Metropolitan, Urban/Rural and PUMA Areas

Missouri Census Data Center, Sept. 2003

Metropolitan Areas

  • The idea is to have county-based geographic clusters corresponding to densely settled areas with economic and cultural ties.

  • In earlier times (2000 and before) these areas required a central city of at least 50,000 people to qualify as a metro area (MA).

  • Now we have both Metropolitan and Micropolitan areas, and these fall under the general category title of Core Based Statistical Areas, or CBSA's. The criteria of a central city is replaced with that of a central population cluster, or urban area, and we have two size levels. Micropolitan areas need a core urban area of 10,000 to 50,000 people.

  • As a result of these changes, Jefferson City, MO now qualifies as a Metropolitan SA, even though the population of Jefferson City remains below 40,000.

  • The definitive web site for information about these concepts is This page has links to other pages which discuss both the general concepts as well as listings of the specific metro area definitions.

  • We have two sets of Metro Area definitions that you need to keep track of and differential between:

    1. The ones defined as of June 30, 1999 (see under the old central city standards. These are the entities that are identified on all 2000 Census products with 4-digit FIPS codes in fields named MSACMSA and PMSA (on our datasets, at least).

    2. The current definitions as published this spring using the new core-based standards. These are the CBSA's and include the new Micropolitan areas. See for details.

Micropolitan Areas

  • Technical point: should Micropolitan Areas even be discussed in a presentation about Metropolitan areas? These entities (as well as the new Metropolitan areas) are assigned new 5-digit FIPS codes which will make it difficult to do any linkage over time.

  • There is little or no data yet (summer 2003) aggregated to these entities that we know of. At least not by the Bureau. The MCDC has done some basic population counts. See, for example, /webrepts/geography/mocbsas03.pdf for a report with 2000 pop counts and 2002 estimates for Missouri CBSA's by county.

Other Metro Entities

  • Metropolitan Divisions are sub-areas of the new (CBSA) metropolitan areas. They are equivalent to the old Primary Metropolitan Statistical Areas (PMSA's). New MSA's that have metro divisions defined within them are thus somewhat equivalent to the old Consolidated Metropolitan Statistical Areas (CMSA's).

  • Combined Statistical Areas are combinations of new MSA's. You might think they would be like the old CMSA's, but they are not. They tend to be very large.

  • New England Town-based Statistical Areas are defined only in New England, and serve as alternate ways of defining metro areas there. This is similar to the old situation where MSA's, PMSA's and CMSA's were defined in terms of towns in New England (only) and we had NECMA's -- New England County-based Metro Areas -- as an alternative definition. The switch here is that now all metro areas in the US are county-based, including New England. But New Englanders can still choose to use an alternate defintion if they choose.

    Urban/Rural Concepts

  • The definitive description is at (The new stuff is at the bottom of the page - the final hyperlink.)

  • Urban is not metropolitan and Rural is not non-metro. Urban/Rural is much more rigorous as it attempts to classify an area as being densely or sparsely populated. It is block-based, while metro/micropolitan areas are based on complete counties.

  • Urbanized Areas and Urban Clusters (the term used for smaller versions of the same concept) are the basis of the new urban/rural definition. This is the definition that is used in classifying 2000 Census data.

  • The gist of the new definition of urban, taken from the official web page at : ... densely settled territory, which consists of:
    • core census block groups or blocks that have a population density of at least 1,000 people per square mile and

    • surrounding census blocks that have an overall density of at least 500 people per square mile
    In addition, under certain conditions, less densely settled territory may be part of each UA or UC.

  • It was the "certain conditions" that took years to pin down, and that only a handful of GIS gurus can totally comprehend. The key is to create densely settled contiguous areas, but not to stop the contiguousness because of a large sparsely populated area (such as a lake or a park) that really should be included as part of the urban entity because it is surrounded by the densely populated parts.

  • The difference between now and 1990 is that there were no Urban Clusters (densely populated areas such as Poplar Bluff or Hannibal) -- only the larger Urbanized Areas. In 1990 you were considered urban if you lived in an Urbanized Area or in a place of 2500 or more people (including a Census Designated Place).

  • The new definition of urban takes advantage of advances in technology that make it much easier to define the urban clusters that are at the heart of the new definition. They also recognize the fact that whether or not you happen to live within a city limit line is no longer a reliable indicator of whether you live "in town" or "in the country". A lot of small-town sprawl takes place just outside the city limits.

    PUMAs: Public Use Microsample Areas

    • The smallest geographic entities that are identified on the PUMS microdata files.

    • If the area for which you want data is comprised of PUMA's you are very much in luck, since it is rather easy to build almost any kind of table you want at the PUMA or PUMA-aggregation level.

    • There are two sets of PUMAs for each state:
      1. The 1% or "Super PUMA"s, used as the only unit of geography on the PUMS 1% sample files. These areas have a minimum population of 400,000.
      2. The 5% PUMA's, which nest within the Super PUMAs, and appear as identifiers on the 5% PUMS sample records. These areas have a minimum population of 100,000.

    • The 5% PUMAs for Missouri are (almost?) identical to those used for the 1990 census.

    • In St. Louis and Kansas City there are large counties that are made up of multiple 5% PUMAs, while in rural portions of the state, like the Northeast corner, there can be a dozen or more counties comprising a single 5% PUMA.

    • Most PUMA's in Missouri are defined in terms of complete counties or census tracts. They can also be comprised of incorporated places.

    Using MABLE/Geocorr2k to Access the Codes, Names and Populations

    • Users can use the MCDC's MABLE/Geocorr2k web application to get detailed information about all of the geographic entities mentioned on this page (except for Super PUMA's).

    • The utility allows you to select different geographic coverages from a twin pair of Source/Target geographic select lists. The program creates output (as a delimited file and/or a report) listing all the geographic coverages selected along with the population of their intersections.

    • You can specify "Entire Universe" as the Target geography to get a simple listing of the specified coverage(s).

    • For example, we ran the program with the following specifications:
      • State = Missouri (default)
      • Source Geocode(s) = County
      • Target Geocode(s) = Urbanized Area/Urban Clister (2000)
      • Weight Variable = Population (default)
      • List output format = pdf

    • The resulting pdf format report was copied to a permanent page on our web site and can be viewed at /webrepts/geography/moco2uauc.pdf#pagemode=none.

    • The example is relatively simple, but you can use geocorr2k to do more involved extractions, to answer questions such as:
      • What portion of each ZIP Code (or School District, or State Senate District, or Regional Planning Commission, etc) is classified as rural?
      • What census tracts or portions of census tracts comprise each State House District.
      • How do Missouri and California compare in terms of what portions of their population live in micropolitan areas.