Sometimes the government passes laws that restrict people for what it claims to be their own good, such as laws that ban drugs that are supposed to be bad for your health. Laws like this are called “paternalistic.”
Libertarians oppose paternalism, but it is not only libertarians who reject it. It is at odds with the whole tradition of classical liberalism. John Stuart Mill famously opposed paternalism in On Liberty. He defended the Harm Principle: “[T]he only purpose for which power may be rightfully exercised over any member of a civilized community, against his will, is to prevent harm to others. His own good, either physical or mental, is not a sufficient warrant.”
Paternalism has in recent years made a comeback, as we see in such absurdities as restrictions on the size of cans of soda. I’d like to look at one argument against Mill’s Harm Principle advanced by the influential lawyer and government administrator Cass Sunstein, in his book Nudge and elsewhere. (In fairness to Sunstein, he says he is a libertarian paternalist, not a paternalist tout court. “Libertarian paternalist” seems contradictory to me, but I will put this aside.)
The argument I want to consider is Sunstein’s response to what he calls the Epistemic Argument: “Because individuals know their tastes and situations better than officials do, they are in the best position to identify their own ends and the best means of obtaining them.” He thinks the Epistemic Argument is the strongest argument in favor of the Harm Principle.
To challenge the Epistemic Argument, Sunstein points to cognitive mistakes that people make. Sunstein is a leading figure in behavioral economics, and he writes about these mistakes with great authority. Following the psychologist (and Nobel Prize-winner) Daniel Kahneman, he distinguishes between two “cognitive systems” in the mind. “System 1 works fast. It is often on automatic pilot. Driven by habit, it can be emotional and intuitive.” By contrast, System 2 is “deliberative and reflective.” When we operate, as we often do, with System 1, we are subject to various sets of mistakes, which count as “behavioral market failures.” With the details of these mistakes, we are not here concerned, but the errors include “present bias and time inconsistencies,” “ignoring shrouded (but important) attributes,” “unrealistic optimism,” and “problems with probability.” What for our purposes is important is the conclusion Sunstein draws: “With respect to paternalism, the unified theme is that insofar as people are making the relevant errors, their choices will fail to promote their own ends. It follows that a successful effort to correct these errors would generally substitute an official judgment for that of choosers only with respect to means, not ends.”
Suppose, for the moment, that we accept Sunstein’s claim that these cognitive mistakes impede people from getting what they want. Does this give one reason to reject the Epistemic Argument? I do not think so. According to the Epistemic Argument, each person is in a better position than government officials to choose the appropriate means to satisfy his ends. This is entirely consistent with people’s making cognitive mistakes. The point of the Epistemic Argument is that people can better judge their situation than officials can, not that their judgment is without error.
Ludwig von Mises fully realized this point, and Sunstein would have benefited from a reading of Mises’s comment in his essay “Laissez-Faire or Dictatorship” on J.E. Cairnes’s objection to laissez-faire: “Let us for the sake of argument accept the way in which Cairnes presents the problem and in which he argues. Human beings are fallible and therefore sometimes fail to learn what their true interests would require them to do. … It is very unfortunate that reality is such. But, we must ask, is there any means available to prevent mankind from being hurt by people’s bad judgment and malice? Is it not a non sequitur to assume that one could avoid the disastrous consequences of these human weaknesses by substituting the government’s discretion for that of the individual citizens?”
There is a further problem with Sunstein’s use of cognitive mistakes to justify paternalistic interventions. He offers no evidence that people who act in ways he wants to modify have fallen victim to cognitive mistakes. Do people who smoke, or consume sodas in large quantities, or fail to buy fuel-efficient cars, suffer from cognitive mistakes? Maybe they do, but the fact that people are susceptible to these mistakes does not show, for any particular example, that they have made these mistakes.
The challenge to the Epistemic Argument thus fails.
