Job Analysis and Pay Discrimination Case Study

Plaintiff’s Report Findings Excluded at Trial

Pay discrimination cases are common in employment law. The premise of these cases is generally that one or more minority groups (such as African Americans, women, older workers) are underpaid compared to the rest of employees at the same company–and that the pay difference is due to discrimination. These cases can be very persuasive, particularly when it looks like the differences in pay aren’t simply random, but also can’t be obviously explained by seniority or skills. 

An added complication is that these cases are often class action suits, meaning that more than one employee (within the same protected class or classes) is making the same allegations.

The Challenge in Pay Discrimination Cases

Pay discrimination cases often bring in economic experts on both sides to examine important questions that may factor into company-pay practices. At the outset, this involves lots of questions: What are “similarly situated” employees–that is, employees of similar experience level and similar job functions–earning, if they are not part of the plaintiffs group? Do all employees in the class share the same characteristics? Are there any other reasons besides race, gender, or age that these employees might be paid differently?

The Pay Discrimination Case Details

All of these questions were important in a recent case wherein Welch Consulting economists served as expert witnesses. A company’s African American employees believed that they were being paid less than “similarly situated” white employees.

The plaintiff’s expert provided Welch Consulting labor economists with a data set to support this claim. The data included employees’ years of experience, job titles, other data collected from 2008 to 2015, all of which seemed to show that African American employees had been paid less than white employees in the same jobs.

READ THE FULL CASE STUDY HERE

Our Economists’ Approach


An important first step for our economists was to look at all the jobs held by the plaintiffs and examine how much employees in each of those jobs were paid.

When we looked at the data provided by the plaintiff’s expert, we noticed one group of employees that had the job title “temporary employees.” The plaintiff’s expert had grouped both African American and white employees into this group. Other employees were grouped into more specific job titles (such as “machinist,” “janitor,” “welder,” or “pipefitter”), so we thought this “temporary employees” category was a bit odd.

When we asked the firm what was going on, we found out that the temporary employees were in fact doing some of the same jobs as permanent employees–but they were new hires, and were still on a probationary period.

We realized that classifying jobs correctly would be an important part of our analysis. So, we decided to assign each employee with “temporary” status to the permanent job title that best fit the type of work they were doing. Then, we re-ran the plaintiff expert’s analysis. 

Our Analysis


Once we corrected the job titles, the data looked completely different.

From 2008 to 2015, both white and African American employees had been grouped misleadingly into this one “temporary” job category, and as a result, the data gave an illusion of pay disparities. The plaintiffs had assumed that these differences in salaries were due to racial bias.

In fact, when we took out the “temporary” job category and correctly grouped workers by their job functions, there were no longer significant differences between African American and white employees’ salaries. In fact, our corrected data even showed that some pay differences were slightly in favor of the African American employees in the firm.

Results of the Pay Discrimination Case

This was only the first step we took to correct the data, which we later learned contained numerous errors. In addition to our jobs analysis, we realized there was a system-wide glitch that had removed relevant years of experience. This further confused the wage data, since the number of years a given employee had been at the company was not taken into account.

We cleaned up the data and made some additional corrections. When we were finished, the data showed no substantial differences in pay between white and African American employees. This negated the plaintiff’s allegations of employee discrimination.

As a result of our data review and corrections, the court ruled to exclude the plaintiff’s expert report findings from evidence at trial.

Connect With Us


Start the conversation today by contacting us below.