Richard Thaler Responds to Critics of Libertarian Paternalism

Economist Richard Thaler, a leading advocate of libertarian paternalism, has briefly responded to some of the points I made in my most recent post on that subject. In that post, I argue that advocates of libertarian paternalism implicitly rely on the assumption that regulators and voters are rational, while consumers and other private sector actors are not:

Ilya Somin gets the discussion off to a unhelpful start by claiming (a) that we consider regulators to be perfectly rational, when again we repeatedly say in the book that regulators are human just like everyone else and (b) that “regulators have no reliable way of estimating the benefits and costs that consumers derive from potentially risky products.” If he means “perfectly reliable” then fine, but that is not a sensible standard. Suppose that a crib is found to strangle babies that sleep in it, as happened to the child of friends of mine. I think it is pretty easy to guess that parents would not want to buy a crib that has strangled a dozen kids. Don’t you?

Thaler’s response, I think, misinterprets my argument. Of course I am well aware that Thaler and other advocates pf libertarian paternalism realize that regulators are “human just like everyone else.” Indeed, in my post I linked some important articles on regulatory irrationality by Thaler’s coauthor Cass Sunstein. The problem is that this recognition of regulators’ “humanity” gets lost in libertarian paternalists’ policy recommendations, where the implicit assumption of regulator rationality plays a crucial role. If the libertarian paternalists built in to their theory the fact that regulators cannot be expected to be more rational than consumers (and, for reasons, I indicated in my post are often likely to be less rational), then it is unlikely that they would continue to advocate government regulation as a good solution for consumers’ cognitive biases.

On his second point, I did not contend that a “perfectly reliable” estimate is required. Rather, I linked an earlier post where I argued that regulators don’t have anything close to an even partially reliable way of estimating the benefits that consumers derive from risky activities such as smoking, drinking, sky diving, and so on. They may know how great a health risk these activities pose, and thus have some way of at least roughly estimating their costs. But they can’t measure the utility that consumers get from doing these things. That is especially true when consumers’ tastes and preferences vary greatly. And it’s hard to have an accurate cost-benefit analysis when at best you can only estimate one side of the ledger.

Thaler’s example of the dangerous cribs somewhat obscures these points because it relates to children, a group we typically do not allow to choose for themselves on a wide range of issues and do not trust them to estimate their own utility. Paternalism with respect to adults is a different matter. Even in the case of the cribs, however, Thaler’s analysis obscures as much as it reveals. Assuming that a dozen children did die, one would still want to know how many that was relative to the total number of users, how good those cribs were relative to their competitors in other ways, and so on. A dozen dead children is, of course, horrible. But many more children than that are injured or die each year from activities such as taking baths and bicycle accidents. Yet that doesn’t mean that the government should forbid parents to allow their children to take baths and ride bikes. Some low-probability risks are sometimes worth tolerating even in the case of children.

The rest of Thaler’s essay is devoted to responding to others who can fend for themselves. However, I want to briefly note the significance of Thaler’s dismissals of the possibilities that paternalistic regulation might lead to “slippery slope” effects and that regulators might use their authority to impose their own personal preferences or those of the majority of voters. Thaler says that that’s not what he and his coauthors want regulators to do; he wants them to “choos[e] the choice architecture that is your best guess of what the participants would choose for themselves if they had the time and expertise to make an informed choice.” But once we recognize that regulators and voters are prone to cognitive biases “just like everyone else,” there is every reason to expect these and other negative impacts of expanded paternalistic regulatory authority to occur. Indeed, the assumption that our own preferences are right for everyone is one of the most common cognitive biases of all. The “best guess” of a cognitively biased regulator as to what others will choose if they were were well-informed is likely to be very similar to the regulator’s estimate of what he himself would do under such conditions. Thaler’s laudable desire that regulators avoid these pitfalls does not prove that they actually can or will do so.

UPDATE: I should note that in my initial post, I wasn’t sufficiently clear about the distinction between what libertarian paternalists understand in their own minds and what is built into their theory. In my view, they clearly do understand that regulators aren’t fully rational in the former sense, but have failed to incorporate it into their analysis in the latter. I am sorry for any misunderstandings this might have caused.