The halo effect is a rating bias in which one strong impression of a person, good or bad, spills over into how you judge their other, unrelated qualities. In HR it shows up wherever a human scores another human, most visibly in the annual performance appraisal and in hiring interviews.

Where the term comes from

Psychologist Edward Thorndike named it in a short 1920 paper in the Journal of Applied Psychology, “A Constant Error in Psychological Ratings”. He looked at army officers’ ratings of their subordinates on separate qualities such as physique, intelligence, leadership and character, and found the scores for one person correlated far more closely than real, independent traits plausibly would. Raters were forming one overall impression and then colouring every individual rating with it. The negative mirror image, where one weakness drags everything down, is usually called the horns effect. Both are different from recency bias, which is about timing (the last few weeks dominating the year) rather than one trait dominating the others.

How it shows up in Indian appraisals and interviews

• The engineer who closed one big client escalation in Q3 gets “exceeds” on documentation, mentoring and process discipline too, even though nobody checked those.
• A candidate from a well-known IIT or IIM, or one with fluent English, is scored higher on problem-solving than their actual answers justify. The horns version: a regional accent or a gap in the CV quietly lowers the technical score.
• In plants and branch offices, the supervisor’s favourite (punctual, never argues) gets a strong rating on output quality without the rejection data being looked at.
• One missed deadline during year-end closing pulls down ratings on teamwork and communication for the whole cycle.

The practical defences are to rate one dimension at a time across all employees (instead of one employee across all dimensions), demand a piece of evidence for every rating, and use behaviour-based anchors such as a BARS rather than vague labels. For hiring, an interview scorecard filled in independently before the panel discusses the candidate does much of the work.

Frequently asked questions

Is the halo effect always conscious favouritism?

No, and that is what makes it hard to manage. Most raters genuinely believe each score is separate, which is why the fix is process design (separate dimensions, evidence per rating, a structured interview format) rather than simply asking managers to be fair.

How can HR spot halo in appraisal data?

Look for managers whose ratings for each employee are nearly identical across every competency, so a person is all 4s or all 2s. Real performance is uneven, so a flat profile is a strong hint that one impression drove the form.

Does calibration remove halo bias?

It can catch extreme cases when other managers challenge a rating, but a calibration meeting has its own social pressures, and it usually reviews the final rating rather than each dimension, so halo inside the form often survives.

Our guide to the performance appraisal process lists the other rater errors worth training managers on, and 360-degree feedback adds raters who did not form the same first impression.