Racial bias exists in the job mar­ket from the moment a job-seek­er hits “sub­mit” on an appli­ca­tion, accord­ing to new research.

Some experts say com­pa­nies can con­sid­er this a reminder to assess their hir­ing prac­tices and make changes when they spot inequities.

New find­ings on racial dis­par­i­ties. Nation­al Bureau of Eco­nom­ic Research econ­o­mists released a new paper detail­ing the results of an exper­i­ment, in which they sub­mit­ted 84,000 job appli­ca­tions to 11,000 jobs at 108 For­tune 500 com­pa­nies between 2019 and 2021.

The researchers sent up to four pairs of appli­ca­tions to entry lev­el jobs that did­n’t require a col­lege degree or much expe­ri­ence. The appli­ca­tions were iden­ti­cal, except for the asso­ci­at­ed names, which were changed to sug­gest the job-seek­er was either Black or white.

Assumed white and female names (like Misty, Heather, and Lau­rie) were con­tact­ed most often, while assumed Black female names (like Lawan­da, Tame­ka, and Latisha) were con­tact­ed the least.

While the experiment—the largest of its kind in the US—discovered that gen­der dis­crim­i­na­tion at the inter­view lev­el is rare, racial dis­crim­i­na­tion appeared preva­lent. Assumed white appli­cants received 9.5% more inter­view offers than assumed Black appli­cants on aver­age.

The find­ings like­ly under­es­ti­mate the fre­quen­cy of racial dis­crim­i­na­tion, since it hap­pens less fre­quent­ly at large com­pa­nies, and oth­er under­rep­re­sent­ed races weren’t part of the research, soci­ol­o­gist Lin­coln Quil­lian told the New York Times.

The worst per­form­ing com­pa­nies, many in the auto indus­try, offered job inter­views to assumed white appli­cants 23% more than assumed Black appli­cants.

Gen­uine Parts and Auto­Na­tion were the only com­pa­nies that received a one star (the low­est) grade. Auto­Na­tion, a used car retail­er, con­tact­ed assumed white appli­cants 43% more than assumed Black appli­cants, and Gen­uine Parts Com­pa­ny, an auto parts com­pa­ny, con­tact­ed 33% more. Con­verse­ly, Avis-Bud­get, Dr. Pep­per, and Charter/Spectrum had lit­tle to no racial dis­crep­an­cies, and received three stars (the high­est).

“We want to bring peo­ple’s atten­tion not only to the fact that racism is real, sex­ism is real, some are dis­crim­i­nat­ing, but also that it’s pos­si­ble to do bet­ter, and there’s some­thing to be learned from those that have been doing a good job,” Patrick Kline, an econ­o­mist at the Uni­ver­si­ty of Cal­i­for­nia, Berke­ley and coau­thor of the study, told the Times.

“We are always eval­u­at­ing our prac­tices to ensure inclu­siv­i­ty and break down bar­ri­ers, and we will con­tin­ue to do so,” Heather Ross, a spokesper­son for Gen­uine Parts direct­ed HR Brew to its state­ment in the Times. Auto­Na­tion did not respond to HR Brew’s request for com­ment.

Tak­ing action. Researchers found that hav­ing a diver­si­ty offi­cer and diver­si­ty train­ing did­n’t appear to reduce the chance of dis­crim­i­na­tion in entry-lev­el hir­ing. How­ev­er, com­pa­nies with cen­tral­ized and for­mal HR offices had less bias, accord­ing to the Times.

Work­place equi­ty schol­ars say that the research is a mixed bag of frus­tra­tion and hope. Ash­leigh Shel­by Rosette, lead­er­ship pro­fes­sor at the Fuqua School of Busi­ness at Duke Uni­ver­si­ty and Acad­e­my of Man­age­ment schol­ar, said that HR lead­ers should look at well-per­form­ing com­pa­nies as mod­els of how to improve their own prac­tices.

“On the oth­er hand, the preva­lence of bias among a sig­nif­i­cant por­tion of the com­pa­nies list­ed in the study is dis­cour­ag­ing but not unex­pect­ed,” Rosette told HR Brew via email. “Find­ings like these sug­gest that the recent surge in resis­tance toward DEI ini­tia­tives lacks sub­stan­tial jus­ti­fi­ca­tion. The neces­si­ty for poli­cies, pro­ce­dures, and pro­grams aimed at mit­i­gat­ing sys­temic bias in hir­ing remains of sig­nif­i­cant impor­tance.”

Adi­na Ster­ling, asso­ciate pro­fes­sor at Colum­bia Busi­ness School and an Acad­e­my of Man­age­ment schol­ar, said that the research is a reminder to HR lead­ers that, regard­less of the Civ­il Rights Act of 1964, Black indi­vid­u­als still receive less inter­est than their sim­i­lar­ly cre­den­tialed white coun­ter­parts.

“I would encour­age HR lead­ers to decide to enact sys­tems-lev­el change to fix this, by for exam­ple, using tech­nol­o­gy and stan­dard­ized selec­tion process­es,” Ster­ling told HR Brew in an email. “I also rec­om­mend that HR lead­ers audit their own firms, so that they can stress-test them to track change.”