This article tries to show that there are many cases in which there are no determinate preferences for public or private policies to follow or match. This could be because we have no stable preferences at all or because our preferences are influenced by the policies under consideration.
Paternalistic policies seek to influence choices for the benefit of the chooser where “influence” can range from prohibition to advice. Non-paternalistic policies do not do this. Thaler and Sunstein argue that there is a class of paternalistic policies that are on an equal footing with non-paternalistic ones when it comes to manipulating choices and preferences. Since this is so, those who are concerned with individual freedom have no reason to prefer the non-paternalistic policies over this class of paternalistic ones. Assuming that paternalistic policies usually achieve their aim of benefitting the chooser, why not adopt them?
I identified what I called strong and weak claims about the relationship between libertarianism and paternalism. I also argued that the reasons why one is a libertarian make a big difference for the authors’ claim that libertarian paternalism is not an oxymoron. These are laid out in the handout.
I tried to make the point that the reasons for being a libertarian matter to our assessments of this paper. I think the authors should have paid more attention to that since their argument is meant to show that libertarianism and paternalism are compatible.
This paper is addressed to libertarians who consider themselves, in some form, utilitarians. They think that minimal government involvement in the economy and social life is the best way of bringing about the best results overall. At the obvious risk of over-generalization and oversimplification, I think it’s fair to say that this is a prominent view in the University of Chicago law school.
But this is not the only reason why one might be a libertarian. One might think that the state lacks the power to interfere in certain kinds of decisions. It’s not part of the social contract, one might say. Or it hasn’t been given those powers under (an appropriately amended) Constitution or democratic vote. Or individuals have rights that protect them against any state involvement in their decisions, even if it is for their own good.
Of course, some laws anchor choices. If the laws don’t make vacation time mandatory, that will effect bargaining between employers and employees, usually in ways less favorable to employees. But it’s not the state’s job to tip the balance of bargaining over employment or any other contracts.** Compare handout §2.3 Defaults and anchors, example 5 or p. 245 in the article.
Martin suggested another version of libertarianism. This one attempts to minimize political interference with individual choices. It may be true that this sort of policy will alter the choices that people will make when compared with an alternative policy. But if our libertarian cares primarily about individual choice, why should this matter to him or her?
Finally, there is a kind of libertarianism sparked by fear of what the government will do if it is allowed to interfere for our own good. The reason for limiting government isn’t that individual choices are especially valuable or rational. It’s that individuals don’t trust the government, or whatever interest group that influences it, to make the choices for them. Libertarian paternalism assumes that the studies of our choices are accurate, that the government will implement them with the sincere intention of benefitting those affected, that it will succeed in implementing good policies, and that it will not abuse its power in malign ways. Those may be fair assumptions to make for us (or not), but they certainly have not always been true.
Incidentally, one of the things that thinking about this has taught me is that I’m only lukewarm about the value of individual choice and the value of autonomy. If government can improve my welfare using the means that Sunstein and Thaler mention, I’m all for it. Let the technocrats rule! Your opinions may differ, of course.