Steve Levitt reviews Nudge

and he likes it!

I could not have been more surprised and delighted when I finally got to read a copy of their new book Nudge. Despite my initial misgivings, I’m halfway through it, and this is a book I love.

The main point of the book (paraphrased) is as follows:

Since people don’t think very hard about the choices they make, it is a lot easier to trick them into doing what you want than to try to educate them or incentivize them to change their behavior. There are many ways to trick people, but one of the easiest is simply by giving thought to the way choices are arrayed to them, or what they call “choice architecture.”

This is an important insight, because as an economist my first instinct is always to jump right to incentives. What they show, pretty convincingly, is that there is often an easier way to get what you want.

What makes the book is the never-ending assault of interesting examples:

Let’s say you want men to stop accidentally peeing on the floor instead of in urinals in an airport bathroom. (Dubner is fascinated with airport bathrooms, so I’m sure he could think of some incentive schemes.) Or maybe someone could invent a new urinal. The choice architects have an easier solution: paint a fly in the urinal. It turns out with something to aim at, “spillage” is reduced 80 percent.

A more substantive example involves 401k plans. It has proven very difficult to get people to make intelligent choices regarding their 401ks: people fail to diversify, hold assets that give them no tax benefit, etc.

Some economists have tried to educate and inform, with little impact. The choice architects take a different approach: almost everyone opts for the default allocation of assets if a default is given. Thus, the answer is simply to make the default choice intelligently given what the choice architect knows about the person. The impact of savings behavior from altering defaults swamps everything else.

Picking and choosing a few examples can’t convey what is most surprising about the book: it is really fun to read.

Academics aren’t supposed to be able to write this well.