“The private sector inevitably nudges, as does the government. No government can avoid some kind of choice architecture. We can object to particular nudges, and particular goals of particular choice architects, but not to nudging in general.” – The Ethics of Influence

Two young fish are swimming along in the ocean. They pass an older fish, who says “morning boys, how’s the water?” The young fish keep swimming along (like many young fish, they are somewhat rude to their elders), until one turns to the other and says: “What’s water?”

This, says Cass Sunstein, is the story of how we are affected by our environment. The effect is so everpresent we often don’t even know it is there, and it is inescapable. At a time when nudges have been enormously popular with governments (see, for example, the UK Behavioural Insights Team, or the US Social and Behavioral Sciences Team), the question of whether they are ethical is very salient.

Perhaps unsurprisingly given Sunstein’s role in popularizing nudges, The Ethics of Influence argues they are. It advances four key arguments: that individual nudges can definitely be unethical, but that nudging qua nudging is not unethical in the same way that taxes qua taxes are not; that objections are pointless, because nudges are inescapable, such as snowy days increasing how many four-wheel drives are purchased and then returned to car dealers; that it’s possible to set the rules of the game, including nudges, without predetermining outcomes (for example, by making a decision easier across the board); and that in surveys, people prefer educational campaigns to direct nudges, but if they are told that education is less effective, support for nudges increases.

Sunstein’s definition of nudges is quite generous, which can sometimes muddy the waters: he includes all educational campaigns as a nudge, for example. That’s not wrong—after all, it is a non-coercive form of affecting behaviour—but it is very different from the examples that most people have ethical concerns about, such as using opt-out to encourage organ donation. At a few points, the book feels closer to salesmanship than analysis: in a chapter on environmental nudges he repeatedly raises the example of setting printers to double-sided defaults, cutting paper use by 40%. It’s a good and useful example, but if you’re already a critic of nudges, that isn’t what you have in mind.

Ethics of Influence is more academic than some of Sunstein’s recent books, and so readers expecting Wiser, Simpler, or Nudge may be a bit taken aback. It is heavily footnoted and abstract, making it somewhat hard to follow at points. It’s also on a very important topic, and one that is making huge waves in government. Probably not a book for the casual reader, but if you’re already interested in the field, Sunstein is a giant in it, and his arguments are relevant and important.

Disclosure: I read this book as an advance reader copy. You can read more reviews and order it here: The Ethics of Influence.