Research Library Article


How Performance Appraisals Mislead

David Earle

Performance Summary
A seven year study of performance appraisals, what they do and how they work, including their associated employment outcomes. 

Employers hire job candidates because of the potential fit between their talents and aptitudes and the work being offered. Pre-hire appraisals (tests, interviews, background checks, references…) are supposed to help determine that probability. Post-hire measurements, i.e. performance appraisals, tell employers whether or not they did. Unfortunately, both appraisal processes are sufficiently imperfect that retention remains a chronic annoyance, and significant financial drag, in most enterprises.

How can competent managers make so many mistakes? Researchers have been pursuing that answer for some time. Already they have documented the many hidden biases that color the judgement of even the best hiring managers. And they have discovered how much the appraisal process varies from manager to manager and from department to department, even in well managed, successful companies.

In this paper, Peter Cappella and Martin Conyon examine what performance appraisals actually do (as opposed to what we think they do) and how well they work as a HCM tool. Overall, their seven year study found that performance appraisals do work as a large-mesh — meaning fairly crude — talent filtering device. Very poor scorers do tend to get fired, low-average scorers do tend to quit, and the high scorers do get rewarded. 

But that broad-brush picture quickly becomes nuanced.

  • Appraisals are generally either about rewarding past performance, as a contract would, or about an ongoing work relationship focused on improvement. The authors found that they were more about the latter. But outcomes were not linear, for example a 10% pay increase for a 10% increase in performance. The common assumption that managers tend to clump merit pay increases in the middle turned out to be incorrect. In reality, the best performers and the worst performers were rewarded or penalized disproportionately.
  • The assumption that good past performers will be good future performers, and bad ones forever bad, turns out to be another incorrect assumption. Appraisals of past performance turned out to be poor indicators of future performance. In only one-third of the cases could last year’s scores predict next year’s.
  • “It’s all about the manager” is a third incorrect assumption. This research showed that changing managers doesn’t consistently affect performance scores.


Everywhere in HCM practice today we see new research that upends yesterday’s assumptions and practices. The takeaways here are that the appraisal systems we have created in the past to assess talent are not producing the results we want. Biases are pervasive, forced rankings cause distortions in managers’ and employees’ behavior, annual performance appraisals and linear reward schemes are insensitive and indiscriminate, and the assumption that most employees will perform in the future the way they have in the past is simply incorrect.

Research  Sources