10th March 2011
A debate with Harold Demsetz on second-hand smoke
I sent an email to Harold Demsetz about my blog post yesterday, and he kindly replied, with a brief analysis of the second-hand smoke issue I had raised. Let me take this analysis further here, noting that this is preliminary analysis, subject to revision as I find time to think about it.
(very slightly edited – for typographical issues – from the version posted in my earlier blog post)
"Take the example of second hand smoke that causes lung cancer. The venue in this case is private, and the guest pays the appropriate market price for entry, but the potential harm caused to the bartender (lung cancer) is simply too large and disproportionate.
"Here, one can identify all buyers and sellers over a long period, and everyone involved knows the precise likelihood of harm, but no one knows which individual causes a particular secondary harm. Allocating property rights perfectly doesn't solve this problem (or at least the markets have never figured out a solution by themselves).
"In my view such a situation needs to be brought under some umbrella of justice, and that is where a government can enter, by imposing a ban on such risky behaviour even where property rights are fully allocated. I've elaborated on this in The Discovery of Freedom."
Harold Demsetz's response:
Sanjeev: I appreciate much the thought and the advertising you have given to my latest paper. I will need to think over your bartender case, but my first reaction is that if the courts or legislature have given people the right to smoke it is up to the bartender to pay smokers to desist. Now, you might say that there will be a long line of people before him puffing cigarettes, waiting to be bought out. True, but, then, has this not become a strategic, free rider problem, one that can be solved by ownership of the slot of space at the bar by the bartender or by having the bartender pay his boss to make the bar no-smoking? — Harold
I'm happy to assume, along with Demsetz, that people have a legal right to smoke. But I note at least three difficult issues:
1. The challenge of defining property rights in air
As I pointed out earlier, "the markets have never figured out a (property-rights based) solution" to this kind of a matter. Who owns which air particle in the bar? In my view it is not sufficient to create property rights in air behind the bar's counter, for smoke-filled air spreads everywhere. The bartender also moves across other's "air space".
2. The smokers will not pay
Assume for a moment that a method to assign property rights in air has been discovered and implemented. One thing is clear. Smokers, who have a legislated right to smoke won't pay the bartender. But even property rights in air to someone (say, to the bartender) won't make them pay the bartender. That is because smokers can escape liability entirely. They know that the harm they individually cause is too small and too difficult to attribute to them. No court of law can penalise them. This, along with their legal right to smoke, means one thing at least: that they will never pay the bartender for the right to smoke.
3. The bartender will not buy out smokers
In this case the bartender can (theoretically) pay. But as Demsetz notes, a queue will quickly form. More problematically, all types of smokers and non-smokers will join the queue, pretending to be heavy smokers. That is only the beginning of the bartender's problems. The marginal price the bartender will have to pay will be determined by the true addicts who (presumably!) won't give up a smoke for any amount of money. And if I, a non-smoker, know that behind me in the queue is standing a true addict who will surely demand $10 million to not smoke, then I too will demand $10 million. The bartender will be a fool to even try to pay smokers. I can't readily prove it mathematically without more thinking, but my guess is that the bartender's optimal strategy, even assuming he can devise a way to prevent cheating, is to NOT buy out the smokers.
In summary: a) property rights can't be readily defined in this case and b) trade in this 'property' (air quality) will fail in the absence of identifiable responsible person/s who can be held to account for pollution. As a result, the bartender's likelihood of death from lung cancer will remain unchanged.
The economic solution does not work.
The key issue of harm; and justice
I raised other issue of justice in my blog post. I believe this is a crucial consideration. The principle of negative liberty states that you are free to do what you wish, so long as your action (and in some cases, inaction) doesn't harm me. The consideration of harm is important.
In DOF I have outlined three categories of harm. Economic harm (loss) figures very low in the list. Some economic harm is purely imaginary, as well (such as harm from price change arising from 'monopoly' or 'cartels'). But this particular example is not about economic harm. I (bartender) can't wash my lungs (like I can wash laundry dirtied by factory smoke). I am dealing with deadly smoke. The potential cost is infinite: death. That is too big a cost to bear – for no fault of mine.
I also know that there is a specific causal action in this case (smoking in confined premises), but I know that no one can be penalised. On such a matter of life and death, where economic solutions will fail, coercive regulation comes to mind.
How does this fit into Demsetz's paper? I don't know for sure. Happy to receive input from you.
ADDENDUM: Now please read Prof. Demsetz's response here.