Across-the-board raises or merit pay?
Equal raises feel fair and keep the peace. They also pay off the wrong people.
Short answer: Across-the-board raises feel fair, but an equal percentage is not equal treatment. It gives the most money to people already paid the most, gives strong and weak performers the same and leaves underpaid people where they are. Most companies should differentiate, but only as far as their ratings and pay data can bear. Where they cannot, fix those first.
Everyone got 3.5% and the wrong people stayed
Picture a 190-person company in Omaha that sells freight audit software to trucking and logistics firms. Last year it gave across-the-board raises: 3.5% for everyone, the full budget, in one line on every letter. The CEO liked it. The review took two weeks instead of eight. Nobody complained to HR and managers were spared hard conversations in a year when the team had pulled together through a difficult launch.
Nine months later the picture is less comfortable. Two of the company's four strongest engineers have left for competitors. In exit interviews both said the same thing in different words: The raise told them their work counted for no more than anyone else's. One of the people who stayed has been on a performance plan since March and received the same raise as the engineer who rebuilt the billing engine.
The finance lead has found something else. The equal raise put $6,300 into the pay of a director on $180,000 and $2,100 into the pay of a support analyst on $60,000. The analyst was already near the bottom of her range. The director was above the top of his. A raise designed to be fair to everyone widened the dollar gap between them by $4,200 and moved neither of them closer to where the company says each should be paid.
Now the budget for next year is 3.5% again and the CEO is asking whether to do the same thing. The managers who liked the easy cycle would vote yes. The people who left would have voted no.
This is a composite of a common situation at companies this size, not a client. The details and figures are illustrative.
Why the choice between equal and merit raises is hard
Two terms first. An across-the-board raise, sometimes called a "peanut butter" raise, gives every eligible employee the same percentage increase. A differentiated raise varies the increase by person, usually by performance and position in the pay range.
Equal raises are not a lazy choice. They answer real problems. They are fast to run. They need no ratings, no calibration and no manager judgment, so they cannot be biased by any of those. They avoid the hardest conversation in a pay review, which is telling someone they get less than a colleague. In a year when a team has pulled together, they send a message of shared effort. They are common in organizations with many hourly employees, where pay often follows a step or a schedule anyway [2].
Differentiation answers different problems. A company that pays everyone the same increase is telling its strongest people that their extra contribution is not worth extra money. It also loses its main yearly chance to move underpaid people toward their range and to slow the growth of pay for people already above it. A compensation data provider's 2026 survey found that a quarter of organizations name perceptions of unfair pay as a leading reason they are losing talent [1].
The survey numbers on how common equal raises are point in different directions. One provider reports that 36% of organizations gave across-the-board increases in 2026 and 32% plan to in 2027 [1]. Another survey of 756 employers found that only 4% gave equal raises to everyone instead of merit increases in 2026 [3]. Both can be true. Many companies give a general increase to everyone and then a merit amount on top, which counts as an across-the-board raise in one survey and not in the other. The safer reading is that pure equal raises are rare, but a flat element inside the raise is common. Either way, a third large survey describes the direction of travel as more precise, performance-driven pay aimed at key roles and skills [4].
The real difficulty is that differentiation only works if the inputs are sound. A merit raise is only as fair as the rating behind it and only as useful as the pay data that says who is underpaid. Many companies at 50 to 500 people are not sure of either. That doubt, more than any belief in equality, is often what keeps them on equal raises.
What each approach buys and what it costs
This table compares the choices companies usually weigh. It shows the trade-offs, not how to set a budget.
| Approach | What it buys | What it costs |
|---|---|---|
| Equal raise for everyone | Speed, simplicity and no rating risk | Strong performers and underpaid people fall behind; gaps grow in dollars |
| Equal general increase plus merit on top | A shared message and some room to reward | The merit part is often too small to mean anything |
| Differentiated by performance only | A clear link between contribution and pay | Ignores who is underpaid; carries any bias in the ratings |
| Differentiated by performance and position in range | Rewards contribution and moves pay toward where it should be | Needs sound ratings, sound ranges and harder conversations |
Five traps with across-the-board raises
"An equal raise is the fair raise"
What it looks like: The company gives everyone the same percentage and describes it as treating everyone alike. Leadership believes it is the neutral choice that nobody can object to.
Why smart teams fall into it: The same number on every letter looks like equality. Percentages feel like a common unit, so nobody stops to look at the dollars.
What it costs: An equal percentage is not an equal amount. It gives the most money to the people already paid the most and widens every dollar gap in the company every year. A company with a pay gap between groups, by gender or by job family, will see that gap grow in dollars under equal raises even though nobody chose it. Equal raises also freeze each person's position in their range. Someone hired low stays low. Someone paid above the top of the range moves further above it. The raise feels neutral, but the result is a decision to keep today's pay order, including whatever is wrong with it.
"The budget is too small to differentiate"
What it looks like: With a budget of 3% to 3.5%, leadership decides that splitting it up is pointless. A strong performer might get 4.5% instead of 3.5%. That does not seem worth the effort or the hard conversations.
Why smart teams fall into it: The spread a small budget allows feels too thin to motivate anyone. Differentiating with little money also means some people get very little or nothing, which nobody wants to say out loud.
What it costs: A small budget is the strongest reason to differentiate, not the weakest. When there is little money, giving it evenly means every dollar is spread across people who are leaving anyway, people who are already overpaid and people who are underpaid and still at risk. Real differentiation with a small budget comes from giving some people little or nothing, not from giving top performers slightly more. Some sectors do this openly. A 2025 to 2026 planning survey of 563 organizations found that in software and consulting services, nearly 55% of companies give high performers 150% to 300% of the average increase [5]. Our note on what a 3.2% merit budget really means looks at why the average hides who should get 0% and who should get 8%.
"Our ratings aren't good enough to pay on"
What it looks like: Leadership knows the ratings are inflated, inconsistent between managers or both. Most people are rated "meets" or above. Rather than pay on ratings it does not trust, the company gives everyone the same raise.
Why smart teams fall into it: It is an honest judgment and often a correct one. Paying on bad ratings is worse than not paying on ratings at all.
What it costs: The trap is treating this as a reason to stay on equal raises for good, rather than a problem to fix. The doubts are well founded. A 2025 field study at a multinational financial services firm found consistent race gaps in manager ratings and that women of color ended up with the lowest final ratings. When self-ratings were hidden from managers, the gaps barely changed, which suggests managers anchored on the previous year's ratings instead [6]. That is why ratings need checking before they drive pay. But a company that never fixes its ratings never earns the right to differentiate. Every year on equal raises costs it the people it most wants to keep. The usual answer is to differentiate modestly while the ratings improve and to check the outcomes by group before letters go out. Our guide on merit matrices and manager judgment covers how each one turns ratings into raises.
"Differentiate on performance and the rest takes care of itself"
What it looks like: The company moves off equal raises and pays only on rating. A strong performer gets 5%, a solid one 3% and a weak one 1%, wherever each one sits in their range.
Why smart teams fall into it: Pay for performance is the reason for leaving equal raises behind. Adding position in range feels like making the message muddier.
What it costs: A raise has two jobs: Rewarding contribution and moving pay toward where it should be. Performance alone does only the first. A strong performer already above the top of the range keeps moving further above it. A solid performer hired well below the midpoint stays underpaid and becomes the next resignation. Differentiating on rating alone can also make compression worse when strong new hires already earn close to strong long-tenured staff. Our guides on spotting pay compression early and splitting the salary budget cover the parts of the budget that merit alone should not carry.
"Call it a cost-of-living raise"
What it looks like: The company gives an equal raise and calls it a cost-of-living adjustment to make it easier to explain. Some companies keep the label for a merit amount they never actually differentiate.
Why smart teams fall into it: The label ends the conversation. Nobody argues with keeping up with prices and nobody asks why they got the same as a colleague.
What it costs: The label sets an expectation the company may not want to keep. Once a raise is called cost of living, people read it as an entitlement linked to prices rather than a decision about their pay. In a later year when the company wants to differentiate, a 1% raise for a weak performer reads as a pay cut in real terms and the company's own words make it harder to defend. The label also hides the real choice. If the company means to protect everyone's buying power, it should say so and decide whether it can afford to do that every year. If it means to reward contribution, the label gets in the way.
What a sound raise decision looks like
A sound pay review does not have to be fully differentiated. It has to be deliberate. At 50 to 500 people, a decision you can defend has these qualities.
The choice is made on purpose each year. Leadership decides whether to differentiate and how far, with a reason in plain words. Equal raises are a valid choice in some years, but they are chosen, not inherited from last year.
The spread matches what the company can stand behind. A company with ratings it trusts and ranges it believes in can differentiate widely. A company still fixing either differentiates less and says why. Nobody pretends to differentiate with a spread so thin that a strong performer cannot tell.
Position in range counts as well as performance. Underpaid people who are doing the job well move toward their range faster. People far above the top of the range get less in base pay, or a lump sum instead, with a clear reason.
Low and zero raises are possible and explained. Some people get little or nothing in base pay, for a reason they can understand. Managers are prepared to explain the system, not only the number.
The outcome is checked before letters go out. Someone looks at how raises fall by gender, by race where the data allows, by job family and by manager. Where a group comes out lower, the company can explain it in terms of ratings and position in range, or it changes the result.
The rules are written and can be shared. The factors that decide a raise are written down and applied the same way to everyone. Under the EU rules, the criteria for pay progression must be objective, gender-neutral and easily accessible to workers [9]. In the US, a merit system can explain a pay difference only if it is a real system with predetermined criteria, communicated to employees and applied consistently [7] [8].
The label matches the purpose. A general increase is called a general increase. A merit increase is called a merit increase. Nothing is called cost of living unless the company means to keep it that way.
The calls only you can make
Whether pay should follow performance. Some companies want pay to follow contribution closely. Others want it to follow the job and the market, with recognition coming in other ways. Both are defensible. Only the leadership team can say which one is yours and that is a question for your compensation philosophy, not for this year's budget.
How hard a conversation you are willing to have. Real differentiation means some people hear that they got very little. Whether your managers can hold those conversations and whether you will back them when they do is a judgment about your culture.
When the ratings are good enough. No rating system is perfect. Deciding when yours is sound enough to pay on is your call. So is how much weight to put on it until then.
What to do in a high-inflation year. When prices rise fast, protecting everyone's buying power and rewarding contribution pull against each other. How to balance them is a values choice. So is whether to say so openly.
How much to share. The EU rules require progression criteria to be accessible to workers [9]. How much of the rest, such as the spread you used, to share with your people is a choice about trust.
The data and the law
- How common equal raises are. A compensation data provider's survey of 1,266 organizations, fielded in May and June 2026, found that 44% were adopting or considering across-the-board increases going into 2026, 36% gave them and 32% plan to in 2027. A quarter named perceptions of unfair pay as a leading reason they are losing talent [1]. The approach is common in organizations with many hourly employees [2].
- A different count. A survey of 756 employers published in May 2026 found that only 4% gave equal across-the-board increases instead of merit increases in 2026. Most use a mix of individual performance and position relative to market or peers [3].
- The direction of travel. A planning survey of more than 34,000 companies, 1,650 of them in the US, fielded from March to May 2026, describes employers moving to more precise, performance-driven pay aimed at key roles, skills and talent segments. US budgets are projected at 3.4% for 2027 [4].
- Bias in ratings. A field study published in 2025 found consistent race gaps in manager ratings at a multinational financial services firm and the lowest final ratings for women of color. Hiding self-ratings from managers did not remove the gaps [6].
- United States, merit as a defense. The Equal Pay Act allows a pay difference made under a merit system (29 U.S.C. 206(d)(1)) [8]. The EEOC's compliance guidance says a merit system is a defense only if it is based on predetermined criteria, communicated to employees and applied consistently and even-handedly to both sexes. A subjective merit system should be strictly scrutinized and the defense covers only as much of the gap as the system explains [7].
- European Union, progression criteria. The pay transparency directive requires employers to make the criteria used to set pay, pay levels and pay progression easily accessible to workers. Those criteria must be objective and gender-neutral (Article 6(1)) [9].
This is general information, not legal advice. Rules and their national versions differ. Check the rules where you employ people.
Our view
We think most companies should differentiate raises and that equal raises are usually a way of avoiding a decision rather than making one. An equal percentage looks neutral, but it widens dollar gaps, freezes underpaid people in place and tells the strongest people that their extra work is not worth extra money. With a small budget, that cost is larger, not smaller.
But we would not tell a company to differentiate on ratings it does not trust. The honest path is to differentiate as far as the ratings and the pay data can bear, check the outcome by group before anyone sees a letter and fix the inputs so the spread can widen next year. An equal raise can be the right call for one year. It should never be the default.
Where we help
Our pay review work helps you decide how far to differentiate this year, builds a merit guide that weighs performance and position in range and checks how raises fall by group and by manager before letters go out. The starting point is pay review cycle design, quoted within one business day. If a conversation would help first, Book a free call.
Sources
- Payscale, 2026-2027 Salary Budget Survey press release (4 August 2026; 1,266 submissions, May to June 2026), checked 2026-10-29
- CFO Dive, Fewer companies plan peanut butter pay increases in 2027 (4 August 2026), on Payscale's salary budget survey, checked 2026-10-29
- HR Dive, Peanut butter pay raises are not mainstream (8 May 2026), on Mercer's QuickPulse US Compensation Planning Survey of 756 employers, checked 2026-10-29
- HR Dive, Companies aim for more strategic pay increases (24 July 2026), on the WTW Salary Budget Planning Report, March to May 2026, checked 2026-10-29
- Pearl Meyer, 2026 compensation planning: Data and insights (November 2025; 2025 to 2026 Compensation Planning Survey, 563 organizations), checked 2026-10-25
- Bohnet, Hauser and Kristal, Can gender and race dynamics in performance appraisals be disrupted? The case of social influence, Journal of Economic Behavior and Organization 235 (2025), summary by Harvard Kennedy School, checked 2026-10-29
- US EEOC, Compliance Manual Section 10: Compensation Discrimination, 10-IV.F.1 (merit and incentive systems), checked 2026-10-29
- 29 U.S.C. 206(d)(1), Equal Pay Act (Cornell Legal Information Institute), checked 2026-10-29
- Directive (EU) 2023/970 on pay transparency, Article 6(1) (EUR-Lex), checked 2026-10-28