Interpersonal Comparisons of Utility

So as I said earlier, I've been getting involved in a whole bunch of debates involving value theory, and I wanted to sketch out a few of my views in order to have a starting point for discussion, so I don't have to keep explaining myself over and over again. In this post, I want to address the idea that because there is no way to objectively measure utility, it is impossible to coherently make claims which involve comparisons of utility between people.

But before addressing this point, I anticipate that it will be necessary to offer something by way of disclaimer. It occurs to me that by suggesting that something can be more valuable to one person than to another person, I will be accused of being a Utilitarian--of claiming that the person to whom the thing would be most valuable should be the one to have it. An example of a view like this can be found in Peter Singer's famous essay, "Famine, Affluence, and Morality," where he wrote:

When we buy new clothes not to keep warm but to look "well-dressed" we are not providing for any important need. We would not be sacrificing anything if we were to continue to wear our old clothes, and give the money to famine relief. By doing so, we would be preventing another person from starving...we ought to give money away, rather than spend it on clothes which we do not need to keep us warm. To do so is not charitable, or generous. Nor is it the kind of act which it would be good to do, but not wrong not to do. On the contrary, we ought to give the money away, and it is wrong not to do so.

I do not wish to comment on exactly to what extent I disagree with Singer, except to say that I think that for very many people, maintaining a good appearance is a very serious interest, and should not be brushed off with such little care. I will also say, in passing, that I find slightly disturbing Singer's view that morality requires us to give to others "...unless in doing so we would be sacrificing something of comparable moral significance..." I simply can not assent to the idea that we have no right to indulge in pleasures of our own (even if we've earned those pleasures!), as long as there are others who are worse off than we are. The source of this objection is obviously Kantian: my life is an end in itself, and I need not justify my pursuit of personal fulfillment by reference to anything else but the notion that I am an individual with a right to live my life according to my own desires. I do, however, find plausible the idea that we have some obligation to help those in need, and that neglect of this duty is immoral (though perhaps, as Nozick pointed out, we would not be justified in enforcing morality in this sense). But all of this is beyond my point.

The purpose of this discussion is not to explain the ethical consequences of interpersonal comparisons of utility. It is simply meant to give some reason for thinking that we are not necessarily incoherent to make claims which rely on comparisons of utility between people. Accordingly, my disclaimer is this: My argument is positive, and not normative. In saying that interpersonal utility comparisons are, in at least some cases, possible, I am not saying anything about what ought to be done in cases where these comparisons show that one person would benefit more than another from some policy, or where one person would gain more than another would lose.

So with that protracted disclaimer out of the way, I'll get on to the issue at hand. Perhaps the most famous reason for thinking that utility can't be compared between individuals is discussed by R.F. Harrod in his essay, "Scope and Method of Economics." He writes:

Whether the nth unit of X has greater or less utility than the mth unit of Y to a given individual may be made the subject of test. He can be given the choice. But there are no scientific means of deciding whether the nth of X has greater or less utility to individual P than the mth of Y has to another individual Q. The choice can never be put.

This point is entirely true. But notice what Harrod says next: "This implies that we cannot in fact decide whether two pense have more utility to a millionaire or a beggar. Yet we may have a shrewd suspicion. But this, we are told, is "unscientific," for lack of a test." He continues, "Can we afford to reject this very clear finding of common sense? Of course great caution must be exercised in not pushing the matter too far. Since the evidence is vague, we must not go farther than a very clear mandate from common sense allows."

This, I think, is the crux of the issue. When I walk home from class, sometimes I feel like grabbing a cup of coffee from Espresso Royale. I can't remember exactly what it costs me for a medium cup of coffee, but however much it is, it wildly eclipses the amount it would cost me to make a cup at home. We have a pretty good coffee maker here, and we grind our own beans. I love coffee, and can definitely tell the difference between good coffee and bad coffee, but I honestly don't think there's a difference in quality between what I can make at home and what I can buy at the coffee shop. The difference is really that I have to wait; I have to walk all the way home and wait for the darn stuff to brew. I also have to grind the beans, put water in the machine, and find my own cup. But to be honest, the difference is pretty negligible; I'm honestly not sure it's even rational for me to ever buy coffee from the shop, and usually I don't.

Then again, sometimes I do. The dollar or two difference buys me a small amount of convenience and instant gratification. But imagine if every time I was going to buy coffee from Espresso Royale, I stopped myself and put the money in an envelope and sent it to help desperately poor people in Africa. For a few dollars, you can buy quite a bit over there, even after those altruistic leeches in the non-profit agencies skim off their "administrative costs." In many countries, people live off of that much money per day. Someone who might have starved would be fed; someone who might have suffered horribly from an easily curable illness would be given the medicine they need.

Quick, tell me which will result in a greater increase in utility: (a) I get my cup of coffee at Espresso Royale instead of making it at home; (b) A child in Sierra Leone doesn't die from tetanus because she is given a vaccine paid for by me. To preempt the inevitable objections: let me reiterate, I am not saying anything about whether or not I should get the coffee. I am only appealing to the notion that it would be pretty ridiculous of me to say that I can't determine whether (a) or (b) results in more utility gained, because there's no scientific way to compare them. As Amartya Sen famously said, "Why must we reject being vaguely right in favor of being precisely wrong?"

That being said, I want to return to something important that I quoted Harrod as saying earlier: "Of course great caution must be exercised in not pushing the matter too far. Since the evidence is vague, we must not go farther than a very clear mandate from common sense allows." It is key to keep in mind that when we make interpersonal comparisons of utility, we are not measuring utility in a cardinal manner for each individual and comparing these measurements to each other. We can not coherently say that vaccinating the child in Sierra Leone produces 50 more utils than me getting my Espresso Royale coffee. We cannot even say that the vaccination produces 10 times more utility than I gain from my coffee. At least not with any hope of accuracy, we can't. Given this inability to precisely compare utility between individuals, it should be clear that while we might justifiably engage in interpersonal comparisons in extreme situations, we must be very careful when we approach the margins.

Hopefully that's enough to establish my point of view. As always, feedback is welcome and appreciated.

UPDATE:

Feedback I asked for, and feedback I received!

Please check out Jonatan Krovitsky's response to this post, and my reply as well. Also, Dmitry Chernikov had a few things to say about this post as well.