Guide · Comp policy

Pay compression: How to spot it early

Most pay compression is visible in payroll a year before anyone complains. Watch new-hire pay against tenured pay every quarter.

Short answer: Pay compression is when pay for newer or more junior people creeps up to pay for experienced people in the same or a higher job. Inversion is when it passes them. Both build quietly, because offers follow the market while raises follow a budget. The way to catch them early is to compare new-hire and tenured pay by job and level each quarter, not once a year.

The question in the one-to-one

Picture a 240-person company in Phoenix that makes scheduling software for dental practices. One of its senior data engineers has been there four years. In her one-to-one she asks a direct question: Is it true that the engineer who joined her team last month is paid more than she is? He is one level below her.

Her manager does not know. He asks the head of people, who pulls the numbers that afternoon. Of the fourteen engineers at the senior level, the five hired in the past year are paid on average about 11% more than the six hired three or more years ago. Two people at the level below now earn more than the leads they report to. Every one of those offers was inside the band. Every one was approved.

Nobody decided to pay the newest people the most. The market for data engineers moved faster than the company's raises did. Recruiters offered what it took to close. Annual raises ran at about 3.5%. Each offer made sense on its own. Together they turned the pay order upside down in one team in about eighteen months.

The finance lead asks the obvious question: How many other teams look like this? Nobody can say without a week of work. The engineer has not resigned. But she has asked and she will want an answer before the next review.

This is a composite of a common situation at companies this size, not a client. The details are illustrative. We see the pattern often because pay compression is almost never a decision. It is the sum of many reasonable offers that nobody compared.

Why pay compression is hard to catch

Two terms first. Pay compression is when the gap between people with different experience or levels narrows until it no longer reflects the difference in their work. Pay inversion is the extreme case, when the newer or more junior person is paid more. Both can happen between new hires and tenured staff in the same job, between adjacent levels or between a manager and the people they manage.

Two good goals pull against each other. The first is hiring at the market now. A company that will not pay what the market asks loses candidates, especially in roles where demand moves fast. The second is paying fairly the people who stayed. Their raises come from a budget set once a year, usually as a percentage of pay. The two move at different speeds. In August 2026, people who stayed in their jobs in the US saw median pay growth of 4.4% over a year, while people who changed jobs saw 7.3%, in one large payroll data set [1]. That gap is the engine of compression. A company that hires steadily in a moving market buys new people at the new price and raises existing people at the old pace.

The picture is uneven, which is what makes it hard to see. In one compensation software provider's 2026 data on more than 10 million US employees, no job family as a whole paid new hires more than tenured staff in the same role. Yet some individual roles did, by as much as 19% [2]. Averages across a company can look fine while one team is inverted.

It is also hard to see from the inside. Employees often find out first. In a UK survey fielded in summer 2026, a third of HR leaders said the pay gap between early-career staff and entry-level hires had narrowed over the past two years. Nearly half expected it to narrow further [3]. Among employees who had learned what new starters earned, 16% found the new starters were paid more than they were [3].

Finally, not every narrow gap is a problem. A new hire may bring skills the role now needs that a tenured colleague does not have. A long-serving person may be paid low in the range because their work is at the low end of the level. Compression is a fault only when the gap no longer matches the work. Telling the two apart is the judgment this guide is about.

Four patterns that look alike

Compression shows up in a few common shapes. Each tends to have a different cause and each needs a different answer. This table is illustrative, not a diagnostic test.

What you seeWhat it often meansWhat it usually needs
Recent hires paid above tenured staff in the same job and levelThe market for that role moved faster than the company's raisesA market review of the role, not one-off raises for whoever asks
People at one level paid close to or above the level aboveOffers placed high in the band or levels set loosely at hireA look at how people were leveled and how offers were placed
A manager paid close to or below someone they manageA specialist market running ahead of the manager band, or a held-back raiseA deliberate decision, recorded, rather than an accident left in place
Long-tenured staff low in their ranges across many jobsRaise budgets below market movement for several yearsA budget and band question for leadership, not a team-by-team fix

The first two rows are the most common in growing companies. The last row is the slowest to build and the most expensive to fix.

Five traps in managing pay compression

None of these comes from carelessness. Each is a sensible default that fails when hiring is steady and the market moves. The cost shows up a year or two later, as resignations, a hard conversation or a pay gap the company cannot explain.

"We'll hear about it if it's a problem"

What it looks like: Nobody checks new-hire pay against tenured pay. The company learns about compression when someone asks, as the engineer in Phoenix did, or when someone resigns and mentions it on the way out.

Why smart teams fall into it: Pay is private in most companies and checking it takes effort. If nobody is complaining, it feels safe to assume the pay order is roughly right. A small HR team has other fires.

What it costs: The people who ask are not a fair sample. Many people who find out say nothing and start looking. In the UK survey, most employees who had learned what new starters earned said they felt undervalued, frustrated or disappointed. One in six said they had accepted a job offer elsewhere and a third had asked for a raise [3]. The ones who leave are often the ones with the best outside options. Waiting also makes the fix bigger, because every quarter of hiring adds new cases. And it is getting harder to rely on secrecy. Posted pay ranges and new pay transparency rules mean staff can often see what a new hire at their level is offered. Compression that was private is becoming visible.

"Every offer was inside the band"

What it looks like: Offers are checked against the band and nothing else. If an offer sits between the minimum and maximum, it goes out. Nobody compares it with the people already in that job.

Why smart teams fall into it: The band is the rule the company wrote, so staying inside it feels like control. Checking each offer against current staff takes time a recruiter does not have.

What it costs: Bands are wide on purpose, wide enough that two people doing the same work can sit far apart in one. An offer can sit high in the band while a tenured colleague doing the same work sits low in it and both are "inside the band". The band is a limit, not a check on fairness between people. The check that matters is where the offer lands against the peers it will sit beside. That check belongs in offer approval, which our guide on who approves a pay exception covers. If most offers for a role land high in the band, the band itself may be out of date, which our guide on when to update salary bands deals with.

"We'll fix it in the annual review"

What it looks like: Compression is noticed in the spring and parked until the pay review in the autumn. The plan is to give the affected people a larger merit raise.

Why smart teams fall into it: The review is where pay changes happen, with a budget and a process. A raise outside the cycle feels like an exception that others will copy.

What it costs: Merit money is the wrong size and the wrong kind. A gap of 10% or more cannot be closed by a merit raise in a cycle where the budget is a few percent. Closing it from merit takes money from everyone else, as our guide on splitting the salary budget explains. It also mixes two messages. A merit raise says "you performed well". A compression fix says "we let your pay fall behind the market". People who get a correction labeled as merit may read it as a reward that could be withdrawn. Meanwhile the six months of waiting are the months when people find out and start looking.

"Tenure should always pay more"

What it looks like: Any case of a new hire paid above a tenured colleague is treated as an error. The fix is to lift everyone with longer service above every newer person in the same job.

Why smart teams fall into it: It feels fair and it is easy to explain. Loyalty should count for something. Inversion is uncomfortable to defend in a one-to-one.

What it costs: Time in a job is not the same as the size of the work. A new hire may bring a skill the role now needs that a long-serving colleague does not. A tenured person may be doing the job at the low end of the level. Lifting everyone by service pays for time rather than work, which is expensive and builds the next problem: The new hires who were paid correctly now look underpaid by comparison. It can also create a legal weakness. Pay differences should rest on factors the company can explain from the work, the person's qualifications or the market for the role. A blanket rule that ignores those makes each gap harder to defend. The aim is not to restore a tidy order by service. It is to make sure each gap has a reason the company would write down.

"The averages look fine"

What it looks like: The company checks compression once, at the level of the whole workforce or a whole department. The average new hire is paid about the same as the average tenured employee. The check is closed.

Why smart teams fall into it: Company-wide figures are what the HR system reports easily. A clean average feels like an answer.

What it costs: Compression lives in specific jobs and levels, not in averages. The provider data above shows the pattern at national scale: Tenure is paid more in every job family overall, yet some single roles pay new hires far more [2]. The same happens inside a company. A hot team can be inverted while a quiet team is fine and the average shows neither. The average can also hide who is affected. If recent hires and tenured staff in a role differ by sex, compression becomes a pay gap between men and women doing the same work. In the US, market value counts as a reason for a pay difference only where the employer can show it actually assessed the market value of the person's job-related qualifications. Prior salary alone does not justify a disparity [4]. In the EU, employers that report a gender pay gap of 5% or more in a category of workers, without a reason based on objective, gender-neutral criteria, must carry out a joint pay assessment [5]. Our guide on which pay gaps you can defend covers how such reasons are judged.

What catching compression early looks like

A company that catches compression early does not avoid it. Markets move and some inversion is the price of hiring well. What it avoids is finding out by accident. A few qualities mark that kind of company.

A few numbers are watched every quarter, by job and level. The most useful ones compare where recent hires sit in the band with where tenured staff in the same job sit, how many cases of inversion exist and whether any manager is paid close to or below a direct report. They are read by job and level, not by department. A one-off gap of a few percent may be noise. A gap that widens two quarters in a row is a signal.

Each offer is compared with its future peers before it goes out. Not just "inside the band" but "where does this land against the people already doing this job?" When an offer will land above tenured staff doing the same work, someone outside the hiring team sees it and decides on purpose.

Every inversion that stays has a written reason. Some gaps are right: A scarce skill, a larger role, a person still growing into the level. Those are recorded with the reason and a date to look again. Gaps without a reason go on the list to fix.

Fixes have their own money. Compression corrections come from a line set aside for market and equity adjustments, not from the merit pool. They are labeled as corrections, so the people who get them know what they are.

The cause gets fixed, not only the cases. If the cause is a band that trails the market, the band moves. If it is offers placed high by default, offer placement changes. If it is raise budgets below market movement, that becomes a leadership conversation about the budget. Fixing the people without the cause means doing it again next year.

Someone owns it. One named person reads the numbers each quarter and brings the cases to whoever decides pay. In a small company that may be the head of people or a part-time compensation lead.

A useful sign is how the company would answer the Phoenix engineer's question. If someone can say within a day whether her case is a known, reasoned gap or a fault with a date to fix it, the company is watching. If it takes a week to find out, it is not.

The calls only you can make

Good monitoring tells you where compression is. Some choices about what to do with that remain the company's own, because they are about what it values.

Does tenure earn anything on its own? Some companies want long service to show in pay. Others pay only for the work and the market. Both are defensible. The answer decides which gaps count as faults.

How much inversion will you accept to hire? In a hot role, refusing any offer above tenured staff may mean not hiring at all. Accepting every one means a fix later. The balance depends on how critical the role is and how fast its market moves.

Fix the people or move the band? Targeted adjustments cost less now. Moving the band costs more but stops the next case. The choice is about cost now against cost later.

How fast do you close a gap? All at once, or over two or three steps? A single step is clearer and costs more this year. Several steps spread the cost but leave the gap in place while people wait.

What do you tell people? Some companies explain a compression fix openly as a market correction. Others make the change quietly. The answer says how much the company trusts staff with the reasons behind their pay.

The data, the law and the numbers

  • Pay growth for stayers and changers. ADP's National Employment Report for August 2026, published 2 September 2026, reported median year-over-year gross pay growth of 4.4% for US job-stayers and 7.3% for job-changers. The base pay figures were 3.0% and 4.7% [1]. These are national figures from payroll data, not pay for matched jobs.
  • New-hire against tenured pay. Payscale's 2026 Flight Risk Report, published 16 June 2026 and based on HRIS data for 10.2 million employees at 4,500 organizations effective May 2026, found no job family where new hires as a group out-earned tenured staff in the same role. In roles where new hires did earn more, they earned 3.6% more on average. In roles favoring tenure, tenured staff earned 6.1% more. For one senior software engineering role, new hires earned 19% more [2]. New hires are people hired within the past year. These are US national medians across employers, not comparisons within one company.
  • What employees and HR leaders see. A Censuswide survey for IRIS Software Group of 511 UK senior HR professionals and 500 UK employees with two to five years of full-time work, fielded 30 July to 5 August 2026, found that 32% of HR leaders said the gap between early-career staff and entry-level hires had narrowed over the past two years and 46% expected bands to narrow further. Of employees who had learned what new starters earned, 16% said new starters were paid more than they were. Only 20% of HR leaders planned targeted salary adjustments [3]. IRIS sells HR and payroll software. These are self-reported survey figures.
  • United States. Under the EEOC's compliance guidance, market value qualifies as a factor other than sex only if the employer proves it assessed the marketplace value of the job-related qualifications. Prior salary cannot, by itself, justify a compensation disparity [4].
  • European Union. Pay structures must allow work to be compared on objective, gender-neutral criteria including skills, effort, responsibility and working conditions (Article 4(4)). Employers must make the criteria for pay and pay progression easily accessible to workers (Article 6). Where pay reporting shows a gap of 5% or more in a category of workers that is not justified by objective, gender-neutral criteria and not remedied within six months, a joint pay assessment follows (Article 10) [5].

Where this work involves market data, we don't provide market data. We use the data you have or buy.

This is general information, not legal advice. Equal pay and pay transparency rules differ by country and state and change over time. Have an employment lawyer review how you document reasons for pay differences between people in the same job.

Our view

We think pay compression is best treated as something to watch, not something to discover. It is the predictable result of hiring at today's market and raising pay at last year's pace. Every growing company that hires steadily will have some.

The fault is not having it. The fault is not knowing where it is, why it is there and whether the company meant it. Compare new-hire and tenured pay by job and level every quarter. Check each offer against the people it will sit beside. Write down a reason for each gap you keep and fix the ones without one from their own budget line.

A company that does this still has gaps. It just has answers when someone asks.

Sources

  1. ADP, National Employment Report for August 2026 (2 September 2026), checked 2026-10-29
  2. Payscale, 2026 Flight Risk Report press release (16 June 2026), checked 2026-10-29
  3. IRIS Software Group, pay squeeze research (Censuswide survey of 511 UK HR leaders and 500 UK employees, 4 September 2026), checked 2026-10-29
  4. US EEOC, Compliance Manual Section 10: Compensation Discrimination, 10-IV.F.2.g (market factors), checked 2026-10-29
  5. Directive (EU) 2023/970 on pay transparency, Articles 4(4), 6 and 10 (EUR-Lex), checked 2026-10-28

Merit, promotion or market raise?

A merit raise, a promotion and a market fix answer different questions. Why one label on all three hides what you pay for and what it costs later.

Compa-ratio or range penetration?

The same pay can be a 0.90 compa-ratio and sit below the band minimum. What each measure answers and where each one misleads.

What does a fractional comp lead do?

A pay structure starts to drift the day it is delivered. What a fractional comp lead does each week, quarter and year and when you do not need one.

See Resources for every guide and the self-check.