Hacker News Reader

Comment by rosser | Hacker News Reader

The marginal utility of a tax dollar is greater to someone who makes less than I do than it is to me.

EDIT: s/value/utility/

> The marginal value of a tax dollar is greater to someone who makes less than I do than it is to me.

Perhaps you mean marginal utility? Strictly in terms of value, the wealthy tend to trade that dollar for assets rather than consumption. Marginal value of any dollar is higher when it generates a return.

If you do mean marginal utility, there is a strong philosophical position that interpersonal utility comparisons are not meaningful. Diminishing marginal utility holds for any one person, but subjective preference and merely ordinal (rather than cardinal) preference confounds any attempt at interpersonal comparison.

Yes, I meant marginal utility.

I think it's going to take rather more than a "strong philosophical position" to convince me that 1/100000 is somehow not bigger than 1/100000000.

I think he was trying to talk about MPC. http://en.wikipedia.org/wiki/Marginal_propensity_to_consume

I'm curious what the philosophical position is that says that $1,000 is not worth meaningfully more in value or utility to a poor person than an equity fund manager, and why that position would be described as "strong".

There is no known measurement of utils and good reason to believe there never will be. Our strongest insight into subjective preference is revealed preference. Without action, preference is not revealed, and where preference is revealed by action, we can only know that one preference is higher than another -- ordinal. We cannot determine that one preference is, say, 10x higher than another. These are the limitations of utility analysis for a single person.

Diminishing marginal utility holds for a single person, and it ought to hold for interpersonal comparisons ceteris paribus, but of course ceteris paribus is a logical impossibility for interpersonal comparison given subjective preference. To hold ceteris paribus for two people is to assert they are the same person, and we're back to diminishing marginal utility for a single person.

Our only guide to interpersonal utility comparison is intuition, which is easily confounded. There is no way to say that my number one preference has more utility than yours, and we have even less reason to say anything meaningful about my number X preference compared to your number Y preference.

I don't think there are many people suggesting that a poor man's dollar has 9.7438393 times more utility to him than does a rich person's. The question is one of more or less not how much more or less. Trying to frame it that way is a distraction, if not a straw man.

So, to clarify your position: absent some specific ratio, is a single dollar worth more to a person of modest means than it is to someone wealthy? Or is even that degree of comparison meaningless in your estimation?

I'm afraid you've misunderstood. I said that we cannot even determine more or less:

> There is no way to say that my number one preference has more utility than yours

== Ninja edit in response to clarification ninja edit: ==

That is correct. We have no reason, other than intuition, to believe that a dollar is "worth more" to a poor person than a rich person. I can easily give some counterexamples:

    * Poor person is a monk
    * Poor person believes in barter rather than purchase
    * Poor person has no counterparties that accept dollars
    * Poor person has material needs satisfied
    * Poor person has wacky preferences

Even if you want to try to hold external circumstances equal, several of these examples show the dollar to be equally worthless to both parties. Separately, the fact of internal circumstances being subjective -- unable to be held equal -- confounds attempts at comparison.

All of your counterexamples seem to me to be speciously unlikely, and so under-representative of reality that I honestly find myself wondering whether you're trolling.

Let me ask the question even more pointedly: to whom would the unexpected receipt of $1000 be more beneficial, a single, paycheck-to-paycheck mother who drives a '90s Accord, or the scion of a squillionaire tycoon, trying to decide between different colors of leather for his new Ferrari?

Do you still maintain those things can't be compared?

> Do you still maintain those things can't be compared?

Only intuition tells us they can be compared. Utility theory given subjective preference gives us no tools to make the comparison nor any way of verifying it. We have only the pretense of knowledge. Compare to e.g. diminishing marginal utility for a single person, which has a clear theoretical and verifiable basis.

Here is why intuition is misleading: we can easily intuit the fact of our own diminishing marginal utility. Then we can say: if I was poorer, diminishing marginal utility makes the marginal dollar worth more to me. However, this does not hold across a different set of subjective preferences. Our intuition is strongly biased against the full recognition of subjective preference because it operates by injecting our own subjective preference onto others. When we imagine how some other person lives, we really imagine ourselves in that position. Unfortunately, we can only know others' preferences via revealed preference. Intuition is a poor guide.

Furthermore, for any specific (and particularly extreme) example, our intuition might be strong enough to dismiss theoretical concerns. However, this does not generalize. Finally, if the mother prefers to spend the $1000 on crack, then any choice of leather is more beneficial.

There's an inherently subjective aspect to money anyway, though. The whole basis of capitalism is that people trade because it's worth it to both parties, and that value is created out of thin air due to the trading. It seems specious to object to progressive taxation on the grounds of "subjective value", while still defending capitalism itself.

As you noted, one of the wonders of capitalism (or free trade, or the right to transact, or the right to refuse to transact) is that every voluntary trade increases net wealth. Only when both parties are made better off does voluntary trade occur. This concept is very mysterious without the understanding of subjective preference.

My objection is only to the specific justification of progressive taxation based on interpersonal utility comparisons. There are other justifications of progressive taxation not subject to my objection.

This has nothing to do with interpersonal utility comparisons. It has everything to do with the very real, undebatable fact that ratios with smaller denominators are larger than ratios with larger denominators provided that the numerators are the same.

> It has everything to do with the very real, undebatable fact that ratios with smaller denominators are larger than ratios with larger denominators provided that the numerators are the same.

Tautologies are true, and identities necessarily hold. You are saying that a single dollar is a larger portion of a smaller net worth than that of a larger net worth, I gather. Of course this is true. The relevance remains unclear. It's the next step that is treacherous.

Reply on news.ycombinator.com