data analysisThe EEOC offers its “Special” data file as a benchmark to guide employers in ensuring their organization’s internal workforce is in line with a relevant local labor pool. Still, great care must be taken when interpreting discrepancies between internal and external labor market measures that seek to establish equity in race and/or gender representation, or when considering questions of compensation equity within a job.  It is critical to keep in mind that unlike narrowly defined company job titles—which group employees who perform similar tasks, requiring similar skills and qualifications, in a comparable workplace—the EEOC’s “benchmark” comparator job group (into which a company’s internal job title falls) is likely to contain a variety of “similar” individual job titles, all of which have been aggregated into a single job group.

To give just three examples of this phenomenon, consider the different job skills that have been aggregated into the EEOC’s Tabulation Codes 2600, 3840 and 3740:

  • EEO Tabulation Code 2600 (Art and design workers)

Floral Designers (SOC code 27-1023) generally do not use computer-assisted design software or have a background in industrial safety that would constitute technical skills comparable to Commercial and Industrial Designers (SOC code 27-1021).

  • EEO Tabulation Code 3840 (Fish and game wardens and parking enforcement officers)

Parking Enforcement Workers (SOC code 33-3041) generally do not have a background in biology or a bachelor’s degree, as compared to Fish and Game Wardens (SOC code 33-3031).

  • EEO Tabulation Code 3740 (Firefighting and prevention workers)

Fire Inspectors (SOC code 33-2020) may not meet all of the physical requirements generally needed to perform the work of Firefighters (SOC code 33-2011).

At an extreme, it could be that a group of company employees within the same job title exactly mirrors the set of all employees with the same skill set and qualifications, and fulfilling the same job requirements, within the local labor market—but that scenario is unlikely.  When comparing the representation of the company’s employees against the EEOC’s closest “Tabulation Code” employees, substantial and meaningful differences in gender and race representation and in compensation among employees may seem evident; however, these differences could be solely due  to the aggregation of similar-sounding job titles whose job requirements are in fact very different.

It is crucial that meaningful distinctions between worker populations be preserved in any comparisons between a company’s internal employee base and external labor market measures. Before employing the EEO data, it is critical to understand the makeup of specific job titles making up the aggregate measure–and when the EEOC changes its aggregates (as has been recently proposed) category changes, these should be validated to assure they do not degrade the accuracy of the data for the job titles they represent. This loss of data precision in government-provided data is especially disappointing, given the trend in data science is going the other direction, as private data firms are providing increasingly granular and customized data for their customers who can pay for it.

Economists at Welch Consulting can incorporate other statistical data sources, such as the U.S. Bureau of Labor Statistics’ Current Population Survey or state employment service data where additional analysis of a job or job group is necessary to obtain accurate availability statistics.

The opinions expressed are those of the author(s) and do not necessarily reflect the views of our firm or its clients.

 

Learn more:

EEOC/OFCCP Audit Support Services

Job Analysis and Pay Discrimination Case Study