TECHNICAL ASSISTANCE FOR FEDERAL AGENCIES IN USING THE 2006-2010 AMERICAN COMMUNITY SURVEY EQUAL EMPLOYMENT OPPORTUNITY TABULATION

Introduction

This document is intended to assist federal agency officials who prepare and submit reports pursuant to EEO Management Directive 715 (MD-715). As will be described in more detail below, beginning with their Fiscal Year (FY) 2013 MD-715 reports, agencies will use the 2006-2010 American Community Survey Equal Employment Opportunity Tabulation (EEO Tabulation), which the U.S. Census Bureau (Census) released on November 29, 2012.

This document will serve two purposes. First, it will alert EEO Tabulation users to the major changes concerning each agency's data and how it should be formatted for MD-715 reporting. It also will highlight the differences between the EEO Tabulation and the 2000 Census decennial Data Tool, each of whichwas designed to provide external benchmarks to assist federal agencies in monitoring employment practices and enforcing workforce civil rights laws. Second, after identifying and describing those tables designed for federal agency use, the document lists those elements of the EEO Tabulation that should not be used for MD-715 purposes, and explains the reasons.

Background

Over the past decade, the Office of Personnel Management (OPM): (1) created new occupational families and the specific occupations within those families; and (2) revised certain occupations within existing families. OPM has also deleted entire occupational families (and all the occupations within them), along with certain occupations in remaining families.

During this same time period, Census revised the way it collects the socio-economic data underlying the decennial reporting cycle behind EEO files covering 1970, 1980, 1990 and 2000. It eliminated the decennial "long form" and replaced it with an ongoing American Community Survey (ACS), which asks the same questions of fewer people each year rather than asking more people less frequently. Also, at the request of EEOC and other federal agencies responsible for EEO monitoring of the nation's civilian labor force ("CLF"), Census created the above-referenced EEO Tabulation based upon ACS data. It can be accessed utilizing Census's American FactFinder (AFF) on-line tool at https://www.census.gov/topics/employment/equal-employment-opportunity-tabulation.html

Using the EEO Tabulation

As noted above, beginning in FY 2013, covered agencies will prepare their MD-715 reports using the EEO Tabulation. In addition to preparing the workforce data tables, agencies, after consulting with EEOC's Office of Federal Operations (OFO), Federal Sector Programs (FSP), may conduct additional, more refined workforce analyses using the EEO Tabulation to sort employees into both EEOC's nine federal EEO occupational groups ("FedSec9") and the agency's major occupations. https://www.census.gov/topics/employment/equal-employment-opportunity-tabulation.html

Understanding the Tables

The EEO Tabulation contains fourteen Table Sets, which are tables related by content that filter data into specific categories. Federal agency users should concentrate on Table Sets 2 and 5 only, which are explained in more detail below. To access the AFF, see https://www2.census.gov/programs-surveys/demo/guidance/eeo/eeo-tabulation2006-2010-5-year-acs-data-tutorial.pdf.

The information contained in Table Sets 2 and 5 are similar, except in one respect. Table Set 2 applies to major occupations that exemplify an agency's mission and role within the federal government; the particular occupations can vary accordingly. Table Set 5 pre-sorts all federal jobs into one of nine (FedSec9) aggregations of occupations having similar skill levels, training, mobility, and other labor market characteristics.

All individual federal occupations are provided on a separate table, which contains the Census equivalent to each federal occupation. That table, EEOC Federal Sector Occupational Cross-Classification Table ("Crosswalk"), can be accessed at https://www.eeoc.gov/federal-sector/eeoc-federal-sector-occupation-cross-classification-table. For each occupation listed, the Crosswalk shows (1) the OPM occupation code and title, (2) the FedSec9 occupational category to which that job is assigned, (3) the external Census title and code, and (4) the Office of Management and Budget ("OMB") Standard Occupational Code ("SOC").

For federal agency users, the similarities between Table Sets 2 and 5 are illustrated in more detail below. Each is divided into three subsets, sorted on the basis of United States citizenship (which is important because most federal jobs are limited to citizens).

The tables are designated either:

Table Set 2 - EEO-ALL02R, EEO-CIT02R, EEO-NCIT02R, EEO-ALL02W, EEO-CIT02W, and EEO-NCIT02W;

Table Set 5 - EEO-ALL05R, EEO-CIT05R, EEO-NCIT05R, EEO-ALL05W, EEO-CIT05W, and EEO-NCIT05W.

After the prefix "EEO", which applies to the entire set of tabulations, each set is sorted by three additional categories:

The letter designation "ALL" covers all members of the CLF enumerated in that table set. The letter designations "CIT" and "NCIT" sort the observations on the basis of citizenship status, between citizens and non-citizens.
The numerical designations "02" or "05" tie the information to its particular table set, individual or FedSec9.
The letter designations "R" and "W" sort the observations on the basis of each worker's residence location or worksite location.

The R and W designations relate to the labor market geography for each worker. Depending on which table is used, the external benchmark of non-federal workers employed in equivalent occupations can vary. On that basis, EEOC regards the selection of a labor market area for MD-715 reporting purposes to be a critical element of each agency's reporting protocol. This is covered in more detail below.

The tables in both Table Sets 2 and 5 adhere to OMB's uniform geographic boundary specifications, known as the Federal Information Processing System (FIPS) codes. Agencies can select levels of geographic specificity that reflect its employment practices, its size, and the breadth of its workforce distribution, in terms of total numbers, as well as the location of designated activities or organizational subcomponents.

Defining the Geography

In the past, some agencies have defaulted to nationwide data to avoid complex and detailed comparisons, and to reflect the presence in all or many states of agency staff. The EEO Tabulation permits, when appropriate, labor markets below the nationwide level, especially when recruitment and hiring is done on a more localized basis and/or where applicant tracking data shows that persons applied only from a specific geographic area.

For EEO comparisons and MD-715 reporting purposes, agencies are instructed to use the Worksite (W) tables when benchmarking within specific geographic locations. Worksite tables reflect the distribution of jobs in a particular area on the basis of where employer facilities are located and the work is actually performed, both in the federal and private sectors. Where recruitment and hiring is done at the national level, agencies should use the Residence (R) tables. Thus, the fact that some agencies previously used Residence-based data because of problems obtaining comparable Worksite information, no longer constitutes a valid basis to do so, beginning with the FY 2013 reporting cycle, other than when dealing with U.S. totals and nationwide labor markets.

Even at the FedSec9 level of occupational aggregation shown in Table Set 5, it is less likely that an agency should use the nationwide option for all of its occupational categories. However, the nationwide (R tables) are appropriate when the agency: employs large numbers of workers in most states; benchmarks against Officials and Managers, and the Professional workers categories, (because these categories are often recruited nationally); or the category constitutes a substantial segment of that agency's workforce. Otherwise, availability can be calculated using an agency's own regional structure and then aggregated into a weighted average, or some other intermediate configuration.

A third, more local option can reflect the more localized labor markets that generally apply to blue-collar wage grade ("WG") workers, as well as technical, administrative support, and service workers in the general schedule ("GS") ranks. As with all aspects of this Guidance, technical assistance is available from EEOC's OFO/FSP staff, contact information provided on the last page of this document.

A few additional factors should be kept in mind when working with the EEO Tabulation. After selecting the appropriate table from the options described above, the next step is to specify the geographic level that will be used (before going to the next step of identifying the particular major occupation(s) or the FedSec9 occupational aggregations). For MD-715 purposes, the range of geographic levels is: nationwide, agency region (FIPS), state, or county of 50,000 people or more. Beyond those options, the most refined level (using Census terminology) is Place, which refers to cities, towns, or other places of 50,000 or more persons.

The importance of Place, when combined with the W, Worksite option, is three-fold. First, because the distribution of residential patterns in terms of race or national origin ("RNO") sometimes differs from the distribution of available work in the same labor market area, benchmark availability estimates often vary between the two. The EEO Tabulation automatically takes into account both those who live and work in a particular location, and those who commute to jobs in that specific location but live elsewhere.

Second, the Worksite tables have a sub-option that lists the locations from which people commute and provides the total commute flow. For each Worksite, this option identifies up to nine specific in-commutes, plus a residual composed of all individuals not otherwise coming from one of those locations. While MD-715 does not require commute-pattern information, such detail helps clarify anomalies that are revealed by analysis at that level. For example, an agency may find that a low participation rate for a RNO group could result from a lack of hiring from one specific contributing flow which contains a significant number of workers from that RNO group. Agencies are encouraged to use this approach with several of their most populous major occupations in the event that low participation rates or other anomalies are present and affect employment by RNO or sex.

Third, the accuracy of external availability estimates is increased when facilities are located in large counties, where there may be a multiple number of places that reflect worksite and commute flows within that larger geographic pattern. Los Angeles County, California, illustrates this point.

The absence of detailed occupational data for geographic areas of less than 50,000 persons on the EEO Tabulation is because of Census Bureau's confidentiality standards, which prohibit repeated cross-tabulations that could reveal the identity of specific individuals in Table Set 2. Because occupations in Table Set 5 are already aggregated into only nine categories, the geographic limit increases to areas of fewer than 100,000 when analyses are done at the FedSec9 level.

As specified above, agencies may only use Table Sets 2 and 5 for MD-715 analysis and reporting purposes. To understand more fully why other Table Sets should not be used, agency EEO and Human Resources ("HR") officials may be aided by the following information:

Whereas Table Set 1 contains much of the data that is found in Table Set 2, it does not sort on the basis of citizenship status. Table Set 1 is thus suited for private sector employer comparisons, but not for the federal sector.

Table Set 1 also contains some occupations that are populated only in the private sector but not in federal employment. Table Set 2 only covers occupations listed with a Census title and code and a SOC code, and that have federal counterparts. Furthermore, because OPM currently lists 681 individual GS and WG occupations, but the EEO Tabulation contains only 487 occupations, some federal jobs must be compared to the same external availability standard.

Table Sets 3, 4 and 6 cover non-federal employment at aggregated occupational level. Table Set 4 sorts the 487 Census occupations into the nine private sector categories used by EEOC's EEO-1 annual report form, while Table Set 3 conforms to a fourteen category alternative EEOC format. Similarly, Table Set 6 is used by the Department of Justice to compare state and local government employment to external availability for public jobs below the federal level, as covered by the EEO-4 survey.

Table Sets 7 through 14 contain occupational information sorted on the basis of age, income levels, industry, and educational attainment. Although EEOC has jurisdiction over the Age Discrimination in Employment Act, age is not a component of MD-715's affirmative employment requirements. The other variables enumerated above affect non-federal employment, especially in the private sector, but do not result in similar differences based on RNO or sex in the analysis of the federal workforce.

One additional limitation should be noted. While both Table Sets 2 and 5 include geographic options for Census-defined "Metro" or "Micro" Statistical Areas, they are not part of the labor market alternative set recognized for MD-715 reporting because of methodological limitations that can result in substantial inaccuracy. Several issues contribute to this problem. These include the assumption that availability by RNO status is uniformly distributed throughout the entire Area, as it would need to be for the estimate to be valid, and the issue of Areas composed of several counties where the worksite versus residence differences, along with the differential in- and out-commute patterns encountered across Metro Area boundaries, distort availability estimates by RNO and sex.

Selecting the Occupation

After defining the geography, agencies must identify its major occupations for benchmarking. The Crosswalk referred to earlier in this Guidance was updated in January, 2013, and provides the appropriate Census occupation code for each federal occupation. When utilizing the Census AFF tool, agencies will use the Census code identified by the Crosswalk to conduct its workforce benchmarking.

EEOC will amend the Crosswalk whenever OPM adds, deletes or modifies an occupational series. It is possible that an agency's current occupational distribution, using OPM GS and WG codes and titles, may include instances where a specific series is no longer active. Technical assistance concerning occupational cross-classification or the Crosswalk can be obtained from staff within EEOC's Office of Federal Operations.

Additional Information

Aggregation and Allocation

The EEO Tab tables follow a uniform format in displaying race and ethnicity data. For MD-715 purposes, agencies must modify the data. When reporting data, agencies must aggregate and report their workforce data as follows:

The two columns under Hispanic or Latino must be combined. These are identified as "White alone Hispanic or Latino" and "All other Hispanic or Latino."
The population within the category of "White and Black" must be combined and included in the "Black or African-American alone" category.
The population within the category of "White and AIAN" must be combined and included in the "American Indian and Alaska Native alone" category.
The population within the category of "White and Asian" must be combined and included in the "Asian alone" category.
The population within the category of "Black and AIAN" must be included in the "Two or more races" category.
The balance of the "Not Hispanic or Latino, two or more races" must be included in the "Two or more races" category.

For more information, please see OMB BULLETIN NO. 00-02 Guidance on Aggregation and Allocation of Data on Race for Use in Civil Rights Monitoring and Enforcement.

Puerto Rico Community Survey

The Puerto Rico Community Survey (PRCS) is the equivalent of the ACS for Puerto Rico, and began data collection in 2005. Data results from both the ACS and the PRCS are released together as a unified American Community Survey dataset. However, the PRCS data is not included in the U.S. totals for the ACS. Agencies should not modify the U.S. totals to include the PRCS data.

Where an agency is benchmarking labor market geography against a specific occupation, and Puerto Rico is a significant source of applicants, the agency should include Puerto Rico as a part of its labor market. Using the AFF tool, Puerto Rico can be added to a labor market through a sort of the "State" options.