The NILG Annual Conference is an excellent barometer for what is happening in equal opportunity, affirmative action and Diversity Equity and Inclusion (DEI). At this year’s Conference, held in Phoenix, the hot (pun intended, it was 110 degrees out) topics included the recent Supreme Court decision concerning college admissions, compensation practice, and Artificial Intelligence (AI).
Acting OFCCP Director Michelle Hodge addressed the Conference. She stated that the Agency is strengthening its regulatory enforcement efforts by hiring more than 100 new employees, many to focus on analysis of federal contractors’ compensation practices. Hodge specified that “transparency” in OFFCP audit practices would “continue” through use of Pre-Determination Notices (“PDNs”, outlining Agency concerns about a review before a violation is found) and Early Resolution procedures. (This writer continues to militate for a robust exchange of information between federal contractors and the Agency during the earlier desk audit stage. Knowing what the Agency is thinking is the best way for all concerned to handle audits efficiently. Unfortunately, the Agency’s Final Rule on Pre-Enforcement Notice and Conciliation Procedures, released on the last day of the Conference, significantly trims the amount of detail the OFCCP will be providing contractors during audit.)
EEOC Commissioner Charlotte Burrows also spoke. Like Hodge, Burrows stressed the Biden administration’s Infrastructure Investment and Jobs Act and Chips Act are means for spurring economic development and ensuring that vulnerable communities have access to “good jobs.” For Affirmative Action practitioners, even those not in construction, this translates to carefully monitoring outreach, hire and promotion practices to ensure that proper records are maintained and that selection practices do not constitute barriers to access in the workplace.
Burrows spoke of Artificial Intelligence (AI) as a new frontier for civil rights, one in which “even its creators say we need a pause.” She stated that AI can “replicate and amplify lack of diversity.” The most telling government comment concerning AI was made by OFCCP District Director Sam Maiden during the OFCCP District Director’s panel discussion. He stated that the OFCCP intends to ask contractors for the data used to “train” AI systems (this is the data that AI systems review to identify the characteristics they correlate to successful choices).
Speakers at Conference workshops outlined best practices for using AI, including asking vendors how the systems work, retaining data, auditing results, and performing traditional “validation” under the Uniform Guidelines on Employee Selection Procedures (UGESP) when the recommendations the systems generate create adverse impact in selection. (EEOC’s recent review of AI under Title VII is a good place to start to review many of the issues that AI presents.)
Burrows also addressed Component 2 pay data reporting. She maintained that the 2022 National Academy of Sciences report on the collection “confirmed what we suspected” — that the data can be a useful tool for identifying pay disparities. (Many at the Conference argued to the contrary, holding that the data collected has very limited utility for government review or employer self-assessment.) Burrows said that the Commission will look for public stakeholder input because it is “important to get it right.” Practitioners should expect another chance to comment on the EEOC’s proposed pay data collection methodologies.
Both Hodge and Burrows addressed the recent Supreme Court decision on Harvard University and University of North Carolina (UNC) admissions practices. They stated the ruling does not affect “properly tailored” company Diversity, Inclusion, Equity and Accessibility (DEIA) and Affirmative Action programs. Conference speakers agree that the race-based admission practices at issue in the case do not carry over into employment, where with very limited exceptions race is not a legitimate factor in selection. Conference speakers made the point that this distinction is lost on some CEOs. Speakers voiced concern that because the Court questioned the accuracy of metrics determining goals and progress towards goals, affirmative action could be subject to future court challenges.
Many sessions focused on compensation review practices. Of note was a panel comprised of consultants to the contractor community and representative of the OFCCP and DOL’s Office of the Solicitor. The panel touched on contractors’ 41 CFR 60-12.17(b)(3) obligations. Consultants stressed that the mandated review can be done without resort to statistics, or by using relatively simple “cohort” review to identify patterns and trends in the pay of similarly situated employees. OFCCP stated that it is likely to regroup employees for analysis even in the early stages of its audit review. While consultants expressed concerns about comparing the pay of employees in different roles, the discussion was as much needed open exchange of views in an ongoing discussion on this important subject.
The NILG Annual Conference remains a unique opportunity for practitioners, consultants, and agency representatives to meet, exchange views and get to know one another. This year’s event was no exception by highlighting what practitioners need to know about using AI, maintaining compliant DEIA and affirmative action programs, and performing mandated pay equity reviews.
1Paul McGovern is a member of the National Industry Liaison Group (NILG) that hosts the Conference. Circa is a Conference sponsor.
2For example, if not used correctly AI can identify attendance at an ivy league college as a predictor of success on the job. (This could favor the hire of whites given that marginalized communities have traditionally been underrepresented at such institutions).
