HBR: managers who go easy on weak performers may be making a rational call. 5 ways to make reviews fair enough to be honest.

A Saturday question for People leaders and CPOs in Dublin, London, Amsterdam, Stockholm and Tallinn: in your last review cycle, how many people sat in the middle rating? And how many of your managers would admit they rated someone higher than they deserved, simply to avoid the fallout?

An HBR article from May, “Why Leaders Should Let Minor Mistakes Slide”, argues that this habit isn’t always a management failure. The piece was written by Avery Forman, a senior associate editor at Harvard Business School Working Knowledge, and features the work of HBS faculty member Henrique Castro-Pires. HBR’s summary is direct: many managers are criticised for showing leniency in performance reviews for low performers, but new research finds that may actually be a rational response to a costly problem, employee retaliation. The full text is published free on HBS Working Knowledge as “When Candor Hurts the Business: The Economics of Employee Retaliation” (1 May 2026), and the facts below come from those two pages and HBR’s follow-up management tip (19 May 2026).

1. What the research actually says

In plain language:

• Retaliation is real, and mostly quiet. The examples include gossiping, working slower, talking back and damaging products. The research cites a 2022 study of a US call centre where workers responded to a pay cut by issuing more customer refunds: refunds rose by 5.8 percentage points, reducing net sales by about $11 an hour. Castro-Pires notes it’s hard to measure because most employees don’t admit to it and most workplaces don’t admit it’s happening.
• Lenient, bunched-up ratings can be a choice. Critics fault reviews for leniency (inflating weak employees’ ratings) and centrality (failing to reflect real variation in performance). Castro-Pires’ model treats both as features: managers expect tough criticism to trigger retaliation, so they avoid poor reviews. As he puts it, reviews “don’t reveal a fine level of difference in performance. But that might be optimal, even from the perspective of the firm itself.”
• Pay is the trigger. In his model, employees retaliate when their pay is lower than they expected, and adding negative feedback on top of the pay hit can make the reaction worse.
• Every option has a cost. “Sticks and carrots” (bonuses and pay cuts) measures effort more precisely but invites some retaliation, and works best for short-term assignments. “Carrots only” rewards strong work and overlooks weaker work, but poor performance becomes hard to tell apart from average, so “you’re paying a little bit more for the same performance.”

Castro-Pires’ suggestions for managers are to let small issues slide, decouple pay from negative feedback, and clarify the reasons for a poor review: “A lot of retaliation comes from this sense of unfairness. It’s important to tell employees what criteria they’re being evaluated on.” For HR teams, he suggests leaning on rewards and using pay cuts sparingly, and asking managers and employees, anonymously, whether they have seen or experienced retaliation. HBR’s tip sums up the risk: when employees feel criticised, unfairly evaluated or financially penalised, frustration can spill into disengagement, resentment or subtle retaliation that hurts the broader team.

Honest caveats. This is a theoretical model, not a survey or a field experiment, so it explains why ratings might bunch up rather than measuring how often it happens. The examples come from the US and Brazil, not from European companies, and the dollar figure is from one call-centre study. The full HBR article sits behind HBR’s paywall; I have only used what is on the HBR page, the free HBS Working Knowledge original it was produced from, and the HBR tip. It’s a useful lens, not a benchmark.

2. The mid-market hurdle: when the bar is vague, honesty feels personal

In a 200–600 person company the pattern is familiar. Most people land on “meets expectations”, a handful get top marks, and almost nobody gets a low rating, even when everyone in calibration knows who is struggling. The People team then has to make pay and promotion calls from ratings that don’t separate anyone, and the managers who did rate honestly feel exposed.

Part of the reason is the one Castro-Pires names: unfairness. If the criteria live in a manager’s head, or in a competency document nobody has opened since it was written, a low rating looks like an opinion about the person. Tie that rating straight to a smaller pay rise and you have exactly the conditions his model describes. Managers aren’t being weak when they soften the rating; they are reading the room.

3. 5 ways to make reviews fair enough to be honest

1. Publish the criteria before the cycle opens. Castro-Pires’ clearest point is that people should know what they are evaluated on. Share the scorecard, the skills expected at each level and the company values with every employee weeks before self-reviews start, not on the day of the conversation. The talent bar by seniority post and the competency matrix templates are a starting point.
2. Describe behaviours, not personalities. A criterion like “raises risks to stakeholders before deadlines slip” can be checked against evidence; “good communicator” can’t. Observable behaviours make a low score easier to explain and harder to read as personal. Run your current form through the free performance review bias checker to spot vague wording.
3. Separate the improvement conversation from the pay conversation. His advice is to decouple pay from negative feedback where you can: “What I’m trying to do here is help you improve.” In practice that can mean a development conversation first and the compensation discussion in a separate meeting, after calibration.
4. Decide what counts as a minor slip, and let it go. Agree in advance with managers what is worth formal criticism and what is coaching in a 1:1. Something only slightly below average, once, doesn’t need to appear in the review. Serious or repeated gaps still do, with a clear plan.
5. Reward strong work clearly, and ask about retaliation. Lean on recognition and rewards for strong performance, and use pay cuts sparingly, as Castro-Pires suggests. Add one anonymous question to your next pulse survey: “Have you seen or experienced retaliation after a performance review here?”

For the PDF, use your work email. The call is a free 15-minute review-cycle teardown: bring your review form and last cycle’s rating spread.

4. Where Revolut People fits, and where it doesn’t

Revolut People started as the system Revolut built to run performance as it scaled to more than 10,000 people across 60+ countries. The part that maps to this research is one thing: the three-pillar review scorecard. Every review uses the same structure: Goals (what someone delivered), Skills (how they did it, taken from a role competency matrix by seniority, with managers answering Yes or No on defined, observable behaviours) and Values (assessed company-wide). Because the matrix is set per role and level, the criteria an employee is reviewed against can be shared with them before the cycle, and each scorecard opens with their goals and previous-cycle data already filled in.

To be straight about the limits: a scorecard won’t stop a manager from softening a rating, and it won’t remove the tension when a review affects pay. That still depends on how you run the conversation and how you separate development from compensation. What a shared, behaviour-based structure does is give managers and employees the same written bar to point at. Related reads: HBR on feedback that crosses the line, Gartner on pay for performance, and publishing a talent bar by seniority.

Bottom Line

HBR’s May piece on Henrique Castro-Pires’ research says managers who go easy on weak performers may be making a rational call, because retaliation is costly and “a lot of retaliation comes from this sense of unfairness.” For a mid-market People team in the UK, Ireland, Benelux, the Nordics or the Baltics, the answer isn’t to push managers to be harsher. It’s to make the bar visible: publish behaviour-based criteria before the cycle, separate development from pay, agree what counts as a minor slip, and reward strong work clearly. Try it with one department before your next cycle.

FAQ

What did the HBR article ‘Why Leaders Should Let Minor Mistakes Slide’ say?

The article, by Avery Forman (11 May 2026), was produced by Harvard Business School Working Knowledge and features HBS faculty member Henrique Castro-Pires. It says many managers are criticised for showing leniency in performance reviews for low performers, but new research finds this may be a rational response to a costly problem: employee retaliation. Castro-Pires’ advice to managers is to let small issues slide, decouple pay from negative feedback, and clarify the reasons for a poor review, including the criteria people are evaluated on.

What are ‘leniency’ and ‘centrality’ in performance reviews?

As HBS Working Knowledge describes them, leniency is inflating weak employees’ ratings for inferior work, and centrality is failing to reflect actual variations in employee performance. Castro-Pires’ model treats both as a choice managers make when tough criticism may trigger retaliation, rather than simply as rating errors.

Is this based on a survey of companies?

No. The research is a theoretical model, published as “Monitoring, Performance Reviews, and Retaliation” in Management Science (February 2026). The examples of retaliation in the coverage (a chocolate factory in Brazil, a 2022 study of a US call centre, a 1972 book) illustrate the idea; they are not a measure of how common retaliation is in European companies. Treat it as a useful lens on why ratings bunch up, not as a benchmark.

Does this mean managers should stop giving negative feedback?

No. Castro-Pires’ suggestions are narrower: let slips that are only slightly below average go, keep the improvement conversation separate from pay where you can, and explain the criteria behind a poor review. Reward strong performance clearly and use pay cuts sparingly. Serious or repeated gaps still need a direct conversation and a clear plan.

Is Revolut People available in the UK, Ireland, Benelux, the Nordics and the Baltics?

Yes. Revolut People is available to businesses with a registered entity in the UK, the EEA or the US, which covers Ireland, the Benelux, Nordic and Baltic countries.