The Skill That Earns a Promotion Isn't the Skill a Promotion Requires

The Best Performer Is Often the Wrong Promotion

Organizations overwhelmingly promote based on demonstrated current-role performance, a criterion that feels objective, defensible, and fair because it's directly observable and easy to justify to everyone involved, including the person not promoted. It also feels like the natural extension of a meritocratic principle most organizations claim to operate by: reward the person who has earned it, and let that reward include greater responsibility. Peter and Hull's (1969) original formulation of what became known as the Peter Principle proposed that in a hierarchy, every employee tends to rise to their own level of incompetence, because promotion continues as long as performance remains strong, and stops only once someone reaches a role their actual capability can't sustain.

For decades, this was treated as a wry observation about organizational life, popular enough to become a piece of common business vocabulary, but rarely treated as a documented empirical pattern worth building policy around. It was cited more often in casual conversation, usually about a specific frustrating manager, than in serious workforce planning conversations about how promotion decisions actually get made. Benson, Li, and Shue's (2019) large-scale empirical study changed that considerably. Using detailed performance data on sales workers across 131 firms, the researchers found direct, quantified evidence consistent with the Peter Principle: firms systematically promoted based on current sales performance, and the workers promoted on that basis frequently underperformed as managers relative to workers with weaker sales records but stronger indicators of managerial potential.

The mechanism the researchers identified is not organizational carelessness or a simple failure of judgment. It's a rational, if flawed, response to a genuine measurement problem: current-role performance is directly observable and easy to defend as a promotion criterion, while managerial potential requires a separate, deliberate assessment that most organizations have never actually built or invested in constructing. The researchers estimate that the resulting costs, in the form of weaker subsequent team performance under the newly promoted manager, are substantial, not a marginal inefficiency that gets easily absorbed elsewhere in the organization's overall results.

The Skill Mismatch Behind the Pattern

The core problem the research identifies is not that strong performers are somehow undeserving of promotion or advancement generally. It's that the skills producing strong individual performance in a given role are frequently different from, and sometimes in genuine tension with, the skills a managerial role actually requires of someone occupying it. A skilled individual contributor is often skilled precisely because of deep, specialized focus and a level of personal ownership over their own output that management, which requires distributing responsibility and coordinating other people's work rather than doing it all oneself, actually works directly against.

Benson, Li, and Shue found that this mismatch produces real, measurable costs, not just for the newly promoted manager personally, but for the entire team they now lead, in the form of lower subsequent sales performance for the whole group relative to teams led by managers who were selected using different criteria. The researchers also found that firms partially, though not fully, adjust for this dynamic in practice: when a managerial role carries meaningfully more responsibility, firms place somewhat less weight on current sales performance in the promotion decision, suggesting some genuine organizational awareness of the mismatch even without a fully developed, systematic solution to it.

The uncomfortable implication is that a promotion process built primarily around rewarding current performance is optimizing for exactly the wrong variable in exactly the moment it matters most: the transition into a fundamentally different role requiring a meaningfully different set of skills, judged by criteria the organization has never actually separated from the ones that produced the original promotion decision. The person being promoted didn't do anything wrong; the process asked the wrong question of them.

A Genuine Competing Explanation Worth Taking Seriously

Not every economist reads the Peter Principle the same way Benson, Li, and Shue do, and the disagreement is genuinely instructive rather than a simple contradiction to be dismissed. Lazear's (2004) influential theoretical paper, published in the Journal of Political Economy, argues that at least part of the observed decline in performance after promotion is a statistical artifact of regression to the mean, not necessarily evidence that firms are making a genuine misjudgment about who actually deserves promotion.

The logic runs as follows: an employee's measured performance in any given period reflects both a stable, underlying ability and a transitory component, essentially good or bad luck within that specific window of time. Someone promoted on the strength of an unusually strong period likely benefited from a favorable transitory draw as well as genuine underlying ability, and since that transitory luck doesn't carry forward into the next period, their subsequent measured performance regresses toward their true, underlying ability level, which looks like real decline even if no genuine mistake was made in the original promotion decision.

Lazear's model shows that firms behaving fully rationally, and fully aware of this regression effect, will still observe apparent post-promotion decline as a mathematically necessary consequence of any promotion rule based on a noisy performance signal, not evidence that the underlying rule itself is broken or poorly designed. His analysis further suggests that well-calibrated firms should respond by deliberately inflating their promotion threshold, requiring performance meaningfully above the bar the role actually demands, precisely to offset the predictable regression that reliably follows any promotion decision made on a noisy signal.

This doesn't cancel out Benson, Li, and Shue's empirical findings; it complicates the interpretation in a genuinely useful way. Their data still shows firms could meaningfully improve managerial outcomes by weighting non-performance indicators more heavily in the decision, a finding regression to the mean alone doesn't fully explain away. But Lazear's framework suggests part of the fix isn't only building a new potential-assessment system from scratch. It's also being honest that a raw current-performance threshold, applied without adjustment for known regression effects, will systematically overpromote people whose apparent excellence was partly a matter of favorable timing rather than durable underlying capability.

Why Organizations Keep Making the Same Choice Anyway

Whichever explanation carries more actual weight in a given organization, and both likely operate simultaneously to varying degrees, the practical persistence of performance-based promotion has a structural explanation that goes beyond simple oversight or a failure to know better. Current performance is genuinely easier to measure, easier to defend in an actual promotion conversation with the people involved, and easier to build organizational consensus around than a genuine, structured assessment of managerial potential, which requires real investment in something that can't simply be read off an existing performance dashboard everyone already trusts.

Promotion also functions in practice as a compensation mechanism as much as it functions as a pure placement decision. Organizations use the prospect of promotion to incentivize strong current-role performance in the first place, which means shifting the promotion criterion meaningfully away from performance would require rethinking how the organization motivates people in their current roles, not just rethinking how it selects the next manager once a role opens up. This creates a genuine tension rather than a simple, easily corrected oversight: performance-based promotion criteria may be doing real, valuable incentive work throughout the organization even while producing measurably weaker managerial outcomes on average, which means the fix isn't as simple as switching criteria outright and hoping the incentive effects take care of themselves.

What a More Evidence-Based Approach Actually Requires

The evidence, taken together across both explanations, points toward organizations needing two genuinely distinct systems operating in parallel rather than one conflated process trying to answer two different questions at once: a system that continues to reward and incentivize strong current-role performance, calibrated with real, explicit awareness that raw performance thresholds will systematically overpromote people who benefited partly from favorable timing, and a separate system that identifies and develops managerial potential based on criteria that are actually distinct from current-role output.

Building a genuine potential-assessment process means identifying specific, observable indicators, how someone handles delegating work rather than insisting on doing it all themselves, how they communicate direction and context to others rather than simply executing tasks alone, how they respond to being accountable for someone else's output rather than only their own, that are meaningfully distinct from the metrics that make someone a strong individual performer, and that can genuinely be observed before someone is actually placed in a managerial role rather than discovered only afterward, at real and often lasting cost to the team now depending on them.

Organizations that get this right are not the ones that have abandoned performance-based reward entirely, which the evidence doesn't actually support doing. They are the ones that have stopped treating current performance and managerial potential as though they were the same underlying question, calibrated their promotion thresholds with genuine awareness of regression effects rather than trusting a single number at face value, and built the separate, deliberate assessment process the evidence, under either competing explanation, consistently shows managerial potential actually requires.