David R. Henderson  

Sunstein Goes Straight to Coercion

Hayek as read by Chinese commu... Did Galbraith and Steinbeck Ev...

Time out from posting on Galbraith to note a current discussion. My next post on Galbraith will appear this afternoon.

Co-blogger Bryan Caplan has posted recently and cogently about libertarian paternalism and outright coercion. As it happens, I have a review of "libertarian paternalist" Cass Sunstein's latest book, Simpler. It will be out in the Fall 2013 issue of Regulation.

Here's an excerpt from my forthcoming review that goes to some of what Bryan blogged about.

Moreover, Sunstein, who has been a strong proponent of "libertarian paternalism" by government, advocates measures that go beyond libertarian paternalism to straight coercion. Many libertarians have feared that some of the "nudges" libertarian paternalists advocate would turn out to be simple coercion. I was an early, though cautious, defender of Sunstein's and co-author Richard Thaler's advocacy of nudges in their book, Nudge. (See my review of Nudge, "A Less Oppressive Paternalism," Regulation, Vol. 31, No. 2, Summer 2008.) But based on Sunstein's book, I must conclude that many of the libertarian critics' fears have turned out to be justified.

In at least three instances, Sunstein crosses the line from advocating nudges to advocating outright old-fashioned coercion: price controls, restrictions on the size of soda containers, and graphic warnings about smoking.

Take price controls. (Please.) One of the worst regulations Sunstein favors is price controls on health insurance. Sunstein, who is obviously economically literate, doesn't seem to feel the need to justify price controls, despite the fact that opposition to price controls and the distortions they cause is one of the things that the vast majority of economists agree on.

Or consider New York City mayor Michael Bloomberg's move in 2012 to limit the size of soda containers to sixteen ounces. Sunstein points out an obvious fact: that limit would not have allowed people to choose a larger size. Yet, in discussing comedian Jon Stewart's negative reaction to Bloomberg's ban of larger containers, Sunstein writes, "Stewart is capturing a pervasive and general skepticism about paternalism in general and nudges in particular." Here, Sunstein himself is incoherent. As he had admitted a few lines earlier, Bloomberg's regulation was a ban, not a nudge. Sunstein might argue that it is a nudge because one can always buy multiple containers. But that is costly. Moreover, what if someone wants 20 ounces of a soda? It's hard to buy a 4-ounce drink. I hate to say it, given his earlier huge investment in advocating what really are nudges, but Sunstein seems to be illustrating precisely what many libertarian critics had feared: one of the primary advocates of nudges and libertarian paternalism seems quite comfortable with old-fashioned coercive paternalism.

Or, finally, consider the graphic warnings that the FDA wants to require on cigarette packs. Such warnings include disgusting pictures of people with bad health due to smoking. Sunstein claims that such warnings "are a distinctive kind of nudge." "However graphic," he writes, "the warnings maintain freedom of choice." It is true that, with the graphic warnings in place, people would still be able to choose to buy cigarettes. But there's more than one choice involved. Another choice is the kind of package people buy their cigarettes in. And the FDA regulation that Sunstein supports would substantially limit people's choice. Call it a hunch, but I think most smokers would rather not buy their cigarettes in such packages. My guess is that the reason Sunstein is oblivious to that lack of choice is that he's not in the market for cigarettes. I wonder how he would feel if, when he ordered a fattening dessert in a fancy restaurant, the server were required to serve it with pictures of people who are in poor health because of overeating such desserts.

Moreover, whether the issue is cigarettes, cars, drugs, or any other good, Sunstein consistently puts a zero weight on the freedom of producers. In discussing the various examples he cites, Sunstein devotes not a sentence of concern for their freedom--or lack of. His indifference to producers' freedom becomes explicit in his discussion of regulations on advertising airline fees. In 2011, notes Sunstein, the Department of Transportation introduced a regulation to require "airlines to disclose prominently all potential fees on their web sites." "Even better," he writes, "airlines have to include all government taxes and fees in every advertised price." Sunstein notes that some airlines sued to invalidate the regulation, "complaining especially about the requirement to include taxes and fees and invoking the First Amendment, no less, to say that the requirement was unconstitutional." Here's what he doesn't tell you: the airlines that sued, in Spirit Airlines, Inc. v. Department of Transportation, wanted to be able to state the government taxes on the ticket in a font as big as the font on the overall price. But the regulation prohibits them from doing so. So passengers won't be as aware of the government's role in high airfares as of the fares themselves. That sounds like a First Amendment case to me. You would think that Sunstein would understand that. After all, his book is a commercial product and he left out this important piece of information. In that sense, he's like an airline that leaves out information about high baggage fees. Yet I bet he would object to a law requiring him to tell the reader the whole story. To his credit, he earlier rejected his own tentative proposal for a "fairness doctrine" for the Internet; he had toyed with the idea of legally requiring bloggers to link to contrary views. That Sunstein even seriously thought of that idea, though, suggests that on the issue of free speech, he has a tin ear.

COMMENTS (15 to date)
Max writes:

Ugh, would be difficult to convey the level of my disgust with mere words. I actually liked Nudge and was hoping it represented a move in the right direction for Sunstein. I guess critics who claimed that it was just a sneaky way of pushing us onto a slippery slope were right.

Dan McLaughlin writes:

The more important issue, even with nudges, is the assumption that the "nudger" knows what is best for everyone else and has a right or responsibility to nudge them in that direction. It is that much more offensive with coercive means.

The fundamental point is that a central planner cannot know what is good for millions of people, cannot know their wants, their needs, their goals, their aspirations.

Sunstein was offensive from the beginning because libertarian paternalism is not libertarianism at all. It is paternalism.

Carl writes:
legally requiring bloggers to link to contrary views

Wow, what a creepy idea. At least Mr Sunstein saw sense on that one!

Ted Levy writes:

One of the classic examples of "nudging" is the recommendation to put cafeteria line desserts in the back and move the salads forward, allowing people to choose what they want but making it easier to choose "what is good for you" and harder to choose "what is bad for you."

I see no difference between this and the cigarettes with upsetting pictures on them.

These cases are purely analogous. In either both cases or neither case is the consumer coerced. It seems straight forward to me that in both cases the consumer is NOT coerced, and thus Prof. Henderson is wrong to include this in his triad of examples.

Granted, in both cases, if the decision is to move the desserts to the back or to put disturbing pictures on the cigarette packs is mandated by the government, the PRODUCER is being coerced, if not the consumer. And this is where the alleged libertarian paternalist always fails. He doesn't come up with SUGGESTIONS for improvements that market participants may consider and reject. He comes up with mandates to make things "better" as he understands matters.

Silas Barta writes:
Ugh, would be difficult to convey the level of my disgust with mere words

And even if you did find the words, Sunstein would ban their use! ;-)

Sunstein's views are irrelevant when it comes to judging libertarian paternalism on its merits. Seems to me like this is a slippery slope argument, the true intention of which is to cast all nudges is a negative light.

Gene writes:

In conversations with people who advocate this or that policy to incentivize ordinary people to change their behavior in order to achieve some "social good," I find it's often useful to ask them to define the limit beyond which they would never go to enforce that change in behavior. And it's helpful to ask them what they would do if the population whose behavior they want to change utterly refuses to cooperate. Would they be willing to admit defeat and walk away, or would their response be to find some higher form of coercion to achieve their goal? In other words, are you willing to give up on your policy idea when it becomes clear that people will not permit the change?

For those who have read Sunstein's books I'd be interested to know if he addresses these questions, and his response.

Brian writes:


I've got to take issue with your third example. Putting warnings on cigarette packages is a clear example of a nudge, not consumer coercion. Your notion that consumers should have a right to choose the kind of package they want is, well, just plain wrong. If you want a given product, you have to accept whatever packaging the producers choose. You may want your cigarettes packaged in aluminum cans, but it's not happening. You deal with it if you want the cigarettes. If you don't like it, you can always repackage your own cigarettes after you buy them.

I do agree that there's some producer coercion involved, but that's always true of government-sponsored nudges. This is where the libertarian has to decide whether any degree of paternalism is possible; some will say no for this very reason. But then one has to wonder whether ANY warning signs could ever be justified. And what of the consumer's right to be fully informed?

Finally, regarding consumer coercion, here's an alternative. The government could allow warning-less packages for an extra surcharge of, say, $1 per pack. That is, producers could choose to remove the warnings by simply charging $1 more and giving it to the government, or choose to use the warnings and avoid the extra dollar. This would seem reasonable if the warnings are shown to reduce smoking and the likelihood of lung cancer--such warnings would therefore have some clear economic value. I'm betting that both producers and consumers would choose to keep the warnings, suggesting that the nudge is not as coercive as it might appear.

Silas Barta writes:


If you want a given product, you have to accept whatever packaging the producers choose

But the producers didn't choose it in this case.

geoih writes:

Maybe you just need to broader definition of nudge. Think of it like the barrel of a gun pressed against your back. It's a kind of incentive.

R Richard Schweitzer writes:

What is the underlying concept of "paternalism" and the Nudge?

Does it not require the conclusion that some have the (special?) capacities to identify and select the correct directions for the conduct and relationships of others?

Once that conclusion is reached, the march down the "Road" begins.

guthrie writes:

@Ted Levy,

I agree that rearranging the placement of food offerings really doesn't coerce the consumer, pictures on cigarette cartons might.

Here's a test for the various points of view... would a smoker would pay more for a picture-free carton? I'm not a smoker, but I think I would at least think about it. What do you think?

MikeDC writes:

I can review Simpler in one sentence.

A book about government, called "Simpler", that doesn't spend at least 95% of it's words on the tax code, education, medical industry regulation, and law enforcement is about as deeply un-serious as can be.

If the goal is to simplify, the solution must come where people actually interact with the government.

Scott Scheule writes:


Given you think pictures on cigarettes don't count as nudges, what do you do think counts as a nudge?

I suspect nothing for you would count since every nudge is a bit of coercion in some sense, and changes the freedom of the purchaser to buy a particular type of product--i.e. everything nudged product is an impingement on freedom if the wish of the consumer is to buy an unnudged product.

If so, you're making two different arguments. 1. The idea of non-coercive nudging is incoherent. 2. Sunstein is inconsistent in his preference for nudging.

Cari Beth writes:

Would Sunstein and company approve of pictures of aborted fetuses on all abortion clinic signs, wall decor, and advertisements?

My guess is he would not because the subject matter is something he approves of.

Nudging (and government as a whole) is about whose ox is being gored. Freedom for me but not for thee...I think I know what is better for you and everyone else.

Comments for this entry have been closed
Return to top