Why Pay Data Matter in the Fight for Equal Pay

Download the PDF here.

Equal pay for women is often touted as an issue that enjoys widespread support; it is popular among the public across gender, race, and political affiliation.1 Few people, including policymakers, want to be viewed as being opposed to equal pay. But supportive words alone are not the same as effective action.

Securing equal pay for all women requires vigorous enforcement of equal pay laws and intentional work to combat discriminatory pay practices, including an in-depth understanding of how these practices play out for diverse groups of women. In order to be most effective, this work requires access to information about how women’s compensation compares with that of other workers, broken down by factors such as race and ethnicity, in order to assess whether everyone is being paid in an equitable and fair manner. Such pay data can reveal critical disparities that otherwise might go undetected and serve as a useful tool to help pinpoint areas where additional enforcement scrutiny is needed.

The lack of access to comprehensive pay data is a long-standing gap in the enforcement toolbox for federal agencies charged with enforcing equal pay laws. Efforts to correct this problem have been adopted but ultimately were short-lived, as opponents have worked to undo improvements and preserve a status quo that consistently depresses women’s wages. The Trump administration is spearheading the latest attacks to roll back the most recent progress on pay data collection, diluting and undermining hard-won gains aimed at strengthening equal pay enforcement. Rebuffing these attacks is critical—not only to improve equal pay enforcement, but also to help address persistent pay disparities that deny women earnings that they sorely need for themselves and their families.

The persistence of the gendered pay disparity

The persistent gap between women’s and men’s wages over decades has resulted in women earning consistently less than men across all races and ethnicities. In 2018, women working full time, year-round earned 82 cents for every dollar earned by men. When broken down by race and ethnicity, this gap was even steeper: For every dollar earned by white men, Black women working full time, year-round earned 62 cents; Latinas earned 54 cents; white women earned 79 cents; and Native women earned 57 cents.2 Asian women earned 90 cents for every dollar earned by white men, but the gap varies widely by subpopulation. Nepali and Cambodian women, for example, are estimated to earn 50 cents and 57 cents, respectively, for every dollar earned by a white man.3 Although there is much less research available, data on the earnings of transgender women and immigrant women also show pay disparities when compared with their male counterparts.4

Moreover, gender wage gap numbers are more than just interesting talking points; they have real-world implications for women in terms of depressed wages and devalued work, harming families who increasingly rely on women’s earnings to make ends meet. There are many factors5 driving this gender wage gap, some of which are quantifiable, such as differences between women and men in hours worked and time spent handling caregiving responsibilities, and others that are not easily measured. Understanding and tackling the portion of the gap that cannot be explained by measurable factors, which leading researchers and experts have attributed to discrimination,6 is particularly important. The work to ensure strong equal pay enforcement is specifically aimed at combating this discrimination head-on. Effectively challenging pay discrimination in the workplace and ensuring strong enforcement require access to the underlying data that can reveal pay differences and practices. Without concrete data, it is nearly impossible to show where disparities are occurring and who is most affected by these disparities. For years, equal pay enforcement has been hampered by the lack of access to comprehensive pay data that could help show these disparities. Although enforcement agencies can access such data when investigating a specific complaint or conducting a compliance review, agencies do not have comprehensive data for all employers. As a consequence, enforcement agencies are not able to look broadly across occupational categories or industries to identify specific trends or potential areas for further scrutiny.

Efforts to provide enforcement agencies with better access to comprehensive pay data7 have been met with stiff opposition from the business lobby and opponents of vigorous equal pay enforcement for years. Nonetheless, the Obama administration undertook a series8 of executive actions to address barriers to equal pay, one of the most important of which was to require more comprehensive reporting of pay data. In a rule9 finalized in September 2016, the Obama administration required employers with 100 or more employees to report their compensation data broken down by race, gender, and ethnicity in their annual filing of a form called the Employer Information Report, or EEO-1, form. The form is used to collect data about the composition of an employer’s workforce by race, gender, and ethnicity across 10 occupational categories, referred to as Component 1 of the form. The Obama rule added a new component—referred to as Component 2—requiring employers to provide information on compensation broken down by race, gender, and ethnicity across the same 10 occupational categories in 12 pay bands for each category. The EEO-1 form is subject to review by the Office of Management and Budget (OMB) every three years under the Paperwork Reduction Act, and the modifications to the form were approved during the regular renewal process for the form.

Unfortunately, the Trump administration has backtracked on the Obama pay data collection rule, along with other important equal pay reforms.10 In August 2017, the OMB released a memorandum directing the Equal Employment Opportunity Commission (EEOC) to halt implementation of the pay data collection rule with only a cursory explanation.11 This action was subsequently challenged by women’s and civil rights legal advocates in federal court and found to be arbitrary and improper.12 The court then directed the EEOC to collect the EEO-1 Component 2 pay data for 2017 and 2018, consistent with the Obama rule. However, the Trump administration has persisted with trying to eliminate the pay data collection requirement. The EEOC published a notice13 in the Federal Register on September 12, 2019, indicating its intent to seek a renewal of the EEO-1 form for another three years but without the pay data component. The notice attempts to justify the discontinuation of the EEO-1 pay data collection requirement by raising concerns about employer burden and overall utility. But the arguments put forward fall far short of providing a sufficient rationale for weakening the EEO-1 form’s existing data reporting structure that includes pay data. Moreover, preserving tools such as pay data collection is crucial to promoting broader, more rigorous equal pay enforcement and accountability. Rolling back this type of progress would be a serious setback for women at a time when stepped-up enforcement is sorely needed to promote equal pay.

Pay data collection is an essential tool to combat pay discrimination

Access to pay data is an essential prerequisite to combating pay discrimination and creating greater accountability for workplace pay practices. Pay data may not answer every question about whether discrimination has taken place, but without it, showing that a violation has occurred is much harder. There are a range of pay data collection tools that could help strengthen equal pay enforcement.

The EEO-1 pay data collection tool represents important progress to strengthen enforcement

The 2016 revision of the EEO-1 form to incorporate the collection of pay data was an important step forward, and that progress is critical to uphold. Opponents have sought to disparage the revisions to the form and the need for pay data overall, but these critics offer no concrete alternatives to promote equal pay or improve enforcement. Instead, their priority seems to focus on protecting employers and the status quo rather than ensuring equal pay and improving outcomes for workers. When analyzing the EEO-1 Component 2 pay data, using measures such as utility and burden, the value and potential uses of the data become clearer.

Utility

One important benefit of collecting pay data using the EEO-1 form is that the form has a proven track record. Long before the pay data component was added, the form for years played an important role in helping to identify disparities in which workers get certain jobs and which workers do not.14 The EEO-1 form was first put into practice back in 1966, during the EEOC’s earliest years, as a tool used to pinpoint discriminatory practices in different industries or jobs. The demographic information gathered through the form was vital to surfacing patterns of discriminatory conduct and helped provide the basis for hearings to explore the scope of the problem. Today, the Component 1 demographic data collected through the form continue to be used by enforcement officials to examine workforce composition and to shed light on workplace practices. There is no reason why pay data collected through the EEO-1 form cannot have similar utility.

Collecting pay data is also particularly important to gain a more in-depth understanding of women’s diverse experiences, especially women of color, who experience the largest pay gaps. Too often, the unique challenges facing women of color get lost in the broader equal pay conversation, because there is a lack of information beyond the basic data documenting the wage gap. While employers are quick to tout their efforts to address equal pay broadly, there is little focus on digging deeper to understand the pay differences of women of color. More analysis is sorely needed to gain a better understanding of why these gaps persist, and the pay data collected through the EEO-1 form could be an important tool. Component 1 data have previously been used, for example, to document where women of color are employed across the private sector.15 Such information could help chart progress but also show where work is still needed. Furthermore, EEO-1 pay data could offer a more nuanced understanding of the experiences of women of color at work and help to identify which types of corrective enforcement strategies could be most effective or useful.

Opponents of pay data collection have argued that collecting pay data is not useful. The EEOC’s September 2019 notice, without supportive analysis, dismisses the EEO-1’s pay data as having “unproven utility” that is outweighed by the burden. But this matter-of-fact assertion delivered before the data have even been analyzed suggests a predetermined conclusion not based on facts. Indeed, a growing body of evidence suggests that pay data reporting can have positive outcomes. A study by Danish researchers found that pay data disclosure was linked to a shrinkage in the gender pay gap.16 Furthermore, greater transparency may help ensure a higher level of discipline and precision as employers evaluate their compensation practices. It also puts employers on notice about expectations of enforcement officials and may incentivize more regular analysis of compensation.

Opponents also have argued that the EEO-1 form pay data would not be useful, because the data would be collected in 10 broad occupational categories rather than by individual occupation. This argument is unpersuasive. As already noted, the Component 1 workforce demographic data have been collected using the same EEO-1 categories for years, and they have proven to be a useful source of information for enforcement purposes. A showing of pay differences within occupational categories is one of the key threshold facts needed at the outset of an investigation to show that there is merit in looking further to determine whether a pay difference actually violates the law.17 The simple fact that a pay difference exists is not per se proof of discrimination, but information about pay differences is essential to laying the evidentiary foundation that is ultimately needed to prove that discrimination has taken place.

Burden

One of the most frequent issues raised about EEO-1 pay data collection involves questions about burden. Most often, these questions focus on potential burdens on employers when collecting the data. But any questions about burden should begin with some historical context to have a better understanding of the 2016 EEO-1 revisions that added the pay data component to the form in the first place. Indeed, these changes did not occur in a vacuum but rather were the culmination of efforts extending back more than two decades. These efforts aimed to identify a mechanism for collecting pay data that would minimize burdens on employers while providing enforcement officials with much-needed information. As officials explored options over the years—including different proposals published for notice and comment, multiple public hearings, and expert studies—one theme that consistently emerged was the need to balance multiple interests in order to devise a fair, reasonable, and efficient data collection method. Largely in response to these objectives and employer concerns, the EEO-1 form was identified as a potential vehicle for collecting pay data precisely because it was responsive to many of the concerns raised by employers, while satisfying the needs of enforcement. The form had been proven to be an effective enforcement tool; at the same time, because it was already in use, the form was familiar to employers and might require fewer systemic modifications. Understanding this context is important, because any new data collection regime will always require some balancing of competing interests. There are certainly more robust strategies that could be pursued to gather pay data from employers, such as requiring employers to submit all of their pay data on an annual basis. But the decision to utilize the EEO-1 form was an effort to strike a balance: Collect data that could expand the enforcement tools available, but do so using a familiar document that could help ease the transition to a new reporting structure.18 

Unfortunately, the Trump administration has sought to put forward an elaborate analysis to argue that the EEO-1 pay data collection is overly burdensome. The EEOC’s September notice states that the EEOC reexamined the previous burden estimates and concluded that it underestimated the actual burden associated with filing the EEO-1 form.19 The notice asserts that the burden now should be calculated differently, based on the number of forms filed by each individual filer rather than calculating an overall burden estimate for each individual filer or employer. Thus, employers who file more than one EEO-1 form—such as large, multiestablishment employers who file an EEO-1 form for each establishment—would be able to argue that the burden they incur in completing the form should be multiplied by the number of forms filed. Using this new assessment, the notice then argues that the number of burden-hours and total burden-hour costs for filing the form are far higher than previously understood. But there is little explanation of the underlying assumptions or how the administration determined the hour estimates. There seems to be an assumption that EEO-1 report filings will take longer for employers who file multiple forms, but this assumption may not be accurate in practice. Many multiestablishment employers are larger and often have greater access to advanced technology that could streamline reporting and reduce, not increase, burden. There also seems to be an assumption that the relevant burden costs would rise rather than fall over time, but that seems unlikely given that the costs would typically be expected to decrease as filers develop a routine.

Moreover, the burden analysis put forward in the notice inexplicably ignores the best evidence available to analyze any burden associated with the form: the real-world experience of employers who have been required to file Component 1 of the EEO-1 form for more than five decades. The notice does not explain how the burden for filing a form that has been in place for decades could suddenly increase, nor does it point to any evidence that indicates that the burden of filing Component 1 data has become more onerous than history has previously demonstrated. Moreover, the fact that many filers already had to submit their 2017 and 2018 forms with the pay data included by September 30, 2019, means that as a practical matter, the impact of any initial burden has already occurred and would be expected to ease with subsequent filings.

Furthermore, the notice ignores other factors that could significantly reduce the burden associated with filing components 1 and 2. Advances in technology and the availability of automated systems to upload and file the EEO-1 form may help to reduce the time and cost of filing. Although it is difficult to calculate with certainty, there are already myriad companies that now offer services to assist with filing EEO-1 forms.20 Potential improvements on the horizon also may include updated software or other tools to facilitate future reporting. Discontinuing the pay data component when such support is available and likely to expand is shortsighted and premature.

While the September 2019 notice focuses on the potential burdens of the EEO-1 data collection requirement on employers, there is virtually no mention of potential burdens on employees. However, it is equally important to consider how the failure to improve equal pay enforcement places ongoing burdens on workers, especially women. The persistent inequality in wages disproportionately affects female workers across all races and ethnicities, resulting in fewer resources for families trying to make ends meet. A Center for American Progress analysis of quarterly median, usual weekly earnings for women and men working full time in 2017 and 2018 found that, compared with a white man working full time during that two-year period, a Black woman earned almost $35,000 less; a Latina earned almost $40,000 less; a white woman earned nearly $19,000 less; and an Asian woman earned nearly $7,000 less.21 These numbers help illustrate some of the financial burden of the pay gap on women and their families. Over that same two-year period, women working full time earned more than $1 trillion less total than men working full time because of the stubborn persistence of the pay gap.22

These disparities have consequences for all families but particularly households with female breadwinners, which rely on women’s earnings to make ends meet. More than two-thirds of mothers are breadwinners in their families, and the numbers are even higher for some women of color.23 For example, more than 80 percent of Black mothers are either co-, sole, or primary breadwinners for their families. While many factors, such as differences in education and hours worked, cause the pay gap, it is essential for the EEOC to make use of every tool at its disposal to address any portion of the gap. Doing so would provide women and their families with much-needed resources to help with mortgage payments, student loans, child care costs, prescription costs, household bills, car repairs, groceries, emergency expenses, and more.

Rather than abandon the Component 2 pay data collection, an alternative and more useful response would be to take proactive steps to support its implementation. Providing extensive technical assistance; exploring public-private partnerships or research efforts to develop new software to help facilitate the collection of the relevant data; creating webinars with leading experts on data collection; and implementing other measures could build on and strengthen data collection efforts. The EEOC should not retreat from deploying a wide range of complementary strategies to support vigorous equal pay enforcement, including the utilization of pay data.

Voluntary action versus vigorous enforcement

Although eliminating pay discrimination requires vigorous enforcement of equal pay laws, many employers argue that their internal compensation analyses are far more sophisticated and superior to any pay data that would be collected through the EEO-1 form. Thus, they argue, the EEO-1 Component 2 pay data would be of little value and should not be utilized. But this argument ignores the basic purpose of such pay data collection: to provide insight into compensation practices and to help identify areas for enforcement officials where more scrutiny may be needed or useful. The tool is not intended to give a precise look into a particular workplace, but rather to surface trends and broad areas of interest or concern. Furthermore, pay data can be used to help expand and refine enforcement efforts, strengthen investigations, and better target resources. Voluntary efforts, regardless of their scope or detail, are not a substitute for robust enforcement of the law, nor should such efforts be used as a justification for weakening the investigatory tools available to enforcement officials.

Future reforms are also needed

Rather than pull back on pay data collection, the Trump administration should examine complementary efforts that could help promote greater pay transparency. Some employers are already subject to some form of reporting about pay. The Securities and Exchange Commission requires many publicly held large companies to report on the pay ratio between the CEO’s annual earnings and the median annual compensation for all employees.24 Employers in the United Kingdom, which include U.S.-based multinational companies operating in the United Kingdom, are now required to publicly report on their gender wage gap on an annual basis.25 These efforts could be extended to cover more U.S. employers and refined to gain a more nuanced understanding of the gap experienced by diverse groups of women. Other reforms could include requiring employers to post salary ranges in job announcements and undertaking equity assessments to quantify the burden for not addressing equal pay.

Conclusion

Pay data collection is critical to robust enforcement of anti-discrimination laws and securing equal pay. The 2016 EEO-1 revisions to collect Component 2 pay data represent long-overdue progress, informed by years of comment and review. Abandoning that progress is a step backward that women, men, and their families can ill afford.

Jocelyn Frye is a senior fellow with the Women’s Initiative at the Center for American Progress. 

Endnotes

  1. See, for example, Thomson Reuters Foundation and The Rockefeller Foundation, “The 5 key issues facing women working in the G20,” available at http://www.womenatworkpoll.com/ (last accessed February 2020); AFL-CIO, “Our Voices: A Snapshot of Working Women” (Washington: 2016), available at https://aflcio.org/sites/default/files/2017-03/1662_WWSurvelReport2_webready.pdf; American Association of University Women, “New AAUW Poll Shows Big Majorities of Americans Fault Structural Biases for the Pay Gap,” March 28, 2019, available at https://www.aauw.org/article/aauw-poll-for-the-pay-gap/.
  2. National Women’s Law Center, “The Wage Gap: The Who, How, Why, and What To Do” (Washington: 2019), available at https://nwlc.org/resources/the-wage-gap-the-who-how-why-and-what-to-do/.
  3. Jasmine Tucker, “Equal Pay for Asian American and Pacific Islander Women” (Washington: National Women’s Law Center, 2020), available at https://nwlc.org/resources/equal-pay-for-asian-pacific-islander-women/.
  4. Kristen Schilt and Matthew Wiswall, “Before and After: Gender Transitions, Human Capital, and Workplace Experiences,” The B.E. Journal of Economic Analysis & Policy 8 (1) (2008): 1–39, available at https://www.ilga-europe.org/sites/default/files/before_and_after_-_gender_transitions_human_capital_and_workplace.pdf; Ann Garcia and Patrick Oakford, “Unequal Pay Day for Immigrant Women,” Center for American Progress, April 9, 2013, available at https://www.americanprogress.org/issues/economy/news/2013/04/09/59471/unequal-pay-day-for-immigrant-women/.
  5. Francine D. Blau and Lawrence M. Kahn, “The Gender Pay Gap: Extent, Trends, and Explanations” (Cambridge, MA: National Bureau of Economic Research, 2016), available at https://www.nber.org/papers/w21913.
  6. Ibid.; Mark Paul and others “Returns in the labor market: A nuanced view of penalties at the intersection of race and gender” (Washington: Washington Center for Equitable Growth, 2018), available at https://equitablegrowth.org/wp-content/uploads/2018/07/080718-WP-intersectionality-labor-market.pdf.
  7. U.S. Department of Labor Office of Federal Contract Compliance Programs, “Government Contractors, Affirmative Action Requirements; Final Rule,” Federal Register 165 (219) (2000): 68022­–68047, available at https://www.govinfo.gov/content/pkg/FR-2000-11-13/pdf/00-28693.pdf; U.S. Department of Labor Office of Federal Contract Compliance Programs, “Affirmative Action and Nondiscrimination Obligations of Contractors and Subcontractors; Equal Opportunity Survey,” Federal Register 71 (174) (2006): 53032–53042, available at https://www.govinfo.gov/content/pkg/FR-2006-09-08/pdf/E6-14922.pdf.
  8. These actions included: prohibiting federal contractors from retaliating against workers who discuss pay in the workplace, requiring a higher minimum wage for workers on federal contracts, adopting minimum wage protections for home care workers, and requiring federal contractors to provide greater transparency about pay. See Obama White House, “Executive Order—Fair and Safe Workplaces,” Press release, July 31, 2014, available at https://obamawhitehouse.archives.gov/the-press-office/2014/07/31/executive-order-fair-pay-and-safe-workplaces.
  9. Equal Employment Opportunity Commission, “Agency Information Collection Activities; Notice of Submission for OMB Review, Final Comment Request: Revision of the Employer Information Report (EEO–1),” Federal Register 81 (135) (2016): 45479–45497, available at https://www.govinfo.gov/content/pkg/FR-2016-07-14/pdf/2016-16692.pdf; Equal Employment Opportunity Commission, “EEOC to Collect Summary Pay Data,” Press release, September 29, 2016, available at https://www.eeoc.gov/eeoc/newsroom/release/9-29-16.cfm.
  10. U.S. Department of Labor Office of the Secretary, “Guidance for Executive Order 13673, ‘Fair Pay and Safe Workplaces’,” Federal Register 82 (213) (2017): 51358, available at https://www.federalregister.gov/documents/2017/11/06/2017-23588/guidance-for-executive-order-13673-fair-pay-and-safe-workplaces.
  11. Letter from Neomi Rao to Victoria Lipnic, “EEO-1 Form; Review and Stay,” August 29, 2017, available at https://www.reginfo.gov/public/jsp/Utilities/Review_and_Stay_Memo_for_EEOC.pdf.
  12. National Women’s Law Center v. Office of Management and Budget, Civil Action No. 17-cv-2458 (TSC) (March 4, 2019), available at https://nwlc-ciw49tixgw5lbab.stackpathdns.com/wp-content/uploads/2019/03/EEO-1-Opinion.pdf.
  13. Equal Employment Opportunity Commission, “Agency Information Collection Activities: Existing Collection, Federal Register 84 (177) (2019): 48138–48142, available at https://www.federalregister.gov/documents/2019/09/12/2019-19767/agency-information-collection-activities-existing-collection.
  14. Equal Employment Opportunity Commission, “Early Enforcement Efforts,” available at http://www.eeoc.gov/eeoc/history/35th/1965-71/early_enforcement.html (last accessed February 2020).
  15. Equal Employment Opportunity Commission, “Women of Color: Their Employment in the Private Sector” (Washington: 2003), available at http://www.eeoc.gov/eeoc/statistics/reports/womenofcolor/womenofcolor.pdf.
  16. Morten Bennedsen and others, “Do Firms Respond to Gender Pay Gap Transparency?” (Cambridge, MA: National Bureau of Economic Research, 2019), available at https://www.nber.org/papers/w25435.
  17. See generally, Equal Employment Opportunity Commission, “EEOC Compliance Manual, Section 10: Compensation Discrimination” (Washington: 2000), available at http://www.eeoc.gov/policy/docs/compensation.html.
  18. Equal Employment Opportunity Commission, “Agency Information Collection Activities; Notice of Submission for OMB Review, Final Comment Request: Revision of the Employer Information Report (EEO–1).”
  19. Equal Employment Opportunity Commission, “Agency Information Collection Activities: Existing Collection.”
  20. Mark Dwyer, “Comment on the Equal Employment Opportunity Commission ‘Agency Information Collection Activities: Existing Collection’” (Washington: Equal Employment Opportunity Commission, 2019), available at https://www.regulations.gov/document?D=EEOC-2019-0003-7048.
  21. Findings based on CAP analysis of data from the U.S. Bureau of Labor Statistics (BLS). The individual earnings gaps reported compare quarterly median usual weekly earnings (Table 2) for full-time wage and salary workers by race and sex from all eight fiscal quarters of 2017 and 2018. See U.S. Bureau of Labor Statistics, “Usual Weekly Earnings of Wage and Salary Workers Archived News Releases,” available at https://www.bls.gov/bls/news-release/wkyeng.htm (last accessed February 2020). Owing to limitations of archived BLS data, the data used for earnings of women overall and men overall for all eight quarters were seasonally adjusted quarterly median usual weekly earnings as reported in Table 1 of U.S. Bureau of Labor Statistics, “Usual Weekly Earnings of Wage and Salary Workers, Fourth Quarter 2018,” Press release, January 17, 2019, available at https://www.bls.gov/news.release/archives/wkyeng_01172019.pdf. Women whose ethnicity is identified as Hispanic or Latino may be of any race.
  22. Ibid. The cumulative earnings gap reported compares quarterly median usual weekly earnings with monthly labor force totals for employed, full-time working women overall and by race as well as for men overall and white men for January 2017 through December 2018. See BLS Beta Labs, “BLS Data Viewer, (Unadj) Employed, Usually Work Full Time, Women,” available at https://beta.bls.gov/dataViewer/view/timeseries/LNU02500002 (last accessed February 2020); BLS Beta Labs, “BLS Data Viewer, (Unadj) Employed, Usually Work Full Time, Women,” available at https://beta.bls.gov/dataViewer/view/timeseries/LNU02500001. Women whose ethnicity is identified as Hispanic or Latino may be of any race.
  23. Sarah Jane Glynn, “Breadwinning Mothers Continue To Be the U.S. Norm” (Washington: Center for American Progress, 2019), available at https://www.americanprogress.org/issues/women/reports/2019/05/10/469739/breadwinning-mothers-continue-u-s-norm/.
  24. Dodd-Frank Wall Street Reform and Consumer Protection Act of 2010, Public Law 111-203, 111th Cong., 2nd sess. (July 21, 2010), available at https://www.congress.gov/111/plaws/publ203/PLAW-111publ203.pdf.
  25. U.K. Advisory, Conciliation and Arbitration Service and Government Equalities Office, “Gender pay gap reporting: overview,” February 22, 2017, available at https://www.gov.uk/guidance/gender-pay-gap-reporting-overview.