Cigarette Taxes and Happiness

TPS stalwart Thomas Johnson sent forth this bit on smoking, it gives a nice outline of a number of interesting paths of analysis when trying to explain smoking in a rational choice framework. It tries to get at the larger question of: Are smokers happy with higher cigarette taxes? In categorizing smokers, one possibility is that smokers are completely rational, know the costs and benefits of smoking, and move forward accordingly. You'd never guess it, but it's two Chicago economists with that theory! Another possibility is that people are helpless of smoking, paying any price to get their fix. Yet another possibility is that smokers know the long term costs but prefer the short term benefits. I like the first at face value-- millions of smokers can not all be categorically irrational. The second is absurd, and the third is an intertemporal explanation that deserves consideration-- after all, smokers have been shown to "irrationally" purchase only single packs of cigarettes when cartons would be much more cost feasible. There are definitely issues of multiple time periods involved in smoking.

Sadly, the author of the article turns to happiness surveys to try to get at the overall issue of whether rising cigarette prices lead to more happiness. I skimmed through the paper cited in the article-- it's pdf-linked there-- and they get at the issue by comparing the effect of higher cigarette taxes on high propensity to smoke groups versus the effect on low propensity to smoke groups, and conclude that higher taxes make the smoking groups better off. Well, it's been my experience that those who don't smoke get particular pleasure out of discouraging others to smoke-- higher taxes would certainly fall into that range. And I don't think those people are going to be distributed across the population in any significantly biased manner-- if anything, those that feel like this probably deal with smokers the most, so the distribution may well follow that of the smokers themselves. I'm not sure the study can avoid this possibility. Nonetheless, in happiness research, you're ultimately stuck comparing the happiness gained to happiness lost, and now Arrow is getting really upset.

I'm trying to come up with a good analogy for happiness surveys...I'll post it when it comes to me.