Monday, May 24, 2010
Smoking and Seat Belts
I used to smoke and I enjoyed doing so very much. I gave up because my doctor told me it might kill me. ‘Might’ was enough to convince me that the pleasure did not outweigh the risks. For all I know it may turn out that his warnings were misplaced. Maybe the correlation between smoking and disease will turn out to be spurious. The result of a common cause. But the uncertainty gets swamped out by the costs of being right. The cost of death to me is high. High enough that even if there is only a small chance of my doctor being right, I want to avoid it. And the cost of avoiding it, foregoing the pleasure, pales in comparison to that: s < p.d That is to say, the forgone pleasure of smoking (s) is less than (<) the probability of smoking causing my death (p) time (.) the cost of my death (d). I used not to wear seat belts. I found them too much of a bother to fuss with and I disliked the way the chafed on my shoulder. I started wearing them because of published data about the superior survival rates of those wearing seat belts in crashes. The data was enough to convince me to put up with the bother of wearing seat belts even though I was not sure of their reliability. For all I know, those who wear seat belts are more cautious that those who don’t. So here too the correlation might be spurious. But here to the cost of death is high and swamps out such considerations. s < p.d where now ‘s’ stands for the foregone convenience of not wearing seat belts. As things go with smoking and seat belts, why do they not also go when it comes to climate change? Here ‘s’ is the cost of significantly reducing greenhouse gas output. ‘p’ is the probability of climate change caused by such output and ‘d’ is the cost of such climate change. If there is a non-zero chance of catastrophic climate change, of ecological collapse, then even if that chance is low, the costs will be so high that p.d will still be high, high enough to outweigh the costs of avoiding it however low the value of p.