Virginia Parks, Ph.D.
Recoding 1990 Employment Data to the 2000 Industrial Classification System.
In 2000, the Census Bureau changed its industrial classification system to accord with the North American Industry Classification System (NAICS), a new categorization system developed by the U.S. Department of Commerce to replace the older Standard Industrial Classification System. Adopted in 1997 for use by all governmental agencies (U.S. Bureau of the Census, 2007), the NAICS standardizes industrial classification in economic data across the U.S., Canada, and Mexico and is meant to better represent sectors of economic activity in the contemporary economy. For example, one significant change is the creation of a new aggregate sector, "information," that includes industries such as newspapers, publishing, and radio and television broadcast.
As a result, employment by industry is not directly comparable between the 1990 and 2000 Census. To address this correspondence problem, Scopp (2003) constructed a crosswalk table that allows researchers to redistribute 1990 employment into the 2000 industry categories or vice versa. I do the former in my research, as I am interested in tracing racial, ethnic, and gender patterns in the "new economy"-representative sectors of which, especially in services and technology, are better captured in the 2000 classification system. To determine 1990 employment within the "new" industries, Scopp's crosswalk reassigns, by percentage, employment within each "old" 1990 industry across the new industry categories. Put differently, Scopp's crosswalk "maps" how jobs were reclassified between 1990 and 2000, depicting where jobs were "moved" from and where they were "moved" to (e.g., the "new" traveler accommodation industry comprises 97% of employment in the "old" 1990 category "hotels and motels" plus 2.5% of employment in the 1990 category "lodging places, except hotels and motels" plus 0.8 % of jobs in the 1990 category "eating and drinking places" plus 1% of jobs in the 1990 category "real estate" plus 2% of jobs in the 1990 category "misc. entertainment and recreation").
I wrote the following SAS program to reallocate 1990 employment into 2000 industrial categories (comments, suggestions, and/or corrections are welcome) according to Scopp's crosswalk. Note that this program reallocates PUMS employment data at the 3-digit industry level and gives employment totals for each industry. Given the uneven racial and gender division of labor and its regional variation, it is highly recommended to reallocate employment separately for each ethnoracial, nativity, and/or gender group and for each metro area that may be of interest to you (as per Scopp, 2003, p. 5). This can be done easily by subsetting your data at the beginning of the program.
For example, in my analyses of the Chicago and other labor markets, I did just this. To continue the previous example, I calculated the employment of native-born Black women in the "new" industry category "traveler accommodation" by adding 97% of all jobs held by native-born Black women in the "old" category "traveler accommodation" plus 2.5% of all jobs held by native-born Black women in the "old" category "lodging places, except hotels and motels" and so on.
Lastly, here's the citation for Scopp's 2003 paper. I highly recommend that you read it before running this program:
Scopp, T. S., 2003, The Relationship Between the 1990 Census and Census 2000 Industry and Occupation Classification Systems. Washington DC: United States Census Bureau.
This material is based upon work supported by the National Science Foundation under Grant No. 0525667. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation.
Click here to download source code for SAS
The School of Social Service Administration
The University of Chicago main home page
vparks@uchicago.edu