In 1920, the psychologist Edward Thorndike asked commanding officers in the US military to rate their soldiers on a set of separate qualities: physique, intelligence, leadership, and character. He expected messy, uncorrelated results, since these are genuinely different things and a strong body implies nothing about honesty.
Instead the ratings were almost absurdly consistent. Officers who rated a soldier highly on physique rated him highly on intelligence, leadership, and character too. Thorndike concluded that the raters were not evaluating four traits at all. They were forming one global impression and then broadcasting it across every category.
He named it the halo effect, and a century of research has confirmed it operates nearly everywhere.
What is actually happening
The brain resists holding inconsistent impressions of a person. Once a general evaluation forms — this person is impressive, this person is unreliable — subsequent judgments get pulled toward consistency with it. This is fast, automatic, and almost entirely invisible from the inside.
Crucially, the halo is not deliberate favouritism. If you ask people whether the attractiveness of a job candidate influenced their assessment of that candidate's competence, they will say no, and they will believe it. The effect operates upstream of the reasoning we can inspect.
The negative version is sometimes called the horn effect, and it works identically: one unfavourable impression — an awkward greeting, a typo in the first sentence, an unfamiliar accent — spreads to unrelated judgments.
Where it shows up
Hiring and interviews
Physical attractiveness, height, vocal confidence, and shared background all reliably inflate ratings of unrelated qualities like analytical ability. Unstructured interviews are particularly vulnerable because the interviewer forms an impression in the first minute and then spends the remaining time gathering material that fits it.
Education
Studies in which the same essay is graded under different attributed authors, or with the student's earlier performance known, consistently show grading pulled toward the existing impression. A student who performed well earlier gets more generous readings of ambiguous work.
Courtrooms and authority
Defendants perceived as attractive or well-spoken receive measurably different treatment in mock-jury research. Confidence in a witness's manner drives perceived accuracy far more than it should, given that confidence and accuracy correlate only weakly.
Brands and products
A company known for one excellent product enjoys inflated expectations of its unrelated products. Attractive packaging raises ratings of taste in blind-ish tests. A well-designed website raises trust in the accuracy of the information it contains.
Leadership and business narratives
This is where the halo does its most expensive work. The business writer Phil Rosenzweig documented how the same company gets described as visionary, focused, and bold while its share price rises — and as arrogant, rigid, and reckless afterwards — with no change in the underlying facts. Business books that identify traits common to successful companies are often measuring the halo of known success rather than the causes of it.
Wealth and expertise
Being rich, famous, or credentialed in one field produces automatic deference in unrelated fields. A brilliant physicist's opinion on economics carries weight it has not earned; a successful entrepreneur's views on medicine circulate widely. The halo is a large part of why celebrity endorsement works at all.
Why it is hard to defeat
Three features make the halo effect unusually stubborn.
It is invisible. You experience the conclusion, not the process. It genuinely feels like you assessed the person's competence directly.
It is partly rational. Traits do correlate somewhat in the real world. Conscientious people do tend to be punctual and organised. The problem is not that the inference exists, but that it is applied far more widely and confidently than the actual correlations justify.
Knowing about it does not remove it. Awareness of a bias reliably fails to eliminate it, and can make things worse by creating a false sense of immunity. The effective countermeasures are structural, not attitudinal.
What actually reduces it
Evaluate one dimension at a time, across people
Instead of assessing candidate A fully, then candidate B, assess everyone on criterion one, then everyone on criterion two. This prevents a global impression from forming and contaminating the rest. Daniel Kahneman used exactly this restructuring in the Israeli army's interview process and found it substantially improved predictive validity.
Structure the process before you meet anyone
Decide the criteria, the weights, and the questions in advance. Structured interviews with fixed questions and independent scoring predict job performance considerably better than free-form conversations, and much of the improvement comes from suppressing halo.
Remove irrelevant information where possible
Blind auditions for orchestras — musicians playing behind a screen — significantly increased the proportion of women hired. Anonymised grading and anonymised code review work for the same reason.
Collect independent judgments before discussion
Have each evaluator score privately first. Once someone influential says "I really liked her," the room converges, and a group halo forms that is far harder to dislodge than an individual one.
Ask for the disconfirming case
Deliberately generating evidence against your impression — what is the strongest argument that this person is not the right hire — partially interrupts the automatic consistency pull.
Separate the outcome from the process
When something succeeded, ask what would have to be true for the same decisions to have been mediocre. This is the main defence against halo in evaluating strategies and companies, where success retroactively beautifies every choice made along the way.
The uncomfortable flip side
The halo effect is not only something done to you. It is something you benefit from and can deploy. Clear writing makes arguments seem more true. A well-organised presentation makes underlying analysis seem more rigorous. Turning up on time makes people assume you are competent at things they have never seen you do.
That is worth knowing in both directions: it is a genuine reason to invest in polish, and a genuine reason to be suspicious of it when you are the audience.
The short version
One salient trait sets a global impression, and that impression then colours judgments of unrelated traits — invisibly, automatically, and in a way that awareness alone does not fix. It distorts hiring, grading, verdicts, purchases, and our entire understanding of why some companies succeed.
The fix is structural. Score one criterion at a time, decide the criteria before you meet anyone, gather judgments independently before discussing, and hide information that should not matter. Do not try to be unbiased. Design a process that does not need you to be.
Be the first to share a thought.