In a previous news blog, I argued that the state of being dead should not come into a health utility analysis.[1] Even if the disutility is discounted, I argued that this is an erroneous practice. Here, I build on this argument: first, with arguments against the practice, some of which summarise my previous argument; then a counter-argument, based on the idea of an intrinsic value of an un-lived life. Finally, I will counter the counter-argument.
Arguments against projecting a utility value of zero throughout an individual’s remaining life in an economic model:
One, the axiomatic argument. As argued previously, it is automatically wrong to ascribe a value of disutility to somebody who is dead because there is no disutility to be experienced. There is a disutility contingent on facing death and this should be accounted for. But not so the utility of the dead state itself. Describing such a utility to a health economic decision is a fallacy that arises from using methods developed for an individual person concerning their own treatment, not for use in aggregate decision-making. The axiomatic error that health economists make here is to conflate a person’s trade-off while they are alive with the disutility of the dead state, relative to other states the person might be in where they are not dead. The issue is compounded by summing the disutility of the dead state over each remaining year of life.
Two, the lived experience argument. While a person cannot experience a loss of utility while dead, they certainly can while alive. Therefore, the disutility of the state in which a person is moving to death should certainly include the dread of their impending death. In fact, in so far as death can be averted at a younger age, the dread of not being rescued when such rescue is possible, should also be factored into the calculus. But it is lazy to use the utility of the state dead as a surrogate for these factors.
Three, the spill-over argument. Currently, the dead person suffers a disutility of one (utility of zero), while members of the family do not lose any utility at all. Consider the health economics of infertility treatment. Here the beneficiaries must include all those who gain utility –the index woman may not even be the person who is treated, nevermind the only beneficiary. The same applies when someone is dying and persists after they have died. It should be accounted for, insofar as it might not be encountered (as in the case of a child), or be attenuated (as in the case of a death deferred to an extreme age).
The counter argument is that there is indeed a value in human life, even if the disutility is not experienced. This whole argument turns on the fact that if the life could be created or preserved, then it would still experience a utility. That is to say, it is the utility forgone that must be included in the model. Indeed, this is Parfit’s ‘repugnant conclusion’ argument mentioned previously in your news blog.
Countering the Counter Argument
It is more coherent to think about decision outcomes in terms of disutility than utility. Failure to do that might leave one in a blind alley, and the above conclusion is indeed repugnant. The repugnant conclusion rubric is at variance with contraception or, indeed, any failure to take advantage of a reproductive option. Once one thinks in terms of disutility, then the argument becomes clear. The optimal decision is simply the one with the least disutility for people who exist or may exist in the future. The only argument left is the population disutility that might be avoided by productivity that may be realised when death of a working age person is avoided –but that is a topic for another day!
To conclude, a new model is needed that focuses only on the disutility of people who are living, including those who are dying, and completely excludes a utility that is not experienced.
— Richard Lilford, ARC WM Director
Reference:
- Lilford RJ. The Value of Lives That Do Not Exist. NIHR ARC West Midlands News Blog. 2023; 5(1): 9.