Will Wilkinson has had an interesting back and forth with Conor Clarke and Matt Yglesias on income distribution. Most of this is based on Will’s paper income inequality, which held that income inequality in and of itself is not that important.

When the paper came out I didn’t say much about, because it matched what I already believed, namely:

  • Consumption inequality is more important than income inequality
  • Consumption inequality between nations is much more important inequality within nations
  • There is no real empirical support for the idea of class interests
  • Consumption at the high end probably yields a lot less direct happiness than consumption at the lower ends

It was this last point that caught the attention of several bloggers. If consumption doesn’t yield much at the high end they said, then it makes a lot of sense to take it away from wealthy people and give it to poor people. Conor went so far as to call Will a liberal hero for pushing this point.

I tend to agree that in and of itself declining marginal utility is an argument in favor redistribution. There are a couple of points that need to be made, however.

First, none of this undercuts the basic argument of Will’s paper which is that concentrating on income inequality is not particularly useful. It is entirely possible for someone to have high levels of income but moderate levels of consumption. Its not immediately clear that this is someone whose income it would be beneficial to redistribute.

Something is driving them to save so much. It could be paying off past debt. High earning professionals might have significant educational debt that they are paying off. It might be that this bit of income is transitory, a one year bonus.  Or, it might be that they have really high risk aversion. This last point, I think, is key.

Some people are hoarding money primarily out of anxiety. They may not voice it this way. They may think that their actions are the prudent thing to do. That people who don’t save so much are “flirting with disaster.” However, at the heart it is anxiety that is driving them to do it. Taxing away their savings exposes them to increasing anxiety and that’s a serious cost. After all, they were willing to forgo so much consumption to avoid it.

People with high income, but not particularly high consumption might also have lower than average happiness. They are driven to earn and save money because they want to avoid a negative emotion. The fact that this haunts them may mean that even at high levels of income they are not happy at all. I suspect that this is behind some of the finding that higher income doesn’t lead to greater happiness. Some people are high income because they are generally nervous people and they need the accumulate wealth to make themselves feel okay.

You might object that the investment banker with a house in the Keys and a her own airplane looks pretty happy. Yes, she does. But, now we are talking about someone with high income and high levels of consumption. Wouldn’t it be much more sensible to tax that consumption? The only people you miss in such a tax are precisely the one’s who for some reason are driven to squirrel away lots of money and never spend it.

Second, to rehash an old point, its typically true that the more we consume the less that happiness we get out of each additional unit of consumption. However, that doesn’t say anything directly about interpersonal comparisons of utility. Its entirely possible that the joy someone that someone receives from buying their 500th pair of shoes is much greater than I would get spending the same money.

I can tell you that many of them look like it gives them a great deal of pleasure. On the other hand I am not that much happier buying stuff even though my overall consumption is probably lower. From a utilitarian perspective it doesn’t make sense to transfer income from them to me. Indeed, this is one of the weaknesses in my view of the utilitarian approach. It doesn’t benefit me enough. No, more seriously under utilitarianism it is possible to justify extremely unequal distributions of happiness, so long as one person is a more efficient producer of happiness than another.

Utilitarianism suggests we should not work too hard on the depressive precisely because its so hard to get the depressive to feel better. Its more efficient to give those resources to the bubbly, who feel even more great because of it.

Those kind of results cut against the moral intuition that supports utilitarianism in the first place and makes me think a more consumption based Rawlsian approach is better.