Friday, December 03, 2004

Interpersonal Comparisons of Utility

We learn in economics that you cannot do interpersonal comparisons of utility because a level of utility has no meaning outside of variations within a person: it is ordinal not cardinal. That is, if I assign a value of $10 of utility to an activity, and you assign a value of $10 of utility to an activity, then we really cannot say how our levels of utility compare. Sure we value it the same in dollar terms, but how do we value dollars. So what's interesting about this article is the claim that you can, in fact, make interpersonal comparisons of happiness -- which we could understand, from an economists' perspective, to be utility. The Times article isn't extensive enough on method to know for sure, but it does seem that the researchers claim that you can make explicit comparisons of happiness across people, else you couldn't come up with all the data they do. For example, the income statistic I cited above. Unless the participants had changes in income, then to claim that variations in income did not have an effect on happiness means that you have to have made such an interpersonal comparison of happiness.

My point is that utility and happiness are essentially subjective measures which have no objective measure. And if you ask people to put themselves on a scale, then for two different people a 7 may mean very different things. After all, we don't know what it means for someone else to be "really" happy or "slightly" happy, we only know what that means for us. Someone might just always be 100 times happier than me, but such a scale wouldn't reflect this. So all this study can say is not that certain things make people more or less happy in some objective sense, but that these activities put people at such and such points on their own happiness scales. Some people or cultures may have a tendency to say that they are less happy, even though they are at the same level in some indeterminate objective sense.

Now Daniel Kahneman who did this study is a hell of a lot smarter than me: he's a Nobel Laureate in economics and has probably thought of these things. Given that economists are so often criticised for being unsubtle about human emotions, it is interesting that a psychologist thinks that it is possible to even present the results as somehow compiling and comparing happiness when economists have concluded that such an idea is ludicrous because each person's happiness is so subjective.

3 Comments:

Blogger brad said...

I wouldn't say that we cannot make interpersonal comparisons of utility. I would say that we choose not to do so because it opens lots of cans of worms on which we cannot make much progress.

However, mark me down as somebody who thinks questions like Do I know that it hurts you more to have your arm ripped off by wolverines than it hurts me to be bitten by a mosquito?" and Do I know that you are more disappointed by watching your child die than I am disappointed to find that the local Starbucks has run out of low-fat coffeecake?" do in fact have unambiguous answers...

7:16 PM  
Blogger Isaac said...

Brad,
You're right that we can subjectively say that a mosqitoe bite hurts less than such and such, but is there some clean, technical way of doing it? Based on Henry's above post, it seems that they have tried, but can it work?

8:25 PM  
Blogger Isaac said...

Brad,
You're right that we can subjectively say that a mosqitoe bite hurts less than such and such, but is there some clean, technical way of doing it? Based on Henry's above post, it seems that they have tried. The point isn't that it's impossible for me to judge the difference, the point is tha it's impossible to express in a numerically meaningful way.

8:25 PM  

Post a Comment

<< Home