Shane Frederick begins his reply with an extended example about a river in which many people like to swim — and in which a few have drowned. His first question is, “What should the warning sign say?” My first question is, “Who owns this place?”
Frederick doesn’t specify the answer in his set-up. (The river is “open to the public,” but lots of private places satisfy that criterion.) But the answer matters. If this were my private property, I might consider banning swimming there entirely because I don’t want the bloated carcasses washing up on my shore. And that would be perfectly libertarian. I can do what I want with my property.
But assuming I allowed people to swim there at all, what sign would I post? I suppose I’d go with something like option #3 (“Extremely Dangerous Currents! Swim at Your Peril”) or #4 (“Danger. 0.000017% of park visitors have drowned here”). And how did I make that decision? Did I just engage in a small bit of paternalism? Only if you classify “honesty” as a form of paternalism, which seems a bit of a stretch to me. We didn’t need behavioral economics to tell us that people generally benefit from accurate information; the rational actor model told us that much.
How would I choose between options #3 and #4? Well, here’s where behavioral economics does tell us something interesting. It turns out different phrasings of the same basic information can induce different choices. And here’s where the new paternalists think they have a “gotcha,” because we have to choose something. They want to choose based on what makes people “better off.”
But better off according to what? Maximizing their lifespans? Bzzt! That would mean banning swimming entirely, regardless of how much people enjoy it. Maximizing satisfaction of their own preferences? Bzzt! That doesn’t work either, because the inconsistency of choice (the fact that some people choose differently depending on the framing) means we cannot identify their “true” preferences.
So it turns out the new paternalism doesn’t really offer a clearer answer to Frederick’s question than what we already had without it. My alternative standard of “providing accurate information” gave me just as much guidance.
But what if the swimming hole is public property? What should the sign say then? If I were allowed to choose, it would say “For Sale.” I’d privatize the land, and then let private providers of swimming areas compete for customers. If for some reason privatization is off the table, then the state is forced to act as a property manager. In that case, I’d like it to make decisions in much the same way a private owner should — in this case, by trying to provide accurate information. There should also be some additional constraints (like applying rules uniformly across citizens).
In any case, this example strikes me as a distraction, much like the cafeteria example so often deployed by Sunstein & Thaler. Is this really all the new paternalism is about? No, it’s about much more. Frankly, I’m perturbed that the new paternalists are so unwilling to defend in public the proposals they’ve made in journal articles. But perhaps Frederick isn’t really one of them (I confess I’ve read less of his work), and putting warning signs in front of rivers is as far as he would go.
The new paternalists would assuredly go further. Christine Jolls and Cass Sunstein, for instance, say that consumers should be exposed to “risk narratives” that go beyond mere provision of accurate information: “Specifically, the law could require firms — on pain of administrative penalties or tort liability — to provide a truthful account of consequences that resulted from a particular harm-producing use of the product, rather than simply providing a generalized warning or statement …” This is fairly close to Frederick’s option (which I don’t think he favors) of showing swimmers the corpses of drowning victims. And this is where the slippery slope potential kicks in. If the government imposes legal liability on firms for failing to tell customers scary stories, how scary is scary enough? There is no way to draw that line, and hence nothing to prevent the gradual growth of liability judgments under “insufficient narrative” tort claims.
Moving on to Frederick’s other points. He illustrates the idea of small-change tolerance with the old boiling-frog analogy — which he quickly points out is factually wrong. Good thing I didn’t use that analogy! Would it help if I introduced the analogy of men whose haircuts evolve into comb-overs? The larger point, which I think stands, is that people are inclined to ignore small changes. This is actually rational because small changes usually make small differences; nevertheless, the cumulative effect can be significant.
Frederick says there might be a feedback mechanism, so that we (like the frog) will eventually jump out of the boiling water. I hope he’s right. But when it comes to the political process, I fear he’s not. Medicare and Social Security head closer to insolvency every year that we fail to reform them, but that doesn’t seem to make much difference. Political feedback mechanisms are notoriously weak, for reasons public choice economists have made clear. And if people are as myopic as the behavioral economists claim, there is every reason to think policymakers will fail to act on problems until they are upon us.
I appreciate Frederick’s straightforwardness in admitting that behavioral economists do not have a means of “adjudicating among preferences that conflict within or across time.” So I will just reiterate that this is a major piece of the new paternalist framework. Whenever new paternalists propose to set a new default rule, thereby using status quo bias or anchoring or the endowment effect for the express purpose of changing behavior, they are effectively adjudicating among conflicting preferences.
Nevertheless, Frederick concludes by saying, “I see nothing pernicious about recommending options that appear to be better, nor with favoring one’s own preferences when doing so.” I applaud his candor in saying it’s his own preferences that he’s advancing. The new paternalists are implicitly doing the same thing, notwithstanding their lip-service to making people better off “by their own standards.”
What remains unclear, however, is whether Frederick’s recommendations would cross the line from voluntary to coercive and from private to public. I, too, see nothing wrong with telling my friends which dishes at a restaurant I’ve enjoyed the most. And if they requested my help in losing weight, I’d be happy to help out however I could. And if I ever start a business, I might well offer my employees a “Save More Tomorrow” option. Nobody I know, libertarian or otherwise, opposes any of these things.
But if that’s all new paternalism is about, then I’ve been snookered by a series of articles that have advocated much, much more. New paternalists want the state to get involved, and therein lies the danger. First come the genial recommendations; then come the rising hurdles to opting out; then come the taxes and legal penalties for disregarding the state’s friendly advice; and if all else fails, why not ban some choices altogether? That’s a serious slope, and new paternalism is the toboggan.
Christine Jolls & Cass R. Sunstein, Debiasing Through Law, 35 Journal of Legal Studies 199 (2006).