The first extensive attempt by the Census Bureau to define areas based on the metropolitan area concept was the identification of industrial districts for the Census of Manufactures of 1905, which showed such districts for New York, Chicago, Boston, and St. Louis. The Census Bureau gave official recognition to the metropolitan area concept for decennial census purposes when it defined metropolitan districts for the 1910 census. These metropolitan districts were defined on a nationwide basis for cities having populations of at least 100,000.
The Census Bureau defined metropolitan districts again for the 1920 census, applying the same criteria that had been used in 1910. Metropolitan districts again were defined for the 1930 and 1940 censuses, but the criteria were modified for these censuses so that metropolitan districts for cities with minimum populations of 50,000 would be recognized. There were 96 metropolitan districts for the 1930 census, and 140 metropolitan districts for the 1940 census.
During the period 1910 through 1940, the Census Bureau defined metropolitan districts in terms of minor civil divisions (MCDs)--county subdivisions such as townships or election districts--and determined their boundaries primarily based on population density.
However, few agencies or organizations outside the Census Bureau compiled data for MCDs. As a result, federal, state, local, and private statistical groups could not readily prepare data and conduct socioeconomic analyses using the metropolitan district as a statistical base. By World War II, some of these groups had developed alternative metropolitan definitions in terms of whole counties that did not coincide with the Census Bureau's metropolitan districts or the metropolitan definitions of other agencies or groups.
Standard definitions of metropolitan areas were first issued in 1949 by the then Bureau of the Budget (predecessor of Office of Management and Budget or OMB), under the designation "standard metropolitan area" (SMA). The term was changed to "standard metropolitan statistical area" (SMSA) in 1959, and to "metropolitan statistical area" (MSA) in 1983. The term "metropolitan area" (MA) was adopted in 1990 and referred collectively to metropolitan statistical areas, consolidated metropolitan statistical areas, and primary metropolitan statistical areas.
The term "core based statistical area" (CBSA) was adopted by OMB in 2000 and refers collectively to metropolitan and micropolitan statistical areas, both of which are defined around urban centers, or "cores." Census Bureau-defined urbanized areas and urban clusters form the cores of metropolitan and micropolitan statistical areas, respectively.
OMB also introduced the concept of combined statistical areas, consisting of groupings of adjacent metropolitan and micropolitan statistical areas, and representing larger regions in which the component CBSAs are socially and economically integrated, but to a lesser extent than territory within individual metropolitan or micropolitan statistical areas.
On June 6, 2003, OMB announced definitions of 362 metropolitan statistical areas and 560 micropolitan statistical areas in the United States; eight metropolitan statistical areas and five micropolitan statistical areas in Puerto Rico; and 116 combined statistical areas in the United States and Puerto Rico. OMB announced updates to the CBSA classification as of December 2003, designating 13 new micropolitan statistical areas, merging two metropolitan statistical areas, and identifying nine new combined statistical areas.