The Psychology of the Salary Number

The raise landed in January: 9 percent, the largest of her career. The product manager in Austin had asked for 8 and expected 5, so by the arithmetic it was a win. It did not feel like one. Two weeks earlier, a teammate had mentioned, almost in passing, what the company paid the newest hire on their team — about 18 percent more than her own number. She had not been able to stop turning the figure over since, and by March she was interviewing elsewhere.

Salary is a number with an emotional life. The paycheck buys the apartment and the groceries, but satisfaction is driven less by the size of the number than by how it sits against the people around you — and by the stories you tell yourself about fairness, rank, and what you were promised. Psychologists and economists have spent three decades measuring that gap between the number and the feeling.

This article reports what the research shows about how people actually experience pay: why a coworker's raise can sting more than your own helps, why the same offer feels different depending on where you rank, and what the findings imply for negotiating your next raise and evaluating an offer. The short version: the number matters, but the comparison matters more.

The neighbor's paycheck

The most consistent finding in the study of money and happiness is that relative pay beats absolute pay. In a 2005 study, the economist Erzo Luttmer analyzed data from the National Survey of Families and Households and found that neighbors' earnings dragged on well-being: people were less satisfied when the people around them earned more, even after controlling for their own income and local prices. The effect was largest among people who spent the most time with their neighbors — the comparisons you actually see are the ones that hurt.

A decade earlier, the economists Andrew Clark and Andrew Oswald had found the same pattern inside the workplace. Working with British household data, they built a measure they called comparison income — what similar workers, in similar jobs and regions, were paid — and showed that it entered job satisfaction equations with a negative sign. Raise a worker's own wage and satisfaction rises; raise the comparison wage and it falls. Your pay, their model said, is experienced as a fraction: what you make, divided by what people like you make.

The pattern appears in the national happiness data too. The economist Richard Easterlin documented in the 1970s that average reported happiness in the United States did not rise as average income rose — a finding that still carries his name — and subsequent work has generally confirmed that gains in absolute income matter most at the bottom of the distribution and least once basic needs are covered. For most working adults, the difference between comfortable and more comfortable is decided in comparison.

None of this is an argument that money does not matter. Money buys security, and security is a happiness input of its own; the research also finds that income losses hurt more than equivalent gains help. The point is narrower: once a paycheck covers the bills, the next dollar is judged against the people around you, not against the budget.

Rank matters more than the total

Educational psychologists have a name for a related effect: the big-fish-little-pond phenomenon. A student of equal ability reports higher academic confidence at a school of modest peers than at an elite school, because confidence is calibrated against the pond, not the fish. Economists have found the equivalent in pay. In a 2008 study of British workers, a team including Andrew Oswald reported that a worker's rank within a comparison group of similar employees predicted job satisfaction better than the size of the wage itself. Two people can earn the same salary and feel entirely differently about it — one at the top of the group, one near the bottom.

Compensation consultants see the effect in retention data. A senior engineer who is the highest-paid member of a small team will often report more pay satisfaction than a director at a larger company who sits in the bottom quarter of the band, even when the director earns twice as much. People leave for rank as often as for dollars, which is one reason titles carry so much weight in job offers: a title is a rank signal, and rank is a salary emotion.

The rank effect shows up in career decisions as well. A data analyst in Cleveland turned down a job that paid 14 percent more than her current salary because the offer letter showed the new range, and the figure sat at the bottom of it. Eighteen months later she still calls it the right call: her employer pays her less in dollars and keeps her at the top of its band, and she says she checks the rank before she checks the total.

The same logic governs the counterfactuals people run in their heads. In a 1995 study, psychologists who analyzed the reactions of Olympic medalists found that bronze medalists appeared happier than silver medalists — the silver medalist imagines gold, the bronze medalist imagines fourth place. Pay works the same way. A bonus of $8,000 feels large or small depending on whether the person next to you received $4,000 or $20,000, not on what the $8,000 buys.

How a raise actually lands

The research on adaptation explains the rest of the feeling. In a 1978 study, researchers interviewed lottery winners and a control group and found that a year after the windfall, the winners were not measurably happier than people who had not won. The pleasure of a raise, like the pleasure of a lottery ticket, fades faster than people expect — which is why a raise that is merely adequate can feel like nothing within a quarter.

What a raise feels like on arrival depends on four things, and the size of the increase is only one of them. The first is the gap between the number and the expectation: a raise that beats what you asked for lands differently from one that matches it, and a surprise raise is remembered for years. The second is the comparison group — the raise is read against what peers received in the same cycle. The third is rank: a raise that moves you from the bottom of the band to the middle feels bigger than one that leaves your standing unchanged, even at a higher dollar value. The fourth is the process: studies of workplace fairness find that how the number is explained — whether criteria were clear, whether the manager could justify it — moves satisfaction as much as the size of the increase.

What shapes how a raise feels, in rough order of influence

Factor	Why it matters
Comparison with peers	The reference point for almost every pay reaction
Distance from expectation	Beating the number you asked for changes the feeling
Rank within the group	Same dollars, different standing, different reaction
Fairness of the process	A clear explanation moves satisfaction as much as the raise
Absolute size	Matters most at the low end of the income scale

The table is a rough composite of the research, not a formula, but it explains a familiar office phenomenon: the same 6 percent can be a triumph for one employee and a grievance for another. The difference is rarely in the arithmetic.

The problem with pay secrecy

None of this would matter so much if people could check their comparisons. They cannot. Pay is the one number in adult life that is both central and hidden, and the hiding does not reduce the comparing — it makes the comparisons worse. In a 2022 study, economists at Harvard and UCLA asked employees at a large Asian bank to guess what their peers and managers earned. The guesses were off by wide margins, and the errors were not random: employees systematically underestimated what their managers made and overestimated how their own pay compared with the group.

The consequences of those misreadings show up in the field. In related work, the same researchers conducted an experiment in which employees were shown their own pay compared with that of their peers. Those who learned they earned below the median of their group became measurably less productive and more likely to quit; those above the median did not change their behavior at all. The information itself moved people — not a new offer, not a manager's comment, just a spreadsheet row.

Companies understand this, which is why pay secrecy is usually deliberate. A few years ago, an employee at a large Seattle-area software company compiled a spreadsheet of thousands of internal salaries and circulated it internally; the file spread within a day. Managers spent weeks fielding questions about gaps they had never explained, and a share of the employees who discovered they were below their peers began interviewing. The company's response was not to narrow the gaps but to tighten the secrecy — a reminder that for many employers, the hiding is the policy. Federal labor law protects most private-sector workers who discuss pay with colleagues, and SHRM surveys find that about a fifth of employers still maintain formal pay-secrecy rules, with many more discouraging the topic informally. The secrecy is a policy choice, not a legal default.

When a peer earns more

The clearest evidence that pay comparisons drive behavior comes from the experiments described above and from the broader research on internal equity. Studies of pay dispersion within teams find that the wider the gap between the highest- and lowest-paid members of a group doing similar work, the more turnover on the low side. The gap does not have to be large; it has to be visible.

Behavioral economists have documented the underlying impulse in the laboratory. In the ultimatum game, one player proposes a split of a sum of money and the other accepts or rejects it; if the second player rejects, neither gets anything. People reject offers they consider unfair even when rejection costs them real money. The same impulse operates in offices, where a smaller raise that is seen as fair can be accepted calmly and a larger one that is seen as arbitrary can be resented for a year.

"People don't experience a salary as a number. They experience it as a comparison — to their peers, to their past, to what they were promised. Every pay conversation is a conversation about standing, with arithmetic attached."

The consultant who said that has advised large employers for two decades and asked not to be identified. He keeps a file of what he calls the leak stories — spreadsheets, shared offer letters, a manager who said too much in a meeting — and he argues that the damage is never the leak itself. It is the silence that follows it, when a company confirms the number but refuses to explain the logic. "You can defend any gap," he said, "if you can tell the story behind it. Most managers can't, because they were never told the story either."

What this means at the offer table

For a job seeker, the research translates into a handful of practical moves. The first is to control the anchor. The tendency for the first number in a conversation to pull the final one — documented by the psychologists Daniel Kahneman and Amos Tversky in the 1970s and replicated hundreds of times since — is among the most reliable findings in behavioral economics. In salary conversations, the person who states a number first sets the anchor. A candidate who answers the range question with a researched figure, rather than with a guess or a deflection, moves the negotiation in a measurable way.

The second move is to replace gossip with data before the conversation starts. The peer comparisons that feel so urgent are usually built on guesses, and the research shows the guesses are wrong in predictable directions. Payscale and Glassdoor publish ranges by role, city, and experience level, and a few minutes with a salary calculator converts any offer into an hourly figure and an annual figure, or a real raise after inflation with the inflation adjustment tool. None of that replaces the research — but it gives the comparison an accurate number to work with.

The third move is to ask where the offer sits within the band, not just what it is. A figure at the top of a published range signals something different from one at the bottom, and the rank predicts how the number will feel in two years, when the adaptation has done its work. The same logic applies to raises: ask for a specific number before review season, so that the eventual figure has an anchor to beat, and ask how it was set. A manager who can explain the number has done half the work of making it land.

The research, in short, suggests three questions before any pay decision:

• What is the range, and where does this number sit in it?
• What will I compare this against a year from now?
• Can the person offering it explain how it was set?

The comparison you control

The reference group is not fixed. People choose their comparisons — the college roommate, the neighbor, the coworker in the next pod — and the choice shapes the feeling more than the paycheck does. The consultant's advice to clients who feel underpaid is to pick the reference class deliberately: compare against market data for your role, your city, and your experience, not against the one friend who got lucky or the colleague whose number you only think you know. The comparison you inherit from your employer — the band, the peer group, the title ladder — is a choice someone else made; the comparison you apply to your own decisions can be yours.

There is a second, less comfortable implication. If satisfaction is relative, then chasing a number is a treadmill by design: every raise resets the comparison, and the peer who was behind you last year is now the reference point ahead. The research does not say the chase is pointless — money buys real security, and the future salary calculator shows how raises compound across a career. It says the chase should be aimed at a number you have chosen, against a comparison you have chosen, rather than at whatever the person next to you happens to make.

What to watch

The direction of the next few years is toward more comparison, not less. Pay transparency laws now require published ranges in more than a dozen states, and job boards surface ranges by default, which means the reference points that used to be hidden are becoming public. Early research on transparency finds that it compresses pay gaps and, for people who learn they are paid fairly, raises satisfaction — while making the unfairness, where it exists, impossible to ignore.

Employers are responding by managing the comparisons deliberately: explaining bands, coaching managers on how to deliver numbers, publishing ranges internally before employees find them elsewhere. Workers, in turn, have more information to negotiate against reality than any previous generation. The habit will not change — people will always compare — but the comparison is becoming a conversation with data in it.

The product manager in Austin did not wait for the market to catch up. She left in June for a job that paid about 22 percent more, at a company that published its ranges and could explain its numbers. The raise she had received in January had been objectively good. It had failed the only test that mattered to her, and the research says the test was not the arithmetic. The number matters. The comparison matters more, and it always has.