Book Review: Nudge

I am admittedly late to the party in reading this popular book by University of Chicago faculty members Richard Thaler and Cass Sunstein. It is a popular interpretation of the work by behavioralists on the effects of framing and choice architecture. The authors endorse thoughtful choice architecture in light of known proclivities of humans to make decisions which might not be in their best long-term interests. The idea is that by providing people with a "nudge," they are more apt to make the "right" decision. The authors are fairly consistent in this message, although a couple of early digressions on daylight savings time, social interactions, and the Rev. Jim Jones confuse the reader into thinking that nudges cause ineffective substitution effects, information cascades and peer effects, and who-knows-what, respectively. After about the fourth chapter the authors settle into a groove that illuminates their central point.

The nudge framework is presented as an alternative to both unbridled (and often overwhelming) choice and to direct command via governmental proscription. Framing effects have been pretty well explored in the academic literature on behavioral and experimental economics. The book was released in 2007 and in paperback in 2008. It was apparently influential in shaping the thinking of some members of Obama's administration, of which Sunstein is one.

I am broadly sympathetic with the attention that could maybe should be paid to choice architecture. If advertisers and marketers have learned the lessons, and there seems to be plenty of evidence that they have, then perhaps others responsible for public and private decision-making (either their own or that of other people) can also benefit from these insights. While the authors do an admirable job of anticipating their critics in a chapter near the end of the book, they don't explicitly address one of my concerns about libertarian paternalism, their ostensibly-oxymoronic moniker for profligate governmental nudging.

Economists are perhaps unique in their uniform objective of efficiency--most other disciplines, including the subset of behavioral economists--acknowledge the richness of other possible motivations for human decisions. So what makes a decision "right?" Which outcome should we nudge towards? To take a concrete example, consider the nudge inherent in proposed changes to laws for labor union formation: instead of an anonymous ballot, advocates endorse changing the choice architecture so that workers need only sign a (non-anonymous) card. This is pretty clearly the type of nudge that Thaler and Sunstein are talking about. Changing the default is a simple way to affect the outcome of infrequent decisions like unionization. But who gets to decide that more unionization is a good thing? On what grounds? The apparent answer is the party in power, implying that nudges can be undone and reversed as political winds change. On these grounds I'm not sure that nudges are likely to "improve decisions about health, wealth, and happiness;" it seems more likely that they are simply a tool to effect different decisions. So far as there is broad agreement about goals (healthier food choices reduce public health expenditures), nudges are useful. When (potentially clandestine) agendas are forwarded, as in the labor unionization case, nudges take on a less Pareto-improving shine.