January 30, 2004

Does the declining marginal utility of money justify progressive taxation?

I'm having a slow day, which I ought to be spending finishing up this book review rather than surfing the blogs, but I'm surfing the blogs anyway.

I found this over at the Blog Delong, which does prove that not everyone over at the Volokh Conspiracy is completely loony, but that wasn't what I found interesting. It was this comment from Jonathan Goldberg, down almost at bottom, that struck me:

I've often heard claims like:

"If nothing else the bracket system stinks just on principle."

I've never understood them. I've always favored progressive taxation on a straight 19th century declining marginal utility of money/equality of pain/the wealthy benefit most from government arguement, with a little 20th century Rawlsianism and equality of opportunity thrown in. I've yet to see this (quite simple) reasoning even confronted in any plausible way (I have seen attempts, but they were obviously dishonest). Would anyone like to do so, or even provide a pointer (NOT to Nozick, please).

I'm going to take a shot at it, but not as an argument for lower incomes taxes. This is going to be an argument that the "declining marginal utility of money/equality of pain/the wealthy benefit most from government arguement" really doesn't justify enough taxation.

Why should different people have different amounts of wealth? We're not talking about money, per se, which is a commodity that can have very different values in different places and circumstances. We're talking about capacity to buy what you want. So, someone who is physically handicapped may need more money than someone who isn't in order to reach the same level of real wealth, or someone who lives in Manhattan needs more cash flow than someone living in Boise. Those are the cases I'm not talking about. I'm talking about how one person can live in a mansion with guards and servants and the next one lives in a roach-infested slum. or, for a less consciously class-specific case, why should you be able to buy a better car than me?

I can think of only three reasons that have ever made any sense to me:

1: Person A genuinely works harder than person B. If they put more of themselves into doing things that serve socially desireable ends - and for the moment we will assume that market value is good metre for deciding which ends are socially desireable - society owes them a bit more.

Note though that I am not claiming that the market value of their labour and/or the fruits of their labour is a good measure of how hard they really work. Even if we assume that it is a good measure of the social desireability of their work, that has nothing to do with how much they have really put themselves into it. The man who makes two million dollars a year can't possibly be putting a hundred times more of himself into his work than the man who makes $20,000 a year working 40 hours a week at menial jobs. There isn't a hundred times more self to give.

This argument justifies taxing the rich man far, far more than anything the marginal utility of money argument could allow for. It justifies redistribution to the point where each is rewarded for their efforts rather than the social value of their work.

2: Person A's work is more socially desireable than person B's, and therefore higher compensation is necessary as an incentive for each person to do the kinds of work that are most socially desireable.

This argument is standard to free-market ideas about wages and compensation, and it's a good argument built on some fairly sound principles. However, it imagines that choosing between the kinds of work we do is as easy as choosing between different brands of potato chips at the grocery store. It would justify radical differences in wealth if I just needed to choose to do the same kind of work that Bill Gates does in order to make what Bill Gates makes. If that were the case, I imagine a free labour market would quickly produce a just and efficient distribution of labour without any need for redistribution or intervention. However, that is patently not the case.

An argument based on a heuristic reward for doing socially desireable work only justifies differences in wealth large enough to efficiently distribute labour. There is very little reason to believe that being a corporate CEO is so onerous and unpleasant a job that people will only do it if we offer them millions of dollars a year. While this argument does justify moderate differences in income on the basis of the kind of work and the quality of the work you do, I can't see anyway to use it to justify the income spreads found even in the most egalitarian countries.

3: Personal wealth is justified not as something the wealthy person has earned, but as a social charge which the holder is bound to use in socially desireable ways. Empirical evidence suggests that no other kind of agent except the individually wealthy person is able to allocate investment so efficiently.

This one is a tricky one, but it's really very close to a lot of arguments in favour of less taxation on the upper income brackets. It's also hard to argue against because the first part is actually in the Old Testament, the Quran and is even implied by Confucius. When all three great Mesapotamian monotheisms agree - and Confucius says so - it's hard to make a case against it.

The classical attack is on the second part, particularly because it is a debatable empirical claim. A great deal of wealth is, in fact, invested by agents other than individual wealthy people, and there is no particular reason to think that those investments have been uniformily worse than individuals' allocations. I'm not thinking of government allocation. Mutual funds, banks and companies, for example, do quite a lot of investing. But, that line of argument gets long and complicated, so I'm going to try another approach:

If personal wealth in excess of what can be justified by other arguments is a social charge, then it should be taxed at a rate proportionate to a reasonable return on investment. For example, we could set wealth taxes in proportion to the interest rate on long-term treasury bonds. If wealth beyond its personal utility has to exist for the sake of economic efficiency, then the wealthy should have to demonstrate their personal skill at efficiently allocating it by paying a tax will take their money away from them if they can't get a better rate of return on it than they could if they gave it to the government to invest.

This is, therefore, an argument not for the progessive taxation of income but for the taxation of privately held capital. Earned money falls under arguments 1 and 2, and argument 3 tries to justify an at least flat-rate taxation of assets directy. Not only does the government take more if you earn more, it takes money just because you have money and you're failing to earn enough with it.

I've been tinkering with this last argument for a little while and looking to see if anyone else has advocated it and what counter arguments there are. So far, no dice, but I'm not very well read in economics. It's a really very tempting approach because it says that if large personal fortunes, private ownership of capital and individual possession of the means of production really are economically efficient, then a tax that redistributes money away from poor investors should raise the efficiency of capital allocation, right? It sort of dares the free-market true believer to put his money where his mouth is by taking it away if he can't make better investments than the state.

Still, I think this line of attack on progressive income taxation leads to the notion that merely progressive taxes aren't taxing enough. As for arguments of the "who gave the government the right to have all the money that isn't really earned?" variety, I would offer this response: A person who consumes more is by definintion consuming more of the goods produced by the government. Even if they do not personally consume them, the value created by government activities is embodied in other things that they consume simply because the other people who make them also consume government services. There is, therefore, ample justification for taxes designed to equalise real wealth.

Only the argument from economic efficiency has any traction these days. No one believes that Bill Gates works a million times harder than the guy who sweeps his office or that just anyone can become a millionaire through hard work. A personally wealthy class can only be justified by economic efficiency, and if that's the case then the tax code ought to reflect it.

Posted 2004/01/30 14:49 (Fri) | TrackBack

"The man who makes two million dollars a year can't possibly be putting a hundred times more of himself into his work than the man who makes $20,000 a year working 40 hours a week at menial jobs. There isn't a hundred times more self to give"

what if i design robots that do the work of 100 people?

Posted by: Shai at February 10, 2004 14:58

That's a claim about the productive value of your work, not the labour you've put into it.

Posted by: Scott Martens at February 10, 2004 15:32

Can you explain these theme more attentievly, please.... I didn't understand...

Posted by: автомобили at April 12, 2004 16:08

the salary people can command has nothing to do with labor but everything to do with the value they generate to the employer.

old socialist ideals like reward being based on just physical use of energy are dead.

Posted by: deus at May 21, 2004 7:21

Kind of depressing that such an interesting blog post got such useless comments.
I wish I had something good to add, but I don't really -- I was just investigating object-valuation ideas for some programming I'm working on, and poking around for information on what people think of marginal utility, and came across this.
Damned oppression of the lower class.

Posted by: Mike at May 21, 2004 23:39
Post a comment

Remember personal info?