Two resumes go out for the same job. They list the same schooling, the same skills, and the same years of experience. The only real difference is the name at the top. One name reads as White, the other reads as Black. On paper the candidates are equal, so the callbacks should be equal too. In a landmark study, they were not.
The study came from economists Marianne Bertrand and Sendhil Mullainathan, published in 2004. They built thousands of fake resumes and mailed them to real job ads in Chicago and Boston. Each resume was randomly given a name that sounded either White or Black, names like Emily and Greg on one side and Lakisha and Jamal on the other. Everything else was shuffled so that skill and name were separated. Then they counted how often each resume got a call back for an interview. The design was clean, and the result was hard to explain away.
The White sounding names received about 50 percent more callbacks than the Black sounding names. Put another way, a job seeker with a Black sounding name had to send out far more applications to get the same number of calls. The researchers also found that a stronger resume helped White names more than Black names, so polishing the paper did not close the gap. Better credentials paid off less if the name signaled Black. That finding cut against the common belief that qualifications alone decide the outcome. The number was steady enough that luck could not explain it.
One study is never the last word, and researchers kept testing it. A large field experiment led by economist Patrick Kline and colleagues, published in 2021, sent tens of thousands of applications to major companies. On average, applicants with White sounding names still heard back more often than those with Black sounding names. The size of the gap varied a lot from one company to the next, which was its own useful finding. Some firms showed almost no difference, while a handful drove much of the bias. The pattern was not everywhere in equal measure, but it was real and it held up.
The reason this design matters is that it removes the usual excuses. When two people differ in schooling or experience, you can point to that as the cause of a different outcome. Here the only thing that changed was a name, assigned at random, so the difference in callbacks points back to the name itself. That is what researchers mean by discrimination in a strict, measured sense. It is not a claim about what any single manager feels. It is a pattern that shows up across thousands of decisions.
For job seekers, the effect is not abstract. A callback gap means more time searching, more applications sent, and more rejection before a fair shot arrives. That extra friction falls on people who often have the least cushion to absorb it. It can push someone toward a lower offer simply because the search dragged on. Over a career, small gaps at the front door add up to real money and lost momentum. The cost is paid in hours, in wages, and in confidence.
The research also points toward fixes that work. Structured hiring, where every candidate answers the same questions and gets scored on the same scale, cuts the room for bias to slip in. Some companies remove names from the first screen so reviewers judge the resume before the person. Clear job criteria set before the applications arrive keep the bar from moving based on who applied. These steps are not about lowering standards, they are about applying one standard evenly. The studies suggest the gap shrinks most where the process is tightest.
The takeaway is not that every employer is unfair. Many are not, and the newer work shows the problem sits heavily with a smaller set of firms. The takeaway is that a name should never be a qualification, yet the data shows it sometimes acts like one. Naming the pattern is the first step to fixing it, because a bias you cannot see is a bias you cannot correct. Better systems beat good intentions, and the numbers back that up. Equal resumes deserve equal calls.
