Economists find discrimination in recruiting by race and gender. We can change it.

Economists from the University of California, Berkeley, and the University of Chicago last week released the results of a discrimination audit of 108 companies. The audit revealed that entry-level applications from candidates with a “Black name” get fewer callbacks than similar applications bearing a “white name.” The resumes were identical except for the names. The good news is that the researchers found that employers with centralized human resources handling job applications tend to discriminate less, suggesting that uniform procedures can help reduce racial and gender bias. We have the knowledge, processes, tools, and AI-powered technology to change it. Do we have the will?

As reported by The New York Times, the researchers began in 2019 to send 83,000 fake job applications for entry-level positions to 108 companies. Most of the companies were in the top 100 of the Fortune 500 list and some of their subsidiaries.

The researchers, Patrick Kline and Christopher Walters of University of California at Berkeley and Evan K. Rose of the University of Chicago, found that distinctively Black names reduce the likelihood of employer contact relative to distinctively white names by 2.1 percentage points, an effect equal to 9% of the Black mean contact rate.

This gap in contact rates varies substantially across the firms in the study. In the study, researchers found most firms exhibited mild discrimination against Black applicants but a few exhibit substantial levels of discrimination. They estimate that the top 20 percent of discriminating firms are responsible for 46 percent of the lost contacts to Black applicants in their study. Looking at the numbers, for every 1,000 applications received, the researchers found white candidates received about 250 callbacks compared with about 230 for Black candidates. However, among top discriminators, the average gap grew to 50 callbacks.

Companies also varied enormously in their treatment of applicant gender. On average, male and female applicants are equally likely to be contacted, but statistical analysis also showed that a few companies discriminate more than others. “Like racial discrimination,” the authors write, “gender discrimination is highly concentrated in particular firms, with the top 20 percent of discriminating firms responsible for 57 percent of contacts lost to gender discrimination.

In contrast to race, gender, and age, the authors found no significant discrimination with callbacks for membership in a Lesbian, Gay, Bisexual, Transgender, or Queer (LGBTQ) club or evidence of heterogeneity across firms.

The authors found that there was no significant concentration of discrimination in regions of the country. However, they did find disproportionately clustered discrimination in customer-facing sectors, including the auto services and sales sector and certain forms of retail. They found large contact gaps favoring women at apparel stores but favoring men in the wholesale durable sector. These discriminating firms are over-represented in the auto sector, general merchandising, and eating and drinking establishments.

The authors found that discriminating companies tend to be less productive. Their conclusions matches the finding of other studies, including economist Gary Becker’s groundbreaking research from the 1950s finding it is costly for firms to discriminate against productive workers.

Angela Hood, the CEO and Founder of ThisWay Global, has had first-hand experience with resume discrimination. “When I was applying for engineering roles, I noticed that when I changed my name on my resume from my full name to just “AL Hood,” I received more callbacks. You know, I wasn’t mad about it. It is just what you had to do in the 1990s as a woman to get your foot in the door. It is sad that we are still there today. We have made such little progress.” Angela started her company, in part, due to the discrimination she faced. ThisWay Global uses artificial intelligence to rapidly find qualified candidates on the internet or in the applicant tracking system based on objective job requirements. The resumes are presented to recruiters redacted, meaning they have no clue as to the candidates’ name, race or gender and only match candidates’ skills to the job requirements. More on ThisWay Global below.

Over the last several years, several studies have found that organizations that are diverse — racially, ethnically, and in terms of gender — achieve better financial results than ones that are not.

The empirical evidence from McKinsey and others, such as economist at MIT and George Washington, shows that having women and people of color on boards, in leadership positions, and more diverse workforces drive improved revenue and profitability. In the latest report, released in May 2020, the McKinsey authors write: “In 2019, top-quartile [financially performing] companies outperformed those in the fourth quarter by 36 percent in profitability, slightly up from 33 percent in 2017 and 35 percent in 2014. As previously found, the likelihood of out performance continues to be higher for diversity in ethnicity than for gender.”

But the overall progress on diversity and inclusion is stalling and uneven, slipping backwards in some companies, and accelerating in others. Again, McKinsey’s authors:

“By following the trajectories of hundreds of companies in our data set since 2014, we find that the overall slow growth in diversity often observed in fact masks a growing polarization among these organizations. While most have made little progress, are stalled or even slipping backward, some are making impressive gains in diversity, particularly in executive teams. We show that these diversity winners are adopting systematic, business-led approaches to inclusion and diversity (I&D).”

The chart below illustrates McKinsey’s findings on those companies making progress with executive diversity, those falling behind, and those “resting on their laurels”:

Clearly, many corporations have stalled progress, but others have shown that biases can be driven out of recruiting and business cultures. Now, is the time to make real progress with diversity and inclusion. The good news is that companies that value diversity and inclusion and treat employees with respect and dignity can achieve success and grow their businesses financially. However, this result requires publicly stated values and business and HR operating systems to power that success.

I have five tips for companies that want to drive bias out of their recruiting, especially for highly skilled roles such as engineers, finance, sales and marketing workers, and leaders.

No. 1. It starts with values from top executives. Executives who value diversity must make their values well known and build them into the operating tenets of their business. They need to make it clear to their employees, subcontractors, and suppliers that they will not tolerate harassing or discriminating behavior. These executives need to fund strategies and tactics to ensure their values are lived daily. They need to measure the effectiveness of their diversity and inclusion goals for recruiting, pay, promotions, and career development.

Now turning to recruiting:

No. 2. Look at your job postings. Your job postings may be your worst nightmare, driving away job applicants (especially female ones) with jargon and needless clutter. If your job posting is more than 250 words and uses gender-biased wording, it is time for an overhaul. Learn more here. As you will see in No. 3, digital technology can also screen for gender-based wording.

No. 3. Use AI technology de-biases candidate search and screening: AI technology used in recruiting exists to identify qualified job candidates based on skills-based criteria. The best of these technologies presents candidates in a redacted way to conceal names, race, ethnicity, and gender to prevent human bias during resume screening. AI-powered tools, such as ThisWay Global, spent years in development before being made available commercially. ThisWay Global’s AI identifies the top candidates in nanoseconds using objective criteria and instantly presents the most qualified candidates. The company has made over 15 trillion matches and has onboarded over 5,000 customers.

Angela Hood says her AI now automatically removes bias and gender neutralizes job descriptions for divisive language that often drives away women and people of color from the job. It screens out language like we are looking for ninja sales workers. “The negative language isn’t intentional by the job description writer,” Angela says. “They are often just cutting and pasting job descriptions. It hurts the candidates and recruiters.” ThisWay Global has seen a significant impact, and her clients love it.

Angela observes that the workforce and middle management in companies are changing, becoming more diverse. “One of the top three criteria used by job candidates in their searches is how much companies value diversity and inclusion. If the candidates don’t see diversity,” Angela warns, “they will select another employer.” Learn more here. Or visit YouTube to watch my 2019 video with Angela Hood on ThisWay Global and the future of AI and recruiting.

No. 4. Use validated assessments

With technology that can screen job candidates without discriminating bias, what can we do to minimize bias once humans enter the recruiting process? Tips No. 4, validated assessments, and No. 5 structured interviews, provide the path forward.

Validated assessments can be useful to screen job candidates if they have been statistically validated to show that they are effective in selecting job candidates that perform well on the job and are not biased.

First, the vendor of the assessment needs to conduct a statistically validated study. Then the company needs to “benchmark” the study against the performance of its workforce to determine if the assessment can identify the best performers among its current employees. Finally, companies need to decide where to set the scoring threshold to move job candidates to the next step in the hiring process.

However, assessments should never be the only tool used in hiring decisions. They are most successful when used with structured interviews and background checks. If the assessment is not validated, companies can be liable for discriminatory hiring practices.

When using assessments for selection, companies also need to track the results of their hiring decisions and look for evidence of bias. Today’s machine learning software in applicant tracking and talent management systems makes this work much easier to achieve.

When deciding on a test to use in hiring, one size doesn’t fit all shoes. Assessments have been designed for different job families and the skills and attributes required on the job. For example, the Hogan Personality Inventory (HPI) has proven to be a valid predictor of job performance for jobs, such as management, customer service, hospital administrators, and police officers. It is one of the most studied and validated leadership assessments ever. I have personally found HPI to be an excellent tool for coaching managers on leadership style.

I have also had great success when selecting sales workers using a different validated assessment, such as the Berke assessment. A modified version of the Berke assessment is excellent at selecting sales leaders. They are different assessments because the skills and personalities of sales leaders are different from sales workers.

No. 5. Use structured interviews. Despite 100 years of empirical evidence, many companies do not put the time and effort into setting up effective interviewing techniques to significantly improve their ability to hire top talent. Instead, they rely on casual, poorly prepared interviews (academic researchers call them “unstructured interviews”) that are heavily affected by first impressions and unconscious bias, which can lead to discrimination. Not a good way to hire.

A structured interview is based on the knowledge, skills, abilities, and competencies to perform the job and align with company values. With structured interviews, the company creates questions to ask each candidate to determine if they have the education, experience, technical skills, social skills, and emotional intelligence to do well on the job. They can significantly remove the bias out of interviewing. Learn more here.

Research since the 1950s reveals that companies that discriminate do not perform as well as companies that thrive at diversity and inclusion. Economists using big data can now identify the industries and businesses where discrimination occurs. Potential exposure to regulators such as the Equal Employment Opportunity Commission is one concern. The more significant incentive, however, is the improvement in productivity, sales, revenue, and profit in business that do not discriminate. We have the knowledge, processes, tools, and AI-powered technology to make a significant change and create a world of work without bias. Are you willing to act?


Victor Assad is the CEO of Victor Assad Strategic Human Resources Consulting, managing partner of InnovationOne, and Sales Advisor to MeBeBot. He works with companies to transform HR, implement remote work, recruit executives, and develop extraordinary leaders, teams, and innovation cultures. He is the author of the highly acclaimed book, Hack Recruiting: the Best of Empirical Research, Method and Process, and Digitization. Subscribe to his weekly blogs at 


Leave a Reply

%d bloggers like this: