by Pierre Lemieux

…what does it mean to say that the loss of life is a cost for an individual?

drug scales.jpg

The report published last month by the Council of Economic Advisers (CEA) on the cost of the opioid crisis raises more questions than it answers. Mainly because it incorporates the estimated value of the lives lost through overdose (using the “value of a statistical life,” in the standard cost-benefit jargon), the reports reaches a humongous figure of 2.8% of GDP, as much as 44 times previous estimates.

The major problem is that the CEA estimates the social cost of opioid consumption without even mentioning its benefits. Cost-benefit analysis, whose methodology the report claims to follow, has no meaning if only the social cost is calculated. The social benefit, which is the sum of individual benefits, must also enter into the calculation. Opioids must have some perceived benefits, and even perceived net benefits, for otherwise nobody would take the risk of consuming the stuff. It is reasonable to think, as Gary Becker‘s theory of rational addiction proposed, that an addict does not enjoy his addiction per se, but that it helps him face pre-existing personal problems that would otherwise be even more difficult to support; which is to say, it provides utility.

The standard objection is that the consumer (of opiods, in this case) does not “internalize” all costs. Unaware of what he is doing, he ignorantly imposes part of the cost on himself as if such costs were negative externalities (cost imposed by others). Exaggerating a bit (perhaps), we must assume that an addict knows what he is doing to himself at least as well as a typical politician understands what he is doing to others.

At any rate, one cannot just ignore benefits. A voluntary loss of life, including through freely assumed ex ante risk, cannot be counted as a cost if the corresponding benefits are not included in the calculation. It is like calculating the cost, but not the benefit, of the “automobile epidemic,” or the cost of life (everybody dies at some point) without including its benefits. Cost are incurred to obtain benefits and lose all meaning if the latter are excluded.

Can we go farther? Looking at the problem ex post, what does it mean to say that the loss of life is a cost for an individual? Since the individual is no more (at least in this life, outside of which costs presumably don’t matter), it would seem that his lost utility is a sunk cost that should not enter into any current calculation.

Of course, the future possibility of losing utility through death does matter for the victim. This is precisely why living individuals typically support criminal laws that deter murder – but not laws that deter them from doing what they want with their own lives. A murder ban deters murderers from making trade-offs between life and death for other people. There is a big difference between a risk freely incurred by the individual himself and a risk imposed on him by others.

Lost productivity (which means lost production) raises similar issues. For example, a smoker who knows that smoking increases his risk of illness and thus of lower productivity and income, and still chooses to smoke, obviously judges that the benefit of smoking is higher. Many old studies on the “social cost of smoking” neglected this.

If we take seriously the basic principles of cost-benefit analysis, an individual’s death is a cost to himself, not to “society.” It can only be conceived as a cost to society in the derived sense that social costs are the sum of individual costs, but then individual benefits must also be incorporated in social benefits. If a slave-master owns a young, productive slave, the latter’s death is a cost to the former. But if the individual owns himself, his death or lost production is a cost only to himself. The normative assumption of self-ownership is a natural one to make for the economist.

But doesn’t somebody’s death deprive others of his previous “contribution to the economy”? The answer is generally no. If an individual is paid the value of his marginal product (what he contributes to production) as economic analysis shows, he does not “contribute to the economy.” He contributes to himself. He reaps all the value he adds to production.

We must admit that some (rare) individuals contribute to the economy by generating a non-insignificant increase in social benefits for others through higher consumer surplus (defined as what consumers get over and above what that they pay). The activities of these create positive externalities for others. Think about Bill Gates, or Steve Jobs, or perhaps Malcolm McLean, inventor of the shipping container. But note that these people’s untimely deaths would have deprived others of utility only for a time, because the same discoveries would have been made independently later on. Interestingly, only artistic creators seem to generate lasting, unreproducible discoveries. Monkeys typing randomly could reproduce the works of Shakespeare, Virgil, and Homer, but this may require an eternity in time or an infinite number of monkeys.

What about personal relations that are not traded on markets, whose value is difficult to estimate? The death or illness of somebody is a cause of sorrow, and thus lost utility, for those who liked him. What Hippocrates said about the physician applies to the loved ones or more precisely the loving ones: “the misfortunes of others bring a harvest of sorrows.”

True enough, but this may be cancelled by the fact that individuals also generate negative externalities. Some may rejoice at somebody else’s death. Albert Einstein provided fabulous new knowledge, but people in Hiroshima and Nagasaki got negative externalities. Some individuals bring love to the world, others bring hate, and many bring both. There is no way to know whether one individual’s net contribution is positive or negative until he is dead. And even then, the evaluation of an individual’s net contribution would require weighing costs against benefits, and such weighting is the major conceptual problem of cost-benefit analysis.

Who knew that, when he was a young child, Adolf would become Adolf Hitler, or that Bill would be Bill Gates? What was the net benefit or the net cost of Paul Samuelson? Was the dead opioid addict going to be the next mass killer or the next Beethoven? Who can predict the long and twisted chain of historical events?

All this reminds us that cost-benefit analysis is a speculative discipline. Its main, non-negligible advantage may be summarized by the strong sense of the word “discipline”: its methodology forces us to follow clear definitions of costs and benefits, and to include both. (Note a further point: government transfers are not by themselves real social costs or benefits; as transfers from some individuals to others, they net out in the process.)

To wrap up, the CEA is wrong to include the value of lives lost or even of productivity lost when calculating the cost of opioid consumption. These probabilistic costs belong to each individual, who nets them out from his expected benefits. Piling up costs without netting them against benefits is meaningless. The CEA – and the government in general – has no scientific basis for overruling individual choices. An addict’s friend may try to persuade him or her that better solutions exist, but coercive public policy is very different.