Considerations for Employers Regarding the Collection of Demographic Data on Employees and Applicants

With the EEOC poised to discontinue EEO-1 reporting requirements (which currently require employers to collect and file reports regarding the race, sex, and ethnicity of employees sorted by various job categories), a recent executive order rescinding agency guidance regarding disparate impact discrimination, and the government’s questioning the soundness of the Uniform Guidelines on Employee Selection Procedures, employers reasonably may be asking, “If we are no longer required by the EEOC to collect demographic information about our workforce, are we permitted to continue to collect this information? If so, should we? And if we do, how can we best ensure that this information is maintained and used only for lawful purposes?”

Although some may perceive an opportunity to rid themselves of a burden that is costly and intrusive, the benefits of continuing to collect this information are worth considering. A number of considerations for employers pondering whether to continue to collect demographic information regarding the race, sex, and ethnicity of applicants and employees are set forth below.

The Uniform Guidelines on Employee Selection Procedures Remain in Place

At least for the time being, the answer to these questions is straightforward. While the future of the Uniform Guidelines on Employee Selection Procedures (UGESP)¹ may be uncertain, at least for now, they remain in place. At a minimum, UGESP permits employers to collect and regularly analyze data on the race, ethnicity, and sex of employees and applicants for employment. Moreover, there is an argument that UGESP actually makes the collection and analysis of this data mandatory.

UGESP was adopted in 1978 by the EEOC, U.S. Department of Labor, U.S. Department of Justice, and the Civil Service Commission (now the Office of Personnel Management). UGESP provides that employers subject to Title VII must collect information on the race, ethnicity, and sex of applicants and periodically analyze that data for indicators of potential adverse impact. UGESP further provides guidance as to how to either remedy or validate selection procedures that appear to have an adverse impact based on race, ethnicity, or sex.

Although UGESP includes language providing that its requirements are mandatory,² the courts and commentators have disagreed over whether UGESP is actually a rule that must be followed or merely hortatory guidance. 

What this means for employers that want to collect and analyze demographic data is that there remains clear legal authority to continue to do so even if EEO-1 reporting should be discontinued. For employers that want to stop the collection of this data, there is an argument that they may do so, but such employers should first determine how courts have treated this issue within the applicable jurisdictions. 

But what if UGESP is merely hortatory or if it should also be rescinded? In that event, should employers continue to collect and analyze applicant and employee demographic data? For the following reasons, the answer may be yes. 

Statistical Evidence Remains Relevant to Suits Under Title VII and 42 U.S.C. §1981 

First, the relevance of statistical evidence in private litigation is not subject to the executive orders or the EEOC’s decision whether or not to pursue a given case, and will continue to be an important element of the most sweeping claims of discrimination. The disparate impact theory of discrimination has been recognized by the U.S. Supreme Court, is codified in Title VII of the Civil Rights Act at 42 U.S.C. 2000e-2(k), and remains available as a method of proof in private litigation unless and until the Court revisits its prior holdings. Additionally, both Title VII and §1981 recognize a “pattern or practice” of discrimination in violation of those statutes, based on proof of “gross statistical disparities,” which may be accompanied by anecdotal evidence of discrimination. Accordingly, statistical evidence will continue to play a substantial role in the most dangerous discrimination cases. 

Is an Employer’s Defense to These Cases Likely to Be Stronger in the Absence of a Database of Demographic Data? 

At first blush it may seem foolish for an employer to continue gathering data that may lead to its demise—the phrase “hoist on one’s own petard” comes to mind—but that may be short-sighted. The absence of a database will not purge statistical considerations from the case. A plaintiff may be able to acquire the demographics of an employer’s workforce by means of a simple interrogatory seeking such information. If the scope of the interrogatory is reasonable, the employer will be obligated to provide the requested information, and if data are not readily available it would have to explain why the burden of obtaining it (presumably by querying its employees or else by observation) exceeds its likely probative value.

A different problem arises regarding unsuccessful applicants and previous employees. These groups are beyond the employer’s control, and an employer that now lawfully has declined to obtain this information should not be put to the burden of collecting it at the behest of the plaintiff. However, courts have approved a variety of proxies for the missing information. For example, some courts have approved “geo-coding” as a method of profiling applicants whose demographics are unknown. This method consists of using the addresses of current employees, first to identify the neighborhoods where they reside, then obtaining public information regarding the neighborhood’s demographics (perhaps from the Census), and then imputing to applicants the demographic characteristics of the neighborhoods where they reside. For example, an applicant who lives in a neighborhood that is 50% Hispanic would be assigned a 50% likelihood of being Hispanic. Were all applicants from similar neighborhoods, this proxy would support the inference that the applicant flow was 50% Hispanic. 

Similar approximations have been approved by courts. In the landmark pattern-or-practice case of Hazelwood School District v. United States, the Supreme Court held that the appropriate comparison was to the representation of Black teachers in the relevant labor market area, although a fact issue existed regarding how it should be defined. Although there are obvious limits to the probative value of these approximations, the point is that courts have approved ways of filling this evidentiary void and the true choice facing employers is whether they are more likely to prove nondiscrimination based on actual applicant data or its proxies.

In the Absence of Data, an Employer Cannot Assess its Exposure to Discrimination Claims Brought Against it  

In the absence of demographic data of some type, an employer cannot assess whether statistical analysis will favorably portray its employment practices—an employer essentially will be flying blind as litigation unfolds and recent executive orders may compound this problem. Because the EEOC no longer will pursue disparate impact claims, it is likely that an EEOC charge, which remains a prerequisite to a lawsuit, will quickly be dismissed by the agency. As a result, an administrative process that previously may have taken years to be completed now may end within days, thereby depriving the employers of the time it previously had to evaluate its data, assuming it did not routinely engage in that analysis. Now, because plaintiffs must file suit within 90 days of the EEOC’s notice, an employer that must assemble its data from scratch will have lost the advantage of knowing before the plaintiff where its data analysis leads. 

States Are not Bound by Federal Executive Orders 

It is reasonable to suppose that the void left by the EEOC will be filled, in part, by state anti-discrimination agencies. Indeed, in December 2025, New Jersey formally amended agency regulations to expressly reference disparate impact discrimination as a cause of action. New Jersey also requires “covered” employers that contract with the state to report annually their workforce demographics. Many other states and local governments also require government contractors to maintain records on workforce demographics. Employers in regulated industries may also be subject to such requirements. California for some time has required all employers with any employees in the state to report on employee demographics and Illinois has had reporting requirements for employers with at least 100 employees in the state. Other states that have recently enacted reporting requirements include Colorado and Massachusetts. Multistate employers in these states may find the cost of carving out demographic data pertaining solely to states more burdensome than creating and maintaining a national database inclusive of demographic data. 

EEO-1 Reporting May Return

The Trump administration’s proposal to eliminate EEO-1 reporting represents a dramatic rejection of a requirement that has enjoyed bipartisan acceptance for 60 years. If the administration moves forward with its proposal to rescind the requirement, there are likely to be legal challenges to the decision that could result in the continuation of reporting at least through the 2025 reporting cycle (which would normally take place in 2026) and, perhaps, even through the 2026 reporting cycle (which would normally take place in 2027). Therefore, at a minimum, employers should not stop collecting and maintaining demographic information until it is clear that EEO-1 reporting will no longer be required. 

Furthermore, even if judicial challenges to the EEOC’s rescission of the reporting requirement fail, it seems very likely that a future Democratic administration would promptly revive workforce reporting.

Ensure that any Demographic Data Collected Is Used Only in a Lawful Manner

Of course, collecting any demographic information comes with the risk that an employer will be accused of using it improperly or making decisions affecting employees and employment in an unlawful manner. In that light, it is worth remembering that it is one thing for employers to continue to collect demographic information on a voluntary basis from their workforce—but far more important is ensuring that this information is used lawfully and for the purposes intended only, such as those discussed above. 

Demographic information regarding race, ethnicity, sex (or, for that matter, any other protected category) cannot and must not be used to inform or influence any particular employment decision. Given this prohibition, employers that continue to collect this information may wish to consider as a recommended practice limiting its dissemination only to those with a genuine need to know; not sharing any demographic information with those involved in selecting and hiring candidates; and ensuring that they have a robust anti-discrimination policy and training in place that makes clear that information about any individual’s race, ethnicity, sex, or other protected trait must not be used directly or indirectly in making any employment decision or otherwise impacting the terms and conditions of employment. 

In short, if, as expected, the EEOC rescinds the requirement for employers to collect and report annually on the racial, ethnic, and sex demographics of its workforce, employers will be faced with a choice: they can either continue to collect and maintain this information data for their own lawful uses or they may dispense with this process entirely. There are pros and cons to each choice, and these should be weighed carefully, relative to the circumstances of each employer. Moreover, given the rapidly evolving legal landscape in this area, employers contemplating these choices are well served to do so with the advice of counsel.