We SHOULD be taxing tall rich men and subsidizing poor short women.
(most ungrammatical post so far)

That's the combined implications of two papers/articles:

Mankiw and Weinzierl
Alesina and Ichino (via Mark Thoma)

Personally, I don't see what the big deal is. What's so crazy about that?

The basic rational goes something like this:

Taxes are distortionary but you gotta raise some revenue.
So how do you max the rev subject to min-ing the dist?
Tax inelastic stuff (like intrinsic ability)
But what if you don't observe the inelastic stuff that matters (like intrinsic ability) but only some inelastic stuff that correlates with the inelastic stuff that matters (like height)?
Well, since we know height is correlated with income and both height and income are observable, height provides information on a person's intrinsic ability (note it does not say that height causes higher ability)
So tax the inelastic stuff that correlates combined with the variable, like income, that you care about.
Also women's labor supply is a lot more elastic than men's. So income taxes on men are more distortionary than on women.

Ergo,
"Optimal Taxation" involves more taxing of the more inelastic folks (men) for the sake of efficiency, more taxing of the richer people for the sake of "equality" and more taxing of the taller folks as a way of getting at the unobservable, inelastic, variable "ability".


Gabriel, and I think Mankiw, want to throw the whole thing out the window at this point.

But
Some taxes are necessary. We can argue about the level. Once we (possibly only hypothetically) agree on that, than it's all about who's gonna pay, there's no way we can avoid making this choice.
Any choice about the distribution of the tax burden essentially involves making a social welfare choice. Even if you say "I don't care who bears this burden" that's just a particular social welfare function. SW(distribution)=constant. Assuming instead that there's diminishing marginal utility of income (dollar to a rich person is worth less then a dollar to a poor person) is not that crazy. If you get that than there's a trade off between efficiency and redistribution if something like labor effort is a choice variable. Add in imperfect observability on the part of tax levying authority of this choice variable and you get the result.

(I've got a bit more thinking to do about the Alesin and and Ichino result - it works for the purposes of maximizing revenue but not necessarily welfare if you take household production into account)

The key thing about the Mankiw and Weinzierl result is that you're not really "taxing height" - you're "imperfectly taxing ability". Once you realize that and you're okay with taxing ability then a good portion of your "fairness" concerns should disappear. It's just Henry George in a brand new suit.


-----
(I've chosen the " " " (i.e. the quotation mark) as the punctuation symbol that I'm gonna wantonly abuse in this post, rather then the paranthases). Well, both. In my defense, I'll just say that Franz Kline is one of my favorite artists.

Of course this view of things could be just due to my economic education making me callous. But hey, logic convinces me and I don't find the premises crazy.

I'm average height. I have nothing to gain or loose on the height thing. I can suck it up and eat it on the gender aspect. This means that I'm a 96.4% objective outsider. Mankiw, who's 6'2'', is obviously suffering from "height guilt".

I hereby claim to be the first blogger ever to have put a quotion mark inside quotation marks. Getting crazy post-modern and all.