Center for Strategic Tax Reform (CSTR) CSTR Logo


Tax Reform

Savings, Retirement and Social Security Reform

Contributing Members Commentaries


How Much Simplification is Enough? Is a Returnless Tax Realistic?

Tax Notes
December 23, 1996
By Ernest S. Christian

Ernest S. Christian was the deputy assistant secretary (tax policy) of the Treasury Department in the early 1970s, and is counsel to the Center For Strategic Tax Reform, Washington, D.C. He is the author of a soon-to-be published book, The American Tax Revolution: Seeking a Higher Order of Integrity and Competence in Government.


In this report, the author worries that fundamental tax restructuring may be defeated by its own unrealistic expectations and overblown rhetoric; suggests that simplification, so often the rhetorical touchstone, is an ancillary product of a tax system that is neutral as between persons, neutral as between labor and capital, neutral as between saved and consumed income, and neutral internationally; and asserts that such a neutral system does not in any way involve resort to radical untested ideas or experimentation, but rather can be achieved by a handful of familiar, long-sought amendments to the current code, nearly all of which already have substantial constituencies in and outside of the 105th Congress.


I. Introduction: Great Expectations

II. Definitions and Overview

III. The Search for a Returnless Tax System

A. The Corporate Income Tax
B. The X-Tax in Two Variations
C. The Flat Tax
D. The USA Tax
E. The Corporate Model Plus Direct Withholding
F. A Schedular System Plus Universal Withholding
G. IRS-Prepared Tax Return
H. The Sales Tax Option

IV. Conjectures and Conclusions


The movement to restructure the American tax system may now have reached an impasse from which it must back up and start again. If so, it is not because the proponents of the current code have prevailed. (The Internal Revenue Code of 1986 has few defenders.) Rather, it is because of some tax reconstructionists' own great expectations, not all of which are necessarily realistic.

In the realistic category, we all want and can expect to achieve a tax system that is fair, even though there are now at least four definitions of fairness: the one that insists on taxing upper income people at a higher rate than others; the one that allows Americans a fair opportunity to save and invest; the one that equates tax fairness with a low rate of tax; and the one most widely applied by the American people to whom "fair tax" may mean "less tax." We can also expect a tax system that is internationally competitive. (The IRC unnecessarily hampers our ability to compete in global markets. We are playing the game with one hand tied behind our back.) We can also expect a tax system that sticks to its knitting and does a good job of the task we have assigned it. (The IRC seems more concerned with altering our behavior than with collecting taxes in an efficient and evenhanded manner.) On top of all these achievements, we can insist upon a much simpler tax. (At present, figuring out how much tax we owe is often more painful than paying it.)

One of the leading contenders does a pretty good job of meeting these basic expectations. The USA Tax preserves "progressivity"; eliminates the double tax on saving and investment; levels the international playing field; and makes few distinctions based on behavior.(1)

On the simplification front, it provides a much smaller, more understandable tax code. Its tax return is of modest proportions compared to the monstrosity of today. The flat tax promises to simplify even more and to reduce the tax return to the size of a postcard.(2) Many experts dispute the reality of that claim, but even if larger than advertised, the flat tax return would be exquisitely simple compared to Form 1040. The flat tax would also drastically reduce the top tax rate. An adept blending of the USA Tax and the flat tax could produce a result that by ordinary standards would be a dream come true: a relatively simple, internationally competitive tax system, with lower rates and no double tax on savings and investment, one that neither adds to the deficit nor increases taxes.

Some tax reconstructionists also expect to eliminate personal tax returns altogether and, in the process, the IRS "as we know it." When imbued with the characteristics of ultimate simplicity and zero bureaucracy, the idea of a returnless system is hard to resist. With a little imagination, it conjures up a highly seductive vision: The IRS is an automaton that, once set in motion, quietly goes about its assigned task. Almost unnoticed, the robot annually collects the right amount of tax from the private economy and transfers it to the Treasury's coffers for public use. There are no hassles with the IRS. Being only an obedient servant and having no agenda of its own, the robot never meddles in the personal and financial affairs of its masters. With the robot in charge, congress is not constantly changing the tax up or down and playing favorites as it decreases tax for some people at the expense of others. (If we could then reengineer the expenditures side of government to operate in the same way, the grand vision would be complete: a portion of what we earn would automatically go into the Treasury account and would automatically come back out to pay for what we want and expect government to provide.)

A returnless tax system may or may not be realistic but, flights of fancy aside, it is certainly realistic to expect that government can be an obedient and competent servant, in taxes as well as in all else. (If not, we are in serious trouble indeed.) Most would-be architects of a new tax system for America's future incorporate into their designs a neutral, evenhanded tax base. In a tax system with few, if any, distinctions among persons or types and uses of income, there is little role for the IRS other than as an honest and competent accountant.

Another expectation common to most would-be architects seems to be that if we could just come up with the "right" tax design and agree on it, we could then freeze it in place (immune from the continuing vicissitudes of politics and other disruptive intrusions) and everyone would live happily ever after. To guard against destructive change, tax alternatives as diverse as those advocated by House Majority Leader Richard Armey, R-Tex., and Minority Leader Richard Gephardt, D-Mo., would include special procedural constraints on future congresses, such as a supermajority or approval by a plebiscite to increase tax rates.(3) One of the strongest advocates of fundamental reform, House Ways and Means Committee Chairman Bill Archer, R-Tex., has called for repeal of the Sixteenth Amendment in order to prevent the future regrowth of any tax system even remotely resembling the federal income tax of today.

Some critics might suggest that special procedural constraints on future revisions would not be necessary if the tax design were "right" to start with, but this kind of carping adds little by way of enlightenment. (A correct tax structure is not rendered incorrect by the addition of a supermajority requirement.) Further, those who would impose special procedural requirements are likely to have in mind something more worthy than trying to muzzle complaints about genuine defects in their tax design. They are likely to be most concerned about the demonstrated proclivity of congresses and presidents, on their own initiative, to mess up something that is already quite satisfactory.(4) Supermajorities are no more a substitute for a correctly designed tax system than budget resolutions are a substitute for responsible execution of the public trust, but their effectiveness (or lack thereof) is not the main point. Their greater significance is in recognizing that congress is a political institution and that politics is about always making things "better." After having enacted an ideal tax system, if such a thing exists, is it realistic to expect that all future elected officials would then be content to stand aside and be a "do nothing" congress insofar as taxes are concerned? Probably not, but anything that encourages them to keep their hands off is worth a try.

To some reconstructionists, the ideal tax system is one with a neutral, evenhanded base and no tax return. We all know that a returnless tax system, in combination with all the other goals of tax restructuring, is a tall order. But why dismiss it out of hand? Why not fully explore the possibility? Some tax reconstructionists will not be satisfied until the feasibility of such a tax system is either proved or disproved. Moreover, dedicated and sincere advocates of fundamental tax restructuring are not the only ones who have raised the standard of "simplicity" to potentially impossible heights. Many opponents of tax restructuring have used extreme definitions of "simplicity" as a way to subvert tax reform.(5) Therefore, it is in everyone's interest, except the opponents of tax restructuring, either to achieve the ultimate simplicity of a returnless system or, if that proves problematic, to revert back to a realistic target where the simplicity quotient of the new tax system is compared to current law, not to zero.

The whole process of tax restructuring involves balancing many conflicting considerations and goals. The no return vs. simple return question is no exception to that rule. Indeed, it goes to the core of two related issues which run throughout this exposition: how much simplification is enough and at what point does the continued pursuit of simplification become counterproductive?


Short of some kind of automatic tax-debit mechanism in some future age where all transactions are in electronic "money," there would appear to be no such thing as a returnless tax system in the purest sense. But a returnless system in that sense is probably not what its advocates have in mind. Instead, we will assume that their goal is to collect tax without individuals having to participate in the process other than as the ultimate payers of tax. (Individuals would, in this sense, be tax payers but not return filers.) Therefore, given the practical necessity for some degree of accounting for tax as between payers (people) and the collector (IRS), some intermediary would be required. As a practical matter, businesses are the available intermediary to "administer" the tax on the private sector side. /6/ It is they who would file returns and deal with the IRS under a tax system which relieved individuals of that responsibility. Broadly speaking, businesses are the source of all the income we receive either as wages and salaries or as interest and dividends, or, if we are self-employed, the equivalent payments we make to ourselves out of our own businesses. Thus, if tax is not to be accounted for by, and collected from, the individual who receives the income, the alternative is to account for and collect the tax at the source, from the business which pays out the income in the first place. If one wishes to think in terms of taxing "consumption" (the expenditure of income after receipt), businesses could serve as intermediaries in that case as well. (Businesses produce and sell everything we purchase when we expend our income.) Given that businesses already serve as private-sector administrators of a vast array of governmental regulations and programs, further concentrating the entire burden of private-sector tax administration on them is not ideal, but there would appear to be no way around it, if individuals are to file no personal tax return at all.(7)

Needless to say, all other considerations aside, tens of millions of Americans would be ecstatic to be relieved of the annual agony of April 15. If, in addition, their taxes were not increased, it would for them be the Declaration of Independence and the Emancipation Proclamation all over again. If the new tax system were also efficient and widely perceived as fair, there would be no way to measure the magnitude of the accomplishment and public gratitude that would ensue. Only H & R Block would be truly unhappy, but tax reformers are unlikely to shed many tears over the plight of professional tax return preparers. Businesses would have mixed emotions if the new tax system added greatly to their burden as private-sector administrators for government, but we can probably assume that tax complexity at the business level is not the primary concern of those tax reconstuctionists who seek to dispense with tax returns. It is people, not businesses, who are thought to be near revolt at having to spend so much of their time and energy on complex tax returns. Form 1040 also requires people to disclose to the IRS the details of their financial affairs and much about their personal lives as well. (The extent to which the broad swath of Americans are offended by this invasion of their privacy is unclear, but some tax reconstructionists are greatly offended as a matter of principle and one suspects that their search for a returnless tax system may be motivated as much by concerns about civil liberties as anything else.)

Even at its best, however, the idea of shifting the entire return-filing responsibility to businesses does not necessarily mean that all individuals would be free from dealing with the IRS or free from having to account to the IRS for what is actually "their" income. For starters, millions of self-employed people in the agricultural and services sectors would come within the "business" category and would still have to file tax returns. Elsewhere throughout the economy, the smaller the firm, the more the business tax return would become the owner's tax return under another name, and the more those individuals would continue to be directly involved with the IRS. Only pure employees (as distinguished from owner- employees) and people who derive their income from passive investments in large enterprises could count on being totally free from tax returns. For all these reasons, it is important that the new tax system be as simple as possible. As ordinarily defined, a "simple" tax is one with few moving parts, i.e., one with a readily ascertainable base to which the rate schedule is applied and which, at most, makes only a few distinctions among taxpayers, types of income, or uses of income. By this definition, several of the business-level taxes discussed later on in this exposition are quite simple. If no other issues are involved, it is not difficult to design a simple tax system where only businesses file tax returns.

Using the same ordinary definition of simplicity, a tax can be simple even though individuals are also required to file a tax return. Later on, when we examine the flat tax in detail, it will be seen as well within the ordinary definition even though a personal tax return is required. When properly understood and viewed objectively, the USA Tax will also be seen as within the outer boundaries of the ordinary definition of simplicity. The availability of these and other return-type alternative tax systems further emphasizes that those who seek to free Americans from the necessity of filing any personal tax return at all have goals in mind that go beyond the desire to eliminate the absurd complexities and wastefully high compliance costs associated with current law. We have already referred to the civil liberties concerns on the part of those reformers who object to the personal and financial disclosures that are inherent in a tax return, no matter how simple. The absence of a personal tax return may also be thought to be a self-enforcing constraint on government's ability to use the tax system as a regulatory device. It is true that, without a tax return, it is more difficult to create special dispensations (i.e., deductions, credits, etc.) that government can either withhold or allow depending upon the behavior or characteristics of the taxpayer. But to say that the absence of a tax return is the reason for the absence of such special dispensations is a little bit of a "chicken and egg" proposition. (Knowing in advance the constraint imposed by a returnless system, congress would enact such a system only if congress had already decided, for other reasons, to enact an evenhanded, nonregulatory tax.) Strictly speaking, about the most that can be said is that, once having been enacted, the returnless system would make it much harder for some future congress to revert back to a behavior- modification tax system such as exists today, especially if the freedom from filing personal tax returns proved to have the enduring public appeal its proponents predict.

Another reason for seeking to eliminate personal tax returns may be more tactical than ideological. Some proponents apparently believe that capturing the public's imagination at the outset with such a dramatic step would make it easier to gain enactment of the "good economics" common to nearly all the leading alternative tax systems. In this respect, they may be right -- assuming, of course, that it is possible to find an economically sensible returnless tax system that does not have other characteristics that render it unenactable or, if enactable, unstable.


It is now time to sort through some structural options to see if any of them (either "as is" or somewhat reconfigured) might be a suitable candidate. To meet all of the great expectations of all advocates in the current tax restructuring debate, it appears that the new tax system would have to meet all of the following requirements, some of which are subject to differing interpretations.

1. It must relieve individuals of the obligation to file a "tax return." In general, this means that individuals must not be required to tote up their income, calculate the tax thereon, and file the results with the IRS. Short of that, does identifying oneself to the IRS and filing some minimal information with it constitute a prohibited "tax return"? Some might say so, but, at some point, rigid insistence on zero contact with the IRS and total anonymity becomes a quibble without substance.

2. It must be simple. (Because a returnless tax system would by definition be simple for individuals and should be simple for businesses as well, this requirement has primarily to do with the category of "nearly returnless" systems and with whether an exquisitely simple tax should be considered returnless even though something is filed with the IRS.)

3. Its top marginal rate must not be too high. At one point in our history, a tax rate over 50 percent was high, but the threshold dropped to 40 percent some time ago and, at current, the range of primary debate is probably between 25 and 35 percent, depending largely on the base of the tax and other factors that make generalization difficult. (Even a rate less than 20 percent may be too high if strictly applied to everyone without exception.)

4. It must be perceived as fair, if not by everyone, at least by a large majority. Some might favor a highly progressive multiple-rate tax. Others might say that only a strictly proportional single-rate tax is fair. But even the flat tax is graduated to some extent. Therefore, we shall assume that some degree of graduation is necessary, even if the tax has a single rate. (At a minimum, there must be some means of shielding the poorest among us from the full impact of the tax.)

5. It must eliminate the double tax on saving and investment. In the case of a business, this means expensing of capital investment. To prevent double-taxation of personal saving (if individuals are separately taxed), individuals must either be allowed to deduct saved income or, if no deduction is allowed, the earnings on previously taxed saved income must be excluded from further tax.

6. It must be evenhanded as between labor income and capital income. In general, this means that both types of income must be taxed alike. (Fully equal treatment of all income is, however, difficult because of the existing FICA payroll tax on most wage income.(8))

7. It must be internationally competitive. In general, this means that the tax must be border adjustable for exports and imports, and that it must be territorial, instead of worldwide, in its application. Some proponents of tax restructuring may consider "territoriality" alone to be sufficient, but, as a practical matter, both international components are necessary.(9)

8. It must replace the current federal income tax, but in doing so, it must not cause either "transitional" dislocations or windfalls of great magnitude.

All or most of the substantive requirements (3. through 8.) can be met by return-type systems, all of which are simpler than current law and some of which are "nearly returnless." The question is whether it is realistic to expect all or most of the desirable substantive characteristics in combination with the total absence of a personal tax return. Eight possible constructs, some with tax returns of varying degrees of significance, are considered below.


If the only goal were to eliminate personal tax returns, and there were no other constraints, there is an easy way to do the job: repeal the personal income tax and apply the current corporate income tax to all unincorporated businesses as well. Collecting all tax from "corporations" (as broadly defined) would certainly eliminate personal tax returns, but if we then began to take other considerations into account, significant adjustments in this model would be necessary to prevent violation of other criteria for a properly restructured tax system. Even then, this primitive corporate model would fall short of the ideal, but let us not dismiss it altogether. It has definite possibilities and is obviously representative of the general idea of a returnless tax system.

The narrow "capital-only" base of the current corporate income tax, where both investment and the earnings on equity investment are taxed (but returns to labor and returns to debt capital are not) can be cured. The tax base could be broadened and made uniform by eliminating the current deductions for interest on debt and for employee compensation. (10) With returns to labor and returns to both equity and debt capital included in the corporate tax base, the corporate model would then function to collect (and pay over to the IRS) tax at the source on all forms of income received by individuals. (11) As a general proposition, with this broader base, a corporate tax rate of roughly 20 percent would be sufficient to replace the revenues produced by the federal income tax of today.(12) Because of its base, the corporate model could also be border-adjusted for exports and imports, thereby permitting it to be made territorial as well.(13)

Nevertheless, even as amended, this "returnless" broad-based corporate model has two major defects. First, with all tax being collected and accounted for solely at the business level, it provides no means by which to mitigate the strict proportionality of the implicit flat rate of tax on all forms and amounts of income received by individuals who are shareholders, debtholders or employees. Second, because it requires businesses to pay over to the IRS a total amount of tax equal to the entire amount of the current individual and corporate income taxes combined (about $750 billion in 1995), it would impose ruinously large cash flow drains on nearly all businesses. Compared to the amount of the current corporate income tax (about $150 billion in 1995), there would be a fivefold increase in cash outflows from businesses to the IRS. These two seemingly inconsistent consequences (on the one hand, an implicit proportional tax on individuals and, on the other, a devastatingly large increase in business tax payments) are the result of the "incidence" theory most likely to be used by economists to determine who bears the economic burden of the tax, as distinguished from who initially pays over the tax to the IRS. Corporations pay the corporate tax and suffer the immediate cash flow consequences thereof, but people ultimately bear the economic burden.(14)

The prevailing and most logical view appears to be that a tax like the broad-based corporate model should be attributed to people's labor and capital factor incomes, i.e., to the corporation's employees, shareholders and debtholders. That is the case despite all the current rhetoric about "consumption taxation," a term that appears to have little to do with the ultimate incidence of a tax paid by a corporate intermediary. For example, an economist might refer to the broad-based corporate model as a "consumption tax," not because it is thought to impose any extra-normal burden on anyone's expenditures for goods and services in the marketplace, but, rather, because it allows businesses to expense capital investment. Therefore, in this instance, the economist would be referring to the base of the tax and, for whatever it is worth, only in the very narrow sense of the National Income and Product Accounts (NIPA) where, broadly speaking, if one reduces gross domestic product by gross private investment, the residual is personal consumption expenditures.(15)

For most of our tax history under the current system, Congress has largely ignored the incidence of the corporate tax. (Some might even say that Congress has managed to "hide" a good portion of the current tax burden inside the corporate wrapper.) But it is necessarily the case that "tax incidence" would be of dominant importance in the future under any new returnless tax system where some intermediary, such as corporations, would pay all tax -- in form and in legal effect on its own account but, in economic reality, on someone else's behalf. Congress would want to know who it is taxing and in what proportion, by income level and otherwise. Therein lies the rub. The tax in the broad-based corporate model is strictly proportional, without any graduation whatsoever and without any exceptions for people at the bottom of the income scale.

The basic idea that people bear the economic burden of the broad-based corporate tax, roughly in proportion to their labor and capital factor incomes, is hard to argue with. It is consistent with the fact that corporations have no capacity either to produce income or to pay tax apart from their employees, shareholders, and other capital-providers. It is consistent with the widely held view that the employer-paid payroll tax is borne by employees. It is consistent with the traditional view that the current corporate income tax (with its capital-only base) is borne by equity shareholders, and it is consistent with the newer view that the current corporate tax is borne by both the employees and the owners.(16) Further, the fact that corporate taxes may to some extent be reflected in some combination of higher prices and a lesser volume of goods and services being produced and sold is not inconsistent with attributing the tax under the broad-based corporate model to labor and capital incomes.

On the other hand, from the immediate perspective of the corporations whose tax payments would be increased by many multiples under the "returnless" broad-based corporate model, it is not reasonable to expect that the factor-income adjustments implied by incidence theory would occur immediately and without friction. They would not and the potential for severe dislocations in the meantime is great. In the long run, the corporate model would serve to precollect individual taxes at the source and, in that sense, can be analogized to withholding of income tax on wages, dividends, and interest, but the implicit tax imposed by the corporate model on factor incomes is not the same as withholding in all respects. For one thing, the transitional cash flow effects are quite different.(17)

Those who have trod this ground before have sought to go beyond the broad-based corporate model. For the most part, they have used it as the starting point for a simple two-level tax with very few moving parts where initial "payments" of tax are made by both individuals and businesses. A two-level tax facilitates the introduction of some degree of graduation in the tax borne by individuals, but obviously reinvolves individuals to some extent with tax returns or, at least, with something similar to a tax return. Most of these two-level approaches could be considered "nearly returnless" insofar as individuals are concerned, although one of them, the USA Tax, involves a full-fledged tax return. For completeness and to illustrate additional considerations bearing upon the search for a returnless tax system, several of these two-level options are discussed next, under the general rubric of David Bradford's renowned X-Tax.(18)


In its pure or root form, the X-Tax is the broad-based corporate model. (All tax is collected at the business level, a deduction is allowed for capital investment, but no deductions are allowed for wages, interest, or dividends, and, therefore, a uniform and proportional tax is precollected on all returns to labor and capital.) But the X-Tax was first proposed at a time in our tax history when highly "progressive" taxes were de rigeur and for that reason, if no other, the X-Tax immediately departed from its roots in the corporate model and first appeared as two variations, both designed to introduce an element of progressivity into what would otherwise be a proportional tax. The first variation fails to solve the transitional cash flow problem inherent in the corporate model, but the second variation solves that difficulty as well. Both variations are nearly returnless and are very simple by ordinary standards.


The first variation would maintain intact at the business level the full labor and capital base of the corporate model, but would use a refundable credit mechanism to ameliorate the implicit flat rate of tax on individuals. For example, with an assumed 20 percent X-Tax rate at the business level, an employee who received a $100 wage from a business would be assumed to have prepaid an implicit flat-rate tax of $20, but the employee would, for example, be allowed a 15 percent (or $15) refundable credit.(19) In that case, net of the refund, the tax rate on this employee would be only 5 percent. If the level of the credit phased down from 15 to 10, 5, and 0 percent as income increased, a substantial degree of progressivity would be achieved. As proposed by David Bradford, the refundable credit applied only to wage income, and, in that sense, the refund-type X-Tax variation embodied the principle of a "schedular" tax system (i.e., wage income is assumed to be associated with lower-income people for whom progressivity is thought by some to be most important, and capital income is assumed to be associated with higher-income people for whom a flat rate is more acceptable, provided it is higher than the rate on most wage income). The refundable credit could, however, also be applied to the combined total of a person's wage income and capital income, although to do so would introduce more complication.

Being graduated, nearly returnless, unbiased against saving, and potentially border adjustable to boot, the refund-type X-Tax represents a high level of achievement in many respects.(20) Employees would still have to file something with the IRS disclosing the amount and source of their wage income (and family size if that were a factor) to claim and receive the refund.(21) But who could really object to the disclosure of basic wage information that the IRS and the Social Security Administration already know in connection with the FICA payroll tax? Besides, the IRS could use the information only to compute and pay a refund. No tax increases or other hassles with the IRS would be involved. Insofar as capital income and one's financial affairs are concerned, individuals would make no disclosure to the IRS at all. As a general proposition, all tax on interest, dividends, etc. would have been precollected and prepaid at the business level at a flat rate of 20 percent.(22)

On the negative side of the refund-type X-Tax, the idea of the IRS making cash distributions to nearly everyone in America would be a major problem. For one thing, it calls to mind, on a much grander scale, the experience under current law with the earned income tax credit which has been afflicted with both fraud and mistake. For another, universal cash payments by the IRS may be too reminiscent of a negative income tax to suit a congress already struggling with other entitlements and still going through the trauma of having reformed welfare. Worse, a tax system premised on the idea of collecting all tax from corporations and redistributing a portion of the take to individuals might, rationality aside, create a constituency for higher "corporate" taxes. (The higher the flat-rate tax on General Motors and its capital owners, the higher the cash "refunds" to most of its rank-and-file employees.)

It is one thing, in designing a sensible tax system, to recognize that people do ultimately bear the burden of corporate taxes. Going so far as to make regular cash "refunds" to people for taxes paid by corporate intermediaries is, however, quite another matter. It is bad civics. When we pay a tax directly to the IRS and then receive a check back from the IRS, the connection between the tax we paid and the refund we received is very clear. We know that we are not net better off than we would have been had we not been required to pay the tax in the first place. On the other hand, when we receive a refund check for taxes paid by a corporation, the connection between the tax detriment suffered and refund benefit received is easily lost. We are likely to see ourselves as net better off, as the result of government generosity. Even if we are told that the IRS check represents a portion of taxes paid by the company for which we work, we are likely to think of it as our government- provided share of the company's profits.(23)


The next X-Tax variation directly modifies the broad-based corporate model by excluding returns to labor from its base. Like the current corporate income tax, payments of compensation to employees would be deductible. Having been deducted at the business level, compensation would be directly included in the personal income tax return of the recipient. On the other hand, interest and dividends would not be deductible and, as under the broad-based corporate model, tax would continue to be precollected at the business level on returns to capital. This bifurcated approach (where businesses "pay" everyone's tax on capital income and individuals "pay" tax on their own wage income) alleviates the transitional cash flow drain on businesses which is otherwise inherent in the corporate model. It also permits multiple and progressive tax rates to be applied to labor income while at the same time avoiding the problems associated with the "refunds" in the other X-Tax variation. For example, under the bifurcated X-Tax, assuming a 20 percent business tax rate (and an implicit 20 percent rate on personal income in the form of interest and dividends), the personal tax rates on labor income might be 0 percent, 5 percent, 10 percent, 15 percent and 20 percent. (The basic idea would be to have personal rates on labor income that progress up to, but not beyond, the implicit flat rate on capital income.)

The bifurcated X-Tax has many positive aspects. Even though the bifurcated X-Tax would "rearrange" the tax base among individuals and businesses, the aggregate base would remain the same as under the corporate model and, therefore, the same as under the refund-type X- Tax. From a simplification perspective, the individual-level tax is the equivalent of a simple, multiple-rate gross receipts tax on wage income. If enacted exactly as proposed and if it remained that way, the bifurcated X-Tax would be extremely simple and "nearly returnless."

On the negative side, the bifurcated X-Tax illustrates how attempts to "rearrange" the labor-capital base of the corporate model to serve some goals of tax restructuring can compromise other goals and/or cause political problems. For example, bifurcating the tax base by taxing wage income directly to employees conflicts with the desire of many tax reconstructionists to have border tax adjustments for exports and imports.(24) Taxing wage income directly to employees also opens the door to specialized personal deductions against wage income which would not be available against capital income.(25) Returnless advocates would probably fear that once having gotten a foot in the door with a few personal deductions, a highly regulatory tax system comparable to current law would quickly reemerge. The other part of the bifurcated base -- taxing capital income at the business level -- also currents difficulties. For example, one might wonder why, after having modified the corporate model to make direct tax payers out of employees, the bifurcated X- Tax did not also allow a business-level deduction for interest and dividends and, thereby, make direct tax payers out of capital owners as well. The most obvious reason for leaving capital owners in the indirect category is to be able to use the simple "yield-exemption" approach to eliminating the double tax on saved income.(26) Otherwise, it would be necessary to use the more complicated "deduction" approach found in the USA Tax. All else being equal, the simplest approach is the best, but as illustrated by real-life experience with the similarly bifurcated flat tax, leaving capital owners in the indirect category gives the politically awkward misimpression that only wage earners are paying taxes.


Despite having adopted the bifurcated X-Tax structure that was designed to permit labor income to be taxed at progressive rates, the Hall-Rabushka flat tax made prominent by House Majority Leader Armey reimposes that flat rate at the personal level on wage income. Viewed solely in light of the preceding discussion of the bifurcated X-Tax and its particular rationale, the single rate under the flat tax seems especially curious. After having departed from the corporate model, where all income is taxed at the business level at the same flat rate and where no personal tax return is necessary, and having recreated the necessity for a personal tax return on wage income, why turn around and reimpose the identical flat rate on wage income? Reimposing the single flat rate at the personal level might seem even more curious given that, after having done so, the flat tax then introduces a substantial amount of graduation by means of very large personal and family exemptions. Further, while the tax-free levels of wage income created by the exemptions are large enough to risk taking a potential majority of voters off the tax rolls altogether (not a good idea from a civics standpoint), the degree of graduation achieved by the exemptions is not sufficient to avoid large middle- class tax increases relative to current law. Why not stick with the original idea of the bifurcated X-Tax, use multiple tax rates, keep nearly all citizens on the tax rolls, and avoid creating impediments to what is otherwise a "nearly returnless" tax system in combination with the good economics of eliminating the double tax on saving and investment?

The answers to these and other questions about the flat tax may be largely ideological. Given their druthers, most proponents of the flat tax would probably prefer a strictly proportional tax: one universal rate and no deductions. (Even though the flat tax is not "flat," at least it does not have the multiple rates that historically have been associated with the kind of highly regulatory and redistributive tax philosophy they are likely to abhor.) Many advocates of the flat tax are also dedicated "supply siders" and tend to equate a single rate with the idea of a low top marginal rate that they believe will unleash an upward spiral of economic growth. Large levels of exempt income -- $31,400 for a family of four under Armey- Shelby -- are an unfortunate concomitant of a single-rate tax, but some flat tax advocates may rationalize the exemptions on the grounds of "simplification" with an ideological twist. ("If we can't get rid of the tax code altogether, at least we can free tens of millions of Americans from its clutches.")

Once one gets beyond puzzling about the single rate and related speculations and, instead, looks at the results, it is fairly obvious that the dominant and distinguishing characteristic of the flat tax is tax-free status for a large and potentially growing group of citizens. The single-rate version of the bifurcated X-Tax is no simpler than the multiple-rate original. The single rate (20 percent) is no lower than the top marginal rate under a multiple-rate version with very low personal exemptions and about the same income class distribution as the flat tax.(27)

It is also obvious that the Hall-Rabushka flat tax has built-in tensions and discordances that tend to make it unstable and subject to evolutionary change for the worse. Human nature and politics being what they are, the exempt levels of income inevitably will grow over time and, as they do, the tax burden will be concentrated on a smaller and smaller minority of voters at ever higher tax rates. Additional sources of upward pressure on the single rate are also obvious. For example, the payroll tax on wage income will need to be taken into account and a limited step in that direction has already been suggested by the Kemp Commission.(28) The likely addition of itemized deductions for charitable contributions and home mortgage interest -- also acknowledged by the Kemp Commission -- would erode the broad base and increase the tax rate.

As discussed in connection with the bifurcated X-Tax, subdividing the labor-capital base of the corporate model between individuals (labor) and corporations (capital) has its own set of problems. For example, advocates of border tax adjustments are likely to seek an efficacious way of having a combined labor-capital base at the business level. On the individual side, critics of the bifurcated structure will point to how easily the flat tax can be demagogued when only wage income is included in personal tax returns. Real life debate authenticates this concern about the political viability of the "yield-exemption" approach. Although there are many versions of the same devastating lampoon, the most well-known says, "Mr. Rockefeller would file no return and pay no tax, but his gardener and his chauffeur would."


Even though it requires a tax return and, in fact, makes no apology for it, the USA Tax has a definite place in any exposition about the search for a returnless tax system with all of the right substantive characteristics. With only two other exceptions, one curable (high rates) and the other debatable (simplicity), it meets the requirements for a restructured tax system. It is graduated by means of multiple rates, neutral as between saved and consumed income, evenhanded as between labor and capital income, border adjustable, territorial, and carefully crafted to take into account all transitional consequences. The USA Tax is also pertinent for reasons of clarity in this exposition. Because it includes both labor and capital income directly in the personal tax return and, therefore, uses the deduction approach (as distinguished from the "yield- exemption" approach) to eliminate the double tax on saving, the USA Tax serves as the analytical bridge between the "nearly returnless" flat tax and other variations of the broad-based corporate model that, as discussed later on, would be "returnless" but for the need to allow a deduction for saving.

Structurally, the USA Tax starts off with the broad-based corporate model where returns to labor and returns to capital are taxed at a uniform flat rate. But, instead of then dismantling the model by removing the labor component from the base and taxing wage income directly to individuals, the USA Tax retains the corporate model intact, at a greatly reduced tax level and rate, as step one in a two-step process of collecting tax on the flow of income from its source (businesses) to its ultimate recipients (people). Because the full labor-capital base of the corporate model is used only to precollect an amount of tax equal to current business tax payments (including the employer payroll tax), the business-tax rate is only 11 percent. (If the business credit for the employer payroll tax were omitted, the USA business rate would be only about 4 percent.) Also, because of the full labor-capital base, the USA business tax is border adjustable. At the same time, however, because of the low rate, the business cash flow problem otherwise present in the corporate model is avoided.

Having, in step one, precollected a uniform and partial tax on labor and capital income at the source, the next step under the USA Tax is to collect the remainder of the tax from individuals when they receive wages and salaries or interest and dividends. Both types of income are directly included in the recipient's personal tax return and, after allowable deductions, are taxed at progressive rates of 8, 19, and 40 percent (or 19, 27, and 40 percent during a 3-year transition period.)(29) Deductions are allowed for relatively small personal and family exemptions. A deduction is also allowed for the full amount of income -- either labor income or capital income -- the person saves by purchasing a financial instrument such as a stock or bond, or puts in a bank account. Having embarked down the road of a full-fledged tax return, the USA Tax also allows deductions for charitable contributions and home mortgage interest. Once tax is computed (by applying the tax rate schedule to income minus deductions), so assiduous is the USA Tax in the pursuit of evenhanded treatment of labor and capital, it then allows a tax credit for the full amount of the employee-paid payroll tax.

The 40 percent top rate under the USA Tax is high, but such a rate is not an inherent feature. The basic USA structure is designed to accommodate any set of rates, from highly progressive to flat or somewhere in between. The 8, 19, and 40 percent rate schedule in S. 722 produces a result that is both revenue neutral and distributionally neutral relative to current law. In this respect, S. 722 was apparently intended to demonstrate that an internationally competitive, savings-oriented tax system need not be regressive or lose revenue in comparison to current law. What might be the best set of rates under a system such as USA, if enacted, is open to debate. If the standard of strict revenue and distributional neutrality were relaxed somewhat, the intermediate and top rates under USA could be lower than in S. 722. At whatever level finally set, the rates under USA will, however, always appear to be higher than they are in comparison either to current law or to some other alternative system such as the flat tax, which does not allow a credit for the 7.65 percent employee-paid payroll tax. For example, insofar as most wage income is concerned, when S. 722 is compared to a 20 percent Armey- Shelby flat tax, either the top rate under USA must be said to be 32.35 percent (40 percent minus 7.65 percent) or the single rate under the flat tax must be said to be 27.65 percent (20 percent plus 7.65 percent).

There is, however, more to the story. Having included both returns to labor and returns to capital in personal income subject to personal tax rates, and having thereby solved many of the discordances of the bifurcated X-Tax and its flat tax cousin, the USA Tax must allow a deduction for personal saving. Otherwise, saved income would be double taxed. (The broad based corporate-only model needs no deduction for personal saving because expensing is allowed against the corporate rate and there is no separate rate on personal income. The bifurcated X-Tax and the flat tax need no deduction for saving because earnings on saving are excluded from tax at the personal level.) The allowance of a discrete, unlimited, and highly visible deduction for personal saving is one of the strengths of the USA approach and, in the minds of many, is its defining characteristic. On the other hand, the need for individuals to account for this deduction is the source of some complexity.

The deduction serves to permit people to shift their savings from one savings vehicle to another without incurring tax. If all employees, shareholders, debtholders, and depositors left their wage income, dividend income, and interest income in the "business" from which derived, and withdrew it only for the purpose of immediately spending it, no deduction for saving would be necessary to prevent a double tax on income that is saved. In that hypothetical case, it would be necessary only to apply a basic "realization" principle familiar to all tax lawyers and accountants: income is taxed only when received and salaries, wages, dividends and interest are received only when withdrawn from the business payor. (For this purpose, sales of stock or securities are the equivalent of withdrawals and gains are taxed only when realized.)

The deferral of tax that occurs both under this broad realization principle and under the deduction approach in the USA Tax can be illustrated by an example where Mr. Shareholder is in a 40 percent personal tax bracket and owns half the stock of Mexxon Corporation, which earns a $100 profit. In all three cases, the operative ingredient is the so-called "corporate shield" under current law (i.e., corporate profits are not taxed to shareholders until paid out to them or until the equivalent of a payout is achieved by a sale of stock).

CASE A: Mexxon plows the $100 back into the business instead of paying a dividend. Therefore, Mr. Shareholder has saved $50 of income. Under both current law and the USA Tax, Mr. Shareholder's tax on this $50 of saved income is deferred. Shareholders are not taxed on corporate profit until realized through distribution of money or property.

CASE B: Instead of retaining the $100 of earnings, Mexxon pays Mr. Shareholder a $50 dividend and he spends the $50. Under both current law and the USA Tax, Mr. Shareholder has realized $50 of income and under both must pay a 40 percent (or $20) tax on the amount realized.

CASE C: Mexxon pays Mr. Shareholder a $50 dividend and he purchases $50 of IBR Corporation stock. In effect, he has moved $50 of saving from Mexxon to IBR. Under current law, Mr. Shareholder must immediately pay a 40 percent (or $20) tax. Under the USA Tax, Mr. Shareholder is allowed a $50 deduction for the purchase of IBR stock which serves to continue the deferral of tax he could have achieved had he left his $50 of savings in Mexxon. When he later withdraws $50 from IBR and spends it, he will pay tax at 40 percent.

In this simple case (which is representative of more complex situations as well), it is obvious that the USA deduction has served to permit Mr. Shareholder to move his $50 of saving from one vehicle to another. Appallingly, it is also obvious that current law would tax Mr. Shareholder on his savings only because he moved it. Had he left the savings in Mexxon, even under current law, he would pay no tax until he withdrew the $50. Worse, current law is highly selective in allowing this deferral. Under the IRC, shareholders can defer tax, but except for narrowly restricted situations, employees cannot defer and save salary free of tax. Depositors in banks must pay tax on accrued interest even though not withdrawn, but owners of insurance policies can defer tax on accrued interest in the form of inside build-up. In contrast, under the USA Tax, all inside build-up in all savings vehicles is tax-deferred under the broad realization principle, and when savings are moved, the deduction serves to continue the tax deferral.

The saving deduction entails a tax return, some calculations, and some recordkeeping. People who borrow large amounts or who had accumulated large amounts of after-tax savings under prior law would have to take those quantities into account in calculating their saving deduction, but most Americans -- probably 90 percent or so -- would be excluded from these complications.(30) In most cases, the saving deduction would be no more complicated than an IRA and would be calculated for customers by banks, brokerage firms, etc. Those people with more complex financial affairs could simplify the net savings calculation by using asset management or "Schwab" accounts, which are already widespread and are likely to become universal. (Under the USA Tax, savings transactions inside brokerage and similar accounts are ignored and only withdrawals need be taken into income.) With some refinements, the saving deduction under the USA Tax could probably be made a virtually automatic, hassle-free process for nearly everyone. This is not the place to argue the case that the USA Tax is far simpler than it is given credit for being. (It is not returnless and the return is not a postcard.) New technologies and data handling techniques are, however, rapidly changing the way in which everyone handles and accounts for money. Tax systems can and probably must change accordingly. So must the definition of "simplicity." What might have seemed complex in the old-fashioned "pencil and paper" world of yesterday might be simple today and tomorrow.


Still proceeding from the same basic two-level approach, where a partial tax is collected from businesses that produce income and the remainder of the tax is collected from individuals who receive income, it might appear that personal tax returns could be eliminated if businesses directly withheld the individual portion of the tax from each recipient's check and remitted the withheld tax to the IRS in essentially the same way as currently occurs with wage withholding. This option is usually among the first techniques that occurs to anyone seeking to design a returnless tax system. At first blush, it is quite plausible. At present, except for pensions and occasional gains on sales of assets, most Americans receive only wage income and wage withholding is already well established. There is no strong reason why tax could not also be withheld from interest and dividends.(31)

Upon closer examination, however, universal withholding on all forms of income is no more the key to total elimination of personal tax returns than is wage withholding under current law. Only if tax were withheld from all individuals at a uniform flat rate (with no personal or family exemptions or other deductions) would personal tax returns be unnecessary as a general proposition. Even then, people with gains or losses on sales of assets would have to file a tax return. With multiple tax rates (or with a single rate and large exemptions for purposes of graduation), all employees who worked for more than one employer, and all investors who received interest or dividends from more than one source, would have to file some type of year-end return. On the other hand, all these "returns" would be exquisitely simple -- assuming that the array of special deductions and credits under current law were eliminated and that the withholding-based system was, therefore, a tax on gross wages, interest and dividends in excess of personal and family exemptions.

The "nearly returnless" nature of a universal withholding system falls apart when other goals of fundamental tax restructuring are taken into account. Having departed from the basic corporate-only model by imposing a separate tax on the receipt of wages, interest, and dividends by individuals, the withholding-based system would double-tax saved income unless individuals were allowed a saving deduction. Once a deduction for saving is allowed, a full- fledged tax return is required and we are back to the basic structure of the USA Tax. The USA Tax already includes wage withholding and the addition of withholding on interest and dividends would hardly seem to be a material improvement.


Although deficient in other aspects, a schedular tax in combination with universal withholding is one way of totally eliminating personal tax returns while still maintaining some degree of graduation in tax as between high-income and low-income people. As a general proposition, a schedular tax system is one that varies the rate of tax by category of income instead of by reference to the recipient's income level or other characteristics peculiar to the recipient -- such as family size, tax-favored uses of income, etc. As the name implies, a schedular system has more than one rate schedule. Typically, there would be two: a lower rate (or set of rates) for wages and salaries and a higher rate (or set of rates) for interest and dividends. For example, in a flat-rate system, the rates might be 10 percent on wages and 25 percent on dividends. In a multiple-rate system, the rate schedules might be as follows.

Wage & Salary Income
Dividends & Interest

Schedular systems have sometimes been used as a last resort in less- developed countries where, because of widespread illiteracy and other reasons, tax returns and self-assessment are unsuitable. In such cases, tax based on the schedular rates is withheld at the source and that is the end of the matter. No further adjustments are made.

Such a system is certainly simple, but it is also highly arbitrary. The fact that a person receives more capital income than labor income, or vice versa, is not a reliable indication of that person's total income level and ability to pay taxes. Many high- income Americans receive more in salary than in interest or dividends. Many lower- and middle-income Americans receive only wage income, but some (retirees, for example) receive only investment income. Further, if strictly applied (without a saving deduction) to achieve a returnless result, schedular withholding would double tax saved income in a way that would seem to be more severe even than current law. Not only would both saved income and the earnings on savings be taxed, but the earnings on savings would be taxed at a premium tax rate.


If we do want to abandon our return-based self-assessment system, we could try to convert the IRS into a quietly competent servant to calculate our taxes for us. For example, the IRS would calculate your taxes for the year and send you a bill for the amount you owe. If you agreed, all you would need to do is pay the bill. No calculations (on your part) and no tax return. If you did not agree and thought the IRS's bill was too high, you could contest it. Provided the IRS did a good job and established a track record of trustworthiness, most of us who have modest amounts of income and pay modest amounts of tax would probably accept the IRS's calculations. In other cases, where large amounts of income and tax are involved, protests and protracted disputes might be routine.

Right now, the IRS would not be able to calculate our taxes with reliable accuracy unless we provided it the requisite information about our personal and financial affairs, in which case we might as well prepare the return ourselves or give the information to H & R Block and let them do it. But, with sufficient additional computer power and unlimited access to all existing public and private databases, the IRS could come very close to correctly calculating taxes for all of us who live on-the-record lives.

We can all think of reasons why an IRS-calculated-tax system would be unworkable or, if workable, undesirable as a matter of principle. Some might point out that the British have recently abandoned a government-calculated tax in favor of a return-based self-assessment approach.(32) Others would be alarmed by the vision of an all-knowing and potentially all-revealing and all-controlling giant computer in the IRS headquarters at 12th and Constitution, N.W., Washington, DC.

Tax reconstructionists who are more concerned about the substance of America's tax system than with the presence or absence of a tax return would likely look askance at the IRS-calculated tax. Other alternatives, such as the X-Tax variations and the USA Tax, achieve simplification and returnless or nearly returnless results as a by-product of desired substantive changes in current law. An IRS- calculated tax, on the other hand, would not necessarily require any major substantive change in current law. Once having freed people from the constant irritant of filing tax returns, it could be the vehicle for perpetuating the bad economics and behavior-modification features of current law.

Total elimination of all personal tax returns by means of an IRS-calculated tax is probably an issue for the future, but the same basic idea can already be applied in a more limited way to many current return filers, especially those with wage income subject to withholding and who have little or no investment income. (For example, Bob Dole's Plan For Economic Growth promised to free up to 40 million wage earners from the need to file a tax return, provided they had less than $250 of investment income.) Given that wage withholding is already fairly accurate for most people, and if the IRS were able to fully and correctly utilize all existing W-2 and 1099-type information as provided to them under current law, it is not implausible to think that 1040-EZ might be dispensed with for employees who are nonsavers and who have no other disqualifying characteristics. One perverse result of this approach would, of course, be to create a kind of "schedular" system -- not a schedular system that applies a higher rate of tax to capital income than to labor income, but one that says that you will be forced to file a tax return unless you abstain from saving and investing, owning your own business, or engaging in any activity other than working as an employee.


Like the broad-based corporate model (pure X-Tax) with which we started, the retail sales tax is returnless in the sense that only businesses would file tax returns. Individuals would be essentially anonymous insofar as the IRS is concerned. (Other parts of the governmental apparatus, as well as private-sector credit bureaus and the like, would know a great deal about everyone's financial and personal affairs, but the IRS would not.) If the income tax were replaced by a 20 percent sales tax, there would be no double tax on or other bias against personal saving. For businesses, the sales tax would achieve the equivalent of expensing by excluding wholesale transactions from tax. From an international competitiveness standpoint, the sales tax would be territorial in its application. Although not explicit, there would be implicit adjustments for exports and imports.

A retail sales tax meets all but one of the stated criteria for a restructured tax system: it is not graduated at all. Essentially the same thing can, of course, be said about the broad-based corporate model. Setting aside transition, it too meets all the stated criteria except for the assumed need for some degree of graduation in tax. In that sense, in our search for a returnless tax, we have come full circle back to the choice between two versions of a flat-rate business-only tax, both of which have the same aggregate tax base. In both cases, businesses are the return filers and are the ones who make tax payments to the IRS, but, in both cases, businesses are acting on behalf of others who the are real payers in the sense of bearing the ultimate economic burden of the tax. Unlike the basic corporate model, however, the sales tax is not normally thought of as a tax paid by a business on behalf of the people who are its owners and employees. Instead, because of its unique point-of-sale feature and the attendant statutory admonition that businesses are on each occasion to collect the tax from the customer, the real payers of the sales tax are usually said to be those people who make retail purchases. The popular assumption that people bear the economic burden of the sales tax only in proportion to their consumer purchases (and not at all in proportion to the amount of income they produce as owners and employees) is highly questionable, but thoroughly ingrained in the definition of a sales tax.(33) The assumption that consumers are the payors is significant only because some people are able to consume substantially more than the income they produce and others are able to consume substantially less than income. (If, for example, all consumers were also producers and if everyone's expenditures were each year exactly equal to income, the sales tax would lose its special character and would, in result, be indistinguishable from a flat-rate income tax with no deductions or exemptions.)

The most interesting and troublesome aspects of the sales tax relate to distributional analysis and to the inevitable attempts to modify the impact of the sales tax on selected people. The first and most obvious targets for relief will be low-income people. (Note the emphasis on "income." Consumption level may determine how much tax a person is assumed to have paid, but income level will determine eligibility for relief.) Priority will be given to low-income people because, without some modification, the sales tax is genuinely "regressive," i.e., tax as a percentage of income tends to decrease as one moves up the income scale.(34) It is possible to shield the poorest among us from the impact of the tax and to introduce some degree of graduation among those people who consume all their income, but the sales tax cannot be made progressive all the way up the income scale.(35)

Because of the assumption that the sales tax is borne solely in proportion to consumption (and not at all in proportion to income), the sales tax will always seem to fall most heavily on those who consume in excess of their income. Welfare recipients, students and others living on loans, and people receiving unemployment compensation consume in excess of income as traditionally defined, but they would be helped by whatever mechanism is used to shield the poor in general. Retirees are the ones who are not necessarily poor but who often consume in excess of current income and who are harshly treated under the sales tax for reasons which have nothing to do with income or wealth levels. Most retirees will eventually dissave, often in large amounts, and when they do, they are likely to be spending saved income that was taxed when earned under prior law at rates up to 40 percent, 70 percent, even 90 percent in some cases. For example, assume a retiree had, back in the 1960s, earned $100X and, after having paid a $70X tax, had invested the remaining $30X in a stock that is now worth $200X. If the retiree sells the stock and spends the proceeds, the entire $200X will be subject to a 20 percent federal sales tax. Double taxation of retirees and others with preexisting savings is an inherent "transition" defect in a sales tax.

As is the case with any flat-rate tax paid to the IRS by business intermediaries on behalf of others, the only way to modify the impact of the sales tax on the real payors is to provide them with an after-the-fact refund of all or part of the tax they are presumed to have paid. But in a world in which the sales tax has replaced the income tax and all the return filing and information reporting that goes with it, how is the IRS going to determine what amount of refund to make to whom? In the case of the refund-type X- Tax, a portion of the tax paid by any particular business intermediary is refunded to its employees and the amount is determined by reference to each employee's wages, but, in the case of the sales tax, all of the presumed "real payors" are anonymous insofar as the business intermediary and the IRS are concerned. The IRS would have no way of knowing how much any of us spend, whether we have a high income or a low income, or whether we are spending borrowed money, gifts, previously taxed savings, public assistance or whatever. Only by reinstituting something very like a fairly detailed tax return could the IRS provide appropriate refunds only to selected people.

The only way to preserve the returnless character of the sales tax is to provide refunds to everyone in a uniform amount and to convey the refunds to the recipients without any of them having to file an application for it. Therefore, short of the IRS simply mailing checks to the last known address of everyone with a social security number, the refund system would need to be appended to other on-going governmental activities that already regularly involve monetary transactions with potential recipients of refunds (e.g., the payment of FICA payroll tax to government and the receipt of social security benefits from government.) The rudiments of this approach are found in the Schaefer-Tauzin sales tax proposal, where employers are to "refund" to employees a portion of the FICA payroll tax withheld from their wages.(36) (In this particular case, the refund is the sales tax rate times an amount of wages up to the poverty level varied by family size.) Refunds similar to those provided to wage earners could be conveyed to nonworkers by increasing social security benefits, welfare payments, and unemployment compensation.

Although it is possible to design a functional, returnless sales tax, there are many reasons to doubt that a national sales tax either would or should be enacted as the replacement for the current federal income tax. Solely as a matter of principle, many members of congress, Republicans and Democrats alike, would find the sales tax hard to accept -- even if the impact on the poorest among us were blunted by a refund procedure. The refunds would, themselves, also be troublesome. Many would question the wisdom of a tax system which necessitates across-the-board cash payments to everyone, especially at a time when we are already struggling with what to do about existing "entitlements." The refunds might start off small (keyed to some assumed poverty level of expenditures), but a flat 20 percent federal sales tax on top of a 6 to 8 percent state sales tax would create enormous pressure for ever larger refunds to further erode the impact on people far above the poverty level. (Every time the refund increased, it would be necessary to raise the tax rate and every time the tax rate increased, the pressure for another boost in the refund level would mount.) After a while, it might be hard to distinguish the sales tax from a guaranteed annual income system financed for everyone by an ever smaller group of taxpayers.

For these and other reasons, the sales tax seems an implausible replacement for the income tax. If the sales tax has a role at the federal level, it is as a supplemental tax. A sales tax in combination with the current income tax is not an appealing prospect.


A tax system where only business "intermediaries" file tax returns is probably not realistic. All variations of the corporate- only model would have to overcome serious structural and philosophical hurdles. In the end, one must also question the need for such a struggle. Simplification and all the substantive goals of tax restructuring can be achieved without eliminating personal tax returns. Tax returns are not even necessarily bad. There is, for example, nothing like a tax return, and writing a check to go with it, to focus our attention on how much tax we pay to the government and to cause us to weigh in the balance, as all citizens should, whether we and society are getting our money's worth. All else being equal, a tax system with a tax return is, in my opinion, better than one without a tax return. The ideal would be a simple tax return that not only reminds us each year of how much tax we have paid, but also of how much we have set aside in savings for future income growth and security.

The goal of "simplification" must be put in proper perspective. Desirable though it is, simplification is not the most important goal of tax restructuring. Even if it were, simplification is not a goal which is readily achievable in isolation. (As we have seen, the technique of starting off with a "returnless" premise and then trying to fit the substance of a tax into that mold tends not to work.) True simplification is a by-product of the most important goal: a tax system that is neutral as between persons, neutral as between labor and capital income, neutral as between saved income and consumed income, and neutral internationally. The IRC of 1986 is not complex because it requires a tax return. It is complex because it violates all those principles of neutrality.

Vigorous pursuit of tax neutrality in enactable form is the surest route to the tax reconstructionists' grand vision of government as a competent servant quietly going about its assigned tasks without interfering in its master's affairs. The leading alternative tax systems, the USA Tax, the flat tax and the sales tax included, are neutral as between saving and consumption and, in varying degrees, share other good qualities as well, but so long as they are abstract alternatives waiting in the wings, their superiority to current law is of no practical importance. One of them must be enacted into law before the grand vision, or some realistic version thereof, can become a reality. It is in this respect that the USA Tax most outranks the flat tax and the sales tax.

To vote for the USA Tax, a member of Congress need only accept two closely related ideas: the pre-1986 IRA, where additions to savings are deducted but taxed when withdrawn; and capital gains "rollover" where, e.g., a stock or bond can be sold tax-free provided the proceeds are reinvested in another stock or bond. Once these two fundamentals are accepted, the rest of the USA Tax automatically falls into place in an easy and familiar pattern. It is not a hard choice.

The basic idea of allowing people a fair opportunity to save, in the manner of the pre-1986 deductible IRA, enjoys much bipartisan support, and in the 105th Congress, probably more than ever. The closely related idea of capital gains tax reduction remains controversial, with some ardently in favor and others adamantly opposed, but all factions ought to be able to unite around capital gains rollover. Surely people should be permitted to move their savings from one stock or bond to another without having to pay a toll charge. Pro-saving tax policies are becoming mainstream as the large constituency of middle-aged boomers now desperately try to save and invest for their retirement. If not already, it will soon become obvious that as social security and other entitlements inevitably decline in importance, Congress must allow Americans -- at all ages and income levels -- a fair opportunity to save and provide for more of their own security. For these and other reasons, members of congress, on both sides of the aisle, are likely to vote for savings and investment, provided that choice is not made too hard for them by extraneous factors.

As a matter of bottom line substance, both the flat tax and the sales tax would eliminate the bias against saving and investment, but, let us face it, both carry loads of extra baggage. To support either, most members of Congress would see themselves as voting against the principle of progressive taxation. A vote for a sales tax would also be seen as a vote against the whole idea of taxing income. Some members of congress might cast both votes with enthusiasm, but a prevailing bipartisan majority would probably balk. For them, the price would be too high, especially given the availability of far more palatable ways of eliminating the bias against saving and investment.


1. The USA Tax Act of 1995, 104th Cong., 1st sess., S. 722, Congressional Record 141, no. 67, daily ed. (25 April 1995): S5664- 74. The USA Tax is primarily identified with its two leading sponsors, Sen. Sam Nunn, D-Ga., and Sen. Pete V. Domenici, R-N.M. Often referred to as "Nunn-Domenici," the USA Tax is distinguished both by its bipartisan support and by its attempt to deal with a broad range of issues, including the payroll tax, cross-border adjustments for exports and imports, and a deduction for saving, all within the familiar structure of a progressive-rate income tax.

2. Freedom and Fairness Restoration Act of 1995, 104th Cong., 1st sess., H.R. 2060, Congressional Record 141, no. 67, daily ed. (19 July 1995): H7256 and S. 1050, Congressional Record 141, no. 67, daily ed. (19 July 1995): S10320-21. In a legislative context, the flat tax is primarily associated with House Majority Leader Richard K. Armey (H.R. 2060) and Senator Richard C. Shelby (S. 1050), both Republicans, but more broadly is identified with two academics, Robert E. Hall and Alvin C. Rabushka, and is often referred to as the Hall-Rabushka Flat Tax. Robert E. Hall and Alvin Rabushka, The Flat Tax (Stanford: Hoover Institution Press, 1985) and Robert E. Hall and Alvin Rabushka, The Flat Tax, 2d ed. (Stanford: Hoover Institution Press, 1995). In both its legislative form and its academic origins, the flat tax is distinguished by its low single rate of tax and the absence of any personal deductions except for large personal exemptions necessary to mitigate the effect of the flat rate.

3. Given that the overall rationale of the flat tax depends on maintaining the combination of a low single rate and a full base undepleted by deductions, the Armey-Shelby proposal (note 2 supra) would require a three-fifths vote of both Houses to increase (or add) a tax rate or to provide for a deduction or exclusion. The Gephardt proposal (Press Release dated 7/7/96) for a 10 percent tax for most people and a 20, 26, 32, or 34 percent tax for the rest, would require that any rate increase be approved by voters in a national referendum. (In the course of the 1996 presidential election campaign, Republican presidential candidate Robert J. Dole proposed a supermajority requirement to guard against future reversals of his proposed income tax cut. Restructuring the American Dream, Bob Dole's Plan for Economic Growth, August 5, 1996, Chicago, Illinois.)

4. Congress has long recognized its own lack of restraint and has, in other areas of its broad writ, tried to curb its excesses with extra-normal procedural requirements. The procedurally constrained budget and spending processes of today, and going all the way back to 1975 in one form or another, are the most prominent example. A failed example? Perhaps. But who can say that the self- imposed discipline of congressional budget resolutions has had no salutary effect? Without them, the federal deficit might be even greater.

5. Some opponents are in philosophical disagreement with the basic tenets of tax restructuring as reflected in most of the current proposals. Instead of a neutral and unintrusive tax system, they would prefer one more like the current system that rations and allocates resources, and that regulates behavior in ways they consider beneficial. Other opponents are motivated more by economic self-interest than by ideology. Burdensome and complex as it is, many large businesses and others have invested enormous effort and resources in finding a fairly congenial path through the labyrinth of the IRC. For them, current law may not be so bad. Bad for their competitors, bad for society as a whole, but not necessarily bad for them. Other self-interested opponents may include the bevy of lawyers, accountants, and other members of the tax industry that has grown up around the IRC. To them, its complexities and vagaries are the source of enormous power and wealth in a society preoccupied with taxes.

6. The Internal Revenue Service also "administers" federal taxes, but in a self-assessment, return-based system such as the current federal income tax, tax administration is primarily performed in the private sector. For the most part, it is the private sector where income and deductions are accounted for, where the arcana of the law are interpreted and applied, and where tax returns are prepared -- either by taxpayers themselves, by commercial preparers, or by professional specialists such as lawyers and accountants. It is in the context of this private-sector tax administration where the current federal income tax does its mischief, both in terms of its indirect costs (the loss of GDP due to economic distortions) and its direct costs (dollar outlays for accounting, interpretation and return preparation). Those costs are huge, variously estimated at between $150 to $200 billion annually. In comparison, the IRS's "operating costs" were about $7.5 billion in 1995. Treasury Department, Internal Revenue Service, Internal Revenue Service 1995 Data Book (Washington, D.C.: GPO, 1995), 26.

7. The extraneous duties of administering governmental regulations and programs interfere with the essential function of business, which is to produce national income and wealth. The annual cost of complying with federal regulations has been estimated at about $700 billion. "Tokyo Bill," The Economist 339, no. 7969 (8 June 1996): 34.

8. Although ignored by all tax restructuring proposals other than the USA Tax and not given great emphasis here to prevent complication of the main point of this exposition, the payroll tax can be viewed as an addition to the marginal rate on covered wages, especially in the case of current workers under age forty where the rate of return on their mandatory social security "savings" is likely to be zero or negative.

9. A case can be made for territoriality on a stand-alone basis, but, as a practical matter, in the absence of border tax adjustments, territoriality leads to the spectre of "runaway" plants. Conversely, when an export exclusion is combined with territoriality, there is no U.S. tax advantage in manufacturing abroad for sale abroad. Similarly, if there is an import tax, there is no U.S. tax advantage in manufacturing abroad for sale back into the U.S. market. The combination of territoriality and border tax adjustments is the best way of achieving the international competitiveness goal of tax restructuring.

10. Ideally, this corporate model would allow a credit for the entire existing 15.30 percent FICA payroll tax (both the employer and employee portions).

11. Absent the broader base, a corporate tax rate of about 150 percent would be required to replace all current income tax revenues. With such a rate, and when the corporate tax is viewed as a way of collecting tax on people's income without their having to file a personal tax return, a confiscatory tax would be concentrated on equity capital and the owners thereof, but no tax would be collected on interest income or wage income.

12. The estimated rate of about 20 percent assumes replacement of current income taxes with a border adjustable corporate model that also includes amortization deductions for the adjusted basis of pre- effective date assets that produce taxable returns on investment after the effective date. If this corporate model also replaced the entire FICA payroll tax (employer and employee), or if a credit were allowed for payroll taxes, the rate would be about 32 percent.

13. A broad-based business-level tax that includes both the labor factor and the capital factor would, as a general proposition, qualify under international agreements. See Gary C. Hufbauer and Carol Gabyzon, Fundamental Tax Reform and Border Tax Adjustments, Institute for International Economics, Washington, D.C., 1996.

14. Many informative insights and discussions by experts on tax incidence and distributional analysis have been complied and edited by David Bradford in a single volume that reveals the severe limitations on the current state of the art and the lack of consensus. David F. Bradford, ed., Distributional Analysis of Tax Policy (Washington, D.C.: AEI Press, 1995). Another informative discussion of distributional analysis is found in Methodology and Issues in Measuring Changes in the Distribution of Tax Burdens, prepared by the Staff of the Joint Committee on Taxation, U.S. Congress, June 14, 1993.

15. The residual is equal to personal consumption expenditures only because GDP is accounted for on the product side, i.e., it is the sum of personal consumption expenditures + gross domestic investment + government purchases of goods and services + exports (-) imports. On the other hand, if the accounting were fully and correctly made on a double entry basis, the product side (personal consumption + gross domestic investment + government purchases + exports (-) imports) would always equal the income side (wage, salaries, interest, dividends, etc.).

16. The traditional idea that the current corporate income tax is borne by equity shareholders has generally been attributed to Arnold Harberger, but in more recent analysis, this distinguished economist suggests that the current corporate income tax is borne by both labor and capital -- despite its capital-only base. The traditional view had assumed a closed economy, but the more recent analysis assumes an open, global economy where capital can flee high domestic taxes and seek a better after-tax return elsewhere. See generally Arnold C. Harberger "The ABCs of Corporation Tax Incidence: Insights into the Open-Economy Case," Tax Policy and Economic Growth (Washington, D.C.: ACCF Center For Policy Research, April 1995), 51- 76. When this same basic and highly logical analysis is applied to the broad-based corporate model with its combined labor and capital base, most economists would probably attribute to employees all of the tax on the labor portion of the base and more than half of the tax on the capital portion of the base.

17. In the direct withholding case, the corporation by law withholds from salaries, dividends, etc. the amount of cash it is required to remit to the IRS, but under the corporate model, the corporation would first have to pay the tax to the IRS on its own account and then seek to recoup it either by employing fewer people and/or by paying less for the capital and labor it does use.

18. David F. Bradford, "An Uncluttered Income Tax: The Next Reform Agenda?" (discussion paper #20 presented at "A Supply-Side Agenda for Germany?" at the Institut der deutschen Wirschaft, Koln, Germany, June 28-30, 1988); David F. Bradford, Untangling the Income Tax (Cambridge, MA: Harvard University Press, 1986). A further discussion and some additional elaboration is found in David F. Bradford, "On the Incidence of Consumption Taxes," in The Consumption Tax: A Better Alternative? ed. Charls E. Walker and Mark A. Bloomfield (Cambridge, MA: Ballinger Publishing Company, 1987).

19. Although this example and examples of other X-Tax variations continue to assume a rate of 20 percent, the rate would need to be increased to take into account either refundable credits for individuals or the use of multiple rates for individuals.

20. The refund-type X-Tax presents a unique and interesting question under the GATT. As in the case of the broad-based corporate model, businesses do pay a tax based on both the labor factor and the capital factor. The question is whether the refund to employees (not to the business) precludes the business tax on the labor factor from being taken into account as a "supply cost." See Hufbauer, note 13 supra, for a discussion of taxes as supply costs.

21. Theoretically, if each employer accurately reported the amount of tax it owed, the amount of wages (direct and indirect) it paid to each employee, and that employee's name, address, and family size, and if the IRS could accurately computer-match all multi- employer cases, it could make refunds without anyone having to apply for them.

22. It is unclear exactly how pension and retirement benefits, previously deducted at the business level under current law and never taxed, would be treated. It is also unclear how amounts paid by owner-employees to themselves would be treated. For example, if an owner-employee of a small business paid himself a "wage," part of the business tax would be refunded. If he paid himself a "dividend," there would be no refund.

23. In the 1970s, a proposal called the "national dividend plan" received some attention in some tax policy circles. John H. Perry Jr., The National Dividend (New York: Ivan Obolensky, Inc., 1964). All voters would have received a pro rata share of corporate tax collections. The theory was that voters with an indirect "equity" interest in profits would support public policies conducive to profits. Perhaps so, but given that their interest in profits was solely a function of the corporate tax rate, they might also have been inclined to vote for higher tax rates.

24. In general, the combined base of the business and individual portions of the bifurcated X-Tax is the same as the labor- capital base of the broad-based corporate model, but it is hard to argue that the individual tax on the labor portion of the base is a "supply cost" for businesses. Multiple rates at the individual level, personal exemptions and the likelihood of additional personal deductions make the task of fitting the bifurcated X-Tax into a GATT- acceptable mold even more difficult.

25. Personal deductions for home mortgage interest and charitable deductions are hard to resist. If allowed, the bifurcated structure could accommodate them only in the case of a person with wage income -- where a personal return is filed. Therefore, the bifurcated X-Tax presents the definite possibility that two people -- one with wage income and another with dividend income, but both having given the same to charity -- would be treated quite differently. The wage earner would be allowed a deduction, but the dividend recipient would not.

26. As noted earlier, the current double tax on saved income can be eliminated either by allowing no deduction for saved income and exempting the earnings on savings or by allowing a deduction for saved income and taxing both the principal amount and the earnings when withdrawn from savings, the former being known as the yield- exemption approach and the latter being known as the deduction approach.

27. Ernest S. Christian, "Kemp Commission Report: The Good, the Less Good, and the..." Tax Notes, Jan. 29, 1996, p. 605.

28. National Commission on Economic Growth and Tax Reform, Unleashing America's Potential, Report of the National Commission on Economic Growth and Tax Reform (Washington, D.C.: 1995).

29. During this transition period, retirees and others of modest means are allowed to deduct in three equal annual increments up to $50,000 of savings accumulated under prior law.

30. In general, the allowable deduction for new saving under USA is reduced by the amount withdrawn from old prior-law savings, but in those cases where old saving (exclusive of retirement accounts and a home) does not exceed $50,000, this calculation need not be made. Similarly, the allowable deduction for new saving must be reduced by borrowing only if debt exceeds $35,000 (exclusive of home mortgage debt).

31. Under current law, suggestions to withhold tax on interest and dividends have met with considerable opposition, but if universal withholding were part of a returnless system and if such a system proved to be as much in demand as its proponents believe, such objections would probably not be given great weight.

32. David Cay Johnston, "British to Adopt an American-Style Tax Filing System," New York Times, Feb. 15, 1996, sec. D, p. 6, col. 1, final edition.

33. After enactment of the sales tax (here assumed to be in the 18 to 20 percent range), the amount paid by buyers to sellers would consist of two components, one called price, the other called tax. The issue is whether the sum of those two components could, in real terms, forever be greater than what the price would be were there no tax. No one suggests that all retailers would rush out the day after the sales tax is enacted and mark down their prices to compensate for the tax. They would not. On the other hand, it is implausible to think all sales volumes and all returns to labor and capital would forever remain undiminished after an 18 to 20 percent tax-induced price increase to consumers.

34. When tax is divided by income, the sales tax is regressive only when (1) income is measured on an accrual or accretion-to-wealth basis, i.e., it is included in the divisor when earned (instead of only when drawn down and used) and (2) tax is assumed to be paid in proportion to consumption. If either condition were changed, the sales tax would be a proportional tax. For example, it would be proportional if the divisor included only consumed income. In the alternative, it would be proportional if tax were assumed to be paid in proportion to income earned.

35. Using the traditional measurements and assumptions (note 34 supra), if everyone were given a $4,000 annual exemption from a 20 percent sales tax, a person who earned and spent $20,000 would pay zero tax (0 percent), a person who earned and spent $30,000 would pay $2,000 (6.66 percent), and a person who earned and spent $40,000 would pay $4,000 (10 percent). A person who earned $80,000 and spent $40,000 would, however, also pay $4,000 and, in that case, tax as a percentage of income would be only 5 percent.

36. National Retail Sales Act of 1996, 104th Cong., 2nd sess., H.R. 3039, Congressional Record 142, no. 29, daily ed. (Mar. 6, 1996): H1775.

About | Leadership | Commentaries | News and Events | Newsletter | Testimonials
Seminars | Contributions | Congressional Testimony | Contact CSTR | Home