But [the case for libertarian paternalism] is much too sweeping, because it overlooks everything the rest of the book says about how the experts are as prone to cognitive biases as the rest of us. Those at the top will be overly confident in their ability to predict the system-wide effects of paternalistic policy-making...While it's right for regulators to be humble about their degree of knowledge about the world, and to be cautious in creating new regulations, there are several reasons why this particular criticism is wrong.
First, we are not comparing like with like. There is no claim that a regulator, when placed in the same situation and making the same decision as a citizen, will come up with a better answer. Instead, we are looking at times when citizens make snap decisions without thinking them through - or, often, make no overt decision at all because they do not notice that there is a decision to be made. In these cases, the regulator's goal is either to say "what would the citizen decide if they did think about it carefully?", or even better, to encourage the citizen to make the effort of thinking it through themselves. If the answer to "what would the citizen decide?" is controversial or ambiguous, the regulator is unlikely to try to intervene.
Second, everyone specialises in something. A lawyer specialises in the law - I wouldn't expect them to be better at making business decisions than me, but where my business decisions have legal ramifications I'd like to have their input. A doctor does not know better than me what I should choose to eat for dinner, but they can give me useful information to help me pick the foods that are right for me. And similarly, somebody who spends their professional life thinking about decision-making and examining the extensive research in this field is likely to be able to help me make decisions that I'll be happier with.
Third, even if regulators are not perfect, a best-effort regulation may well be better than none at all. The absence of regulation does not mean the absence of nudging. As Thaler and Sunstein point out in Nudge, our decisions are going to be influenced by context, framing and defaults no matter what. If the government takes no part, then the influences will be random, or chosen by private companies (whose interests are sometimes opposed to mine, though not always). If a democratically accountable government can help to move from one default frame to another that is more likely to be in my interest, then why would I not prefer that one?
There are people who read public choice theory as implying that most government policies are likely to be corrupt and wrong-headed. These people may well oppose Nudge-style regulations as they oppose most other regulations. But I think that view is not a mainstream one.
Indeed, mainstream economics has a clear place for regulations - where the action of a private party imposes an unpriced cost or "externality" on others, an appropriate regulation will actually make the market function more like a genuine free market, not less. One could argue that the framing carried out by private companies acts as an externality by imposing hidden costs on the consumer - in which case a better market outcome will be achieved by reversing these frames.
Finally, Nudge policies are designed to be optional. Anyone who does not trust government to provide a default which is in her own interest is welcome to ignore the government's recommendation and make her own choice. Those who would prefer to make the tradeoff of trusting the government's (carefully researched, democratically supervised) recommendations can simply take the default and save themselves the effort of thinking it through, and reduce their risk of making a mistake.
Easterly certainly raises an interesting challenge for behavioural research: the "system-wide effects of paternalistic policy-making" are indeed not well-understood. The fact that the policies are optional for the citizen is (I suspect) likely to mean that the system exhibits stability rather than instability, with any effects of Nudge regulations being damped rather than amplified by the interactions of the system. But it's not certain yet. My own research focuses partly on this area, as I think it's important to find out how a more accurate picture of human behaviour, scaled up to the systemic level, will affect our understanding of how markets work.
I notice that I have written on this subject before, so I'll finish by quoting myself:
We don't expect aeronautical engineers to be immune from the law of gravity. Yet we still trust them to design planes that can help us transcend our own gravitational problems.