On September 29, 2016, the Equal Employment Opportunity Commission (“EEOC”) introduced the revised Employer Information Report or “EEO-1,”[1] which for the first time contains a pay data collection component. With the change in presidential administration and significant opposition from employer groups, many expected it to disappear. Meanwhile, nothing has changed, and the deadline for employers to submit the new report is less than nine months away—March 31, 2018. Under the new rule, the EEOC also will collect information about W-2 earnings and hours worked, organized by EEO-1 category, sex, race, and ethnicity from employers with 100 or more employees, in addition to the demographic information previously required. In this Alert, we describe the EEOC’s revised rule, and identify key concerns and recommendations for employers.

I. Background

The EEOC’s new rule was an outgrowth of President Obama’s 2010 National Equal Pay Task Force. It followed an EEOC-commissioned pilot study from the National Academy of Sciences (“NAS”) to make recommendations on employer pay data collection, feedback from employer groups and other experts, as well as comments to its Notice of Proposed Rulemaking in 2016. According to the EEOC, the objective is to better position federal agencies to enforce pay discrimination laws, while respecting employer concerns about confidentiality and minimizing employers’ data collection burdens. The EEOC noted in its Notice of Proposed Rulemaking that until this rulemaking, pay discrimination had gone largely undetected due to lack of information about employee compensation and employer and establishment-specific data which could verify allegations of pay discrimination. The new tool was intended to remedy this, but employers and other stakeholders alike were and remain skeptical of its utility.

Many forecasted that the Trump administration would shift EEOC’s priorities and rollback EEOC’s pay data reporting in response to employer opposition. Thus far, however, the administration has taken no steps to change course on this.[2] Therefore, at this point employers should begin planning for the March 31, 2018 deadline.

II. Requirements for the EEO-1 Report under the Revised Rule

By March 31, 2018, employers with 100 or more employees are required to provide summary W-2 compensation and hours worked data for all employees by gender and race/ethnicity. The revised EEO-1 Report has two new elements: (1) summary pay data, and (2) aggregate hours worked data.

A. Summary Pay Data

For reporting summary pay data, employers must categorize all employees into one of ten well-established job categories (e.g., Officials and Managers, Professionals, Sales Workers) and report the number of employees whose compensation falls within each of the 12 defined pay ranges (or “pay bands”) defined by the Bureau of Labor Statistics for the Occupation Employment Statistics survey. The pay bands range from the lowest, covering employees paid $19,239 and below, to the highest, covering employees paid $208,000 or more. To identify the proper pay band, employers will rely on the pay reported for income tax purposes for that year in Box 1 of the employee’s W-2 form. Employers should note that this income reflects employee choice (e.g., employees’ elective participation in overtime) and supplemental pay, such as signing bonuses, which may unfairly distort the picture presented by the data. Employers should be prepared to explain any apparent discrepancies caused in this instance by non-discriminatory factors in Section F—“Remarks” of the EEO-1 Report Form.

For summary pay data, employers report the total number of full and part-time employees they had during a calendar year in each of the 12 pay bands listed for each EEO-1 job category. After tallying the total employees in each pay band by job category, the employer will be required to enter this tally in the appropriate columns of the EEO-1 Report based on sex and ethnicity/race of the employees.

B. Aggregate Hours Worked Data

For reporting aggregate hours worked, employers will tally and report the number of hours worked that year by all the employees accounted for in each pay band. For non-exempt employees, employers will consult and report the aggregate number of hours worked for all employees in each pay band. For exempt employees, the employer may either: (1) report 20 hours per week for each part-time employee and 40 hours per week for each full-time employee; or (2) report the actual number of hours worked by exempt employees, full or part-time.

C. Recommended Statistical Analysis

The NAS pilot study recommended several statistical techniques to test within job categories. The EEOC and OFCCP plan to develop a software tool that will allow investigators to conduct an initial analysis by looking at W-2 pay distribution within a single firm or establishment by comparing the establishment’s data to aggregate industry or metropolitan area data. We recommend using a distributional analysis called the “rank sum” test to determine whether disparities exist. The rank sum analysis tests the hypotheses that two independent samples (such as the pay of men and the pay of women) are from populations with the same pay distribution. The test combines both men and women in each of the 12 pay bands, ranks the employees from lowest pay to highest pay, then assigns a numerical value starting with one to the lowest paid rank. For example, the lowest paid employee is assigned a rank of one, the second lowest paid employee is assigned the rank of two, etc. For any employees who are tied, the average rank is used. If there are six women and four men in the lowest pay group, each of these employees would have a rank of 5.5, the average of 1-10. The numerical value assigned to all ranks of women are summed up and compared to the sum of the expected ranks of all women if assignment to the pay group were unrelated to gender. If the sum of the actual ranks is less than the sum of the expected ranks for women, this indicates women are both ranked, and paid, lower than men making the comparison “adverse” to women. If the difference between the actual and expected rank sums for women is greater than two standard deviation (.05 probability) then the difference is considered statistically significant (or not likely to have occurred by chance). As discussed below, there are material flaws in this analytic approach which will create both false positives (adverse results for a protected group) and false negatives (neutral or favorable results for a protected group).

III. Additional Changes

Previously, the EEO-1 Report required covered employers to identify the type of work activity at each of its locations based on the North American Industry Classification System (“NAICS”). The revised EEO-1 form requires that all reports, except the Type 2 Consolidated Report, provide a description of the major activity of the establishment. The EEOC specified that the major activity should be “sufficiently descriptive to identify the industry and product produced or service provided.” If the establishment engages in more than one activity, the major activity description should include the activity in which the greatest numbers of employees work. It is unclear whether or not using the appropriate NAICS code will suffice.

IV. Timing for Reporting

The new EEO-1 Report must be filed by March 31, 2018. Employment reporting data must include all full-time and part-time employees who were employed during the payroll period selected by the employer between October 1 and December 31, 2017 (“workforce snapshot period”). This workforce snapshot period has changed from the previous rule which had a reporting period between July 1 and September 30. The employer may choose any pay period during this three month snapshot period to count its full- and part-time employees. A later snapshot may reflect important late-year developments (such as promotions which could alter an employee’s job category or pay band), but other year-end obligations may make it administratively easier to choose an earlier window, especially in light of the March 31, 2018 filing deadline. Of course, the annual W-2 reports that will be required to identify in which of the 12 pay buckets each employee should be counted in are typically not finalized until late January, so the window for completing the new EEO-1 forms will be tight.

V. Confidentiality Concerns

As previously discussed in our February[3] and July 2016[4] alerts, employers may have legitimate concerns about confidentiality of the revised EEO-1 required compensation data. As reported, the compensation data may practically disclose employees’ pay levels where there are small numbers of employees in an EEO-1 job group and pay band. The EEOC is prohibited by Title VII from making “public in any manner whatever any information obtained by the Commission . . . prior to the institution of any [Title VII] proceeding . . . involving such information.” However, the EEOC can provide the OFCCP with the summary pay data (including hours worked) from federal contractors and subcontractors subject to Executive Order 11246. The OFCCP is restricted in disclosure as to non-contractors. With regard to federal contractors and subcontractors, the EEOC has stated that the OFCCP will hold this data confidential “to the maximum extent possible under the Freedom of Information Act (“FOIA”) and Trade Secrets Act.” If a FOIA request is received, the OFCCP will notify the contractors, providing them with an opportunity to object to the disclosure. If the contractor objects, and the OFCCP determines the objection is valid, the data will not be disclosed.[5]

VI. Compensation Data Reporting Concerns/Challenges

The revisions to the EEO-1 reporting requirements may also serve to complicate, rather than improve pay equity enforcement. Collection of pay data in the form of W-2 compensation data by EEO-1 job groups may lead to false indicators of pay disparities. The revised EEO-1 job categories fail to distinguish between different types of jobs (e.g., human resources professionals generally earn less than engineering professionals), and the type or level of work performed by the employee in each job category (professionals included both highly experienced individuals as well as individuals who just completed their degree). Job and type/level of work are usually key explanations for pay differences under a job category umbrella. Additionally, as described above, W-2 compensation comprised of bonuses and severance payments may unfairly distort the picture presented by the proposed EEO-1 compensation data. The EEOC’s requested compensation data set may identify false positives, where employers appear to have suspect pay distribution when in reality they have perfectly legitimate non-discriminatory explanations for the detected disparities.

Additionally, coordinating the collection of this compensation data may pose a challenge because HR and payroll systems tracking this information may not be compatible to report the data needed. Diligent employers should be inclined to begin collecting the required data now, to ensure that they are prepared to file by the March 31, 2018 filing date.

VII. Next Steps

We encourage employers to coordinate across their IT, HR, Compensation, and Legal departments to prepare for the EEO-1 compensation data reporting and perform a “test run” to evaluate the data results prior to the EEO-1 filing date. We recommend employers prepare a self-critical, privileged, proactive pay analysis using the “rank sum” test in order to determine the scope of any pay disparities that the analysis may reveal, and take steps to mitigate, remediate, or determine other methods to address any pay disparities before filing their EEO-1 Report.