the liberal case against progressive taxation

In the afterglow of election the Bush administration, clearly feeling its oats, hinted publicly at a historic shift in national tax policy: the abolition of the graduated income tax in favor of a flat income tax or, possibly, a national sales tax. It has lately backed away from those hints, suggesting instead a tax reform that would radically cut or even eliminate taxes on savings and investment—presumably by eliminating taxes on capital gains and dividends and possibly by expanding tax breaks for business investment. Liberals are breathing a reflexive sigh of relief that the graduated income tax is out of danger. They should not.

In truth, the President’s new proposal—made to appear deceptively modest by the flat-tax gambit— may result in a more regressive and inequitable tax system than flat-tax advocates ever dreamed of. And if the Democrats are to thwart this move, they must do the unthinkable: call the President’s bluff, abandon a century of tax tradition, and embrace a modified flat tax.

Despite its name, the progressive income tax as we know it poorly reflects progressive values, and indeed cannot be justified in terms of those values. A truly progressive tax system, in the political sense, could and should accommodate a flat tax for the great majority of taxpayers even while serving two core progressive principles: protecting the poor and preventing accumulations of great wealth that threaten the integrity of the democratic political process. Such a regime also offers a politically viable alternative to the deeply regressive reforms proposed by President Bush.

Attacking Progressivity: Efficiency and Fairness

The graduated income tax has been with us, on and off, since 1862, and has been a foundation of progressive governance since the Wilson administration. Challenges to it have been anathema to Democrats for ninety years. Challenges there have been, though, and strong ones: the Republican Party included anti-progressivity planks in its platforms in at least five successful Presidential campaigns since 1920. Only one of those— in 1984—explicitly called for a national flat tax, but the implication has always been clear. What’s remarkable is the consistency of the arguments made by opponents of progressivity over that period. Here are excerpts from two documents calling for tax reform:

A sound [tax] policy equally demands the . . . reduction of the tax burden which may be achieved by substituting simple for complex tax laws . . . prompt and certain determination of the tax liability for delay and uncertainty; tax laws which do not, for tax laws which do, excessively mulct the consumer or needlessly repress enterprise and thrift.

Instead of being simple, the current tax system is needlessly complex, making it susceptible to abusive tax avoidance schemes. Instead of being efficient, it punishes hard work, discourages savings and investment, and hinders the international competitiveness of U.S. firms. Instead of being fair, it is out of line with our basic values and undermines our sense of fairness. Instead of being predictable, it is highly unpredictable and uncertain.

The first is from President Harding’s campaign, in 1920. The second is from President Bush’s campaign, in 2004. The political case against the progressive income tax, that is, remains essentially unchanged since the early years of Prohibition. Progressive taxation is too complex; it breeds uncertainty; it discourages economically desirable behaviors like savings and investment; it harms the economy by suppressing growth; and it is, in some deep and moral way, unfair.

Note that there are two very different kinds of arguments here. The first four are arguments about economic efficiency. The progressive income tax leads to economically inefficient results—in the sense of economic outcomes that are distorted relative to a hypothetical tax-free world— because of its effect on two key variables: costs and incentives. The enormous technical complexity of the tax code imposes high compliance costs on every taxpayer and therefore on the national economy itself: although there is no universally agreed method for measuring tax compliance costs, academic estimates suggest that those costs run to about 10-15% of total tax collections—over $100 billion a year. Uncertainty also imposes costs, measured in both taxpayer anxiety and over- and under-payment, which require additional effort and costs to correct. Progressive income taxation also creates perverse incentives, or so the argument goes: by taxing income instead of consumption it encourages the latter, to the detriment of national investment and savings. By raising rates as income rises progressivity also is thought to reduce desirable incentives to productive economic behaviors: why work an extra week if your return on that work drops because of increased marginal taxation?

These economic arguments turn out to be surprisingly weak. With regard to complexity, no one can argue that the tax system is anything but phenomenally complex or that the costs of that complexity are real and substantial. It is much less clear, however, that progressivity is responsible for very much of that complexity. Certainly some of the revenue code’s labyrinth of rules are designed to capture strategies for avoiding progressive rates, like interpersonal wealth transfers, but a great deal more results from the inevitable temptations of political pork, as favored constituencies receive deductions, credits, phase-outs, and exemptions from Congress. Richard Epstein and others have argued that progressivity is to blame for those temptations, but it is hard to understand why: there are manifold ways to favor a constituency other than simply altering its marginal rate, including deductions, exemptions, credits, and phase-outs—all of which would be equally available in the flat-tax world Epstein favors. Uncertainty, too, seems to be less a function of graduated rates than of the public-choice problems inherent in legislative taxation: the determination of the actual rate of taxation on income is perhaps the easiest part of tax compliance.

The strongest economic argument against progressivity put forth by its opponents is usually that, as Epstein has put it, “higher marginal brackets mean less production from the most productive people” in society. Graduated rates, in this view, punish productivity by discouraging economically valuable behavior. This is intuitively plausible; the trouble is that there isn’t much evidence that it’s actually true. A series of recent studies of the behavioral effects of progressive taxation demonstrated that “real behaviors, such as labor and savings, seem much less affected by taxation” than conservative economists suppose. The reason is in part that high-income Americans “are unlikely to decrease hours worked as tax rates go up—either because they find the work inherently satisfying . . . or because they focus more on maximizing their after-tax income.” People work for all kinds of reasons, that is, of which income maximization is only one—and possibly not the most important one. Tax rates of themselves appear not to be sufficient to create the incentive problems conservatives expect, at least not at the moderate marginal rates currently in vogue. (We might very well find a different answer at very high rates, like the 91% top rate that applied in the Eisenhower administration, but such rates seem a political impossibility today.)

The economic arguments against progressivity, then, are underwhelming. In any case, they are all grounded on a dubious premise: that efficiency and wealth maximization are the sole, or at least the most important, values served by a national revenue policy. That has never been true, historically. Taxation serves a number of broad social purposes, one of which is, and has always been, equity. We not only want a bigger pie; we also want to divide the pie in a way that comports with our understanding of fairness. This is why the conservatives’ fairness-based arguments against progressivity are far the strongest, both politically and as a matter of policy. If the economic arguments against progressivity are easily countered, the fairness arguments are much less so. If progressive taxation has a future in the United States, it is because a case can be made that it is fair—fairer, in the end, than nonprogressive alternatives.

The core of the fairness critique is simple and powerful. Even if we accept—and everyone, even the most conservative tax critic, does—that wealthy people should pay more in taxes than poor people, that principle doesn’t require progressive taxation. It requires only proportional taxation, such as that imposed by a flat tax. I pay, say, 10% of my income; Larry Ellison pays 10% of his income; and all other things being equal, Larry Ellison pays vastly more of the national tax bill than I do. Progressive taxation requires another, larger leap: that wealthy people should pay more, not only in raw dollars, but also in the proportion of income they pay. If we assume that Larry Ellison’s income is 1,000 times mine—it’s probably more than that—a progressive scheme requires that Ellison contribute not 1,000 times what I contribute but, say, 1,300 times.

Critics of progressivity argue that this simply isn’t fair, by any conventional understanding of equity. The wealthy already pay more; by what rationale do we ask them to pay more than more? It’s a good question. In other contexts we generally equate fairness with proportionality. Corporate shareholders’ “fair share” of corporate profits is distributed according to their proportional holdings. Joint ventures assess cash calls to their members according to proportion of membership interests. Partners in a partnership are liable for their proportional share of its outstanding debts. Order pizza with your friends and you’re unlikely to charge them different amounts based on their ability to pay. It’s hard to think of any circumstance, in fact, in which fairness is commonly thought to require progressivity.

This explains, I think, the intuitive appeal of the fairness critique. At the same time, however, the argument for progressivity is also grounded in notions of fairness—so much so that Robert McIntyre, writing in Washington Monthly only a year ago, could describe the progressive income tax of 1916 as “a huge victory for tax fairness” without explaining why. Perhaps he didn’t need to: perhaps modern liberals support progressivity reflexively, because the redistributive impulse it represents is rooted deeply in the constellation of liberal values. But the truth is that all taxation is redistributive to some extent, and the Republican platform suggests that the presumptive fairness of progressive taxation is no longer a given. If progressive income taxation is to survive, arguments must be made to defend it. At the end of the day, those arguments aren’t adequate to justify the system we have today.

The “Uneasy Case” for Progressive Taxation

Nearly every federal income tax ever passed by Congress, from 1862 until today—and every one since 1913— has featured a graduated rate structure in which wealthier people pay more, not just in raw dollars but in the percentage of their income that they pay. It’s worth asking how such a policy ever came about in the first place, because it seems counterintuitive: why is it that our income tax structures require more than a simple proportional—that is, flat— levy?

A common answer to that question relies on an economic concept known as diminishing marginal utility. The more of something you have, the theory says, the less each additional unit of that something is worth to you. If you buy candy at the movies, you’re much more likely to give some to your friend if you have twenty M&M’s than if you have just two. Applying that theory to income, some economists believe that as my wealth increases, the value I place on each additional dollar of income falls. Thus, to equalize my tax burden as my income rises, my effective tax rate must also rise to keep pace. A flat 10% levy on a teacher’s $30,000 salary in fact imposes a much greater burden on the teacher than would a similar levy on a lawyer’s $300,000 salary, because that $3000 (assuming a drastically simplified taxation scenario) is in fact worth more to the teacher than the $30,000 is to the lawyer.

There’s an intuitive appeal to this, and it’s clearly true at the baseline: the dollars we spend on absolute necessities—food, housing, medical care—are worth more to us than the dollars we spend on BMWs and Burberry topcoats; we know this because we would never trade necessities for luxuries in a voluntary exchange. This doesn’t prove much, though, because every tax reform currently being considered—except the national sales tax—contemplates an exemption for low incomes, so that the dollars used for true necessities, narrowly defined, aren’t really subjected to income taxation at all.

Whether the diminishing marginal utility of money applies at higher income levels is anybody’s guess. Such things are inherently difficult to measure, and even if we could establish conclusively that in general the wealthy value their additional income less than do the middle class, we would still be faced with substantial interpersonal variations—that is, there’s probably a lot of people out for there for whom it isn’t true. For those people, progressive taxation looks profoundly unfair, if our justification for it is diminishing marginal utility theory. And there’s some persuasive evidence—which I’ll return to a moment—that the very wealthy actually place more value on marginal income increases than other people.

There’s a bigger problem, too. Even if we were confident that the utility of money falls as income rises, that’s really the wrong question. Progressive taxation is only justifiable if the rate at which the utility of money drops is greater than the rate at which tax rates rise. This is an unprovable proposition; we may think it intuitively plausible, but as I’ve suggested already, intuitive judgments no longer seem sufficient to sustain the policy commitment to progressivity. Better reasons are required.

In any case, American income taxes have been (mostly) progressive since 1862; the theory of diminishing marginal utility did not achieve wide distribution even among economists until the publication of William Jevons’s Theory of Political Economy in 1871. Why, then, did Lincoln’s administration choose to impose not just an income tax but a progressive income tax? How did we end up with a progressive income tax in the first place?

Historically, advocates have offered four basic justifications for progressivity in addition to diminishing marginal utility theory. The first, commonly cited by proponents of Lincoln’s income tax, was that taxation should be based on the taxpayer’s ability to pay. In The Wealth of Nations, Adam Smith’s first canon of taxation states that “the subjects of every state ought to contribute towards the support of the government, as nearly as possible, in proportion to their respective abilities; that is, in proportion to the revenue which they respectively enjoy under the protection of the state.”

Smith is a fearsome authority for economic principles, of course (although Lincoln was no free-marketeer), but what’s interesting here is that Smith’s canon doesn’t really explain the progressive feature of the Revenue Act of 1862. Ability to pay simply requires that those who have more, pay more. Smith’s law suggests not progressivity, but proportionality. The ability to pay theory, by itself, can’t justify the progressive income tax.

The other three major justifications for progressive taxation arose in the run-up to the ratification, in 1913, of the Sixteenth Amendment to the Constitution, which amended Article I, Section 9 of the Constitution to overrule the Supreme Court’s 1895 decision striking down the income tax on constitutional grounds. All three are grounded in ideas about fairness; but each represents a very different approach to what fairness means, and why we might choose one kind of fairness over another.

The first was a simple fairness argument couched in terms of proportionality. Congressman Clyde Howard Tavenner, for example, a Democrat from Illinois, won office in 1912 on a platform calling for a progressive income tax that would “require the rich to bear a fair proportion of the burden of taxation, which they are not now doing.” The 1913 income tax had a threshold exemption of $4,000 for married couples, which meant that less than 2% of American households owed any tax. Rates began at 1% and rose progressively thereafter to 7% for the very wealthiest taxpayers. Progressivity, that is, applied only to the highest incomes. That system was “proportional,” to people like Rep. Tavenner, because the rest of the populace was already bearing an enormous portion of the tax burden in the form of the high prices produced by the labyrinth of protective tariffs then in effect. Progressivity, that is, was “fair” because it helped to counterbalance an existing revenue program that was systematically regressive. Although the modern revenue system still features significant regressive elements—notoriously in the form of payroll taxes—those elements no longer require a progressive income tax to achieve a measure of horizontal equity among taxpayers.

The second argument falls under the general rubric of “benefit theory,” and it has roots in the more communitarian Progressive thinkers of the early twentieth century—men like Robert Lee Hale and Henry Carter Adams. Benefit theory holds that the wealthy benefit disproportionately from society’s collective activities—protection of private property, national defense, rule of law, stable fiscal policies—all of which, it is argued, make possible the circumstances that allow individuals and business associations to earn and accumulate wealth. Those who benefit most should pay most. This was a common trope among early Progressives, including Theodore Roosevelt, who in 1906 defended progressive taxation by arguing that “the man of great wealth owes a peculiar obligation to the state, because he derives special advantages from the mere existence of government.” This view retains some intellectual force—Cass Sunstein is one of its prominent intellectual heirs, and it is a useful corrective to the facile possessive-individualist view that dominates much American thinking today—but it is not clear why benefit theory requires anything more than proportional, rather than progressive, taxation. We really have no sound basis for believing that, even if the wealthy do benefit more from public goods, they do so at a disproportionate rate.

The last major argument for progressive taxation, and the most powerful, is that redistribution of great wealth it effects is necessary to the functioning of the democratic state. Professor Avi-Yonah of the University of Michigan Law School has argued that the progressive taxation of wealth—forcing the wealthy to pay not just more but more than more—is a necessary foundation of political democracy. Because vast wealth creates vast social and political power; and because “the modern democratic state is built on the assumptions that all power should ultimately reside in the people and their representatives,” Avi-Yonah says, redistributive taxation of that wealth is crucial to preserving a fair political process.

Avi-Yonah’s argument is an expression of a long tradition of Progressive antipathy toward great accumulations of power, of which the progressive income tax and the Sherman and Clayton antitrust acts are products. Both progressive taxation and trust-busting were part of a broader effort to achieve fairness—not in the narrow context of taxation and business law themselves but through the use of taxation and business law for the advancement of more fundamental democratic political principles.

Those political principles, Avi-Yonah argues, are again endangered, as they were in Teddy Roosevelt’s America, by concentrations of wealth and power. “The belief in progressive income taxation,” he concludes, “must ultimately rest on the same conviction that animated the reformers of the 1890s: that extreme concentrations of wealth in the hands of private individuals . . . is an unhealthy phenomenon in a democracy.”

I agree, and I hope that most Americans do, too. But notice that even Avi-Yonah’s argument-from-democracy doesn’t really justify the graduated income taxation as it exists today. It only justifies higher tax rates for a very limited number of individuals and giant corporations whose wealth can plausibly be thought to distort the democratic process. The income tax we have today imposes all its progressive features on income between zero and $319,100 for married couples. All income over that substantial but hardly overwhelming figure—from $400,000 to $2 million to $20 million and beyond—is taxed at 35%. The vast majority of wealthy individuals’ income, that is, is taxed at a flat rate.

If we accept the democratic approach to tax fairness it is very hard to understand why modest rises in income for a two-earner family earning, say, $120,000 a year—a family that, given current household debt loads, is probably not much further from bankruptcy than its neighbor who earns half that much—result in higher tax rates, but massive increases in income for the nation’s 2.27 million millionaires do not.


Achieving Tax Fairness: A Modified Flat-Tax Proposal

The history of the progressive income tax is a history of political approaches to fairness. Everyone believes that the tax system should be fair, but fairness is not a monolithic concept. Affirmative action, for example, is fair from one point of view; color-blindness is equally fair from another. The real question is what deeper value we use to choose the fairness principles that guide our tax regime. Choosing among values is a hopeless task—unless a value choice exists that is necessary to the existence of some other value, the importance of which is universally agreed. Here, I believe that Avi-Yonah’s argument from democracy provides such a value. Taxation is a fundamental element of government, and our system of government requires a level of political and social equality that is threatened by extreme poverty and great wealth. The tax code is one means by which democratic government can create the conditions requisite to its own existence.

What would such a tax code look like? It would have four basic features: a high exemption for low-income families to prevent unfair burdens on the poor; a flat-rate, that is, proportional, tax on the vast majority of the citizenry; a high progressive rate that kicks in only at levels where wealth begins to confer disproportionate social and political power on its owners; and a steep inheritance tax on large estates to prevent excessive intergenerational wealth accumulation.

Remember that in such a system the modestly wealthy still pay more—much more—than their lower-income cousins, and only the very, very wealthy are forced to pay more than more. The truly poor pay nothing—and indeed might receive federal tax credits to offset the regressive effects of state-level sales and excise taxes. The overall effect would be to shift wealth from the wealthiest to the poorest, while seeking rough equality of treatment for those similarly situated—the middle and upper middle classes who are neither disenfranchised by poverty nor imbued with disproportionate power by wealth.

President Bush’s plan would have precisely the opposite effect. Because the vast majority of savings and investment income belongs to the wealthy—and disproportionately to the very wealthy—the effect of eliminating taxes on capital gains and dividends would be to reduce, substantially, the proportion of the national tax burden owed by the wealthy. It represents the equivalent of reducing the marginal tax rate on very high incomes to below the level paid by average incomes. If, as President Bush has repeatedly said, any such reform would be revenue-neutral, the decrease in tax receipts would have to be made up by additional taxes, explicit or otherwise, on the middle class and the poor. Advocates of the President’s plan, of course, deny this, claiming that eliminating taxes on savings and investment will produce such a rush of economic activity that overall tax receipts would actually rise. There is little evidence for this, however, and ultimately the reform would produce little but modest economic growth and a massive wealth transfer from working people to the leisure class.

That’s an acceptable result to conservatives, because they take a very different philosophical approach to tax policy—one that is entirely geared to overall wealth maximization. If a tax regime results in an increase in the size of the pie, they think it’s irrelevant that the size of the slices becomes more and more inequitable. That is their choice, of course, but if we believe—as I do—that great disparities of wealth threaten not just general notions of fairness but the viability of our political system itself, then a tax system that encourages such disparities is simply unacceptable.

There was a time when such views were politically powerful. In 1913 Speaker of the House Champ Clark said that the progressive income tax would

introduce into the taxing system of this country that we will tax what a man has instead of what he has to buy to live on . . . I believe that if you had a secret vote, 95 per cent of the people of the United States would vote for the [progressive] income-tax proposition in this bill . . . . If you gentlemen on the Republican side ever get into possession of the Government again you will not more dare to repeal it than you would attempt to jump off the top of the Capitol. The people will not permit it.

Speaker Clark may have been right, but the progressivity in force today bears no resemblance to the progressivity of Mr. Clark’s tax, which exempted all but the wealthy and imposed its graduated rates only on very high incomes. Such a tax structure is impossible today—we have no tariff to fill the national coffers—but we could return to a model in which the progressive features of the income tax are aimed at the truly wealthy, rather than the middle class, and in which the poor are generously exempted from paying income tax at all. The exemption can no longer apply to the middle and upper-middle classes—the nation cannot afford it—but a flat tax for that great swath of American taxpayers would inject a much-needed dose of fairness into the tax system without endangering liberal principles. Do liberals have the courage to try it? We shall see.