In talent reviews, a senior evaluator raises hedged doubts about high-potential people who are not politically aligned, and the doubts compound across cycles into a recorded pattern of concern.
An example
In the third cycle, the HR business partner who ran talent calibration for a fixed-line telecom operator's enterprise division noticed a habit. The senior vice president of enterprise sales would say, about certain high-potential managers, that he liked them a lot but had questions about how they handled ambiguity at the next level. Nobody in the room could test the remark, and it went into the notes. She pulled two years of notes and found the same kind of comment on five managers, all of whom had opposed his reorganization of the account teams, and on none of his own direct reports. Three of the five had dropped off the high-potential list. None had been told why. Their sales numbers were above the division average.
Why it works
Dropping the candidate looks easier because committees are risk-averse about high-potential bets: a wrong promotion is public, while a downgrade goes unseen. A credible voice framing the record as ambiguous makes the candidate seem a weaker bet, and keeping them becomes risky, since anyone who defended them will seem to have ignored a warning if anything later goes wrong. In a closed room where the candidate is absent and consensus rules, the doubt costs almost nothing to plant. It needs a senior voice, calibration norms that weight stated concerns over recorded results, no sponsor in the room, and concern language that resists rebuttal.
The thinking behind it
Daniel Kahneman Judgment under Uncertainty: Heuristics and Biases (1974), by Amos Tversky and Daniel Kahneman, describes anchoring: people estimate by adjusting from a starting value, and the adjustment is usually too small. A senior evaluator's hedged doubt, entered first, becomes the starting point from which a committee assesses the candidate, and later evidence moves the judgment less than it should. Recording each concern with its evidence gives the committee a better anchor.
Amos Tversky Loss Aversion in Riskless Choice: A Reference-Dependent Model (1991) extends loss aversion to ordinary choices: losses and disadvantages have greater impact than gains and advantages, though how general that asymmetry is remains disputed. A calibration committee that sees a wrong promotion as a loss weighs one planted doubt heavily against a strong record. A sponsor in the room restores weight to the other side, what the company gives up by passing the candidate over.
George A. Akerlof The Market for "Lemons": Quality Uncertainty and the Market Mechanism (1970) shows how a gap in information distorts a market: buyers who cannot verify quality discount everything on offer. In calibration the candidate never learns what was said, and the committee cannot easily check the doubt. Sharing recorded concerns with the candidate closes the gap from both sides.
Warning signs
Doubts that cannot be measured, about judgment, presence, readiness, or resilience, come from the same evaluator about several unaligned candidates across successive cycles. The concerns are recorded but never reach the candidates. Candidates drop off high-potential lists while their measured results hold steady or improve. The evaluator's own allies attract no comments of the kind.
How to defend yourself
Have a sponsor in the room who will dispute the framing, and brief them before each calibration with your results.
- Ask your manager after each cycle whether any concern was raised about you, and what evidence it rested on.
- Keep your record measurable and current, so calibration has numbers to anchor on.
- Push for a rule that any concern raised about a candidate is recorded with its evidence and shared with the candidate.
If you run calibration, pull the notes across several cycles and look for one evaluator raising the same kind of doubt about people whose only common trait is having disagreed with that evaluator.
How a leader shuts it down
Anchor calibration to objective results and require evidence for every concern raised. Share recorded concerns with the candidate and give them a chance to respond. Audit calibration notes across cycles for patterns by evaluator, and make sure each high-potential candidate under discussion has a sponsor or advocate present.