In 2006 the Global Deception Research Team surveyed people in seventy-five countries, asking an open question: how can you tell when people are lying?
The most common answer, in almost every country surveyed, was that liars avert their gaze. Across cultures, across languages, across legal traditions and religious traditions and levels of development, humanity has converged on the same belief.
It is wrong.
Meta-analyses of behavioral cues to deception find no reliable relationship between gaze aversion and lying. If anything, the relationship runs slightly the other way: motivated liars, who know the stereotype as well as everyone else, often maintain deliberate eye contact in order to appear honest, and the effort of doing so is one of the few things that occasionally shows.
A belief this widespread and this false requires an explanation, and there are three components.
It is culturally transmitted. Children are told to look at people when they speak to them, and told that people who will not meet your eye are hiding something. The belief is taught before it can be examined.
It is intuitively satisfying because it has a story attached. Shame makes people look away; liars ought to feel shame; therefore liars look away. The story is plausible and the inference is unsound, since guilt about a lie is only one of many things a liar may feel and often not the dominant one.
And it is almost never corrected by feedback, which is the crucial part. To learn that a cue is invalid you would need to make judgments and then find out whether you were right, repeatedly. Almost nobody ever does. You decide someone was lying and you never learn otherwise; the case where you find out is rare and memorable, and memorable cases feel like evidence. Without feedback, a belief cannot be updated by experience however much experience you accumulate — which explains why police officers with twenty years of interviews are no more accurate than students, and much more confident.
The harm is concentrated on people whose baseline gaze behavior differs from the expected norm, and it is severe.
Autistic people frequently find sustained eye contact uncomfortable or effortful. In many cultures, including much of East Asia, the Middle East and among Indigenous peoples across several continents, sustained direct gaze at an authority figure is disrespectful rather than forthright, so the respectful behavior reads as evasion. Traumatized people avoid eye contact. Depressed people avoid eye contact. Frightened people, and people who have had bad experiences with police, avoid eye contact.
Every one of these groups is systematically misjudged as deceptive by interviewers, jurors, border officers and employers applying a rule that does not work on anyone.
The practical instruction is unusually blunt for this guide, and it is the right one. Discard the cue entirely. Not weight it less — discard it. A cue with no validity is not a weak piece of evidence to be combined cautiously with others; it is noise that will correlate with ethnicity, neurotype and trauma history, and including it makes your judgment worse than excluding it. The same applies to fidgeting, hesitation, nervousness, and self-touch as indicators of lying specifically.
See Defense: discarding folk deception cues