What You Should Know: The National Academies’ Evaluation of Compensation Data Collected Through the EEO-1 Form

On July 28, 2022, the National Academies of Sciences, Engineering, and Medicine issued a Consensus Study Report evaluating the U.S. Equal Employment Opportunity Commission’s historic, first-time collection of pay data from certain private employers and federal contractors, which was completed in February 2020.  The report finds that the data the EEOC collected may be used to prioritize investigations and the agency’s allocation of resources to identify potential pay discrimination.  It also offers a number of recommendations for improving pay data collection by the EEOC if undertaken in the future.

Background on the EEOC’s Pay Data Collection

1.     What is the EEOC’s authority to collect pay data from private employers?

The EEOC’s mission is to prevent and remedy unlawful employment discrimination, including pay discrimination, under certain federal laws including Title VII of the Civil Rights Act of 1964 (Title VII) and the Equal Pay Act of 1963.  Title VII prohibits pay discrimination based on race, color, religion, sex, and national origin.  The EPA prohibits pay discrimination based on sex.

Pursuant to section 709(c) of Title VII, EEOC regulations require certain employers and other entities covered by Title VII to report workforce demographic information (referred to as “EEO information”) to the EEOC.  Since 1966, the EEOC has collected workforce demographic information from certain private employers and federal contractors through the EEO-1 form (now known as “Component 1” of the EEO-1).  In 2016, the EEOC added a pay data collection to the EEO-1 form for the first time, known as “Component 2.” The EEOC collected 2017 and 2018 pay data from private employers through the EEO-1 Component 2 between July 2019 and February 2020.

2.     Why did the EEOC decide to collect pay data from private employers?

Preventing and remedying pay discrimination in the workplace is core to the EEOC’s mission.  It’s been almost 60 years since the Equal Pay Act and Title VII went into effect.  Yet women who work full-time in the United States still make just 83 cents for every dollar paid to men.[1]  The pay gap is even wider for women of color.  Not all pay disparities result from discrimination, but discrimination in pay and promotions, as well as other discriminatory factors such as race and gender segregation in jobs and assignments, contribute to the problem.

The lack of access to pay data absent a specific charge of discrimination has been a longstanding barrier in the agency’s efforts to enforce federal laws prohibiting pay discrimination.  In 2010, the National Equal Pay Enforcement Taskforce recommended that the EEOC consider collecting pay information to support the agency’s law enforcement efforts.  Over the next several years, the EEOC engaged in an extensive, multiyear, public process that included two independent studies, a public hearing during which the Commission received written and oral testimony from experts with a wide variety of views, and two rounds of notice and comment from stakeholders and other members of the public.  In response to its notices proposing pay data collection, the EEOC received and considered hundreds of public comments from stakeholders, including employer and worker advocates.  The Commission considered this extensive public input in revising its proposal to collect pay data.  After considering the hearing testimony and written public comments, as well as the two independent studies, the EEOC ultimately decided in 2016 to collect aggregated data on pay and hours worked from certain private employers and federal contractors through Component 2 of the EEO-1 form.

3.     Who did the EEOC collect pay data from?

The EEOC collected 2017 and 2018 pay data from private employers and certain federal contractors with 100 or more employees.  By the time the pay data collection closed in February 2020, the EEOC had collected pay data from approximately 70,000 employers in each collection year, covering over 100 million workers.

4.     Why did it take the EEOC so long to collect pay data?

In September 2016, the EEOC announced approval from the White House Office of Management and Budget (OMB) for its pay data collection.  The pay data collection was scheduled to start in March 2018, giving employers 18 months to prepare for the new collection.  The EEOC conducted informational webinars for employers and stakeholders on the new data collection; set up a webpage dedicated to the pay data collection; and developed forms, factsheets, and Q&As to assist employers in complying with the new collection.

However, in 2017, OMB stayed the EEOC’s pay data collection before it began.  The stay was challenged in court, and a federal judge ultimately reversed it as unlawful and ordered the EEOC to implement the pay data collection that had been authorized and approved in 2016. The EEOC began collecting 2017 and 2018 pay data from employers in July 2019 and competed the collection in February 2020.

5.     What did the EEOC do with the pay data it collected from private employers?

Given the unique circumstances of the first-time data collection and the importance of pay data for enforcement purposes, the EEOC contracted with the National Academies of Sciences, Engineering, and Medicine (National Academies) in 2020 to study the pay data EEOC collected.  By unanimous, bipartisan vote, the Commission asked the National Academies to independently evaluate the pay data collected from private employers and provide recommendations for future data collections. 

6.     Is the EEOC currently collecting pay data from private employers?

No.  The EEOC voted in 2019, with White House approval, to discontinue the EEO-1 Component 2 pay data collection in the future.

Background on the National Academies and the Study Process

7.     Why did the EEOC select the National Academies to review the pay data collected from private employers?

The EEOC chose the National Academies to review the pay data the EEOC collected because it “provides independent, objective advice to inform policy with evidence.”[2]  The National Academies bring together “knowledge and expertise across disciplines to study complex and sometimes contentious issues, reach consensus based on the evidence, and identify the best path forward."[3]  National Academies reports “are viewed as valuable and credible because of the institution’s reputation for providing independent, objective, and nonpartisan advice with high standards of scientific and technical quality.  Checks and balances are applied at every step in the study process to protect the integrity of the reports and to maintain public confidence in them.”[4]  For each study, it selects panel members who serve as independent experts, not as representatives of organizations or interest groups. [5] Reports produced by the National Academies undergo a rigorous review process that includes review of the draft report by a second, independent panel of experts and then a final review by the National Academies to ensure the report meets its standards and that all review comments were carefully considered.[6]

8.     What are the qualifications of the panel members selected to study the pay data?

All study panels convened by the National Academies are created to be balanced and independent. All panelists are experts in their fields and are checked to ensure that they do not have conflicts of interest before their final selection.  Panel members are also regularly screened for conflicts throughout their participation. 

In this case, the National Academies selected 13 highly qualified experts to serve on its report committee.  The 13 experts represent a broad array of disciplines, including labor economics, sociology, statistics, survey design and methodology, employment law, race and gender equality studies, and diversity and inclusion evaluation.  The panel members come from different industries and backgrounds and represent different perspectives.  For example, one of the panel members works for the Association of Chamber of Commerce Executives, which typically represents business and employer interests.  Another member comes from the National Employment Law Project, which typically advocates for employee interests. 

All panel members have experience and background in understanding how data can be used to identify and fight discrimination.

9.     What sources of information did the National Academies consult in reaching its conclusions?

The National Academies analyzed the data the EEOC collected from private employers for reporting years 2017 and 2018, reviewed the scientific literature, and heard from stakeholders and subject matter experts over a period of more than a year.  The study panel held a series of public meetings to collect input from worker advocates, employer representatives, and state and federal agencies.  The Panel also deliberated in closed meetings to develop draft findings and recommendations free from outside influences.[7] 

Conclusions and Recommendations of the National Academies

10.  What did the National Academies find?

A summary of the National Academies’ conclusions and recommendations can be found beginning on p. 228 of the report.  Among the key findings for the EEOC’s purposes:

• Federal pay data collection is a necessary tool to find and fight pay discrimination effectively.
  • The lack of pay data has been a longstanding gap in enforcement of our nation’s pay discrimination laws and advancement of pay equity. (Report, p. 14-18).
  • Lack of transparency about wages “contributes to difficulties identifying cases of pay discrimination, enforcing of antidiscrimination laws, and charting progress toward more equal pay at the societal level.”  (Report, p. 13).
  • A number of hurdles contribute to underreporting of employment discrimination generally and pay discrimination in particular.  As the Panel put it, “some employees may be unaware that they are experiencing discrimination; some may be afraid of retaliation; and some may lack the data to justify a complaint.  Thus, there is a need for better and more complete data.” (Report, p. 15).
  • Collecting pay data is necessary to assess pay practices and differences in compensation by sex, race, and ethnicity. (Report, p. 28).
• The pay data EEOC collected may help EEOC better focus its resources to identify potential pay discrimination.
  • The EEOC pay data collection completed in February 2020 is unique and no other federal data collection captures the same information from private-sector employers. (Report 1, 3, 14-15).
  • The pay data EEOC collected (after “data cleaning”) may be used to help prioritize investigations and the allocation of resources; identify systemic discrimination; and analyze national, regional, and industry-based pay gaps. (Conclusion 7-1, Report, p. 6).
  • Standing alone, the pay data the EEOC collected are not sufficiently detailed to definitively identify employers engaged in pay discrimination but are useful to help the agency focus its resources, identify potential outlier employers, and sharpen and refine its investigations. (Report, p. 7).
• Using Component 2 pay data collected by the EEOC, the National Academies panel found significant pay gaps worthy of further investigation within Silicon Valley firms.
  • Using the EEOC pay data, the Panel identified specific, unnamed employers in the Silicon Valley tech sector with extreme pay gaps based on race, sex, and/or ethnicity compared to their industry counterparts.  
  • The Panel found that one employer, for example, had a –51.3% pay gap for Black male professionals compared to White male professionals. (Report, p. 214, 216-217).
  • Another employer had a –52.3% pay gap for Hispanic female professionals relative to White female professionals, which was substantially higher than the pay gap in the local labor market. (Report, p. 214, 220).
  • While these raw disparities do not tell the whole story, the Panel concluded that analysis of such pay gaps could be used to inform next steps in the investigation of charges of discrimination by, for example, assigning priority and resources to the case, requesting more information from the employer about any legitimate reasons for the pay differences, or closing the investigation. (Report, p. 29, 221).
  • The EEOC could also use pay data to identify industries, occupational groups, or geographic regions with significant disparities in pay by race, sex, or ethnicity for use in focusing its public outreach, education, training, and compliance assistance. (Report, p. 30).
• The EEOC achieved remarkably high response rates—about 90%, approximately 70,000 employers in each collection year —for this historic, first-time data collection that occurred under unique circumstances.  
  • The EEOC’s historic, first-time pay data collection took place under uniquely challenging circumstances. (Report, p. vii).
  • Despite these significant challenges, the EEOC achieved remarkably high response rates from employers submitting pay data for both years—about 90% —which exceeded the court-mandated response rate.  The Panel found the high response rates “notable” for a new data collection. (Report, p. 89).
  • Some industries, including utilities, finance, and insurance, achieved response rates approaching 96%. (Report, p. 91).
  • Even the lowest industry response rates—for accommodation and food services, administration and support, and retail—were still at 86%. 
  • Overall, the Panel found that the response rate for the new pay data collection was “at least as high” as the response to the EEOC’s longtime EEO-1 Component 1 demographic data collection. (Report, p. 112).
  • The Panel noted, however, that not every eligible employer was included on the EEOC’s master list of employers, and recommended that the EEOC take steps to identify additional private sector employers that meet the criteria for responding. (Report, p. 4).

11.  What recommendations did the National Academies make for future pay data collections?

The Panel recommended that the EEOC take steps to broaden and strengthen its data collections to advance pay equity and suggested short- and long-term refinements to any future pay data collections to enhance the EEOC’s ability to identify and remedy pay discrimination. (Report, p. 3, 10, 243). Many of the recommendations complement efforts already underway at the EEOC as part of the agency’s ongoing modernization of its EEO data collections and data analytics.  For example: 

• The Panel recommended that the EEOC expand and improve its “master list” of employers that are required to provide data to the agency.  (Report, p. 239).
  • The EEOC is actively working to improve the coverage of its data collections and those efforts have already borne fruit. 
  • Over 91,000 employers responded to the EEOC’s most recent EEO-1 Component 1 data collection—a significant increase over the number of employers that responded to the 2017-2018 pay data collection that the Panel studied. 
• The Panel also recommended that the EEOC take steps to improve data collection from professional employer organizations (PEOs). (Report, p. 238). Beginning with the 2021 EEO-1 Component 1 data collection, the EEOC updated the reporting procedures for third-party human resource organizations (including PEOs) to include among other updates, separate reporting for each client-employer of a PEO and client-employer certification of their own reports.  The updated procedures will allow the EEOC to collect more accurate employee demographic data in support of the agency’s mission to prevent and remedy unlawful employment discrimination and advance equal opportunity for all in the workplace. 
• The Panel also recommended that the EEOC eliminate the option for Type 6 (Establishment List) Reports for establishments with fewer than 50 employees. (Report, p. 238). Beginning with the 2021 EEO-1 Component 1 data collection, the Type 6 Establishment List Report was discontinued in order to collect more accurate employee demographic data at the establishment level in support of the agency’s mission to prevent and remedy unlawful employment discrimination and advance equal opportunity for all in the workplace.  For the 2021 EEO-1 Component 1 data collection, filers no longer had the option to choose either a Type 6 Establishment List Report or a Type 8 Report for establishments with fewer than 50 employees.  Rather, Type 8 Reports were required for such establishments.   
• The Panel also recommended that the EEOC collect pay data at the same time as general workforce demographic data to reduce the burden on employers and improve the accuracy and reliability of the data. (Report, p. 60).
  • That recommendation aligns with how the EEOC originally envisioned and approved the pay data collection. 
  • The pay data collection was separated from the traditional EEO-1 demographic data collection (Component 1) only due to the White House’s decision to stay the pay data collection in 2017 and the subsequent court order reversing the stay. 

The Panel made other recommendations for the EEOC to consider in any future pay data collections to reduce potential burden on employers and enhance the value of the data to the EEOC and the public.  For example: 

• Collect individual-level wage data from employers instead of collecting that information in pay bands because employers already provide individual-level wage data to state and federal agencies. (Report, p. 69).
• If the EEOC decides to continue collecting pay data within pay bands instead of collecting individual-level pay, consider adopting additional and narrower pay bands to better capture differences in pay, particularly for low-wage and high-wage earners. (Report, p. 242).
• Use W-2 Box 5 income instead of W-2 Box 1 income to better reflect employees’ total compensation. (Report, p. 240).

Next Steps

12.  Will EEOC use the 2017 and 2018 pay data it collected from private employers?

The Panel found that, after cleaning, the pay data the EEOC collected could be used to help prioritize investigations and the allocation of resources; identify systemic discrimination; and analyze national, regional, and industry-based pay gaps. Because of the unique circumstances of the collection, which was completed under court order, and the Commission’s decision to ask the National Academies to study the data, the EEOC did not implement standard post-collection cleaning of the data before turning the data over to the National Academies.  The EEOC would review and clean the data before deciding whether to use it for any of the purposes studied by the National Academies. 

Importantly, the EEOC would not, and never intended to, use pay data, by itself, as the sole evidence of pay discrimination.  Rather, consistent with the intended use of the data outlined by the EEOC in 2016, and as the Panel recommended, the EEOC would use pay data, together with other information, to inform next steps in a charge investigation, such as assigning priority and resources to the case, requesting more information from the employer about any legitimate reasons for any pay differences, or closing the investigation.

13.  Will the EEOC collect pay data again?

The EEOC is examining the report from the National Academies and will use it to guide its approach to any future pay data collection.  Before deciding whether or how to proceed with a future pay data collection, the Commission will seek and carefully consider the views of employees, employers, unions and members of the public.  Any new pay data collection would be preceded by public notice, a public hearing, approval by the Commission, and robust opportunities for public input.