more than intermediate inputs prices, so industry value-added prices fell 7.0 percent. Productivity increases enabled the industry to lower its unit labor costs 11.6 percent and to lower its unit capital costs 2.6 percent, and as a result, its real value added increased 12.3 percent.

Contributions to real GDP growth

An industry's contribution to real GDP growth depends on both its real growth rate and its relative size.5 In 1995-2000, private goods-producing industries contributed 1.0 percentage point to the 4.0-percent average annual growth rate of real GDP, and private services-producing industries contributed 3.4 percentage points (table B). In 2001, private goods-producing industries contributed -1.0 percentage point to the 0.3percent growth rate of real GDP, and private services-- producing industries contributed 1.2 percentage points.

The slowdown in real GDP growth from 3.8 percent in 2000 to 0.3 percent in 2001 can be examined in terms of changes in the industry contributions to real GDP growth. The contribution of private goods-producing industries fell 1.8 percentage points-from 0.8 percentage point in 2000 to -1.0 percentage point in 2001. The contribution of private services-producing industries fell 2.3 percentage points-from 3.5 percentage points to 1.2 percentage points. The decrease in the goods-producing industries primarily reflected a decrease in the contribution of manufacturing from 0.8 percentage point to -0.9 percentage point. The decrease in the contribution of the services-producing industries was spread across several industry groups: Transportation and public utilities decreased from 0.6 percentage point to 0.0 percentage point, wholesale trade and retail trade combined decreased from 1.1 percentage points to 0.4 percentage point, FIRE decreased from 1.2 percentage points to 0.6 percentage point, and services decreased from 0.7 percentage point to 0.2 percentage point.

In 2001, the 0.3-percent growth in real GDP reflected largely offsetting contributions from the private goods-producing industries (-1.0 percentage point) and the private services-producing industries (1.2 percentage points). In the goods-producing industries, manufacturing-the sector hardest hit by the recession-made the largest negative contribution (-0.9 percentage point) to real GDP growth. Durable-goods manufacturing and nondurable-goods manufacturing each contributed -0.5 percentage point. In the private services-producing industries, large positive contributions were made by FIRE (0.6 percentage point), retail trade (0.4 percentage point), and communications (0.3 percentage point).

Shares of current-dollar GDP

An industry's share of current-dollar GDP is a better indicator of the industry's relative size in the economy than its share of real GDP, because the shares of real GDP depend on the choice of the reference year. The 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 1998-2001, the share of current-dollar GDP that was accounted for by private goods-producing industries declined, while the shares accounted for by private services-producing industries and by government rose (table C and table 2). The share of goods-producing industries decreased from 23.2 percent in 1998 to 21.6 percent in 2001; it decreased each year, and the largest decline was in 2001. The share of services-producing industries increased from 64.6 percent to 66.8 percent; it increased each year. The share of government increased slightly from 12.6 percent to 12.7 percent. The increase in the share of services-producing industries was mostly accounted for by services and FIRE. The share of services rose 1.3 percentage points; business services increased the most (0.4 percentage point). The share of FIRE rose 1.1 percentage points; the largest increases were in real estate (0.4 percentage point) and nondepository institutions (0.3 percentage point).

In 2001, the share of goods-producing industries dropped 1.3 percentage points. The drop was attributable to a 1.4-percentage-point decline in the share of manufacturing. Durable-goods manufacturing decreased 0.9 percentage point, and nondurable-goods manufacturing decreased 0.4 percentage point. Shares of each of the detailed manufacturing industries either decreased or did not change. The shares of the two largest industries decreased-industrial machinery and equipment by 0.3 percentage point and electronic and other electric equipment by 0.2 percentage point.

The share of services-producing industries increased in 2001. The share of services increased 0.6 percentage point, and the share of FIRE increased 0.5 percentage point. In contrast, the share of wholesale trade decreased 0.3 percentage point.

The share of government increased 0.3 percentage point in 2001. The increase was entirely accounted for by a 0.4-percentage-point increase in the share of State and local government. The Federal Government's share decreased 0.1 percentage point.

Components of income for current-dollar GDP by industry

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 GDP by industry-compensation of employees, indirect business tax and nontax liability, and property-type income. The labor share of production is approximated by 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 by property-type income, which consists of corporate profits and proprietors' income, inventory valuation adjustments, rental income of persons, net interest, private capital consumption allowances, business transfer payments, the current surplus of government enterprises less subsidies, and government consumption of fixed capital.6

For the total economy, the share of GDP accounted for by labor increased 1.4 percentage points in 19982001, while the share accounted for by capital decreased 0.6 percentage point (table 4).7 For the private goods-producing industries, the labor share increased 1.8 percentage points, while the capital share decreased 2.2 percentage points. For the private services-producing industries, the labor share increased 1.1 percentage points, while the capital share decreased 0.6 percentage point.

In private goods-producing industries, manufacturing's labor share increased 3.4 percentage points, while its capital share decreased 4.1 percentage points. The labor share increased and the capital share decreased in both durable-goods manufacturing and nondurable-- goods manufacturing. In durable-goods manufacturing, which accounts for nearly three-fifths of manufacturing output, the labor share increased 5.0 percentage points, and the capital share decreased 5.2 percentage points. This shift in shares indicates that spending on capital inputs leading up to, and during, the recession fell relative to spending on labor inputs. In 2000, compensation of employees in durable-goods manufacturing industries grew $40.7 billion, while property-type income fell $8.9 billion. In 2001, spending for both components fell, but property-type income fell $3.1 billion more than compensation of employees (table 3).

Measures of Industry Prices

The growth rate in an industry's price index indicates the extent to which its prices of labor and capital are changing. An industry's contribution to GDP price change indicates the extent to which the industry is affecting overall prices. Changes in an industry's unit costs indicate the extent to which the cost structure of the industry is changing.

GDP-by-industry prices

The price index for GDP by industry for an industry or an industry group represents the price of the industry's primary factors of production (the value-added inputs of labor and capital).8 In 1995-2000, the GDP price index grew at an average annual rate of 1.7 percent, and the price index for private industries grew 1.4 percent. The price index for private goods-producing industries grew 0.6 percent, and the price index for private services-producing industries grew 1.7 percent (table D and table 7A).

In 2001, the GDP price index increased 2.4 percent. The price index for private goods-producing industries increased 1.3 percent, and the price index for private services-producing industries increased 2.1 percent. The slower growth in prices for goods-producing industries largely reflected continued declines in durable-goods manufacturing prices-particularly in prices for IT-related products. The price index for industrial machinery and equipment (including computers) fell 4.1 percent, and the price index for electronic and other electric equipment (including semiconductors, digital telephone switching equipment, and LAN equipment) fell 17.8 percent.

Prices of petroleum-related products increased much less in 2001 than in 2000. In the mining industry group, the price index for oil and gas extraction decelerated from 62.0 percent to 0.2 percent. In nondurable-goods manufacturing, the price index for petroleum and coal products decelerated from 49.1 percent to 28.2 percent.

The growth rates in an industry's gross output prices and in its intermediate inputs prices can provide insight on the growth rates in its value-added prices. For example, the price index for the gross output of durable-goods manufacturing decreased 2.5 percent in 2001 (table 11A), while the price index for intermediate inputs decreased only 2.0 percent (table 13A). Because output prices decreased more than intermediate inputs prices, value-added prices fell more (3.3 percent) than output prices.

Contributions to GDP price change

An industry's contribution to the growth in GDP prices depends on the growth rate of its price index and on its relative size.9 In 1995-2000, private industries contributed 1.2 percentage points to the 1.7-percent average annual growth rate in the GDP price index; private goods-producing industries contributed 0.1 percentage point, and private services-producing industries contributed 1.1 percentage points (table E). Among private goods-producing industries, durablegoods manufacturing contributed -0.3 percentage point. Among private services-producing industries, services was the largest contributor-at 0.7 percentage point.

In 2001, private industries contributed 1.6 percentage points to the 2.4-percent growth in the GDP price index; private goods-producing industries contributed 0.3 percentage point, and private services-producing industries contributed 1.4 percentage points. Among private goods-producing industries, durable-goods manufacturing contributed -0.3 percentage point. Among private services-producing industries, services contributed 0.9 percentage point, but communications and wholesale trade each contributed -0.2 percentage point. The negative contributions by durable-goods manufacturing and by communications partly reflected declines in the prices of IT-related products.

Unit costs

In the GDP-by-industry accounts, the price index for an industry represents the price of its primary factors of production, so an industry's price index, combined with its current-dollar components of GDP by industry, can be used to assess each component's contribu

9. An industry's contribution to GDP price growth is the product of its share of current-dollar GDP and the growth rate in its GDP-by-industry price index.

tion to the total labor costs and capital costs for an industry. The measures of unit costs are computed by dividing current-dollar GDP by industry and its components by real (chained-dollar) GDP by industry. The resulting quotients are the GDP-by-industry implicit price index and the part of the price index that is associated with each component. Unit cost measures by private industry group are presented in table 14. If the percent change in the unit cost for a component is greater than the percent change in the GDP-by-industry price index, the relative importance of that component in the industry cost structure has increased.

In 1995-2000, the total cost per unit for private industries increased 1.4 percent (table F). Unit costs for compensation of employees (unit labor costs) increased 1.8 percent. Unit costs for property-type income (unit capital costs) increased 1.2 percent, and unit costs for indirect business tax and nontax liability decreased 0.2 percent. The larger increase in unit labor costs indicates that labor costs became a larger part of total unit costs during the period. Increases in unit labor costs were reported for all industry groups except durable-goods manufacturing, wholesale trade, and retail trade. The largest decrease in unit labor costs was in durable-goods manufacturing (-3.2 percent).

In 2001, total cost per unit decreased in durablegoods manufacturing, in wholesale trade, and in mining. In durable-goods manufacturing and wholesale trade, the decreases in total cost per unit continued the 1995-2000 trend. The 3.3-percent decrease in durablegoods manufacturing resulted from a 0.5-percent decline in unit labor costs, which accounted for almost three-fourths of total current-dollar cost per unit, and an 11.6-percent decline in unit capital costs. The decrease in wholesale trade resulted from declines in both unit labor costs and unit capital costs. In mining, total unit costs decreased slightly as unit capital costs decreased and unit labor costs increased.

Revisions to the Estimates of GDP by Industry

The estimates of GDP by industry for 1999-2000 were revised to incorporate this year's annual revision of the NIPA's and new and revised source data for gross output and prices. The revisions to current-dollar GDP for 1999 were smaller than those for 2000; GDP was revised up $5.7 billion for 1999 and was revised down $48.3 billion for 2000 (table G). Real GDP growth was not revised for 1999, and it was revised down 0.3 percentage point for 2000. The revisions to real GDP for some industry groups were substantial but largely offsetting, so their effects on real GDP growth were small.

The revisions to the current-dollar estimates of GDP by industry reflected the incorporation of the revised NIPA estimates of the components of gross domestic income and the industry distributions of these components. Private industries was revised up $6.1 billion for 1999 and was revised down $49.6 billion for 2000. For 1999, upward revisions to the statistical discrepancy ($33.9 billion) and wholesale trade ($11.8 billion) were partly offset by downward revisions to manufacturing ($15.5 billion), FIRE ($11.8 billion), transportation and public utilities ($6.7 billion), and services ($3.7 billion). For 2000, downward revisions to services ($48.2 billion), manufacturing ($46.3 billion), and transportation and public utilities ($15.7 billion) were partly offset by upward revisions to FIRE ($40.5 billion), wholesale trade ($22.7 billion), and mining ($6.0 billion).

The revisions to the growth rates of real GDP by industry reflect the revisions to current-dollar GDP by industry and the incorporation of new and revised source data for gross output and prices. For 1999, the growth rate of real GDP was not revised; the growth rate of goods-producing industries was revised down 0.8 percentage point, and that of services-producing industries was revised down 0.3 percentage point.10 For 2000, the growth rate of real GDP was revised down 0.3 percentage point; the growth rate of goods-producing industries was revised up 0.7 percentage point, and the growth rate of services-producing industries was not revised. By industry group, the largest revisions for 1999 were to wholesale trade (up 3.0 percentage points) and mining (up 2.3 percentage points). The largest revisions for 2000 were to mining (up 3.8 percentage points), wholesale trade (up 3.1 percentage points), nondurable-goods manufacturing (down 2.2 percentage points), and services (down 1.8 percentage points).

Changes to the Methodology and Presentation

This revision to the GDP-by-industry accounts incorporated several changes to the methodology and the presentation. These changes included the incorporation of improved methodology from the NIPA annual revision, the incorporation of inputs data from the 1998 annual input-output (I-O) accounts, and the inclusion of percent-change tables for all price and quantity indexes. In addition, because the GDP-by-industry accounts are currently based on the Standard Industrial Classification (SIC) system, BEA converted the source data that are based on the North American Industry Classification System (NAILS) to an SIC basis.

The following improvements that were made as part of the NIPA annual revision were incorporated. The industry estimates for security and commodity brokers incorporated the improved price indexes for personal consumption expenditures (PCE) of brokerage and investment counseling that were developed from the Bureau of Labor Statistics' producer price index for brokerage services. The industry estimates of profits incorporated the improved measures of imputed commissions on equities transactions.

The intermediate inputs by industry for 1999-2001 are based on the industry distributions of inputs from the 1998 annual I-O accounts, the latest available I-O accounts.

The GDP-by-industry tables were expanded to include percent-change tables for the chain-type quantity and price indexes for outputs (tables 10A and 11A) and intermediate inputs (tables 12A and 13A). This expansion allows for greater consistency and comparability across the GDP-by-industry measures of industry performance.

This revision incorporated new and revised source data from private industry associations and from other Federal Government agencies. The source data from several Federal Government agencies are now based on NAICS rather than on the SIC, including Census Bureau data from the annual survey of manufactures, the services annual survey, and the annual trade surveys and the Internal Revenue Service data from the Statistics of Income. In addition, the BEA estimates of change in private inventories are on a NAICS basis. Because the GDP-by-industry estimates remain on an SIC basis, these NAICS-based source data were converted to an SIC basis by the source agency or by BEA using information provided by the source agency. Implementation of NAICS for the GDP-by-industry estimates will occur in the next comprehensive revision to the GDP-by-industry accounts, which is scheduled for release in 2004. At that time, NAICS-based data from both the NIPA's and the benchmark 1-0 accounts will be incorporated.

Tables 1-14 follow.