How should you split the salary budget?
Fund the fixes first, or merit quietly pays for them.
Short answer: Split the salary budget into separate lines before managers see a number: Structure and below-minimum fixes, pay equity corrections, promotions, market adjustments and merit. Size each line from its own evidence and give it an owner. When everything comes from one merit pool, the urgent fixes get paid first and merit becomes whatever is left, with nobody having decided that.
One number, five jobs
Picture a 180-person health software company in Boston, planning its January pay review. The CFO approves a salary budget of 3.5% of payroll, in line with what US employers are planning for 2027 [1]. Everyone calls it the merit budget.
Then the requests arrive. The people team has just moved the salary ranges up after buying new survey data, so eleven employees now sit below the minimum of their range. A pay equity check has flagged two women in customer success paid well below men in the same role. Fourteen people are up for promotion. The data team, hired in 2022, is now below market and two of its engineers have offers elsewhere. And the managers expect to give raises for performance, because that is what the budget is called.
No one decides how to divide the money. Each request is approved on its merits as it comes in. The range fixes go first, because people below a minimum look like an error. The promotions follow, because promotion letters are already drafted. The data team gets its market adjustment, because losing two engineers costs more than paying them. When the managers finally see their pools, the money left for merit works out to a little over 2%.
The strong performers who were not promoted, not flagged and not on the data team get the smallest raises in the company. Nobody chose that outcome. It is what a single pool does when it is asked to do five jobs.
This company is a composite of situations common at this size. It is not a client and the details are illustrative.
Why splitting the budget is hard
A salary budget is one decision on the finance side and several decisions on the people side. Finance needs a single number to plan cost. The pay review needs to pay for different things that have different causes and different owners.
Five kinds of money usually sit inside it. Merit rewards performance in the current job. Promotions pay for a bigger job. Market adjustments move pay that has fallen behind the market for a role. Structure adjustments move the ranges themselves. Someone then has to decide what happens to people who fall below a new minimum. Pay equity corrections fix gaps that cannot be explained by the work.
The tension is between simplicity and honesty. One pool is easy to approve, easy to explain to the board and easy to hand to managers. Separate lines take more work and invite arguments about each line's size. But one pool hides what the company is paying for. A raise of 6% might be strong merit, a promotion or a correction for an old mistake. In a single pool nobody can tell which.
The surveys reflect this blur. The headline US numbers for 2027 are a merit budget of about 3.2% inside a total of about 3.5% [1]. One of the largest US budget surveys defines its total as merit plus general increases plus "other" increases. It asks respondents to say what the "other" money is for, from range adjustments to compression to market fixes [2]. Promotion increases are not one of the named parts of that total. So a company that compares its "merit budget" with a survey figure may be comparing different things.
The size of the pot makes it worse. With 3% to 4% to spend, any one large need can take a visible share of the total. A pay equity fix for a few people or a market move for one team can take a large share of it. When that money comes out of merit, the cost lands on everyone else.
Five ways the budget split goes wrong
Most of these traps have the same shape. A real need gets paid from the merit pool because there is no other line to pay it from.
"It's all merit"
What it looks like: One budget is set and called merit. The company lets every kind of increase draw on it. Range fixes, promotions and market adjustments are approved one at a time. Merit is whatever remains.
Why smart teams fall into it: It is how the budget was approved. Finance gave one number, so the pay review spends one number. Splitting it feels like bureaucracy. Each individual request is reasonable on its own.
What it costs: Merit becomes a residual. Its size is set by how many other needs turned up that year, not by any decision about how much to reward performance. In a year with many fixes, strong performers in stable jobs get the least, as in the Boston company. They are also the people a company can least afford to lose quietly. The practice is common enough that one major budget survey asks employers directly whether they fund pay equity fixes out of the merit budget "even though the merit budget is not inflated" to cover them [2]. And because nobody wrote the split down, nobody can explain afterwards why merit was smaller than the number everyone was told.
"Promotions come out of the same pool"
What it looks like: Promotion increases are paid from the department's merit pool. A manager with two promotions on the team has much less left for everyone else.
Why smart teams fall into it: Promotions happen at the review, so paying them from the review budget feels natural. It also seems to keep promotions honest, since a manager who promotes pays for it.
What it costs: Promotions are large and uneven. A compensation software provider reported a median increase of 8.6% for promoted employees in the first half of 2026, with 8.6% of all employees promoted [3]. US employers expect to promote about 8.4% of their staff in 2027 [1]. Spread across a company, that is a meaningful slice of a 3.5% budget. Concentrated in one team, it can take most of the pool. The same provider notes the effect directly: When promoted employees get larger increases, the pool for everyone else compresses [3]. Managers learn the lesson fast. Some delay promotions to protect their team's raises. Others give promoted people a token increase so the rest of the team is not penalized, which leaves the promoted person paid below the new range. Either way, the timing and size of a promotion now depend on the budget arithmetic of one team, not on the job.
"Moving the ranges fixes the market problem"
What it looks like: The company updates its salary ranges to match new market data. The ranges move up. Pay does not, except for the people who now sit below a minimum.
Why smart teams fall into it: New ranges feel like the market fix. The structure is current, the job ads can show the new numbers and the work looks done.
What it costs: Moving a range moves the yardstick, not the pay. US employers plan to move their salary structures by about 2.6% on average for 2027 [1]. If ranges move by that much and a typical raise is a little over 3%, most people barely move within their range in a year. People hired at the bottom stay near the bottom. The company then faces two hidden costs. Employees who fall below the new minimum need money to reach it. That money has to come from somewhere. And if the merit guide gives larger raises to people low in their range, the move pushes many people down the range at once and the merit guide starts paying for the structure change. Range moves are a structure decision. They need their own line, sized from the number of people they affect, not a hope that merit will absorb them. Our guide on pay above the range maximum covers the opposite edge of the same problem.
"Market money goes to whoever asks"
What it looks like: Market adjustments are made case by case, usually after a resignation threat or a competing offer. The teams with the hottest skills and the most vocal managers get them.
Why smart teams fall into it: It feels efficient. Money goes where the retention risk is visible. The company does not pay to move people who were never going to leave.
What it costs: Case-by-case market money rewards the threat, not the gap. People who are equally far below market but less likely to complain stay behind. Over time that pattern can follow lines the company did not intend, for example by gender or by team. There is also a compounding problem with hiring. Among US employers, 36% report hiring at higher salary ranges and 32% are raising starting ranges, while 34% are increasing retention bonuses or spot awards for key people [4]. When new hires come in above people already in the job and only those who ask get moved, the people who stayed become the cheapest staff the company has. A market line works when it is sized from a review of who is actually below market, not from who has an offer this month.
"We'll fix pay equity if there is money left"
What it looks like: A pay equity review finds gaps. The fixes are approved in principle and then wait for the budget. In a tight year, they slip to the next cycle, or they are folded quietly into the affected people's merit raises.
Why smart teams fall into it: Equity fixes feel discretionary in a way that promotions and range fixes do not. Folding them into merit seems discreet and avoids a separate conversation.
What it costs: A gap the company knows about and defers is harder to defend than one it never found. In the EU, for employers that report under the pay transparency directive, a gender pay gap of 5% or more in any category of workers that the employer cannot justify and has not remedied within six months of reporting leads to a joint pay assessment with worker representatives [5]. The assessment must include measures to address the gaps [5]. Our guide on what triggers a joint pay assessment covers that process. Folding fixes into merit has its own cost. It makes the corrected person's raise look like a strong merit raise, which distorts the merit pattern the company will later review. And it hides the correction from the record, so the company cannot show it acted. A pay equity line, sized from the review's findings and spent before merit, avoids both problems.
What a sound budget split looks like
A sound split is not a set of fixed percentages. The right size of each line changes every year with the company's situation. In our view a sound split has these qualities.
- The lines exist before managers see a number. Merit, promotions, market, structure and pay equity are named and sized when the budget is approved, not discovered when the money runs out.
- Each line is sized from its own evidence. Structure money follows how many people the range move affects. Equity money follows the review's findings. Promotion money follows the expected number of promotions. Market money follows a review of who is below market. Merit is the only line set mainly by policy.
- Fixes are funded first. Below-minimum and pay equity corrections are paid before merit is set, so they cannot quietly eat into it. Merit is then a deliberate number, not a residual.
- Promotions are paid centrally or with their own pool. A team's raises do not depend on how many of its people were promoted.
- Each line has an owner. Someone is accountable for structure money, someone for equity, someone for promotions. Managers own merit within their pool.
- Some money is kept for the year. Promotions and market problems do not wait for the review. A company that spends everything in January ends up funding off-cycle increases from next year's merit.
- The split is reconciled after the cycle. Finance and the people team compare what each line was meant to pay for with what it actually paid for. Any line that ran over is investigated before the next budget is set.
- The split can be explained. If an employee asks why the merit raise was smaller this year, a manager can say what else the money paid for and why.
None of this is a template to copy. Each quality is a test a reviewer can put to last year's cycle.
A short test helps. Take three raises from the last cycle: The largest, a typical one and the smallest given to a strong performer. For each, ask which line paid for it and who approved that line. If the answer to either question is "the merit pool" or "nobody", the split did not exist in practice.
The calls only you can make
Dividing the money well depends on priorities that no survey can set for you.
How small merit can get in a year of big fixes. A year with a large range move or a big equity correction can leave little for merit if the total is fixed. Some companies accept a thin merit year to fix the structure once. Others protect a merit floor and spread the fixes over two cycles. Either is defensible if it is chosen.
Whether the total grows to pay for fixes. The alternative to a thin merit year is a bigger total. That is a finance decision about how much a cleaner structure is worth this year.
How fast to close equity gaps. Closing every gap in one cycle is expensive and sends a clear message. Phasing the fixes is cheaper and leaves gaps open longer. Where EU reporting applies, the law narrows this choice [5].
Whether market money goes to roles or to people. Moving everyone in a role that has fallen behind treats people alike. Moving only those at risk is cheaper and rewards the people who ask.
Who controls promotion money. A central pool keeps promotions consistent across teams. Department pools give leaders more control and more room for drift.
How much of the split to share. Telling employees what the budget paid for builds trust in a thin merit year. It also invites questions about each line. Our guide on merit matrices and manager judgment covers what managers need to explain the merit part.
The numbers
- Total and merit. US employers plan merit increases of about 3.2% and total salary increase budgets of about 3.5% for 2027, according to a July 2026 survey of about 1,000 organizations [1]. Another survey of 1,266 organizations, fielded May to June 2026, also puts the 2027 total at 3.5% [6].
- Structure moves. Employers in the same July 2026 survey plan to move salary structures by about 2.6% on average for 2027 [1].
- Promotions. US employers expect to promote about 8.4% of staff in 2027 [1]. A compensation software provider reported a median promotion increase of 8.6% and a promotion rate of 8.6% in the first half of 2026 [3].
- Overall budgets beyond merit. A September 2025 survey of 243 technology and life sciences companies found a median US merit budget of 3.5% and a median overall budget of 5% of base payroll for 2026. The overall budget includes money for promotions, pay equity fixes and market adjustments [7].
- What the totals include. One of the largest US budget surveys defines its total as merit, general and "other" increases. It asks what the "other" money is for, from range adjustments and compression to market fixes. It also asks whether pay equity fixes are budgeted separately or taken from merit [2].
- Targeted pay. In a 2026 survey of 1,650 US organizations, 36% were hiring at higher salary ranges, 32% were raising starting ranges and 34% were increasing retention bonuses or spot awards for key employees [4].
- EU law. Under the EU pay transparency directive, an unjustified gender pay gap of 5% or more in a category of workers, not remedied within six months of reporting, leads to a joint pay assessment that must include measures to address it [5]. This is general information, not legal advice.
Survey figures are averages across many companies. Overall budgets at technology companies tend to run higher than all-industry totals. Early projections often drift down by the time cycles run [7].
Our view
We think the budget should be split before anyone spends it. The fixes should be funded first. Range moves, below-minimum corrections and pay equity gaps are debts the company already owes. If they come out of merit, merit stops being a decision about performance and becomes whatever the fixes leave behind.
That does not mean merit always loses. It means the size of merit should be chosen. A company may decide on a thin merit year to clean up its structure, or it may decide to grow the total. Both are sound if someone makes the call and can explain it.
We would argue hardest against paying promotions and equity fixes from team merit pools. Both create the same distortion: The timing of a promotion or a correction starts to depend on a team's budget. The people who stayed and performed pay for it.
Where we help
Our pay review work sizes each part of the budget from your own data, sets the order in which fixes and merit are funded and builds the reconciliation that runs after the cycle. We use the market data you have or buy. The starting point is pay review cycle design, quoted within one business day. If a conversation would help first, Book a free call.
Sources
- Mercer, Results of the July 2026 QuickPulse US Compensation Planning Survey (31 August 2026), checked 2026-10-25
- WorldatWork, Salary Budget Survey 2026-2027 US questionnaire and top-level results (15 July 2026; 1,799 organizations), checked 2026-10-25
- Pave, merit increase data for the first half of 2026 (26 June 2026), checked 2026-10-25
- HR Dive, Companies aim for more strategic pay increases (24 July 2026), on the WTW Salary Budget Planning Report, March to May 2026, checked 2026-10-25
- Directive (EU) 2023/970 on pay transparency, Articles 9 and 10 (EUR-Lex), checked 2026-10-25
- Payscale, 2026-2027 Salary Budget Survey press release (4 August 2026), checked 2026-10-25
- Pave, How big of a raise can you expect in 2026? (27 October 2025; survey with Alpine Rewards of 243 companies, September 2025), checked 2026-10-25