From a legal perspective, there are several laws that provide a regulatory framework for promoting equal employment opportunities and reducing adverse impact on protected groups. These laws establish guidelines and standards that employers must follow to ensure fair and non-discriminatory practices in the workplace. Some of the key laws in this regard include:
Title VII of the Civil Rights Act: This federal law prohibits employment discrimination based on race, color, religion, sex, and national origin. It prohibits both intentional discrimination (disparate treatment) and practices that have a discriminatory impact (disparate impact) on protected groups.
Uniform Guidelines on Employee Selection Procedures (UGESP): These guidelines, issued jointly by several federal agencies, provide a framework for evaluating the validity and fairness of employee selection procedures. They emphasize the importance of minimizing adverse impact and ensuring that selection criteria are job-related and consistent with business necessity.
EEOC Fact Sheet on Employment Tests and Selection Procedures: The U.S. Equal Employment Opportunity Commission (EEOC) provides guidance on employment tests and selection procedures to assist employers in complying with anti-discrimination laws. The fact sheet highlights the need to use valid and reliable tests that do not disproportionately exclude or disadvantage protected groups.
These laws and guidelines play a crucial role in promoting equal opportunity and preventing discriminatory practices in the workplace. Employers are encouraged to familiarize themselves with these legal requirements, implement fair and inclusive policies, and regularly assess their practices to mitigate adverse impact on protected groups.
New Legal Milestone: NYC Local Law 144
In recent years, the fairness of automated employment decision tools (AEDTs) has received increasing attention. The New York City Council passed Local Law 144 in November 2021, which mandates bias audits for companies using automated employment decision tools (AEDTs) to screen candidates or promote employees. Effective from July 5, 2023 (initially scheduled for January 1, 2023, and then postponed to April 15, 2023), this groundbreaking legislation requires independent audits of AEDTs. The Department of Consumer and Worker Protection (DCWP) has proposed rules specifying that bias must be assessed using impact ratios based on sex/gender and race/ethnicity categories, at a minimum.
NYC Local Law 144 is a significant step in regulating AI in employment decisions in New York City. It introduces requirements for bias audits, promoting fairness and transparency. Employers utilizing AI-based tools must comply with independent audits, public summaries, and disclosures to ensure accountability and prevent discrimination. This law sets a precedent for the regulation of AI in employment, fostering ethical practices and equal opportunities.
Adverse treatment and adverse impact explained
In this context adverse and disparate are used interchangeably.
Adverse treatment is intentional employment discrimination. For example, a requirement for a college degree that applies only to Latinx.
Adverse impact is unintentional employment discrimination. For example, a requirement for a college degree constitutes adverse impact.
This means that most job postings out there with a college degree requirement (or even a prior experience requirement) run afoul of EEOC regulations. Luckily this is not the case, because unlike adverse treatment (which is always prohibited), selection procedures with an adverse impact are allowed when they are justified (UGESP § 1607.3.A and Title VII).
Justified in this context means that the selection process results are related to the performance of the role in question and that no other, better selection process is known. For example, a French-language assessment would not be justified for all Sales roles in a company; it would be justified for those roles that are specifically targeting French-speaking customers.
In such cases, the adverse impact may be deemed justifiable if it can be demonstrated that the practice or policy is necessary for the effective operation of the organization and there are no less discriminatory alternatives available.
Fairness First: Bryq - Empowering Bias-Free Hiring Decisions
Bryq is designed to mitigate adverse impact by default. As an objective tool, it eliminates common unconscious bias factors that are often present in resume screening, such as name, gender, age, and education. This ensures a fair and unbiased assessment process. It has been shown that the usage of systems, such as Bryq, which contribute to enhanced reliability and validity, are also viewed by the courts as enabling the interview to protect against unlawful employment discrimination (Williamson, L. G., Campion, J. E., Malos, S. B., Roehling, M. V., & Campion, M. A. (1997). Employment interview on trial: Linking interview structure with litigation outcomes. Journal of Applied Psychology, 82(6), 900–912.).
While Bryq does not control the job description provided by the employers, it ensures a bias free AI-generated profile by frequently conducting statistical analysis to validate zero significant difference among EEOC protected attributes. Results of these recurring analyses are published on our website i.e. example from a study focusing on gender differences.
Finally, Bryq acted proactively to guarantee that our customers will have access to a tool that not only fosters inclusivity but also avoids any form of discrimination against protected groups. In particular, we used an esteemed independent auditor, Holistic AI, to conduct an impartial analysis and summary of results is also available to our customers upon request. That way, we managed to remain compliant and uphold the highest standards of regulatory adherence.
Local validation study: A justified selection process
There is ample evidence in I/O Psychology research that a combination of cognitive skills and personality traits in an assessment is the best predictor of future job performance. (For example, see Schmidt, Frank & Hunter, John. (1998). The Validity and Utility of Selection Methods in Personnel Psychology. Psychological Bulletin. 124. 262-274. 10.1037/0033-2909.124.2.262.).
Our AI proposes a weighted average between cognitive attributes and personality skills that will determine the optimal balance to maximized predictive performance while maintaining compliance and inclusiveness.
However, in certain situations with ample data and evidence from real employee data, local validation studies may reveal a different association of one of the two parameters the Bryq Assessment measures i.e. cognitive attributes and personality skills resulting in higher performance prediction for a specific role.
💡 A local validation study can potentially justify adverse impact if it demonstrates that a particular employment practice or assessment, which may have a disparate impact on a protected group, is valid and job-related.
If the study provides evidence that the assessment or practice in question accurately predicts job performance or success, and there are no other equally effective, less discriminatory alternatives available, it may be considered justifiable adverse impact. This means that despite the disparate impact on certain groups, the practice is deemed necessary and valid for effective job performance within the specific organization.
These studies are based on the examination of high-performing employees within your organization and are validated through the actual use of the optimized profile in the recruiting process. Consequently, local validation studies may identify a justifiable adverse impact based on real employee data.
Diversity insights
The Bryq score is not affected by the EEOC data. Nevertheless, in case certain groups are underrepresented in the data what our internal controls are monitoring and flagging is the following:
Display the EEOC survey in a prominent place within the Candidate Portal to prompt users to respond;
Diversity mix table view in our Dashboard to encourage DE&I strategy for our customers;
Quarterly checks to monitor the representativeness of each EEOC attribute.
Resources
SHRM Avoiding (and measuring) adverse impact
Uniform Guidelines on Employee Selection Procedures
EEOC Fact Sheet on Employment Tests and Selection Procedures