alternatives, not mutually exclusive. We could reduce speed limits and enforce them; everyone would be adversely affected (time is a cost), but there would be some reduction in air pollution. We could modify cars to reduce damage from collisions or crashes, making them heavier; everyone would be adversely affected, and there would be a slight increase in air pollution. We have required seat belts and their use; there was a small cost to nearly everyone. We can require air bags; the cost is much greater than seat belts, however, and their additional contribution to lives and injuries reduced is modest. The cost per life saved is estimated at $1 million. 53 We could raise the driving age, concentrating the cost on a few with the highest risk of accidents. We could cancel driver licenses for years of all those causing crashes, driving under the influence, or both, thus concentrating costs on those whose behavior indicates they are high-risk drivers. We have to consider driver responses in calculating benefits: seat belts and air bags may make drivers less cautious; so may heavy cars well protected against collisions. As for raising the minimum age for driving, is it just age or the fact that inexperienced drivers, whether sixteen or eighteen, are accident prone? Thus estimating benefits is not a cut-and-dried proposition. Costs, in contrast, are highly diverse; how do we value time lost by lower speed limits? The inconvenience of postponing or denying the right to drive? Or the distribution of costs, among accident prone and accidentproof alike, versus their concentration on the former? In making policy decisions, I doubt that all these alternatives were carefully compared, and priorities selected on a benefit-cost basis. Each policy has its avid proponents, its lobby. Victory goes to the best-organized lobby.
What does saving a life actually mean? We all die eventually. To make comparisons, costs should be expressed in terms of an additional year of life, a measure that also reflects the age at which a life is saved. That is one reason why vaccinations are so cost effective; for each individual life preserved, the average number of years saved could be seventy. The benefit estimates place far too much stress on lives saved, or prolonged, because they are much more easily measured than illness and injury avoided. In most instances, most of the benefit is in the latter. Cancer is the main exception, because of its high mortality rate.