Analysis of NAICS Code Changes over Time for Project tx2414 - “Racial Residential Segregation and the Distribution of Health-Promoting Establishments in Neighborhoods over Timeâ€

This Census project broadly examines the relationship between changing demographics and the spatial distribution of business establishments over time. As such, this project uses several restricted Census products. It links the Longitudinal Business Database (LBD) and the County Business Patterns Business Register (CBPBR) to the Decennial Census and the American Community Survey (ACS) from 1992-2020. The goal is to better understand the associations between changing residential demographics, racial/ethnic residential segregation, and the spatial distribution of formal health-promoting organizations over a twenty-eight year period. To accomplish this, it relies on longitudinally-linked data on organizations as organized by industry codes. One major consideration is that in 1997, the Census Bureau started using North American Industry Classification System (NAICS) codes in lieu of Standard Industrial Classification (SIC) codes to classify industries. This was done in order to more accurately represent the activities of an establishment, and to better reflect the changing nature of industry and employment in the United States. However, this would affect the comparability of the data from 1992-1996 as compared to the data from 1997-2020 available in the data set. The Census Bureau recently completed the process of adjusting the earlier data to include both SIC and several NAICS coding schemes in a longitudinally consistent fashion, and is available as an additional file to the LBD. In 2021, the agency also adjusted the data to include the 2017 vintage NAICS code versions. The goal of this technical report is to provide benefit to the Census by analyzing these codes over time for consistency using the CBPBR and the LBD, especially for those industries and services that are related to health and well-being. From the analysis presented here, the NAICS codes appear to be relatively stable over time. Very few organizations experience changes over time using the 2017 NAICS code vintage over the 28 year period. Moreover, where these changes occur, they appear to be relatively stable over time (a similar amount of change from year to year), and not particularly concentrated in any one industry. Finally, any changes observed in the NAICS codes over the different vintages appear to be primarily related to changes in the coding schemes themselves for the different NAICS vintages. Thus, we conclude that the normalization over time for the NAICS codes, at least for this 28 year period and for this group of health-related organizations, appears to be consistent.