How Anchoring Biases Distort Performance Appraisals When Managers Review Historical Ratings First

Human resources directors and department heads argue that reviewing an employee’s historical ratings before writing a new appraisal is essential for tracking long-term growth. They view past scores as a necessary baseline that prevents erratic jumps in performance data and ensures year-over-year consistency across large teams.[2]

Behavioral economists and organizational psychologists view that exact same baseline as a contaminant. They argue that exposing a manager to a previous rating replaces the objective observation of recent work with an unconscious cognitive drag. The past score becomes a gravitational pull that distorts the current evaluation.[1]

This tension between longitudinal tracking and independent evaluation sits at the center of modern performance management. When a manager opens a file and sees a 4.0 out of 5.0 from the previous year, that specific number immediately sets the boundaries for the current conversation.[4]

The stakes for this structural design choice are entirely financial. Performance appraisals dictate salary increases, bonus pools, and promotion velocity. When a past rating artificially caps a current score, it directly suppresses an employee's earning potential and alters their long-term career trajectory within the firm.[2]

The Mechanics of the Anchor

Anchoring bias is a cognitive heuristic where the human brain relies disproportionately on the first piece of information it receives. In a performance review, the anchor is typically a numerical score or a summary judgment from a previous cycle that sets an inescapable mental reference point.[1]

"The anchoring bias is a pervasive cognitive bias that causes us to rely too heavily on information that we receive early in the decision-making process," explains Dr. Sekoul Krastev of The Decision Lab. The brain uses this initial data to save mental energy during complex evaluations.[1]

Once that anchor is set, all subsequent evidence is evaluated relative to that initial number rather than against an objective standard. If a manager is uncertain whether an employee's recent project warrants a "meets expectations" or an "exceeds expectations," a historical "meets expectations" rating will resolve the ambiguity downward.[4]

This cognitive shortcut operates entirely outside of conscious awareness. Even managers who are explicitly trained to recognize the anchoring effect, and who know the foundational behavioral science behind it, still adjust their current estimates toward the historical number when forced to make a subjective judgment.[1]

The effect is particularly pronounced when a new manager inherits an existing team. The incoming leader's first impression of each team member is anchored to the predecessor's scores, transferring any historical biases directly into the new reporting relationship before a single new project is completed.[4]

As a result, the performance review never actually starts from zero. The previous year's rating acts as a ceiling for improving employees and a floor for declining ones, compressing the overall distribution of scores into a narrow, historical band that ignores recent reality.[4]

Quantifying the Distortion

The scale of this cognitive distortion is measurable across large corporate datasets. According to 2024 Gallup research, leniency, severity, and anchoring biases collectively distort up to 35 percent of performance ratings in large organizations, rendering a massive portion of the data statistically unreliable.[4]

That figure means that more than one in three performance reviews is skewed by a systematic rating error rather than reflecting actual employee output. When 33 percent of the data is compromised, the downstream talent decisions based on that data are equally flawed and inequitable.[4]

Anchoring bias specifically restricts rating mobility over time. High performers who secure an early top rating can coast on old wins, as managers unconsciously anchor to their historical excellence even if their current output drops significantly below the required standard for their pay grade.[4]

Conversely, newer employees or late bloomers struggle to break out of average bands. If an employee received a mediocre rating during their onboarding year, that initial anchor drags down their subsequent evaluations, even after they have mastered the role and increased their daily output.[4]

This inertial pull fundamentally undermines the purpose of an annual review. If the system is designed to measure discrete performance over a 12-month period, carrying over a historical anchor transforms the appraisal into a trailing average rather than a current snapshot of an employee's value.[2][4]

The Self-Assessment Trap

Historical ratings from managers are not the only anchors that distort appraisals. Employee self-assessments, when submitted and reviewed before the manager writes their own evaluation, exert an identical gravitational pull on the final score, shifting the baseline before the manager even begins to type.[4]

Research published by the Harvard Kennedy School in December 2025 revealed a striking correlation in appraisal sequencing. When managers read employee self-evaluations first, their final managerial scores closely tracked the employees' self-ratings, proving that the first number seen dictates the final number given.[4]

This sequencing creates a severe structural disadvantage for specific demographics. The same Harvard Kennedy School research found that women and workers of color consistently rated themselves lower on self-assessments than their white male peers, introducing an immediate downward anchor into their performance files.[4]

Because the managers anchored to those deflated self-assessments, the final performance scores mirrored the demographic confidence gap. The system effectively penalized employees for humility while rewarding those who aggressively overstated their contributions, baking demographic disparities directly into the company's compensation structure.[4]

To interrupt this specific bias, organizational psychologists recommend a strict sequencing firewall. Managers must complete and lock their independent evaluations before they are permitted to view the employee's self-assessment, ensuring that the manager's score is derived from observation rather than an employee's self-doubt.[4]

The Contrast Effect and Team Size

Anchoring bias frequently interacts with the contrast effect, another cognitive distortion where a manager evaluates an employee against the previous person reviewed rather than against a fixed rubric. The preceding employee's score becomes a temporary anchor that skews the subsequent conversation.[4]

If a manager reviews a standout performer who hit 110 percent of their targets, that exceptional performance becomes the temporary anchor. The next employee, who may have delivered a solid 98 percent, is penalized simply for following a top-tier peer in the review schedule.[4]

This effect amplifies as the manager's span of control increases. The more ratings a manager must complete across teams of 25 to 150 people, the more they rely on peer-to-peer comparisons as a mental shortcut, abandoning objective criteria in favor of relative rankings.[4]

Order effects also contaminate the data. Employees evaluated first in a large review cycle consistently receive higher ratings than those evaluated last, as managers unconsciously anchor to a baseline of leniency before experiencing evaluation fatigue and becoming more critical as the cycle drags on.[4]

For employees in large teams, this means that the scheduling of the review conversation can impact the final score. Being pushed to the end of a manager's calendar increases the statistical probability of receiving a lower rating, entirely independent of the employee's actual work.[4]

Structural Solutions and BARS

Mitigating anchoring bias requires structural changes to the appraisal process, as awareness training alone fails to override the cognitive heuristic. The most effective intervention is removing the anchor entirely by hiding historical ratings within the human resources software during the drafting phase.[1][4]

Organizations are increasingly shifting to Behaviorally Anchored Rating Scales to force objective evaluation. Instead of asking a manager to rate an employee on a generic 5-point scale, this method ties each rating level to specific, observable actions that must be documented and verified.[2][4]

By requiring managers to document how many times an employee demonstrated a specific behavior, the system grounds the rating in recent evidence. This shifts the cognitive load from subjective memory to objective frequency counting, making it much harder for an anchor to survive.[4]

While Behaviorally Anchored Rating Scales require significant upfront investment to define the behavioral examples for every role, they dramatically improve consistency across different managers and strip the historical anchor of its power by demanding proof for every single rating level.[2][4]

Calibration sessions serve as the final structural defense. Before ratings are finalized, managers defend their scores in peer groups, forcing them to justify their evaluations with concrete examples rather than relying on past precedent or leaning on a deflated self-assessment.[4]

The Financial Implications of Bias

The financial consequences of uncorrected anchoring bias compound over the course of a career. A suppressed rating in a single cycle reduces that year's base salary increase by an average of 2.5 percent, which lowers the compounding baseline for all future raises.[3][4]

Furthermore, talent management systems often use historical ratings to filter candidates for leadership development programs. An employee trapped by a mediocre anchor may be systematically excluded from high-potential tracks, permanently altering their career trajectory because of a single biased evaluation years earlier.[4]

For the enterprise, anchoring bias results in misallocated compensation budgets. When managers overpay declining performers anchored to past glory and underpay rising talent anchored to past struggles, the organization loses its most effective retention tool and risks losing its highest actual producers.[4]

Ultimately, a performance appraisal system is only as valuable as the accuracy of its data. By severing the link to historical ratings during the evaluation phase, companies can restore the integrity of their talent metrics and ensure that employees are paid for the work they are doing today.[2][4]