Job interviews are supposed to measure competence. A 2020 paper in the Journal of Personality and Social Psychology suggests they often measure something else: how comfortable a person is with sounding certain. Peter Belmi, Margaret A. Neale, David Reiff and Rosemary Ulfe studied the relationship between social class and overconfidence, and argued that class-based inequality can be reproduced through a deceptively simple mechanism. Higher-class people may be more likely to overestimate their own performance, and observers may mistake that overestimation for competence.
The title of the paper is unusually direct: “The social advantage of miscalibrated individuals.” That word, miscalibrated, is doing a lot of work. It does not mean stupid, fraudulent or incapable. It means a person’s confidence and their actual performance are not lined up — and that the gap itself can be socially rewarded.
The finding is worth taking seriously, but it should not be read as the final word. This is one body of work, not settled consensus. The useful point is narrower and more uncomfortable: in some hiring contexts, what looks like superior ability may partly be superior comfort with sounding certain. Confidence is not distributed only by ability. It is shaped by class, culture, family history and the social environments in which people learn what kind of self-presentation is acceptable. Some people are trained early to sound assured before the evidence is complete. Others learn that overstating themselves is risky, rude or socially costly.
The interview problem
In everyday hiring, miscalibration can be rewarded. A candidate who speaks fluently about their talent, answers with certainty and presents themselves as obviously capable can appear stronger than a candidate who is more measured. The second person may be just as capable, or more so, but their accuracy can read as hesitation.
That is the quiet class issue in the finding. The lower-class or working-class candidate is not framed as less able. The issue is that they may be less inflated in how they assess and present themselves. If an interviewer treats inflation as evidence, the system begins to reward the wrong signal.
This is not a small distinction. Many workplaces already treat interviews as moments of personal revelation. A candidate has an hour to prove not only that they can do the job, but that they can narrate themselves in the right register. The polished answer, the clean anecdote, the easy claim of leadership and the lack of visible doubt can become part of the score, even when nobody says so.
Class as a style of certainty
Social class is often discussed through money, education and occupation. Those things matter, but class also travels through behaviour. It teaches people what to expect from institutions, what they are allowed to ask for, how much space they can take up and how harshly they may be judged for getting the tone wrong.
For people from more privileged backgrounds, self-promotion is often safer. They may have had more practice being asked for opinions, encouraged to advocate for themselves and rewarded for presenting ambition as natural. Confidence can become a familiar social language. For people who grew up working class, the lesson can be different. Modesty may be valued. Bragging may be punished. Certainty without proof may be treated as arrogance. A person may learn to understate their strengths, wait to be asked, or qualify their claims because that is what counted as respect in the environment that formed them.
None of this means every higher-class person is overconfident or every working-class person is modest. Class never works that neatly. But the paper points to a pattern that hiring systems often ignore: self-presentation is learned, and interviewers may confuse a learned style with a personal trait.
What overconfidence buys
The most troubling part of the finding is not that some people overestimate themselves. Humans do that all the time. The troubling part is that observers can reward the overestimation.
In a job interview, overconfidence can look like command. It can make answers sound cleaner than they are. It can turn a limited example into a story of obvious potential. It can make a person seem like they belong in the room before they have shown much evidence that they can do the work. That is why the paper is not really about confidence as a private feeling. It is about confidence as a social signal. A signal only matters if someone else reads it. If interviewers read inflated confidence as competence, they are not simply observing candidates — they are helping convert class-coded self-presentation into opportunity. The effect can be especially powerful because interviewers rarely think they are selecting for overconfidence. They think they are selecting for leadership, clarity, presence, communication, ownership or potential. Those words are not meaningless. Some jobs genuinely require calm communication under pressure. But they can also become soft containers for class advantage when they are not anchored to job-relevant evidence.
Why structure matters
This is where the hiring literature becomes useful. A 1994 Journal of Applied Psychology review and meta-analysis by Michael A. McDaniel, Deborah L. Whetzel, Frank L. Schmidt and Steven D. Maurer examined the validity of employment interviews. One of the enduring lessons from that literature is that interviews are not all the same. More structured interviews, built around consistent questions and job-related evaluation, tend to be more defensible than loose conversations in which impressions roam freely.
That matters because an unstructured interview gives overconfidence more room to work. If every candidate is asked different questions, if the scoring criteria are vague, and if the interviewer is left to judge “fit” by feel, the confident performance can expand to fill the space. The interviewer may remember the person who sounded decisive, not the person whose evidence was strongest.
A structured process does not remove bias. It does not make hiring pure. But it forces a basic question that unstructured interviews often avoid: what exactly are we measuring? If the answer is the ability to do the work, the process should make candidates demonstrate job-relevant judgement, skill and reasoning, not only the ability to speak about themselves with ease.
Work samples, practical tasks, consistent scoring rubrics and interview questions tied to the actual role do not solve class inequality. But they reduce the amount of hiring that depends on the interviewer’s intuitive reading of confidence. That is already a meaningful shift.
The wrong lesson
The wrong lesson from this research is that working-class candidates should simply become more overconfident. That turns a structural problem into a personal performance burden. It also asks people to adopt the very distortion the paper is warning about.
A better lesson is that employers should become more suspicious of certainty when it appears alone. Confidence can be useful. It can help someone communicate clearly and act under pressure. But confidence without evidence is not competence. A candidate’s ability to sound convincing should not be allowed to stand in for their ability to do the job.
There is also a moral trap here. Many hiring cultures say they want humility, learning ability and honesty. Then they reward the candidate who speaks as if there is no gap between current ability and future performance. The system praises self-awareness in theory while selecting for self-inflation in practice.
That contradiction does not only harm individual candidates. It changes who gets promoted, who is seen as leadership material, who is interpreted generously and who is asked to keep proving themselves. Over time, small misreadings of confidence can become large differences in opportunity.
What interviewers keep mistaking
The most useful reading of the Belmi paper is not that interviewers are foolish. It is that interviews are socially dense situations. People are trying to make high-stakes judgements with limited information, under time pressure, while using signals they may not fully understand.
That is exactly why class can hide inside merit. Nobody needs to say they prefer privileged candidates. Nobody needs to consciously penalise working-class candidates. The process only has to reward a style of self-presentation that privilege made easier to learn.
When that happens, the interview is no longer just a test of capability. It is also a test of calibration, ease and entitlement to sound certain. The person who grew up learning to be accurate about themselves may lose to the person who grew up learning that confidence would be taken as proof.
The paper’s uncomfortable implication is that some interviews may be selecting for the wrong kind of signal while calling it potential. The working-class candidate is not necessarily less capable. They may simply be less overconfident. And in a hiring culture that keeps mistaking overconfidence for competence, that difference can be enough to change who gets the job.
