The smoking debate redux
For the first substantive post on this blog I’ve gone with a topic
that’s an oldie (for NZers), but a goodie: banning smoking in public
places. Britain’s currently going through the growing pains of a
conversion to smoke-free public places and economist Tim Harford
thinks that the economics of the change just don’t stack up. He pulls
out the well-worn argument that if people wanted non-smoking pubs then
the market would have provided them. Well, game theory tells us that
the answer to this riddle is really quite straightforward.
Smokers want to be able to relax in a pub at night with a beer in one
hand and a cigarette in the other, surrounded by friends. Non-smokers
want exactly the same but minus the cigarette. This is a clear example
of a co-ordination game between smokers and non-smokers. Both are
better off when they go to the same pub whether that be a smoking or a
non-smoking pub. No non-smoker wants to alienate his smoking mates by
heading off to a separate pub and the same is likely true for
smokers. The fact that we’re currently co-ordinated on an equilibrium
where people go to smoking pubs rather than non-smoking pubs is an
artifact of the historical norm in favour of smoking. Thus, the
lack of non-smoking pubs shows that nobody individually has an
incentive to deviate. What it doesn’t show is that this outcome is
best or even that it is preferred by everybody.
This is where the government comes in: they must decide whether it
is welfare improving to pass laws that re-organise society around a
non-smoking norm. Certainly the non-smokers will be better off, but
what about the smokers? They may well be worse off, but not by
as much as they would like you to think.
Plenty of research on time inconsistency has shown that smokers
themselves can be made better off if the government forces them to
smoke less (see Gruber and Koszegi’s work for instance). Thus, some
reduction in their smoking is welfare improving for them as well as
for the non-smokers. Does the decreased reduction in smokers’ utility
mean that their losses are outweighed by the non-smokers’ gains? That
is a matter for empirical research, but it is likely that the British
government’s actions will be far less traumatic for British smokers
than Mr Harford anticipates.
Interesting. I notice you mention the concept of time inconsistence for smokers, maybe you could explain that in more detail ;).
Are you sure that this is socially optimal though. During my first year I gained a lot of satisfaction from having a drink and a smoke at the same time, as the two are complements, and I’m not even a smoker. Furthermore, no when you go to bars it smells like farts and BO, I preferred it when it smelt like smoke.
Surely, if individuals hated being in bars with smokers so much, non-smoking bars would have appeared. After all if non-smokers value not having smoke around, they would be willing to pay a premium, and the bar would sell a differentiated good. If the non-smoker had a smoking friend they would negotiate. If not smoking was so important, the smoker would sacrifice smoking indoors (and maybe choose a worse alternative for smoking outside) for some compensation (e.g. a drink). As this transaction didn’t take place (which would have been a version of the Coase theorem) I doubt the externality was severe enough to require a solution.