468 November 51 1985 THE CASE AGAINST A VALUE-ADDED TAX Bruce Bartlett John M. Olin Fellow INTRODUCTION I It long has been conventional political wisdom that any politician who calls for a U.S. value-added tax (VAT) is committing political suicide. Former House Ways and Means Committee chairman A1 Ullman D-OR for example, is widely believed to have been defeated in 1980 largely because of his support for a VAT.l Yet it now appears that C o ngress is intent on making U.S. taxpayers pay some form of VAT. The Senate and the House Ways and Means Committee have already approved a form of VAT to finance toxic waste cleanup in the Superfund reauthorization legislation.2 Senator William Roth (R-DE) is.obtaining significant support on the Senate Finance Committee for a VAT to be included in the tax reform bi11.3 And These developments have revived debate on the VAT, which was assumed to have been closed permanently. A VAT, however, should be as unacc e ptable today as it ever was. The difficulties that killed VAT in the past have not gone away and deserve serious and careful examination before congressmen embrace this new form of taxation As Congressman Byron Dorgin (D-ND) recently remarked The last guy to push a VAT isn't working here any more Jeffrey Birnbaum Tax Plan Backers Seek to Regroup After Two Setbacks The Wall Street Journal October 21, 1985.

H.R.2005, which passed the Senate on September 26, 19

85. The House Ways and Means Committee adopted a similar measure on October 17, 19

85. See Broad Corporate Tax Backed to Finance Bigger Superfund The Washington Post, October 18, 1985.

National Journal, October 5, 1985, p. 2235 2 THE UNEASY CASE FOR A VAT A value-added tax is a form of sales tax levi ed at each stage of production. Manufacturers and distributors of goods and service providers pay to the Treasury a tax on the cost they add to the product or service (hence "value-added at each stage in the process. At each stage, the tax paid by the bus i nessman is added onto the price of the good or service. Ultimately, there fore, the whole tax effectively is paid by the consumer at the retail level. The VAT has been a major feature of European tax systems for 20 years and frequently.has been proposed f or the u.s.4 The two most common cases made for a VAT are: 1) it would encourage saving because it is a tax on consumption; and 2) it would help the balance of trade because the tax would be applied to imports but rebated at the border on exports.

The four main arguments against a VAT are: 1) it is inflation ary, because it leads to a jump in the price level when introduced 2) it is regressive, falling more heavily on those with lower incomes who consume a higher percentage of their income; 3) it i s administratively complex and expensive to collect; and 4) it is a hidden tax and thus is easily increased once established.

The Effect on Saving and Investment Those arguing that a VAT would increase saving and investment assume that the VAT would replac e the entire current.tax system or at least the corporate income tax. Were this the case, then America's low saving rate probably would indeed increase, because the current income tax system essentially taxes savings twice. A consumption tax would remove this bias against saving.5 A version of the VAT was proposed for the United States as early as 1921 by tax theorist T.S. Adams. See T.S. Adams Fundamental Problems of Federal Income Taxation Quarterly Journal of Economics, August 1921 pp. 527-5

56. But the federal government has never adopted any form of VAT, although it was adopted by the Senate as an amendment to the Revenue Act of 19

21. See John F. Due The Value-Added Tax Western Economic Journal, Spring 1965, p. 1

65. The State of Michigan, however, does have such a tax. See The Michigan Single Business Tax: A Different Approach to State Business Taxation [Washington, D.C Advisory Commission on Intergovernmental Relations, March 1978).

An illustration may make this clear. Suppose there is an income t ax of 50 percent and the rate of interest is 10 percent. To obtain $1,000 either to save or consume, one must therefore earn $2,000 to start. If one consumes the money the cost is $100 per year of additional income.

But taxes apply to that as well. In ord er to earn a return of $100 per year one would have to save $2,00O,'which would require him to earn 4,000 in the first place. In this way, income taxes make saving twice as costly as consumption 3 The trouble is that recent discussions of the VAT no longe r envision it as a wholesale replacement for the federal tax system or even the corporate tax--but rather as an extra tax atop all others. If, as in the case of the proposed Superfund tax, it were a new tax to increase the overall level of taxation, then t here is no reason to believe that saving would be stimulated.

Indeed, to the extent that any portion of the tax fell on capital saving and investment would fall, not rise.6 On the other hand, if a VAT were offset by cuts in other taxes, it is uncertain wha t the overall impact would be. If the taxes cut were those that fall largely on capital, then it is likely that saving would be stimulated. But increasing saving per se, is not the sole aim of policy to increasing national output per capita. Though increa s ed saving and capital formation are a means to that end, they are not the only means Consumption, after all, is two-thirds of GNP. If consumption were to be heavily taxed, the demand for goods would fall and GNP actually might fall in the short run It is r eally just a means This effect on consumption is not necessarily an argument against the VAT saving in the U.S. and that increasing the rate of saving would lead to a more rapidly rising standard of living over time for all Ameri~ans.~ The point is simply that lawmakers need to consider all consequences of a VAT--good and bad. And there is no evidence that the VAT in Europe has had any effect on saving and investment.8 Many economists believe that there.is too little The Effect on the Balance of Trade The s uggested positive impact of the VAT on the balance of trade is probably the most misunderstood and dubious argument made for the tax. On the surface it seems common sense that, if a tax is imposed at the border on imports and rebated on exports then expor t s will benefit and imports will suffer. In fact this is not necessarily the case at all. This mechanism merely means that a U.S. VAT will impose the same burden on domestic and foreign goods. Explains a Treasury study The export rebate and import tax allo w ed for the value added tax are merely border tax adjustments required to put the value-added tax on a destination basis. The export rebate merely allows exporters to enter world markets free of value-added tax, not at a subsidized At least in the short ru n, introduction of a VAT will unambiguously reduce saving, because during the period between the time a VAT is passed by Congress and the time it takes effect, consumers will raid their savings in order to purchase goods before the VAT is imposed.

Martin Feldstein Does the United States Save Too Little American Economic Review, February 1977, pp. 116-121.

Henry J. Aaron, The Value-Added Tax: Lessons from Europe (Washington D.C The Brookings Institution, 1981), p. 13; Stephen M. Brecher et al., The Economic Impact of the Introduction of VAT (Morristown, New Jersey: Financial Executives Research Foundation, 1982), p. 48 4 price below the pre-tax price. Similarly, imposing a value-added tax on imports merely places imports on an equal footing with domesticall y produced goods; it does not penalize imports. A comparison with state retail sales tax is illustrative; in any particular state charging retail sales tax on a Toyota does not make a Chevrolet more competitive in that state, because the same sales tax app l ies to both automobiles. Nor would the Chevrolet be more competitive abroad just because it could be exported free of sales tax As with a retail sales tax, the imposition of a value-added tax with no offsetting change in any other taxes, would not directl y improve the U.S. trade balan~e The result: no competitive edge is provided to the exporters.

Only in a few special cases would a VAT give the U.S. some trade advantage. Among them 1) If a VAT replaced an existing--.tax---or---taxes ther-e could be a bene fit if the taxes replaced had been incorporated into the prices of goods. In this case, the prices of domestic goods would not rise as much as those of foreign imports.

VAT might help U.S. exports. But in a flexible system, a change in the flows of export s and imports resulting from an export rebate would simply cause exchange rates to adjust to reflect different price levels, with no ultimate advantage--even if case No. 1 held true 2) In an international system of fixed exchange rates, a 3) If foreign co u ntries did not retaliate in some way against a tax designed'to block their exports to the U.S. and to subsidize U.S. exports.into their countries, the U.S. could enjoy some export gains.10 The premises on which these special cases rest are shaky It seems u nlikely, for example, that all taxes are incorporated into prices the same way. The burden of the corporate tax, for example, falls largely on shareholders and has never been shown to be a significant factor in prices.ll Also, the sensitivity of Tax Refor m for Fairness, Simplicity, and Economic Growth: The Treasury Department Report to the President, vol 3, Value-Added Tax (Washington D.C Department of the Treasury, November 1984 p. 22 lo See Comptroller General of the United States, The Value-Added Tax--W hat Else Should We Know About It Washington, D.C General Accounting Office PAD-81-60, March 3, 1981 pp. 18-20 Arnold C. Harberger The Incidence of the Corporation Income Tax,"

Journal of Political Economy, June 1962, pp. 215-2

40. And as a widely used textbook notes The Harberger model is probably the most widely accepted view of the effects of the corporation income tax Edgar K Browning and Jacquelene M. Browning, Public Finance and the Price System 2nd ed New York: Macmillan, 1983 p. 3

70. In short, if the corporate tax, which does not fall on consumers, is replaced by a form of value-added tax, which falls entirely on consumers, then prices would be expected to rise even if the VAT replaced the corporate tax dollar-for-dollar l1 5 consumers to price changes (or the Itelasticity of demand for goods) varies. Some prices are easily raised without cutting revenue, others are not. Thus, a VAT could be passed through on certain goods, but not at all on others because of falling dem and.

Most studies show that, with a VAT, prices would increase roughly dollar for dollar unless it were to replace'a direct sales tax. Since the U.S. does not have a national sales tax,to replace, the benefit described in case No. 1 above would not be achi eved.

As for case No. 2, the U.S. does not trade in a fixed-exchange rate environment and adjust quickly to any tax-generated changes in price levels.

This would eliminate any trade advantages from imposition of a VAT, even if case No. 1 held true The ma jor currencies of the world freely float In case No. 3, history teaches that foreign countries retali ate against measures intended to tax their exports and subsidize exports of competing nations in Europe provides no evidence that a VAT improves a countr y 's Though much has been made of the possible salutary effects on the balance of payments from adopting the value-added tax, there is no evidence that it had any material effect on the balance of trade. Irl As in the case of savings and investment, the use of a VAT trade balances. Explains Henry Aaron of the Brookings Institution The major arguments for the VAT thus are weak. There is little reason to believe that such a tax would lead to an increase in saving and investment, and no reason to conclude that i t would improve the U.S. balance of.trade. Of course, if the VAT were to be a substitute for existing taxes on capital and, therefore help give the U.S. an advantage in savings and investment compared with its competitors, it eventually would tend to impr o ve America's competitive position in the world the tax simply stacked on top of existing taxes quick-fixIf for the trade deficit But this would not occur were A VAT is not a THE EFFECT ON PRICES It is generally assumed that a VAT increases the prices of g oods roughly by a dollar for every dollar of tax. There are many VAT advocates, however, who argue that this is not necessarily the case. The historical evidence on this point, in fact, is ambiguous.

To assess the likely price effects of a VAT requires cer tain assumptions. If the VAT were imposed as an additional tax, for l2 Aaron, op. cit. 6 instance, it probably would have a more significant inflationary impact on a wide range of goods, due to llcost-pushll pressures than if it were imposed as a replacem e nt for existing taxes closer the tax replaced is to a sales tax, the less likely the inflationary impact. In Europe, for example, the VAT often replaced a nationwide retail sales tax the introduction of a VAT had a negligible impact on prices. But since t h e U.S. has no nationwide sales tax to replace, the effect on prices of substituting a VAT for other taxes is much less clear The Under such circumstances Assumptions also must be made about the role of monetary policy. Most economists believe that the qua n tity of money is the most important influence on the general price level. Former Treasury Under Secretary Norman Ture points out that the Federal Reserve Board could counteract a possible impact of the VAT on prices through a restrictive monetary policy.1 3 While this may be theoretically true, in reality it is mainly assumed that the Federal Reserve Board would accommodate the introduction of a VAT with easier money. Consequently, most economists generally accept the view that introduction of a VAT would p u sh the price level up to a new plateau that a 10 percent VAT would raise the general price level by 8 percent.14 The Treasury Department estimates The historical evidence on this point, however, is unclear due to the varying circumstances.under which coun t ries have introduced a VAT and to the fact that general inflation was usually rampant at the time these taxes were established. An International Monetary Fund study found that in some cases the VAT was inflationary and in others it was not. It concluded t hat there is no inherent reason why a VAT should be inf1ati0nary.l Of course, if a VAT has no effect on prices, this raises important questions about who really bears the burden of the tax.

It is generally. g ss~ed ma Lthe tax is. borne ultimately by the c onsumer at the retail level. But if prices do not rise when the VAT is introduced, the tax then is borne largely by producers and distributors. This in turn raises questions about the presumed regressivity of the VAT and its potential impact on savings an d investment l3 Norman B. Ture, The Value-Added Tax: Facts and Fancies (Washington D.C of Taxation, 1979 p. 40 W. Hafer and Michael E. Trebing The Value-Added Tax--A Review of the Issues," Federal Reserve Bank of St. Louis Review, January 1980, p. 9 Washin gton, D.C 80/75, November 3, 1980 The Heritage Foundation and Institute for Research on the Economics l4 Tax Reform for Fairness, p.

21. See also Aaron, op. cit p. 12; and R l5 Alan A. Tait, Is the Introduction of a Value-Added Tax Inflationary?

Internat ional Monetary Fund, Departmental Memorandum 7 REGRESSIVITY AND ADMINISTRATION If normal assumptions about the effect of VAT on prices are made, then.a VAT is regressive. However, most countries with a VAT counter its regressivity by levying different rat e s on "neces exempted entirely from VAT, while such luxury items as jewelry may bear very high rates. In some cases, of course, the VAT replaced taxes that were already regressive; this made the VAT's impact on existing income distribution negligible. The s ame would be true in the U.S. if the VAT were used to replace a regressive tax such as the Social Security tax. sitiesl' and lfluxuries.lf Certain foods, for example, may be The big problem with making the VAT less regressive by exempting certain items an d imposing varying tax rates is that it increases enormously the administrative complexity, and thus the cost, of the tax. It becomes much harder to prevent cheating and requires endless distinctions to be made between similar classes of goods. Example: ch o colate in the form of a bar might be considered a food while chocolate as icing might be considered a luxury. Such distinctions increase the cost of administration and undermine support for'the tax system because the distinctions must often be arbitrary a nd sometimes appear bizarre.16 since every businessman becomes a tax collector under a VAT system, he must also make himself aware of these distinctions adding to his costs and headaches.

The'cost that must be expended by the government in terms of money a nd manpower to collect a VAT is no small sum. The Internal Revenue Service estimates that it would require 20,000 additional personnel at a cost of $700 million per year.17 Anything that makes the VAT more complex will push this cost still higher. As such , redressing the regressivity of a VAT perhaps ought to be considered on the spending side through some form of negative income tax, rather than by having varying tax rates And A HIDDEN TAX For many conservatives, the main objection to a VAT is what others see precisely as its greatest virtue: it is easier to collect than the income tax and thus easier to raise. Americans might not even be aware that their taxes were being hiked because the VAT would be incorporated into the prices of goods, rather than lev i ed on them directly as higher income tax rates. When income taxes are raised, moreover, taxpayers may reduce their l6 See Carlson, op. cit pp. 55-56; and Charles E. McClure, Jr Value Added Tax: Has the Time Come in Charls E. Walker and Mark Bloomfield eds ., New Directions in Federal Tax Policy for the 1980s (Cambridge Massachusetts: Ballinger, 1983 pp. 193-194 l7 Tax Reform for Fairness, p. 128 I I 8 liability through deductions, credits, and other preferences.

And low income families specifically can be exempted from tax increases. Rich and poor alike can avoid a VAT only by not consuming goods and services.

For good reason there is concern that, even if a VAT initially were introduced in a revenue-neutral form, it would be soon raised to pay for addition al ,government spending. According to economists James Buchanan and Geoffrey Brennan If such a proposal is adopted it may be predicted that, ultimately, the value-added tax would be used to generate revenues greatly in excess of the revenue reductions und e r other taxes.Itl8 Since each percentage point of a general VAT would raise approximately $24 billion very large sums of money might be raised with little immediate palpable pain the Treasury Department The experience of VAT countries confirms such fears. Notes For nearly all European countries with a value-added tax, total taxes have increased as a percentage of national output since the introduction of the value added tax.lS Adds Brookings economist Henry Aaron The value-added tax in Europe was intended a s a substitute for other taxes, but it has been associated with an increase in taxation.tt20 For this reason, even many VAT supporters, such as former Deputy Secretary of the Treasury Charls Walker, support it only in combination with some form of balance d budget/expenditure limitation amendment to the Constitution to prevent it from leading to higher levels of government spending THE SUPERFUND TAX Despite these problems with a VAT, the U.S. Senate already has passed a version of value-added tax as part of the Superfund extension. A similar tax also has been reported by the House Ways and Means Committee. The current Superfund, which pays for toxic waste cleanup, is largely funded by an excise tax on crude oil, imported petroleum products, eleven petrochemi cals, and 31 inorganic chemicals. Some general revenues also underwrite the l8 Geoffrey Brennan and James M. Buchanan, The Power to Tax (New York Cambridge University Press, 1980 p. 49 l9 2o Aaron, op. cit., p. 15.

Tax Reform for Fairness, pi 23. 9 Superfu nd. However, more than 90 percent of the tax was paid by just 78 companies out of some 876 liable for some tax.21 Thus the current tax base for Superfund is extremely narrow and could not possibly bear the major increase in tax revenue necessary to financ e the proposed toxic waste cleanup costs scheduled to climb from the current 1.6 billion to $7.5 billion over the next five years. Moreover, the chemical industry is already suffering from the burden of the existing Superfund tax.

The industry argues convincingly that it is not solely responsible for toxic waste. For these reasons, it was generally agreed that a broader-based tax was necessary to pay for the Superfund. The Senate chose a 0.08 percent tax on all manufacturing compan ies with sales above $5 million per year, based on gross sales.

The Washington Post correctly calls this a value-added tax. All the arguments against a VAT, therefore, apply to this new Superfund tax as Though the Senate simultaneously adopted a resolu tio n urging the conferees to find some other form of tax to pay for the Superfund, unless general revenues are used it is hard to see what sort of tax they will be able to come up with that raises sufficient revenues. And current concern about the budget def icit essentially eliminates the use of general revenues.

There are a number of problems with the Senate's new Superfund tax. By imposing the tax, for example, on many companies and individuals who have no conceivable responsibility for creating hazardous t oxic waste dumps, those who do create such problems would be less likely to cease doing so. They would know that much of the cleanup cost would fall on others. As Professor Richard Epstein of the University of Chicago puts it, general taxes on production, rather than on toxic waste itself, Ilwill only coerce parties who safely handle dangerous substances to subsidize those who do not.1f23 For these reasons, a waste-end tax, combined with general revenues to pay for sites where no culpable business exists o r can be found, plus increased efforts to identify those responsible for creating toxic waste would be a better method of dealing with toxic waste cleanup 21 Rashida Belal Environmental Taxes 1981-83," Statistics of Income Bulletin, Spring 1985, pp. 61-

67 . See also Joint Committee on Taxation, Background and Issues Relating to the Reauthorization and Financing of the Superfund (JCS-ll-85 April 24 1985; idem Background and Issues Relating to House Bills for Reauthori zation and Financing of the Superfund ( JCS-13-85 May 8, 1985; Congres sional Budget Office, Hazardous Waste Management: Recent Changes and Policy Alternatives, May 1985, pp. 64-90.

Editorial A Bad Tax for Superfund Washington Post, May 19; 1985, p.

D6.

Case of Superfund The Cat0 Journal, Spri ng 1982, p. 33 Superfund and Hazardous Waste 22 23 Richard A. Epstein The Principles of Environmental Protection: The 10 Perhaps the greatest SuperfundDAT danger is that it easily could be transformed into a general revenue source for programs unrelated t o toxic waste cleanup. As Congressman Tom Downey D-NY) recently warned Any time we need money, we'll go after it.1t24 This was similarly emphasized by Robert Dole (R-KS Senate Majority Leader, in his dissent to the Superfund bill. He warned The fact that a new tax starts out with a low rate and a limited purpose is no guarantee it will stay that way. The entire income tax system started out with similar limitations, and that did nothing to stop its expansion. 25 THE ROTH PROPOSAL Although the Superfund tax i s not intended to become a general value-added tax, such a general tax has been making its way through the Senate Finance Committee. Proposed by Senator William Roth (R-DE a 10 percent tax would be levied on net business receipts. The revenue would permit a number of changes in the Reagan tax reform proposal. Among them 1) a reduction in individual marginal tax rates from the proposed 15-25-35 schedule to a 15-20-25 schedule 2) a reduction in the proposed corporate tax rate from 33 percent to 30 percent 3) an alteration in depreciation schedules to achieve de facto I1expensinglt (the deduction of all business expenses in the year they are incurred rather than over the useful life of the asset and 4) the institution of a new IRA-like savings account with sig nificantly higher contribution limits and no penalties for withdrawal.

This proposal has not yet been formally introduced, but it is evidently based on the provisions of two earlier Roth bills S.243 and S.11

02. The former, known as the "Broad-Based Enhan ced Savings Tax Act of 1984,Il contains provisions for expensing depreciation and institutes a new savings account. The latter known as the IfBusiness Transfer Tax Act of 1985" contains the revenue-raising features of the new proposal.26 24 Quoted in Birn b aum Tax Plan Backers op. cit 25 Senate Report 99-73, 99th Congress, 1st Session, May 23, 1985, p. 32 26 For details, see the Congressional Record, February 6, 1985, pp. S1186 S1193; May 8, 1985, pp. S5675-S5680; and August 1, 1985, pp. S10994-S10995. 11 W h ile not widely known or discussed, Roth's proposal has quietly gained significant support in the Finance Committee and may clear the committee. Whether it will survive floor considera tion is uncertain Although labeled a Ilbusiness transfer tax," the Roth proposal is in essence, a VAT. It would be a flat-rate tax on "net receipts,Il defined as gross receipts of all businesses less purchases of raw materials and other inputs. Salaries, interest payments, and dividends, however, would not be subtracted from t he tax base, although it would be deductible from the corporate income tax base. Consumers would not explicitly pay the tax as they do with a European-style VAT, but firms surely would attempt to pass-through as much of the tax as possible, making it a de facto VAT. This is true for the Superfund tax as well There is much to recommend the Roth proposal. It would eliminate much of the tax system's bias against productive acti vity. Marginal rates would be slashed, and savings and investment would be stimula t ed. But the proposal cannot be viewed in isola tion It must be evaluated in the political and economic context of the tax system and tax reform generally. Most seriously it poses the basic problems of any VAT: it is administratively complex, would raise p r ices, and probably would soon be increased above its initial level. There is no guarantee, moreover, that the Congress will accept Roth's proposal as a package deal. The U.S. could end up with a VAT without getting the offsetting benefits of a 25 percent top individual tax rate or the other elements in the Roth plan.

On.balance, the dangers of such a proposal outweigh its potential economic benefits. It would be easier to accept if the Congress had already acted upon a balanced budget/expenditure I limitat ion amendment or if the VAT were a wholesale replacement for some other tax, rather than being a new tax on top of all the others-albeit with lower rates the constant and continuing efforts by many to raise taxes rather I than cut spending, the risk that a VAT would quickly lead to a major tax increase is just too great at this time I Given the .deficit p-roblem and I CONCLUSION What is good economics, regrettably, is not necessarily good politics. Yet it is extremely dangerous to look only at the economic s while ignoring the political dimension. Such may be the case with a VAT. As a replacement for high marginal income 27 On the similarity between the business transfer tax and a value-added tax, see Michael Schuyler, Financing Tax Reform With the Business T ransfer Tax (Washington, D.C Institute for Research on the Economics of Taxation Gnomic Report No. 33, September 24, 1985 pp. 7-10. i 12 tax rates and other taxes on capital, it probably would improve the efficiency of the tax system and the economy as a whole, even without the false claims for VAT's impact on foreign trade.

Nevertheless, without some overall constraint on the government's ability to tax and spend, it would seem inferior even to today's hodge-podge of taxes. And there is absolutely no just ification for a VAT as a new revenue-raiser, as in the case of Superfund An economically neutral tax system, one that removes impedi ments to work, savings, and investment, while not distorting investment decisions, is clearly desirable. The VAT potential ly could play an important role in the design of such a tax system.

Its virtues are many. Unfortunately, those very virtues make it politically unacceptable. As a hidden tax whose burden is incor porated into the prices of goods and services, it is too tem pting a source of revenue for a government unrestrained by constitutional limits on taxing and spending. Therefore, a VAT should not be part of the U.S. tax system in any form--regardless of its apparent short-term benefits, or its attractiveness to acade mics.