Gross state product by industry, 1963-86.

Renshaw, Vernon ; Trott, Edward A., Jr. ; Friedenberg, Howard L. 等

Gross State Product by Industry, 1963-86

In this article, BEA introduces annual estimates of gross state product (GSP) by component and by industry for each State and the District of Columbia for the period 1963-86 (tables 1 and 2). These estimates are the most comprehensive measures of production available for States and will improved the basis for analyzing and forecasting trends in States economic activity.

GSP is the gross market value of the goods and services attributable to labor and property located in a State. It is the State counterpart of the Nation's gross domestic product (GDP).

BEA prepares GSP estimates for 61 industries. For each industry, GSP is composed of four components: (1) Compensation of employees (hereafter termed "compensation"); (2) proprietors' income with inventory valuation adjustment and capital consumption allowances ("proprietors' income"); (3) indirect business tax and nontax liability ("IBT"); and (4) other, mainly capital-related, charges ("capital charges"). For the farming, mining, construction, and manufacturing industries, BEA directly estimates total GSP and three components--compensation, proprietors' income, and IBT--and then subtracts the three components from GSP to get capital charges. For the other industries, BEA directly estimates each of the four components of GSP and then sums the components to get GSP.

Previously, earnings by place of work-estimated in connection with State personal income--was the only part of GSP that BEA published regularly by industry. Earnings includes most of the compensation and proprietors' income GSP components, but excludes capital charges and IBT. (Table A shows in detail how GSP corresponds to earnings and GDP.) The capital charges component reflects capital stocks and profit rates by State. The IBT component reflects liabilities charged to business expenses, most of which are sales and property taxes levied by State and local governments.

In the absence of State estimates of capital charges and IBT, earnings (or wages and salaries) have often been "blown up" to approximate GSP in nonfarm industries. This procedure assumes that each State's share of total GSP for the Nation in an industry equals its share of earnings in the industry. That is, blowups assume away State-to-State differences in capital stocks and rates of return to capital and in tax structures and rates. Blowups are particularly prone to error where earnings are a small portion of GSP, such as in the real estate, oil and gas extraction, petroleum refining, and other capital-intensive industries. BEA's GSP estimates overcome, for the most part, the limitations of blowup estimates.

BEA estimates GSP in both current and constant dollars. Current-dollar GSP estimates reflect changes in the command over resources associated with production and are particularly useful for analyzing the differential regional effects of large changes in relative output prices, such as the changes in energy and agricultural prices in the 1970's and 1980's.

Constant-dollar GSP estimates reflect changes in the physical volume of production and are particularly useful for comparing regional trends in labor productivity or for projecting the volume of industrial output. Consequently, the constant-dollar GSP estimates will be used in the set of BEA regional projections to be published in 1990 (when the GSP estimates will be updated).

The constant-dollar GSP estimates are now based on national price deflators by industry. At some point, it may be possible to develop State price data to improve the constant-dollar estimates. Such data would improve the estimates for those industries--such as energy, construction, real estate, and State and local government--in which prices vary regionally.

Analyzing Regional Growth

Patterns Using GSP

This section focuses on changes in regional shares of national totals for GSP and its components, emphasizing what the GSP estimates show about regional growth patterns that is not shown by compensation--the component that corresponds most closely to the measures (earnings or wages and salaries) commonly used in blowups. The discussion is based on current-dollar estimates for economic census years 1967, 1977, and 1982, and for 1986 (the most recent year for which estimates have been made).

From 1977 to 1986, the share of economic activity generated in the Nation's interior regions--whether measured by GSP or compensation--declined more than 3 percentage points, while the share generated by regions along the Atlantic and Pacific coasts increased (table B). Relative weakness in the interior regions was apparent in the sharp declines in the manufacturing and farming industries of the Great Lakes and Plains regions in the late 1970's and early 1980's and then spread to the Southwest, Rocky Mountain, and interior Southeast regions after 1982, as declining oil prices adversely affected the regions' energy-oriented industries. In both the Great Lakes region in 1977-82 and in the energy-oriented regions in 1982-86, the capital charges component of GSP had substantially larger relative declines than did compensation.

Manufacturing in the Great Lakes region

The Great Lakes region dominated the relative economic decline of the interior regions in both manufacturing and all industries combined from 1977 to 1982. Its share of GSP for manufacturing declined 5.5 percentage points (to 23.0 percent of the Nation), and its share of GSP for all industries declined 2.9 points (to 16.9 percent). The other interior regions combined, in contrast, increased their share of both manufacturing GSP (by 1.9 points) and all-industry GSP (2.6 points) over this period.

The Great Lakes' relative loss of both manufacturing and all-industry GSP occurred in part because the region's manufacturing profitability declined, relative to other regions, from the late 1960's. From 1967 to 1982, capital charges in manufacturing fell as a percentage of manufacturing GSP by 8.9 points (to 16.4 percent) in the Great Lakes region; in all other regions combined, the percentage fell only 1.8 points (to 21.8 percent). In addition, the region's loss reflected its heavy dependence on manufacturing industries that were hard hit by the 1980 and 1981-82 recessions and by increasing competition from foreign producers.

Forecasts and projections--including those of BEA--made in the late 1970's substantially underestimated the Great Lakes' relative decline from 1977 to 1982. Apart from not anticipating the severity of the 1981-82 recession, the forecasts were based on trends in compensation only. While the Great Lakes' share of compensation in manufacturing had been nearly constant in the decade preceding the forecasts, its share of capital charges in manufacturing had declined substantially.

Although the Great Lakes region dominated the decline of the interior regions from 1977 to 1982, it had the smallest decline in the all-industry GSP share of any interior region from 1982 to 1986. Some of the Great Lakes' manufacturing industries that had been hard hit--such as the motor vehicles industry--recovered well from the 1981-82 recession. In addition, other manufacturing operations were restructured to improve their relative profitability. In the meantime, the interior regions that had been gaining GSP share began to decline.

Mining in three interior regions

The relative economic decline of the interior regions from 1982 to 1986 was dominated by decline in regions dependent on energy-related mining. Each of the three interior regions with large mining sectors lost share of all-industry GSP from 1982 to 1986 after gaining share from 1977 to 1982.

Energy-related mining--that is, oil and gas extraction and coal mining--dominates the U.S. mining industry, and, even with growth of oil production in Alaska, energy-related mining is concentrated in the interior regions. The Southwest, Rocky Mountain, and interior Southeast regions combined accounted for 75 percent of the Nation's mining GSP in 1977 and still accounted for over 70 percent in 1986.

Large changes in world oil prices affect the States with energy-related mining because they bring large changes in the value of production (GSP) in mining. These changes in production, in turn, can lead to large changes in oil and gas exploration and in related industries--for example, in production of oil drilling equipment and in selected distributive and service industries.

Because compensation is generally small relative to IBT and capital charges in energy-related mining (especially in oil and gas extraction), compensation alone provides an inadequate basis for analyzing the economic effects of the ups and downs of oil prices. Substantial oil price increases in 1973 and 1979 (asociated with the OPEC oil embargo and disruption of oil production in Iran, respectively) increased the value of U.S. oil, gas, and coal production and stimulated domestic exploration up through the 1981-82 recession. As a result, most of the Southwest and Rocky Mountain STates and several interior Southeast STates (especially Louisiana) experienced a growing share of all-industry GSP. In all three regions, the share of all-industry GSP increased faster than the share of all-industry compensation from 1977 to 1982, reflecting the substantial increase in profits for many producers involved in energy-related mining. In the interior Southeast, the share of all-industry GSP increased while the share of compensation actually declined slightly. In the Southwest, where three of the four States have important oil and gas extraction industries, the share of GSP increased by more than 2 percentage points.

Oil prices peaked in 1981, and their subsequent fall began to eliminate the gains in share of all-industry GSP in the Southwest, Rocky Mountain, and interior Southeast regions. The Southwest and Rocky Mountain regions each had declines in share from 1982 to 1986 that amounted to more than one-half of their gains from 1977 to 1982, and the interior Southeast experienced a decline more than twice its earlier gain. As in the case of the 1977-82 "boom," the 1982-86 "bust" resulted in larger changes in share of all-industry GSP than compensation for all three regions.

Appendix--Sources and

Methods

The GSP estimates presented in this article are an extension of the benchmark estimates published in a 1985 BEA staff paper. The new estimates include (1) updates of the 1963, 1967, 1972, and 1977 benchmark-year estimates published previously, (2) 1982 benchmark-year estimates, and (3) annual estimates for nonbenchmark years. To make the previously published benchmark-year estimates consistent with the 1982 estimates, they are adjusted to incorporate both the 1985 revisions to the national income and product accounts and recently completed revisions to the State personal income series.

Compensation and proprietors' income

Annual estimates by State and industry of two components of compensation--wages and salaries, and other labor income--as well as of proprietors' income with inventory valuation adjustment (IVA) are from BEA's State personal income series. Wages and salaries, in turn, is part of the basis for assigning to States the components of compensation not measured in the personal income account--employer contributions for social insurance--and proprietors' income with IVA is the basis for assigning noncorporate capital consumption allowances.

Capital charges

For the benchmark years, the sources and methods for capital charges differ among industries.

Goods-producing industries.--For 27 agricultural, mining, construction, and manufacturing industries, BEA estimates capital charges by first estimating total GSP and then subtracting compensation, proprietors' income, and IBT. Economic census data on value added in production, adjusted to conform to BEA's income and product definitions, are the basis for estimating total GSP.

Regulated distributive and service industries.--For seven transportation, communication, utility, and finance industries, data contained in financial reports filed by firms with regulatory agencies are the basis for estimating capital charges. BEA employs indicators of capital stock or its use--for example, airline boardings--to assign capital charges for multistate firms to States.

Real estate industry.--For this industry, BEA mainly uses data from the population and housing censuses and the U.S. Department of Agriculture to assign capital charges to States in accordance with the location of real property.

Unregulated distributive and service industries.--For 23 transportation, trade, finance, insurance, and service industries, BEA uses economic census data on business receipts or sales and data on wages and salaries to assign capital charges to States.

Government.--For Federal Government enterprises, BEA uses data specific to each enterprise to assign capital charges--that is, surplus or deficit--to States. For State and local government enterprises, BEA uses data on current revenues and expenses, by type of enterprise, from the census of governments to assign the surplus or deficit.

For the nonbenchmark years, capital charges in all industries--except in farming, in real estate, and in manufacturing for 1983 and 1984--are interpolated or extrapolated using movement in wages and salaries and in national control totals. Farm estimates for all years are directly estimated based on U.S. Department of Agriculture data. Real estate estimates for intercensal years are based on data developed in the course of estimating the rental income of persons in the State personal income series. Manufacturing estimates for 1983 and 1984 are based on data from the Census Bureau's Annual Survey of Manufactures (ASM). As resources permit, BEA expects to incorporate ASM data for additional years (extending backward as well as forward) and other annual data, particularly that contained in regulatory agency reports.

IBT

For the benchmark years and for the years 1983-85, IBT estimates are based on the following data: (1) Taxes collected, broken down by State and type of tax, from the census of governments (for State and local IBT) and the Internal Revenue Service (for Federal IBT) and (2) taxes collected, broken down by industry and type of tax for the Nation, from BEA's National Income and WEalth Division.

For the nonbenchmark years prior to 1982, estimates for IBT for all levels of government and types of taxes by industry are interpolated, based on movement in compensation of employees, proprietors' income, and national control totals. In the absence of 1986 information, the 1986 estimates were derived using the 1985 distribution by State.