Thursday, April 3, 2014

Confusion Regarding Paternalism versus Negative Externalities: The Case of Cigarette Bans

I quit (or took a break from) smoking about half a month ago. During that time, I’ve come to understand some of the dynamics of pubs that allow smoking versus those that do not. I’ve realized that these places provide an informative case study for crystallizing a frequent confusion that arises in law and economics discussions, namely the labeling of paternalistic arguments as externality based arguments.

[For those readers who are not familiar with these terms, here are brief and informal definitions. Paternalism essentially refers to the limiting of people’s freedoms to act under the theory that without such limitations these individuals will act in ways that are contrary to their self-interests. Negative externalities refer to costs incurred by (third) parties who did not engage in the activity that caused those costs.]

In my law and economics class, when I ask my students to question whether there is a utilitarian justification for a given rule or practice, they rarely rely on paternalistic arguments. There’s a simple reason for this: I do not teach them much about paternalism, and I do not provide them with any analytical tools to study paternalism. (I explain the concept and discuss how it may affect analyses etc., but never formally cover it.) However, I do teach them about negative externalities.

I have noticed over the last few years that some students, and unfortunately also some law professors, mistake paternalistic arguments as arguments based on negative externalities. Some may call this an unimportant semantic problem;  I believe that this problem not only makes it hard for two people to communicate with and understand each other, but it also disguises an argument that relies on the existence of some behavioral problems as one that could exist in a world where every individual is acting completely rationally. Due to this ‘disguising effect’ the rule for which the justification is provided appears to be desirable under a broader set of circumstances than is actually the case.

The policy of banning cigarettes or smoking in pubs provides an excellent example to illustrate this point. Consider two potential rules: (i) the owner of the pub gets to decide whether or not people can smoke in the pub, and (ii) smoking is prohibited in pubs. Imagine that I ask my students to provide an efficiency based justification for rule (ii). That is, are there any reasons to believe that rule (ii) would generate higher social welfare than rule (i)?

The following (hypothetical) dialogue is close to what I had with some of my students in the past (S=student, M=Murat)

M: “Is there a good justification to choose rule (ii) over rule (i)?”

S: “By banning smoking in pubs we’re eliminating the negative externality that smokers impose on non-smokers in those pubs where the owner would otherwise allow smoking.”

M: “Okay. So, let’s step back a second. What is the pub-owner presumably trying to maximize?”

S: “Profits.”                          

M: “Under rule (i) will he not choose that policy that maximizes his profits?”

S: “I guess so.”

M: “If the negative externalities imposed by smokers on to non-smokers are sufficiently high, would you not, as a profit maximizing pub-owner, choose to compete with smoker-pubs, by opening up a non-smoking-pub, and attract all individuals who suffer great negative externalities in smoker-pubs?”

S: “Probably.”

M: “So under rule (i), if we observe pubs that allow smoking, does this not imply that (1) non-smokers who go to those pubs do not suffer great negative externalities, and/or (2) the competitive process leads to a good distribution of smoker/non-smoker pubs in town?”

S: “Maybe. But, you are making an implicit assumption that need not be true.”

M: “Most likely. What is that assumption?”

S: “That people can accurately tell how much negative externalities they suffer. And more importantly, when there are smoking-pubs, more people tend to become smokers. We’re incentivizing bad-habits.”

M: “Those are excellent points. But, both of these points rely on people not accurately knowing themselves. The first point you raise is basically that people do not accurately estimate the cost of becoming a second-hand smoker, and the second point is that people do not accurately estimate the cost of being a smoker. Is that right?”

S: “Sounds right.”

As this dialogue demonstrates, the hypothetical student starts off by making a negative externality argument and then quickly switches to a paternalistic argument. The distinction is important. Negative externality arguments are valid to the extent that we believe regulatory responses are legitimate ways of protecting people from others. Furthermore, negative externalities can exist in situations where all individuals are rational. Paternalistic arguments, on the other hand, are valid to the extent that we believe regulatory responses are legitimate ways of protecting people from themselves. Perhaps more importantly, problems that give rise to paternalistic arguments commonly require individuals who act irrationally.

| Permalink

TrackBack URL for this entry:

Listed below are links to weblogs that reference Confusion Regarding Paternalism versus Negative Externalities: The Case of Cigarette Bans:


As much as irrational persons (or people with bounded willpower; I don't want to smoke, but...) some claim the perceived impossibility to reach an agreement on the core rational of certain issues may result in paternalism, and it is good to reduce social conflict.

'“Minimalism” is the name of an approach to constitutional interpreta- tion theorized by Sunstein during the 1990s. Minimalism in constitutional law is the view that the courts should take small steps, decide cases—as the title of Sunstein’s book on the topic puts it—“one case at a time” and, in doing so, avoid theoretically ambitious pronouncements (Sunstein 1999). Minimalism thus rejects, among other things, grand theories such as origi- nalism or what Sunstein refers to as the “liberal perfectionism” of Ronald Dworkin.
In particular, minimalist judges seek “shallow” rulings, those that avoid taking a stand on important foundational questions about, say, when life begins or whether the death penalty is morally defensible. In many circum- stances, the shallowness of a ruling is based on what Sunstein (1995b) has called an “incompletely theorized agreement,” an agreement on a concrete outcome without agreement on the abstract rationale underlying the result.
For Sunstein, one key advantage of being a minimalist is that it reduces social conflict. How so? Shallow rulings help democratic societies negotiate reasonable yet irreconcilable disagreements that are inevitable in a pluralistic society. They do so in a way that “tend[s] to promote social peace,” because shallow rulings “show a high degree of respect to those who disagree on big questions. In a heterogeneous society, it is generally valuable to assure citi- zens, to the extent possible, that their own deepest commitments have not been ruled off-limits. By accomplishing this task, shallow rulings reduce the intensity of social conflicts. . . . The practical point is supplemented by the fact that those who seek shallowness are demonstrating respect for competing foundational commitments” (Sunstein 2008b, 365, citing Rawls 1993).'


Posted by: Meric S | Apr 3, 2014 10:54:08 AM

Murat, your underlying point is interesting and a full response deserves its own post. But just to quibble with the facts you present on non-smoking bars, suppose it were the case that serving smokers are more profitable than catering to non-smokers. If entry to the bar market were costless, we would still expect some non-smoker bars, since at equilibrium firms will enter the smoker market until returns are driven down to the non-smoker levels. But most towns constrain entry with liquor licenses. Isn't it possible that in this non-competitive equilibrium it wouldn't pay to open a non-smoking bar?

Posted by: BDG | Apr 3, 2014 11:20:48 AM

Brian, thanks for the comment. True, licensing can cause frictions that do not exist under perfect competition (in addition to licensing, I believe, here in Tallahassee we have territorial restraints, e.g. you cannot open up a new bar within whatever distance of an older one...). So let me clarify my point. My point is not that one cannot find a non-paternalistic utilitarian argument for cigarette bans. It's that people frequently make paternalistic arguments thinking that they are arguments based only on the existence of externalities. This is problematic, because the nature of the arguments we use affect the conditions under which the rule being defended is socially desirable.

Posted by: Murat C Mungan | Apr 3, 2014 11:44:21 AM

"most towns constrain entry with liquor licenses. Isn't it possible that in this non-competitive equilibrium it wouldn't pay to open a non-smoking bar?"
Yes, but that would be efficient. Suppose just one bar is allowed. The owner would do a calculation of which policy is favored by the majority of his customers. The calculation might turn out either way.

Posted by: Eric Rasmusen | Apr 4, 2014 8:11:07 AM

Regarding second-hand smoke, paternalism actually argues for the illegality of non-smoking pubs. Most people overestimate, not underestimate, the danger of second-hand smoke, due to misleading government propaganda. Smoking is dangerous (though probably people overestimate that too, nowadays), but the evidence for the danger of second-hand smoking is feeble. Thus, we might have people who choose a nonsmoking pub even though they otherwise wouldn't like it as well, because they have a false fear of smoke.

I think, actually, that people fear the health danger of second-hand smoke much less than they say they do. What people really don't like about other people smoking is that it smells bad. That's an equally valid externality from the economist's point of view, but people feel they have to bring in a health risk or their desire to constrain others is morally illegitimate.

Posted by: Eric Rasmusen | Apr 4, 2014 8:15:35 AM

Post a comment