  |
[*pg 1145]
TAXING ENDOWMENT
INTRODUCTION
I. UTILITARIANISM AND ENDOWMENT TAXATION
II. LIBERAL EGALITARIANISM AND ENDOWMENT TAXATION
A. The Prima Facie Case for Endowment Taxation
B. The Talent Slavery Objection
C. Refining the Definition of Endowment
D. Endowment Taxation as "Distinctly Illiberal": The Hasen Objection
III. SLOUCHING TOWARD ENDOWMENT TAXATION
CONCLUSION
FOOTNOTES
 |
INTRODUCTION
Writing in 1888, Francis A. Walker, the first president of the American Economic Association, claimed that the ideal tax base was neither wealth, nor income, nor consumption, but rather "faculty, or native or acquired power of production."1 According to Walker, only this tax base -- what people could earn, rather than the often lesser amounts they actually earn -- could achieve an equitable distribution of the tax burden:
We must, I think, conclude that . . . to tax revenue instead of faculty is to put a premium upon self-indulgence in the form of indolence, the waste of opportunities, the abuse of natural powers; and that a faculty tax constitutes the only theoretically just form of taxation, men being required to serve the state in the degree in which they have ability to serve themselves.2
Although he conceded that practical difficulties made adoption of a pristine faculty tax impossible, he urged that it should nevertheless be "held in view, as furnishing the line from which to measure all departures from the equities of contribution, as one or another form of taxation . . . comes to be adopted for meeting the wants of government."3
Writing almost a century later, the economist David Bradford reached a similar conclusion.4 According to Bradford, tax burdens should be sensitive to differences in people's opportunities, but not sensitive to the different choices they might make. Ideally, each person would be subject to a one-time tax, at the beginning of his working years, on the present value of "his initial endowment of financial wealth and of future earning power."5 A tax imposed annually rather than once in a lifetime could approximate the effects of the single tax, if (1) financial wealth was subject to a consumption tax rather than an income tax,6 and (2) people were taxed on each year's potential labor earnings, rather than on their actual labor earnings.7 For Bradford, the practical policy significance of the endowment tax ideal related to the first point -- that a consumption tax was superior to an income tax because it more closely approached the endowment ideal.8 As a theoretical matter, however, he also insisted on the desirability of taxing potential earnings. In support of his claim, he posited a professional football quarterback and an individual without any high wage skills. Ideally the quarterback should be taxed more heavily than the nonathlete, because of his better opportunities, even if the nonathlete actually earns more money:
It might be that the nonathlete works very hard at a menial job and manages to earn more than the athlete, who chooses to work only on Saturday [sic] afternoons. If the athlete has the option of doing the same thing as the nonathlete but in addition has the opportunity to earn a high reward as a quarterback, the fact that he chooses to use the better opportunity in part to enjoy more leisure does not [*pg 1147] alter the conclusion that his opportunities are better or the theoretical case for assigning him the larger burden.9
Generalizing the point, Bradford explained, "What one might like to have is a tax based on an individual's potential earnings: the individual's tax would then be unrelated to his choice of what to do."10 Only the unfortunate inability of the tax administrator to observe potential earnings dissuaded Bradford from actually advocating such a tax.11
Despite the longstanding and understandable lack of congressional interest in taxing unexercised earning capacity, the idea of endowment taxation -- in the sense of a tax based on potential rather than actual earnings -- has attracted increasing attention in recent years from economists, philosophers, and academic tax lawyers. The dialogue among academic tax lawyers is especially lively at the moment, with important recent contributions by Daniel Shaviro,12 Kirk Stark,13 and David Hasen.14 With endowment taxation being addressed from several disciplinary frameworks, and with most contributions focusing on only one or a few aspects of the topic,15 it is not easy for a reader to grasp all the scholarly threads relevant to endowment tax analysis. The problem is exacerbated by the fact that some recent articles with important implications for endowment taxation have not been identified by their authors as contributions to the endowment tax literature,16 and so have gone unnoticed by those who are explicitly pondering endowment taxation.
The aim of this review Essay is to provide a reasonably comprehensive critical guide to the state of the endowment tax discussion. Because the utilitarian arguments for and against endowment taxation are quite different from the liberal egalitarian arguments, this Essay considers the merits of endowment taxation from both perspectives. Part I explains that an endowment tax base has no inherent normative significance for a utilitarian -- that is, endowment is not necessarily the very thing a utilitarian would want to tax in an ideal world. On the other hand, if it were practical to measure endowment, a utilitarian might support an endowment tax out of a belief that it would produce a higher sum of individuals' utilities than any other tax system (because it could accomplish utility-enhancing redistribution with minimal inefficiency costs). Finally, however, the strange distributional effects of a utility-maximizing endowment tax -- the fact that after-tax utility is inversely related to ability in such a system -- may cause some to reconsider their philosophical commitment to utilitarianism.
Part II considers the liberal egalitarian responses to endowment taxation. It is considerably longer than Part I, because the liberal egalitarian analysis of endowment taxation is (frankly) more interesting than the utilitarian analysis, and because the scholarly development of the liberal egalitarian concerns has been more extensive than the development of the utilitarian concerns. In general, liberal egalitarians would find an endowment tax attractive for reasons congenial to Walker and Bradford -- because of its potential for redistributing on the basis of differences in brute luck without also redistributing on the basis of differences in choices -- but for the risk of talent slavery posed by endowment taxation. The talent slavery concern is that a person who could earn a very high wage, but only by working full time at a job he despises, may be forced into that hated employment by the need to pay the tax. After evaluating the merits of the talent slavery objection, this Essay considers various ways an endowment tax might be circumscribed or the definition of endowment might be revised, so that the threat of talent slavery would be reduced or eliminated. In the process, this Essay examines ways in which a narrow focus on potential earnings may disregard important differences in the true endowments (or net brute luck) of different persons. Finally, Part II describes and evaluates the claim that the basic premise of endowment taxation is "distinctly illiberal," entirely apart from concerns about talent slavery.
Part III turns to some practical implications of endowment tax
theory. Endowment taxation -- perhaps with endowment not defined solely
on the basis of potential earnings -- would have much to commend it, if
the problems of measurement and talent slavery could be overcome. The
question thus arises whether some modest steps could and should be
taken in the direction of endowment taxation -- slouching toward an
endowment tax -- if those steps would not implicate either of the
problems associated with a full-fledged endowment tax. This Essay
describes and evaluates several proposals for partial implementation of
endowment taxation. It also suggests that several aspects of current
law -- both within and outside the Internal Revenue Code -- can
plausibly be viewed as instances of limited endowment taxation. This
Essay concludes that, although Congress will not and should not enact a
full-fledged endowment tax, endowment tax principles can usefully
inform tax policy analyses, especially if the concept of endowment is
broadened from the usual narrow focus on earnings potential.
| |
I. UTILITARIANISM AND ENDOWMENT TAXATION
A classic utilitarian's goal in designing a tax-and-transfer system is to maximize the sum of individuals' utilities. If the base of the tax is to be income, the crucial question is how best to balance the utility gains from redistribution of income from higher-income to lower-income persons (on the assumption that income has declining marginal utility) against the utility losses from the substitution effects of the tax.17 Substitution effects result when taxpayers change their behavior to avoid a tax, substituting untaxed (or less heavily taxed) behavior for the taxed behavior. Suppose, for example, that a taxpayer's wage rate is $100 per hour, that his marginal income tax rate is 30 percent, that he values leisure time (at the margin) at $80 per hour, and that he is deciding whether to spend the next hour working or at leisure. In the absence of tax, he would choose to work, because he values his leisure at less than his $100 wage rate. In the [*pg 1150] presence of the income tax, however, he will choose $80 of untaxed leisure over $70 of after-tax income. He views the $30 tax that would be imposed on his work as a positive externality -- a benefit to society but not to him, which he disregards in making his labor supply decision. The result of the substitution effect in this example is a deadweight loss of $20 -- the difference between the $100 value of his work and the $80 value of his leisure.
If it were possible to replace the income tax with a tax on potential earnings -- that is, an endowment tax -- the utility gains from redistribution could be achieved without the deadweight loss associated with an income tax. The taxpayer would have no tax incentive to substitute leisure for labor because his tax liability would be unaffected by the amount he actually earns.18 A head tax -- a tax imposed in an equal dollar amount on each person -- would also avoid the deadweight loss of the substitution effect, but it could not achieve the utility gains from redistribution possible with an endowment tax. At first glance, then, it seems that an endowment tax would be the choice of a utilitarian, if the tax administrator is able to observe potential wage rates. Upon closer examination, however, the utilitarian case for an endowment tax becomes shaky, even on the heroic assumption that potential earnings are observable.
First, the conventional wisdom that head taxes and endowment taxes have no substitution effects may not be quite right. In a recent article, Terrence Chorvat claims that the fact that head taxes and endowment taxes must be paid in cash means the taxes are less than perfectly efficient: "While lump-sum taxes are commonly argued to be perfectly non-distortive, we can see that such taxes can distort relative prices between monetary and non-monetary goods, due to [their] requirement that [they] be paid in money. Therefore, even a lump-sum tax has deadweight loss."19 To make the point more concrete, suppose a particular person with a high wage rate would choose, in the absence of any tax, to spend all his time as a beachcomber -- because he hates the work that would produce the high wage, because he loves beachcombing, or both. If he is subject to a high endowment tax based on his potential earnings, and if he can earn the high wage only by working full time (i.e., the labor market is very "lumpy" in his case), then he has two choices: he can work full [*pg 1151] time at the hated job and pay the tax in cash, or he can decline to work and be imprisoned. The latter choice can be viewed as an in-kind tax, equal to 100 percent of his time, if he insists on not fully exploiting his earning potential. If, faced with these choices, he understandably chooses to work at the hated job, the change in behavior can be understood as a substitution effect of the in-kind tax, with resultant deadweight loss. As in standard accounts of substitution effects, he has changed his behavior to avoid the heavy tax imposed on refusing to work.20 Although this means that an endowment tax does not achieve efficiency nirvana, the substitution effects of an endowment tax are certainly less significant than those of an income tax. Thus endowment taxation should continue to be attractive to utilitarians as far as deadweight loss is concerned.
The next question for a utilitarian pondering an endowment tax is whether transfers of cash from persons with higher potential earnings to those with lower potential earnings are necessarily utility enhancing (setting aside questions of deadweight loss). In the context of an income tax, utilitarians are generally willing to make the assumptions about interpersonal comparisons of utility necessary to conclude that taking money from a person with more income and transferring it to a person with less income benefits the lower-income person more than it hurts the higher-income person. Is it also reasonable to conclude that redistribution of cash based on earnings potential improves overall utility, because a person with higher potential earnings has a lower marginal utility of cash than does a person with lower potential earnings? That conclusion would be reasonable, if having a higher wage rate is associated with having higher earned income.21 On the other hand, if people differ in respects other than their wage rates -- in particular if they differ with respect to their degree of work aversion (disutility of labor) or materialism (utility derived from purchased consumption), or both -- then taking money from those with higher wage rates to finance transfers to those with lower wage rates does not necessarily increase the sum of [*pg 1152] individuals' utilities.22 Imposing a heavy endowment tax on a person who can earn a high wage only by working at a job he hates may cause a massive utility loss for that person, especially if the labor market is lumpy (so that he must work full time at the high-wage job or not at all).23 This is the talent slavery scenario, which is at the heart of most liberal egalitarians' objections to endowment taxation.24 The common liberal egalitarian response is to reject an endowment tax if it would result in talent slavery of even one person, regardless of how attractive the tax might be in all other respects. A utilitarian, by contrast, might be willing to accept a system producing talent slavery for a few persons, if their utility losses were outweighed by the utility gains the system produced for others.25
In sum, the utilitarian view of endowment taxation is rather complicated. An endowment tax would be the utilitarian ideal if people were identical in all respects other than their earnings capacities. On the other hand, as Daniel Shaviro remarks, utilitarians would find "no clear justification for imposing a straightforward endowment tax, even if we could measure endowment. . . . once we allow for taste differences and lumpy labor markets."26 Finally, if endowments were measurable but other differences (in work aversion and materialism) were not, a utilitarian might conclude that the net effects of an endowment tax -- redistribution of cash mostly in utility-enhancing directions, with much less deadweight loss than an income tax -- were superior to the results achievable under any other tax system.27
It may be, however, that when push comes to shove there are few
utilitarians. Several decades ago, James Mirrlees considered the design
of the utility-maximizing endowment tax-and-transfer system, assuming
the tax administrator could observe wage rates and that [*pg 1153] people differed in their
wage rates but were identical in all other respects. He demonstrated
that the optimal system would leave each member of society with the
same number of after-tax-and-transfer dollars (and thus with the same
amount of purchased consumption), but that higher-wage persons would
work more hours than others.28
The results can be explained intuitively. First, redistribution of cash
(and so of consumption opportunities) should continue until utility
gains from further redistribution are not possible, and that point will
be reached only when marginal utilities of cash have been equalized
across the population by equalizing after-tax-and-transfer cash
receipts. Second, for any given amount of foregone leisure, the work of
the more able produces more value -- and hence more consumption
opportunities translatable into utility -- than does the work of the
less able; it follows that utility is promoted if the more able work
more hours than the less able. Under this utility-maximizing endowment
tax, persons of higher ability are worse off than persons of lower
ability. Each person's total utility is the sum of his utility from
consumption and his utility from leisure. Higher wage persons have the
same consumption as lower wage persons while having less leisure, and
so have lower total utility. To a thoroughgoing utilitarian there is
nothing wrong with that result, since it produces a higher sum of
individuals' utilities than any other possible system. Few people may
remain utilitarians, however, if that requires accepting the conclusion
that the ideal tax-and-transfer system would make the most talented
members of society the least well-off.
| |
II. LIBERAL EGALITARIANISM AND ENDOWMENT TAXATION
This Part considers the prima facie liberal egalitarian case for endowment taxation, the "talent slavery" objection to endowment taxation, and the possibility of designing an endowment tax so as to avoid talent slavery. Talent slavery might be avoided by limiting the redistributive ambition of the tax. It might also, and perhaps more interestingly, be avoided by the use of more nuanced concepts of endowment, focused on welfare rather than potential earnings, and taking into account the disutility of work. Finally, this Part considers a claim that endowment taxation is fundamentally "illiberal," even when talent slavery is not an issue.
A. The Prima Facie Case for Endowment Taxation
Although there is great diversity in liberal egalitarian theories of distributive justice, the moral distinction between people's "brute luck" and the results of people's choices is central to most liberal egalitarians.29 Brute luck -- including the distribution of abilities in the natural lottery -- is morally arbitrary. People do not deserve their brute luck, either good or bad, and so the government should act to ameliorate the disadvantages of those who have suffered from bad brute luck. Once there has been appropriate redistribution with respect to brute luck, however, the government should be neutral with respect to the free choices made by different individuals pursuing different visions of the good life. Thus the government should not redistribute among people who, faced with the same opportunities, make different choices and experience different outcomes.
Although people do, of course, differ in their endowments of financial wealth, the most significant differences in brute luck are differences in human endowments -- that is, differences in abilities based on the morally arbitrary distribution of genetic inheritances and the benefits of parental care and attention. Also morally arbitrary are the differing rewards the economy happens to confer on people possessing different skills. If some people are endowed with great earnings potential and others are not, the egalitarian urge to lessen or eliminate the effect of brute luck would suggest redistribution from those with above-average ability (i.e., earnings potential) to those with below-average ability; and since ability cannot be redistributed in kind,30 the redistribution might take the form of cash transfers to [*pg 1155] the less able financed by taxes (payable, of course, in cash) on the endowments of the more able. If we want to redistribute with respect to endowment, then the ideal base for the tax to finance the redistribution is necessarily endowment.31
In theory, such a tax might be designed as either a one-time tax -- imposed perhaps at birth or upon reaching the age of 18 or 21 -- on the present value of one's lifetime earnings potential, or as an annual tax on the maximum wages one could have earned in the current year. Under certain stylized assumptions (most notably, a single tax rate and the absence of any uncertainty about one's lifetime earnings potential), the two approaches are economically equivalent. To avoid piling the impracticalities any higher than necessary, the remainder of this discussion will assume the endowment tax under consideration is of the annual variety. Imagine, then, two people who have the same high earnings potential for the current year, but who have made different choices -- one to work at the highest available wage for as many hours as humanly possible, the other to earn much less by working at a more pleasant (or less unpleasant) lower wage job, by working fewer hours, or both. A liberal egalitarian might conclude that they should bear the same tax burden (to finance redistribution to those of lower ability), despite their very different cash incomes, since they had the same opportunities. If taxation is to be sensitive to the distribution of brute luck, but insensitive to differences resulting from different people making different choices, then the tax system should not treat these two equally able persons differently merely because one chose to work at a higher wage rate (at a less pleasant job) or to work more hours.
In short, the prima facie liberal egalitarian case for a tax on potential earnings, rather than on actual wages, is strong. Nevertheless, liberal egalitarians generally end by rejecting endowment taxation.32 The reasons for this rejection are described and evaluated below.
B. The Talent Slavery Objection
1. The Objection.The most prominent liberal egalitarian objection to endowment taxation is that it would inevitably result in "slavery of the talented."33 Imagine a person with one skill very highly valued by the economy who could earn $10 million in one year by fully exploiting that skill. Suppose also that the market for that skill is as lumpy as possible -- i.e., he must either exploit that skill fully or not at all -- and that his next highest potential wage (doing a very different sort of work) is $100,000. If he is taxed on his $10 million earnings potential, at a rate of, say, 20 percent, he will owe $2 million tax. The only way he can generate the money with which to pay the tax is to spend the entire year working at his maximum earnings potential. That is fine if he is happy (or even merely not unhappy) with that work, but if he happens to hate the work the result is the slavery of the talented. On pain of being consigned to debtors' prison or exile or perhaps the wilderness, he must spend the entire year doing work he detests. As much as liberal egalitarians value equality of opportunity, they also value autonomy, and many of them have concluded that the interference with the autonomy of the talented produced by endowment taxation is simply unacceptable.34 With the recent and important exception of Daniel Markovits (whose views are discussed below),35 the liberal egalitarians who have objected to endowment taxation on grounds of talent slavery have done so quite summarily; the objection has seemed so apparent and so conclusive as to require little discussion.
Recently, however, several commentators have scrutinized the talent slavery objection and found it wanting. The most detailed and [*pg 1157] most unequivocal rejection of the objection has come from Kirk Stark.36 Professor Stark begins with Robert Nozick's well-known libertarian objection to the taxation of wages actually earned, as "forcing the [taxed] person to work n hours for another's purpose," and as therefore "on a par with forced labor."37 Stark notes that Professor Nozick's involuntary servitude objection to income taxation bears a marked resemblance to the liberal egalitarians' talent slavery objection to endowment taxation. In fact, he claims that "there is no difference in kind or in degree between the interference with liberty occasioned by the two types of taxes," and that liberal egalitarians must either accept Nozick's critique of income taxation or "embrace endowment taxes as a moral ideal."38
Professor Stark makes his argument with the aid of a hypothetical, based on a similar example offered by Professors Murphy and Nagel, involving a sculptor subject to an endowment tax and a corporate lawyer subject to an income tax.39 Each could potentially earn $500,000 annually by practicing corporate law. The sculptor would prefer to spend all his time sculpting pieces he will not sell. The corporate lawyer would prefer to spend all his time practicing law, in order to earn enough to buy a $500,000 painting he covets. If the sculptor is subject to a 20 percent endowment tax (on $500,000 of potential earnings), and the corporate lawyer is subject to a 20 percent income tax (on $500,000 of actual earnings), they will [*pg 1158] be -- according to Stark -- equally burdened.40 True, the sculptor will have to spend 20 percent of his time practicing corporate law to generate the $100,000 necessary to satisfy his tax obligation, but the corporate lawyer will be similarly burdened by being able to buy only a $400,000 painting. Each loses 20 percent of his preferred consumption bundle -- the sculptor by losing 20 percent of his desired time to sculpt, and the lawyer by losing 20 percent of the desired value of his art acquisition. Stark concludes that there is no difference in kind or in degree between the impact of the endowment tax on the liberty of the sculptor and the impact of the income tax on the liberty of the lawyer.41 To Stark, if it is acceptable for the income tax to commandeer the value of 20 percent of the lawyer's time, it should also be acceptable for an endowment tax to commandeer the value of 20 percent of the sculptor's time.
Professor Stark mentions -- but only in a footnote -- that his analysis is based on the assumption that the sculptor could earn $100,000 by spending 20 percent of his time practicing law. He concedes, "If labor markets are 'lumpy' and a particular wage rate can only be achieved by working full-time, then the analysis may be different."42 If labor markets were completely nonlumpy, as Stark supposes in his example, his conclusion would have considerable appeal. It is not easy to view as the moral equivalent of slavery an endowment tax that leaves the sculptor free to spend 80 percent of his time sculpting. In such a world, the talent slavery objection would lose most of its force, although a narrow "conscientious objector" exception might still be needed for taxpayers who honestly considered the best-paying use of their time to be not merely unpleasant, but immoral (a view some sculptors might hold of corporate law practice). One might also object that even in a world of nonlumpy labor supply Stark's endowment tax on the sculptor seems acceptable only because of its relatively low rate; a tax at the rate of 50 percent, requiring the sculptor to spend half his time practicing law, is instinctively much more troubling. Stark could reply, however, that his assumed rate is reasonable, because a 20 percent endowment tax might raise as much revenue as the current income tax (or even more), given its considerably larger base.
In their original version of the sculptor-and-lawyer hypothetical,
Professors Murphy and Nagel implicitly make the opposite assumption
about labor supply lumpiness: "[A] trained corporate lawyer taxed
according to $500,000 in potential annual earnings will find that the
time for sculpture is reduced almost to zero."43
They are obviously assuming that the only way he can earn $100,000 is
by working full time and earning $500,000; if they supposed he could
satisfy his tax obligation with 20 percent of his time they would not
claim he would have "almost zero" time left for sculpting. Thus, the
disagreement about the talent slavery implications of an endowment tax
appears to be based not on any difference of principle, but on
different assumptions about the extent to which highly talented persons
can exploit their talents in the market on a part-time basis. Certainly
Professor Stark could point to a few situations in which labor markets
approach the perfect nonlumpiness he assumes: if Tom Cruise wants to
devote most of his time to sculpting he could satisfy his tax
obligation by spending ten weeks each year making one movie. It seems,
though, that similar options are not available for highly talented
people in most other professions. If the extreme lumpiness assumption
of Murphy and Nagel is much closer to the real world than the no
lumpiness assumption of Stark, then Murphy and Nagel have the better of
the argument. If the endowment tax forces the would-be sculptor to
spend all of his time practicing corporate law, that
interference with his liberty is of a higher order of magnitude than
the income tax forcing the corporate lawyer to make do with a $400,000
painting.
2. Designing an Endowment Tax that Avoids Talent Slavery.
As
this discussion of the disagreement between Stark and Murphy and Nagel
suggests, the liberal egalitarian talent slavery objection is not a
fundamental objection to endowment taxation. If the labor market were
not lumpy, an endowment tax would not imply talent slavery. Even
assuming lumpiness in the labor market, liberal egalitarians might
accept -- or even embrace -- an endowment tax modified as necessary to
avoid talent slavery. One such modification has been suggested -- as an
aside -- by Louis Kaplow. Professor Kaplow notes that "an ability tax
with an upper limit of 90 percent of wages or income actually earned
would almost completely achieve the ability [*pg 1160] tax result without
literally violating the forced-labor constraint."44 He is certainly right that this hybrid
endowment-income tax would not raise any talent slavery concerns
greater than those raised by a simple income tax with a 90 percent top
rate, and one supposes that most liberal egalitarians would not view
high rates under an income tax as constituting talent slavery. Kaplow
is on weaker ground, however, when he claims that the hybrid tax would
produce results closely resembling a pure endowment tax. If taxpayers
like the one described earlier (with $10 million of potential earnings
in one field but only $100,000 actual earnings in a different field)
really exist, a $90,000 tax on $100,000 wages actually earned would not
"almost completely achieve the ability tax result" of a $2 million tax
on $10 million of potential earnings. In some cases, then, the hybrid
tax is likely to be much more of an income tax, and much less of an
endowment tax, than Kaplow supposes.
The most thorough consideration of how an endowment tax might be designed to avoid talent slavery comes from Daniel Markovits.45 His analysis begins by supposing that people know the distribution of wage rates in their society, but that no one knows his own particular wage rate.46 (People know their own skills, but they do not know the wage rate those skills will command in the market.) In this imagined original position, people might plausibly agree to a tax-and-transfer system as a form of insurance against having been unlucky in the talent lottery. If a low wage rate is the risk to be insured against, the obvious policy tool would be an endowment tax -- a tax on the highly talented to finance cash transfers to the less talented.47 The question then becomes how much insurance people would agree to in the imagined original position. Professor Markovits argues convincingly that people would not agree to the highest possible level of insurance, under which a person with a below-average wage rate would receive a cash transfer equal to the excess of the average (mean) annual wage for full-time work over his own [*pg 1161] potential earnings.48 The transfers required under that system could be made only if each person with above-average earnings potential worked full time at his highest possible wage rate, with all wages in excess of the average being taxed at the rate of 100 percent for transfer to those with below-average skills. But this would result in talent slavery not just for a highly talented few, but for everyone with an above-average maximum wage rate: "[S]he must always work flat out and only at that job which, given her talents, pays most. She would be forced to work at this job, and this job only, even if she hated the work involved . . . ."49 No one in the original position would agree to the level of insurance needed to eliminate the effect of brute luck on the distribution of wage rates, when such insurance necessarily imposes a high risk of talent slavery.50
That much is clear, but a harder question remains: At what lower level of wage-rate insurance, and associated endowment taxation, would the risk of talent slavery be removed so that people in the original position would agree to the insurance program? Professor Markovits notes that there is no general answer to that question, but he also explains that the "maximum talent-slavery-safe level of insurance" will fall well below the average wage rate in a society in which the distribution of earnings potential has a pronounced tail at the high end:
[A]s the talent distribution becomes more dispersed, and particularly as the median and the mode of the distribution fall further below the mean (as the distribution develops a tail at the high end), insuring even at levels below the mean will require people to pay a greater proportion of their maximum wage as premiums, premiums a smaller range of jobs will enable them to support. For both reasons, the maximum talent-slavery-safe level of insurance will in such cases fall further below the mean.51
Performing some back-of-the-envelope calculations based on the 1999 wage distribution in the United States (featuring an impressive tail at the high end), Markovits plausibly concludes that people aware of that wage distribution but unaware of their own wage rate "would [*pg 1162] insure only up to talent levels that fall significantly -- more than 25 percent -- below the mean."52
Endowment taxation and talent slavery are not the central themes of Professor Markovits's article (and perhaps for that reason his article has previously gone unnoticed by endowment tax commentators). Rather, he uses his discussion of endowment taxation to support his major thesis, which is that the two goals of many liberal egalitarians -- elimination of inequalities due to brute luck, while respecting differences resulting from different people having made different choices -- are logically incompatible. The significance of endowment taxation to Markovits's thesis is that endowment taxation can eliminate brute luck only under a full insurance scheme that people would reject if given the choice. Thus one must either abandon the goal of eliminating inequality of brute luck in order to respect choice, or abandon the goal of respecting choice in order to eliminate inequality of brute luck.53 Quite apart from Markovits's thesis, however, his analysis has important implications for the ongoing debate over endowment taxation and talent slavery. It provides an attractive account of endowment taxation as wage rate insurance, and situates the talent slavery concern as a constraint in the design of the endowment tax system, rather than as an argument against endowment taxation in any form.
C. Refining the Definition of Endowment
The talent slavery objection is aimed at an endowment tax which
defines endowment as one's maximum wage rate multiplied by the maximum
number of hours one could work, without regard to the utility or
disutility associated with that level of work effort. There are two
possible refinements of the concept of endowment, however, each of
which arguably does a better job of measuring real differences in brute
luck than the simple notion of potential earnings. Each of these
refinements can be defended as being preferable to potential earnings
as an ideal tax base, and each has the additional benefit of
ameliorating the talent slavery problem.
[*pg 1163]
1. The Musgrave Proposal.
The first refinement has been
offered by Richard Musgrave.54
According to Professor Musgrave, an ideal tax-and-transfer system would
redistribute not with respect to potential income, but with
respect to potential welfare.55
He illustrates his point with an example involving two persons with
different wage rates, who are identical in all other respects. The
welfare of each person is a function of his hours of leisure and his
consumption of goods. The two persons have identical utility functions,
so each will enjoy the same level of welfare as the other from any
given combination of hours of leisure and dollars of consumption. In
the absence of a tax-and-transfer system, the person with the higher
wage rate (H) will have a higher level of welfare, as he can
select a combination of leisure and consumption that will place him on
an indifference curve higher than any indifference curve reachable by
the person with the lower wage rate (L).56
Under the standard view of endowment taxation, complete elimination of
the brute luck difference in their wage rates would require a transfer
from H to L sufficient to give the two persons equal
net-of-transfer potential consumption.57
For example, if H's hourly wage is $60, L's is $10, and
there are 4,000 potential work hours in a year, the required transfer
would be $100,000.58 Although
this will equalize the potential consumptions of H and L,
it will leave L with a higher level of welfare than H.
Suppose, for example, that faced with the $100,000 tax-and-transfer
program, H decided to work 3,333 hours (generating $200,000 of
pretax income) and L decided not to work at all. Each would
have $100,000 available for consumption after the transfer of $100,000
from H to L, but L would have 4,000 hours of
leisure while H had only 667 hours.59
Quite apart from the question of whether H is a victim of talent slavery, a tax-and-transfer program which systematically leaves more talented persons with lower welfare than less talented persons seems inconsonant with liberal egalitarian conceptions of justice. Accordingly, Professor Musgrave's ideal system would require a smaller transfer from H to L, designed not to equalize their potential incomes, but instead to equalize their welfares by putting them at different points on the same indifference curve. Musgrave describes this approach as "a tax on income and leisure, imposed at the proper rate by an omniscient referee."60 Suppose, for example, that both H and L (who by hypothesis have identical utility functions) would be indifferent between having $140,000 of consumption and 1,000 hours of leisure, or having $40,000 of consumption and 4,000 hours of leisure. If H responds to a $40,000 lump sum tax by working 3,000 hours and L responds to a $40,000 transfer by not working at all, then H will have a package of $140,000 consumption ($180,000 wages minus $40,000 tax) and 1,000 hours of leisure, and L will have a package of $40,000 consumption and 4,000 hours of leisure. Although H has more consumption and L has more leisure, the two are equally well-off, and the goal of eliminating welfare differences attributable to brute luck has been achieved. And it has been done by a transfer of only $40,000, compared with the $100,000 transfer that would have been required to equalize their potential incomes. As Musgrave explains, "The taxpayer's [H's] loss of leisure due to more work reduces his welfare just as the transfer recipient's [L's] gain of leisure raises it, thus narrowing the remaining need for income transfer."61
Although Professor Musgrave makes a strong argument that welfare is a more appropriate object of redistribution than potential income as a matter of first principles, he also claims that basing the tax on welfare has the attraction of neatly sidestepping the talent slavery problem raised by taxation of earnings potential:
While the high-ability person gives up both goods and leisure, this is merely the outcome of his free choice between them, made in response to the lump sum tax. . . . The prior concern for leisure, it [*pg 1165] appears, is not needed in the context of a just solution, although it is appropriate if relative positions are viewed in terms of goods only.62
To put Professor Musgrave's point in Professor Markovits's terms, persons in the original position would not view with abhorrence the possibility of turning out to be a high-ability person under Musgrave's welfare tax. Since the tax is designed to leave high-ability persons with as high a level of welfare as anyone else, the result cannot be reasonably viewed as talent slavery.
As presented by Professor Musgrave, his approach assumes (as in the
illustration) that people are identical in all respects except wage
rates. Differences in individuals' utility functions -- including
differences in the utility derived from any given dollar amount of
consumption and differences in the disutility associated with work --
presumably would be ignored in the construction of a Musgrave-inspired
tax system. (This would seem to be a necessity, if the system is to be
merely extremely difficult to administer, rather than absolutely
impossible.) The reliance on a single standardized utility function
means the Musgrave approach would not really equalize welfare.
If the disregarding of individualized utility information were combined
with lumpy labor supply, it is also possible that talent slavery
problems would remain under the Musgrave approach, despite Musgrave's
claim to the contrary. In the illustration above, for example, things
work out neatly only because H has the option of working 3,000
hours at $60 per hour. If the only way H can earn $60 per hour
is by working 4,000 hours at a job he hates, and if his alternative
wage rate with fewer hours worked is much lower, the problem of talent
slavery has not been eliminated. It remains true, however, that talent
slavery concerns would be much diminished in a Musgrave-inspired
tax-and-transfer system, compared with the problems under a tax based
simply on potential earnings.
2. Factoring in the Disutility of
Labor Effort.
The above discussion suggests that the definition of
endowment as potential earnings is deficient because it disregards the
significance of hours of leisure to overall welfare. It can also be
argued that the definition of endowment as potential earnings is
deficient because it disregards the utility or disutility people
experience from work. Consider a person who could earn a high wage, but
chooses instead to work at a [*pg 1166] different job for a much lower wage. Such
persons are objects of great solicitude for opponents of endowment
taxation.63 But why would a
person who could work at a higher wage decide to work at a lower wage
(assuming the time demands of the two jobs are the same, so there is no
question of more or fewer hours of leisure)? It must be because of
differences in the utility (negative or positive) associated with
supplying his labor in the two situations. One possibility is that he
hates the high wage job, while he neither likes nor dislikes the low
wage job. If the disutility associated with the high wage job causes
him to take the low wage job, then his welfare if he took the high wage
job would be no higher (and would probably be lower) than the welfare
he experiences with the low wage job. If endowment is defined not
narrowly as potential earnings, but more broadly as potential earnings
plus or minus the utility or disutility associated with the work
necessary to generate those earnings, then this person's true endowment
is properly measured by his actual low wages. His actual low wages plus
zero labor utility from the low wage job must equal or exceed his
potential high wages minus the labor disutility from the high wage job.
Under this concept of net endowment, an income tax would produce for
this person the same result as a tax on his net endowment, whereas a
tax on potential earnings would tax him on more than his net endowment.
If all persons who turned their backs on high wage jobs did so because they hated those jobs, and if a tax system were based on this concept of net endowment, the problem of talent slavery would be solved, as net endowment could never be higher than wages actually earned. A person who would not do high wage work because he hated it would be treated just like someone who could not do high wage work because he lacked the skill. As a solution to the talent slavery problem this amounts to killing the endowment tax in order to save it, since the best way of implementing this vision of an endowment tax would be with a traditional income tax.
Although this position -- that an income tax is actually a good approximation of an endowment tax, if endowment is defined to take into account labor disutility -- has considerable appeal, it is subject to two objections. First, a distaste for high wage work can be viewed as a kind of expensive taste (because the alternative uses of one's time are [*pg 1167] expensive in terms of foregone income), and there is a prominent liberal egalitarian argument, associated primarily with Ronald Dworkin, that distributive justice should make no allowance for expensive tastes. Working from a model of tax-and-transfer systems as insurance against bad brute luck, Professor Dworkin concludes that insurance against expensive tastes is inappropriate, because one's tastes are an integral part of one's personhood.64 As there is no reason to suppose people would want to insure against being other than who they are, there is no reason to suppose people would want to insure against having different tastes than they actually have. This view has been convincingly refuted, however, by Daniel Markovits. He suggests that a person in the original position, who knows his tastes but not whether they are expensive, might well want "insurance not against having the taste, but rather against turning out to live in a world in which it is expensive."65 By the same token, a person who hates a particular job might want to insure against the possibility that the hated job pays a high wage and thus would generate a huge tax bill under a tax system based on potential earnings. The simplest way of providing such insurance might be simply to abandon the taxation of potential earnings in favor of taxing only actual earnings.
The second objection to the idea of an income tax as the best embodiment of a net endowment tax is less easily answered. In the preceding illustration, the income tax works well as a net endowment tax only because the example assumes the person turns down the high wage job because he hates it, rather than because he loves the low wage job. Suppose instead that the person neither likes nor dislikes the high wage job, loves the low wage job, and decides to take the low wage job. If so, taxing him on the potential earnings from the high wage job may be appropriate, since his utility from his current situation is as high as the utility associated with his potential earnings. In the previous example, his net endowment was no greater than his actual wages from the low wage job, but in this example his net endowment is at least as great as the potential wages from the high wage job. Even if the effect of a tax on potential wages were to force him to take the high wage job, it would be hard to view that as talent slavery if he does not dislike the high wage job.
Although this second objection has force, it may not be powerful enough to rehabilitate taxation of potential wages. If one's ideal tax base would be net endowment (i.e., potential wages plus labor utility or minus labor disutility), but for the inability to measure labor utility or disutility, then the available second best choices are an income tax or a tax on potential earnings. The choice, then, would be based on (1) which tax would produce more errors, (2) the magnitudes of the errors produced by each tax, and (3) whether the type of errors produced under one tax is worse than the type of errors produced under the other. The income tax produces errors when a person chooses low wage work because he loves it, whereas the tax on potential earnings produces errors when a person avoids high wage work because he hates it. My suspicion is that hatred of high wage work is more common than love of low wage work, in which case the income tax would produce fewer errors. Even if empirical investigation proved that suspicion wrong, one might still favor the income tax on the basis that the errors produced by a tax on potential earnings result in talent slavery, whereas the more numerous errors produced by the income tax would have less dire consequences. In sum, if redistribution would ideally be done with respect to net endowment -- defined as potential wages adjusted for the utility or disutility of work -- it is difficult to conclude that a tax on potential wages is superior to an income tax in its approximation of the ideal.66
So far the discussion in this section has abstracted from Professor Musgrave's concern about differences in hours of leisure, by assuming that persons differ in their earnings and in their love or hate of their work, but not in the division of their time between work and leisure. A comprehensive concept of endowment, however, might include both a Musgravean consideration of leisure and a consideration of differences in the utility or disutility of work. Designing a tax system reflective of that concept of endowment would be a daunting task, but a point made by Louis Kaplow is worth noting in this regard. Professor Kaplow has demonstrated that if two persons with different wage rates nevertheless have the same earned income, because the person with the higher wage rate experiences greater labor disutility and so works fewer hours, then the two persons will also have the same total utility (assuming they are identical in all other respects).67 Compared with the lower wage person, the greater utility the higher wage person derives from more leisure is precisely offset by the greater disutility he suffers from working. The two persons are equally well-off under a comprehensive definition of endowment (reflecting their available packages of consumption, labor utility or disutility, and leisure), and a simple income tax on their identical incomes would appropriately subject them to equal tax burdens. It does not follow that an income tax always magically replicates the results that would be achieved by a tax based on a broad concept of endowment,68 but one could do worse than an income tax from this perspective.
D. Endowment Taxation as "Distinctly Illiberal": The Hasen Objection
The standard objection of liberal egalitarians to endowment taxation (in the usual sense of a tax on earnings potential) is that it will result in talent slavery. David Hasen, however, has recently lodged a more fundamental objection.69 According to Professor Hasen, endowment taxation should be unacceptable to liberal egalitarians, quite apart from any risk of talent slavery. As is common [*pg 1170] in liberal egalitarian accounts of distributive justice, Hasen asks what sort of social contract people might agree to from "a 'pre-governmental' bargaining situation."70 He believes that the social contract would reflect a strong concern for "autonomy, understood as the right to posit and realize one's own ends with minimal interference from others."71 People in the original position would agree to constraints on their autonomy only to the extent those constraints were needed in order to make possible the benefits of social cooperation. Under this view, "the point of taxation is to allocate the costs of producing benefits that membership in a system of social cooperation confers,"72 and the natural tax base is therefore the benefits of social cooperation. Because earned income is the result of social cooperation but potential earnings are not, actual income is an appropriate tax base but potential income is not:
[I]t is the specific nature of income -- that it is produced through cooperation -- that confers the right to tax it in the first place. It is the production of actual, not possible, income that political society facilitates and for which the participants must pay; the system does not . . . produce income-earning capacity.73
If interference with autonomy is justified only to the extent necessary to facilitate social cooperation, and if unexploited earnings capacity is not produced through social cooperation, it follows that "the value of autonomy largely rules out endowment taxation."74 This goes far beyond a concern about possible talent slavery: "[I]n order to make endowment taxation legitimate within a liberal framework, one would need to locate some additional constitutive principle that could support it, because autonomy as an organizing principle leaves essentially no room for it."75
Professor Hasen's claim is subject to three quibbles, and to one major objection. The first quibble, which Hasen notes himself, is that even people who choose to earn little or no income -- such as the oft-invoked beachcomber -- may benefit from social cooperation to the extent they are protected by the police and fire departments, and by national defense. Hasen concedes that endowment taxation "might be perfectly appropriate" as to those benefits.76 Second, Hasen does not consider the possibility that unused earnings capacity might be a measure of noncash benefits of cooperation. If a law professor could have earned much more as a corporate lawyer, but chooses a teaching career because of its intangible joys, those joys are benefits of social cooperation, and they could be reached through a tax on potential earnings. Third, Hasen does not acknowledge the autonomy case for an endowment tax rather than an income tax. The substitution effect of an income tax pushes taxpayers away from earning income and toward nontaxed uses of their time. If a proper respect for autonomy requires the government to be as nearly neutral as possible among different ends, conceptions of the good, and uses of time, it is not clear that an income tax is preferable to an endowment tax. True, the requirement that an endowment tax be paid in cash may push people into the workforce (in the extreme case resulting in talent slavery), but because one's endowment tax does not depend on how one spends one's time, the endowment tax avoids the favoritism for leisure (and for low-wage, high-pleasure employment) inherent in an income tax.
The major objection, however, is that Professor Hasen's argument
depends on a constricted view of what qualifies as liberal egalitarian
thought. His account of distributive justice and the social contract is
a plausible one, and he is right that endowment taxation would be ruled
out under that account. But could not one imagine a different version
of the social contract, under which people in the original position
would agree to endowment taxation? Hasen asks this question himself:
"Why can't or wouldn't the parties to the original bargaining position
agree to make some or all of their endowments the subject of their
agreement [and thus subject to tax]?"77
He answers, "[W]hile such an approach is possible, it is also
distinctly illiberal."78 It is
illiberal because by "compel[ling] more than [*pg 1172] a minimal amount of social
cooperation" it does not sufficiently respect autonomy.79 Despite Hasen's claim that his argument
applies with respect to "liberal political theories in general,"80 across a wide range of
liberal egalitarian thought, his conclusion follows only under a narrow
view of what constitutes liberal egalitarianism. Certainly a number of
other thinkers, liberal egalitarians by self-identification and by
common understanding -- including Ronald Dworkin, Daniel Markovits,
Liam Murphy, and Thomas Nagel -- have found endowment taxation
compatible with liberal egalitarianism, apart from talent slavery
concerns.81 The appeal is
based on the idea, pioneered by Professor Dworkin and developed by
Professor Markovits, that people in the original position might favor
endowment taxation as low-endowment insurance, at least if it could be
done without talent slavery.82
This follows from the value liberal egalitarians place on ameliorating
differences among individuals attributable to brute luck. Although that
value must be weighed against the interference with autonomy caused by
endowment taxation, it is not clear why anyone who thinks the insurance
positives of endowment taxation (limited as necessary to avoid talent
slavery) outweigh the autonomy negatives should be declared a heretic
and excommunicated from the liberal egalitarian church.83 Unless one is willing to deny the liberal
egalitarian credentials of the likes of Dworkin, Markovits, Murphy, and
Nagel, Hasen's argument is only a liberal egalitarian case
against endowment taxation, not the liberal egalitarian case.
| |
III. SLOUCHING TOWARD ENDOWMENT TAXATION
Despite David Hasen's claim to the contrary, a liberal egalitarian might reasonably conclude that redistribution from those with better luck in the natural lottery to those with worse luck is normatively attractive. Such a liberal egalitarian would support something along [*pg 1173] the lines of an endowment tax, although he would insist that it be designed to avoid talent slavery and he might believe that mere earnings capacity (not taking into account, for example, differences in labor disutility) is inadequate as a measure of a person's net brute luck. Thus Daniel Shaviro has suggested that liberal egalitarians should employ endowment as an "orienting idea" in thinking about tax-and-transfer policy, and that "[t]he idea [of endowment taxation] deserves greater prominence and acceptance, even though we will never see, and probably do not want to see, a literal, direct endowment tax."84
The question then becomes to what extent it might be both practical and desirable to take steps in the direction of endowment taxation, without adopting a full-fledged endowment tax. One possibility has already been mentioned -- Louis Kaplow's suggestion that the talent slavery problem could be solved by a tax on potential earnings, subject to the limitation that a person's tax liability shall never exceed 90 percent of his actual earnings.85 Kaplow does not actually endorse this idea, and it is obviously a political nonstarter. In addition to the formidable administrative difficulties involved in determining each person's maximum wage rate, even with the ceiling the tax would be opposed by many as unduly burdensome to those who fail to realize their maximum earning potential because doing so would require them to work at jobs they would hate.86
There may, however, be smaller steps in the endowment tax direction which are not subject to such serious objections. A thoughtful discussion of practical ideas for partial implementation of endowment tax principles appears in a recent article by Kyle Logue and Ronen Avraham.87 The significance of the article to the endowment tax debate is somewhat obscured by the fact that the article is framed not as a contribution to the endowment tax literature but as a contribution to the literature on the appropriate use of taxation vis-à-vis nontax legal rules. The article is a response to the claim of some law and economics scholars -- most notably Louis [*pg 1174] Kaplow and Steven Shavell -- that redistribution should be accomplished solely through the tax-and-transfer system, rather than also through nontax legal rules.88 The article's thesis is that many differences in brute luck are not captured by differences in income-producing capacities, that a liberal society may desire to redistribute with respect to some of these non-income-related inequalities, and that legal rules will sometimes have important advantages over tax rules in accomplishing such redistribution.89
Their most compelling example relates to genetically-determined diseases.90 If we want to redistribute from those without disease-related genes to victims of genetic bad luck, it would be theoretically possible to impose a tax directly on the good luck of not having a particular disease-related gene, and to transfer the resulting tax revenue to those with the gene. A similar result could be reached much more simply, however, by a nontax law prohibiting health insurers from discriminating against persons with the gene (either by refusing to offer them any coverage, or by charging them higher premiums). Such a nondiscrimination rule would redistribute from those without the disease-related gene to those with the gene, and this redistribution would be accomplished almost effortlessly compared with the alternative of (1) allowing insurers to discriminate against persons with disease-related genes and (2) using the tax-and-transfer system to offset the effects of that discrimination. The nondiscrimination rule approach thus has the uncanny effect of redistributing on the basis of a particular aspect of genetic endowment without the need for determining anyone's actual genetic information.
Although the fact that the insurance nondiscrimination rule is a nontax law is crucial in the debate between Professors Logue and Avraham and Professors Kaplow and Shavell, the significance of the Logue-Avraham analysis for endowment taxation lies elsewhere. As Logue and Avraham themselves point out, "at the margins, these categories -- tax rules and legal rules -- begin to break down."91 [*pg 1175] Suppose, for example, that the United States moved to a system of universal government-sponsored health insurance, financed by taxation, and that the system was designed (on both the tax and the insurance sides) to avoid discrimination on the basis of disease-related genes. This would accomplish the same redistribution as the nondiscrimination rule suggested by Logue and Avraham, but would it be redistribution by way of tax rules or by way of nontax legal rules? For present purposes, the answer is that it does not really matter. What is important is that Logue and Avraham have identified a situation in which the government could and arguably should redistribute on the basis of one (nonincome related) aspect of endowment; whether that redistribution is better accomplished through tax or nontax means is a lower order question. In any event, if the redistribution occurs in the context of a government-sponsored insurance program financed by taxation, it will be accomplished through the tax system. In fact, something like this is already being done. Medicare is a massive tax-and-transfer system, which redistributes from healthy participants to sick participants (in addition to redistributing along a number of other dimensions). To the considerable extent that good or bad health among the elderly is a matter of brute luck rather than lifestyle choices, Medicare redistributes on the basis of one aspect of endowment.92 Similarly, one of the dimensions along which Social Security redistributes is longevity; the system benefits long-lived recipients at the expense of the short-lived. Longevity can be viewed in this context, somewhat counterintuitively, as a negative aspect of endowment (because of the costs associated with it), against which people in the original position might want to insure through a government-sponsored annuity program. Under this view, Social Security constitutes another form of existing redistribution on the basis of a particular component of endowment, accomplished through a tax-and-transfer program.93
Does this sort of partial endowment taxation make sense? Is it justifiable to redistribute on the basis of one or more selected components of endowment in isolation, without regard to the total endowment of the persons involved in the redistribution? Would it be fair, for example, for a billionaire with a disease-related gene to benefit from the limited endowment tax inherent in a health [*pg 1176] insurance nondiscrimination mandate? Professors Logue and Avraham argue that it would be fair, for two reasons. The first reason they label the "independence assumption."94 The idea is that some types of inequality of brute luck may be appropriate subjects for redistribution, without regard to the overall distribution of brute luck. Thus, redistribution within the health insurance system to those with disease-related genes is justified, even to persons who happen to be rich, if we believe that the genetic "source of inequality is problematic, not as a proxy for well-being, but in and of itself."95 Even if the independence assumption is rejected, however, Logue and Avraham claim it would still make sense to use different redistributive tools to redistribute with respect to different components of overall endowment, with each tool aimed at the component "for which that instrument has a comparative advantage at either observing [the component of endowment] or redistributing [with respect to that component] or both."96 Thus, although a single tax on a comprehensive definition of endowment (i.e., net brute luck) is impossible, the closest practical approximation of that ideal might include, among other things, (1) an income tax as a very imperfect means of redistributing with respect to earnings capacity, (2) health insurance subject to a nondiscrimination mandate as redistribution with respect to the presence or absence of disease-related genes, and (3) Social Security as redistribution with respect to longevity luck.
Bruce Ackerman and Anne Alstott have proposed slouching toward endowment taxation from a different direction.97 They would replace Social Security retirement benefits based on earnings histories with "citizen's pensions" payable in equal amounts to all elderly citizens.98 They would finance these pensions not with payroll taxes, but with a new "privilege tax." Each person would be required to pay the privilege tax annually, from the age of twenty-one to the age of sixty-seven,99 with the size of the annual tax bill depending on one's parents' income during one's childhood.100 The tax would identify [*pg 1177] three levels of privilege: taxpayers with parents in the top 20 percent of the income distribution would pay the highest tax (Ackerman and Alstott suggest $3,800 per year), taxpayers with parents in the middle 60 percent of the income distribution would pay a lower tax (they suggest $2,090 per year), and taxpayers with parents in the bottom 20 percent of the income distribution would pay the lowest tax (they suggest $380 per year).101
Professors Ackerman and Alstott explicitly deny that their proposal constitutes an endowment tax, for two reasons. First, they claim that their proposal is narrower than an endowment tax, because it focuses on a single aspect of endowment -- economic privilege during childhood (what they call "social privilege") -- while disregarding all other differences (including genetic endowment and in-kind parental care and attention).102 But this merely means that their proposal is -- in the spirit of Professors Logue and Avraham -- for a partial endowment tax, rather than a comprehensive endowment tax. This limited form of endowment tax has two significant attractions. First, social privilege (as defined by Ackerman and Alstott), unlike genetic endowment and parental care and attention, is susceptible to reasonably accurate measurement. Second, many persons who believe that society has no valid claim on the benefits of a person's genetic endowment (which society arguably had no role in creating) may concede that society has a legitimate claim on the benefits of economic privilege during childhood (privilege which could not have existed in the absence of the state).
The second distinction offered by Professors Ackerman and Alstott between their proposal and an endowment tax cuts in the opposite direction: they claim their proposal is broader than an endowment tax because an endowment tax is concerned only with earning power, whereas the privilege tax is aimed at the "intangible social and psychological advantages" conferred by childhood social privilege, in addition to "market advantages."103 Although this does make the privilege tax different from an endowment tax as usually defined, in fact it moves their proposal closer to a comprehensive definition of endowment as net brute luck.
Professors Ackerman and Alstott have an interesting answer to the anticipated talent slavery objection -- or, in their words, the concern that the privilege tax "interferes with the liberty of privileged children to pursue low-paying careers."104 According to them, this objection "gets the point precisely backward: the obligation to contribute to society in accordance with privilege is prior to one's liberty to choose careers."105 Although not all readers will be satisfied with this easy dismissal of the talent slavery concern, their response is plausible in the context of their proposed privilege tax, whereas the same response would clearly be inadequate in the context of a proposed full-fledged tax on potential earnings. The priority of the tax obligation over free choice of career is defensible because of two design features of the privilege tax. First, the low maximum amount of annual tax (less than $4,000) means even a would-be full-time beachcomber would have to work only a modest number of hours, with a wide range of occupational choices, in order to pay the tax. Second, it is more plausible to place the obligation to pay for one's social privilege (measured in dollars of parental income) ahead of one's ability to choose a career, than it is to give the same priority to an obligation to pay for one's genetic endowment; this is because society played a greater role in the creation of one's social privilege than it played in the creation of one's genetic endowment.
None of this is to say that the privilege tax is necessarily a good idea. For a tax proposed as a progressive replacement for "the regressive payroll-tax categories"106 currently used to finance Social Security benefits, the tax is strangely regressive. If childhood privilege is appropriately measured by parental income during childhood, a tax that takes from the children of Bill and Melinda Gates less than twice the amount it takes from the children of median wage parents, and only ten times what it takes from the children of the poorest parents, is highly regressive with respect to privilege. In fact, the privilege tax appears to be distributionally quite similar to the regressive payroll tax it would replace. Whether or not one favors the privilege tax, however, the major point for present purposes is that the privilege tax is another example of how we might slouch toward an endowment tax in ways that avoid many or all of the theoretical and practical objections associated with a full-fledged endowment tax.
The taxation of "indicator goods" as a supplement to a tax on labor income can also be understood as slouching toward endowment taxation. This concept, which comes from the economics literature, has been introduced to the legal literature in a recent article by Joseph Bankman and David A. Weisbach.107 An indicator good is one which is consumed differentially by persons of different ability levels, even when those persons have the same earned income.108 Suppose, for example, that high culture (such as Shakespeare and opera) is an indicator good positively associated with ability, while cigarettes are indicator goods negatively associated with ability. In that case, a person with $100,000 of earned income who subscribes to the opera and does not smoke probably has higher innate ability than a person who earns $100,000, smokes, and hates opera. If the income tax were supplemented by taxes on indicator goods associated with high ability and subsidies for indicator goods associated with low ability, that would be a move in the direction of endowment taxation.109 Emmanuel Saez has suggested, rather provocatively, that savings is an indicator good (i.e., at any given level of labor income, persons of higher ability will typically save more than persons of lower ability), and that the taxation of investment income under an income tax can thus be viewed as a move toward endowment taxation, relative to a tax on only labor income.110
Once one begins to think in terms of partial endowment taxation, one may see various aspects of current law as unacknowledged forms of limited endowment taxation. Under Professor Saez's analysis, all taxation of investment income may be so understood. A view of Medicare and Social Security as partial endowment taxation has been noted earlier.111 Another example, on a small scale, is the additional standard deduction for the blind in the federal income tax.112 Or consider the rule, common to the Temporary Assistance for Needy Families (TANF) programs of many states, that a person must [*pg 1180] comply with specified work requirements as a condition of receiving cash benefits, unless one is disabled.113 From one perspective -- admittedly not the usual one -- this resembles an endowment tax on the ability to earn a low wage, rather than no wage at all. If two persons who do not work apply for TANF benefits, and the disabled applicant receives benefits while the able-bodied applicant does not, the denial of benefits to the able-bodied applicant is the equivalent of a tax on her ability to earn a wage.
One may also imagine approaches to partial endowment taxation not embodied in current law. Consider, as an alternative to the privilege tax of Professors Ackerman and Alstott, a similarly tiered tax, similarly imposed on working-age adults, but based on Scholastic Aptitude Test (SAT) scores achieved during high school rather than on parental income. The necessary information would be simpler for the taxing authorities to obtain than eighteen years' worth of parental income, and, given the relatively low level of tax burdens contemplated by Ackerman and Alstott and the importance attached to SAT scores in the college admissions process, nobody is likely to take a dive on the SAT as a tax avoidance strategy.114 In addition, if SAT scores depend on some combination of social privilege and genetic endowment, a tax based on SAT scores might come closer to a tax on overall endowment than a tax based on social privilege alone. On the other hand, as noted above, the failure of the privilege tax to reach genetic endowment is arguably an attraction of the privilege tax, which would be lost if it were replaced by an SAT tax. The privilege tax might also be superior to the SAT tax in that its reach seems limited to aspects of true brute luck, whereas the SAT tax would be based on some combination of brute luck and effort.
Finally, a very different sort of slouching toward endowment
taxation is suggested by Mickey Kaus's contention that government
should seek to produce "social equality," rather than "money equality,"
and that requiring a year of government service from every
eighteen-year-old would promote that goal by "mix[ing] the classes in [*pg 1181] a
common endeavor."115 Kaus
does not describe his proposal in terms of endowment taxation, but it
can be viewed as an in-kind endowment tax. Although it resembles a head
tax in that it requires an equal number of hours from each person to
which it applies, it obviously takes larger amounts of potential
earnings from those with higher wage rates. Thus, it can reasonably be
understood as a proportionate endowment tax (e.g., taking 2.5 percent
of each person's lifetime earnings potential, on the assumption that
each person has forty working years).116
| |
CONCLUSION
Redistribution in favor of those with bad brute luck is attractive
on both utilitarian and liberal egalitarian grounds. This has caused
some to conclude that an endowment tax, defined as a tax on potential
earnings, would be an appealing -- perhaps even an ideal -- tax base,
if only measurement problems could be overcome and talent slavery could
be avoided. It is somewhat surprising, however, that potential earnings
has been so readily accepted as an adequate measure of brute luck or
endowment, when people's luck differs in so many ways not captured by
the measurement of potential earnings. It does not follow, however,
from the inadequacy of potential earnings as a full measure of
endowment, that a stand-alone income tax is the best that can be done
by way of redistributing with respect to differences in brute luck.
Various steps in an endowment tax direction -- including insurance
nondiscrimination rules, taxation of "social privilege," and even
taxation on the basis of SAT scores -- may be both feasible and
desirable complements to income taxation. If the endowment tax
discussion can break free of the assumption that the ultimate measure
of endowment is potential earnings, endowment tax thinking may have --
and may deserve to have -- considerable influence on tax-and-transfer
policy.
| |
FOOTNOTES
Copyright © 2006 by Lawrence Zelenak.
| † | Pamela B. Gann Professor of Law, Duke University School of Law. |
| 1. | Francis A. Walker, The Bases of Taxation, 3 POL. SCI. Q. 1, 14 (1888). |
| 2. | Id. at 15. |
| 3. | Id. at 16. |
| D | AVID F. BRADFORD & U.S. TREASURY TAX POLICY STAFF, BLUEPRINTS FOR BASIC TAX REFORM 36 (2d ed. 1984). |
| 5. | Id. |
| 6. | Id. at 36-38; DAVID F. BRADFORD, UNTANGLING THE INCOME TAX 162-66 (1986). |
| 7. | BRADFORD, supra note 6, at 154-56. |
| 8. | "A consumption tax amounts to a tax on lifetime endowment. . . . An income tax will tend to assess tax burdens in a way presumably correlated with lifetime wealth, but because it depends upon matters of timing, the correspondence is nowhere near as close as . . . under a consumption base tax." BRADFORD & U.S. TREASURY TAX POLICY STAFF, supra note 4, at 39. If two people begin life with the same financial wealth, but one consumes the wealth immediately, while the other saves and invests it to finance greater future consumption, an annual consumption tax will mirror the effects of a one-time tax on financial endowment, in that both the consumption tax and the endowment tax will impose equal burdens on the two persons (in present value terms). By contrast, an income tax will impose a heavier burden on the saver. BRADFORD, supra note 6, at 163. |
| 9. | BRADFORD, supra note 6, at 154. |
| 10. | Id. at 156. |
| 11. | Id. |
| 12. | Daniel Shaviro, Endowment and Inequality, in TAX JUSTICE: THE ONGOING DEBATE 123 (Joseph J. Thorndike & Dennis J. Ventry Jr. eds., 2002). |
| 13. | Kirk J. Stark, Enslaving the Beachcomber: Some Thoughts on the Liberty Objection to Endowment Taxation, 18 CANADIAN J.L. & JURISPRUDENCE 47 (2005). |
| 14. | David Hasen, The Illiberality of Human Endowment Taxation (Aug. 10, 2005) (unpublished manuscript, http://papers.ssrn.com/abstract=779545 (last visited Apr. 4, 2006)). |
| 15. | The discussion in Shaviro, supra note 12, is perhaps the widest ranging in the literature. At least five significant contributions to the literature postdate Shaviro's work, however, and thus are not considered by him. LIAM MURPHY & THOMAS NAGEL, THE MYTH OF OWNERSHIP: TAXES AND JUSTICE (2002); Kyle Logue & Ronen Avraham, Redistributing Optimally: Of Tax Rules, Legal Rules, and Insurance, 56 TAX L. REV. 157 (2003); Daniel Markovits, How Much Redistribution Should There Be?, 112 YALE L.J. 2291 (2003); Stark, supra note 13; Hasen, supra note 14. |
| 16. | E.g., Terrence Chorvat, Taxing Utility, 35 J. SOCIO-ECON. 1 (2006); Logue & Avraham, supra note 15; Markovits, supra note 15. |
| 17. | How to achieve this goal, under various assumed conditions, is the subject of optimal income tax analysis. See generally MATTI TUOMALA, OPTIMAL INCOME TAX AND REDISTRIBUTION (1990) (a comprehensive monograph on optimal income taxation); J.A. Mirrlees, An Exploration in the Theory of Optimal Income Taxation, 38 REV. ECON. STUD. 175 (1971) (the seminal work in the field); Lawrence Zelenak & Kemper Moreland, Can the Graduated Income Tax Survive Optimal Tax Analysis?, 53 TAX L. REV. 51, 51-62 (1999) (a nontechnical introduction to optimal income tax analysis). |
| 18. | For statements that an endowment tax has no substitution effect, and therefore no deadweight loss, see Stark, supra note 13, at 54, and Hasen, supra note 14, at 8. |
| 19. | Chorvat, supra note 16, at 6. |
| 20. | See, e.g., Joseph Bankman & Thomas Griffith, Social Welfare and the Rate Structure: A New Look at Progressive Taxation, 75 CAL. L. REV. 1905, 1919-21 (1987) (explaining the substitution effect of an income tax, which is a "change in the willingness of individuals to sacrifice leisure for consumption"). |
| 21. | See Shaviro, supra note 12, at 137 (arguing that redistribution on the basis of endowment is utility enhancing, assuming a higher wage rate results in higher earnings); Stark, supra note 13, at 54 (same). |
| 22. | See Shaviro, supra note 12, at 137-40 (noting that the results may change if one takes these individual preferences into account); Stark, supra note 13, at 54-55 (same). |
| 23. | Shaviro, supra note 12, at 138. |
| 24. | On the talent slavery concerns of liberal egalitarians, see infra text accompanying notes 33-43. |
| 25. | The willingness of utilitarians to make tradeoffs of this sort is the object of John Rawls's criticism that utilitarianism "does not take seriously the distinction between persons." JOHN RAWLS, A THEORY OF JUSTICE 27 (1971). |
| 26. | Shaviro, supra note 12, at 139. |
| 27. | Shaviro and Stark reach similar conclusions. See Shaviro, supra note 12, at 140 (concluding that endowment is a better tool for measuring inequality than the "proxy standards" of income, consumption, or wealth); Stark, supra note 13, at 54-55 (concluding that endowment taxation could redistribute cash in utility-enhancing directions). |
| 28. | Mirrlees, supra note 17, at 201. |
| 29. | See, e.g., MURPHY & NAGEL, supra note 15, at 119 (describing "equal libertarianism" as calling for redistribution to the extent necessary to produce "equality of opportunity in a market economy"); G.A. Cohen, On the Currency of Egalitarian Justice, 99 ETHICS 906, 916 (1989) (arguing that egalitarianism calls for the elimination of "involuntary disadvantage," but not for the elimination of the consequences of different individuals' different choices); Ronald Dworkin, What is Equality? Part 2: Equality of Resources, 10 PHIL. & PUB. AFF. 283, 292-95 (1981) (distinguishing between "brute luck" and "option luck," and suggesting that redistribution with respect to the former is appropriate while redistribution with respect to the latter is not); Markovits, supra note 15, at 2300 (labeling this view "responsibility-tracking egalitarianism"); Eric Rakowski, Can Wealth Taxes Be Justified?, 53 TAX L. REV. 263, 285 (2000) ("Liberal egalitarians like myself believe that justice asks a good deal of beneficiaries of [brute] luck . . . ."). |
| 30. | That is, one person's innate ability cannot be taken from that person and given to another person. On the possibility of an endowment tax payable in labor rather than in dollars, see infra notes 115-116 and accompanying text. |
| 31. | Cf. Shaviro, supra note 12, at 123 ("If we want to redistribute 'income,' then of course [income] is the right thing to tax . . . ."). |
| 32. | See, e.g., MURPHY & NAGEL, supra note 15, at 121-25 (concluding that "[t]he principal ethical objection to endowment taxation [is that] it would constitute a . . . radical interference with autonomy" and therefore "it would not be a serious option"); JOHN RAWLS, JUSTICE AS FAIRNESS: A RESTATEMENT 158 (2001) (rejecting an endowment tax because it "would violate our basic liberties"); Rakowski, supra note 29, at 267 n.10 (noting that "[v]irtually nobody" advocates a tax based directly on a person's abilities). |
| 33. | For an early use of talent slavery terminology in this context, see Dworkin, supra note 29, at 312, 322. The fullest and most thoughtful discussion of the taxation of earnings potential and talent slavery is in Markovits, supra note 15, at 2305-13. |
| 34. | See, e.g., MURPHY & NAGEL, supra note 15, at 123 ("[T]he problem is not so much that endowment taxation forces people to do what they would prefer not to do, but that it may leave people with literally one option in life."); RAWLS, supra note 32, at 158 (concluding that an endowment tax-which Rawls, somewhat confusingly, refers to as a "head tax"-"would force the more able into those occupations in which earnings were high enough for them to pay off the tax in the required period of time; it would interfere with their liberty to conduct their life within the scope of the principles of justice"); Rakowski, supra note 29, at 267 n.10 ("[Taxing people on the value of unexploited earnings potential] would effectively enslave the able, by forcing them to put their highly taxed talents to some lucrative employ, on pain of sitting in a debtors' prison, however unpalatable the person found richly compensated work"). |
| 35. | See infra text accompanying notes 45-53. |
| 36. | Stark, supra note 13. Daniel Shaviro has also suggested that the concern about talent slavery may be overstated. Professor Shaviro notes that the current tax system penalizes the decision to work for a wage, by taxing wages while not taxing leisure or imputed income from services, but that liberal egalitarians do not view this inducement not to work as violating any fundamental liberty interest. He suggests that liberal egalitarians who are unconcerned about the inducement not to work produced by current law should not be so concerned about the compulsion to work produced by an endowment tax: "Although compulsion is merely a strong version of inducement (suggesting that they differ only in degree, not in kind), the inclination [among liberal egalitarians] to put them in separate categories is apparently strong." Shaviro, supra note 12, at 143. Shaviro is perhaps insufficiently sensitive to the possibility that at some point a difference in degree may become a difference in kind. As Murphy and Nagel remark, being forced to work full time at a particular occupation (because only by so doing can one generate enough cash to pay one's endowment tax liability) is an "extreme interference with autonomy," which has no counterpart under a tax imposed only on income actually earned. MURPHY & NAGEL, supra note 15, at 123. |
| 37. | ROBERT NOZICK, ANARCHY, STATE, AND UTOPIA 169 (1974), quoted in Stark, supra note 13, at 48. |
| 38. | Stark, supra note 13, at 49. |
| 39. | Id. at 59. For the original version of the hypothetical, see MURPHY & NAGEL, supra note 15, at 123. |
| 40. | Stark, supra note 13, at 59. |
| 41. | Id. |
| 42. | Id. at 59 n.64. |
| 43. | MURPHY & NAGEL, supra note 15, at 123. |
| 44. | Louis Kaplow, Human Capital Under an Ideal Income Tax, 80 VA. L. REV. 1477, 1506 n.71 (1994). |
| 45. | Markovits, supra note 15, at 2305-13. |
| 46. | Id. at 2306. |
| 47. | Professor Markovits borrows from Professor Dworkin, Dworkin, supra note 29, at 323-34, the model of endowment tax as insurance against the risk of having a low wage rate. Markovits's development of the model, however, is considerably more thorough and thoughtful than Dworkin's. |
| 48. | Markovits, supra note 15, at 2307-09. |
| 49. | Id. at 2308. |
| 50. | Id. at 2308-09. |
| 51. | Id. at 2310. |
| 52. | Id. at 2313. |
| 53. | Id. at 2323. |
| 54. | Richard Musgrave, Maximin, Uncertainty, and the Leisure Trade-Off, 88 Q.J. ECON. 625 (1974). |
| 55. | Id. at 630 & nn.10-11. |
| 56. | An indifference curve indicates all the combinations of leisure and consumption which would produce the same level of utility for a person. Thus, a person would be indifferent as to the choice among leisure-consumption combinations located on the same indifference curve. |
| 57. | This is, of course, the sort of tax which Professor Markovits concludes would never be agreed to by persons in the original position. See supra text accompanying note 50. |
| 58. | H's net-of-transfer potential wage would then be (4000 x $60) - $100,000 = $140,000, and L's net-of-transfer potential wage would then be (4000 x $10) + $100,000 = $140,000. |
| 59. | This is closely related to James Mirrlees's observation that a utility-maximizing tax on potential earnings will leave the more talented with lower utility than the less talented. See Mirrlees, supra note 17, at 201 ("It is worth noticing that . . . with the particular welfare function specified . . . utility will be less for more highly skilled individuals."), discussed supra in text accompanying note 28. |
| 60. | Musgrave, supra note 54, at 631 n.11. |
| 61. | Id. at 632. |
| 62. | Id. |
| 63. | See, e.g., MURPHY & NAGEL, supra note 15, at 123 (considering the example of a potential sculptor who must work as a corporate lawyer to pay his taxes under an endowment tax). |
| 64. | Dworkin, supra note 29, at 303. |
| 65. | Markovits, supra note 15, at 2315. |
| 66. | The discussion in the text only scratches the surface of the
possible relationships between potential wages and the utility or
disutility of work. As a more elaborate example, consider the following
three persons, all of whom work the same number of hours: A works at a high wage job
(HWJ) which he neither likes
nor dislikes.
If the ideal tax base is net endowment (i.e., potential wages plus work
utility or minus work disutility), then B has a net endowment
at least as high as A, and C has the same net
endowment as B. As the discussion in the text indicates, a net
endowment tax could reasonably tax B on his high potential
wages, even if it meant B had to move from the low wage job to
the high wage job. The problem for the net endowment tax is C,
who is just as well off in net endowment terms as B (same wages
and same utility from work), but who unlike B does not have
the ability to cash out his love of his job by moving to a high wage
job. A comprehensive tax on net endowment may not be possible, then, if
it would impose a higher tax burden on C than he could
possible pay in cash. The response to this problem might be to revise
the concept of net endowment, so that one's utility should be included
in the tax base only to the extent one could convert it to cash. The
tax would then be neither a pure tax on utility nor a pure tax on
potential earnings, but a hybrid of the two.B could work at the same HWJ as A, which he would neither like nor dislike, but works instead at a low wage job (LWJ), which he loves. C does not have the skills for HWJ, so he works at the same LWJ as B, which he loves as much as B loves his LWJ. |
| 67. | Louis Kaplow, The Theory of Taxation and Public Economics ch. 5, at 17-18 (2006) (unpublished preliminary draft, on file with author). |
| 68. | For example, consider individuals A and B in the illustration in note 66. Under a comprehensive definition of endowment, B's endowment is at least as great as A's, but B's income tax liability would be much less than A's. |
| 69. | Hasen, supra note 14. |
| 70. | Id. at 13. |
| 71. | Id. at 15. |
| 72. | Id. at 30. |
| 73. | Id. |
| 74. | Id. at 17. Mark Kelman appears to have taken essentially the same position, in a much less fully-developed form, a quarter century earlier. Mark G. Kelman, Personal Deductions Revisited: Why They Fit Poorly in an "Ideal" Income Tax and Why They Fit Worse in a Far from Ideal World, 31 STAN. L. REV. 831, 842 n.33 (1979). |
| 75. | Hasen, supra note 14, at 19. Professor Hasen does not discuss the possibility of an ultimate exit option under an endowment tax-i.e., that a person who did not want to be taxed on his unused earnings potential could escape the tax by leaving the country, or perhaps by remaining in the country but living on his own in the wilderness. Presumably he would consider the protection of autonomy afforded by an exit option to be insufficient. |
| 76. | Id. at 30 n.94. |
| 77. | Id. at 27. |
| 78. | Id. |
| 79. | Id. at 28. |
| 80. | Id. at 30. |
| 81. | MURPHY & NAGEL, supra note 15, at 121-25; Dworkin, supra note 29, at 311-12; Markovits, supra note 15, at 2305-13. |
| 82. | See supra notes 45-53 and accompanying text. |
| 83. | According to Professor Hasen, autonomy is a "constitutive" principle of liberal egalitarianism, rather than a mere "regulative" principle, with the result that it cannot be balanced against other values. Hasen, supra note 14, at 17-19. Apparently, anyone who disagrees with Hasen on this point is not, in Hasen's view, a true liberal egalitarian. |
| 84. | Shaviro, supra note 12, at 125, 144. |
| 85. | See supra note 44 and accompanying text. |
| 86. | Conversely, endowment tax purists-if any exist-would object that the 90 percent-of-actual-wages ceiling prevents the hybrid tax from being a close approximation of a true endowment tax, for those who earn far below their potential. See supra text accompanying note 44. |
| 87. | Logue & Avraham, supra note 15. |
| 88. | LOUIS KAPLOW & STEVEN SHAVELL, FAIRNESS VERSUS WELFARE (2002); Louis Kaplow & Steven Shavell, Fairness Versus Welfare, 114 HARV. L. REV. 961 (2001); Louis Kaplow & Steven Shavell, Why the Legal System is Less Efficient than the Income Tax in Redistributing Income, 23 J. LEGAL STUD. 667 (1994). |
| 89. | Logue & Avraham, supra note 15, at 252-57. |
| 90. | Id. at 208-18. |
| 91. | Id. at 208. |
| 92. | Professors Logue and Avraham briefly note this point. Id. at 226-27. |
| 93. | Id. at 227 n.211 (describing Social Security as "produc[ing] transfers from the short-lived to the long-lived"). |
| 94. | Id. at 245. |
| 95. | Id. |
| 96. | Id. at 246. |
| 97. | See BRUCE ACKERMAN & ANNE ALSTOTT, THE STAKEHOLDER SOCIETY 155-77 (1999) (proposing a "privilege tax" to finance the Stakeholder Society). |
| 98. | Id. at 129-54. |
| 99. | Id. at 168-69. |
| 100. | Id. at 166-67. |
| 101. | Id. at 159, 166-68. |
| 102. | Id. at 259 n.10. |
| 103. | Id. |
| 104. | Id. at 264 n.52. |
| 105. | Id. |
| 106. | Id. at 169. |
| 107. | Joseph Bankman & David A. Weisbach, The Superiority of an Ideal Consumption Tax Over an Ideal Income Tax, 58 STAN. L. REV. 1413, 1453-55 (2006). |
| 108. | Albert L. Nichols & Richard J. Zeckhauser, Targeting Transfers Through Restrictions on Recipients, 72 AM. ECON. REV. 372, 375 (1982). |
| 109. | From this perspective, the current practice of taxing cigarettes and subsidizing opera seems to be precisely backwards. |
| 110. | Emmanuel Saez, The Desirability of Commodity Taxation Under Non-Linear Income Taxation and Heterogeneous Tastes, 83 J. PUB. ECON. 217, 227-28 (2002). |
| 111. | See supra text accompanying notes 92-93. |
| 112. | I.R.C. § 63(c)(3) (2000). |
| 113. | E.g., CAL. WELF. & INST. CODE § 11320.3(b)(3)(A) (West 2001 & Supp. 2006). |
| 114. | James Mirrlees has suggested, rather casually, that a tax based on I.Q. scores might be attractive for a similar reason. High I.Q. scores "may be sought after so much for prestige that they would not often be misrepresented." Mirrlees, supra note 17, at 208. Because the practical advantages of a high SAT score are significantly greater than those of a high I.Q. score, in the United States today, an SAT tax would be preferable to an I.Q. tax. |
| 115. | MICKEY KAUS, THE END OF EQUALITY 81 (1992). |
| 116. | The idea of in-kind endowment taxation could be taken in other directions as well. For example, a person with high earnings potential who chooses to work at a low paying job producing clear social benefits-such as a public school teacher-might be considered to have satisfied some or all of her obligation under an endowment tax in kind, whereas an equally talented person earning an equally low wage job bereft of social benefits would not have satisfied any of her tax obligation in kind. However, identifying the jobs producing sufficient social benefits would be a daunting-perhaps impossible-task. |