Skip Main Navigation Skip To Navigation Content

Equal Employment Opportunity (EEO) Tabulation

Skip top of page navigation
You are here: Census.govPeople and HouseholdsEqual Employment Opportunity (EEO) Tabulation MainAbout › › User Note About the Effects of Rounding in the EEO Files (Revised on 1/5/2004)

User Note About the Effects of Rounding in the EEO Files (Revised on 1/5/2004)

The Census 2000 Special Equal Employment Opportunity (EEO) Tabulation was created as a special tabulation that was contracted by and paid for by a consortium of four Federal Agencies, consisting of the Equal Employment Opportunity Commission (EEOC), the Department of Justice (DOJ), the Department of Labor (DOL), and the Office of Personnel Management (OPM).

The Census Bureau’s internal Disclosure Review Board (DRB) is responsible for reviewing specifications for all census data products, and determining that no product format is approved that contains any degree of disclosure risk. To meet the DRB’s disclosure avoidance standards, the Special EEO Tabulation was created using approved rounding rules, that are described in the “Protecting Privacy” appendix located in the technical documentation and also on the EEO main web site. The rounding rules were applied to the Special EEO Tabulation estimates, as presented to the consortium and available on CD-ROM.

In the EEO residence and worksite files (#1-7 and #20-24), each occupational group contains combinations of characteristics that are individually rounded (for example, the sex – race group variables in file #1, such as White non-Hispanic males) and also a total civilian labor force variable that was constructed based on the unrounded total and was then rounded (named “total” on the files). This approach provides users with occupational group totals that are closer to the actual unrounded estimates. However, the total civilian labor force value may not equal the sum of the individually rounded characteristic variables for each occupational group. (This rounding procedure occurs for each observation in the file; for example, each geographic unit’s occupational group). In addition, the totals for a particular summary level of geography (state) will likely not equal the sum of the geographies within that summary level (counties).

In the EEO worksite/residence (also known as workflow) files (#8-19), the rounding of the total number of people employed at work (named “people_r” on the files) occurs for each observation in the file, which is a combination of occupational groups by race by sex for every place of work by place of residence occurrence.

How the rounding affects the EEO web tool output:

While the Special EEO Tabulation on CD-ROM contains no totals above the individual occupational groups, the EEO web tool does include those totals at the request of the contracting consortium of agencies. However, some totals in the EEO web tool output may not agree between tables for the same geographic areas and/or occupational groups, because of the required rounding of estimates in the underlying EEO files. For example, because the user can select from various EEO residence files that contain a different number of occupational groups (each individually rounded), the total civilian labor force for a geographic area may differ in the web tool output, depending on the table selected.

Data from the Special EEO Tabulation will not agree with estimates from other Census 2000-based products (such as Summary Files 3 and 4) because of universe differences and rounding. Use published SF-3 and SF-4 data products to obtain estimates of the total civilian labor force, total number of employed, or high-level occupational groups for the civilian employed population. The Special EEO Tabulation gives users access to more detailed occupational group data.


Source: U.S. Census Bureau | Equal Employment Opportunity (EEO) Tabulation |  Last Revised: 2013-03-08T12:44:59.662-05:00