This website presents metropolitan-level (PMSA) 1990 Census occupation counts converted to their equivalent Census 2000 categories for a consistent metropolitan region geography, and for “Management, Professional, and Related Occupations” we give 1990 counts based on an alternative crosswalk developed at the Mumford Center prior to the release of the Census Bureau’s SF-3 Occupation Table Crosswalk. We also show the effect of the Census and Mumford conversions on the 1990 counts (as the percent change between the conversion counts and the 1990 STF-3 Occupational Categories “Executive, Administrative, & Managerial Occupations” plus “Professional Specialty Occupations”) and the 1990-2000 percent change between the Census 2000 SF-3 count of “Management, Professional, and Related Occupations” and the Census and Mumford 1990 conversion counts of this occupation class.
The comparability of occupational classifications across censuses is important for analyzing long-term trends in employment and other characteristics of workers. Unfortunately, Census 2000 uses a different standard occupational classification system than did the 1990 Census. Census occupational codes in 1990 were based on the 1980 Standard Occupational Classification (1980 SOC) system that organized occupations into a hierarchical structure wherein knowledge, skill level, and the experience considered necessary for new entrants to an occupation determined an occupation's place in the classification system. In contrast, Census 2000 occupational codes are based on the 1998 Standard Occupational Classification (1998 SOC) coding structures in which occupations are grouped according to "job families." The general concept behind job families is to combine people who work together producing the same kinds of goods and services regardless of their skill level. For example, doctors, nurses, and health technicians are all members of a job family (SOC User Guide, 2001). This conceptual change makes it more difficult to identify the highly qualified occupations that were traditionally placed in the management/professional category at the top of the job hierarchy. In addition to these changes in classification structure, the 1998 SOC has more professional, technical, and service occupations and fewer production and administrative support occupations than the 1980 SOC, reflecting advances in information technology, the shifts to a service-oriented economy, and increasing concern for the environment (BLS Report 929, 1999). Added and deleted occupations make it more complicated in many cases to know what occupation in one year is the equivalent of an occupation in the other year.
The Census Bureau’s website American FactFinder (AFF) now provides data from both the 1990 Census and Census 2000 and a separate website (http://www.census.gov/hhes/www/sf3occcross_menu.html) provides an occupation crosswalk template, documentation, and detailed instructions on downloading occupation counts from AFF and using the crosswalk template the 1990 occupational classification to an equivalent 2000 classification. The Census crosswalk links detailed 1990 occupations to their aggregated 2000 equivalents via weights derived from “double-coding” samples of census data into both the 1990 and 2000 industry and occupation classifications. An alternative method used by the Mumford Center matched occupation titles for both censuses using the Dictionary of Occupational Titles to guide the matches. This effort was restricted to occupations corresponding to the Census 2000 SF-3 occupation category “Management, Professional, and Related Occupations” and is documented in a Mumford Center report (http://brownS4.dyndns.org/cen2000_s4/report.html).
The Census Bureau’s AFF website makes no attempt to define a consistent geography in metropolitan regions between census years. For example, some towns included in a 1990 metropolitan definition are excluded from the 2000 definition, while (more commonly) towns included in the 2000 definition were excluded from the 1990 definition. As a consequence, 1990-2000 change in counts of equivalent occupations as shown by AFF may be an artifact of metropolitan definitions rather than an indication of change in the distribution of occupations. The occupation counts available at this website are for 331 metropolitan regions with consistent 1990 and 2000 geographic definitions. The 1990 STF-3 Occupation Categories to 2000 SF-3 Occupation Categories conversion counts based on the Census method given here differ from those available at the Census Bureau’s AFF only by differences in metropolitan definitions. Using a consistent 1990-2000 metropolitan definition generally results in a larger 1990 count and a smaller 1990-2000 percent change than given by AFF.
description of the census data files used for the analysis on these
pages can be found here.
(c) 2002 by The Lewis Mumford Center for Comparative Urban and Regional