Review: Nudge by Richard H. Thaler and Cass R. Sunstein

A book with some interesting ideas in it, but who is it for? Who is the target audience? It covers topics from weight loss to retirement planning to privatizing marriage to government policy on health care, so it feels a bit unfocused. Moreover, the book repeats a lot of the information about cognitive biases from “Thinking Fast and Slow” and “Predictably Irrational,” but not nearly as effectively as those two books.

If you can get past all of that, there are a few valuable takeaways:

• Libertarian paternalism: the author’s preferred political stance is that you should always give people the freedom to choose (libertarian), but that it is acceptable for the choice architect to try to exert influence towards the choices that will help people be happier and healthier (paternalism). In other words, it’s OK for the person laying out the choices to nudge people towards “good” decisions, but ultimately, if someone wants to pick something “bad,” they should always have the freedom to do so.

• Choice architecture: when you give people a list of choices, how you present those choices has a major impact on what people ultimately decide. Example #1: in Germany, where organ donation uses an opt-in system (default = you are not a donor), only 12% of citizens give consent; in Austria, where organ donation uses an opt-out system (default = you are a donor), 99% of citizens give consent. Example #2: if a doctor says, “90 out of 100 patients who have this surgery are still alive and well 5 years later,” you might think that this is a pretty good survival rate; but if the same doctor says, “10 out of 100 patients who have this surgery don’t survive past the 5 year mark,” you might have a very different reaction, even though it’s the same information.

• Choice architect: if you are the person who is presenting the list of choices (the “choice architect”), you can’t avoid choice architecture. You have to present the list of choices in some manner—taking into account what information is visible, what defaults you pick, and many other factors—so the argument in the book is that you should make this decision consciously and following the idea of libertarian paternalism (i.e., you can nudge people towards good choices, but always make all choices available).

• Some interesting examples of nudges:

  • People are heavily influenced by what other people are doing. If you want to encourage someone to do X, tell them about all the other people who are doing X. E.g., tell them about all the other people buying your product, or all the other people already voting for a certain candidate (e.g., John Kerry may have won later states in the 2004 primary simply because he won a few states early on).

  • Near measurement effect. Asking people about how they plan to act makes them more likely to act that way in the future. For example, if you ask people “will you be voting tomorrow?”, and that person says yes, they are more likely to actually vote the next day than if you had never asked.

  • There is a lot of value in merely disclosing the relevant information. For example, credit card companies should be required to send you a summary at the end of the quarter or year that shows you a clear breakdown of your purchases and payments, including, most importantly, where most of your payments went (e.g., how much you paid in fees and interest). This would be incredibly valuable in nudging people towards better spending habits; it would also make it easier to compare different credit card companies, as you could more clearly see what sorts of fees and interest they would’ve had against the same set of purchases. The same sort of information disclosure would be valuable with mortgage payments (how much did you spend during the year; how much off that went to interest and fees; how would that compare to other mortgage options). Disclosure for companies and products is also very valuable. For example, the energy star label, had a powerful effect on the entire appliance and electronics industry, as companies freely started to try to improve enough to earn that label (though they still had the choice not too, as per libertarian paternalism). Similarly, companies that handle toxic chemicals should be required to publicly disclose what chemicals they have and in what quantities; customers would take notice, and this would exert positive pressure to reduce the use of these chemicals (again, with no legal requirement to force them to stop).

  • Cooling off period: certain risky activities should have a required cooling off period. This is especially useful for rare activities that involve intense emotions. For example, it makes sense for divorce to have a cooling off period to ensure that couples who get in a nasty fight have some time to think things over (“sleep on it”) before short-term emotions lead them to a snap decision they’ll regret in the long term. This is a nice example of libertarian paternalism, as the choice is always still there (e.g., to get a divorce), but you present it in such a way to nudge people towards better choices.

• A few other interesting, random ideas from the book:

  • Privatize marriage. The government should not be involved in marriage, as with marriage, it’s impossible to allow for both (a) freedom of religion and (b) equality. Instead, marriage should be treated as any other business arrangement. Two people are free to sign a contract together, similar to a business partnership, and that may confer a small number of legal or tax implications. Marriage itself should be handled solely by private institutions, and they should be allowed to impose whatever rules they please for that.

  • Slippery slope counter-argument: one common criticism of libertarian paternalism is the “slipper slope argument” that says that it’s too easy to go from nudges to hard requirements. The book argues against this as follows: first, does the basic proposal behind libertarian paternalism have merit? If not, then the proposal should be fixed. But if the proposal does have merit, then instead of avoiding the concrete benefits of the proposal due to the fear of some hypothetical slippery slope, let’s instead find ways to put sand on that slope and make it less slippery.

  • The publicity principal: governments and companies should never adopt policies that they wouldn’t want to defend publicly.

  • Self control: the book talks a lot about how much people struggle with self control, such as with trying to eat less, get up on time, or better manage your money. It argues for the value of self control tools that nudge you towards healthier habits. For example, eating from a smaller plate can help in eating less; alarm clocks help you get up on time; maintaining mental accounts and budgets (e.g., christmas savings account with almost no interest). Many of these tools are inherently inefficient (e.g., your money doesn’t care what “account” it’s in and it would be more efficient to have it all in an account that maximizes returns), but we’re willing to accept these inefficiencies to help us have control over our lives. Discussions of cognitive biases, choice architecture, and the lack of self control always make me wonder, do we really have free will or not?

Rating: 3 stars