Bias, noise and nudges

Over the past 50 years, psychologists and economists have documented many common errors of judgment. When the heuristic judgment processes of System 1 produce inadequate answers, the result is a predictable, systematic average error, or bias. But not all error is bias. When the average error is removed, what remains is variability in judgments, both from one person to another and within the same individual at different times. This unwanted variability, or noise, may in some cases be just as costly as bias, and its reduction should receive the same priority. Can nudges help reduce noise? To the extent that nudges are meant to steer people in a particular direction, they seem more appropriate to debiasing (i.e., correcting a predictable error) than to de-noising (i.e., reducing random error). Some nudges, however, may help address noise, by prompting decision makers to adopt better decision-making practices, or decision hygiene. While this approach to noise reduction is promising in principle, its effectiveness still has to be measured empirically.