By David Gordon
Some economists, such as the 2017 Nobel Laureate Richard Thaler and his colleague Cass Sunstein, have proposed an unusual justification for government interference with people’s choices. They do not intend, they say, to override the preferences that people have. They don’t want to tell people what they “should” want, according to an external standard that people don’t accept.
They claim, however, that accepting the actual preferences people have still leaves room for government intervention. How is this possible? Their answer is that people often choose in an irrational way. They make mistakes in reasoning and choose impulsively. People don’t “really” want what they choose irrationally, so government intervention that pushes people to choose rationally is consistent with respect for people’s preferences.
One way to challenge this view is to deny that people who choose irrationally aren’t “really” choosing. What you would choose if you had full information and weren’t making mistakes in reasoning may be an interesting question, but the answer to it does not tell us what people want. If it does not, there is no room for government intervention that respects people’s preferences, contrary to Thaler and Sunstein’s assertion.
Rizzo and Whitman ask, “What is the evidence that people choose irrationally?” They find this evidence unconvincing.
Some people might say, “Isn’t it obvious that people sometimes make irrational choices? For example, people often sign up for costly gym memberships and then wind up attending the gym fewer times than they thought they would. They would have saved money if they had paid by the visit. Isn’t the government helping people get what they want if it mandates a period of time for them to cancel long-term gym memberships?”
Rizzo and Whitman aren’t convinced.
The analysis is fundamentally static, and thus the crucial observations that individuals plan to go to the gym more than they actually do and that they delay canceling inappropriate contracts are interpreted as partial naiveté. Indeed the individuals may be naïve to begin with, but does that explain where things end? To answer yes would seem implausible. Consider that the people in this study were new gym members and therefore likely inexperienced … Why should we expect inexperienced individuals to know how much self-discipline they will have in going to the gym? The only way they will find out is by getting feedback on their initial optimistic expectations. And this will not happen all at once. Inevitably there will be a period during which they will be paying for visits they did not use. The more patient they are about learning, the longer this period will be. Patience in acquiring the knowledge necessary for self-regulation can be confused, ironically, with present bias.
Rizzo and Whitman’s strategy here is subtle. They are not in this example accepting the view of biased behavior held by the behavioral economists, but rather they are asking whether, given this view, it has been shown that people are choosing irrationally. Further, they aren’t here claiming that people for the most part do choose rationally by this standard, though it’s clear from other things in the book that they think a good case can be made for this. Their limited claim here is that it has not been shown that people act in a biased way.
They use the same strategy in analyzing other studies that claim to demonstrate biased choice. In doing so, they confront many examples of alleged biased choice which have become notorious in the popular literature.
Their dissent from one common example of irrational choice illustrates the depth of their analysis. According to one standard account, employees who can choose to participate in a retirement savings program will choose differently, depending on what the “default” option is. That is to say, if people have to choose to join the program, fewer will sign up than will remain in the program when they have to “opt out” in order to leave it. Surely, the argument goes, an important decision like participating in a retirement savings program should not depend on so trivial a matter as the default option. Isn’t this strong evidence for biased choice?
It comes as no surprise that Rizzo and Wittman are unconvinced.
Employees face a complicated decision about whether and when to enroll as well as what savings contribution to make … The cognitive cost of considering options and reaching a decision is immediate, while the benefits are in the future. Present-biased agents seek to put off the immediate cognitive burden; ‘let me think about this tomorrow.’ … So we must ask; is the fact that many agents eventually opt in explained by their learning about their own bias and then reducing it, or by their learning more about the situation context (including their preferences and the investments options available)? … When they cannot be distinguished, learning looks like procrastination.
Once more, the authors do not claim to have proved that behavioral economists err in asserting that people choose irrationally. Their claim is that irrationality has not been proved to exist.
In other words, we in fact know much less about the prevalence of irrational choice than some behavioral economists think we do. What follows from this, so far as government intervention is concerned? Cass Sunstein answers, “Not all that much.” In the face of objections to claims of irrational choice, he maintains that
it is not enough to offer an array of theoretical, conceptual, and empirical arguments against behavioral paternalism. Rather, we must offer a broad and comprehensive argument that is sufficient to decisively rule out any form of paternalism whatsoever.
Only someone eager to impose his allegedly superior wisdom on the “irrational” masses could take this demand seriously, and the rest of us will join with Rizzo and Whitman in rejecting it.
David Gordon is Senior Fellow at the Mises Institute, and editor of The Mises Review.