The Targeted Raise: Flat Pay Budgets, the Retreat From Across-the-Board Increases, and How Leaders Can Differentiate Pay Without Breaking Trust
Download the article here:
Listen to a review of this article:
Abstract: As organizations finalize 2027 compensation plans, U.S. salary increase budgets have settled at roughly 3.5% for a second consecutive year, while employers are moving away from uniform, across-the-board raises toward targeted increases for top performers, scarce skills such as artificial intelligence, critical roles, and retention risks. This article examines that shift through the lens of research on pay dispersion, pay-for-performance, organizational justice, and pay transparency. It argues that differentiation can improve the performance return on a flat budget, but only when it rests on credible performance measurement, explicit criteria, routine pay equity audits, and managers who can explain decisions. Drawing on 2026 survey evidence and documented practices at organizations in technology, retail, healthcare, financial services, and higher education, the article offers five evidence-based responses and three long-term capabilities for leaders who want to reward contribution without eroding the trust, fairness, and legal compliance that expanding pay transparency laws now make visible.
Every autumn, compensation teams sit down with a number. For the 2027 planning season, that number is remarkably familiar. The Conference Board reports a median U.S. salary increase budget of 3.5% for 2027, unchanged from 2026 (The Conference Board, 2026). Mercer's July 2026 pulse of 1,001 U.S. organizations projects the same 3.5% total increase, with merit accounting for 3.2% (Mercer, 2026). WTW expects 3.4% after an actual 3.5% in 2026 (McLean, 2026b), and WorldatWork's survey puts the 2027 median at 3.5% as well (WorldatWork, 2026). By the standards of the post-pandemic years, when actual U.S. increases peaked above 4% in 2023 (McLean, 2026b), this is a flat budget.
What has changed is not the size of the pot but the way organizations intend to divide it. Payscale's 2026–27 survey found that 32% of U.S. companies plan across-the-board raises for 2027, down from 36% that gave them in 2026 (McLean, 2026a). The report described uniform increases as potentially demotivating because everyone receives the same raise regardless of performance. WTW's rewards data lead summarized the shift plainly: employers are moving away from broad-based increases toward pay that targets key roles, skills, and talent segments (Mayer, 2026). The Conference Board's analysts were even more direct, arguing that stable budgets call for greater differentiation rather than uniform allocation, and identifying AI skills as the leading skill category driving base-pay adjustments, cited by 37.8% of organizations (The Conference Board, 2026).
On paper, the logic is compelling. If the budget will not grow, the only way to improve its return is to spend it more deliberately. In practice, differentiation is where compensation programs most often lose the confidence of the people they are meant to motivate. A 3.5% budget that gives a top performer 6% necessarily gives someone else 2% or nothing, and in a year when private-sector real wages and salaries fell 0.4% (Bureau of Labor Statistics, 2026), a 2% raise is a pay cut by another name. Employees will read those outcomes through the lens of fairness, and increasingly they will have the data to compare. Connecticut's expanded pay transparency requirements, effective October 1, 2026, require good-faith wage ranges in both internal and external job postings (Fisher Phillips, 2026), joining a growing list of U.S. jurisdictions and the European Union's Pay Transparency Directive, whose transposition deadline passed in June 2026 (European Parliament and Council of the European Union, 2023).
In my consulting work with organizations during compensation season, I hear the same tension expressed from two directions. Executives say they cannot afford to keep paying everyone the same increase while their best people leave for 15% raises elsewhere. Employees and many frontline managers say the moment pay becomes differentiated, it starts to feel political. Both are right. The research literature on pay dispersion, merit pay, and organizational justice explains why, and it also points to a set of design choices that let leaders differentiate without breaking trust.
This article examines the 2027 shift toward targeted pay, what the evidence says about its likely consequences for organizations and individuals, and five practical responses that leaders can implement this cycle. It closes with three longer-term capabilities that determine whether differentiated pay becomes a source of competitive advantage or a recurring source of grievance.
The Targeted Pay Landscape
Defining Targeted and Differentiated Pay
Several terms recur throughout this discussion, and they are often used loosely. A salary increase budget is the percentage of base payroll an organization sets aside for pay increases in a given cycle. It typically includes a merit budget, which funds increases tied to individual performance ratings, along with money for promotions, market adjustments, and other actions. The Conference Board's 2027 figures illustrate the distinction: the median total budget is 3.5%, the median merit budget is 3.1%, and the gap is earmarked for actions such as retention adjustments and promotions (The Conference Board, 2026).
An across-the-board increase, which practitioners often call a "peanut butter" raise because it is spread evenly, gives every eligible employee the same percentage or flat amount regardless of performance, market position, or role. Differentiated pay is the opposite: allocation varies by individual or group characteristics that the organization has decided matter. These usually include performance, the market value of a role or skill, flight risk, and position in the pay range.
Within differentiated pay, it helps to separate four distinct mechanisms. Merit differentiation varies increases by performance rating. Skills premiums pay more for specific capabilities that are scarce or strategically valuable, such as AI, data science, or leadership. Market or critical-role adjustments raise pay for jobs whose external rates have moved faster than the internal structure. Retention adjustments are targeted increases or awards for individuals the organization judges to be at risk of leaving and costly to lose. Each mechanism answers a different question, and conflating them is one of the most common sources of perceived unfairness.
Finally, pay dispersion refers to the degree of variation in pay among employees, usually within a unit or job level, and pay compression occurs when the gap between employees with different experience, performance, or levels narrows to the point that it no longer reflects meaningful differences. Shaw (2014) distinguishes dispersion that is explained by legitimate inputs such as performance or job requirements from dispersion that is unexplained or perceived as arbitrary, a distinction that turns out to be central to whether dispersion helps or harms. Pay equity, in turn, refers to the absence of pay differences that cannot be explained by legitimate factors and that correlate with protected characteristics such as gender or race.
Prevalence, Drivers, and Distribution
The flatness of 2027 budgets is striking for its consistency across surveys. Across five large surveys with different samples and methods, the total budget for 2027 clusters tightly between 3.4% and 3.5%. Where merit is reported separately, it sits at 3.1% to 3.2%, leaving roughly 0.3 to 0.4 percentage points for everything else.
That context matters. The ECI shows private-industry wages and salaries rose 3.1% in the 12 months ending June 2026, but after adjusting for consumer prices they fell 0.4% (Bureau of Labor Statistics, 2026). Employees are therefore entering the 2027 cycle with a recent experience of slipping purchasing power. A SHRM review of 2027 planning cited a Resume Now survey in which 69% of working Americans said they feel underpaid (Mayer, 2026). Payscale noted that before the pandemic, average increases of about 3% ran against inflation near 2%, while annual increases peaked at 4% in 2023 and fell to 3.6% in 2024 (McLean, 2026a). In other words, the same nominal raise buys less margin than it once did.
Industry variation is meaningful even within a flat environment. Mercer's 2027 merit projections range from 3.8% in high tech and 3.7% in banking down to 3.0% in healthcare and retail and 2.9% in consumer goods (Mercer, 2026). These spreads reflect differences in labor market tightness, margins, and the share of the workforce in roles where skills are scarce.
The drivers of differentiation are visible in the survey data. Payscale's decline in across-the-board raises sits alongside a sobering finding: a quarter of companies reported losing talent in 2026 because of insufficient pay increases (Ewen, 2026). WTW found that 36% of U.S. companies were hiring at higher salary ranges, 34% were increasing retention bonuses or spot awards, and 32% were raising starting salary ranges (McLean, 2026b). Each of these is a targeted lever, aimed at a segment of the workforce rather than everyone.
Skills are a growing part of the story. In The Conference Board's survey, AI skills led all skill categories in driving base-pay adjustments at 37.8%, followed by leadership and people-management skills at 30.7% and data science and advanced analytics at 30.4% (The Conference Board, 2026). This is a notable development. For most of the past two decades, skills-based pay was a niche approach used in manufacturing and some technical environments. Its re-emergence as a mainstream budget allocation lever reflects both genuine scarcity and a widespread belief that AI capability will shape future productivity.
A third driver is the cost of the alternative. When budgets are flat and spread evenly, the employees with the most external options, who are often the strongest performers and those with scarce skills, receive the smallest raise relative to their market value. Payscale's commentary warned that workers may be more open to exploring the market for higher pay (McLean, 2026a). Employers that remember the turnover of 2021 and 2022 are reluctant to wait for that pattern to repeat.
The distribution of differentiation is uneven. It is most advanced in large technology companies, financial services, and professional services, where variable pay and equity are already significant and performance calibration is institutionalized. In unionized settings, public sector organizations, and many healthcare systems, collective agreements and step structures limit individual differentiation, and targeted pay tends to operate at the level of roles and specialties instead. Smaller organizations often lack the job architecture and pay range infrastructure needed to differentiate defensibly, which is precisely where the legal and trust risks concentrate.
The regulatory environment is shifting in parallel. Pay transparency laws were once concentrated in a handful of states, but each year adds jurisdictions and expands requirements. Connecticut's October 2026 changes are instructive because they extend posting requirements to internal postings and add a general description of benefits (Fisher Phillips, 2026). Internal disclosure matters for differentiation: when current employees can see the range advertised for their own job or the job next door, an increase that leaves them low in the range, or below a new hire, invites an immediate question. In Europe, the Pay Transparency Directive adds rights to information about pay levels and pay progression criteria, along with gender pay gap reporting for larger employers (European Parliament and Council of the European Union, 2023). Multinational organizations designing a single global merit process will find that their least transparent market no longer sets the standard.
Organizational and Individual Consequences
Organizational Performance Impacts
The case for differentiation rests on two mechanisms that compensation scholars call incentive effects and sorting effects. Incentive effects occur when the prospect of higher pay leads current employees to work harder or smarter. Sorting effects occur when pay structures attract, retain, and repel different kinds of people. Gerhart and Fang (2014) argue that sorting is frequently the larger of the two, and that pay-for-performance systems shape who joins and stays at least as much as how hard people work once there.
The meta-analytic evidence on incentives is broadly supportive. Garbers and Konradt (2014) synthesized studies of individual and team-based financial incentives and found positive effects on performance for both, with individual incentives showing a meaningful average effect and team incentives depending on factors such as team size and the equity of the distribution rule. Rynes et al. (2005) reached a similar conclusion in an earlier review, while noting that the effects depend heavily on the quality of performance measurement.
The form of differentiation matters, not just its presence. Nyberg et al. (2016) followed more than 11,000 employees at a large insurance company over five years and found that both merit increases and bonuses predicted higher future performance, but bonuses had a substantially stronger effect. The authors attributed this to the salience of a lump sum and to the fact that a bonus must be re-earned. Bonuses also avoid the compounding cost of permanent base-pay increases, a consideration that matters when budgets are tight. However, they cautioned that relying on bonuses alone can weaken an organization's market position on base pay, which affects recruitment and retention.
Retention is where differentiation shows its clearest organizational benefit. Trevor et al. (1997) found that the relationship between performance and voluntary turnover is curvilinear: both low and high performers are more likely to leave than average performers, and salary growth reduces turnover most among high performers. Shaw and Gupta (2007) found that pay systems emphasizing performance were associated with greater retention of good performers and higher quit rates among poor performers. Seen through this lens, an across-the-board raise in a flat-budget year is a subsidy from the high performer to the low performer, and the high performer is the one with options.
Yet differentiation can also backfire. Bloom (1999), studying more than 1,600 Major League Baseball players over eight seasons, found that teams with more hierarchical, dispersed pay performed worse on field and financial measures, and that players at the lower end of the distribution performed worse individually. Trevor et al. (2012) helped reconcile these findings by showing that pay dispersion explained by performance inputs tends to improve performance in interdependent work, whereas unexplained dispersion harms it. Shaw (2014) and Downes and Choi (2014) reached similar conclusions in their reviews: the effect of dispersion depends on whether employees see it as legitimate, on how interdependent the work is, and on whether the basis for differences is transparent. The practical implication is clear. Differentiation produces performance gains when it tracks real contribution in a way people can see; it produces resentment and coordination failures when it appears arbitrary.
There is an organizational cost on the legal and reputational side as well. Differentiation that relies heavily on manager discretion is fertile ground for bias. Castilla (2008) found, in a large service organization, that women, minorities, and non-U.S.-born employees received smaller salary increases than white men with identical performance ratings. Castilla and Benard (2010) documented a "paradox of meritocracy": in a series of experiments, when an organization was presented as explicitly meritocratic, participants acting as managers awarded larger bonuses to men than to equally performing women. In an era of expanding pay transparency, these patterns become visible faster, and the remediation costs rise.
Individual Wellbeing and Stakeholder Impacts
For individuals, pay outcomes are filtered through comparison. Adams's (1965) equity theory proposed that people evaluate their pay by comparing their ratio of outcomes to inputs with that of relevant others, and that perceived inequity produces tension and efforts to restore balance, including reduced effort or exit. Six decades of research have refined that model, but its core insight has held. Card et al. (2012) provided evidence from a field experiment at the University of California, in which some employees were informed about a newly searchable database of coworkers' salaries: those paid below the median for their unit and occupation reported lower job satisfaction and greater intention to search for a new job, while those above the median showed little change. The asymmetry matters for leaders. Differentiation creates more people below the midpoint who notice than people above it who feel rewarded.
Compression is a related and often hidden source of grievance. When organizations raise starting ranges to compete for new hires, as nearly a third of WTW respondents reported doing (McLean, 2026b), while giving incumbents a modest merit increase, experienced employees can find themselves paid close to, or even below, recent hires in the same role. Few experiences undermine perceived fairness more quickly, because the comparison is concrete, local, and easy to discover. A targeted pay strategy that funds market competitiveness at the point of hire without a parallel mechanism for incumbents can create precisely the unexplained dispersion that the research associates with lower commitment and higher turnover.
Organizational justice research adds a crucial nuance. Colquitt et al. (2001), in a meta-analysis of 183 studies, found that procedural justice, the perceived fairness of the processes used to make decisions, predicted outcomes such as trust, commitment, and job satisfaction even after accounting for distributive justice, the perceived fairness of the outcomes themselves. Folger and Konovsky (1989) showed that procedural justice in pay raise decisions was particularly important for organizational commitment and trust in supervisors, whereas distributive justice was more closely tied to satisfaction with the raise itself. People can accept a smaller raise if they believe the process was sound. They rarely accept even a reasonable raise if they believe the process was rigged.
Pay satisfaction carries consequences beyond mood. Williams et al. (2006), in a meta-analysis of the antecedents and consequences of pay level satisfaction, found that it was related to organizational commitment, job satisfaction, and turnover intentions, and that perceptions of procedural fairness and the actual pay level both mattered. For employees already feeling squeezed by real wage declines, a below-average raise without a credible explanation is likely to register as a signal of how the organization values them.
The interaction with pay transparency complicates matters. Bamberger and Belogolovsky (2010) found in an experimental study that pay secrecy was associated with lower task performance, partly because it weakened perceptions of the link between performance and pay. Transparency, in other words, can strengthen the motivational effect of merit. But it can also constrain differentiation. Obloj and Zenger (2022), analyzing pay data for nearly 100,000 academics in U.S. public universities over 14 years, found that transparency was associated with a smaller gender pay gap and lower pay inequality, but also with a weaker link between pay and research performance. Mas (2017) found that after California cities began publishing municipal managers' salaries, top managers' pay fell, consistent with public aversion to high pay. Cullen (2024), reviewing the broader evidence, concluded that horizontal transparency, which reveals coworkers' pay within a firm, narrows pay gaps but can encourage counterproductive peer comparisons and harder employer bargaining, whereas vertical and cross-firm transparency can raise worker expectations and wage competition.
For frontline and lower-paid workers, the stakes are different again. In a flat-budget year, the choice to differentiate often means that steady, solid performers in roles without scarce skills receive increases at or below inflation. Their wellbeing depends heavily on whether the organization maintains a credible floor, provides a clear path to higher pay through skill development and promotion, and avoids the impression that differentiation is a euphemism for austerity at the bottom. Managers, finally, are a stakeholder group whose burden is easy to overlook. Differentiation asks them to make harder decisions and to explain them in more uncomfortable conversations, often with little support. Salary.com's 2026 pay practices research found that only about 52% of organizations provide any formal training on conducting pay discussions (Moody, 2026), and a Korn Ferry study found that only 16% of organizations are confident their managers can explain AI-related pay and work changes (Goforth, 2026).
Evidence-Based Organizational Responses
The evidence does not support a simple choice between uniform and differentiated raises. It supports differentiation that is structured, explained, and audited. Table 1 summarizes five responses that together make that possible, followed by a fuller discussion of each.
Table 1: Evidence-based responses for differentiating pay within flat 2027 budgets
Response | Evidence Base | Key Practices | Illustrative Examples |
Separate the budget into purpose-built pools | The Conference Board (2026); Gerhart & Fang (2014); Trevor et al. (1997) | Distinct merit, market, retention, promotion, and equity pools; explicit criteria for each | Walmart (store and market manager pay); Amazon (base pay cap); Costco (tiered hourly increases) |
Make performance differentiation credible | Nyberg et al. (2016); Trevor et al. (2012); Garbers & Konradt (2014) | Calibrated ratings; bonuses for strongest differentiation; protect a base-pay floor | Google (rating multipliers); Meta (Checkpoint); Microsoft (2023 salary freeze) |
Price skills and critical roles with discipline | The Conference Board (2026); Shaw (2014); SullivanCotter (2025) | Validated skill criteria; time-limited premiums; role-level market data | Meritus Health and Beaufort Memorial (hybrid models); nursing specialty differentials |
Audit pay equity before and after the cycle | Castilla (2008, 2015); Castilla & Benard (2010) | Pre-cycle gap analysis; post-cycle review of increases; remediation budget | Salesforce (equal pay assessments); Citi (raw and adjusted gap disclosure) |
Explain pay and equip managers to do it | Colquitt et al. (2001); Folger & Konovsky (1989); Cullen (2024) | Published ranges and criteria; manager scripts and training; review channel | Buffer (open salary formula); Purdue University (recurring and recognition components) |
Separate the Budget Into Purpose-Built Pools
The single most useful structural change most organizations can make in 2027 is to stop treating the salary increase budget as one undifferentiated pot. The Conference Board (2026) recommends using the 3.5% median as a planning anchor rather than a uniform strategy, and funding promotions and retention adjustments through separate pools so they do not erode the merit budget. This approach has a sound theoretical basis. Gerhart and Fang (2014) emphasize that pay systems communicate what an organization values. When performance, market, and retention money are blended, employees cannot tell which signal they received, and managers lose the ability to explain decisions coherently. The turnover research cited earlier (Trevor et al., 1997) suggests that targeted salary growth for high performers reduces exits, but only if that money is not diluted across everyone.
Effective approaches:
Size pools to strategy: Allocate the total budget across merit, market and skills, retention, promotion, and equity correction based on workforce data rather than habit.
Write criteria for each pool: Define who is eligible and why, so a retention adjustment is never mistaken for a merit reward.
Use range position: Give larger increases to strong performers who sit low in their range and smaller ones to those already above the midpoint.
Protect a floor: Decide in advance what minimum increase, if any, solid performers will receive, and say so.
Reserve off-cycle funds: Hold a small portion for mid-year market or retention actions rather than spending everything in the annual cycle.
Walmart offers a clear example of role-targeted investment. In January 2025, the company raised average base pay for U.S. store managers from $117,000 to $128,000 and increased stock grants and bonus potential for market managers, who oversee about a dozen stores each (Genovese, 2025). The changes concentrated new money on two specific leadership roles, and they combined base pay with variable components: store manager bonuses can reach 200% of base salary when targets and profit metrics are met.
Amazon took a different structural lever in 2022, more than doubling its maximum base salary for corporate and technology employees from $160,000 to $350,000 and citing a particularly competitive labor market in an internal memo (S. Soper, 2022). Amazon had long relied on stock awards to supplement relatively low base pay, and the memo paired the higher ceiling with larger-than-typical compensation increases and higher pay ranges for most jobs (S. Soper, 2022). Raising a structural cap is a quieter form of targeting than an across-the-board raise: it changes what is possible for the employees whose market value exceeds the old limit without changing pay for everyone else.
Costco's 2025 wage increase illustrates differentiation within an hourly structure. Facing a possible Teamsters strike, the company announced that hourly workers at the top of the pay scale would receive $1 raises in each of the following three years, bringing the top rate to $30.20, while entry-level workers received a 50-cent increase (Selyukh, 2025). The design weighted the largest increases toward employees at the top of the scale rather than spreading a single percentage across all employees, and the union context underscores that structure choices are rarely made in isolation from labor relations.
A common pitfall is creating pools without changing the conversation. If managers still announce a single number to each employee, the separate pools exist only in the spreadsheet, and employees continue to interpret every increase as a judgment of performance. The pools have value only if the communication explains which pool each component came from.
Make Performance Differentiation Credible
Merit differentiation is the most familiar form of targeted pay and the one most dependent on the quality of performance measurement. Nyberg et al. (2016) found that pay-for-performance improves future performance, but that the effect is strongest when performance is measurable and when the reward is salient. Garbers and Konradt (2014) similarly found that incentive effects depend on design features. Trevor et al. (2012) add that dispersion helps only when it is clearly linked to performance inputs. In practice, this means that the credibility of the rating process determines whether differentiation motivates or alienates. Calibration, in which managers review and reconcile ratings across teams, is the most common mechanism for improving consistency, though it introduces its own risks of status-driven negotiation.
Effective approaches:
Calibrate with evidence: Require documented examples for top and bottom ratings, and track whether calibration shifts outcomes systematically for any group.
Use variable pay for the sharpest differences: Reserve the largest differentiation for bonuses and equity, which must be re-earned, and keep base-pay merit spreads moderate.
Avoid rigid forced distributions: Use guidance on rating mix rather than hard quotas, which can penalize strong teams.
Make the middle meaningful: Ensure that a solid, fully competent rating still produces a respectable outcome.
Separate development from pay: Hold at least one performance conversation each year that is not about money.
Google announced changes in 2025, effective in 2026, to strengthen what it called its performance culture, giving managers more room to award top ratings and a larger discretionary budget for high performers, funded in a budget-neutral way by slightly reducing bonus and equity multipliers for middle ratings (Broomfield, 2025). Google stressed that its most common rating remained a strong one that yields an above-target bonus. The design concentrated differentiation in variable pay rather than base salary, which is consistent with the research on bonuses.
Meta's Checkpoint program, announced in January 2026 for the 2026 performance year, uses a four-tier rating scale with explicit bonus multipliers, from 200% for the roughly 20% rated Outstanding to 0% for the roughly 3% rated Not Meeting Expectations, plus a separate award carrying a 300% multiplier for a small number of employees (Mediaweek, 2026). Meta moved to two review cycles a year, with equity refreshers based on the average of the two ratings and merit increases remaining annual. The transparency of the multipliers gives employees a clear line of sight, though the expected distribution also sets visible limits on how many people can receive top outcomes.
Microsoft's 2023 decision shows a different side of differentiation in a constrained year. The company announced it would not provide salary increases for full-time salaried employees, while continuing to invest in promotions, bonuses, and stock awards, though without overfunding the bonus and stock budget as it had the year before (T. Soper, 2023). The move preserved performance-linked variable pay while freezing base salaries, effectively shifting differentiation from merit to bonuses and equity during a period that also included significant layoffs.
These cases share a design principle that the research supports: the sharpest differences sit in pay that must be re-earned, while base pay moves more gradually. They also raise a question many organizations avoid, which is what to do at the bottom of the distribution. A zero increase for employees rated as not meeting expectations frees budget for others, but it should be the end of a documented performance process rather than a surprise delivered in a pay letter. If an employee learns of a performance problem for the first time when the raise is withheld, the organization has failed on procedural justice regardless of the merits of the rating.
A common pitfall is assuming that a rating scale with more categories produces more differentiation. Without calibration, managers tend to cluster ratings at the top, and without a clear explanation of what each rating means for pay, employees treat any rating below the highest as a criticism. Credibility comes from consistent standards, not from the number of boxes.
Price Skills and Critical Roles With Discipline
Skills premiums and critical-role adjustments are the fastest-growing area of targeted pay, and the one most prone to improvisation. The Conference Board's (2026) finding that AI skills drive base-pay adjustments at 37.8% of organizations suggests that many employers are already paying for capabilities that their job architecture does not formally recognize. Shaw (2014) cautions that dispersion based on criteria employees cannot see or verify is likely to be perceived as unexplained, which is the type most associated with negative outcomes. The healthcare sector offers useful evidence on what disciplined role-based targeting looks like. SullivanCotter's 2025 nursing compensation survey found that median base pay increases varied by role, from 3.0% for RN leaders to 1.7% for staff RNs, and that organizations were more selective in awarding larger adjustments, with 90th percentile increases for supervisors, charge nurses, and staff RNs ranging from 5.7% to 8.2% (SullivanCotter, 2025). Specialties such as critical care led median increases.
Effective approaches:
Define skills observably: Tie premiums to demonstrated, assessed capabilities rather than job titles or self-reported proficiency.
Make premiums reviewable: Set skills premiums as separate, periodically reviewed components so they can be adjusted if scarcity eases.
Use role-level market data: Base critical-role adjustments on time-to-fill, turnover, and external rate data, not anecdote.
Open pathways: Offer a visible route for any employee to acquire the skill and qualify for the premium.
Watch for proxies: Check that skills criteria do not unintentionally favor groups who have had more access to training or high-profile projects.
Several health systems have adopted hybrid models that combine broad increases with targeted rewards. Becker's Hospital Review reported in August 2026 that Meritus Health and Beaufort Memorial Hospital in South Carolina have kept broad raises while adding merit rewards for top performers, and that health system HR leaders are directing pay dollars to hard-to-fill roles such as nursing, pharmacy, respiratory therapy, and behavioral health using time-to-fill and turnover data (Kuchno, 2026). The hybrid design acknowledges that in a workforce where most employees share a mission and depend on one another, a meaningful floor matters for cohesion, while targeted money addresses genuine scarcity.
A common pitfall is paying an AI premium for enthusiasm rather than capability. When premiums are awarded to people who volunteer for AI pilots or who hold certain titles, the organization pays for signals that are weakly related to the value it hoped to buy, and other employees quickly notice who was chosen and who was not.
Audit Pay Equity Before and After the Cycle
Any increase in differentiation increases the opportunity for bias to enter pay decisions. Castilla's (2008) finding that equal performance ratings translated into unequal raises for women and minorities, and Castilla and Benard's (2010) evidence that merit-focused cultures can amplify bias, are directly relevant to the 2027 cycle. The encouraging evidence comes from Castilla (2015), who studied a large organization that introduced accountability and transparency procedures into its pay decisions, including a committee that reviewed pay decisions and the sharing of aggregate information about pay outcomes. After the changes, previously observed gaps in performance-based raises for women, minorities, and non-U.S.-born employees were substantially reduced. The lesson is that equity is not achieved by good intentions in a merit system; it is achieved through process design and monitoring.
Effective approaches:
Audit before allocating: Run a statistical pay equity analysis before the cycle so managers know where unexplained gaps exist.
Fund remediation separately: Use a dedicated equity pool so corrections do not compete with merit increases.
Review increases by group: After recommendations are entered, compare increase distributions by gender, race, age, and other characteristics, controlling for rating and range position.
Hold decision-makers accountable: Require explanations for outliers and share aggregate results with leaders.
Report outcomes: Communicate what the audit found and what was done, even if only internally.
Salesforce has conducted regular equal pay assessments since a 2015 commitment. In its 2020 update, the company reported assessing about 50,000 employees, adjusting pay for 7% of them, and spending $2.1 million in that cycle, bringing its cumulative spending on addressing unexplained pay differences to more than $12 million (Salesforce, 2020). The company integrated the assessment into its annual compensation planning, which is the timing the research suggests matters most.
Citi took the further step of publishing both adjusted and unadjusted pay gaps. In 2018, it became the first U.S. bank to share a statistically adjusted pay gap review, and in 2019 it began disclosing raw median gaps, which showed women's median pay at 71% of men's, even though its adjusted analysis had found women paid about 99% of what comparable men earned (Kline, 2024). Publishing both figures distinguished unequal pay for comparable work from differences in who holds the highest-paid roles. Yet in 2024, American Banker reported that Citi had paused its annual pay gap review during an organizational restructuring, indicating it planned to resume once roles were realigned (Kline, 2024). The episode shows both the value of disclosure and the fragility of equity processes during periods of change.
A common pitfall is treating the audit as a compliance exercise run by legal counsel after decisions are final. By then, the opportunity to prevent gaps has passed, and the organization is left choosing between costly corrections and uncomfortable silence.
Explain Pay and Equip Managers to Do It
The organizational justice evidence points to explanation as the most underused lever in compensation. Colquitt et al. (2001) found that procedural and informational justice predicted trust and commitment beyond outcome fairness, and Folger and Konovsky (1989) showed that fair procedures in raise decisions shaped trust in supervisors. Transparency research adds that the type of transparency matters: Cullen (2024) concludes that vertical transparency, showing how pay grows with level, and process transparency can strengthen motivation without the peer-comparison costs of fully open individual salaries. With Connecticut now requiring wage ranges in internal postings (Fisher Phillips, 2026), more employees will see the ranges for the jobs around them, making a coherent explanation of how pay decisions are made a practical necessity.
Effective approaches:
Publish the logic: Share the pay philosophy, the structure of the budget pools, and the criteria for each type of increase.
Show ranges and progression: Give employees access to the range for their own role and the next level.
Train managers to explain: Provide talking points, practice sessions, and data on each employee's range position before conversations.
Say no clearly: When there is no money for an increase, explain why and what would change the answer.
Offer a review route: Create a channel for employees to question a decision without fear of retaliation.
The hardest conversations in a differentiated system are not with top performers. They are with solid contributors who receive an increase below inflation, and with employees who learn that a colleague received a skills premium or retention adjustment they did not. In those conversations, what helps most is specificity: where the employee sits in the range, which pool their increase came from, what criteria governed the other pools, and what concrete steps would change the outcome next time. Employees may not like the answer, but a specific answer signals respect in a way that a vague reference to "the budget" does not.
Buffer, a social media software company, has published its salary formula and individual salaries since around 2013. Its 2017 update described a formula built from a market benchmark, a cost-of-living multiplier, a role multiplier, and an experience factor, and explained the company's decision to remove an automatic annual loyalty raise while honoring existing salaries so no one's pay would drop (Buffer, 2017). Buffer's approach is unusually open and is not a template for most organizations, but it illustrates how explicit criteria make individual outcomes explainable.
Purdue University offers a more conventional example. For fiscal year 2026, its Board of Trustees approved a 2% recurring salary increase plus a 0.5% nonrecurring component described as recognition of exceptional employee contributions (Purdue University, 2025). Separating recurring base pay from a one-time recognition element communicated two different messages in a single, publicly announced structure, which is precisely the clarity that pooled budgets aim to create.
A common pitfall is assuming that managers already know how to have these conversations. The Salary.com finding that only about half of organizations train managers on pay discussions (Moody, 2026) suggests that many managers will default to blaming the system, which undermines both their credibility and the organization's.
Building Long-Term Pay Capability
The five responses above can be implemented within a single compensation cycle. Their effect, however, depends on organizational infrastructure that takes longer to build. Three capabilities determine whether differentiation becomes sustainable.
Job Architecture and Pay Ranges as Infrastructure
Differentiation is only as defensible as the structure underneath it. A job architecture is the system of job families, levels, and titles that defines how roles relate to one another, and pay ranges attached to that architecture provide the reference points for every increase decision. Salary.com's 2026 research found that only about half of employers have a formal job architecture and that 22% do not use job leveling at all (Moody, 2026). Without these foundations, organizations cannot calculate range position, identify compression, comply credibly with posting requirements that call for good-faith ranges, or explain why two people in similar roles are paid differently. The Conference Board (2026) emphasizes differentiation across performance, market pressure, pay equity, critical roles, and scarce skills, but each of those dimensions requires a stable framework against which to measure.
Building this infrastructure is unglamorous work. It involves leveling roles consistently, refreshing market data on a regular cycle, setting range widths that allow meaningful progression, and documenting the rules for movement within and between ranges. Organizations that have done this work find that differentiation becomes easier to explain, because each decision can be located on a map that employees can also see. Those that skip it find that every targeted increase invites a comparison they cannot answer.
The architecture also needs to accommodate skills. If AI and data skills are going to carry premiums for several years, they should eventually be reflected in job definitions, career frameworks, and range placement rather than remaining as ad hoc adjustments. That transition allows skills pay to become part of the structure rather than an exception to it.
Manager Capability and Pay Governance
The second capability is the human system that makes decisions and explains them. Differentiated pay shifts significant discretion to managers, and the research on bias in merit systems (Castilla, 2008; Castilla & Benard, 2010) makes clear that discretion without accountability tends to reproduce existing inequalities. At the same time, pay communication research suggests that managers are the most important channel through which employees understand pay. The Korn Ferry finding that only 16% of organizations are confident their managers can explain AI-related pay and work changes (Goforth, 2026) signals a broad capability gap.
Strong pay governance includes clear decision rights, so managers know what they can decide and what requires review; calibration forums with documented reasoning; second-level review of outliers; and periodic analysis of each manager's pay decisions by group. Manager capability includes not just technical knowledge of ranges and budgets but the interpersonal skill to deliver disappointing news honestly. In my experience, the managers who handle these conversations best are those who understand the logic well enough to explain it in their own words and who are willing to acknowledge when the organization's constraints are genuinely tough.
Investing in this capability pays off well beyond compensation. Managers who can explain pay can also explain performance expectations, career pathways, and organizational priorities. Pay conversations are often the most consequential conversations an employee has with a manager in a year, and treating them as a skill to be developed rather than an administrative task changes the tone of the entire relationship.
Measurement, Learning, and Course Correction
The third capability is the discipline to measure whether differentiation is working. Most organizations track whether they spent the budget. Far fewer track whether the spending achieved its purpose. Useful indicators include regretted turnover among high performers and employees with critical skills; the distribution of increases by rating, range position, and demographic group; changes in compression within levels; employee perceptions of pay fairness and process fairness; and the time it takes to fill critical roles. Salary.com's research found a 31-point gap between the share of HR professionals who believe employees are paid fairly and the share who believe employees agree (Moody, 2026), suggesting that many organizations have little direct evidence of how their pay decisions are received.
Measurement enables course correction. If skills premiums are not reducing turnover among the employees who receive them, the organization should revisit their size or criteria. If calibration consistently lowers ratings for a particular group, the process needs redesign. If high performers continue to leave despite larger increases, pay may not be the binding constraint. The research on dispersion (Shaw, 2014; Trevor et al., 2012) suggests that the same pay structure can produce different results depending on how it is perceived, so perception data belongs alongside financial data in every review.
Over time, measurement also builds organizational learning about which forms of differentiation suit the organization's work. Highly interdependent teams may benefit more from team-based rewards or moderate dispersion, while roles with measurable individual output may support sharper differentiation. These are empirical questions, and the organizations that answer them with their own data will make better decisions than those relying on survey benchmarks alone.
Conclusion
The 2027 compensation season presents leaders with a deceptively simple constraint: roughly 3.5% to spend and a workforce whose real wages have recently slipped. The move away from across-the-board increases is a rational response to that constraint. Uniform raises fail to recognize the strongest contributors, leave scarce skills underpriced, and, as a quarter of companies discovered in 2026, can cost organizations the people they most need to keep. The evidence on pay-for-performance, sorting effects, and turnover supports the idea that targeted pay can produce a better return on a flat budget.
But the same body of evidence warns that differentiation succeeds only when it is perceived as legitimate. Dispersion that tracks visible contribution can improve performance; dispersion that appears arbitrary damages cooperation and trust. Merit systems without accountability tend to reproduce bias. Transparency strengthens the motivational link between pay and performance, yet it also constrains how far organizations can differentiate before employees push back. Procedural fairness, the quality of the explanation and the process, matters as much as the size of the raise.
The organizations profiled in this article illustrate different points on this spectrum. Walmart and Amazon targeted specific roles and structural ceilings. Google and Meta sharpened differentiation through variable pay and explicit multipliers. Microsoft froze base salaries while preserving performance-linked bonuses and stock. Health systems such as Meritus and Beaufort Memorial combined broad increases with targeted rewards. Salesforce and Citi built equity audits and disclosure into their processes, with Citi's pause showing how fragile those commitments can be. Buffer and Purdue demonstrated, in very different ways, how explicit structures make outcomes explainable.
For leaders preparing 2027 decisions, the practical message is that differentiation is a design problem, not a budget problem. The question is not whether to differentiate but whether the organization has the structures, processes, and capabilities to do so in a way employees can understand and accept.
Divide the budget deliberately: Separate merit, market and skills, retention, promotion, and equity pools, with written criteria for each.
Differentiate most through variable pay: Use bonuses and equity for the sharpest distinctions, and keep base-pay merit spreads moderate and defensible.
Price skills with evidence: Tie AI and other skills premiums to demonstrated capability, review them periodically, and keep pathways open.
Audit equity before and after: Analyze gaps before allocating, review increases by group afterward, and fund remediation separately.
Protect a credible floor: Decide what solid performers will receive and explain it, especially in a year of declining real wages.
Explain the logic: Publish the pay philosophy, ranges, and criteria, and prepare for wider visibility as transparency laws expand.
Invest in managers: Train managers to explain decisions, give them range data, and hold them accountable for consistent outcomes.
Measure what matters: Track regretted turnover, increase distributions, compression, and perceived fairness, and adjust the design each cycle.
Research Infographics
References
Adams, J. S. (1965). Inequity in social exchange. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 2, pp. 267–299). Academic Press.
Bamberger, P., & Belogolovsky, E. (2010). The impact of pay secrecy on individual task performance. Personnel Psychology, 63(4), 965–996.
Bloom, M. (1999). The performance effects of pay dispersion on individuals and organizations. Academy of Management Journal, 42(1), 25–40.
Broomfield, B. (2025, May 1). Google boosts comp for top performers – by trimming bonuses for middling staff. HR Grapevine.
Buffer. (2017, December 6). The next evolution of transparent salaries: Our new remote-first formula and updated salary calculator. Buffer.
Bureau of Labor Statistics. (2026, July 31). Employment cost index – June 2026 [News release]. U.S. Department of Labor.
Card, D., Mas, A., Moretti, E., & Saez, E. (2012). Inequality at work: The effect of peer salaries on job satisfaction. American Economic Review, 102(6), 2981–3003.
Castilla, E. J. (2008). Gender, race, and meritocracy in organizational careers. American Journal of Sociology, 113(6), 1479–1526.
Castilla, E. J. (2015). Accounting for the gap: A firm study manipulating organizational accountability and transparency in pay decisions. Organization Science, 26(2), 311–333.
Castilla, E. J., & Benard, S. (2010). The paradox of meritocracy in organizations. Administrative Science Quarterly, 55(4), 543–576.
Colquitt, J. A., Conlon, D. E., Wesson, M. J., Porter, C. O. L. H., & Ng, K. Y. (2001). Justice at the millennium: A meta-analytic review of 25 years of organizational justice research. Journal of Applied Psychology, 86(3), 425–445.
Cullen, Z. (2024). Is pay transparency good? Journal of Economic Perspectives, 38(1), 153–180.
Downes, P. E., & Choi, D. (2014). Employee reactions to pay dispersion: A typology of existing research. Human Resource Management Review, 24(1), 53–66.
European Parliament and Council of the European Union. (2023). Directive (EU) 2023/970 to strengthen the application of the principle of equal pay for equal work or work of equal value between men and women through pay transparency and enforcement mechanisms. Official Journal of the European Union, L 132.
Ewen, L. (2026, August 6). Payscale: Merit increases remain popular as 'peanut butter' raises trend downward. HR Dive.
Fisher Phillips. (2026, September 18). Employer cheat sheet for workplace laws taking effect in October 2026. Fisher Phillips LLP.
Folger, R., & Konovsky, M. A. (1989). Effects of procedural and distributive justice on reactions to pay raise decisions. Academy of Management Journal, 32(1), 115–130.
Garbers, Y., & Konradt, U. (2014). The effect of financial incentives on performance: A quantitative review of individual and team-based financial incentives. Journal of Occupational and Organizational Psychology, 87(1), 102–137.
Genovese, D. (2025, January 24). Some Walmart managers get pay bump, pushing compensation over $600K. FOX Business.
Gerhart, B., & Fang, M. (2014). Pay for (individual) performance: Issues, claims, evidence and the role of sorting effects. Human Resource Management Review, 24(1), 41–52.
Goforth, A. (2026, August 3). Study: Managers can't explain AI-driven pay and work changes. HR Executive.
Kline, A. (2024, March 20). Citi puts pay equity analysis on hold amid cuts to its workforce. American Banker.
Kuchno, K. (2026, August 6). 2026 health system pay strategies are getting more targeted: 5 takeaways. Becker's Hospital Review.
Mas, A. (2017). Does transparency lead to pay compression? Journal of Political Economy, 125(5), 1683–1721.
Mayer, K. (2026, August 18). 2027 pay raises expected to hold largely flat amid cost pressures. SHRM.
McLean, D. (2026a, August 4). Fewer companies plan 'peanut-butter' pay increases in 2027: Payscale. CFO Dive.
McLean, D. (2026b, July 24). Companies aim for more strategic pay increases: WTW report. HR Dive.
Mediaweek. (2026, January 14). Meta revamps reviews with bigger bonuses for top staff. Mediaweek.
Mercer. (2026, August 31). US employers stick with moderate pay increases for 2027 [Press release].
Moody, K. (2026, April 17). Are workers paid fairly? It may depend on who you ask. HR Dive.
Nyberg, A. J., Pieper, J. R., & Trevor, C. O. (2016). Pay-for-performance's effect on future employee performance: Integrating psychological and economic principles toward a contingency perspective. Journal of Management, 42(7), 1753–1783.
Obloj, T., & Zenger, T. (2022). The influence of pay transparency on (gender) inequity, inequality and the performance basis of pay. Nature Human Behaviour, 6(5), 646–655.
Purdue University. (2025, April 4). Trustees endorse 14th consecutive tuition freeze while approving 2%+0.5% faculty and staff salary increase [News release].
Rynes, S. L., Gerhart, B., & Parks, L. (2005). Personnel psychology: Performance evaluation and pay for performance. Annual Review of Psychology, 56, 571–600.
Salesforce. (2020, March 9). 2020 Salesforce equal pay update. Salesforce.
Selyukh, A. (2025, January 31). Costco raises pay above $30 an hour for many, as Teamsters union threatens strike. NPR.
Shaw, J. D. (2014). Pay dispersion. Annual Review of Organizational Psychology and Organizational Behavior, 1, 521–544.
Shaw, J. D., & Gupta, N. (2007). Pay system characteristics and quit patterns of good, average, and poor performers. Personnel Psychology, 60(4), 903–928.
Soper, S. (2022, February 7). Amazon is raising base salary cap to $350,000 from $160,000. Bloomberg.
Soper, T. (2023, May 10). Microsoft won't give raises to full-time salaried employees this year. GeekWire.
SullivanCotter. (2025, December 16). 2025 nursing compensation data reveals targeted pay increases [Press release].
The Conference Board. (2026, September 3). US salary increase budgets 2026–2027. The Conference Board.
Trevor, C. O., Gerhart, B., & Boudreau, J. W. (1997). Voluntary turnover and job performance: Curvilinearity and the moderating influences of salary growth and promotions. Journal of Applied Psychology, 82(1), 44–61.
Trevor, C. O., Reilly, G., & Gerhart, B. (2012). Reconsidering pay dispersion's effect on the performance of interdependent work: Reconciling sorting and pay inequality. Academy of Management Journal, 55(3), 585–610.
Williams, M. L., McDaniel, M. A., & Nguyen, N. T. (2006). A meta-analysis of the antecedents and consequences of pay level satisfaction. Journal of Applied Psychology, 91(2), 392–413.
WorldatWork. (2026, July 15). Salary budget survey 2026–2027: Toplines. WorldatWork.
Jonathan H. Westover, PhD, Chief Research Officer (Nexus Institute for Work and AI); Co-Founder & Chief Workforce and Learning Officer (Future State University); Founder & CEO (Human Capital Innovations); Professor of Organizational Leadership & Change (UVU). Read Jonathan Westover's executive profile here.
Suggested Citation: Westover, J. H. (2026). The Targeted Raise: Flat Pay Budgets, the Retreat From Across-the-Board Increases, and How Leaders Can Differentiate Pay Without Breaking Trust. Human Capital Leadership Review, 41(1). doi.org/10.70175/hclreview.2020.41.1.2
