6 Ways to Make Performance Reviews More Fair

Many studies have exposed the degree to which bias is a factor in the performance-review process. Because the criteria for evaluation are so often vague and open-ended, it’s dangerously easy for patterns of bias to creep into the process and for managers to be guided by implicit biases. Expectations are often gendered; different standards of behavior apply to different groups of people; feedback can reflect negative stereotypes. All of these factors can lead to inappropriate assessments of performance, which in turn can prompt talented employees to leave — especially when they already have their eyes on the door, as is so often the case today, given how fierce the competition for talent is. Companies simply can’t afford to keep a review system in place that is biased against certain employees, misrepresents their skills and abilities, and even prompts them to seek jobs elsewhere.

In my academic work as a behavioral and data scientist, I’ve studied the problem of performance-review bias extensively, and in my work as a consultant I’ve applied what I’ve learned to help companies “de-bias” their review process. This work has allowed me to identify six behavioral nudges — four for managers and two for employees — that will make the process fairer.


1) Consider practice scenarios

Try this nudge about a week before employee performance reviews. Guide groups of six to eight managers through multiple hypothetical assessment scenarios. These scenarios can be based on tweaked, anonymized files of real employees. Have the managers review the files, make their own assessments, and then have the groups share and discuss their reviews.

To take the exercise up a notch, have people compile their assessments before the group discussion, and present how they compare to one another on a whiteboard or slide deck for all to see during the exercise. For each employee assessed, does everybody agree on their assessments of success and underperformance? If not, why not? What sorts of outside factors might be affecting how employees have performed, and what sorts of behavioral contexts affect how managers assess performance? More specifically, ask managers to reflect on the following questions:

• How and when does this employee contribute? Are their contributions valuable?

• What was their output compared with what they were asked to deliver?

• What specific data and observations justify your assessment?

• Would you have given the same feedback if the employee had been a person of a different color and/or a different gender?

2) Contextualize women’s absence

During the past couple of years, as parents have found themselves working at home with their children, many have had to adopt flexible hours — and in their annual reviews, women have been more frequently criticized for doing this than men. To help managers neutralize this sort of bias, make sure that managers ask themselves these questions when they’re making their review assessments:

• Was an employee’s absence due to a personal problem? If so, what did you do to help them find a solution?

• When employees were online and working, how did you feel about their performance and contribution?

• If you had interacted in person with this employee, how might your assessment have been different? In other words, how much has working remotely affected not only them but also you?

3) Use “calibration committees”

A calibration committee meets before ratings are disseminated to examine assessment scores and make revisions as need be to help ensure standardized scores across an organization. Large companies will require multiple such committees, in which case an overseeing calibration committee should be named. To bring as many views to the table as possible, each committee should have members who span generations, genders, races, functions, and time spent at the company. A calibration committee can convene every six months, but each month the representatives should also each allocate several hours to review new data and ensure that the calibration is consistent. For example, if a new employee has one month of data when the calibration committee met at the beginning of the year, as more data comes in managers can ask themselves if they made the right calibration at the time of the committee meeting and can make changes as necessary.

4) Focus on in-group belonging

People are drawn to individuals who are like them, often at the expense of those who are not. This kind of bias, sometimes referred to as similarity bias or like-me bias, generates in-groups and out-groups, often based on such superficial proxies as gender, skin color, and religious affiliation. Managers tend to regard themselves as inclusive or neutral, but nonetheless they tend to favor in-group employees and not fully appreciate the work and skills of out-group employees. Because white men continue to dominate upper management in most companies, they tend to get preferential treatment in the review process.

To work against this bias, try asking leaders to find something in common with employees before conducting their reviews. They might begin by discussing a work event where both participated, or by asking employees about some company “good news” that affects everybody. These may seem like minor acts, but they can lead to important changes, because they can help to create an in-group and create some psychological certainty upon which the review can then be built. The result: Instead of focusing — consciously or unconsciously — on in-group affiliation, managers can focus more fairly on ideas and achievements. And if they regularly work in this way to identify common ground with employees throughout the year, in-group belonging can become a key element of company culture.


1) Keep a journal

Recency bias affects employees and managers alike in the review process. The idea is simple: When people look back at the year and try to take stock, what they remember best — and what they therefore focus on most — is what has happened most recently. The brain ascribes predominance to actions and events that are easier to recall. Managers fall prey to this kind of bias, because the performance that they’re reviewing is not their own. But employees fall prey to it too. It’s easier and more natural than you might think to forget the details of tasks that you performed weeks or months ago — and then, lacking this information, to assess your own performance in ways that undervalue or misrepresent your performance.

To counter recency bias, suggest that employees keep a weekly journal in which they document their activities, record the feedback they receive from managers, and so on. They can even make these journals accessible to managers, so that they’re able to see and address problems and concerns as they arise. But employees who don’t want to share access to their journals obviously don’t have to. They’re a private tool, to be used however employees see fit.

Journals of this sort can be extremely helpful at review time, because they allow employees to reconstruct months of activity and document their performance in real time across the whole review period. In general, if employees keep good journals and share them with their managers, this can lead to fewer surprises and fewer instances of miscommunication at review time, so that both parties can focus on prospects for the future rather than spending time trying to patch together the past. Employees can also revisit their journals and highlight their achievements at promotion time.

2) Review your job description

Surprisingly often, employees underrate themselves on self-evaluations. Research has shown that this is particularly true for underrepresented groups — notably, women and people of color.

One helpful way of working against this kind of self-imposed bias is to provide a context anchor — that is, a standard against which peers’ behaviors, activities, and results are judged. Employees can do this during the review process by reviewing their job descriptions (the context anchor) before embarking on their self-reviews, with an eye to understanding how the work they’ve done during the review period aligns with what is officially expected of them. This exercise injects transparency and fairness, because it makes employees take stock of what they’re being asked to do, and it forces managers to think harder about evaluating their employees based on how they’re performing the jobs they were hired to do.

This nudge, too, allows a forward-looking focus on how employees are growing in their jobs and expanding their duties and responsibilities. After managers and employees assess their performance based on their actual job descriptions, they can then discuss the ways employees have met unexpected challenges, how consistently they have gone beyond their written responsibilities to achieve new team and department goals, and how creatively they’ve generated new ideas or collaborated across traditional divides. To the extent that they’ve done these things well, they can be acknowledged in their reviews and perhaps recommended for a raise or a promotion.

 * * *

To be effective and fair, performance reviews require intentionality and preparation from both managers and employees. The six nudges described in this article can help on both fronts — in ways that will help everybody in your organization achieve their goals.