The pay gap isn't about who asks. It's about who you penalize for asking.

Someone on your team asked you for more this year — a raise, a bigger territory, a title. You remember how it landed. From one person, the ask read as confidence: they know their worth, good for them. From someone else, the same ask — same words, same timing, maybe even a smaller number — read as pushy, or entitled, or not quite a fit. You probably didn't put it in those terms. But you felt something different, and you acted on it.

A new study in the Proceedings of the National Academy of Sciences gives that feeling a name and a price tag. "The social-class gap in negotiation: Lower-class individuals negotiate less and face more backlash," by Ying Lin of the Hong Kong University of Science and Technology, Jackson G. Lu of MIT Sloan, and Michele J. Gelfand of Stanford, runs five separate studies: a nationally representative sample of 7,883 Australian employees, 634 U.S. MBA graduates tracked by parental education, a 699-person field experiment on an online labor platform, a stratified survey of 995 U.S. full-time workers, and a controlled experiment with 1,133 U.S. HR professionals evaluating fictitious candidates. More than 11,000 people across all five. Social class is measured through established proxies — educational attainment, parental education, income — and much of the negotiating-or-not data is self-reported, people recalling past decisions on a survey. The HR-professional study is a lab-style experiment with fictional candidates, not real hiring files. Read it as strong, converging evidence across methods, not one definitive field trial.

The finding: people from lower social-class backgrounds are consistently less likely to negotiate than people from higher-class backgrounds — for starting salary, benefits, a promotion, a raise, even the content of their job. In the survey of full-time workers, negotiating a starting salary was worth an average of $3,878 more. Employees who asked for a promotion got one 95 percent of the time; those who didn't ask, 48 percent of the time. And when the researchers tested why the gap exists, part of the answer wasn't ambition or confidence. It was a correctly calibrated fear of backlash: in the HR-professional experiment, identical negotiation requests were judged as less cooperative when the candidate was signaled to come from a lower-class background than when the same request came from a higher-class one.

"Moreover, when lower-class individuals do negotiate, they face stronger social backlash than their higher-class counterparts." — Ying Lin, Jackson G. Lu & Michele J. Gelfand, significance statement, Proceedings of the National Academy of Sciences

Here is why this lands differently for you than for almost anyone else in the building. A biased HR department is one input among several. A biased person at the top is the ceiling. There's no second opinion above you, no committee to catch the moment you read one person's ask as savvy and another's as a problem. Every negotiation that reaches you gets judged by your instinct alone, and your instinct — like every instinct measured in this study — was built on the same associations everyone else's was.

The part the research doesn't say, but that I'll say plainly: the people least likely to ask you for more are not your most content employees. They're the ones who have already run the calculation on what asking will cost them socially, and decided the number isn't worth it. If your strongest performers in that group are quiet, don't read the quiet as satisfaction. Read it as a bet they've placed on you — a bet that the value they're not asking for will show up on its own, in a raise you initiate before they ever have to negotiate for it. Most leaders never test whether that bet is paying off. They just notice, a year or two later, when the person leaves for an offer that didn't require them to ask.

One thing to do this week: pull your last four or five compensation or promotion decisions and sort them by who initiated the conversation and who you approached first. Don't sort by performance — you already track that elsewhere. Just look at who asked. If the pattern tracks anything other than performance — tenure, background, how someone tends to communicate — you've found the gap this study describes, inside your own company, before it costs you someone good.

This is a bias problem before it's a compensation problem, and bias problems live in how you make calls under pressure, not in your written policies. If you want a clearer read on your own decision patterns — where your judgment is sharp, and where it's running on autopilot — the Leadership Profile is built for exactly this kind of self-audit.