Tax height?

[Surlethe]

Paper by Greg Mankiw [pdf]

This paper can be interpreted in one of two ways. Some readers can take it as a small, quirky contribution aimed to clarify the literature on optimal income taxation. Others can take it as a broader effort to challenge that entire literature. In particular, our results can be seen as raising a fundamental question about the framework for optimal taxation for which William Vickrey and James Mirrlees won the Nobel Prize and
which remains a centerpiece of modern public finance.

More than a century ago, Francis Y. Edgeworth (1897) pointed out that a Utilitarian social planner
with full information will be completely egalitarian. More specifically, the planner will equalize the marginal
utility of all members of society; if everyone has the same separable preferences, equalizing marginal utility
requires equalizing after-tax incomes as well. Those endowed with greater than average productivity are
fully taxed on the excess, and those endowed with lower than average productivity get subsides to bring
them up to average.

William S. Vickrey (1945) and James A. Mirrlees (1971) emphasized a key practical di¢ culty with
Edgeworth’ solution: The government does not observe innate productivity. Instead, it observes income, which is a function of productivity and effort. The social planner with such imperfect information has to
limit his Utilitarian desire for the egalitarian outcome, recognizing that too much redistribution will blunt
incentives to supply e¤ort. The Vickrey-Mirrlees approach to optimal nonlinear taxation is now standard.
For a prominent recent example of its application, see Emmanuel Saez (2001). For extensions of the static
framework to dynamic settings, see Mikhail Golosov, Narayana Kocherlakota, and Aleh Tsyvinski (2003),

Vickrey and Mirrlees assumed that income was the only piece of data the government could observe about
an individual. That assumption, however, is far from true. In practice, a person’ income tax liability is a function of many variables beyond income, such as mortgage interest payments, charitable contributions,
health expenditures, number of children, and so on. Following George A. Akerlof (1978), these variables
might be considered "tags" that identify individuals whom society deems worthy of special support. This
support is usually called a "categorical transfer" in the substantial literature on optimal tagging (e.g., Mirrlees
1986, Ravi Kanbur et al. 1994, Ritva Immonen et al. 1998, Alan Viard 2001a, 2001b, Louis Kaplow 2007).
In this paper, we use the Vickrey-Mirrlees framework to explore the potential role of another variable: the
taxpayer’ height.

The inquiry is supported by two legs— one theoretical and one empirical. The theoretical leg is that,
according to the theory of optimal taxation, any exogenous variable correlated with productivity should be
a useful indicator for the government to use in determining the optimal tax liability (e.g., Saez 2001, Kaplow
2007).1 The empirical leg is that a person’ height is strongly correlated with his or her income. Judge and
Cable (2004) report that “an individual who is 72 in. tall could be expected to earn $5,525 [in 2002 dollars]
more per year than someone who is 65 in. tall, even after controlling for gender, weight, and age.” Nicola
Persico, Andrew Postlewaite, and Dan Silverman (2004) ...nd similar results and report that "among adult
white men in the United States, every additional inch of height as an adult is associated with a 1.8 percent
increase in wages." Anne Case and Christina Paxson (2008) write that "For both men and women...an
additional inch of height [is] associated with a one to two percent increase in earnings." This fact, together
with the canonical approach to optimal taxation, suggests that a person’ tax liability should be a function of his height. That is, a tall person of a given income should pay more in taxes than a short person of
the same income. The policy simulation presented below con...rms this implication and establishes that the
optimal tax on height is substantial.

Many readers will find the idea of a height tax absurd, whereas some will find it merely highly unconven-
tional. We encourage all readers to consider why the idea of taxing height elicits such a response even though
it follows ineluctably from a well-documented empirical regularity and the dominant modern approach to
optimal income taxation. If the policy is viewed as absurd, defenders of this approach are bound to o¤er
an explanation that leaves their framework intact. Otherwise, economists ought to reconsider whether this
standard approach to policy design adequately captures people’ intuitive notions of redistributive justice.

The remainder of the paper proceeds as follows. In Section I we review the Vickrey-Mirrlees approach to
optimal income taxation and focus it on the issue at hand— optimal taxation when earnings vary by height.
In Section II we examine the empirical relationship between height and earnings, and we combine theory
and data to reach a ...rst-pass judgment about what an optimal height tax would look like for white males
in the United States. We also discuss whether a height tax can be Pareto-improving. In Section III we
conclude by considering some of the reasons that economists might be squeamish about advocating such a
tax.

I think the idea has some merit. After all, as the paper points out, if all else is equal a one-inch increase in height leads to a 1.8% increase in income. Taxing height makes sense, if only out of a sense of fairness: tall people have an obvious social advantage over short people, so tall people should be handicapped and short people subsidized.

[haard]

Oh yes, and anyone above average intelligence should be taxed and stupid people subsidized. After all, intelligent people may spend less effort to reach the same productivity.

If the policy is viewed as absurd, defenders of this approach are bound to offer an explanation that leaves their framework intact.

Sure. I'm not utilitarian. I view it as absurd that redistribution and fairness should be viewed as a goal of it's own.

[Surlethe]

Oh yes, and anyone above average intelligence should be taxed and stupid people subsidized.

Doesn't the law cater to stupid people anyway? I mean, for example, we have anti-smoking laws, anti-drunk-driving laws, and seat belt laws just to protect people from their own stupidity. In any case, if one uses fairness as a criterion, why wouldn't one tax above-average (inherent) intelligence? After all, stupid people are at a decided disadvantage; if we're willing to tax rich people and give it to poor - ostensibly to level the playing field - why not do the same with height or intelligence or other attributes people can't control?

[haard]

A) Anti-smoking and anti DIU-laws is or should be to protect others from you; not to protect you from yourself
B) Well, as I said, I'm not really all that hot for levelling the playing field.

I'm for having a government that supplies good education, health care, and infrastructure (while keeping private alternatives), since (I believe) that it promotes a society that produces more wealth and has less friction than the alternatives; I'm not really into redistributing wealth out of fairness.

[Edi]

I'm not really into redistributing wealth out of fairness.

Probably more accurate to say that you are for redistributing wealth to some degree of fairness but not to the degree of leveling everyone to the same position regardless. That's the outcome of your position. Note that this is not an indictment of it, I agree with it mostly without alteration.

More important is that opportunity is equalized as much as possible, because what happens after that depends largely on the individual and what he makes of the opportunity.

[Darth Wong]

Sure. I'm not utilitarian. I view it as absurd that redistribution and fairness should be viewed as a goal of it's own.

So what is your moral framework, then? Gut instinct?

[haard]

Sure. I'm not utilitarian. I view it as absurd that redistribution and fairness should be viewed as a goal of it's own.

So what is your moral framework, then? Gut instinct?

Of course 'gut instinct' is at the root of it.

I'm leaning more towards negative utilitarianism. I am aware that taken to it's logical end NU is... quite emo (the whole 'destroy the world'-thing), but I find that when tempering it with the recognition that positive utility does have a value, it is more in line with my gut instinct, as you put it. I have never found a hard-and-fast ethical theory or framework that I can honestly say I would support fully. The world not being ideal, I don't expect I'll ever find one.

[Solauren]

If we start taxing a single physical attribute that people have no real control over, will it stop with just one?

Suppose brunettes make more money then other hair colors. Will we tax that?
Suppose people with blue eyes make more moeny than other eye colors. Will we tax that?

I'm sorry, the idea of taxing someone on something like physical attributes... it almost sounds like a form of prejudice to me.

[Darth Wong]

Sure. I'm not utilitarian. I view it as absurd that redistribution and fairness should be viewed as a goal of it's own.

So what is your moral framework, then? Gut instinct?

Of course 'gut instinct' is at the root of it.

I'm leaning more towards negative utilitarianism. I am aware that taken to it's logical end NU is... quite emo (the whole 'destroy the world'-thing), but I find that when tempering it with the recognition that positive utility does have a value, it is more in line with my gut instinct, as you put it. I have never found a hard-and-fast ethical theory or framework that I can honestly say I would support fully. The world not being ideal, I don't expect I'll ever find one.

In other words, you are incapable of making consistent ethical arguments, and your contributions on such subjects should be ignored for the worthless no-thinking tripe that they are.

[Starglider]

I'm leaning more towards negative utilitarianism.

That makes literally no sense.

Utilitarianism is unfortunately often given an overly narrow meaning by people trying to claim that their particular preferences are some kind of ultra-rational ideal. In fact utility theory is simply a consistent and provably rational way of making choices, given some arbitrary set of goals. Not all goal systems can be specified as a (standard) utility function, but frankly the ones that can't are basically esoteric logic problems of interest only to philosophers. What utility theory does is force you to admit how much arbitrariness there is in your goal system.

Utilitarianism as a philosophy, rather than utility theory in the field of decision theory, attempts to maximise the benefits experienced by the population as a whole, adjusting for declining marginal utility and assorted other factors. However it does not rule out the notion of 'deserving' a better standard of living; in fact only absurdly ultra-communist versions do. Simply apply a weighting to an individual's utility in some proportion to their economic output and you will arrive at a reasonable goal system. Personally I would regard income as a rather poor indicator of economic output - an unfortunate flaw in our society. Things like property rights are factored in relatively easily by assigning negative utility to violations of them, such that taking something away from someone and giving it to someone else is bad even if the amount of utility lost by the first and gained by the second is equal (personally I would apply this to actual possessions but not money and similar impersonal liquid assets). Finally you have the long term strategic concern of maximising economic growth versus addressing current inequality; again solvable by declining marginal utility.

It's generally impractical for humans to deal with all that but this is how the goal system in a (rational) general artificial intelligence works.

[Surlethe]

If we start taxing a single physical attribute that people have no real control over, will it stop with just one?

Suppose brunettes make more money then other hair colors. Will we tax that?
Suppose people with blue eyes make more moeny than other eye colors. Will we tax that?

I'm sorry, the idea of taxing someone on something like physical attributes... it almost sounds like a form of prejudice to me.

No, it would correct prejudice. If someone with blue eyes makes more money than other eye colors, since there's no conceivable connection between eye color and merit, the wage discrepancy must be the product of underlying social discrimination. A tax on these attributes, therefore, would act to correct the discrimination.

[Patrick Degan]

If I'm reading Mankiw's piece right, this is a half-assed attempt to parody the concept of progressive taxation. Evidently, he thought he was being Jonathan Swift that day and probably patted himself on the back for his alleged cleverness when he hit the "send" button.

[haard]

I'm leaning more towards negative utilitarianism. I am aware that taken to it's logical end NU is... quite emo (the whole 'destroy the world'-thing), but I find that when tempering it with the recognition that positive utility does have a value, it is more in line with my gut instinct, as you put it. I have never found a hard-and-fast ethical theory or framework that I can honestly say I would support fully. The world not being ideal, I don't expect I'll ever find one.

In other words, you are incapable of making consistent ethical arguments, and your contributions on such subjects should be ignored for the worthless no-thinking tripe that they are.

That's retarded. The fact that I don't belive that I'll find the magic algorithm to make my decisions for me is not an indication that I cannot make consistent arguments.
It means that I acknowledge the fact that at least none of the philosophies I've evaluated so far satesfies me to the point that I can say: "this is what I believe, and this will guide my actions" since they all have flaws that means I am not prepared to take them to their ultimate ends. The fact that I don't expect to ever be satisfied in this regard is a result of that I dont actually spend a significant portion of my life trying to formulate a moral framework, or even reading what others have already formulated, and my experience is that finding or coming up with a complete moral framwork where I can live with all the consequences is far from trivial. Also, it is not neccecarily so that it is possible, since mind's rule over emotion is imperfect at best.

[haard]

I'm leaning more towards negative utilitarianism.

That makes literally no sense.

Utilitarianism is unfortunately often given an overly narrow meaning by people trying to claim that their particular preferences are some kind of ultra-rational ideal. In fact utility theory is simply a consistent and provably rational way of making choices, given some arbitrary set of goals. Not all goal systems can be specified as a (standard) utility function, but frankly the ones that can't are basically esoteric logic problems of interest only to philosophers. What utility theory does is force you to admit how much arbitrariness there is in your goal system.

Utilitarianism as a philosophy, rather than utility theory in the field of decision theory, attempts to maximise the benefits experienced by the population as a whole, adjusting for declining marginal utility and assorted other factors. However it does not rule out the notion of 'deserving' a better standard of living; in fact only absurdly ultra-communist versions do. Simply apply a weighting to an individual's utility in some proportion to their economic output and you will arrive at a reasonable goal system. Personally I would regard income as a rather poor indicator of economic output - an unfortunate flaw in our society. Things like property rights are factored in relatively easily by assigning negative utility to violations of them, such that taking something away from someone and giving it to someone else is bad even if the amount of utility lost by the first and gained by the second is equal (personally I would apply this to actual possessions but not money and similar impersonal liquid assets). Finally you have the long term strategic concern of maximising economic growth versus addressing current inequality; again solvable by declining marginal utility.

It's generally impractical for humans to deal with all that but this is how the goal system in a (rational) general artificial intelligence works.

I am quite aware how rational agents and utility systems work.
Your (purely economic?) 'progressive' utilaritism does not remove the problems with mere addition, or that both total and avarage has it's own distinct and well-known problems, but I will conceed that with sliding utility functions and enough utility assigned to some rights - like the right to live - rule (or ideal information act) utilaritism has a lot of good points, and not that many bad. Maybe I'll be converted yet, but that's perhaps straying a bit far from the topic?

[Solauren]

If we start taxing a single physical attribute that people have no real control over, will it stop with just one?

Suppose brunettes make more money then other hair colors. Will we tax that?
Suppose people with blue eyes make more moeny than other eye colors. Will we tax that?

I'm sorry, the idea of taxing someone on something like physical attributes... it almost sounds like a form of prejudice to me.

No, it would correct prejudice. If someone with blue eyes makes more money than other eye colors, since there's no conceivable connection between eye color and merit, the wage discrepancy must be the product of underlying social discrimination. A tax on these attributes, therefore, would act to correct the discrimination.

I'm sorry, but without a detailed breakdown on a large cross-section of the working population, that accounts for things like qualifications, work habits, personal habits, time with job (as in hiring date), time at job (as in average hours per week), age, previous work experience, accomplishments at work, and a load of other things, this smells like bullshit to me.

[Samuel]

The idea is ridiculous for a very simple reason- we tax people on their proven ability to pay, not their possible ability to pay. I might be able to pay a million dollars in taxes... or just 750. How much I pay is decided on how much money I actually make, not how much I might be making.

[Darth Wong]

In other words, you are incapable of making consistent ethical arguments, and your contributions on such subjects should be ignored for the worthless no-thinking tripe that they are.

That's retarded. The fact that I don't belive that I'll find the magic algorithm to make my decisions for me is not an indication that I cannot make consistent arguments.

Bullshit. You just admitted that you ultimately rely on gut instinct if you can't come up with a rationale. In other words, whatever rationalized ethics system you use is just window-dressing on your arbitrary personal judgments.

[Darth Wong]

The idea is ridiculous for a very simple reason- we tax people on their proven ability to pay, not their possible ability to pay. I might be able to pay a million dollars in taxes... or just 750. How much I pay is decided on how much money I actually make, not how much I might be making.

This is the best point. Or, to put it more succinctly: the only reason someone would come up with a scheme like this is discomfort with straightforward progressive taxation, which is a far more accurate way of targeting people with greater means to pay.

[Starglider]

That's retarded. The fact that I don't belive that I'll find the magic algorithm to
make my decisions for me is not an indication that I cannot make consistent arguments.

In which case your decisions are actually made by a random mess of unexamined (and probably quite ugly) heuristics, and furthermore you have no rational basis for trying to convince other people to agree with your judgements.

It means that I acknowledge the fact that at least none of the philosophies I've evaluated so far satesfies me to the point that I can say: "this is what I believe, and this will guide my actions" since they all have flaws that means I am not prepared to take them to their ultimate ends.

Please give examples, showing that you have actually given this some thought.

Also, it is not neccecarily so that it is possible, since mind's rule over emotion is imperfect at best.

You can get away without any formal moral framework if you never have a position of significant power or influence, assuming you like being ignorant of how your own mind works. Again though you lose the ability to make any rational arguments about the morality of other people's actions (all you can say is 'yeah I like it' or 'no that seems wrong').

Your (purely economic?) 'progressive' utilaritism does not remove the problems with mere addition

Which are?

or that both total and avarage has it's own distinct and well-known problems

Which are?

I will conceed that with sliding utility functions and enough utility assigned to some rights - like the right to live - rule (or ideal information act) utilaritism has a lot of good points, and not that many bad

Rule-based utility is not consistent or rational; only outcome-based expected (probabilistic) utility is demonstrably consistent (and optimal, for most sane definitions of the word). Most attempts to assign utility to actions and rules can be expressed in saner outcome-based terms with a little effort. Though I guess you may be using an excessively narrow definition of 'outcome' that refers to economic outcomes only.

This is immediately apparent in AI, where a goal system that assigns utility to actions (or other, even more sloppy systems such as spreading activation from attractors) not only perform worse, they are also get very unpredictable as the complexity of the system increases. Acting based on local heuristics and local preference functions over actions produces an inconsistent mess that easily produces bizarre emergent results (mostly pruned out of humans by millions of years of evolution, but we'd still look fairly bizarre to a purely rational intelligence). By contrast expected utility generally produces smooth, predictable, sensible outcomes (though the actions taken to reach that point may be quite surprising).

[Darth Wong]

The problem haard has is that he thinks he's being reasonable by merely understanding various moral concepts, even if he does not grant them any authority. The problem is that he judges moral concepts by whether they suit his personal instincts, rather than judging his instincts by how they match up against moral concepts.

[Surlethe]

No, it would correct prejudice. If someone with blue eyes makes more money than other eye colors, since there's no conceivable connection between eye color and merit, the wage discrepancy must be the product of underlying social discrimination. A tax on these attributes, therefore, would act to correct the discrimination.

I'm sorry, but without a detailed breakdown on a large cross-section of the working population, that accounts for things like qualifications, work habits, personal habits, time with job (as in hiring date), time at job (as in average hours per week), age, previous work experience, accomplishments at work, and a load of other things, this smells like bullshit to me.

I'm taking for granted that such a breakdown has established a wage discrepancy in the case of blue eyes, as it actually has for height.

[Surlethe]

The idea is ridiculous for a very simple reason- we tax people on their proven ability to pay, not their possible ability to pay. I might be able to pay a million dollars in taxes... or just 750. How much I pay is decided on how much money I actually make, not how much I might be making.

How does this contradict the idea of taxing height? It's not like (in an idealized system) a poor six-foot-nine person is going to be asked for $5000 every April, or this tax is being proposed to supplant a progressive tax code.

There's something a little bit subtler than simply measuring proven ability to pay here: this is, in some sense, a measure of how hard has someone worked to gain what he earns. It seems to me that if a six-foot man and a five-foot man, otherwise identical, both make $180000 per year, the six-foot man has not had to work as hard to attain his job because there are social advantages attached to being of above-average height. So if we wish to move the situation closer to a meritocracy, it makes sense to tax the six-foot man and subsidize the five-foot man so that their incomes reflect the effort they have to put into earning money. (Note that this would augment an already-progressive system; the tax code would already target those who have ability to pay.)

[Dark Hellion]

The problem haard has is that he thinks he's being reasonable by merely understanding various moral concepts, even if he does not grant them any authority. The problem is that he judges moral concepts by whether they suit his personal instincts, rather than judging his instincts by how they match up against moral concepts.

While true in haard's case, there is a level of subjectivity that will always be included within moral judgement, regardless of the adherence to moral concepts. Primarily, this comes from the fact that morallity is a 'science' that requires so much contextualization that one cannot render a perfectly rational set of concepts that is capable of operating over all ethical dilemmas; even simple dilemmas are fully capable of confounding extremely well thought out and exacting frameworks of moral concepts.

For a simple example, lets take a look at a purely utilitarian view of rape, in this case aquantance rape. Say that we know that the culprit only did so because they were intoxicated and in any normal state of mind would not. Let us also say that while tramatized the victim is understanding of the circumstance and the effect upon their long-term societal productivity is minimal. While nuanced, such a situation is not incredibly unbelievable and most people can tell annecdotal stories with a similar storyline. Under pure utility, there is little incentive to punish the culprit, who is still a productive member of society who will not repeat the offense and the victim will recover the large majority of their social faculties in the fullness of time. But, the culprit is a rapist, who's act is still heinous, and there is a gut instinct that justice is not served unless they are somehow punished for their indiscretions against the social order.

And thus we run into a recurring theme in ethics, I call it the Condemnation Problem, and others have chosen myriad names for the same situation. Under perfectly rational morallities, there is an issue that it is very hard to actively condemn activities with the sort of blanket condemnation that seems logically necessary. If the offender will not repeat, and punishment will serve no deterent effect there seems to be no incentive to punish murderers, rapists, thieves, etc. Yet, I doubt most would argue that a crime of passion murderer should recieve no sentencing, or an aquantance rapist no punishment.

However, this is quite a nitpick which I don't think truely applies in haard's case, as I prefaced, because although he seems to have some intuition that such a problem does exist in mechanistic style morallity, he does not seem to attempt to actively combat it and instead relies on the comically outdated notion of moral intuition.

[Darth Wong]

There's something a little bit subtler than simply measuring proven ability to pay here: this is, in some sense, a measure of how hard has someone worked to gain what he earns.

In short, it is an attempt to create a tax system based on some moral precept of how deserving an individual is, rather than a straightforward attempt to tax based on ability to pay. That is going to be the mother of all quagmires. Why don't we also adjust taxes by profession, since certain professions have a work:reward ratio that is totally different from others?

It seems to me that if a six-foot man and a five-foot man, otherwise identical, both make $180000 per year, the six-foot man has not had to work as hard to attain his job because there are social advantages attached to being of above-average height. So if we wish to move the situation closer to a meritocracy, it makes sense to tax the six-foot man and subsidize the five-foot man so that their incomes reflect the effort they have to put into earning money. (Note that this would augment an already-progressive system; the tax code would already target those who have ability to pay.)

What about family connections, which have an enormous impact on wealth? How about physical attractiveness, which is also correlated to success? Should we tax beauty? How about race? Should we have a special "white person" tax?

To steal the language of the Iowa court decision on gay marriage, this idea suffers from being under-inclusive. It uses a general logic which casts an incredibly wide net (ie- "those who have unfair advantages should be penalized through taxes for the sake of fairness") and then targets only a tiny subset of that group (those whose special advantage is height, as opposed to myriad other possibilities).

[Alferd Packer]

What about family connections, which have an enormous impact on wealth? How about physical attractiveness, which is also correlated to success? Should we tax beauty? How about race? Should we have a special "white person" tax?

To steal the language of the Iowa court decision on gay marriage, this idea suffers from being under-inclusive. It uses a general logic which casts an incredibly wide net (ie- "those who have unfair advantages should be penalized through taxes for the sake of fairness") and then targets only a tiny subset of that group (those whose special advantage is height, as opposed to myriad other possibilities).

And how does a tax assessor objectively and fairly determine which of the myriad criteria are applicable? Let's say your six-footer has an IQ of 150, also happens to come from a rich, well-connected and respected family, and he also happens to be super handsome. Do you just tax everything? Or just that which can be measured (in this case, the IQ, the height, and wealth of his family)? It'd be pretty tough to tax beauty objectively, after all.

And if you really want a headache, how about attempting tax those in the modeling industry, where physical beauty essentially determines your income(or rather, plays a large role in determining it)? I can see headaches emerging from numerous other industries, especially those in the arts. Do you tax Buckethead more because he can shred faster than, say, Tom Morello? Or do you attempt to quantify the better guitarist some other way?