Thomas Schelling

Choice and Consequence

Among the poignant issues that policy has to face, explicitly or by default, are some that seem to pit finite cost against infinite value. What is it worth to save a life? How much to spend on fair trial to protect the innocent against false verdicts? What limits to put on the measures, some costly in money and some in anguish, to extend the lives of people who will die soon anyway or whose lives, in someone's judgment, are not worth preserving?

These issues are ubiquitous. They arise in designing a national health program. They are directly involved in decisions for traffic lights, airport safety, medical research, fire and Coast Guard protection, and the safety of government employees. They are implicitly involved in regulation for occupational safety or safe water supplies, in building codes and speed laws, even helmets for motorcyclists—because somebody has to pay the costs.

It is characteristic of policymakers, especially at the federal level, that they usually think of themselves as making decisions that affect others, not themselves. Hurricane and tornado warnings are for those living where hurricanes and tornados strike; mine safety is a responsibility of legislators and officials above-ground concerning the lives of people who work underground. Policies toward the senile, the comatose, the paralyzed, and the terminally ill are deliberated by people who are none of the above. Occasionally the legislator debating a 55-mile speed limit pauses to think whether the benefits in safety to his own family will be worth the added driving time, but if he or she is conscientious even that calculation may be surreptitious.

The situation is different when a small community considers a mobile cardiac unit or a new fire engine. The question then is not what we ought to spend to save someone else's life but what we can afford to make our lives safer. Spending or stinting on the lives of others invites moral contemplation; budgeting my expenditures for my own benefit, alone or with neighbors for the school safety program, is less a moral judgment than a consumer choice, a weighing of some reduction in risk against the other things that money will buy.

There is a suggestion here. Maybe we can reduce the unmanageable moral content of that paternalistic decision at the national level by making it more genuinely vicarious. Instead of asking what society's obligation to them is, we should ask how they would want us to spend their money. In deciding how much to require people to spend on their own seat belts, smoke alarms, fire extinguishers, and lightning rods, it is easier to be vicarious and it is legitimate to get our bearings by reflecting on how much we might reasonably spend on our own safety. The question still may not be easy, but it is less morally intimidating.

Surely, if we were all similarly at risk and in like economic circumstances, this would be the way to look at it, whether for the town bandstand or the town ambulance. On a national scale it is less transparently so, but nevertheless so, that we should want our appropriations committees to think of themselves as spending our money in our behalf. We want them neither to skimp where it really counts nor to go overboard to prepare at great expense—our expense—for the remotest of dangers. We want them to be thinking not about what concern the government owes its citizens for their safety but how much of our own money we taxpayers want spent for our safety.

With that perspective it is remarkable how quickly the issue, now collectively self-regarding instead of other-regarding, drops the ethical content that was only a construct of the initial formulation. We can still find ethical issues, but not the one that seemed so central.

We could call this the contractual approach to social obligation. In the absence of an understanding, I may owe you, in your extremity, unbounded attention and concern, comfort and livelihood, room and board, and the best medical attention in perpetuity, and feel guilty when I stop to wonder whether you are worth the burden you are putting on me. When it is my turn of course I'll expect the same from you (or from whoever has the corresponding responsibility toward me that I had toward you), feeling a little guilty perhaps but not enough to relinquish my claim. But if we could sit down together at an early age in good health and legislate our relation to each other, specifying the entitlements we wished to obtain between us, recognizing equal likelihood of being beneficiaries or benefactors, we could elect to eschew exorbitant claims. And it would not strike us as an ethical issue.

The contractual approach can help with some of those other tantalizing dilemmas, like which planeload of passengers to save, the big plane with lots of passengers or the one with mostly empty seats, if both are at risk and one at most can be saved. What I should do in the control tower if that Godlike decision were mine is an ethical dilemma that for some thoughtful people has no easy answer. But if I am an airline passenger answering a questionnaire for the FAA on what rule I want the control tower to follow in emergencies, the issue is neither ethical nor a dilemma.

I cheated a little in supposing that we were all similarly situated with respect to some risk and alike in our ability to afford protective measures. We usually are not. But the value of this conceptual approach, of considering what the safety is worth to the people who are safer, may still be salvaged. If you are more at risk than I—let's say you are at risk and I am not at all—and the rule is that we share the cost of reducing the risk and we are purely self-regarding, I shall find the measures worth nothing, while you find the measures twice as attractive as if you had to pay it all yourself. If you fly a plane and I do not, the new runway lights at the local airport will cost us each $1,000, and may be worth it to you but not to me. If they are worth more than $2,000 to you, they ought to be bought, at least if you'll pay for them. (Whether I ought to pay half is a separable issue.) But suppose you wouldn't pay more than $1,500 for the slight contribution the runway lights make to your safety. I propose that they shouldn't be bought.

Even though we divide on the issue of whether, paying $1,000 apiece, we ought to buy the lights, the economist in me formulates the problem this way: If, because your personal safety is involved, you are entitled to my contributing $1,000 toward the purchase of runway lights, and I acknowledge your claim and put up my money for your exclusive benefit, do you really want to buy the runway lights or would you rather take my $1,000 and keep it? I don't care. I may be annoyed at having to put up the money when you don't want the lights after all, but I might be more annoyed that you are spending my $1,000 on something that you didn't consider worth the cost. In fact, we could both be better off if instead of purchasing the lights I just gave you $750 and we forgot the lights.