Le biais d'omission — pourquoi ne rien faire semble plus sûr qu'agir

The hidden cost of doing nothing — and six ways to make inaction a visible choice

What is omission bias in decision-making?

Omission bias, documented by Spranca, Minsk, and Baron (1991), is the tendency to judge harmful inactions as less morally blameworthy than equally harmful actions — and to choose inaction even when acting would produce better outcomes. It is driven by the moral asymmetry between doing and allowing: a harm you caused feels authored, and therefore attributable to you, while an identical harm you merely permitted feels like something that happened. Because the felt blame differs while the outcome does not, the bias systematically underweights the real costs of not acting — and it is strongest exactly where the stakes are highest, in medical decisions, safety choices, and any situation where the passive default is quietly the worse option. Recognizing it does not remove the asymmetry; it converts a default you never noticed into a choice you can actually weigh.

Spranca, Minsk, and Baron (1991) documented omission bias through scenarios showing that people consistently judged harmful inactions as less blameworthy than equivalent harmful actions, even when outcomes were identical. The bias is particularly consequential in medical decisions (refusing a vaccine vs. accepting one that causes the same risk), policy debates (passive harm vs. active harm), and personal choices where the safe-feeling default is actually worse than the uncomfortable alternative. Jonathan Baron and colleagues developed this line of work across a research programme in judgment and decision-making, and the finding has held up as one of the more robust asymmetries in the field. The reason it is so hard to correct is that the asymmetry is not a reasoning error you can simply notice and drop — it tracks a real feature of moral responsibility, since causing and allowing genuinely are different in most ethical accounts. The failure is one of calibration rather than of principle: the discount we apply to inaction is far larger than the moral difference warrants, and it operates automatically even when we would reject it on reflection. That is why the practices here do not argue you out of the intuition. They target the specific mechanism instead: making the costs of inaction as vivid, concrete, and explicitly stated as the costs of action, so that the comparison happens on equal footing.