A Follow Up on the Cato Unbound Conversation on New Paternalism

Josh Wright —  25 April 2010

Two weeks ago I highlighted the promising looking Cato Unbound forum on the new paternalism kicked off by Glen Whitman, with follow up posts and responses from the King (or co-King along with Cass Sunstein) of Nudge, Richard Thaler, along with Jonathan Klick and Shane Frederick.  I was really excited about the forum, because I have research interests in this area and consider myself a “skeptic” of the new paternalism generally (hey, the Weekly Standard says “prominent skeptic,” but even I don’t go that far).  So — now that the exchange is over — I find myself, well, better off for having read it but disappointed.  I’m going to blog about the disappointing part.  Don’t get me wrong, it started off really well.  Glen came out swinging, Thaler responds (there no slopes, paternalism is inevitable, and by the way, a really odd choice of example for the lack of evidence in favor of slopes: prohibition), Klick and Frederick chime in.

It was the conversation following the initial postings that left me disappointed.  Whitman started off with a post responding to Thaler that pointed out that Thaler ignores a number of his key points.  The most interesting of these points, or at least the one most in need of a serious and thoughtful response, was about the role of opt-out in the new paternalism.  If new paternalism involves simple “choice architecture” that nudges individuals to make choices that are welfare-improving from their own preferences, one must ask the question about what happens when the individuals don’t respond to the nudge!  This need not require a slope mechanism.  For example, one can introduce a “plain vanilla” requirement that requires those selling credit products to consumers to offer a certain, regulator-approved version and disclose risks of selecting non-approved products.  Of course, these non-vanilla products are very much welfare improving for some individuals.  So the real question is how costly it will be for consumers to opt-out from the nudge.  Or if the costs imposed on lenders to satisfy the “nudge” requirement are in fact to costly as to remove the new products from the market or make them relatively more expensive, thus dampening their beneficial effects.

Here’s Whitman raising the opt-out issue:

Whenever they are challenged, the new paternalists place heavy emphasis on the inevitability argument. Don’t fall for it! It is not inevitable that the state must alter longstanding rules of contract law to reflect political judgments about what people “really” want. It is not inevitable that some opt-outs will be subject to onerous conditions. It is not inevitable that certain kinds of contractual terms will be outlawed entirely. It is not inevitable that the state will impose cooling-off periods on certain purchases, or sin taxes on tempting goods. Again, every one of these proposals appears in Thaler’s own work.

Here’s Klick raising the same issue in his second round response:

In some ways, Thaler’s critics are probably projecting the sins of the paternalist paternalists onto him. Thaler’s nudges are generally premised on the assumption that opt-out costs will be trivial. If we have to have a default rule anyway, why not pick the one that makes people best off, and for those with different preferences, they can simply opt out? Under that premise, Thaler is completely reasonable in suggesting that we should not fear when nudges move from private choices to public policies.

As a practical matter, however, most paternalism is not accompanied with these costless opt out provisions in real world public policy. When it is, and large numbers of people do actually opt out, undercutting the policy goals of the paternalists, a common impulse is to foreclose the possibility of opting out. In my previous comment, I painted national Prohibition as following exactly this pattern. Similar stories can be offered regarding more recent smoking bans and a host of other paternalistic interventions. Beyond simply asserting that their suggestions are default rules, leaving people free to make different decisions, perhaps the libertarian paternalists can spend some time discussing how the opt-outs can be preserved in the face of these tendencies.

I’ve searched Thaler’s two responses for an answer to these questions, which strike right at the heart of the new paternalism enterprise, but have found nothing.  Well, that’s not exactly true.  There is an odd debate over whether Thaler 2003 disagrees with Thaler 2010 (Thaler says no, but Thaler 2010 is a better writer).  There is also lots of aggressive tone to go around (“Whitman’s biggest gripes are with the policies that he only imagines that we favor, ones that involve coercion. What part of the term libertarian doesn’t he get?”).   But I did not read anything responsive to the critical point about the costs of opt-outs.

Of course, in many ways the opt-out issue is one about regulatory design.  And, well, its been said by my friend David Zaring that “when you listen to economists” on such matters, “you are listening to amateurs.”   But even if that were true, I’m not sure that is any excuse for economists to stop talking.  But Jonathan Klick makes a better point.  Economists have a, as Jon puts it, “unfortunate tendency of assuming that you can simply implement a policy with all of the important academic nuances intact without worrying about how the policy will interact with other legal and political forces.”  I agree with him that economists that choose to delve into the area of public policy and actual regulation carry the burden of moving beyond the workshop realm of academic musings and even laboratory and field experiment results, and into practical questions of regulatory design.   Thus, it is no defense to say that others are responsible for the analytics behind the design of opt-outs from new paternalism-based regulation.