Redefinition of the BEA economic areas.

Johnson, Kenneth P.

This article presents the new regional economic areas defined by the Bureau of Economic Analysis (BEA) and discusses the procedures used to arrive at this disaggregation of the Nation on an economic basis.(1) The new disaggregation has 172 economic areas, and it replaces the 183-area disaggregation that BEA first defined in 1977 and then revised slightly in 1983 (table 1 and charts 1 and 2). The redefinition was undertaken in 1993 largely to incorporate newly available information on commuting patterns.(2)

To facilitate regional economic analysis, BEA provides geographically detailed economic data by economic area, as well as by State and by local area. BEA assembles economic area data on earnings by industry, employment by industry, total personal income, population, and per capita personal income. These data may be used to analyze local area economic activity, local interindustry economic relationships, and interarea population movements. In addition, the areas are used as major units for BEA'S local area economic projections.(3) Historical and projected economic area data are used by government agencies for planning public-sector projects and programs, by businesses for determining plant locations and sales territories, and by university and other research groups for doing regional economic studies.

Each economic area consists of one or more economic nodes - metropolitan areas or similar areas that serve as centers of economic activity - and the surrounding counties that are economically related to the nodes. The main factor used in determining the economic relationships among counties is commuting patterns, so each economic area includes, as far as possible, the place of work and the place of residence of its labor force. The decision to redefine the areas reflects substantial changes in the commuting patterns, as indicated by data from the 1990 Census of Population, and changes in the definitions of metropolitan areas.(4)

In general, the redefinition procedure has three major elements. The first element is the identification of nodes. The second element is the assignment of counties to relatively small economic units known as "component economic areas" (CEA'S); each CEA consists of a single economic node and the surrounding counties that are economically related to the node.(5) The third element is the aggregation of the CEA'S to the larger economic areas. For a diagrammatic representation of the redefinition procedure, see chart 3.

Identification of nodes

Economic nodes are metropolitan areas or similar areas that serve as centers of economic activity. Of the 3,141 counties in the Nation, 836 are metropolitan counties that make up the 310 metropolitan areas; each of these areas was identified as the node of a CEA.(6) In addition, in parts of the Nation remote from metropolitan areas, 38 nonmetropolitan counties were each identified as a node.

Identification of most of the nonmetropolitan nodes was a four-part process. First, analysis of commuting data for the Nation's 2,305 nonmetropolitan counties showed that 1,112 of these counties are not closely related to a metropolitan area. Second, of these 1,112 counties, 130 are locations of newspapers.(7) Third, of these 130 counties, 68 have populations of more than 50,000, or their newspapers are widely read in at least five counties, or both. Fourth, only 35 of the 68 counties qualified as nodes of CEA'S that could contain at least five counties. The CEA of each of these 35 nodal counties was named for the city in which the county's major newspaper is published.(8)

In addition, three nonmetropolitan counties were identified as nodes of CEA'S because the county contained the largest city in the CEA. These CEA'S, which are characterized by their relative economic isolation, are the Alaska panhandle, western Oklahoma, and northern Michigan.

Assignment of counties to component economic

areas

Of the 3,141 counties in the Nation, 836 counties constitute the 310 metropolitan area nodes, and 38 counties are identified as nonmetropolitan nodes; together, these 874 counties constitute 348 nodes. Each of the remaining 2,267 non-nodal counties was analyzed to determine the node to which it is most closely related. About three-fourths of these counties were preliminarily assigned to nodes on the basis of their largest county-to-county commuting flows, according to journey-to-work data from the 1990 census. In many instances, the assignment reflected commuting flows to non-nodal counties already assigned to nodes rather than commuting flows to nodal counties. Most of the other counties were preliminarily assigned to nodes on the basis of the locations of the regional newspapers that are most widely read in those counties, according to newspaper circulation data.(9) For all preliminary assignments, the non-nodal counties had to be contiguous to either the nodes or to non-nodal counties already assigned to the nodes.

The preliminary assignment of non-nodal counties to nodes - based on data at the county level - resulted in a preliminary set of CEA'S. Data at both the county and CEA levels were then analyzed to ensure that, to the extent possible, each county was assigned to the CEA to which it has the largest commuting flow. This analysis resulted in the reassignment of 373 counties and in the definition of the final set of 348 CEA'S.

Aggregation to economic areas

The 348 CEA'S were used as "building blocks" for the new 172 economic areas. The CEA'S were aggregated to economic areas so that (1) each economic area includes, as far as possible, the place of work and the place of residence of its labor force and (2) each economic area is economically large enough to be part of BEA'S local area economic projections program.(10) In general, the aggregation had two parts. First, the 59 CEA'S with primary metropolitan statistical areas (PMSA'S) as nodes were combined into 17 economic areas, which mainly correspond to the 17 consolidated metropolitan statistical areas (CMSA'S) that comprise the PMSA'S.(11) Second, each of the 143 CEA'S that do not meet criteria for minimum size, for moderate commuting across CEA boundaries, or for both, was combined with the CEA to which it has the largest commuting flow.(12)

By definition, the labor force of an economic area should work and reside in that area, so commuting across boundaries should be limited. An evaluation of journey-to-work data from the 1990 census indicated that net numbers of commuters across the new economic area boundaries are indeed relatively low.(13) About 80 percent of the 172 areas have net commuting rates of 1 percent or less.(14) In contrast, again according to the 1990 journey-to-work data, only about 60 percent of the 183 areas defined in 1977 have net commuting rates of 1 percent or less.(15)

Availability of Additional Information

The codes, names, and numbers of the counties in each economic area and CEA and of the CEA'S in each economic area are available electronically on the Economic Bulletin Board (EBB) from the Commerce Department's STAT-USA. To access the EBB, use a personal computer and modem, dial (202) 482-3870, and follow the instructions. To access the EBB through Internet, use Telnet address "ebb.stat-usa.gov" for remote login, and download the file named "eacodes.exe." For prices and other information about these services, call (202) 482-1986.

The economic area information is also available on a 3 1/2-inch, high-density diskette for $20. When ordering, please specify the BEA Accession Number 61-95-40-101. Send your order, along with a check or money order payable to "Bureau of Economic Analysis," to Public Information Office, Order Desk, BE-53, Bureau of Economic Analysis, U.S. Department of Commerce, Washington, DC 20230. For further information or to order using MasterCard or VISA, call (202) 606-3700.

(1.) See "Proposed Redefinition of the BEA Economic Areas," Federal Register 59 (November 7, 1994): 55,416-20; and "Final Redefinition of the BEA Economic Areas," Federal Register 60 (March 10, 1995): 13,114-18. (2.) See "Intent to Revise the Boundaries of the BEA Economic Areas," Federal Register 58 (March 9, 1993): 13,049-50. See also Kenneth P. Johnson and Lyle Spatz, "BEA Economic Areas: A Progress Report on Redefinition," Survey of Current Business 73 (November 1993): 77-79. (3.) See Regional Economic Analysis Division, "BEA Economic Area Projections of Income, Employment, and Population to the Year 2000," Survey 70 (November 1990): 39-43. (4.) The redefinition reflects the changes in the metropolitan-area definitions issued in June 1993 by the Office of Management and Budget for statistical purposes; the definitions of metropolitan areas used by BEA are the county-based definitions. The 310 metropolitan areas consist of 240 metropolitan statistical areas, 59 primary metropolitan statistical areas (PMSA'S), and 11 New England county metropolitan areas (NECMA'S). (BEA treats the New Haven-Bridgeport-Stamford-Danbury-Waterbury, CT NECMA as a PMSA.) (5.) Data for the CEA'S can be used by government agencies for administering regulatory programs for small areas and by businesses for developing marketing programs for small areas. (6.) The 3,141 counties are those defined as of January 1, 1990; they consist of counties and of areas classified as county equivalents for the 1990 census. (7.) Data by county on newspaper publication and circulation are from the Audit Bureau of Circulations, an organization whose membership accounts for about 98 percent of U.S. newspaper circulation. (8.) The cities are Flagstaff, AZ; Jonesboro, AR; Idaho Falls, ID; Twin Falls, ID; Quincy, IL; Manhattan, KS; Paducah, KY; Bowling Green, KY; Salisbury, MD; Traverse City, MI; Marquette, MI; Mankato, MN; Worthington, MN; Hattiesburg, MS; Meridian, MS; Tupelo, MS; Greenville, MS; Missoula, MT; Butte, MT; Grand Island, NE; North Platte, NE; Norfolk, NE; Scottsbluff, NE; Lebanon, NH; Hobbs, NM; Farmington, NM; Minot, ND; Pendleton, OR; Aberdeen, SD; Watertown, SD; Cookeville, TN; Lufkin, TX; Staunton, VA; Clarksburg, WV; and Bluefield, WV. Hattiesburg, MS was defined as a metropolitan statistical area by the Office of Management and Budget in mid-1994, after the redefinition was under way (see footnote 4). (9.) The preliminary assignment of a small number of counties with special features, such as unusually small populations, was based on other procedures. (10.) In its forthcoming set of regional projections, BEA plans to publish projections for States in the summer of 1995 and projections for the new economic areas and for metropolitan areas in early 1996. (11.) A CMSA has more than 1 million residents and comprises two or more PMSA'S. (12.) The criteria for minimum size were developed from a combination of data on land area, on number of employed residents, and on number of counties, and the commuting criteria were developed from journey-to-work data from the 1990 census. (13.) The net number of commuters is the difference between the number of in-commuters (nonresidents who commute to work in an economic area) and the number of out-commuters (residents who commute to work out of an economic area). (14.) The net commuting rate is the difference between the in-commuting rate and the out-commuting rate; the rate of in-commuting (or out-commuting) is the number of in-commuters (or out-commuters) as a percentage of the number of employed residents, regardless of their place of work. (15). In the early 1980's, when definitions of the 183 areas were confirmed on the basis of commuting data from the 1980 census, about 80 percent of the 183 areas then had net commuting rates of 1 percent or less.