Identical Résumé, Different Verdict: What the John and Jennifer Study Means for Your Authority
They liked Jennifer more. It bought her nothing.
That is the finding sitting in the middle of the most-cited gender bias experiment in science, and it is the one almost nobody quotes. It is also the one that should change how you spend your effort.
What the study actually did
In 2012, Corinne Moss-Racusin and colleagues sent a single application for a laboratory manager position to 127 biology, chemistry, and physics professors across six research-intensive universities. Same experience. Same coursework. Same recommendation. One variable changed. Half the faculty saw the name John. Half saw Jennifer.
John was rated significantly more competent and more hireable. He was offered more career mentoring. And he was offered a starting salary of $30,238 against Jennifer’s $26,508, a gap of roughly $3,730 on an application that was word for word the same document. The effect sizes ran from 0.60 to 0.75, which is moderate to large.
Female faculty did this at the same rate as male faculty. The result held across discipline, age, and tenure status.
Now the part that gets left out of the retellings. Those same faculty reported liking Jennifer more than John, 4.35 against 3.91 on a seven-point scale. The warmth was real. It converted into nothing. Not the job, not the salary, not the mentoring. The authors ran a mediation analysis and found that Jennifer lost the position specifically because she was read as less competent. Being liked had no path to the outcome at all.
The methods note that changes the reading
Here is the line worth sitting with, from the study’s own methods section. The application was deliberately built to reflect “high but slightly ambiguous competence,” so evaluators would have room to vary.
That was a design choice, and it tells you exactly where the bias lives. Not everywhere. In the band where the record is thin enough that someone has to fill the gap with judgment.
The wider literature supports that reading. The same direction showed up in 1999, when Steinpreis and colleagues found psychologists more willing to hire an identical CV carrying a male name. But in 2023, Stephen Ceci, Shulamit Kahn and Wendy Williams published an adversarial collaboration in Psychological Science in the Public Interest, synthesizing two decades of evidence across six domains. Three researchers who had disagreed publicly for years worked through the literature together.
Their conclusion was mixed in a specific way. On tenure-track hiring, grant funding, and letters of recommendation, where records are thick and criteria are written down, women now do as well as or better than comparable men. On teaching evaluations, where the judgment is subjective and the record is thin, women are still rated lower while teaching just as effectively.
Same pattern in both directions. The gap tracks ambiguity, not the presence of women.
Where this puts your authority work
Authority is often described as being taken seriously, which is true and not very actionable. The research points at something you can actually build. Authority is the project of narrowing the band where someone else’s assumption gets to do the work.
You are not trying to be more likeable. The data says warmth does not convert. You are not trying to be more qualified either, because Jennifer was exactly as qualified as John and it made no difference. You are trying to leave the evaluation less room to guess. That is a different job, and you control it.
Three moves
Make one question unambiguously yours. Pick one valuable question, narrower than a field and specific enough to be checkable, and build a visible track record on it over a couple of quarters. A thick record on one question beats a thin record on ten, because the thin ones are exactly where the guessing happens.
Get the criteria written down before the work is judged. Uhlmann and Cohen showed in Psychological Science in 2005 that evaluators quietly redefine what the job requires to match whichever candidate they already prefer. The important half of that finding is the fix: when people committed to the criteria before they saw who the applicant was, the discrimination disappeared. So ask what success looks like at the start of a project, in writing, in the shared document. That question sounds like diligence, and it is doing something else entirely.
Stop paying for warmth you are not collecting on. Many capable women invest heavily in being easy to work with, and it is not wasted socially, but the evidence is clear that it does not convert into competence judgments on its own. If you are managing how you land more than you are making your reasoning visible, the ratio is backwards. Redirect that effort into the interpretive layer: the call you made, the risk you named, the tradeoff you owned. That is what moves the read on you, as Managing Perception When You Use AI at Work works through in practice.
The study is fourteen years old and still gets shared as proof that the situation is hopeless. Read the methods and it says the opposite. It says the gap opens where the record is thin, which means the record is the lever.
Want to know whether authority is your actual constraint right now? Take the Impact Architecture diagnostic. It scores all four dimensions and names the one holding you back.
Sources
- Moss-Racusin, C. A., Dovidio, J. F., Brescoll, V. L., Graham, M. J., & Handelsman, J. (2012). Science faculty’s subtle gender biases favor male students. PNAS, 109(41), 16474-16479.
- Uhlmann, E. L., & Cohen, G. L. (2005). Constructed criteria: Redefining merit to justify discrimination. Psychological Science, 16(6), 474-480.
- Ceci, S. J., Kahn, S., & Williams, W. M. (2023). Exploring gender bias in six key domains of academic science: An adversarial collaboration. Psychological Science in the Public Interest, 24(1), 15-73.
