Emily & Lakisha 10 Years On: Are Employers Still Biased About Applicants' Names?

You may have heard of an influential study from 2004 with the provocative title: "Are Emily and Greg More Employable than Lakisha and Jamal? A Field Experiment on Labor Market Discrimination."  In it, researchers found that candidates with "African-American-sounding" names had to send out 15 resumes to get one call back, while candidates with "white-sounding" names only had to send out 10 to achieve the same result. They extrapolated that having a "white-sounding" name was the same as a black candidate having eight additional years of experience.

Flash forward 10 years and researchers at the University of Missouri-Columbia decided to repeat the experiment in a new study, unsexily titled "An Updated Analysis of Race and Gender Effects on Employer Interest in Job Applicants." Have hiring practices changed in the intervening decade?

Before we answer that, we should note the researchers made a few changes to the original study. They sent out a lot more resumes and covered a wider geographic area: 9,000 fictitious resumes in response to online ads in seven states. All candidates were made to seem young, with high school graduation dates of 2010. And Hispanic applicants were included as well as black and white ones.

Most importantly, they stayed away from "stereotypically black" first names, because other studies had shown that employers often interpreted them as belonging to people of low socioeconomic status. So "Lakisha Washington" of the 2004 survey was renamed "Chloe Washington." (The other African-American candidate was named "Ryan Jefferson".) The white candidates were called "Megan Anderson" and "Brian Thompson" and the Hispanic candidates were dubbed "Isabella Hernandez" and "Carlos Garcia".

So what did the results show? This time, the researchers found that there was very little difference in the response rates to the black, white and Hispanic candidates. Whites had a slight advantage, but statistically it was a dead heat. "We do not find evidence to suggest that employers systematically discriminate against race and gender groups when responding to resumes from relatively young applicants," the study authors wrote in the paper.

So what accounted for the turnaround? Study co-author Cory Koedel, an associate professor of economics and public policy at the University of Missouri-Columbia, has two theories. "One is the recency of our data. Consistent with this explanation, others have argued that the significance of discrimination is declining," he says via email.

"The second explanation is that we chose to signal intended African-American applicants differently than in previous studies," he says. "Previous studies have used distinctly African American sounding first names, whereas we relied on last names that are strongly associated with race ... Employers may have been less likely to recognize that 'Washington' and 'Jefferson' are strong indicators of [African-American] race in our study, and may not have received the intended signal."

In other words, the employers may not have realized that "Chloe Washington" was supposed to be black even though the U.S. Census shows 90 percent of the people with that surname are African-American. Koedel adds that there doesn't seem to be a socioeconomic distinction for "Hispanic-sounding" first names. 

As to whether he would have got similar results to the 2004 study if he'd stuck with "black-sounding" first names, Koedel thinks the results would have probably fallen somewhere between this current study and the older one. This new study didn't look at whether there was discrimination in other parts of the hiring process such as the interview, so more research would be necessary to understand how that affects minorities and women.