Restaurant Smoking Bans Redux

In a previous post ("Restaurant Smoking Bans"), M.D. Fatwa argued that smoking bans probably make sense, because there's very clearly a negative externality and a market failure. MDF suggests that the political popularity of smoking bans is evidence of a market failure. This is, to say the least, a deeply problematic statement. A proposal that would take $20 away from one person, burn half of them, and then split the rest among ten other people might be extremely politically popular -- over 90%! -- but the market's failure to provide this "service" is an argument in favor of markets and against government. But let's set this aside: it's not his main point.

Nor should we really concern ourselves with why, if MDF is right, hypersensitive nonsmokers don't constitute sufficiently large demand to create a market for nonsmoking restaurants. After all, MDF is a country boy living in the midwest, which as any good eastern elite like myself knows, is a vast wilderness where there are dragons and only a very small number of tiny villages with very few people. 10% of the population wouldn't be large enough to create a market for non-smoking restaurants -- that's, like, 1 person. In Chicago, maybe 2 or 3 -- totally not enough to support a restaurant. So we have to cut him some slack here.

The real problem with MDF's model of market failure is that ... he's not proposing a model of market failure. Market failure is when there exist costs or benefits that are not borne by the decision makers AND this alignment leads to over- or underproduction. It's kind of odd that MDF imagines the first requirement is being met -- apparently positionally sensitive nonsmokers are imagined to go to restaurants just as often and be willing to pay just as high a price when there is some smoking as when there's not. (If they went less, restaurants would get less profits, in which case they're internalizing the costs of their decision, right?) But okay, fine: let's suppose they don't go less and so there are external costs. But these external costs aren't translating into decreased restaurant attendance or food consumption (again, this is a necessary consequence of the stipulation that the costs of smoking borne by nonsmokers don't affect the restaurants' bottom lines). So where's the market failure? The market is producing the correct amount of output -- all we're talking about is the division of the value of the output.

The only way out is to say that positionally sensitive nonsmokers make their restaurant-attending decisions premised on the assumption that they will definitely not be affected by smoking, but then they get to the restaurant, sit next to the smoke, and think, "Crap! I forgot that there's such a thing as a smoking section! I really need to write that down -- or maybe we should just pass a law, because I can't be expected to remember stuff like that." Which is to say, they're irrational (which I suppose one could believe: I'm just not sure why, given that one believes some nonsmokers are irrational, one would also believe that those same nonsmokers should be allowed to make decisions about what kind of restaurant they're going to eat at ... or, much more importantly, what kind of restaurant I am going to eat at).

But leaving aside the possibility of irrationality, the only possibilities are that nonsmokers never decide to stop going to restaurants because of smoking, or that they do, but that the decrease in restaurant attendance by nonsmokers is more than made up by the increase in smoker attendance because smoking is allowed. Either way, there's no market failure.

Okay, so if there's no market failure, why are smoking bans so politically popular? Well, as MDF pointed out, 80% of the population doesn't smoke. So if all smokers vote against smoking bans and all nonsmokers vote for bans, how does that work out? Again, this doesn't mean there's a market failure -- elections don't count intensity of preference. (In other words, a vote of 2 yes's and 1 "no, oh no, please God, no!" still goes to the "yes" people.) So what we might have here is a plain old-fashioned story of rent-seeking: smoking bans might have higher costs than benefits, but why would a non-smoker care so long as the costs were borne by other people?

This might run into problems if voting for smoking bans will make even nonsmokers worse off (which it might very well). But who said people vote in a narrowly self-interested fashion? Some people who haven't been to a bar in years vote on whether we should ban smoking in bars -- how on earth could either outcome matter to them enough to justify going down to the voting booth? Well, it might matter if they also care if you are smoking in bars. So how's that one likely to break -- are there more people who (a) think that smoking is bad and therefore you should be kept from smoking (or kept from working around smoke), or (b) who think that it's none of their business whether you smoke or work in a bar? If you think (b), you need to get out more.

Finally, this could just be a matter of expressive voting. Look: unless your name is "Sandra Day O'Connor" or "Anthony Kennedy," your vote doesn't count. You will not ever break a tie. But you're not stupid and you know this -- so why would you vote as if the election turns on your decision? The answer is you probably don't. There are people in Texas who voted for Nader in 2000. Did they think Texas might not go to Bush? Of course not -- a lot of them would have voted for Gore if they thought their vote mattered. Given that your vote doesn't matter, you might just vote the way that makes you feel good. So, what, you want to be the kind of person who votes for blowing smoke in children's face? What about proposition "Let's not kick puppies?"

So, fine, let's ban smoking. But please let's not pretend that there's a market failure or that there's some sort of greater public interest story here. Nonsmokers have a majority and they can do what they want, or think they want, or like to tell other people that they want. End of story.