The behavioural turn in policy circles came in part from a simple observation: people, faced with a default option, accept it far more often than they would actively choose it if presented with the same alternatives by other means. From this followed the proposal that governments and firms might 'nudge' decisions in directions held to be welfare-improving, by adjusting the structure of choice — what is opt-in and what is opt-out, what appears first on the menu, how vividly a consequence is named — without restricting any of the choices themselves. The promise was attractive: better outcomes at the cost of mere arrangement, with paternalism kept off the books.
A first complaint, made by libertarians, is that the proposal disguises paternalism rather than retiring it. To select which default is welfare-improving is to make a judgement about welfare that the chooser may not share; to call the arrangement 'libertarian' because the option remains formally open underestimates how much defaults do the work of decisions in busy lives. The complaint has force, but it cuts less than it intends: nearly every consequential institution sets defaults of some kind, and the choice is not between defaults and no defaults but between defaults chosen with attention and defaults chosen carelessly.
A second complaint, made from the other side, is more serious. If nudges work because people are not, in fact, choosing on the carefully informed model that liberal political theory presupposes, then the same psychological levers that allow benign nudging allow malign nudging, and the latter may be the more remunerative use. A choice architecture sufficient to push savers toward higher contributions is sufficient to push borrowers toward worse credit, and is more likely, on present incentives, to be deployed for the second purpose than the first. The asymmetry between who can afford to design choice architectures and who is subject to them complicates the political picture.
None of this requires abandoning the original observation. People do accept defaults; defaults will be set; some defaults will be set better than others. What it requires is honesty about what is going on. The nudge is not a neutral technology, applicable wherever convenience suggests; it is an exercise of small power, and like other small powers, it deserves the kind of public reasoning that attends to who benefits, who decides, and what protections are owed to those whose decisions are being arranged on their behalf.
The author's attitude toward the use of nudges in policy is best described as: