Nudges Fade After Bias Correction

Recent overviews claim that once publication bias is addressed, generic nudges show little to no average effect, and very large, real‑world trials report much smaller impacts than the published record. If 'one‑size‑fits‑all' nudges underperform, the case for personalized, context‑specific interventions (with explicit moderators) grows. — This challenges the evidence base behind government 'nudge units' and argues for preregistration, transparency, and a pivot toward targeted designs before scaling behavioral policy.