Tracking the BEA regional projections, 1983-86.

Johnson, Kenneth P. ; Friedenberg, Howard L. ; Renshaw, Vernon 等

Tracking the BEA Regional Projections, 1983-86 By KENNETH P. JOHNSON, HOWARD L. FRIEDENBERG, and VERNON RENSHAW

BEA's projections of population growth for 1983-86 are within narrow margins for error in most regions, but the projections of employment growth show wider margins of error in most industries and regions. These results are from comparisons of projections with measured changes. The projections, which are based on growth trends through 1983, were published in 1985. Because trends change, the projections are tracked to alert users to the industries and regions in which the projections have missed the mark and to help BEA prepare the next set of to help BEA prepare the next set of projections.(1)

(1) The projections for 1990 and 2000 appeared in the May 1985 SURVEY OF CURRENT BUSINESS. Projections for 1986, based on interpolation, are available from the Regional Economic Analysis Division, BE-61, Bureau of Economic Analysis, U.S. Department of Commerce, Washington, DC 20230. The measured employment and population levels through 1986 are available from the Regional Economic Measurement Division, BE-55, at the same address. (Employment estimates for 1987 have not been completed.) Earlier articles tracking BEA's regional projections appeared in the April 1976 and May 1983 issues of the SURVEY. The next set of BEA long-term regional projections is scheduled for publication in 1990.

Employment growth

For the Nation, projected growth in total employment is 2.8 percentage points less than measured growth (tables 1 and 2). The difference reflects partly offsetting industry patterns: In service-type industries, projected employment growth is 6.1 percentage points less than measured, and in goods-producing industries, projected growth is 6.3 percentage points more than measured (table 3). The offset is partial because the number of jobs in goods-producing industries is substantially less than in service-type industries.

Growth in service-type employment was underprojected because the projections did not fully account for the continuing strength of the service economy. In addition, projected and measured growth were based on different concepts of self-employment.(2) The difference in concept contributed to shortfalls between the projected and the measured growth of self-employment, mainly in the real estate, personal services, business services, and miscellaneous repair services industries. Goods-producing employment, in contrast, was overprojected because the projections did not fully anticipate the employment effects of declines in farm and energy prices and the uncharacteristically slow recovery of durable goods manufacturing from the 1981-82 recession.

(2) The 1986 comprehensive revisions to the State personal income series (see the August 1986 SURVEY OF CURRENT BUSINESS) introduced a job count measure of self-employment. In 1985, when the projections were published, self-employment was mainly a count or persons. The person count measure is lower than the job count measure because of multiple job holding.

The industrial patterns of under-projection and overprojection have different effects on regions, inasmuch as regions' industry specializations differ. In general, the underprojection of service-type employment, while affecting all regions, most affects regions along the Atlantic and Pacific coasts (including the part of the Southeast region on the Atlantic coast), and the overprojection of goods-producing employment most affects interior regions (including the other States in the Southeast region).

Coastal regions. -- Unexpectedly, employment growth in service-type industries and construction far exceeded growth in other industries during the recovery from the 1981-82 recession; in each coastal region, projected employment growth in these industries -- and in the all-industry total -- fell substantially short of measured growth. In the Far West and coastal Southeast, the national defense buildup early in the expansion benefited research and development and other services as well as construction. The national defense buildup also benefited New England and the Mideast. In addition, these regions provided business and professional services to growing national and international markets, and the gains in these service industries stimulated construction activity. (New England and the Mideast had not participated in the late 1970's construction boom, which had been centered in interior regions;, where sharp price increases for farm commodities and oil stimulated growth.)

In the Mideast, despite the large underprojections of employment growth in service-type industries and construction, total employment growth was underprojected less than in any other coastal region. The major reason was a partly offsetting, large overprojection of employment growth in durables manufacturing. The overprojection for Pennsylvania, in particular, reflected the failure of the primary metals (steel) industry to recover from the 1980 and 1981-82 recessions in the face of import competition.

Interior regions. -- Largely reflecting unexpected developments in goods-producing industries, the Rocky Mountain, Southwest, and interior Southeast regions show overprojections of total employment growth, and the Plains and Great Lakes regions show relatively small underprojections.

In the rocky Mountain and Southwest regions, total employment, projected to grow faster than in the Nation in 1983-86, grew more slowly. The overprojections were mainly due to the direct and indirect effects of declines in energy-related mining. Declining oil prices -- especiallly the sharp drop in 1986 -- substantially curtailed oil and gas exploration, coal mine development, and related activities such as construction and financial services; strength in these activities had contributed to rapid employment growth in these regions before 1983.

Large overprojections of total employment growth due to declines in energy-related mining extended to States outside the Rocky Mountain and Southwest regions -- notably, Alaska, North Dakota, Louisiana, and West Virginia. The energy-related overprojections in Louisiana and West Virginia were especially large and d offset underprojections of total employment growth in other States in the interior Southeast region.

In the Plains region, total employment -- although projected, largely reflecting unexpected, weakness in the farm economy. In the 1970s, a long-term decline in the number of farm proprietors appeared to be bottoming out, as high prices encouraged farmers to retain, and even to expand, their businesses. In the first half of the 1980's a falloff in farm prices and high interest rates led to increased bankruptcies and continued decline in the farm economy. Employment in farm-related construction and trade, in turn, was adversely affected.

In the Great Lakes region, employment growth in goods-producing industries -- in particular, durables manufacturing -- was overprojected. Durables employment in the region was weak in the 1981-82 recession, as is typical in recessions. Uncharacteristically, durables employment -- except in motor vehichles -- did not recover strongly after the recession; in the face of competition from imports and from regions with lower labor costs, firms laid off workers and closed plants to cut costs. An underprojection of employment in service-type industries more than offset the overprojection in goods-producing industries, resulting in a total employment underprojection about equal to the national average.

Population growth

Projected growth in population for the Nation (2.8 percent) was close to measured growth (2.9 percent). All the regions projected to growth at a slower rate than a Nation did so, and all the regions projected to grow at a faster rate than the Nation -- except the Rocky Mountain region -- did so (table 4). The rocky Mountain region, projected to grow faster than any other region in 1983-863, grew more slowly than the national average. The overprojection was mainly due to sharp declines in job opportunities in energy-related industries.

The other regions for which BEA overprojected population growth are the Plains, interior Southeast, New England, and Mideast regions. In each, population in 1983-86 grow more slowly than in the Nation. In New England, the overprojection occurred despite large gains in employment; this overprojection reflected a wider gap than was expected in the responsiveness of population migration growing job opportunities in a region with a longstanding trend of relatively slow employment growth.

The regions for which BEA underprojected population growth are the Far West, coastal Southeast, and Southwest regions. In the Southwest, the underprojection occurred despite weakness in employment, reflecting -- in the converse of the New England case -- a wider gap than was expected in the responsiveness of population migration to reduced job opportunities.

TABLE: Table 1. -- Projected and Measured Percent Changes in Employment by Industry, 1983-86, United States

 TABLE: Projected Measured Difference(1) Projected Measured
 TABLE: Differences(1)

(1) Percentage-point difference between projected and measured percent changes, 1983-86. A negative difference indicates an underprojection, and a positive difference indicates an overprojection.

TABLE: Table 2. -- Projected and Measured Percent Changes in

 TABLE: Total
 TABLE: Line Projected Measured Difference
 TABLE: Agriculture, forestry, fisheries, Mining
 TABLE: Line and other
 TABLE: Projected Measured Differences Projected Measured Diffence
 TABLE: Construction Nondurables goods
 TABLE: Line Manufacturing
 TABLE: Projected Measured Difference Projected Measured Difference

TABLE: Table 3. -- Percentage-Point Differences Between Projected and Measured Percent Changes in Employment by Industry, 1983-86, United States and Regions

 TABLE: Services-Point
 TABLE: Total Finance, transpor-
 TABLE: Total Services insurance, Govern- tation and
 TABLE: and real ment Trade Public
 TABLE: estate utilities
 TABLE: Goods-producing
 TABLE: Agricul-
 TABLE: Durables ture Nondura-
 TABLE: Total Mining goods Forestry, ble goods Construc-
 TABLE: Manufac- fisheries, manufac- tion
 TABLE: turing and other

TABLE: Table 4. -- Projected and Measured Percent Changes in Population, 1983-83, United States, Regions, and States.

TABLE: Projected Measured Differences (1)

(1) Percentage-Point difference between projected and measured percent changes, 1983-86. A negative difference indicates an underprojections, and a positive difference indicates an overprojection.