Christopher: I read your interesting paper "Smoking, Drinking, and Income". At the end of the paper you write: "smoking causes significantly larger losses in income than either drinking abstention or heavy drinking, but is is difficult to hypothesize plausible causal mechanisms which could generate effects of this magnitude." Have you seen "Cigarette Smokers as Job Risk Takers" by Kip Viscusi and Joni Hersh in *The Review of Econ. & Stat.*, May 2001? Viscusi and Hersh find that smokers are more hazard prone than nonsmokers; they are much more likely to get injured on the job and at home. This raises the cost of employing a smoker and employers adjust their wages accordingly. They also find that smokers take more higher-risk jobs than nonsmokers, but get paid less for doing them. I don't have the article at hand, so I can't quote the percent difference in wages that they say this accounts for, but I recall that it was substantial.
Yours, Asa Janney Applied Statistical Associates, Inc. Oakton, Va. -- The socialist society would have to forbid capitalist acts between consenting adults. -- Robert Nozick, Anarchy, State, and Utopia, 1974