Iatrogenics: The Harm of the Healer

Iatrogenics is the medical term for harm caused by treatment. Taleb extends it as a general concept: the systematic harm done by interventions whose benefit is overestimated and whose downside is hidden. The classical case is medicine — for most of its history, doctors killed more patients than they saved through procedures that felt obviously helpful. The general case includes most management, most policy, most parenting, and most coaching.

The bias toward intervention is structural. Doing something looks active and competent; doing nothing looks lazy and irresponsible. The intervention claims credit for any good outcome and the situation gets blamed for any bad one. Over time, intervention accumulates beyond the point where it produces net benefit, because the negative outcomes are dispersed and invisible while the positive ones are concentrated and credited.

The practical correction is the rule Taleb calls via negativa: more good is often done by removing harmful things than by adding helpful things. Subtract the things that are causing problems before adding the things that might help. The subtraction is often unfashionable, undramatic, and far more effective than the addition would have been.

The closing implication for the reader is to audit the interventions running in your own life — diets, meditations, productivity systems, therapies — and ask which of them are net-positive evidence vs. which are present out of activity-bias. The iatrogenic ones should be removed; the small remainder is what actually works.

Iatrogenics — harm caused by the healer — is Taleb's master concept for the costs of intervention whose benefits are visible and overestimated while whose damages are hidden and ignored. Medicine supplies the literal case: for most of its history doctors killed more patients than they saved through confident procedures that felt obviously helpful, and the harm was invisible because no one counted the patients who would have recovered untreated. He generalizes it to management, policy, parenting, and coaching, all biased toward action because intervening is visible and rewarded while restraint looks like negligence even when it is wiser. The corrective is via negativa: the largest and most reliable gains usually come from subtraction — removing smoking, sugar, debt, noise, and meddling — rather than from adding new cures, because removal has fewer hidden side effects than addition. He proposes a skewed decision rule: intervene aggressively only when the situation is genuinely severe, where the expected benefit is large enough to swamp the hidden downside, and leave mild or self-correcting situations alone, since there the iatrogenic harm of acting tends to exceed the gain. Wisdom, in this account, is largely knowing when not to act — a discipline modern institutions, rewarded for visible activity, are structurally poor at.