Written Testimony of Lynn A. Clements Director, Regulatory Affairs Berkshire Associates Inc. Columbia, MD

Meeting of November 20, 2019 - EEOC Convenes Public Hearing on the Proposed Revision of the Employer Information Report (EEO-1)

Introduction

Chair Dhillon and distinguished Commissioners, thank you for the opportunity to speak with you today about my firm's experiences helping employers prepare and file Component 1 and Component 2 of the Standard Form 100, or EEO-1 Report. My name is Lynn Clements, and I currently serve as the Director of Regulatory Affairs at Berkshire Associates Inc.

My firm, Berkshire Associates, is a human resources consulting and technology firm specializing in affirmative action compliance, applicant data management, and pay equity. We also help employers prepare and file EEO-1 Reports. Berkshire's clients vary in size from small employers who file one EEO-1 Report to nation-wide employers with thousands of employees covered by hundreds of EEO-1 Reports. During the 2018 reporting cycle, Berkshire prepared Component 1 reports for over 225 clients and Component 2 reports for approximately 200 clients, covering more than 930,000 employees.[1] Berkshire also regularly assists employers in conducting self-assessments of their compensation systems. In 2019, we completed over 900 salary equity analyses covering almost 400,000 employees.

Berkshire has submitted detailed written comments on the Commission's September 2019 PRA Notice.[2] In my testimony today, I want to focus on my firm's actual experience in helping clients prepare and file Component 1 and Component 2 EEO-1 Reports this year.

EEO-1 Component 1 Report

The EEOC's September 2019 Paperwork Reduction Act (PRA) Notice seeks approval to continue collection of Component 1 of the EEO-1 Report.[3] We agree that the current benefits of collecting Component 1 data outweigh the current burden of collection. As the EEOC noted in its September 2019 PRA Notice, "collection of Component 1 data is a long-held practice that has occurred for over 50 years and has already proven its utility . . ." Component 1 data (race/ethnicity and gender for all employees by EEO category) is also data that is regularly used by employers for other equal employment opportunity purposes, such as in the preparation of annual affirmative action reports or as part of an employer's diversity and inclusion efforts. 

We also applaud the EEOC for re-evaluating the burden of filing Component 1. Without additional detail about how the EEOC calculated its revised burden estimate, it is difficult to assess the reliability of these estimates. However, EEOC's revised burden estimate for Component 1 may still be too low, when accounting for the extensive collection, verification and validation process that employers undertake to be able to file and certify their Component 1 Reports. In particular, we are not aware of any employer who can collect, verify, validate and report Component 1 data in 45 minutes, even if the employer is filing only one Component 1 report.[4] Despite what some have suggested, Component 1 reports are not generally filed by merely clicking a report button in a Human Resource Information System (HRIS).

EEO-1 Component 2 Report

Component 2 collects two data points - Box 1 of an employee's W-2 wage statement and the employee's total hours worked in that year. During the Component 2 approval process, employers repeatedly stated that the data being collected and the way that data was grouped (within broad pay bands in broad occupational categories) was not useful for identifying or remedying sex- or race-based pay disparities. The experience of preparing the Component 2 reports has only served to underscore these concerns about the utility -- and indeed basic reliability -- of the collected data.

The problems with relying on these two data points have been well-documented in the record before the Commission, but they bear repeating. First, the original intent of the EEO-1 Report was to provide a demographic snapshot of an employer's workforce during a given payroll period. Component 2, however, requires employers to gather and synthesize what happened to employees over a 12-month period. This creates a level of complexity to gathering the data that was severely underestimated. For example, employees may experience a personnel movement (promotion, transfer, hire) within the reporting year. As a result, these employees may transition from one EEO Job Category to another, move between establishments, be reclassified from hourly to exempt, or be added to the employer's payroll through a merger or acquisition. The assumptions/manipulations required due to the Component 2 full year reporting format increase the burden on employers while simultaneously decreasing the utility of the reported data.

Second, the stated goal of the Component 2 collection was to allow EEOC and OFCCP to use the pay data to identify employers with existing pay disparities, as well as encourage employers to self-monitor their pay practices and identify pay inequities. However, using annual W-2 earnings does not allow for this. An employee's W-2 earnings over a 12-month period can be impacted by factors that are totally out of the employer's control, such as elected benefits and deductions or excused leaves of absence. That the W-2 earnings can be so heavily impacted by employee choice completely undermines the stated purpose of identifying race- or gender-based pay disparities that occur as a result of employer pay decisions.

The collection of the hours worked data was also proven to be flawed during the preparation of Component 2 reports. Timekeeping systems can be very complex and can vary greatly in how they are used even within the same company. This resulted in many employers making assumptions on how to present the hours worked data given the information available to them. Additionally, the W-2 earnings included wages for time employees didn't work (paid time off, vacation, jury duty, etc.), while the hours worked figure did not include this time. This inconsistency was one of many reasons that employers Berkshire worked with had serious concerns about the accuracy of the data being submitted, even though it was being reported 'correctly' based on EEOC's instructions and FAQs.

As further evidence of the lack of utility of using these data points for this report, Berkshire has never once had an employer ask us to use W-2 wage information, either alone or in combination with hours worked, when conducting a salary equity analysis. Employers are interested in identifying pay disparities that might be attributed to race- or gender-based discrimination or bias, and W-2 earnings are not going to show them this, for the reasons stated before. The experience of using the W-2 earnings and hours worked to complete Component 2 has not changed this. In order to create the most useful salary equity analysis an employer wants to use information that reflects their compensation decisions, as well as their unique compensation systems and practices.

In fact, employers reported to Berkshire that in some cases the substantial amount of time and resources that had to be directed to the Component 2 preparation took away from other efforts to ensure pay equity. This is the exact opposite effect of the stated goal of the Component 2 report, which was to encourage employers to self-monitor their pay practices and address pay inequities. These are employers who care about the principles of pay equity but the significant burden of time and resources required to complete the Component 2 report meant that these employers could not use those time and resources for more meaningful analyses.

While we believe the EEOC's revised burden estimates in its September 2019 PRA Notice are more in line with employer's actual burden, we continue to believe, as EEOC has recognized, that the burden of collecting Component 2 data far outweighs any utility of the data. Our written testimony provides detailed information about our clients' filing experiences this year. In the interest of time, I will share two examples with you here.

A client with approximately 5000 employees filed its Component 2 data reports using the electronic filing system (i.e. manual keying, not a batch upload). This client estimated that it took 160-200 hours to gather, verify and file the Company's Component 2 data reports. An employer with about 7000 employees in the U.S. estimates that it took approximately 65 hours, to gather and report the required payroll and hours worked information, even though much of this information was in the same system. The employer spent at least one full day trying to "verify" the data gathered, although the employer reported this exercise was nearly impossible given the size of the report files, and the aggregated nature of those files. Additional time was spent preparing and filing the Component 2 data.

As these examples, along with our written testimony, demonstrate, the burden associated with collecting, verifying, and filing the Component 2 data is substantial. During this process it has been argued that this process will be less burdensome in future years as the data has already been collected, and only 'modest reprogramming' by employers would be required. Many of the issues that caused the significant burdens to employers - such as the large number of employee movements between roles/job categories, changes in exempt/non-exempt status, the impact of mergers and acquisitions or the use of multiple different systems to record the information being collected - will remain in future filing years. Despite the significant amount of time and resources employers invested in this exercise this year, employers will not be able to simply "push a button next year" and generate the required Component 2 data.

Conclusion

In closing, I would like to commend the EEOC on the important role it plays in protecting the civil rights of individuals in the workplace.  Berkshire and its clients share that goal. The current National discussion on pay equity is important and one that many of us relate to on a personal level. Collecting pay data in the manner and format specified under Component 2 of the EEO-1 Report has not, however, proven useful to our shared goals.

Helping employers prepare and file Component 2 reports over the past several months has amplified the significant concerns that were raised about the data collection when it was proposed. Reporting W-2 income and total hours worked by broad pay bands in broad EEO-1 job categories is a paperwork exercise, and a burdensome one at that. No employer with whom we worked found the information useful - or even usable - to its efforts to ensure nondiscrimination in pay. The burden on employers of collecting and reporting the information required is far greater than what the agency had originally estimated, in large part because Component 2 reporting deviates from the snapshot approach of Component 1 reporting. This significantly, and unnecessarily, complicated the collection and reporting of the requested pay data. Given the lack of utility of the Component 2 data, and the significant burden of collection, we agree with the EEOC's decision to not renew the PRA approval for Component 2 of the EEO-1 Report.

Thank you. I appreciate the opportunity to share my perspective with you today and would be happy to answer any questions.