Performance management bias is a systematic error in how managers rate, coach, and promote employees, driven by mental shortcuts instead of observed results. It shows up as inflated scores, vague feedback, and uneven career opportunities. Structured criteria, ongoing documentation, and calibration reduce it, though no process removes it completely.
Here is why this matters. In one SHRM-published study, only 15% of women and 24% of men managers had confidence in their company’s performance evaluation process, and most viewed it as subjective and highly ambiguous. If the people running reviews don’t trust them, employees won’t either.
Managers also carry a lot of weight. Gallup estimates that managers account for at least 70% of the variance in employee engagement scores across business units. That is variance in scores, not a guarantee, but it shows how much one manager’s judgment shapes a team’s experience.
Below are 12 biases that quietly distort reviews in US workplaces, with an example and a fix for each.
Performance Review Biases at a Glance
Bias
What it looks like
Fastest fix
Central tendency
Everyone lands in the middle
Rate against role criteria with evidence
Leniency
Scores run too high to avoid conflict
Require examples for every high rating
Severity
Scores run too low
Calibrate against peer managers
Halo
One strength lifts every rating
Score each competency separately
Horn
One flaw drags every rating
Score separately and add multi-rater input
Recency
The last few weeks dominate
Keep a running log all year
Primacy
First impressions stick
Review the full period at set intervals
Similar-to-me
Favoring people like you
Use behavior-based criteria
Confirmation
Noticing only evidence that fits your view
List disconfirming examples
Stereotype
Judging by group assumptions
Use anchored rubrics and audit results
Proximity
Favoring who you see most
Rate outcomes, not visibility
Contrast
Judging against the last person, not the standard
Review against criteria, and vary review order
1. Central Tendency Bias: The Middle Ground Trap
Central tendency bias occurs when managers rate all employees in the middle or “satisfactory” range, regardless of their actual performance. This often happens when a manager evaluates many employees and unconsciously starts giving similar scores to everyone to avoid making difficult judgments.
Example: Imagine a manager overseeing 20 employees. Instead of recognizing individual strengths and weaknesses, the manager gives nearly all of them a “satisfactory” rating. This not only demotivates top performers who aren’t recognized for their efforts but also overlooks underperformers who need improvement.
How to Avoid It
To prevent central tendency bias, focus on each employee’s individual performance and how it aligns with the expectations of their role. Collect performance data at multiple points throughout the year rather than relying on a single evaluation period. Aligning performance with OKRs and goals makes evaluations more objective.
This provides a more comprehensive view of each employee’s contributions and ensures that standout performers receive the recognition they deserve.
Leniency and severity are related rating errors, so let’s look at those next.
2. Leniency and Severity Bias: The Extremes of Evaluation
Leniency bias occurs when a manager rates all employees too positively, while severity bias happens when the manager rates all employees too harshly. Both extremes can distort the performance review process and lead to frustration.
Example: A manager who wants to avoid conflict may give all employees high marks (leniency bias), even when some are clearly underperforming.
Conversely, a manager trying to motivate employees might rate everyone low (severity bias), hoping that tough evaluations will encourage improvement.
Unfortunately, leniency bias creates a false sense of accomplishment, while severity bias can lead to disengagement.
Removing the middle option helps, but it won’t fix things on its own. Without shared standards, managers just pick “slightly above” for everyone. Pair the scale with calibration sessions, where managers compare ratings across teams and explain the evidence behind each score before ratings are final.
How to Avoid It
Quick check before you submit ratings. If most of your team landed in the same rating band, pull the written evidence for each person and confirm it supports the score. If you can’t point to specific work, the rating is probably a default.
To counter these biases, establish clear evaluation criteria and use a consistent rating scale. If using a 5-point scale, consider eliminating the middle or neutral option, forcing managers to make a definitive judgment about performance.
By creating distinct rating categories, managers are encouraged to think critically about each employee’s achievements and areas for improvement.
3. Halo and Horn Bias
Halo and horn bias occur when managers allow a single trait or characteristic of an employee to disproportionately influence the entire performance review.
Halo Bias happens when a manager gives an employee an overly positive evaluation based on one strong trait, such as their punctuality, or even unrelated factors like supporting the same sports team. This singular focus can overshadow areas where the employee may need improvement.
Horn Bias is the opposite, where a manager gives an employee a negative review based on one disliked trait or past mistake, even if the employee excels in other areas. This bias can manifest as a result of personal preferences or even unconscious discrimination, such as sexism or racism.
Why It’s Problematic: No employee is perfect, and focusing on just one aspect of their performance—whether positive or negative—means overlooking other key contributions or challenges. This can lead to unfair evaluations, with high-performing employees going unrecognized or employees being penalized for one-off issues.
How to Avoid It
To avoid halo and horn biases, managers need to adopt a structured and objective performance evaluation process. The most reliable fix is structure. Score each competency or goal separately, using the same questions for every employee, so one standout trait or one bad week can’t set the tone for the whole review.
Evaluating employees across multiple metrics ensures that no important qualities are overlooked, and it helps uncover faulty logic, such as cherry-picking evidence to fit a preconceived conclusion. Many organizations also use multi-rater feedback to ensure broader evaluation inputs.
4. Recency and Primacy Bias: The Influence of Time
Recency bias occurs when a manager focuses primarily on the most recent work or interactions they remember with the employee, allowing these events to overshadow their overall performance throughout the evaluation period.
For example, if an employee closed a big deal just before the review, they may receive a high rating, even if their performance was inconsistent or underwhelming earlier in the year.
Primacy bias, on the other hand, is the tendency to give more weight to an employee’s initial performance, often overlooking their more recent achievements or struggles.
A manager might continue to rely on their first impressions of an employee’s past successes or failures, regardless of their current work.
The spillover effect also plays a role here. This happens when a manager assumes that an employee’s past performance trends – whether positive or negative – are continuing without thoroughly evaluating recent work.
For instance, if an employee has consistently performed well in the past, their manager might assume they’re still doing well and neglect to carefully assess their recent contributions.
Why It’s Problematic: Both biases skew the accuracy of evaluations, leading to unfair assessments. Recency bias can result in overrating short-term successes, while primacy bias can lead to outdated assessments that don’t reflect an employee’s current abilities or efforts.
How to Avoid It
Make it a habit instead of a project. Spend five minutes at the end of each week or project noting one win, one miss, and the context. At review time, you’re reading a record instead of trusting your memory.
To reduce the impact of recency and primacy biases, managers need to assess performance over the entire evaluation period, not just based on recent or early impressions.
Documenting regular real-time feedback throughout the year and reviewing an employee’s contributions at multiple intervals ensures a more balanced and fair evaluation. When employees work in teams, be sure to evaluate their contributions to get an accurate picture of their performance.
By maintaining a structured, consistent review process, managers can prevent these biases from skewing the performance appraisal and ensure that evaluations reflect an employee’s true capabilities over time
5. Similar-to-Me Bias: Liking What’s Familiar
Similar-to-me bias occurs when managers give higher ratings to employees they perceive as being similar to themselves, whether in terms of interests, personality, or background.
Example: A manager who shares a hobby or alma mater with an employee might give them a better evaluation because they feel a connection, even though other employees are performing just as well, if not better.
How to Avoid It
To reduce similar-to-me bias, set specific, measurable criteria for performance evaluations. Ensure that all employees are assessed based on their achievements and contributions rather than personal connections or similarities.
Using a structured, objective review process helps maintain fairness and focus on performance rather than personal traits.
6. Confirmation Bias & Stereotype Bias
What it is
Confirmation bias is when evaluators favor information or examples that confirm their existing beliefs about an employee, ignoring evidence to the contrary.
Stereotype bias involves allowing stereotypes (e.g. gender roles, age, race) to shape judgments of performance rather than actual behaviors. Role congruity theory shows how women in leadership are judged more harshly for traits viewed as “not fitting” stereotypes.
Why it’s problematic
Employees may be under-credited because of preconceived notions, or praised unfairly for confirming the manager’s expectations.
Stereotype bias can lead to systemic inequities in ratings, promotions, and development.
How to avoid it
Encourage evaluators to list disconfirming examples (instances that contradict prior impressions).
Use structured rubrics and behavioral anchors for rating, rather than relying on subjective impressions.
Include diversity, equity & inclusion training that surfaces unconscious stereotypes.
Rotate or anonymize parts of evaluation when possible (e.g. blind peer assessments) to reduce identity cues.
7. Proximity Bias
Proximity bias is the tendency to favor the people you see and talk to most. On hybrid and remote teams, that often means in-office employees get more credit, more stretch assignments, and higher ratings, even when remote colleagues deliver the same results.
How to avoid it
Judge outcomes against written, measurable goals, no matter where someone works.
Ask every employee to submit a short summary of their results before the review.
Compare ratings for remote and in-office employees in the same role.
8. Contrast Effect
Contrast effect happens when you rate someone against the last person you reviewed instead of against the standard. A solid performer looks average after a star, and an average performer looks great after a struggling one.
How to avoid it
Review against written criteria, not against peers.
Rate one competency across all employees before moving to the next.
Change the order you review people in.
Bias in AI Performance Tools in 2026
AI can make reviews more consistent, but it can also repeat and scale existing bias. A tool trained on past ratings learns whatever patterns those ratings contained. Treat AI output as one input to a manager’s judgment, never the final decision.
Where bias creeps in
Historical data bias, where models trained on biased ratings repeat unfair patterns.
Feature bias, where the metrics tracked favor certain roles, styles, or work modes, such as remote versus in-office.
Context blindness, where the tool misses personal circumstances or team dependencies and penalizes people unfairly.
What to do about it
Audit outputs regularly by gender, age, role, and location.
Keep a named human accountable for every rating.
Ask vendors how the model was built and tested for bias.
Tell employees when AI informs their review, and give them a way to question the result.
The US legal picture is also moving. Colorado’s governor signed SB 26-189 on May 14, 2026, replacing the earlier Colorado AI Act, with employer obligations scheduled to begin January 1, 2027. According to Proskauer’s summary, employers using covered automated tools to materially influence employment decisions would face notice, disclosure, recordkeeping, and review obligations. Other states and cities have their own rules, so check where your employees work and talk to counsel.
A Simple Process to Reduce Bias All Year
Awareness alone doesn’t fix bias, but consistency helps. In one site covered in a SHRM-published viewpoint, applying criteria consistently narrowed gender gaps in ratings and removed the overrepresentation of men in the top performance category and women in the middle. That was one organization, so treat it as encouraging, not universal.
Define criteria and rating anchors before the review cycle starts.
Collect evidence continuously through check-ins, notes, and goal progress.
Add more than one perspective, such as peer or multi-rater input.
Hold calibration sessions before ratings are final.
Audit the results by group, then fix the process, not just the individual rating.
Final Thoughts
Bias in performance reviews is normal, which is exactly why process matters. Clear criteria, ongoing documentation, multiple perspectives, and regular calibration give managers something sturdier than memory and gut feel. The payoff is fairer ratings, better conversations, and teams that trust the review process.
What are the most common types of bias in performance reviews?
The ones HR teams cite most often are central tendency, leniency, severity, halo, horn, recency, primacy, and similar-to-me bias. Confirmation, stereotype, proximity, and contrast effects also show up, especially on hybrid teams. Most of them share the same root cause, which is managers relying on memory and gut feel instead of documented evidence.
How can managers reduce bias in performance reviews?
Use written criteria for every role, collect evidence throughout the year, add input from more than one rater, and hold calibration sessions before ratings are final. Then compare rating patterns across groups to spot problems. Awareness helps, but structure does most of the work.
What is the difference between the halo effect and the horn effect?
The halo effect lets one strong trait lift the whole review. The horn effect does the opposite, letting one flaw or past mistake pull everything down. Scoring each competency separately is the simplest fix for both.
How do you prevent recency bias in performance reviews?
Keep a running record of wins, misses, and context all year, and look back over the full review period before rating anyone. A five-minute weekly note gives you evidence to draw on instead of whatever happened last month.
Can AI remove bias from performance reviews?
Not on its own. AI can make scoring more consistent, but it learns from past data that may already contain bias. Audit results by group, keep a human accountable for final ratings, and tell employees how AI is used. Some states are also adding rules for automated decision tools in employment, so check with legal counsel.
Can biased performance reviews create legal risk for US employers?
Yes, they can. If ratings, promotions, or terminations track protected characteristics such as sex, race, age, or disability instead of documented performance, an employer may face discrimination claims under federal and state law. Written criteria and consistent documentation make reviews easier to defend. This is general information, not legal advice.
Srikant Chellappa is the Co-Founder and CEO at Engagedly and is a passionate entrepreneur and people leader. He is an author, producer/director of 6 feature films, a music album with his band Manchester Underground, and is the host of The People Strategy Leaders Podcast.
Innovative
