Flatness and Fairness

The Flat Tax
by Robert Hall and Alvin Rabushka
Hoover Institution Press, 1995

In a 1961 speech to the Orange County Press Club, Ronald Reagan declared:

There can be no moral justification of the progressive tax. Perhaps that is why the bureaucrats pretend it is proportionate taxation. Proportionate taxation we would gladly accept on the theory that those better able to pay should remove some of the burden from those least able to pay. The Bible explains this in its instruction on tithing. We are told that we should give the Lord one tenth and if the Lord prospers us ten times as much, we should give ten times as much. But, under our progressive income tax, computing Caesar's share is a little different.

Twenty years later, Robert Hall and Alvin Rabushka, economists at Stanford University, took up where Reagan left off. In a 1981 Wall Street Journal article, expanded two years later into a book, they proposed replacing the federal income tax with what they somewhat misleadingly termed a "flat tax." The name and idea stuck, and their book -- revised and reissued in 1985 as The Flat Tax -- won a small but devoted following.

Until recently, however, no one believed that Hall and Rabushka's ideas were politically viable. Things have now changed. Steve Forbes's abortive one-issue presidential campaign put the flat tax on the map for the larger public. But Republicans have been working quietly for radical tax reform since the 1994 Republican Congressional sweep. Last July, House Majority Leader Dick Armey, a longtime committed flat taxer, submitted a bill to enact a flat tax modeled on Hall and Rabushka's. This past January, a Republican commission on tax reform, appointed by Dole and Gingrich and chaired by Jack Kemp, issued a report endorsing a flat tax. Short on details, Kemp's proposal also appears to be modeled on Hall and Rabushka's. Variants on the flat tax have been proposed, too, by Pete Domenici and Sam Nunn, and House Ways and Means Chairman Bill Archer. Archer has begun hearings on his plan -- to repeal the income tax entirely and replace it with a flat-rate sales tax -- and has announced his intention to bring the plan to a vote next year. Republican National Committee Chairman Haley Barbour has endorsed radical tax reform along flat-tax lines, careful to remain uncommitted among the Armey, Archer, and Nunn/Domenici plans. His position likely foreshadows the 1996 Republican platform. Bob Dole himself has remained cagey, saying only that he supports a "fairer, flatter, simpler tax system." Though much in his past record on tax policy suggests he would oppose key features of a flat tax, he probably does not care enough about those principles to hold out against a strong push from Armey and Archer, if they can garner broad public support.

It is too early to tell whether all this commotion will amount to anything. As sweeping statements of fiscal principle give way to detailed proposals, flat tax schemes may suffer the fate of the Clinton health care program and most other major policy reforms. Still, if Dole wins in 1996, a major overhaul of our federal income tax system along Hall-Rabushka lines has a credible chance of success. And considering Bill Clinton's well-attested "flexibility" on budgetary matters, the chance is not nil even with a Democratic victory. If enacted, a flat tax would represent (for better or worse) the most revolutionary change in the country's fiscal policy since the income tax came into general use in the 1910s.

Sensing this political opening, Hall and Rabushka have issued an expanded version of The Flat Tax. Written with the energy of true believers, it is the best available statement of the theory behind flat tax proposals.

Hall and Rabushka's flat tax plan, like similar proposals now before Congress, promises three fundamental changes in our federal income tax system. First, it would replace our current income tax with a tax on personal consumption. Second, it would tax all consumption above a set exemption level at a flat (or, to use Reagan's term, proportionate) rate. Each dollar of goods consumed by Bill Gates and a factory worker would be taxed at the same rate (say, 20 percent), irrespective of differences in annual income, consumption, or wealth. In contrast, our current progressive income tax system has taxpayers paying higher rates as their income rises. Third, it eliminates all personal deductions -- including deductions for child care, medical care, charitable contributions, state and local taxes, and home mortgage interest. These three changes are analytically distinct and politically separable. Congress could, for example, adopt a consumption tax without changing the present rate structure or personal deductions, or vice versa. This point is crucial to understand. Flat tax proponents tout the elimination of deductions as a way to ensure that the rich finally pay their "fair share," and this feature is arguably most responsible for the plan's enthusiastic reception among the middle class. But when flat taxers rally public support for flat rates or a consumption tax by highlighting the repeal of personal deductions that benefit the rich, they fly under false colors. There is no reason deductions that benefit the rich couldn't be eliminated within our current income-based tax system, with its progressive rate structure.

Separating the three changes, then, let's consider them in turn, beginning with the least well-known and most radical: the switch from an income to a consumption tax.

Taxing Consumption
A consumption tax taxes people on income not as they earn it, but only as they spend it on food, housing, entertainment, and other items of consumption. The easiest way to administer such a tax is by a sales tax, which taxes consumers at the moment of purchase on the value of their purchase (Bill Archer's proposal). The value-added tax (VAT), widely used in Europe, is a variant on that, measuring value added at each level of production rather than upon final sale. Because a sales tax or VAT is levied on purchases without regard to the identity of the purchaser, it is very difficult to adjust its burden to take into account the particular circumstances of taxpayers (family size, income level, etc.). To solve that problem, most consumption tax proposals, including the Hall-Rabushka plan, adopt instead a "cash-flow" tax. A cash-flow tax operates much like a huge IRA account for every person, but without any restrictions on the timing of withdrawals. Each person is taxed on her current income, with a deduction for any portion saved. When, down the road, she withdraws accumulated savings to consume them, she is taxed on every dollar she withdraws. Because any amount of current income we do not save we (by definition) consume, the effect is to tax annual consumption indirectly.

A right to deduct savings up front gives an enormous tax preference to savings and the investment income it produces. Under some assumptions, the preference is equivalent to exempting investment income from taxation entirely. This rough equivalence has led some analysts to refer to a consumption tax as a "wage tax" -- a tax entirely on wages and not at all on investment income. Under any set of assumptions, a consumption tax reduces the effective tax rate on investment income to a level far below that on wage income. Whereas Hall and Rabushka's plan limits the preference to capital invested after a flat tax is enacted, other plans extend it to existing investments as well.

Proposals for a consumption tax have been around for a long time. Proponents argue that it would increase savings and investment and simplify the tax law -- later, we will present some doubts about the magnitude of these effects. For most people, however, the downside is its distributional impact. Reducing the effective tax rate on capital naturally benefits savers, and in any given year, only those with high incomes save. Some savers are only temporarily in the top income bracket: middle-aged individuals at the height of their earning capacity, who are putting aside funds for retirement. Overall, though, a consumption tax disproportionately benefits those who are wealthy throughout their lifetimes. To counter this distributional impact, previous consumption tax proposals have generally coupled the shift to a consumption tax base with an increase in the progressivity of the rate structure, an increase in estate taxes, or both.

But Hall and Rabushka, along with all other flat tax proponents, have moved in the opposite direction. Under our current progressive rate structure, income above the exemption level is taxed at 5 rates, ranging from 15 percent to 39.6 percent. The Hall-Rabushka plan taxes all consumption above an exemption level at a single rate of 19 percent. While the Kemp commission report didn't specify a rate, Kemp himself has indicated support for a rate "around 19 percent." Dick Armey has adopted a 17 percent rate, to be phased in over three years. Hall and Rabushka also favor eliminating the estate tax. While they have not included that repeal in their proposal, Armey has incorporated it in his. How, then, do they address concerns about fairness?

The flat tax seems like it will disproportionately benefit the rich because it will, and handsomely so.

What About Fairness?
As Hall and Rabushka candidly admit, shifting to a consumption base with flat rates will be "a tremendous boon to the economic elite from the start." In the short run at least, the cost will be borne by a significant portion of the lower-middle or middle class (depending on precisely where exemption levels are set). In an appendix to the 1983 edition of their book (omitted from subsequent editions), Hall and Rabushka estimated that their flat tax proposal as then configured would increase the tax bill for the lowest income families by 78 percent, and decrease it for the very richest families by 41 percent.

Those figures are in line with a January 1996 report by the Treasury Department, which estimated that the current Armey bill, if modified to generate the same tax revenues as the current income tax (that is, to make it "revenue-neutral"), would increase the aggregate tax bill for family income groups below $200,000, and decrease it for groups above $200,000. If the bill were made revenue-neutral by raising the flat rate, the Department estimated that the tax liabilities for the bottom 80 percent will increase anywhere from 80.9 percent (for the poorest 20 percent) to 9.8 percent (for those in the 60-80 percent income group). The tax liabilities of the top five percent will decrease by 21 percent, and the top one percent by 36 percent. If the bill were made revenue-neutral by lowering exemption levels, the estimated increase in tax burdens for the poor, and decrease in tax burdens for the rich, will be even greater.

Effects of the Armey-Shelby Flat Tax (revenue neutral)

		20.8% Rate	20.8% Rate	17% Rate	17% Rate
		% change in after tax income	% change in total federal taxes	% change in after tax income	% change in total federal taxes
Family Income Quintile
Lowest		-6.7	80.9	-7.6	92
Second		-5.4	37.7	-7.2	50.7
Third		-2.9	13.7	-4.7	22.3
Fourth		-2.4	9.8	-3.8	15.1
Highest		2.1	-7.4	3.5	-12
Top 10%		4.2	-14	6.2	-20.7
Top 5%		6.4	-21.3	8.9	-29.5
Top 1%		11.7	-36.1	14.7	-45.2

Source: Department of the Treasury, Office of Tax Analysis.
20.8% rate reflects Armey-Shelby Plan adjusted to be revenue-neutral.
17% rate makes Armey-Shelby plan revenue-neutral by lowering tax exemption levels.
The second quintile begins at $15,600; the third at $29,717; the fourth at $48,660; and the highest at $79,056. The top 10% begins at $108,700; the top 5% begins at $145,400; and the top 1% at $349,450.

Proponents recognize the vulnerability of flat tax schemes to the charge of unfairly benefiting the rich, but are split on how best to respond. For most economists supporting the flat tax, the charge is exasperating not so much because it is false, but because it reflects this society's fetishistic "obsession" (Hall and Rabushka's word) with questions of fairness. But as Hall and Rabushka concede, they have learned over the past 15 years that the indifference of most economists to issues of distributive fairness will not play on a larger political stage. If flat tax proponents lose the fairness battles, they lose the legislative war.

Accordingly, most flat taxers have gone on the offensive, purporting to demonstrate "once and for all," as Hall and Rabushka put it in their new edition, that "the flat tax is the fairest tax of all." By what definition of fairness? Two answers are given, which, though often offered side-by-side, cannot be reconciled.

The most popular line of defense implicitly endorses the fairness of a progressive tax, but suggests that the flat tax will not alter the distribution of income and wealth, and indeed might even work out to be more progressive than our current tax system, with its nominally progressive rate structure. The explanation offered for this seemingly implausible result is that the current tax system is so riddled with tax preferences for the rich that they will end up being taxed at a higher effective tax rate under a 17 percent or 19 percent flat tax with no preferences than under our current system.

The argument, which has enjoyed surprising political appeal to date, is simply rubbish for the vast majority of wealthy taxpayers. Much can be said in favor of eliminating many of these provisions, although the case is less clear than flat tax proponents suggest (a point we will return to later). But the claim that the overwhelming majority of the wealthy -- those now in the 39.6 percent marginal bracket -- will be worse off giving up deductions and exclusions, in exchange for a 17 or 19 percent marginal tax rate on wage income, is simply false, as every distributional analysis of the flat tax (including Hall and Rabushka's) shows. No misleading appearances here: The flat tax seems like it will disproportionately benefit the rich because it will, and handsomely so.

For most of the first half of their book, Hall and Rabushka adhere to the party line, conjuring visions of multimillionaires legally avoiding any tax liability by exploiting loopholes in the current tax code, and who will finally be made to pay their fair share. But ultimately, they throw their lot in with the second, somewhat less popular, line of defense. Conceding that the wealthy will disproportionately benefit from shifting to a flat-rate consumption tax, they argue that such a result is only just, because progressive taxation is the moral equivalent of stealing from the rich.

Variants of this second argument are appearing with increasing frequency, often delivered with that air of tough-minded maturity -- "I'm grown up enough to speak the truth and you are finally grown up enough to hear it" -- in which attacks on affirmative action and other tenets of 1960s liberalism are now routinely couched. Hall and Rabushka, alone among current flat tax proponents, have undertaken to defend it. For that reason, if no other, their discussion in The Flat Tax bears scrutiny.

What is their case for a flat tax on fairness grounds? Like virtually all flat tax proponents, Hall and Rabushka discuss only the shift from progressive to proportional (flat) rates, ignoring the shift from an income to a consumption tax base -- a curious omission, given that arguments for the fairness of a consumption tax, while ultimately not compelling in our view, are weightier than any offered for flat rates.

Their case for proportional rates begins, inauspiciously enough, with the dictionary entry for "fairness." Anyone who seeks to resolve ethical controversy by heading to the dictionary is in trouble. But the choice of moral arbiter is only the beginning of Hall and Rabushka's problems. Favoring the definition of "fairness" as meaning "everyone should receive the same, or equal, treatment," they conclude that fairness requires a flat tax, in which every taxpayer pays the same percentage of his income, and not "any multiple-rate system that discriminates among different classes of taxpayers." The problems with this argument are obvious to anyone who has thought at all about equality-based norms. After all, if fairness means that we all should receive "the same, or equal, treatment" then we should not imprison murderers unless we imprison everyone. Suppose instead we say that fairness means that people who are the same in relevant respects should be treated equally: as applied to taxation, that people who are the same in relevant respects should bear the same tax burden. That injunction is unobjectionable. But it is also useless, because it tells us nothing about which respects are relevant. In the debate over progressivity, the central question is whether a sound tax system would regard differences in income levels or wealth as relevant in setting tax burdens. The same-or-equal-treatment principle doesn't answer that question. It merely begs it. The answer must lie in a deeper theory of justice that will inevitably provoke moral disagreements that cannot be resolved by recourse to the dictionary.

Hall and Rabushka show no interest in developing such a theory, and appear unaware of the monumental efforts of others to do so. What would a morally serious inquiry into the question of a just rate structure look like? Ironically, though almost any seriously entertained rate scheme can find support in one foundational theory of distributive justice or another, the one exception may be the choice flat tax proponents have fixed on: proportional rates. Let's consider the case for progressivity on fairness grounds, and then return to the case for a proportional tax.

Hall and Rabushka suggest that the "redistributionist approach to achieving tax fairness" is about a half a century old, the brainchild of bleeding-heart Democratic New Dealers, the arch villains of this book. In fact, its lineage traces to 15th century Florence, and its more recent expositors have included Karl Marx (a cite Hall and Rabushka would undoubtedly be happy to add), as well as Jeremy Bentham and Thomas Paine (ones that will please them somewhat less). Moreover, the lineage is far broader if one expands the inquiry to arguments supporting income redistribution through any combination of tax and transfer programs.

Four principal lines of argument have been advanced in support of progressivity -- though in this short space, we can present only the briefest sketches of them.

For many, the most compelling argument for progressivity begins from the egalitarian premise that all people are endowed with a basic right to be treated as equals. But many of the differences in our social positions, including differences in income and wealth, arise from morally arbitrary aspects of the natural endowments we are born with or the social circumstances we are born into -- differences that do not redound to our credit or discredit. A just state, committed to treating its members as equals, is obliged to reduce those differences -- to ensure that our fates in life are not fixed by conditions that are so adventitious from a moral point of view. To be sure, people disagree on how much the differences are to be reduced, as well as on the relevant dimensions to be equalized. And progressive tax rates are not strictly necessary to any egalitarian scheme for income redistribution. One could, for example, employ high exemption levels and nominally flat rates above that level, and use much of the revenue collected to finance cash and in-kind transfers to the less well-off. But a progressive rate structure is a likely feature of most egalitarian schemes.

A second line of argument has supported progressively on utilitarian or social welfarist grounds -- that is, on the premise that public policy should promote the aggregate well-being of society. The starting point for all such arguments is what economists call the "declining marginal utility of income": the wealthier you are already, the less additional happiness (utility) one more dollar will bring. The intuition behind that insight is straightforward: If you have $100 to give away, the money can be expected to do more good for a single mother struggling to feed two small children than for Ross Perot.

Originating in Bentham, this insight was formalized in modern neoclassical economics by (among others) Francis Edgeworth, Thomas Carver, and A.C. Pigou. Its implications are radical indeed. Assuming that individual welfare is a function of income, it suggests that (all else being equal) we will maximize aggregate welfare by taxing the better-off exclusively until we have leveled their wealth down to the next income group, then taxing those two groups until we have leveled them down to the third, and so forth, until we have equalized remaining income.

Of course, all else is not equal, as utilitarian proponents of progressivity have always recognized. Willingness to work and save is responsive to expected rewards. Tax rates reduce those expected rewards, and if rates are set too high, they will create strong disincentives to productive activity among the well-off, thus shrinking the pie for everyone and defeating efforts to improve overall social welfare. The optimal combination of tax rates and transfers, then, is a complicated empirical question, to which there is no certain answer. It is generally assumed, however, that the optimal rate structure on utilitarian grounds would be progressive.

A third line of thought defends progressivity as a way for the community to claim its rightful share of each individual's earnings -- rightful, because the value of what we produce depends not only on individual effort, but also on the entire social, economic, and legal infrastructure in which we labor. A fourth defends it on the ground that huge disparities in wealth, whatever their sources, corrode our social and political lives, undermining the sense of commonality and equal membership required to sustain a democracy.

All these arguments have difficulties. But they are not, as Hall and Rabushka and other flat taxers suggest, merely a cover for a "soak the rich ideology" or "the politics of envy."

What of the flat tax? "`[P]roportionality is a principle,'" declared F.A. Hayek 35 years ago, "`but progression is simply hateful arbitrariness.'" But exactly what principle is it? Shifting from dictionary definitions to tautologies, Hall and Rabushka answer: "The principle of equity embodied in the flat tax is that every taxpayer pays taxes in direct proportion to his income." Virtually all defenses of the fairness of a flat tax reduce to this tautology -- which is, in any case, the wrong tautology, since under a flat tax people pay taxes in direct proportion to consumption, not income.

Still, the intuition that a proportional tax represents a fair distribution of collective fiscal burdens is an ancient one, reflected in the biblical doctrine that each person should tithe (that is, pay a tenth of their income) to the church. More recently, a proportional tax has appealed to diverse scholars, from John Rawls to Milton Friedman, and appears to be on the verge of capturing that elusive commodity, the popular imagination.

Notwithstanding its ancient pedigree and recent popularity, the idea that everyone should pay the same percentage of their income in taxes is a mystifying one, difficult to ground in any coherent theory of fairness. The philosophically most compelling case for it has come from libertarians, and there is ample reason to believe that the true affections of many of the flat taxers lie here. (Hall and Rabushka, for example, routinely describe tax collection as coercion; they come close to describing taxation as theft.)

Libertarians begin with the assumption that we have a natural property right, derived from our Lockean rights to our own person, to whatever the market will pay us for our labor or capital. Based on that premise, they conclude that government may not take any of those market earnings without our express consent. Most libertarians make a limited exception for taxes raised to finance the costs of a "minimal state" (national security, highways, etc.), on the ground that we cannot rely on voluntary, unanimous agreements -- that is, on the free market -- to produce these goods at their optimal level. But if that is the sole rationale for taxation, it suggests (as most libertarians believe) that tax rates should equal the market value of the benefits each individual derives from government services -- just what they would have to pay if there were a functioning private market for such services. This so-called "benefits taxation" leads to a proportional tax only on the highly implausible assumption (a "not clearly inappropriate assumption" is the best Milton Friedman can do) that people use government services in proportion to their income. Assuming that in pricing its services the government is limited to charging the cost of providing them, and may not extract the surplus value the wealthy enjoy from those services, the far more likely answer is a highly regressive tax, with poorer citizens paying a much higher proportion of their income for government services than others.

If no coherent philosophical premise implies a flat tax, what accounts for its popularity? A suggestion that emerges from between the lines is that supporters like Hall and Rabushka believe that it is fair (as they state) that "[t]hose who earn more should pay more," but not as much more as liberal Democrats seem to think; and that the "more" extracted under a flat-rate tax is a reasonable compromise. That reading is powerfully borne out by the fact that all flat-rate tax proposals in circulation, including the Hall-Rabushka plan, are not really proportional taxes at all. As noted above, they all exempt a certain amount of family consumption. In the Hall-Rabushka plan, exemption amounts range from $9,500 for a single person to $25,500 for a family of four. Only consumption above the exemption level is taxed at a flat rate. As Hall and Rabushka repeatedly note, such tax schemes (sometimes referred to as "degressive") are progressive in effect. The overall tax rate for any taxpayer reflects the 0 percent rate on the exemption amount and the flat rate on the excess, and therefore rises as consumption increases. Suppose the exemption level is $20,000 and the rate is 20 percent. Then, for consumption of $15,000, the overall tax rate is 0 percent; for consumption of $25,000, the overall rate is 4 percent. And for consumption of $1,000,000, the overall rate is 19.6 percent.

Despite their attacks on progressivity, Hall and Rabushka, like virtually all other flat tax proponents, trumpet the actual progressivity of their flat tax, comparing it favorably to a sales tax or VAT for that reason. What they fail to see is that their solicitude for the poor expressed in the form of high exemption levels, like Reagan's solicitude in his 1961 speech, undercuts any fairness case they might have against progressive rates above the exemption level. If we have an obligation to those less well-off than us -- if, in Reagan's words, "those better able to pay should remove some of the burden from those least able to pay" -- and if that obligation is sufficient to justify requiring us to pick up the poor's entire share of the tab for running the government, then why stop there?

Perhaps the best answer one can give is some version of the one suggested above -- that something approximating the real configuration of a flat tax represents a reasonable compromise among different, warring goals: protecting the poor absolutely; equalizing wealth to some extent across all income brackets; but leaving most of the income of the rich untouched, because they deserve to keep most of it and/or they will not work as hard if we don't let them keep it. There's nothing wrong a priori with such a non-ideal compromise. But if this is the rationale, proportional rates above an exemption amount are at best an arbitrary compromise with no independent moral weight. Thus, they invite the same attacks made by Hayek and echoed by Hall and Rabushka on the "hateful arbitrariness" of any particular progressive rate structure. Why not 18 percent on the first $10,000 of income over an exemption level, 19 percent on the next $10,000, and 20 percent on everything thereafter? Or, for that matter, why not 15 percent, 28 percent, 31 percent, 36 percent, and 39.6 percent on additional increments (our current rate scheme)?

Eliminate Deductions?
The final feature of most flat tax plans, including Hall and Rabushka's, is to eliminate all personal deductions and credits beyond a standard per capita exemption. Historically, hostility to deductions has been a liberal cause, championed most notably by the late Harvard Law Professor Stanley Surrey. Hall and Rabushka and other flat taxers take up Surrey's crusade, but with a twist. The flat tax, as we noted earlier, dramatically expands deductions for business investments. So only so-called "personal" deductions and credits are left to face the ax: child care and earned income credits, charitable and medical deductions, and so on.

This feature of flat tax proposals has its virtues. Deductions and credits increase the complexity of the tax system, and sometimes distort taxpayers' incentives in undesirable directions. But flat taxers' Draconian response runs into the same difficulties Surrey faced three decades ago. Many deductions, such as the medical deduction, arguably help reach a fair definition of disposable income that should be subject to tax. Still others correct for other distortions introduced by the government, or encourage socially desirable behavior. Proponents of eliminating all deductions have always opposed using the tax system for this sort of "social engineering," to use Hall and Rabushka's disdainful phrase, but their opposition is hard to fathom. All tax systems are forms of social engineering. For example, the favorable treatment of savings under the flat tax is designed to engineer greater investment. The trick is to figure out what incentives we want to create, and whether they are best created through the tax system or other means.

The existing tax credit for childcare expenses illustrates all these difficulties. Arguing that the credit "fails to focus its benefits in an area of particular social need," Hall and Rabushka conclude that "the special problems of helping families with child care and other responsibilities should be attacked specifically within the welfare system, not with the scatter-gun of the tax system." In fact, the childcare allowance is pretty well designed. It counteracts disincentives built into our tax system (and the flat tax) for both parents to work outside the home. It is a credit rather than a deduction, and drops as income rises, thereby helping most those who need it most. It reflects the coherent and widely shared view that providing for children is a societal and not purely personal obligation (a view reflected in Hall and Rabushka's own high exemption levels for dependents). Finally, if the government is going to help working parents with children, there is much to be said for doing it through the tax system rather than the "generous welfare system" that Hall and Rabushka propose as an alternative. Either way, it costs the government the same -- in forgone revenues (through a tax credit), or explicit appropriations (through welfare). But a tax credit gives parents autonomy to choose the sorts of care they want, and it is administratively cheap, compared to transfer payments or in-kind provision of childcare services. In any event, Hall and Rabushka's reliance on the welfare system to pick up the slack is hopelessly unrealistic. A Congress that enacts a flat tax is not about to expand the welfare system or subsidize childcare. The authors must know (if not agree with) this political reality, lending an air of disingenuousness to the entire proposal.

Arguments of comparable weight support many other deductions and credits that would be eliminated under the flat tax. Even those tax preferences that would be almost impossible to justify if we were starting from scratch -- for example, the mortgage interest deduction and other preferences for homeowners -- may be difficult to repeal altogether without imposing hardship on those who have invested in reliance on them. To be sure, lots of particular cases require subtle judgment. But what seems clear is that the issue is far more complicated than flat tax proponents think, and that a blanket repeal of all deductions and credits is a bad idea.

Economic Benefits
In the end, proponents' arguments for the fairness of a consumption-based/flat-rate/no-personal-deduction tax system are unlikely to persuade anyone who does not already share their peculiar, undefended notions of distributive justice. But any perceived unfairness must ultimately be weighed against the economic benefits of the flat tax, to which we now turn.

The standard argument for the flat tax is that it will generate economic gains that will ultimately leave everyone a winner (although the wealthy bigger winners than anyone else). The bulk of The Flat Tax is, not surprisingly, devoted to this argument.

Some of the projected gains from the flat tax are entirely unsubstantiated -- for example, those Hall and Rabushka anticipate from reduced tax evasion. Still others, like reducing the costs of complying with the federal income tax (time spent in record keeping and filing, costs of advice, etc.) are real but almost certainly exaggerated. Any tax system can be made to sound simple if we describe only its broad principles. The problems arise in translating those broad principles into positive law. Within weeks of Armey's announcing his plan, smart tax lawyers had figured out how to manipulate the business tax portion of his flat tax to produce a negative rate of tax on capital investments. This particular loophole can probably be closed, but only at the cost of complicating the business tax. And this is just the start. Joel Slemrod, an expert on compliance costs, estimates relatively modest savings from a VAT, in the range of 0.6 percent of gross domestic product. The Hall-Rabushka cash-flow tax is significantly more complicated to administer than a VAT, and hence will produce far lower savings in compliance costs.

A slightly stronger case can be made for economic gains from increased savings and work effort under a flat tax. One leading economist has estimated an increase in welfare equal to (depending on age cohort) a 0.1 to 1.5 percent increase in gross domestic product. Changing the proposal to compensate taxpayers disadvantaged by the switch to a flat tax (something for which there will be significant political pressure) would forfeit much of that gain.

Thus, reasonable projections put the aggregate economic gain from a cash-flow flat tax at between 1 and 2 percent of gross domestic product. Not chopped liver, but hardly sufficient to explain the obsessive zeal with which supply-siders have pursued tax reform. As Paul Krugman recently pointed out, comparable gains can likely be realized through any number of other reform measures (easing traffic congestion, rationalizing the allocation of scarce water or scarce airwaves, for example) that have failed to arouse any interest at all from most flat taxers. Moreover, most of that gain could be achieved through longstanding proposals to reform our existing progressive income tax system. Simply eliminating the current, unjustifiably higher tax rate levied on corporate income than other business income, for example, is projected to achieve significant economic gains, with negligible distributional consequences. Other substantial gains could be achieved by phasing out housing and other unjustifiable preferences without severe economic dislocations, reducing compliance burdens through electronic filing and consolidating federal and state filings, and taxing fringe benefits by denying businesses a deduction for them -- the last a sensible part of Hall and Rabushka's plan that does not require a flat tax to administer.

None of this is likely to faze flat tax supporters, who believe that equity alone demands drastically lower tax rates on the wealthy, quite apart from effects on economic efficiency and growth. But for those who do not share their sense of justice, the at-best modest economic gains from a flat tax are unlikely to seem worth its distributional cost.

Another Agenda?
We have focused thus far on the Hall-Rabushka plan. The flat tax plans currently before Congress raise all of the issues it poses, plus an additional one -- the impact of a flat tax on the federal budget. Hall and Rabushka set out to design a revenue-neutral tax. To their credit, the best guess of impartial observers is that they succeeded. The same cannot be said for their Congressional allies. Armey's version of the flat tax lowers the tax rate to 17 percent (phased in over 3 years), while raising exemption levels significantly above those in the Hall-Rabushka plan. The two changes together are projected to reduce current tax revenues substantially -- almost $200 billion a year for the original version of the Armey proposal released in 1994, $138 billion a year for the revised bill he submitted in July 1995.

The most charitable explanation for the budgetary discrepancy in the Armey bill is that backers believe the sunniest supply-side projections of economic growth under a flat tax. But, taking past experience as guide, those optimistic projections will prove wrong, and it will be impossible to fund government programs at their present levels. The most likely candidates for spending cuts will be redistributive social programs, thus adding to the overall burden of tax reform on the poorer segments of our society. It's hard to avoid the conclusion that flat tax proponents are content to be optimists because the latter outcome does not particularly disturb them. Most proponents of the flat tax, including Hall and Rabushka, also favor shrinking the public sector drastically. Given this other agenda, a call for additional spending cuts to fund flat tax reform, should that come to pass, is (as one friendly critic put it) "more of an opportunity than a problem."

We have been here before. In 1985, Senator Moynihan described Reagan-era tax cuts as "a deliberate decision to create deficits for strategic, political purposes." Flat tax proponents are entitled to their views about the ideal size of the public sector. But for those who do not share them, caveat emptor.