Interpersonal comparisons of utility

As we have no immediate experience of what other men feel, we can form no idea of the manner in which they are affected, but by conceiving what we ourselves should feel in the like situation. Though our brother is upon the rack, as long as we ourselves are at our ease, our senses will never inform us of what he suffers. They never did, and never can, carry us beyond our own person, and it is by the imagination only that we can form any conception of what are his sensations. Neither can that faculty help us to this any other way, than by representing to us what would be our own, if we were in his case. It is the impressions of our own senses only, not those of his, which our imaginations copy.

— Adam Smith, The Theory of Moral Sentiments, §1.1.1

It is not possible to measure, for each person, a cardinal quantity that would correspond to the notion of his subjective “level of pleasure”, or of “satisfaction”, and then to compare such quantities between persons to determine which persons are currently feeling better or worse, in general.

But why not? And then, so what?

0. Contents

1. Difficulties with measurement
2. Assuming equality
3. Biochemical individuality
4. Impossibility of measurement
5. Implications

1. Difficulties with measurement

There would be no difficulty if each person’s subjective “level of pleasure”, or of “satisfaction”, corresponded directly to some single measurable material quantity, such as the proportion of a certain molecule in the bloodstream. This does not happen, as far as we know. So, if measurement were possible, it would seem to have to depend upon various factors about each individual.

The variety of such factors is, by itself, a great difficulty with measurement. Even supposing that we could know and measure every external cause that acts upon each person, we know that persons may be very differently affected by such causes. For comparison, the subjective feeling of heat or cold does not vary uniformly with external temperature, but rather, some persons are more or less sensitive to it. It is similar with pleasure and pain.

Bentham, the first utilitarian, was conscious of this. In his Introduction (§6.5), he attempted to enumerate all of the individual “circumstances influencing sensibility” to pleasure and pain:

1. Health. 2. Strength. 3. Hardiness. 4. Bodily imperfection. 5. Quantity and quality of knowledge. 6. Strength of intellectual powers. 7. Firmness of mind. 8. Steadiness of mind. 9. Bent of inclination. 10. Moral sensibility. 11. Moral biases. 12. Religious sensibility. 13. Religious biases. 14. Sympathetic sensibility. 15. Sympathetic biases. 16. Antipathetic sensibility. 17. Antipathetic biases. 18. Insanity. 19. Habitual occupations. 20. Pecuniary circumstances. 21. Connexions in the way of sympathy. 22. Connexions in the way of antipathy. 23. Radical frame of body. 24. Radical frame of mind. 25. Sex. 26. Age. 27. Rank. 28. Education. 29. Climate. 30. Lineage. 31. Government. 32. Religious profession.

Assuming that this division is exhaustive, it would be very difficult to know all of these factors about each individual involved in any situation, and it does not seem that any measurement or comparison could properly take place without knowing any one of them.

2. Assuming equality

Some persons have attempted to sidestep the above difficulty by assuming that, regarding various of their “circumstances influencing sensibility”, all individuals are equal. Lionel Robbins called this the postulate of “equal capacity for satisfaction”:

I do not believe, and I never have believed, that in fact men are necessarily equal or should always be judged as such. But I do believe that, in most cases, political calculations which do not treat them as if they were equal are morally revolting.

Robbins correctly points out that he cannot make this assumption as an economist, since it is not “value-free”. But even for an ethicist, this assumption is implausible and reckless. Individuals are highly variable, and to disregard this is to willfully make judgments that ignore reality. In order to emphasize this, I will spend the next section quoting at length from an author that wrote some books on the subject.

Before this, though, I will note that at least one leading utilitarian has notably refused to resort to such assumptions. R.Y. Chappell, who is a co-author of Utilitarianism.net, wrote a paper arguing against Robert Nozick’s famous “utility monster” thought experiment. Instead of trying to defend some assumption of human equality, Chappell simply agreed that a “monster” really should be prioritized in utilitarian decision-making, and only attempted to show that our “intuitions” otherwise are wrong, and may be corrected by considering a case where the monster is much more sensitive to suffering, in particular, rather than pleasure – a Negative Utility Monster.

3. Biochemical individuality

The richness of individual human variability was most emphatically described by the biochemist Roger J. Williams. Williams described his journey as follows:

One of the students who came to me for advanced study already had an M.D. degree from Northwestern; he told me of the newly published Atlas of Human Anatomy by Dr. Barry Anson of that institution. Otherwise I might have remained in ignorance of it for some time.

This book was specialized, factual and technical but carried a real punch for me. Many normal anatomical variations were depicted and I was stimulated to find many others, so that after a time I became fully convinced that each of us is built in a highly distinctive way in every particular, and that this is the basis of individuality. Dr. Anson’s book led me to realize—with the certainty that objective facts justify—that human bodies can’t be averaged and that an adequate single picture of the human body, or any of its major parts, cannot be drawn. A picture which purports to show the human body is bound to be misleading and may be vicious in its effects. (You Are Extraordinary, p. 9)

Williams described the variability of human anatomy as follows:

The gap between higher animals and man surely is in the direction of more individuality, not less. Much biological evidence can be cited to support this conclusion.

The ways in which human beings exhibit marked individuality are literally so numerous as to be overwhelming. Any attempt to mention them all would lead inevitably to important omissions. In my book The Human Frontier many of these have been brought together for the first time, and the cumulative evidence has been impressive especially to the many who were informed only regarding certain areas.

Individuals differ from each other even in the minutest details of anatomy and body chemistry and physics: finger and toe prints; microscopic texture of hair; hair pattern on the body; ridges and “moons” on the finger and toe nails; thickness of skin, its color, its tendency to blister; distribution of nerve endings on the surface of the body; size and shape of ears, of ear canals, of semicircular canals; length of fingers; character of brain waves (tiny electrical impulses given off by the brain); exact number of muscles in the body; heart action; strength of blood vessels; blood groups; rate of clotting of blood—and so on almost ad infinitum. [...]

We now know a great deal about how inheritance works and how it is not only possible but certain that every human being possesses by inheritance an exceedingly complex mosaic, composed of thousands of items, which is distinctive for him alone. (Free and Unequal, pp. 17, 23)

Most relevant to our subject, Williams also wrote about the variability of human sensitivity to external stimuli, and to pain in particular: 

It is an inescapable fact that in the light of all we know that every individual is highly distinctive with respect to the numbers and distribution of nerve endings of all kinds—in eyes, ears, noses and mouths, as well as in all areas of the skin. This has tremendous meaning because our nerve endings are our only source of information from the outside world. If the nerve endings are different in number and are distributed differently, this means that the information we get from the outside world is somewhat distinctive for each of us. (You Are Extraordinary, p. 40)

The subject of pain is a very complicated one on which we can touch only briefly, but the existence of a high degree of variability with respect to pain sensitiveness can hardly be questioned. [...]

No one method can be expected to give a complete picture of the pain thresholds, much less the more subtle matter of pain sensitivities. Presumably tests for the pain-inducing effect of heat applied to intestinal tissue of different individuals would yield uniformly negative results, but this, of course, does not mean that individuals are uniform in their pain sensitivity.

Whatever mechanisms may be involved in producing variable sensitivity to pain, they must have a genetic origin, and it would appear that all gradations of sensitivities from zero up to the highest level exist in the tissues of different individuals. (Biochemical Individuality, p. 139–140)

In light of the adduced facts, we must judge that, even if human satisfaction is assumed to be measurable, it could only be measured with extremely detailed knowledge about every single person, such as no one on earth has. And then, this knowledge would have to be frequently updated, since various of the “circumstances influencing sensibility” are variable for the same person over time, as well as between persons.

4. Impossibility of measurement

I have refrained from mentioning this up to now because it is more controversial than the rest, but I believe that even with continuously-updated perfect knowledge of all physical factors about all of humanity, it would not, in fact, be possible to measure and compare the “level of satisfaction” of different persons.

This is because physical arrangements of molecules do not have intentionality, the characteristic “according to which they have an object or content and are thus about something.” (Bunnin&Yu, 2009) Nothing about our body is capable of communicating the content of our mental states, which is strictly incorporeal. So, knowing everything about every human body could not communicate each person’s satisfaction, let alone allow comparison.

Besides, if someone managed to also know about every incorporeal human soul, such a person would realize that the human soul is simple rather than composite, and therefore there is no quantity at all within it, whether of pleasure or of pain or of anything else. The concept of cardinal utility is wrongheaded from the start.

So, measurement and comparison of satisfaction is impossible in principle. Not even God can do it. And supposing that it were possible, I have explained in the previous sections that it would nevertheless be impossibly difficult in practice.

5. Implications

The impossibility in principle of measuring the satisfaction of different persons in a common unit, meaningfully susceptible to aggregation and other arithmetical operations, implies that we ought to stop talking of such units. Any general-welfare analyses ought to speak rather of persons’ desires and preferences, which may be speculated about variously, but only truly known by observing their actions.

The impossibility in practice of intersubjective comparisons of utility, whether such comparisons are possible in principle or not, requires the suspension of judgment about such comparisons. The claim that the sacrifice of one person provided a greater benefit to another cannot be founded upon a sound understanding of welfare.

This does not make all general-welfare analyses of personal actions and political systems impossible, since not all possible actions and systems involve any detriment to anyone involved. So, actions through which someone’s welfare is improved, and no one is made any worse off, may truly be said to be justified by the principle of utility; other actions may not. This principle, which Rothbard called the Unanimity Rule, was first stated by Vilfredo Pareto, as follows:

Let us consider any arbitrary position, and let us suppose that we deviate very slightly from it, in a manner consistent with the constraints. If, by doing so, the welfare of every individual in the community is increased, the new position is obviously of greater advantage to each of them; and conversely, it will be of less advantage if the welfare of every individual is decreased. The welfare of some individuals may, moreover, remain constant without these conclusions being changed. But if, on the contrary, this slight movement causes the welfare of some individuals to increase and that of others to decrease, one can no longer affirm that it would be advantageous for the whole community to undertake it. (Manual of Political Economy, p. 363)

Pareto’s statement is ecumenical; those of us holding to the impossibility in principle of utility measurement would like to emphasize, however, that since a person’s actions cannot show him to be indifferent between two outcomes, we may never rightly assert that anyone’s welfare remained constant. All that we know is that the persons who fulfilled their desires have thereby had their welfare increased, and those who were prevented from doing so have thereby had their welfare decreased; nothing may be asserted about any others. With regard to this, Rothbard’s statement of the rule is more precise:

We can only say that “social welfare” (or better, “social utility”) has increased due to a change, if no individual is worse off because of the change (and at least one is better off).

Given what has been said, no judgments about welfare can be rationally justified in disregarding the Unanimity Rule. The consequences of this for ethics and politics have already been drawn by Rothbard, and I will take them into account in my future judgments.