The Uneasy Case for Capital Taxation


I write not to praise all tax-reductions but to bury one particular set of taxes.

Over a decade ago I began writing on the subject of comprehensive tax reform, in a piece titled The Uneasy Case for Wealth Transfer Taxes,[2] echoing the classic work on tax-rate progression by the law professors Walter Blum and Harry Kalven.[3] While I have never viewed the case for progressivity in tax burdens as especially uneasy, I have now come to see that the case for any tax on capital—that is, any direct tax on capital, a qualification to be made clear, below—is. Simply put, my argument is that there is no compelling reason of fairness or justice to tax capital qua capital, that is, the mere possession of material resources in the hands of an individual or, even more strongly, in the hands of an entity such as a corporation. We can and should abolish capital gains and other capital taxes under the income tax, wealth and wealth transfer taxes such as the gift and estate tax, and, perhaps especially and paradigmatically, corporate income taxes. As this is a claim that I suspect many if not most readers will find surprising, I shall limit my argument here to a fairness or moral case against direct capital taxation; suffice it to say that others, writing in an economics tradition, have raised considerable doubts as to the (possibly related) efficiency of these levies as well.[4]

To clarify from the outset: mine is not any of the three most familiar arguments against capital taxation.[5] It does not follow from the argument, owing to Hobbes, that capital represents a “common pool” of social resources that ought not to be taxed until individuals withdraw from it.[6] To the contrary, mine is not an argument about the aggregate capital stock; it does not depend on the idea of savings as a public good, or on any instrumental quest to get more social capital. Nor is mine an argument of “horizontal equity,” of failing to treat likes alike or (equivalently) equals equally, such as the one most famously sourced to Mill, namely that any “income” tax—any tax that includes both the initial receipt of wealth and the subsequent yield to capital in its base—is a “double tax” on wealth or resources that are not immediately consumed, penalizing savers over spenders, noble Ants over spendthrift Grasshoppers (taken to be ex ante equals by Mill and the traditional view of tax).[7] To the contrary, I accept that, under some circumstances, savers ought to be taxed more, and other times less, than spenders, and, like other thoughtful commentators in the increasingly sophisticated tax policy literature, I eschew a naïve, formalist, horizontal equity approach to tax. [8] Nor, finally, is my argument against capital taxation the simple, lay argument against “double” taxation, come what may, although capital can often be triply or quadruply taxed, at the individual level both when received and when invested, at the corporate or entity level, and again at the individual level when transferred. But many dollars are taxed multiple times in the flow of funds, and the number of times an element of value is taxed is simply less important than the rate at which it is taxed. A single high rate tax can be more burdensome than a panoply of trivial rate ones. My argument happens to be for a one-time tax on individuals, at the moment of spending or ultimate private preclusive use, but this is because I argue that this one time is the right time to make social judgments over the appropriate level of taxation, and not on account of any foundational constraint on the number of times individuals or elements of value can be taxed.

Rather my argument is about progressivity, the very thing that Blum and Kalven found “uneasy.” The central, animating question is when judgments about progressivity in tax burdens should be made, which necessarily runs out to questions about how capital and its yield fit into a normatively attractive account of the fair distribution of tax burdens. I argue that, given that we are going to have a progressive tax system, in which the better-able-to-pay pay more, in percent terms, than the less-better-able-to-pay do—a proposition that I happen to accept as both factually accurate and normatively compelling, but, more to the point, one I simply assume to move forward the analysis of this essay—the next question for policymakers is when to levy such a tax. The goal of any comprehensive progressive tax is to effect a fair distribution of burdens from individuals; this calls for individuated judgment. “Genuinely progressive taxation is necessarily personal taxation,” as the Nobel laureate William Vickrey began his classic 1947 treatise.[9] When should we decide how much individuals ought to share with the body politic or, perhaps better put, when should we decide what is private and what public, what is a fair distribution of resources, in the first instance?[10] The key insight is that we ought to tax people when they use their wealth—that is, spend—and not when they save, give, or die: our ordinary and reflective moral intuitions ought to consistently run out to the uses of material resources, and to not their sources. Capital is presently unused, unconsumed, wealth. It can be put to different uses at different times at the individual level. Society can reasonably make different judgments about the propriety of taxing different uses of capital. We can and should, that is, wait and make judgments about capital in the hands of individuals when their ultimate private preclusive uses of that capital become manifest.

These thoughts lead out naturally enough—with the aid of insights gleaned from tax policy tradition—to a specific form of comprehensive individuated tax, namely a progressive postpaid consumption one, a progressive spending tax in short. These are terms that I shall make clearer in due course. The critical understanding is to see that such a tax is a tax on capital, at the individual level, when (but only when) capital is used to finance enhanced lifestyles or greater consumption of material resources—spending—and not when capital is used simply to move around in time, within or between generations, uneven labor market earnings. This is a compelling moral endpoint for tax. Once we get the major comprehensive tax system down right—from a strictly moral point of view—there is no longer any compelling reason for, and there are good reasons against, any of the traditional direct taxes on capital.

It is time to better explain and defend these claims.

A. The Traditional View

Much of tax policy in the United States and elsewhere has been consumed with debating the relative merits of an income versus a consumption tax.[11] This debate has been framed by the so-called Haig-Simons definition of “Income,” which holds, in essence, that Income equals Consumption plus Savings (I = C+ S).12 This accounting identity—a mere tautology—tells us no more and no less than that all Sources equal Uses or, even more simply put, that all material resources (Income) are either spent (Consumption) or not (Savings).[13] This is hardly profound. But great wisdom can be built on simple truths. The Haig-Simons definition of income has been enormously influential in tax policy. Perhaps most important, it has been used, through a simple rearrangement of terms, to show the essential difference between an income and a consumption tax. If Income = Consumption + Savings, then Consumption = Income – Savings. It appears as if the difference between an income and a consumption tax is that the former includes, and the latter does not, capital or savings in its base.

But there are in fact three major choices of broad-based tax systems in ideal theory: the income tax, prepaid consumption taxes, and postpaid consumption taxes. We get to the three-view perspective when we see that the two broad types of consumption taxes are not created equal under progressive rates. First let us understand the traditional view of matters, which equates the two forms of consumption tax.

An income tax, as we can see from the Haig-Simons definition, applies to all inflows into a household or taxpaying unit, whether from capital or labor (or, indeed, beneficence). This means, as Mill pointed out in his 1848 treatise, Principles of Political Economy, that savings are “double taxed”: in order to have principal to invest, one has to have paid tax on some prior receipt, but the yield to capital is taxed again. In Mill’s words, this is to “tax the same portion of the contributor’s means twice over”; if a taxpayer “has the interest, it is because he abstains from using the principal.”[14]

Consumption taxes, in contrast, are single taxes on the flow of funds into and out of a household. This way of putting the matter allows us to comprehend the two basic forms of consumption tax, which depend on the time when the single tax is levied. In one model, the tax is imposed up-front, and never again: a wage tax, or a so-called pre-paid or yield-exempt consumption tax. “Roth” IRA’s in the United States work on this model (pay tax now, never again).[15] The second form of consumption tax imposes its single tax on the back-end, when resources flow out of households: this is a sales tax, a postpaid, cash-flow or “qualified account model” consumption tax. Traditional IRAs in the United States work this way (no tax now, only later).

Under flat or constant tax rates, the two principal forms of a consumption tax are in fact largely equivalent, a result that can be proven in relatively simple algebraic terms.[16] This equivalence has led to a confusion in the traditional view of tax, an over-
quick equation of prepaid and postpaid consumption taxes. To see this equivalence and also to consider further Mill’s celebrated “double tax” argument against the income tax, a simple numeric example proves illustrative.

Suppose that Ant and Grasshopper each earns $200 in wages, the tax rate is 50 percent, and the interest rate on savings is 10 percent. Grasshopper, as is his way, spends all of his available money at once. Under any tax—income, prepaid or postpaid consumption—the government takes its 50 percent cut, or $100, and Grasshopper consumes the remaining $100. This demonstrates an important point: the choice of income versus consumption taxation has no direct impact on most taxpayers, for the simple reason that they do not save. (If I = C + S, and S = 0, then I = C). But, of course, patterns of progressivity, saving, borrowing, and spending across the entire income and wealth scale matter to all citizens, even if not directly so.

Ant, in contrast, does save, and so the choice of tax does matter directly to her. Suppose that Ant saves her initial wage-earnings for two years, at the conclusion of which she consumes these initial wages plus any interest received on account of her saving them. How do the three different taxes treat her?

An income tax reduces Ant’s $200 to $100 right away, which she puts in the bank. Ant earns 10 percent on her savings, or $10, in Year 1, but the income tax hits this, too—Mill’s double tax—taking away $5, leaving her with $105 at the end of Year 1. In Year 2, this $105 again earns 10 percent, or $10.50; again the income tax strikes, taking $5.25; this leaves Ant with $110.25 to consume at the end of Year 2. If the 10 percent interest rate simply compensated Ant for inflation—if the cost of goods were rising at 10 percent per year—Ant would be losing real value, actual purchasing power, over time under the income tax:

$110.25 at the end of two periods of 10 percent inflation is worth—has the same real purchasing power as—$91 at the start of the two periods.[17]

Consider next the two forms of consumption tax. First, the prepaid model: Ant is taxed on initial receipt under this system, reducing her $200 to $100. But she is not taxed again: recall that consumption taxes are single taxes, escaping Mill’s double-tax label. The $100 grows by the full 10 percent interest rate, to $110, after Year 1. In Year 2, the $110 earns another 10 percent, or $11, to $121, and Ant is left to consume this much at the end of Year 2. Unlike the case with the income tax, this end of Year 2 consumption is worth the same as $100 at the start of Year 1, under a 10 percent inflation or discount rate.

Under the postpaid consumption tax model, Ant pays no tax up-front and so can save her entire $200. This grows by 10 percent, or $20, in Year 1, to $220. The $220 grows by another 10 percent, or $22, to $242 in Year 2. When Ant goes to consume this, the government collects its 50 percent share, leaving Ant with $121 to consume. This is just as under the prepaid model. And it is more than the income tax. There is no smoke and mirrors here. There are only two critical assumptions needed to make out the equivalence of prepaid and postpaid consumption taxes: that the interest and tax rates have stayed constant in the two periods.[18]

The Ant-Grasshopper example, or something rather like it, stands at the center of the traditional view of tax. The income tax is a double tax on value that is not immediately consumed, which has led many conservatives to oppose it as an unfair burden on the noble Ant, but liberals to support it as a necessary means of capturing some of the return to capital, the nearly exclusive domain of the wealthy. Both forms of consumption tax get put on the other side of a divide, as not reaching the yield to capital at all. It becomes a matter of either indifference or administrative convenience which of the two forms is chosen. And, in this traditional view, the debate over income versus consumption taxation is an all-or-nothing one over whether to tax capital (all capital) at all.

This traditional view of tax is flawed. B. A New Understanding

The traditional view’s equivalence of prepaid and postpaid consumption taxes does not hold under non-constant or progressive rates. Once we assume at least some progression in the rate structure, the traditional understanding of consumption taxes is no longer accurate.

Progressive rates under most comprehensive tax systems work through a series of marginal rate brackets, that form, in mathematical terms, a step-function. To have a simple and illustrative structure in mind, suppose that no tax is paid on the first $10,000; followed by a 15% marginal rate on the next $40,000; a 30% rate on the next $50,000, and so on.

Income or Consumption Marginal Tax Rate
$0 -10,000 0%
$10,000-50,000 15%
Over $50,000 30%

Table 1 Simple Marginal tax Rate Schedule

Such a system effects progression in average or effective tax rates. A taxpayer who has $100,000 subject to this tax, for example, will pay total taxes of $21,000 ($6,000, or 15% of $40,000, plus $15,000, or 30% of $50,000), for an average tax rate of 21%; this is a higher tax rate than some one who makes $50,000, who pays $6,000 (15% of $40,000), or 12% in average tax.

The two forms of consumption taxes differ in their effects under progressive rates. Now there are three—not two—alternatives for the tax policymaker to choose. The differences come in when the tax falls, and how this impacts choices of work, savings, education, and so on, and, most important, in how the tax redistributes material resources. The time-path of earnings and spending, inflows and outflows—and with them, the role of capital transactions—now matters critically to the total tax burden. Consider each tax in turn.

One, an income tax falls on all labor market earnings and on the yield to savings, at the time they come into a household. Savers are hurt by the “double taxation” of savings, whatever their intended or actual use. Individuals, like the athletes, artists, and the highly educated, who see their earnings come in relatively short, concentrated, bunches, are also hurt by the timing of the imposition of progressive rates.

Two, a prepaid consumption tax falls on labor market earnings alone, again at the time they come into a household. Once more, people whose earnings profiles are uneven throughout their lifetimes are hurt by the timing of the imposition of the progressive rate structure. But—and here is the rub for most liberals and even moderates—those who live off the yield to capital are never taxed.

Three, a post-paid consumption tax does not come due at the time of initial inflows, but rather at the time of outflows, when money is spent in consumption. This means that a progressive postpaid consumption tax stands between an income tax, which double taxes all savings, and a prepaid consumption tax, which ignores all savings. A consistently progressive postpaid consumption tax treats savings differently depending on their use, as I shall continue to explore in the following sections.


Before continuing on with an exploration of how individuals in the normal course of their lives use capital and its yield, critical to the new three-tax perspective on tax policy, let us pause and reflect on the norms—our ordinary moral intuitions—about capital. Mill’s claim that the income tax is a double tax on savings is descriptive, an analytic fact. It is true both within the income tax’s own base, where savers are penalized vis a vis spenders, and relative to a hypothetical no-tax world, where the income tax destroys the equivalence, in present value terms, between savers and spenders, Ants and Grasshoppers. Yet neither of these facts exert a strong pull on our moral intuitions; it is hard to get from Mill’s is to any compelling ought. This tension featured prominently in an extended debate begun in the 1970s between the American law professors William Andrews and Alvin Warren, over the income versus consumption tax. Andrews first pressed Mill’s position, arguing that “the most sophisticated argument” for consumption taxation rested on preserving the pretax equality of spenders and savers. [19] This is an argument sounding in horizontal equity (comparing savers and spenders) and also taking an ex ante perspective (looking at the moment of decision to save or spend as the right time to make social judgments about fair taxation). Warren counterpunched by taking both an ex post (after the distribution of capital market returns) and a vertical equity

perspective, arguing that those with greater “ability to pay” or (equivalently) more material resources ought to pay more than those with less: Ant has more material resources in the later, second period than Grasshopper does, so why shouldn’t we tax her more?[20]

Warren’s arguments had prevailed, decades before he actually made them, at the dawn of the creation of comprehensive individual tax systems in the U.S. in the early years of the 20th century, and elsewhere, later in the century. Reformers actively desired an income tax because it included the yield to savings, and thus would impose an added burden on financiers and the like.[21] Those were, however, simpler times. As the income tax expanded in both scale, becoming a higher burden and more steeply sloped in its rate progression, and scope, reaching the majority of earners in the U.S. and elsewhere, things changed.[22] Lawmakers began to have second thoughts about double-taxing the yield to savings, anywhere and everywhere. A near century of experience with a so-called income tax in the U.S. and elsewhere in the developed world has by now shown a deep split about the normative propriety of taxing the yield to capital. More and more exceptions to the income tax’s theoretical commitment to double-taxing savings have been piled on one another, whether by happenstance, inertia, deliberate policy plan, or mere mistake; examples include tax-favored medical, education, and retirement savings accounts, the nontaxation of “unrealized” appreciation, and the rather systematic exclusion of the financial gains from personal residences.[23] The result is that we now observe “hybrid” taxes, perched—typically, uneasily—between an income tax model, with its double tax, and a consumption tax, with its principled nontaxation of savings. Trouble is, the compromises to bring about this state of affairs have been effected without suitable normative or practical reflection, resulting in a tax system in which the well-endowed—the capitalist class—can live well and consume away, tax-free. We are neither favoring savings nor effecting a fair distribution of tax burdens across taxpayers; individuals who can live off the yield to capital quite simply need pay no tax.[24]

Consider a simple example, drawn from my longer work.[25] Take the case of Artful Dodger, who happens to have the sum of one million dollars. It does not matter much for the illustration how Dodger got this wealth. If he earned it via wages, he would have paid income and payroll taxes on it; if he received it as a gift or as the proceeds of life insurance, for example, he would never have paid tax on it, and, depending on the circumstances of his benefactor, it is possible that no one ever paid tax.[26 ] The point is that Dodger need never pay tax again. He can invest his million in non-income producing property, such as growth stocks; such property rises in value without producing taxable cash each year. Dodger can borrow, income tax-free, to finance his lifestyle. When he dies, his heirs can sell their inherited property, tax free, and pay off Dodger’s debts, continuing the pattern with any value that is left over. The details vary, but the basic point is that those who live off the fruits of financial capital need pay no federal taxes in the United States, at least, while those who live off human capital—those who get paid for their labor—are hit, and hit hard, by income and payroll taxes, combined.

On reflection, the schisms in contemporary tax systems between income and consumption tax elements are not random. Looking back to the Andrews-Warren, income-versus-consumption debate, settled reflection reveals that ordinary moral intuitions in fact reasonably reach different normative judgments about different uses of savings. On the one hand, many citizens are indeed sympathetic to the noble Ant, especially when she is manifest as a middle-aged wage-earner, struggling to make ends meet while paying her taxes and setting aside some funds for her later retirement, or medical or educational needs within her family. Why should we punish her, with a second tax, for her prudence? And so we observe tax-favored retirement, medical, and educational savings accounts. On the other hand, many citizens are also bothered by the specter of the socially privileged, such as a second or third-generation rich child, like Artful Dodger, living well off the fruits of someone else’s prior capital accumulation. Surely this “trust fund baby” should be taxed, at least as much as the hard working Ant? Surely his income, in the form of rents, royalties, interest, dividends, and the like should count in the tax base, at least as much as the product of noble Ant’s blood, sweat, and tears?

These simple insights and intuitions in fact cash out into two discrete norms about capital. Not all uses of savings are the same. One norm I call the ordinary-savings norm: capital transactions (borrowing, saving, investing) that are simply used to move around uneven labor market earnings in time, allowing people to save for their retirement, or for periods of high spending needs/low earnings, such as times of education or medical urgency, should not be double-taxed or otherwise discouraged and burdened. The second norm I call the yield-to-savings norm: capital that enables a higher, better—more costly—lifestyle than the yield to labor, alone, should bear a burden, one at least commensurate with normal wage earnings.

Ordinary moral intuitions thus agree with both Andrews’s horizontal equity position, and Warren’s vertical equity one, through these two norms, the ordinary savings and yield-to-capital ones. Savers should not be penalized for saving, not consuming, in the ordinary course of their lives, for rainy days or times of greater need or urgency, but the yield to capital is also an increment of value that should not be simply and completely ignored in the tax base. The trick is to design a tax system that implements both norms, simultaneously, without undue complexity. It turns out that, with the right understanding of tax—mirabile dictu—this is surprisingly easy to do.


With this new normative vocabulary in hand, we can now return to the discussion of the uses of material resources, focusing especially on the uses of capital. Consider in financial terms how most of us live out our lifetimes. As any parent knows full well, we spring forth into the world nearly fully formed as consumers: we cost money from the get-go. But (as any parent also knows) we do not earn anything for quite some time. When we do start earning, we have to earn more than we spend (let us hope!), to pay off the debts of our youth, including school loans, and to set aside funds for our retirement, so that we do not have to keep working all the days of our lives. Our lives look like one fairly steady consumption profile, from cradle to grave, financed by a lumpy period of labor market earnings concentrated in midlife. If we lived as islands, unto ourselves, we would have to balance the books on our own account, borrowing in youth, first paying off our debts and later saving for retirement in our mid-life, spending down in old age. Financial intermediaries such as banks and insurance companies would help us to effect these results. In practice, many families work as more or less informal annuities markets, between generations. Thus our parents pay for our youths, and we pay for our children’s youths; we also stand ready to pay our parents back, should their needs exceed their resources in their old age.[27] And so on.

In this stylized depiction of a typical life, note two broad uses of savings. One is to smooth out consumption profiles, within lifetimes or across individuals—to translate uneven labor market earnings into even consumption flows. We do this by borrowing in youth and saving for retirement—and/or other times of special need, such as health and education demands— in mid-life. We can do this using third party financial intermediaries, or within the family, as noted above: perhaps we pay for our children’s youth, and they pay for the youths of their children, our grandchildren, in a recurring “overlapping generations” model.

A second use of savings is the analytic complement of smoothing: capital transactions can shift consumption profiles, up or down. An upward shift occurs when the fruits of our own or another’s savings (via beneficence) allow us to live a “better” lifestyle than we could on the basis of our own labor market earnings, alone, smoothed out over time. Suppose that the noble Ant had gotten wondrously lucky on the $200 of wage-earnings that she had saved; some investment yield her millions of dollars. Ant can then, quite simply, enjoy a “better”—more costly—lifestyle than other ants or grasshoppers earning the same amount of labor market wages. The progressive postpaid consumption tax would thus tax her, on her spending. A downward shift, in contrast, occurs when our Ant’s beneficence or bad fortune means that she will live at a lower—less costly—lifestyle than she otherwise could, all measured off the baseline of her smoothed out labor market earnings profile alone.

The two norms considered in the prior section map up perfectly with these two uses of capital. Smoothing effects the ordinary-savings norm; shifting the yield-to-capital one. Ordinary moral intuitions, reflected in a near-century of experience with actual tax systems, suggest that society ought not to burden smoothing transactions with a double tax, but that the yield to capital is an element of value that can properly be taxed when used to enable a “better,” more expensive lifestyle. This is not envy. It is not, that is, that the rich should be penalized, or that those who earn wealth from capital should be brought down and laid low. It is, rather, the sensible thought that the yield to capital is an increment of value, that deserves to be counted in one’s resources available to pay tax, except when savings are used simply to move values around in time. Such movements in time are one thing, greater material enjoyment is another thing. It is all, in essence, about the fair timing of tax.

Now return to the three basic tax systems, income, prepaid consumption, and postpaid consumption. Under progressive rates, the three basic tax systems discussed above affect different patterns of savings and spending differently. An income tax double taxes all savings, come what may, and makes it judgments of progression on the basis of inflows, however uneven. A prepaid consumption or wage tax ignores all capital transactions, again whatever their use, and also makes its judgments of the fair degree of progression on the basis of inflows, burdening the uneven wage earner. But a consistent progressive postpaid consumption tax, wondrously enough, implements the ordinary-savings and yield-to-capital norms, simultaneously, seamlessly, and by design.

A simple example helps to make points clearer. One taxpayer, Steady Earner, makes and consumes $50,000 a year for the relevant years of comparison, say beginning in her early 20s. A second taxpayer, Lumpy Earner, stays in school until he is 30, and then makes $100,000 a year. But Lumpy Earner spends $50,000 a year, too, using prudent borrowing and saving to effect this result.[28]

Finally, Trust Fund Baby lives off his parents’ fortune, getting and spending

$50,000 a year (which represents a 5% yield off a trust corpus of $1,000,000, small change for the rich today). How do the three taxes affect these three individuals, under the simple progressive rate structure posited above?

An ideal progressive income tax burdens all three taxpayers, but falls most heavily on Lumpy Earner, because of the timing of the imposition of the progressive rates. In the simple rate structure posited above, Lumpy pays 21% of his earnings in income tax (as calculated from Table 1), whereas Steady Earner and Trust Fund Baby each pay 12%.

A progressive prepaid consumption tax also burdens Lumpy Earner most heavily, at a 21% level given the same rate structure, continues to tax Steady Earner at the 12% level, but altogether ignores Trust Fund Baby—taxing him at 0%, thereby accepting the tax, if any, in some prior generation of labor earnings as sufficient for his contribution to society.

A progressive postpaid consumption tax, in contrast, falls equally on all three taxpayers, at the 12% level.

Note that, should any of the three taxpayers get lucky in the capital markets—win the lottery, say, or simply get extra high returns from investments—then the progressive postpaid consumption tax stands ready, at the wait, to tax that good fortune when and as it is used to enhance, or elevate, their lifestyles.

In sum, whereas an ideal income tax double taxes all savings, whatever their use, and a prepaid consumption tax ignores all savings, again whatever the use, a consistent progressive postpaid consumption tax splits the difference, in a principled way, and by design. It allows taxpayers to lower their taxes by smoothing, but it falls on the yield to capital when such yield is used to enhance

lifestyles. This reflects simple, commonsensical attitudes about life, income, and savings. These attitudes are reflected imperfectly under the status quo in the United States and other advanced Western democracies, with a nominal income tax rife with pro-savings provisions for retirement, health, and education.


The better understanding of the analytics of tax that we have now attained can lead to a dramatically simpler tax system that is at the same time far fairer, one that perfectly incorporates the ordinary moral intuitions about savings—namely that savings for some purposes, which we can broadly call smoothing, should not be burdened twice over, but that savings that enable a higher material lifestyle can and should be subject to tax: a tax that is, some positive tax burden commensurate with all other sources of material enjoyment or private preclusive use. The central insight is that a consistent progressive postpaid consumption tax is a tax on the yield to capital, under just the circumstances in which ordinary moral intuitions suggest taxing such yield, and no other. Such a tax can be implemented easily enough, taking advantage of the rearrangement of the Haig-Simons definition of Income noted above: Consumption = Income – Savings (C = I – S). This handy form shows that a consistent postpaid consumption tax can be obtained by tallying up sources of income as under current law, and systematically subtracting savings (and adding in dissavings).[29]

The tax law would thereby, however surprisingly, look much like it does today, in the United States at least. There would be annual wage reporting from one’s employers and annual income tax returns like dreaded the 1040 forms used in the U.S.. Instead of the myriad of tax-favored accounts we observe today, however, there would be a single, unlimited savings account for every individual, which we can call, to make points clear, a Trust Account. All contributions to these Accounts would be deductible, and all withdrawals from them would be includible in taxable “consumed income.” Debt that was used to consume, as dissavings, would likewise be taxable; repayments of principal, as positive savings, would be deductible. In such a fashion, the law would work out the logic of “Income minus Savings.” Ordinary saving for retirement would lower the burden of taxation, by changing the time of taxation from one’s high-earning, midlife years, to the smoother, lower levels of one’s consumption. Saving for medical needs or other special circumstances, such as education, could also lower the burden of one’s taxes, especially if the law imposed a lower (or no) taxation on these uses, parallel to the United States’ current, limited, deduction for extraordinary medical expenses.[30] Lumpy Earner, for example, who makes $100,000 in wages but saves $50,000 for his retirement, would pay tax on $50,000 (100,000 – 50,000). Trust Fund Baby would pay tax on the $50,000 he withdrew from his Trust Account. And Steady Earner, who neither borrows nor saves, would pay tax on her earnings, which equal her consumption, $50,000. Note that there is no need whatsoever to tally up particular items of expenditure; this is all quite general. To this base, a progressive rate structure can apply, like that illustrated by Table 1.

The important point to see is that this progressive postpaid consumption tax would, in and of itself, make for an individuated tax on capital, when, but only when, capital is used to enhance lifestyles. No other tax on capital would therefore then needed—and, in part because any other tax on capital is not so individuated, and hence risks falling on ordinary savings as well as the yield to capital, all “direct” taxes on capital should be eliminated.

Consider first the role of “second” taxes on the yield to capital under the basic individuated tax system, such as capital gains under the income tax. These are simply not needed under a consistent progressive postpaid consumption tax. If a taxpayer sells an asset and reinvests the proceeds, she has continued to save, and there is no reason to tax her—yet; this can all be done inside her Trust Account. On the other hand, any mechanism to finance her lifestyle—wages, the ordinary yield to capital (interest, dividends and the like), someone else’s beneficence, the proceeds of sales of capital assets or, for that matter, borrowing against present assets or future earnings—is taxed, at the moment of private preclusive use, when withdrawn from the Account. Whether or not to sell an asset can be left to the personal decisions of investors, for efficiency; how to tax the proceeds of investments can be left to the moment of consumption, when society can better judge what kind of lifestyle these investments enable.

Consider next the gift and estate tax. The current system in the United States at least aims to “backstop” the income tax, which tax is (in ideal theory) supposed to burden savings, by levying a hefty tax on those decedents who die with large estates or those persons making large inter vivos gifts. This tax is obviously desired as a matter of fairness. But its very existence encourages the rich to consume more, and die broke, whether they spend on themselves or their heirs.[31] In contrast, a consistent progressive postpaid consumption tax never taxes savings directly. Assets saved in the Trust Accounts thus have a zero “basis” in technical tax terms, meaning that they have not yet been taxed, and thus all proceeds from their sale or disposition are subject to tax, if and when consumed.[32] The Trust Accounts can therefore be passed on to heirs on life or at death, without the moment of transfer itself triggering tax. On the other hand, and at a different time, spending by the heirs will generate tax, and under a progressive rate structure, on withdrawal from the Accounts. A consistent progressive postpaid consumption tax does not need, in principle, a separate gift and estate tax, because the very design of the tax entails an accessions or inheritance tax—Trust Find Baby pays the progressive spending tax. Note, by the way, that intergenerational transfers, just as within-generation ones, can effect smoothing or shifting: parents can help to equalize spending across generations, or can self-sacrifice to allow their children to live better. In the latter, shifting, case, the familial burden will increase under a consistent progressive spending tax; in the former, smoothing, case, it will decrease. This pattern has normative appeal.

Finally, parallel─though, indeed stronger—arguments can be made against a separate corporate income tax. The problems with this tax begin with its uncertain incidence: since corporations are not real people, they do not really pay taxes. They must pass these on. A corporate tax falls on workers and consumers, on capital generally, or on some combination thereof.[33] To the extent it falls on ordinary workers and consumers, a corporate income tax’s claims to fairness are fairly obviously questionable. But even to the extent such a tax falls on capital, it cannot do so in any individuated way. Savers bear the burden of the corporate income tax whether they are rich or not, saving for lifetime needs or emergencies or to support a high-end lifestyle. Once again, under a consistent, progressive, postpaid consumption tax—which falls on the yield to capital as a source of personal consumption, making individuated judgments at that time—such a tax is not needed.

The elimination of these other taxes follows from the principle of a consistent progressive postpaid consumption tax: to tax individuals as they spend, not as they work, save, give, or die. Such a tax will enhance simplicity, transparency, and efficiency while promoting fairness. Specifically in terms of capital, the tax would apply to the yield to capital, when but only when it is appropriate to do so. The rich would not be let off the social hook; their tax would come due when, as, and if they spent wealth on themselves. Progressivity could be maintained, even strengthened.


An argument that supports direct capital taxation, manifest in a recent rise in scholarly reflection over a separate freestanding wealth tax,[34] pivots on the idea that capital itself—the mere possession of material resources, unspent—is a phenomenon that bears taxing. I have long written that this argument, as an objection to a consistent progressive postpaid consumption tax, is confused.[35]

To see why requires us to take a deeper, better look at what it is that might be troubling to the wider society in the stock of capital’s resting in private hands. Two potential problems come to mind.

One possibility is that the capital today represents potential consumption, or spending tomorrow: capital qua potential use. But the consistent progressive postpaid consumption tax, best understood, exerts a present lien on capital for any future use. Such a tax redefines property rights, what it means to be “private” and “public.” A taxpayer cannot use “her” capital without running through the gauntlet of the tax system; that tax system stands in the ready to exercise a greater toll on higher, less urgent private use. This argument rests on a confusion.

A second possibility is that capital today represents use, or power, today; there is the problem, as I have called it, of capital qua present use. This is a question of what one does with “his” capital. It is possible to save in a way that exercises power, or confers pleasure, as by buying up newspapers, sports franchises, or, for that matter, elected offices. But so understood, the response to this problem tracks the response to the first. Once again, the tax system defines or redefines property rights, what it means to be private or public. Problems with how capital is used are best met by regulation, as we have today in all advanced capitalist countries, for “private” endowments, pension plans, IRA accounts., and so on Simply forbidding taxpayers to use “their” money in their Trust Accounts to affect political outcomes without first withdrawing the funds from the Account and paying a tax will go a long way—and farther than any mechanism we have in place in the U.S. today—towards curbing the power of capital qua capital. And so this argument, too, is misplaced as a criticism of a consistent progressive postpaid consumption tax. In practice and in theory, such a tax creates a structure in which capital is, indeed, a common pool, and can be regulated as such—though it bears noting, again, that this is not the only or the best justification for the tax, Hobbes notwithstanding. It is, rather, a happy byproduct of doing the right thing—getting the fair timing of tax down right on an individual level.


Another argument for direct capital taxation, pressed particularly vis a vis the corporate income tax, bears noting. It is that these taxes are desired precisely because they are hidden. People do not notice the true incidence of the corporate tax, and this allows governments to have a higher revenue base than they otherwise might.

This may at first seem a weak argument to press in the name of fairness, resting as it does on trickery, and coming in the face of the near-certain regressive incidence of the tax. But recent research that I and others have conducted does indeed suggest that perhaps the best way to effect redistribution of material resources to the poor is to have large relatively flat taxes, accompanied by progressive redistributive expenditure programs.36 It is, after all, the net of tax and transfer programs that matters to any robust and compelling sense of distributive justice. So perhaps we want capital taxes, bad as these levies are, to get the money with which to effect social justice, on the spending side of the government’s tax and spend scheme.

In the end, this might be a compelling practical political argument for corporate taxation, though not for gift and estate taxes or capital taxes more generally, and though it bears noting that corporate taxes have been declining as a source of revenue in all advanced states, and may not, in the end, be worth the candle. But we, as philosophers and scholars, should know this argument for what it is, or label it as such. It bears noting that “hidden” taxes have real costs: the corporate income tax affects prices, distorts decisions, and effects no compelling distributive goal, once a suitably designed progressive spending tax is in place at the individual level. And so maintaining it, even to get funds for doing “good,” is not without cost; in a perfect world, we could generate all the revenues we need for just social spending programs by just social taxation schemes. In public finance as in life, we can pay a dear price for our illusions.


Advocates for fairness in taxation have long supported an income tax, precisely because it gets at the yield to capital, and because, they think, consumption taxes do not. In fact, a better understanding of the analytics of tax shows otherwise. Under progressive rates, the two canonical forms of consumption taxation, prepaid and postpaid, are not equivalent. An income tax is a double tax on all savings, come what may. A prepaid consumption or wage tax does indeed ignore the yield to capital, everywhere and anywhere. But a postpaid consumption tax splits the difference, by design. It falls on the yield to capital when but only when this yield is used to elevate lifestyles, not when used to smooth out in time, within or between generations, uneven labor market earnings.

It turns out that this is the right thing to do. Not only can we derive that from first principles, and the ordinary-savings and yield-to-capital norms, but we can also observe it from a century of practice with a so-called income tax. Whatever one thinks of ideal taxation, we ought to note well the fact that we have never had, and almost certainly never will ever have, an ideal income tax in practice, or anything rather too close to it, at all. The real debate in practical tax politics is and always has been over what form of consumption taxation to have. And here the stakes are large and dramatic for the fate of progressivity in tax, and point towards a consistent progressive postpaid consumption tax.

The final insight is that, once we have gotten the comprehensive tax system down right, from a strictly fairness point of view—by adopting a consistent progressive postpaid consumption tax—we no longer need any direct

taxes on capital. This is not because capital per se is good, or because of a naïve horizontal equity approach to policy. Rather it is because we are now taxing the yield to capital, in an individuated way, at the right time. We can and should repeal all capital taxes under the income tax, the separate gift and estate tax, and corporate taxes of all forms. This will add considerably to the simplicity, administrability, and efficiency of the tax system. But these have not been the point, here. It is, rather, the fair thing to do.


[1] I thank the editors of Social Philosophy & Policy, my research assistants Alex Baskin and Nina Kang, and all of the participants at a conference on Taxation, Economic Prosperity, and Distributive Justice for their helpful comments.

[2] Edward J. McCaffery, “The Uneasy Case for Wealth Transfer Taxes,” Yale Law Journal 104, no.2 (1994): 283-365. (hereafter cited as “The Uneasy Case”). This Article advocates the abolishment of the gift and estate tax and the establishment of a progressive consumption-without-estate tax.

[3] Walter J. Blum and Harry Kalven Jr., “The Uneasy Case for Progressive Taxation,” University of Chicago Law Review 19, no.3 (1952): 417-520 (later published as a book). Blum and Kalven found that most arguments for progressivity in tax burdens, such as those resting on the idea of the declining marginal utility of money income, are “uneasy” or unpersuasive; in the end, the case rests on little more than aesthetics and subjective value judgments.

[4] Arnold C. Harberger, “The Incidence of the Corporation Income Tax,” The Journal of Political Economy 70, no.3 (1962): 215-240; Martin Feldstein, “The Welfare Cost of Capital Income Taxation,” Journal of Political Economy 86, no. 2 (1978): S29-S51; Joseph E. Stiglitz, “Notes on Estate Taxes, Redistribution, and the Concept of Balanced Growth Path Incidence,” The Journal of Political Economy 86, no.2 (1978): S137-S150. The “Fair Tax,” a proposal for a progressive national sales tax, would also entail the elimination of all direct taxes on capital, including the gift and estate tax and corporate income taxes. See Laurence J. Kotlikoff, “The Case for the ‘Fair Tax,’” Wall Street Journal, 7 March 2005; see also

[5] See generally Barbara H. Fried, “Fairness and the Consumption Tax,” Stanford Law Review 44, no. 5 (1992): 961-1017.

[6] Thomas Hobbes, Leviathan (1651; New York: Penguin Classics, 1985); Barbara Fried, in “Consumption Tax,” argues against this Hobbesian position, calling it the “foundational argument”; see also Liam Murphy and Thomas Nagel, The Myth of Ownership: Taxes and Justice (New York: Oxford University Press, 2002) (hereafter The Myth of Ownership).

[7] John Stuart Mill, Principles of Political Economy, Books IV and V (1848; reprint, New York: Penguin Classics, 1985), chap. 2, 162-69.

[8] See Louis Kaplow, “Horizontal Equity: Measures in Search of Principle,” National Tax Journal 42, no.2 (1989): 139-154; but see Murphy and Nagel, The Myth of Ownerhsip; C. Eugene Steuerle, Contemporary U.S. Tax Policy (Washington, DC: Urban Institute Press, 2004).

[9] William Vickrey, Agenda for Progressive Taxation, (New York: The Ronald Press Company, 1947), 3.

[10] See Murphy and Nagel, The Myth of Ownership.

[11] Noel B. Cunningham, “The Taxation of Capital Income and the Choice of Tax Base,” Tax Law Review 52, no. 1 (1996): 17-44; Barbara Fried, “The Consumption Tax”; David A. Weisbach and Joseph Bankman, The Superiority of a Consumption Tax Over and Income Tax (draft on file with author), Murphy and Nagel, The Myth of Ownerhsip, Joel Slemrod and Jon Bakija, Taxing Ourselves 2nd Edition: A Citizen’s Guide to the Great Debate over Tax Reform. (Cambridge: The MIT Press,
2001). For a brief look abroad, to Canada, see David Duff, “Tax Policy in a Libertarian World: A Critical Review,” Canadian Journal of Law & Jurisprudence, 18, No. 1 (2005): 23-45, and to the developing world, see Richard Bird and Eric Zolt, “Redistribution in Developing Countries,” UCLA Law Review 52 no. 6 (2005): 1627.

[12 ] Henry C. Simons, Personal Income Taxation: The Definition of Income as a Problem of Fiscal Policy (Chicago: University of Chicago Press, 1938), 50.

[13] Edward J. McCaffery, Fair Not Flat: How to Make the Tax System Better and Simpler (Chicago: University of Chicago Press, 2002) (hereafter cited as Fair Not Flat).

[14] Mill, Principles of Political Economy, 165.

[15] There are two types of Individual Retirement Accounts (IRAs), tax-favored savings accounts in the U.S. for individuals’ retirement contributions: traditional and Roth. A traditional IRA is an account to which individuals can make tax-deductible contributions. Tax on earnings is deferred until withdrawal from the account at the age of at least 59 1/2. Contributions to a Roth IRA, on the other hand are not tax-deductible, but withdrawals are not taxed. For more information on traditional IRAs, see the Internal Revenue Code (“IRC”), 26 U.S.C. §408; for Roth IRAs, see IRC §408.

[16] Consider what happens to a principal sum, P, invested over time, for n periods, at a rate of return r. Untaxed, the sum grows at the rate of (1 + r), which gets compounded by the n periods. A single tax, t, is taken away from the taxpayer at one time, leaving her with (1 – t). Now it does not matter, under the commutative principle of multiplication, which holds that ab = ba, where, or, better put, when, the (1-t) is levied. Assuming constant t and r —assumptions to be discussed in the text—the following identity holds: {(1 – t) P} (1 + r) n = {P (1 + r) n }(1 – t).

[17] Since the future value (FV) of a present sum, P, invested over time at a rate of return r is P(1 + r) n , the present value of any fixed future sum can be obtained by dividing such future sum by (1 + r) n . Dividing Ant’s $110.25 of Year 2 dollars by (1 + .10)2 or 1.21 yields approximately $91 in start of Year 1 value.

[18] For a fuller discussion, see Edward J. McCaffery, “A New Understanding of Tax,” Michigan Law Review 103 no. 5 (2005): 825-26. (hereafter cited as “A New Understanding”). It may seem as if the need for constant inflation is a third assumption, but inflation is built into the interest rate; in the running example, if inflation were running at 10 percent per year, then Ant would be breaking even with inflation, before tax, but losing real value after it. See note 16 and accompanying text.

[19] William D. Andrews, “A Consumption-Type or Cash Flow Personal Income Tax,” Harvard Law Review 87 no. 6 (1974): 1113-1188.

[20 ] Alvin C. Warren Jr., “Fairness and a Consumption-Type or Cash Flow Personal Income Tax,” Harvard Law Review 88 no. 5 (1975): 931-46; see also Alvin C. Warren Jr., “Would a Consumption Tax Be Fairer Than an Income Tax?,” Yale Law Journal 89 no. 6 (1980): 1081-1124. See also Barbara Fried, “Ex ante/Ex post,” The Journal of Contemporary Legal Issues 13 no. 1 (2003) 123-160

[21] Consider for example the argument of Senator William E. Borah in the debates leading up to adoption of the 16th Amendment in 1913, permitting the modern income tax in the United States. Referring to the founding fathers, Senator Borah
claimed that “it was a republic they were building, where all men were to be equal and bear equally the burdens of government, and not an oligarchy, for that must a government be, which exempts property and wealth from all taxation.” Quoted in Steven R. Weisman, The Great Tax Wars (New York, Simon & Schuster, 2002) at page 224. See also Steven A. Bank, “The Progressive Consumption Tax Revisited,” Michigan Law Review101 no. 6 (2003): 2238-60.

[22] Carolyn C. Jones, “Class Tax to Mass Tax: The Role of Propaganda in the Expansion of the Income Tax During World War II,” Buffalo Law Review Fall (1988/1989): 685-737.

[23] For an extended discussion of these and other deviations from the United States’ income tax’s theoretical commitment to taxing savings, see McCaffery, “A New Understanding,” 885-907.

[24] See Edward J. McCaffery, “A Voluntary Tax? Revisited,” National Tax Association Papers and Proceedings (2000): 268-74.

[25] See McCaffery, Fair Not Flat, 33-34.

[26] IRC § 101 exempts the proceeds of life insurance from income taxation; IRC §m102 exempts gifts.

[27] See e.g., Laurence J. Kotlikoff & Avia Spivak, “The Family as an Incomplete Annuities Market,” Journal of Political Economy 89 no. 2 (1981): 372-391, Laurence J. Kotlikoff and Lawrence H. Summers, “The Role of Intergenerational Transfers in Aggregate Capital Accumulation,” Journal of Political Economy 89 no. 4 (1981): 706-732.

[28] My simple example ignores interest, which I discuss at greater length in McCaffery, “A New Understanding,” 864, note 159 and accompanying text

[29] See McCaffery, Fair Not Flat, 97-111.

[30] IRC § 213.

[31] See Stephan M. Pollan and Mark Levine, Die Broke (New York, N.Y: Harper-Collins, 1997). For my own thoughts on estate or wealth transfer taxation, see McCaffery, “The Uneasy Case”; Edward J. McCaffery, “The Political Liberal Case against the Estate Tax,” Philosophy & Public Affairs 23 no. 4 (1994): 281-312; McCaffery, Fair not Flat, 62-77.

[32] IRC § 1011 (basis); I define “basis” in McCaffery, Fair Not Flat, 161, as after-tax dollars.

[33] See generally sources cited above in note 4.

[34] See generally articles in Tax Law Review 53 no. 3 (2000): 263-498; Tax Law Review 53 no. 4 (2000): 499-696.

[35] McCaffery, “The Uneasy Case”; McCaffery, “The Political Liberal Case.”

[36] See Jonathan Baron and Edward J. McCaffery, “Unmasking Redistribution (Or its Absence),” in Edward J. McCaffery and Joel Slemrod, editors, Behavioral Public Finance (New York: Russell Sage Press, forthcoming); Bird and Zolt, “Redistribution in Developing Countries,” supra note 11.

Previously published by the University of Southern California Law School

Robert C. Packard Trustee Chair in Law and Political Science, USC Gould School of Law, Los Angeles, USA