What is Disparate Impact? Unintended Discrimination and Company Policies

Disparate impact is a form of discrimination prohibited by law. The EEOC defines disparate impact as ‘a facially neutral employment practice that has an unjustified adverse impact on members of a protected class.’ Disparate impact is illegal under Title VII of the Civil Rights Act at the federal level and also under most state and local levels. Disparate impact is unintentional discrimination. Unlike disparate treatment, it results when an employer’s policies or practices:

• Inadvertently exclude or cause harm to a protected class of worker(s)
• Benefit those in non-protected classes greater than those who are not in non-protected classes

Many organizations are unaware they are causing disparate impact until it’s too late. A best practice is to understand what disparate impact is, how it may occur, and how to change policies to avoid it.

Understanding protected classes

Under Title VII, 9 categories of workers (and applicants) enjoy protected status. In the U.S., the main protected classes are:

• Race
• Color
• Religion
• National origin
• Sex – which includes sexual orientation and gender identity
• Age — 40 and above
• Mental or physical disability
• Veteran status
• Genetic information

There are state and even local governments which have additional protected classes defined. Contact your local Department of Labor to learn about additional protected classes where you do business.

Disparate impact occurs when company policies or procedures disadvantage one or more people in one of these groups.

These classes are protected from discrimination and harassment in the workplace based on their status as individuals or groups. Disparate impact occurs when company policies or procedures disadvantage one or more people in one of these groups. It can occur during the recruitment and screening process, interviews and selection, and throughout the employment relationship.

What are examples of disparate impact?

An historical traditional practice of police and fire departments provides one common example of disparate impact. In the past, most set a minimum height requirement for officers and recruits. That policy tended to exclude women, generally not as tall as men. Today, height requirements have been eliminated from most hiring and training practices for these jobs removing the disparate impact they caused.

Can you measure disparate impact?

Disparate impact can occur in several ways. On an objective level, some types are pre-employment testing or testing requirements for advancement in the company. In these cases, organizations use scored tests (often included with screening software) to assess general knowledge. These tests may be problematic if they skew in favor of non-protected applicants. When scored testing is not involved, assessing and analyzing results can uncover disparate impact. These findings measure the outcomes of employment practices to determine if they are having a disproportionate impact on protected classes.

What is the 4/5 rule?

The EEOC created Uniform Guidelines on Employee Selection Criteria to determine whether scored tests adversely impact candidates. Commonly known as the 4/5^th or 80% rule, these guidelines help businesses measure whether or not testing (or other practices) creates disparate impact. Scoring criteria can easily reveal whether policies or practices give favor or advantage to non-protected classes of workers. The rule finds an adverse impact if members of a protected class are selected at a rate less than 4/5^th (80 percent) of a non-protected group. An example can be pre-employment testing. A non-passing ratio would be 25/50, or 50%, when a test is administered wherein 50% of white males receive a passing grade, but only 25% of African-Americans pass. This rate would violate the 4/5^th rule. A company may promote to fill 20 similar positions from within; 50 men and 100 women apply. If the organization promotes 10 men and 10 women, they may believe there has been no disparate impact. However, based on the number of applications, not transfers, they would be mistaken. One in 5 men, or 20%, were promoted, whereas one in 10 women, or 10%, were promoted. Women were advanced at half the rate of men, or 50%, far below the 80% or 4/5^th cutoff.

Does this affect employment tests?

Your company may administer cognitive tests pre- or post-employment to assess:

• Memory
• Perceptual speed
• Reaction times

These tests may adversely impact more senior applicants and employees based on completion results. Physical and mental fitness test requirements may adversely impact persons with disabilities. This group may be able to perform the primary functions of a position with or without accommodation. As a result, group members may be excluded at a higher than allowable rate due to testing requirements. For some positions, English proficiency tests may discriminate against non-native-born Americans. These tests may not be justified if the job does not require high-level language skills, like editing or proofreading.

Non-scored, or subjective, criteria

Tests are not the only way organizations can measure whether or not there is disparate impact. The numbers often speak for themselves. The ratio at which you select, interview, and hire, based on the number of applications or resumes received, can reveal disparate impact.

The ratio at which you select, interview, and hire, based on the number of applications or resumes received, can reveal disparate impact.

The promotion rates, disciplinary action, negative or positive performance evaluations, or terminations may outline disparate impact. Also, measure wages, compensation, and bonuses to assure equity across all categories of employees. Use the same 4/5^th rule to assess these metrics and ensure no disparate impact.

‘Like me’ bias

A more subjective criterion used in hiring and promoting may be in the interview process. During these interactions, affinity bias may be skewed in favor of or against workers. Often called the ‘like me’ bias, recruiters and hiring authorities tend to gravitate toward people with whom they find commonalities or similarities. It might not be based on a protected class, like gender or race. There might be a predisposition to someone who went to the same school or enjoys the same hobbies. It could be a bias toward people who ‘look like me’ or ‘sound like me.’ It can be deliberate or unconscious.

How this goes astray

For some managers, the gateway to hiring or promotion is an applicant or employee that’s ‘comfortable’ or a ‘good fit.’ They aced the interview because they were at ease. This may be the result of an affinity bias. The interviewer may have felt more at ease based on their ‘like me’ preference, allowing the candidate to do so, as well.

"Like me" bias can be deliberate or unconscious.

These subjective interviews can be difficult to quantify, but the results can be revealing. Interviewers who ask more questions to explore common ground with a certain candidate lead them to believe the applicant is a cultural fit. The problem is the interviewer didn't pursue the same level of questions with all prospects. If the net result is more hires in non-protected classes, there may be disparate treatment. One way to minimize affinity bias in hiring (or promotion) is to use a structured interview. A structured interview process asks (and scores) all applicants the same predetermined set of questions. Whether or not the interviewer had an extended or enjoyable conversation with the candidate is irrelevant to the hiring criteria. Hiring is based on structured interview responses alone. They may have found a long-lost friend, but it shouldn’t influence a candidate’s hiring suitability.

Is disparate treatment ever allowed?

Disparate treatment is allowed in very few cases. When an organization can demonstrate a legitimate business necessity to include or exclude a category of worker, it may be justified. An example may be hiring only men to supervise a men’s locker room. To make a case for a justifiable business necessity, there must be legitimate, subjective proof that another category of worker is unsuitable for the role.

Disparate impact is serious

Disparate impact is a result. It can be the outcome of practices that aren’t intended to be discriminatory. It can be the result of deliberate discrimination. Ultimately, whatever the intent or motivation, disparate impact can occur. A best practice is examining policies and practices to ensure no intentional discrimination is necessary. Look beyond to ensure even the most well-intended or neutral policies don’t have a disparate impact. The numbers will show whether you’ll need to make changes and where.