science entered policy through a modest claim.
Small changes in how a choice is presented can change what people choose.
Automatic in pension schemes is the strongest example.
Participation rose from roughly half to over nine in ten, with no .
Reminder messages for appointments and vaccinations produce smaller but reliable gains.
Enthusiasm then outran the evidence.
Published effects were often large; effects in were consistently smaller.
Part of the gap is publication , and part is the difference between a trial and a system.
A well-designed letter cannot fix a queue that is three months long.
The sharper criticism is political rather than statistical.
Nudges are attractive to governments precisely because they are cheap and avoid confrontation.
A can be adjusted without debating who pays.
Critics argue that this displaces structural measures with ones.
Defenders reply that the two are not substitutes and that nothing prevents doing both.
In practice, budgets are finite, and the cheap intervention is the one that survives a review.