Love and Intrapersonal Utility Comparison

I come down firmly on Jonathan's side of the IUC debate, but (in addition to my comments on Brian's post), I have a different example to give.

Here is my theory: I believe that married people (or those in long-term relationships) are more likely to believe in the possibility of intrapersonal utility comparison (IUC). For what is love if not the embracing of your spouse's utility function, the integration of their happiness into your own? Husband and wife are no longer selfish individual maximizers, but instead maximizing some weighted combination of their interests.

And after some practice ("But I thought you would like a lime green couch"), most couples manage to do pretty darn well at it. It should be obvious that to do this, to effectively maximize mutual utility, IUC's are indispensable. A good spouse does them constantly, in fact, as part of every decision that affects both parties.

Now, you can argue solipsistically that no one can every truly *know* anyone else's feelings, and I grant that there is a great deal of truth to it. It is also true that the accuracy of any measurement of position and momentum is restricted to be at most:

heisenberg's uncertainty principle

But y'know, with this limited degree of accuracy granted to us by Heisenberg, we are able to measure position and momentum well enough to launch satellites, send men to the moon, and fire missiles thousands of miles with great accuracy. All this despite the fact that we can never really know exactly where those satellites, spaceships, and missiles are.

A measurement can have some error, and still be damn useful. And while I hate to make such an ad hominem argument, I can't help but wonder how many of the single men arguing against IUC will still believe it after a few years of marriage.


Related posts:

Why IUCs?
Cardinal Schmardinal, Ordinal Schmordinal
Encoding Happiness
IUCs and the Law of Large Numbers
No Soul Suggests IUCs
Futilitarianism
I-CDDFP
I-CUP
What color does a submarine weigh? (True or False?)
Exploding IUCs on the roadside
Interpersonal Utility Comparisons
Pareto Efficiency and Justice
Can the Paradox of the Non-Comparability of Interpersonal Utility be Resolved?

Share this

Becker did some work on one

Becker did some work on one person's utility function showing up in another person's function. That's not IUC.

The fact that I know what you prefer, that I know your preferences, doesn't mean that I can say who prefers something more than the other.

Preferences are not what makes people happy, they are not related to intensity of feeling or anything like that. Samuelson's original terminology was "selected over" rather than "preferred" when it came to revealed preference and the construction of the utility function.

I can select an apple over an orange. You can select an orange over an apple. But between us, there's no selection.

The billionaire and beggar example is hopelessly loaded. It's not that you don't know who would get more of a kick from the money, you do. The problem is that the language of welfare economics doesn't allow you to meaningfully formulate your hypothesis. Who selected what over what? If you can't formulate your hypothesis as an answer to that question then you're outside the realm of subjectivist welfare theory. It's not an issue of accuracy but of meaningfulness.

A happy marriage is not an

A happy marriage is not an example of useful utility comparisons, it's an example of how people can engage in mutually beneficial activities without needing to make such comparisons.

I can prefer to do something that makes you appear happy without any claim that I know your relative weighting, and certainly without any claim that my preference for your happiness is objectively correct.

Patri, You seem to be making

Patri,

You seem to be making it up as you go along. (nothing wrong with that, substitute any other criticism-indicating phrase that you may prefer)

If your purposeful action is to save your wife at the cost of your own life, whose value scale is being used?

There are no limits whatever on the content of your subjective value scale. It is entirely imaginable for your highest ranked subjective values to involve the long term survival of the human race and your purposeful actions to demonstrate those preferences.

Regards, Don

Two points: a) I don't think

Two points:

a) I don't think the denial of interpersonal utility comparisons is best understood as epistemological. It's not that there's some mysterious thing there that we can't know. Rather, utility is an ordinal ranking of preferences rather than a cardinal quantity and there's just no such thing as a comparison between two distinct ordinal rankings. (It's like asking "are you hotter than I am tall?")

b) On the other hand, I think it's a mistake to take the impossibility of interpersonal comparisons of utility to imply that we can never say that one person is better off than another. Well-being, while closely related to preference-satisfaction, is not the same thing as it. If I say that I'm better off than the person being tortured on the rack, I don't mean I have more utils (there are no such things as utils) or that I'm closer to some point on the preference scale than he is (our preference scales have no points in common) but rather that his life lacks some important components of Aristotelean-type flourishing that mine has.

Meaningful Love I have

Meaningful Love
I have previously defined meaningful statements: they are assertions that are either true by definition, known as tautologies, or they derive their meaning by way of testability. ?A triangle has three sides? is a tautology. ?I can fly? is a statement...

What is a marriage filled

What is a marriage filled with strife? Surely it is one in which the husband's preferences are in conflict with the wife's. For a preference is nothing but another way about talking about a choice, and if two people have conflicting preferences this means that the choices they make are in conflict.

Therefore for a marriage not to be filled with strife, the preferences must not be in conflict. There are of course many different ways to achieve this lack of conflict. One way is for the husband always to do whatever his wife says, regardless of what his preference would be in her absence. In terms of preferences, this means that his overarching preference is to do whatever his wife says, and everything other ordering is subordinate to that.

Another way is for the husband and the wife to align their preferences, so that he comes spontaneously to prefer what she prefers, and vice versa. This does not mean that if she likes to read romance novels then he now also likes to read romance novels. What it means is that if she likes to read romance novels then he likes *her* to read romance novels. And conversely, if he likes to do crosswords then she likes *him* to do crosswords.

In that case what you have is a couple each of whom are continually doing nice and thoughtful things for each other and anticipating what they like. Suppose that she sees a cookie. She might take the cookie and eat it. If he has the same preferences as she does, then if he saw the cookie first, he would take it and give it to her to eat. But on other occasions she might see the cookie and give it to him to eat.

Of course that is an idealized situation, but a husband and wife that approach this ideal will have events such as the husband spontaneously giving the cookie to his wife. Does it mean that the marginal utility for her is greater than it is for him? No, I think it means that he prefers that she have it than that he have it.

Here's a specific reason why I don't think we want to talk about comparing utilities: different couples might strike different balances. For example a more beautiful wife might have more the upper hand over the not so handsome husband, and so he might choose to give her the cookie - and she might choose to accept the cookie - more frequently than if she were plain and he were strikingly handsome.

Roderick, I agree 100%. I

Roderick,

I agree 100%. I tried to make both your points (and a lot of others). I look forward to responses.

The fact that I know what

The fact that I know what you prefer, that I know your preferences, doesn’t mean that I can say who prefers something more than the other.

As I pointed out in the comments to Brian's post, if I know your utility function, or a good approximation to it, then all I need is a single common point, a situation where I believe that you and I are receiving the same utility, and now I can compare the scales. Marriages and friendships both give many opportunities to find such calibrating points. It is not an exact science, but so what?

A happy marriage is not an example of useful utility comparisons, it’s an example of how people can engage in mutually beneficial activities without needing to make such comparisons.

But the better a couple makse such comparisons, the better they can engage in mutual beneficial activites. If I have no idea what my wife's utility function is, then every time I am individually planning a mutual activity, I have to ask her. "Dear, how do you want your steak cooked? What would you like to drink?" etc.

But if I learn her preferences, I can guess what she would prefer, and make good decisions for us even when she is not around. Add to those preferences some calibration, as described above, and wham, you have IUC.

Roderick - It all depends on what you mean by utility. For example, I am surprised that you distinguish it from well-being. If we assume that people prefer what maximizes their well-being, and we define utility as the ordering revealed by preferences, then they are the same thing.

I think saying things like "are you hotter than I am tall?" is pointless, since it is assuming your conclusion. Of course if my utility and my wife's utility are totally different things, then we can't compare them. But that is the question at hand, so stating that they are different is irrelevant.

Please explain what is so controversial about my saying "There is only one cookie left, and I believe my wife would enjoy it more than I."

Please explain what is so

Please explain what is so controversial about my saying “There is only one cookie left, and I believe my wife would enjoy it more than I.”

The controversy is that a statement like this is perfectly reasonable, but doesn't say what you think it does. Nobody is claiming that we can't each have models of each other's preferences, nor even that these models aren't comparable. We're (or I am) saying that the model is perfectly fine for making personal decisions, and is in fact all you have. But there is no basis for using any given model as a justification for violent taking of property or liberty.

"I think I'd rather see my wife express happiness than to enjoy the taste of the cookie" is completely internal, and there is no controversy. You can make that choice however you like, and using your model of yours and her preferences is great.

"I think she'd enjoy a cookie I don't own more than the shopkeeper would enjoy the $2 he's asking" may fit your model, but it's an invalid comparison.

My own model of my own current preferences is cardinal, but it's clear to me that my actual choices are ordinal (actually not even that, there is "choice taken" and "all other choices"). The fact that I KNOW my model doesn't match reality doesn't stop me from using it, but it does stop me using it as justification for forcing someone else to do something.

But there is no basis for

But there is no basis for using any given model as a justification for violent taking of property or liberty.

Since no one in this discussion has said that, isn't this a strawman?

Roderick - a thought I just had- I find it strange that you give “are you hotter than I am tall?" as an example of a meaningless statement, when there are perfectly reasonable ways of interpreting it. For example, if we normalize by subtracting the mean and dividing by the standard deviation, we can easily compare hotness and tallness.

I'm not saying this bears on any discussion of IUC. I just find it interesting that you give as an example of something incomparable something I think is comparable. I wonder if some of the difference here is between J & P personality types - those who want exact proofs and those comfortable with general statistical statements.

Ultimately I don't think any

Ultimately I don't think any of these are IUCs. The comparison you're really making with the cookie is that you'd get more enjoyment (or sex) out of giving it to your wife than you would from eating the cookie yourself. I mean, I only give my wife cookies in exchange for sex, for example. That doesn't mean I don't love her.

There is no basis for using

There is no basis for using any given model as a justification for violent taking of property or liberty.

Since no one in this discussion has said that, isn’t this a strawman?

You haven't made this argument, but it seems to be the point for Don's BillG island scenario, Sean's pareto efficiency post, and it's Brandon's least favorite argument against redistribution schemes. (note: I don't claim they're supporting mass redistribution of wealth, just exploring how to value such possible distributions).

If you're just saying "my model of my own and other's preferences is good enough for my decisions", I'm completely with you. If you're saying "my model is testably objectively correct", I think you're mistaken. And it's the second claim that often leads to coercive schemes.

Dagon, You haven’t made

Dagon,

You haven’t made this argument, but it seems to be the point for Don’s BillG island scenario, Sean’s pareto efficiency post, and it’s Brandon’s least favorite argument against redistribution schemes. (note: I don’t claim they’re supporting mass redistribution of wealth, just exploring how to value such possible distributions).

Your disclaimer to the contrary, I take objection to your characterization.

Distribution is both a state and an action. Even if there is 100% agreement that a given distribution state may appear preferable, that is no justification for any method of an active distribution other than MBVE.

Regards, Don

The comparison you’re

The comparison you’re really making with the cookie is that you’d get more enjoyment (or sex) out of giving it to your wife than you would from eating the cookie yourself.

No. I am saying that I believe she would gain more utility from eating it than I would. That fact is one input into my decision.

To make it clearer, there might be a case where doing A would make me a little happy and her very sad, doing B makes me indifferent and her a little happy. Clearly B is the better choice to maximize our combined utility, hence I might well do it, despite it making me less happy. In a long-term relationship, if both parties seek to maximize combined utility, they will likely come out better than if each seeks to maximize individual utility, or even than if they have explicit trades. Trades have transaction costs.

Strange. The fact that I

Strange. The fact that I have known my wife as intimately as I have for twenty years and can still buy a bad gift is a sign to me that IUCs don't exist.

I may be naive, but isn't the situation thus?

1) An Intrapersonal Utility Comparison may be proposed.
2) If it is correct, then each person validates it by choosing for it.
3) A validated IUC shows an overlapping range of differences in utility ("I thought the baker's cookie was worth MORE than my $2.00 and he thought it was worth LESS, so we swapped"). It does not show equality ("We both agreed the cookie's value was $2.00").
4) A large number of transactions throughout time or space or different parties lets you converge on a good statistical estimate of what would be equal intrapersonal utility.
5) Some unfortunate situations arise where two parties could not each validate a transaction in advance, so an arbiter may have to use the good statistical estimate.
6) Some people, through sloppiness or perhaps deceit, try to promote IUC to satisfy their own ends ("Jane just sold her house next door for $500,000, so mine is worth the same.")

When I buy a gift for someone, I often am trying to demonstrate to that person that I am familiar enough with them--in other words that I have seen enough significant choices they have made--that I know what is of value to them. If IUCs were guaranteed to be objective, it just wouldn't be as much fun.

I am saying that I believe

I am saying that I believe she would gain more utility from eating it than I would.

If you believed that your wife would gain more heavenly grace, that would not mean that there was such a thing as heavenly grace.

If you're going to reply, "but no one in this discussion has rejected that utility exists", part of the problem, I think, is that the word "utility" is being used in different ways, so that even if utility(1) exists, that does not mean utility(2) exists.

utility(1) is ordinal

utility(1) is ordinal utility, utility(2) is cardinal utility.

Another problem is where

Another problem is where should your $1000 fit in my utility function. Should it be substituted for the utility I find in my marginal $1000, or should it be measured as the utility I find in your having $1000. If I like you then I may like you to have $1000 and so I might make a gift of $1000 to you should I have lots of money, however I may hate you and so your having $1000 may irk me so much that I might spend my own $2000 to make sure you lose your $1000.

Should the ordinal rankings be from my perspective or yours, or some third parties?

Any attempt to establish a universal utility function assumes universal beneficience on the part of all participants. That is probably a bad assumption. Utility is based on individual self-interest which can be in conflict.

Patri, "Since no one in this

Patri,

"Since no one in this discussion has said that, isn’t this a strawman?"

If you can make good interpersonal utility comparisons then wouldn't it follow that a utilitarian would favor using force if he judged it would increase total utility based on those comparisons?

All other things being equal wouldn't you favor A using force against B if you judged that A benefitted more than B suffered, thus increasing net utility?

Patri, "Clearly B is the

Patri,

"Clearly B is the better choice to maximize our combined utility, hence I might well do it, despite it making me less happy."

If you choose B doesn't your choice reveal *your* utility?

All other things being equal

All other things being equal wouldn’t you favor A using force against B if you judged that A benefitted more than B suffered, thus increasing net utility?

Well, that does seem to be the unstated assumption of some of the pro-IUC participants (and anti-IUC), though a few of the pro-IUC participants have specifically, explicitly denied this implication. If your moral theory reduces morality to utility and if your theory of morality allows IUCs, then you are a utilitarian - i.e., you advocate hurting B to benefit A if the benefit to A is greater than the hurt to B. That is utilitarianism in a nutshell. But if your moral theory is not based on utility but on something else, then you can believe in IUC all you want and that doesn't make you a utilitarian.