Please submit a public comment opposing this rule change by August 24. I have my public comment available at the end of this post along with a link to the proposed rule change and the place to submit a public comment.
The Equal Employment Opportunity Commission is charged with enforcing your Title VII Civil Rights against discrimination in the workplace. A key tool in their toolkit has been mandating that large companies (over 100 employees) and those with federal contracts (technically 50 employees and over $50k in contracts) must keep records on demographic data regarding all personnel decisions and turn those in to the EEOC annually. The EEOC can use large datasets to determine if company’s selection systems are resulting in adverse impact (hiring/firing/promoting/training employees of one group within a protected class at a rate that is less than 80% the rate of another group within that protected class). If this is the case, the organization has an affirmative duty to show that their personnel decisions are necessary for organizational functioning and that there exists no equally good tool for making these decisions that wouldn’t cause this adverse impact.
Under the current administration, the EEOC is proposing an end to the requirement that companies collect, keep, and report these data. The proposed rule change says that merely collecting these data encourages companies to engage in illegal discriminatory behavior against majority groups (i.e., white people, men). This claim is not rooted in fact, and the data broadly bare out that when adverse impact occurs it harms minoritized groups. This is especially going to be the case at large companies with HR specialists and legal teams that understand how current EEOC rules work (e.g., you cannot have a quota for certain groups, you cannot give bonus points in applications for certain groups, and you cannot treat groups differently).
This proposed rule change suggests that data can be revealed upon the rise of a complaint against an employer, but if the data are not required to be collected there is no guarantee that these data would be available. This makes it nearly impossible for individuals who are the victims of personnel systems that result in differential outcomes based on group membership to enforce their rights if they suspect they were on the receiving end of discrimination. It also makes it more difficult for companies to prove they did not cause harm because although people may feel they were treated differently, the data may suggest otherwise.
The proposed rule change also says that the EEOC cannot continue to pay consultant fees to collect and maintain these data, as an already small-budget agency. To this, I say, fully fund the EEOC so it can actually do its work, rather than continuing to kneecap it.
There is so much more to say about this proposed rule change. I think we may do a podcast episode on it if there’s enough interest. Regardless, I think we have an imperative to speak out against this change. I have copied my public comment below and encourage you to write your own (but feel free to pull points from mine or tweak mine as you see fit). I have also included a link to the proposed rule change below.
Read the proposed rule change here: https://www.regulations.gov/document/EEOC-2026-0034-0001
Submit your comment here: https://www.regulations.gov/commenton/EEOC-2026-0034-0001
My public comment:
I hold a doctoral degree in Industrial-Organizational Psychology and have been conducting research in this field for 13 years. I am writing in my personal capacity to strongly oppose the proposed rule change regarding the requirement that large employers collect, keep, and report demographic data regarding personnel decisions.
The Society for Industrial-Organizational Psychology (SIOP) released a public statement (May, 2025) in response to Presidential Executive Order EO14281 discussing in detail the nature of protected classes in the context of selection. As SIOP suggests, “disparate impact, where a sub-group of people is shown to be adversely affected by an apparently neutral employment process, is not illegal if the process is shown to be job-related”. Disparate treatment, on the other hand, in which individuals are treated differently based on their perceived or actual membership of a group within a protected class is already illegal, as both SIOP and much of the background included in the proposed rule change note. This includes the use of quotas, preferential hiring, and other policies that show favoritism to one group over another regardless of that group’s majority status. However, disparate impact cannot be shown without the organization-level data that is currently mandated to be collected, kept, and reported by the EEOC. Removing the requirement to collect, keep, and report these data will not only hinder the EEOC’s ability to proactively enforce Title VII civil rights for all individuals, regardless of majority status, but will also make it less likely that individuals (regardless of majority status) can successful pursue individual claims of disparate impact. Moreover, organizations who stop collecting these data may be more susceptible to suits regarding disparate treatment or impact, failing to have a clear record that exonerates them.
One of the arguments raised by the proposed rule change is that the mere collection and reporting of demographic data encourages organizations to engage in discriminatory behavior. Again, disparate treatment is already illegal and both the EEOC and private citizens have pathways to pursue enforcement of Title VII in this instance. Eliminating the reporting requirement because it may (though there is no systematic data suggesting it does) result in disparate treatment of majority groups is equivalent to throwing the baby out with the bathwater. Moreover, in its cost analysis, the proposed rule change notes that most organizations rely upon an HR specialist to manage data collection and reporting. These HR specialists will certainly be trained in appropriate use of demographic data, including the illegal nature of using demographic information in personnel decisions. For example, the Society for Human Resource Management which provides professional certification for HR specialists includes knowledge of appropriate use of EEOC reporting data in their certification exam (SHRM Body of Applied Skills and Knowledge, 2026). Thus, it is unlikely that this concern mentioned in the proposed rule change is occurring at a meaningful level, particularly in employers large enough to be impacted by the EEOC requirement.
Another argument raised by the proposed rule change is that collection of demographic data proactively rather than in response to filed complaints is not appropriately narrowly tailored or necessary. The proposed rule change not only eliminates the reporting requirement, but also the record keeping requirement. Thus, in the instance of individuals or the EEOC pursuing complaints against an organization, the necessary data to exonerate or implicate the organization’s selection systems may not be available. This is likely to make legal proceedings murkier, relying less on objective statistical analyses of disparate impact but more on subjective reports of fairness and disparate treatment.
Although the proposed rule change notes “virtually every employment practice has at least some adverse impact on some protected group”, recent meta-analytic evidence clearly suggests that there are known selection systems that have higher validity (are better predictors of job performance) and lower levels of adverse impact than others (Sackett et al., 2022). Incentive to use these selection systems wanes if the EEOC does not proactively monitor hiring demographic data. This may, as SIOP puts it, allow “job-irrelevant factors to inhibit opportunities or influence employment decisions, [thus] harm[ing] the effectiveness of both the organization and individual workers” (SIOP, May 2026).
The proposed rule change also notes the dynamic nature of demographic categories as a reason to not collect these data. First, a narrowly-tailored change would be to expand or alter the nature of demographic data collected. That said, many demographic subgroups may be too small to have meaningful statistical examinations (Aguinis et al., 2010). There exist guidelines for researchers on collecting demographic data appropriately (e.g., Call et al., 2023) that may be adopted by the EEOC if this is truly a concern.
Lastly, concerns around disparate impact rather than disparate treatment are increasingly poignant given the rise of artificial intelligence in selection systems (Tippins et al., 2021). When the decision architecture is unknown, as is the case in many AI systems, the only evidence of discrimination may come from evidence of disparate impact.
For all these reasons, I strongly oppose implementation of the proposed rule change. I echo Justice Ginsburg’s dissent in Shelby County v. Holder, that this would be “like throwing away your umbrella in a rainstorm because you are not getting wet.” However, should a version of the proposed rule change move forward, I suggest that at a minimum, organizations continue to be required to collect and keep demographic data while eliminating the reporting requirement.
Aguinis, H., Culpepper, S. A., & Pierce, C. A. (2010). Revival of test bias research in preemployment testing. Journal of Applied Psychology, 95(4), 648 –680. https://doi.org/10.1037/a0018714
Call, C. C., Eckstrand, K. L., Kasparek, S. W., Boness, C. L., Blatt, L., Jamal-Orozco, N., ... & Scholars for Elevating Equity and Diversity (SEED). (2023). An ethics and social-justice approach to collecting and using demographic data for psychological researchers. Perspectives on Psychological Science, 18(5), 979-995.
Sackett, P. R., Zhang, C., Berry, C. M., & Lievens, F. (2022). Revisiting meta-analytic estimates of validity in personnel selection: Addressing systematic overcorrection for restriction of range. Journal of Applied Psychology, 107(11), 2040–2068. https://doi.org/10.1037/apl0000994.
SHRM (2026). SHRM Body of Applied Skills and Knowledge. https://www.shrm.org/certification/decide/Pages/body-of-applied-skills-and-knowledge.aspx
SIOP (May 2025). Title VII and Job-Relevant Employment Practices. https://www.siop.org/post/siop-issues-statement-on-title-vii-and-job-relevant-employment-practices/title-vii-and-job-relevant-employment-practices/
Tippins, N. T., Oswald, F. L., & McPhail, S. M. (2021). Scientific, legal, and ethical concerns about AI-based personnel selection tools: A call to action. Personnel Assessment and Decisions, 7(2), 1. https://doi.org/10.25035/pad.2021.02.001
No posts
Comments
Nothing yet. Say the first thing.
Sign in to join the conversation.