Thursday, March 30, 2017

Customer Preference-Based Discrimination in the Information Age

Customer preference is never a justification for discrimination in employment.  That’s been well-settled law for decades.  A male ultrasound technician cannot be terminated because patients prefer female technicians.  An employer in “redneck country” cannot refuse to promote a qualified employee based on the belief that customers won’t accept an African American account manager. (EEOC v. Frontier Materials Corp., No. H-03-856 (S.D. Tex. Mar. 2, 2004)). A manager’s perception that Hispanic customers prefer to be served by Hispanic salespeople is not a lawful basis for forcibly transferring staff. (EEOC v. AutoZone, Inc.No. 1:14cv5579 (7th Cir. complaint filed July 22, 2014)).

Employers must base their personnel decisions on non-discriminatory business reasons, and may not consider the biased preferences of customers, clients, or patients.  The EEOC has provided guidance on the topic of customer preference as it relates to national origin discrimination, race or ethnicity-based discrimination, and gender discrimination; federal regulations make clear that the refusal to hire an individual based on the preferences of clients or customers is not included in the “bona fide occupational qualification” exception to anti-discrimination laws.

For most employers, this is a familiar concept.  Personnel decision-making policies and processes may not directly address what to do about customer bias, but employers understand their obligation not to discriminate, and most would not knowingly rely on clearly biased customer complaints or feedback.  But what responsibility, if any, does an employer have to identify customer bias?  This is the age of on-line surveys, customer satisfaction questions at the end of telephone transactions, and anonymous electronic feedback.  If such direct customer input is a significant part of an employer’s evaluation of employees, does it matter that customers’ implicit or explicit bias may affect that input?

Even though the law is clear about employers’ liability for acting on discriminatory customer preferences, we don’t yet know how the courts or the EEOC might treat a claim based on an employer’s reliance on anonymous customer feedback.  Although it’s not easy reading, a recently published law review article by Professor Dallan F. Flake, titled When Should Employers Be Liable for Factoring Customer Feedback Into Employment Decisions? offers a detailed and thought provoking exploration of the history, current state, and possible future of the law in this area.  This and other commentary suggest challenging questions.  For example, if customers provide negative feedback about a call center employee with an accent, is it because the customers don’t like people they perceive to be “foreign” or “immigrants,” or is it because the employee isn’t helpful?  If patients rate their experience with Dr. Deep as unsatisfactory, is it because Dr. Deep is Sikh and wears a turban, or because Dr. Deep is disrespectful?  Employers cannot reasonably know (or inquire into) the biases of unidentified customers, and are extremely limited in their ability to discern bias even when customers can be identified.  As on-line, anonymous, and aggregated ratings by customers become more common, and to the extent that they play a significant role in employee evaluations, the existence and potential impact of customer bias may become an important consideration.

The so-called “gig economy” has created special challenges related to customer bias.  For some thoughtful discussions of bias in online survey platforms and the impact of customer bias on gig economy workers, read here and here.

We will monitor cases and developments related to customer bias and urge our clients and readers to do the same.  In addition, we have suggestions that we think make sense, given what we know and what we don’t know about the potential for employer liability:
  • If customer bias is clear in an employee rating, discount the rating or limit its impact.  Employers should be wary of customer ratings that overtly state a discriminatory preference or bias.  Such a rating may contain other important information that can be considered, but employers should note, and keep a record of, their recognition of express bias.
  • If you ask for customer comments, read them and take note of biased statements.  Particularly when employers must deal with a high volume of customer ratings, comments can be overlooked.  If requested, comments should be considered; if they are not considered, it may be best not to solicit them at all.
  • Ask for specifics about the customer experience.  The more specific the questions asked of customers who rate employees, the more job-related the feedback is likely to be, and the less likely it is to be driven by discriminatory attitudes.
  •  Ask for explanations of low ratings.  A customer’s explanation of a low rating may reveal bias that is not apparent in a numerical rating, allowing the employer to take the existence of bias into consideration.
  • Limit the use of customer ratings in the evaluation process.  Supervisor observation, quantitative measures of efficiency, and other evaluative techniques not reliant on customer feedback should be incorporated into the evaluation process.  When customer ratings are used, they can be provided to the employee and used as a basis for discussion and improvement, rather than being the only basis for evaluation.
Someday, employers may be able to screen customers’ feedback for bias through the use of algorithms and big data.  For now, the reasonable employer must rely on the good judgment of those who hire, fire and evaluate to recognize obvious customer bias and limit its impact on employees.

Posted by Judy Langevin