212 September 21, 1982 THE FL A T TAX CHALLENGE INTRODUCTION Any observer of the tax policy scene certainly has been struck by the sudden appearance, early this year, of interest in a fl at-rate tax and by the momentum that has developed for enact ing a flat-rate tax into law. The idea of a flat-rate tax has been around for a long time; the current proposals are novel only in the variations of the basic outlines that are suggested It some w hat strains credulity to be asked to believe that the cur rent surge of interest is attributable to the sudden discovery that the existing income tax is unfair, distortive, hideously complex, expensive to comply with, and frightfully costly to enforce. We have all known this for ages. Could it be that the eruption of interest this year reflects an urgent concern to find some way to increase federal revenues in such a way as to con vince taxpayers that good things will result for them even while additional t axes are extracted from them? If this is, in fact the objective, if the motivation behind the present thrust toward flat-rate taxes really is to increase taxes in a relatively painless manner, then I think we should avoid these proposals like the plague. I ndeed, any such proposal should move toward enactment only if some constitutional or statutory safeguard is provided to limit revenue increases be pursued by a properly designed flat-rate tax the idea of a flat-rate tax must be seriously considered and an alyzed On the other hand, there are quite legitimate objectives to For this reason Our present income tax is, in effect, a collection of excises.

The man in the street readily-and correctly identifies the nature of an excise in terms of its principal effec t--to raise the cost of the thing subject to the excise compared to other things. An excise on gasoline raises its cost. People respond by buying 2 less gasoline, shifting their purchases to other (now) relatively less expensive things. With less gasoline sold, less is pro duced, less production resources are devoted to gasoline produc tion, and less income is generated by that production.

Any and all taxes have this excise effect of increasing the cost of some thing(s) relative to the cost of other things.

Taxes change the relative costs that would prevail in the absence of taxes. Taxpayers respond to these changes in relative costs by changing their behavior. These behavioral changes result in changes in the composition of economic activity, in the alloc a tion of the economy's production capability, and in the income claims generated by production. The greater the excise effect the greater the effect on relative costs, and the less neutral more distortive the tax.

The present income tax is a hodgepodge o f such excises. Its weightiest excise effect is in raising the cost of working rela tive to the cost of leisure (all those uses of time and resources other than the ones for which there is a market determined compensation The income tax also levies a heav y excise on saving investment).l To be sure, both of these excise effects were materially reduced by the Economic Recovery Tax Act (ERTA of 1981, although recent proposals to'increase taxes go far toward restoring the pre-ERTA bias against saving.

These ar e the basic excise effects of the income tax. At a secondary level, many provisions of the tax law act to differen tiate the burden of the tax according to a particular activity industry, or taxpayer characteristic. The income tax, in other words, imposes quite different excises on various taxpayers depending on their activities or other attributes. These excise effects-alterations in relative costs--distort the operation of the market mechanism in allocating production capability among the almost countles s alternative uses.

Reducing these excises and their distortive effects on the tax system, thereby improving the efficiency of the economy's use of its productive resources, should certainly be the primary goal of tax policy. A flat-rate tax is widely beli eved to be more neutral and less beset with excise characteristics than the present income tax. At least in the abstract, it is certainly possible to design a tax that would alter relative costs, parti cularly the cost of saving relative to the cost of co n sumption I For an extended discussion of these excise effects--of the tax bias against saving--prevailing before ERTA, see Norman B. Ture and B. Kenneth Sanden, Effects of Tax Policy on Capital Formation (New York: Financial Executives Research Foundation , 1977 and Ture, "Supply Side Analysis and Public Policy," in David G: Raboy; ed Essays in Supply Side Economics (Washington, D.C Institute for Research on the Economics of Taxation [IRET 1982), pp.9-28. far less than the present income tax ity in this sen s e should be the principal objective of any propos al for a flat-rate tax. The extent to which the proposed tax would serve this neutrality objective should be the foremost criterion for its design Moving toward tax neutral REDUCING COMPLIANCE COSTS AND TH E NEED FOR ENFORCEMENT RESOURCES Any law that rewires government to the economy. services that to compliance minimize this regulation, or public institutional arrangement the citizenry to incur costs in complying and the commit resources in enforcing impos es a burden on This burden is the loss of the output of goods and might have been produced by the resources devoted and enforcement A sensible tax policy should burden.

The present income tax has a track record, virtually unbrok en over the years, of const antly increasing complexity. Year by year, this complexity has expanded compliance and enforcement costs. One of the claims made by flat-rate tax proponents is that such a tax would be simpler in design; presumably, it would be easier to comply with and w ould require the allocation of far fewer resources to government enforcement activities.

Certainly, simplification is an important reason for shift ing from the present income tax to a flat-rate tax. For the most part, proponents of flat-rate taxes fail to point out that it is the change in the tax base contemplated by their proposal, not the flatness of rate per se, which is to achieve this simplifica tion and reduction in compliance and enforcement costs. This view is subject to important qualifications, .

First, any such simplification and the cost savings it might provide must be weighed in terms of what they cost to achieve, in terms of shortfalls in attaining other objectives. The altera tions in the tax base proposed in many of the flat-rate tax propo sals would increase the cost of saving relative to consump tion. This would not be wise. Increasing the excise effect of taxes on saving is too high a price for simplification.

Second, much of the complexity in the income tax is the result of efforts to c onstrain the availability of tax shelters and their effectiveness in reducing tax liabilities. Taxpayers pay a price for these tax shelters in the form of obtaining lower pretax returns on their saving An efficient shelter-using taxpayer will allocate his saving to such investments only if the after-tax return exceeds that which he can obtain from a nonshel tered investment In other words, he will undertake the shel tered investment only if the marginal tax rate on it is lower than that on nonsheltered inv estment, and at least enough to offset the higher pretax rate of return obtainable on the latter.

This search for shelters is a result of marginal bracket) rate graduation It is the possibility of reducing the marginal rate that 4 provides a significant pa rt of the inducement to find deductions exemptions, deferrals, etc. Flattening the rate structure, in itself, reduces the payoff on tax shelters. A single or flat rate would contribute enormously to simplification, without any alteration in the statutory tax base, merely because, having been made relatively more costly, the sheltering provisions would be used less extensively.

Third, even if the tax base revisions contemplated in flat rate tax proposals were, indeed, to afford simplification when fully imp lemented, an enormous price in additional complexity might have to be paid to get from the present status to the fully implemented flat-rate tax. The ultimate savings in compliance and enforcement costs might well exceed the costs of transition but the la tter should not be ignored in assessing the gains expected from moving to a flat-rate tax.

GREATER UNIFORMITY IN TAX TREATMENT The most appealing argument to many people is that a flat tax rate would be fairer than the present income tax It is obvious that, for many of its proponents, the gains in fairness are to be achieved not from flattening marginal tax rates--indeed this is widely perceived as resulting in a loss of equity--but from the tax base changes their proposals would effect. This illustrates the fact that few tax policy concepts are more ambig uous and less useful than equity as a p ractical guide for policy.

For this reason, uniformity of tax treatment should be substituted for fairness as the described objective in replacing the income tax with a flat-rate tax It does not necessarily follow that more nearly uniform tax treatment of taxpayers is fairer treatment, but greater uniformity is attainable while greater fairness, given its conceptual vagueness, is far more elusive. Greater uniformity may be justified in the interests of simplification, but, as in the case of simplicity, its priority must be conditioned on its consistency with the primary objective of neutrality.

FLATNESS OF RATE The broadening enthusiasm for a flat-rate tax might lead to the belief that no significant issues are raised by adopting such a tax. In fact, severa l of the most basic issues of tax policy are involved, and good policymaking requires that these be care fully identified and resolved by consensus.

As widely used, the term, flat-rate tax, is a misnomer. Few of the proposals call for a truly flat-rate ma rginal rate--a single rate applied to the tax base. are concerned more with broadening the tax base than with a flat tax rate. Flatness of rate and broadness of base are not neces Most proposals, in fact, I1 5 sarily tax policy buddies; we may well have o ne without the other. Quite different issues are raised by each.

The matter of how flat the tax rate structure should be addresses a conflict between considerations of economic efficien cy and of fairness. The major reason for providing a single rate to be applied to the tax base is to minimize the excise effect of the taxes. Marginal bracket) rate graduation, reduces increasingly the net return to the earner from each additional dollar of income he produces, whether as compensation for labor services or a s return on saving It becomes more and more costly for him to increase his income, whether by working more or by saving more. By the same token, graduated marginal rates levy an excise on increasing individual productivity In fact, the cost of progressive t ax rates is a less progressive, less efficient economy, in which working, saving, and investment for the advance of productivity are penalized by the tax efficiency resulting from marginal rate graduation two standard answers. One is that the payoff is wh a t many call a fairer tax--one that conforms more closely than others with the public's perception of "ability-to-pay I' The other is that graduation of tax rates can serve as an instrument for redistribu ting or equalizing income and wealth. Neither answe r is accept able The question is what do we get in exchange for this loss of There are As far as ability-to-pay is concerned, there is a virtual consensus among tax theorists that the notion is too vague and elusive to warrant attempting to shape tax polic y around it.

There is a broadly held and solidly based view that, whatever the conceptual construction (and whatever the utility-maximizing function that is assumed there is little reason to consider annual income an adequate measure of taxpaying ability; consump tion is deemed by some to be far better, while others hold out for wealth No matter which is used, there are extraordinary problems of definition to be resolved if there is to be confi dence that the chosen economic variable has anything to do wit h ability-to-pay In any event, it does not follow that graduation of marginal rates is needed to satisfy any operational view of ability-to pay. Indeed, all that is required is that tax liability increas es with income, consumption, or wealth, or whatever b ase is deemed acceptable with ability-to-pay as a criterion. And even if this requirement is construed as calling for more than propor tionate increases in tax liability as the individual tax base increases, it does not follow that the rate applied to the base must be graduated. Indeed, for this purpose, it is the effective tax rate--the quotient of tax divided by tax base--that is rele vant, not the marginal rates. Substantial graduation of effec- tive rates is readily achieved with the imposition of a si n gle or flat marginal rate simply by exempting the first X number of dol 6 lars of the base from the tax personal exemption system or by providing a zero-rate bracket in the tax base This may be achieved'with a This will-o'-the-wisp character of vertical e q uity was noted and documented very early on in the development of tax theory It is seldom, if ever, addressed in rigorous discussions of the proper shape of the tax rate structure. Indeed, Henry C. Simons who probably had.the most persuasive influence on c ontemporary thought about such matters, often asserted that the real and only purpose to be served by an income tax with graduated marginal tax rates is to assist in equalizing the distribution of income and wealth. We should not need a reminder that ther e is far from a substantial consensus that equality of income and wealth distribu tion are appropriately served by public policy contrary were true, we should be brought up short by the fact that marginal rate graduation itself has obviously been almost if not completely, ineffectual for this purpose.2 Disregarding philosophical reservations and the negative empirical evidence about equalizing income distribution, one must ask why marginal rate graduation is needed for income redistribu tion purposes As in t he case of ability-to-pay, it is not the shape of the marginal rate structure that is relevant in this regard; it is the shape of the effective rate structure. If the tax is to be used, however ineffectually, for leveling the distri bution of income, this calls at most for graduated effective rates, which, as shown, can be readily provided by a system of personal exemptions or a zero-rate bracket and a single or flat marginal rate Even if the There clearly is no objection on grounds of fairness or income d i stribution against a flat or single marginal rate. This issue should favor a single rate, with no graduation of marginal rates whatever. Any departure from a single rate almost certain ly will lead to more and more graduation through time easily foresee b u dgetary circumstances akin to those we now face exerting pressure to steepen the graduation as a means of raising revenue without offending all taxpayers. This is, of course diametrically opposed to good public policy, which calls for offending everyone w h en taxes must be raised One can The ultimate results of equality of income achieved by the tax subsystem and the reasons why the graduated income tax has made no significant progress toward such equality is explored in Norman B. Ture, "Taxation and the Di s tribution of Income," principal paper in Wealth Redistribution and the Income Tax (Lexington, Massachusetts D.C. Heath and Company 1978). 7 BROADENING THE TAX BASE Issues concerning the tax base are virtually independent from those pertaining to flatness o f the tax rate structure. But as in the case of the tax rate issues, a seeming conflict arises between considerations of economic efficiency and those of fair ness As suggested earlier, the existing income tax is properly characterized as a mix of differe n t excises. To some extent, the source of.the variance in rate from one excise to another in the tax is difference in the statutory rates. But more important than explicit rate differentials is the difference in the extent to which various expenses and rec e ipts are recognized for tax purposes, as well as the timing of such recognition. By far the most consequential of the excise differences are the differen tially heavy rates imposed on saving compared with consumption uses of income and working versus leis ure.3 The efficiency concern focuses attention, in any proposal for redefining the tax base, on minimizing, if not eliminating, these excise differen tials.

Although there is general agreement in this regard, there is less of a consensus as to the prioriti es to be assigned the various excise differentials as targets for reduction. Those who prefer an expanded .income tax base are prepared to accept-often they simply ignore--the anti-saving bias that is intrinsic to such a tax. They emphasize eliminating or reducing differences in the tax treatment of income derived from various saving outlets as well as the differences between the tax treatment of income derived from capital and that obtained from providing labor services. Many of the proponents of this app r oach perceive the limited) neutrality goal of the measure as indistinguishable from an equity goal often articulated as equal tax treatment of equally situated taxpayers. Implementing this approach would result in adding to the income tax base and fully e x posing, to whatever tax rate structure is adopted, substantial amounts of saving or the returns thereto, which are only partially taxed under present law uniformity of excise effect among the various capital uses of saving while significantly increasing t he excise on all saving compared with that on consumption.

The alternative approach to broadening the base places the emphasis where it properly belongs--on reducing the basic excise differential against saving Some of the designations of the tax base resu lting from this approach--e.g., the "consumption based income tax the "expenditure tax"--are misleading or actually This would very likely result in greater Explanations and illustrations of these biases are to be found in Ture and Sanden, 2. cit.; and Tu r e Supply Side Analysis and Public Policy pp. 9-28. 8 pejorative in connotation. In fact, the basic attribute of this tax base is that it results in the same percentage increase in the cost of saving and of consumption; it is, in other words neutral betwee n these alternative uses of resources. To avoid the unfamiliarity of new terms, let us call this tax base the expenditure tax base.

Without detailing the design of the expenditure tax, its basic attributes can be delineated. Neutrality of excise effect bet ween consumption and saving requires that either 1) all saving (that is, reservation of income from consumption uses or equivalently, all purchases of sources of future income) be excluded from the tax base while all of the gross returns thereto including the gross proceeds from the disposition of the capital instruments to which the saving is committed) be included in the tax base, or 2) saving is included fully in the tax base but all the returns thereto are excluded. These alternatives are perfect equiv a lents; each equally well would eliminate the pre sent excise differential against saving. The choice between them should rest on practical considerations of compliance and enforce ment costs. Most proponents of the expenditure tax prefer the first alterna tive.

With either alternative, the resulting tax base is more nearly neutral between saving and consumption than is the expand ed income base, which, indeed, is likely to intensify the exist ing tax bias against saving.4 An additional advantage of the llex penditurell tax over the expanded income tax is that several of the principal sources of tax complexity would simply vanish Two obvious examples are capital gains and capital recovery pro visions of all sorts. With the exclusion of current saving from the tax base, there would be no occasion to compute capital gains or losses; all of the proceeds from the disposition of assets would be included in taxable income, not merely the gains or losses in the proceeds. Again, by reason of the exclusion of saving (i . e the purchase of sources of future income) from the tax base, there would.be no reason to attempt to allocate the recovery of the investment against the income it generates over time. The exclusion of saving is precisely the same as expensing of capital o utlays, obviating any additional depreciation, deple tion, or other capital recovery.5 To cite a single example, presumably an expanded income tax base would include as part of a covered employee's taxable income his employer's contribution to a pension p l an on his behalf. If exception were to be made on this score, it is more than likely that other exceptions would proliferate. The ultimate outcome might well be a larger tax base than the present one, but with little less arbitrariness in its composition.

With this treatment of saving capital outlays) and the returns there to, there is clearly no reason to distinguish the tax treatment between new and used assets, as some flat-rate tax proposals would differentiation would alter the relative prices of new and used assets and thereby introduce a needless unneutrality and distortion of invest ment decisions Any such 9 Fully implementing this approach not only would remove virtually all of the differentially heavy tax burden on saving it also would eliminate v irtually all of the tax differentials among alternative forms of saving It would, in short, achieve the second level tax neutrality among saving outlets aspired to by proponents of the expanded income tax base, while eliminating the basic bias against sav ing, which the expanded income tax base most likely would intensify.

Another result would be the elimination of tax shelters.

The expenditure tax approach would automatically eliminate any tax differential in the determination of net returns among alter n ative investments. Tax sheltered investments would have to make it on their own and would survive, if at all, in substantially smaller volume than at present. This result, moreover, would be obtained without explicit legislative prohibition of such invest ments.

The expenditure tax raises a fairness challenge which, when closely examined, confounds arguments about the uses of income with those about who the users are. The ability-to-pay adhere nts maintain that income from capital has at least the same taxpaying capacity as income from labor; according to such reasoning there should be no distinction in tax treatment on the basis of where the income comes from or how it is used. The point that i s overlooked in this assertion is that saved income ends up being taxed far more heavily than is consumed income; income from capital is taxed more heavily than income from labor It is difficult to understand why it is fair to tax income that is saved mor e than income that is consumed or why it is fair to tax the returns on provision of capital services more heavily than the compensation for providing labor services.

This challenge is usually finessed by those who advance the fairness argument by turning t o the empirical question of who does the saving It is certainly true that the expenditure tax would shift tax liabilities between those who do and those who do not save, compared to the distribution of liabilities under present law It is also highly likel y that the relatively few people in the upper end of the income scale save more of their income than those at the lower. But this is a minor matter.

Individuals at the bottom or lower end of the income scale can be substantially relieved of most tax liability under a really flat-rate expenditure tax by an adequate zero-rate bracket.

Those at, the top will reduce their tax liabilities only insofar as they continue to be big savers, with the resultant beneficial effects on the entire economy. challenge to fl atness of marginal rates, there is less to the fairness argument than meets the eye.

As in the case of the fairness CONCLUSION Not all flat-rate taxes are born equal. If the current thrust is to produce constructive results rather than tax back10 sliding, it will be necessary to discriminate carefully among the increasing number of proposals In doing so, the principal criterion should be the contribu tion of the proposed tax alternative to greater tax neutrality.

The focus should be on the big picture--el iminating the basic tax bias against saving and eliminating differentials in tax on the returns of different forms of saving. Adherence to this criter ion calls for a changeover to an expenditure tax, not an expand ed income tax, and for insistence on a t ruly flat marginal tax rate.

Giving priority to the neutrality criterion certainly does not ignore either simplification (reducing costs of compliance and enforcement) or fairness A truly flat-rate expenditure tax would be far simpler than the present inco me tax, but it would not be totally free of complexity. Simplicity, however, must take its place in line as a tax criterion. The ultimate in tax simplicity would be a head tax, but few policymakers, if any would urge it as the basic tax in our system.

Similarly, no one would deliberately design an unfair tax.

Even with the best intent and greatest effort to produce the fairest possible tax, however, there is likely to be little confirmation of success. If for no reason than that we do not know what tax fa irness really is, it should take a back seat to other criteria, principally neutrality, in the design of a flat- rate, broad based tax.

Let us not delude ourselves that a flat-rate, broad based tax will be easy to come by the tax so much as it is in deter mining how to get from here to there without serious injury to the innocent--the taxpayers themselves. To a huge extent, present business and household arrangements, transactions, and conduct of daily affairs are designed to accommodate the existing tax r e gimen's exigencies with a minimum of pain and cost. Any abrupt change would prove economically costly. The effort to implement a flat-rate, broad based tax will require a careful, probably extended transition which will present a great many challenging pr o blems The difficulty is not in designing Finally, there is the matter of social security financing If Flat-rate taxes per se offer no solution to these problems. it were deemed appropriate, the financing of the social security system, in whole or in part, could be folded into the general revenue system, whatever the character of the taxes in that system security's financing problems than do present taxes ed flat-rate tax system for affording a tax environment far more nearly neutral and therefore far less r epressive of economic efficiency than the one we now have Flat-rate taxes offer no more magic in solving social In the abstract, there is great promise in'a properly design But we should avoid extrava11 gant claims about what, in this real world, we can e x pect. We need a hardheaded, in-depth examination of the proposals now offered, plus deliberate and careful progress, not the pell-mell intemperate tax legislation we saw in July 0 Dr. Norman B. Ture Visiting Fellow The Heritage Foundation Norman Ture, a H eritage Visiting Fellow, is Chairman of the Institute for Research on the Economics of Taxation (IRET).

July 27, 1982, testimony before the Joint Economic Committee.

This paper was adapted from his For.further analysis, see the soon-to-be-released "An Evaluation of Flat Rate Tax Proposals: Efficiency, Tax Burden and Tax Base the first in a series of IRET studies on the flat-rate tax.