Typical conclusions drawn in peacetime concerning the disadvantages of a general excise or sales tax rest on an assumption that the alternative to this tax is either a progressive tax or a curtailment of government spending. In the special circumstances of wartime finance, when the choice is between a general excise or sales tax and an increase of inflationary borrowing, quite different considerations must apply.

In this section and the next, the term "general sales" taxes will mean "general excise or sales taxes" considered as a generic term without specific reference to the retail sales tax or the business excise tax of the "value-added" type, for example. Section IV will be concerned with the relative merits of these two individual types of tax.

In the current situation war expenditures are the dominant factor and are "given". Borrowing from the public must be heavy. Under the tax program existing and proposed in the Budget, borrowing from the public other than refunding is set at $33.6 billion for fiscal 1943. since estimates of war expenditures must be revised upward in all likelihood, these borrowing figures are probably too low -- even if the proposed tax and social security programs should be enacted in full now. Supply and national income are determined by the war situation and are not very responsive to tax changes. Operations are at capacity and the demand for consumer goods continually tends to out-strip the supply at the existing price level. The basic question becomes: Would the price level rise more if we borrowed the entire $33.6 billion or if we raised, say $5 billion of the total by a general sales tax and borrowed the rest? To put the question differently, we might ask, which is less inflationary under the conditions mentioned, a general sales tax or additional borrowing?

Inflation is characterized by a price rise resulting from an excess of monetary demand for goods beyond their supply at the existing price. Ordinarily such a discrepancy causes an increase in price which acts as a corrective to bring demand and supply together. In a situation like the present, however, where monetary demand exceeds supply by a very large amount, the price movement is not permitted to bring supply and demand into a stable balance. Rather, when the initial price rise occurs, some groups of consumers who are in a stronger bargaining position pass on the price rise by demanding and enforcing increases in incomes, in profits, in farm prices, or in wage rates. This rise in incomes which raises costs will, in turn, cause a new wave of price increases, and the curtailment will finally pass on to those groups in the population who are in weaker bargaining positions. A price rise which restores equilibrium by inducing increased production or by curtailing demand directly should not be called inflation. A price rise which causes additional price rises, which induces increases in incomes and costs, and which passes on the necessary curtailment to weaker shoulders -- this is the essence of inflation.


The fact that business men will try to add the amount of a general sales tax to the price of individual items does not in itself warrant a conclusion that a general sales tax is inflationary. Such a conclusion is based on an unrealistic set of assumptions regarding the present situation and on a misconception of the nature of inflation. In a period like the present, when government borrowing dominates the monetary influences, a rise in the price level of both war and nonwar goods is inevitable. Whether a given dollar amount is raised by borrowing alone or by borrowing plus a general sales tax, the DIRECT increase in the price level for nonwar goods will be approximately the same. The tax will absorb windfall profits resulting from the price rise, however. The choice between borrowing alone or borrowing-plus-taxes will not affect the volume of government spending. But the sales tax absorbs for government a part of the money resulting from a price rise which otherwise would accrue to business and some of which would be paid out to civilian consumers. Any increase in the price level occasioned by a general sales tax is therefore self-limiting as compared with the self- accelerating effect of additional borrowing.

It is possible that use of a general sales tax as a partial alternative to borrowing would induce some changes in consumers' spending-saving patterns. A general price rise does not affect all prices in the same way. Some rise and others fall, depending on flexibilities of demand and supply. Further, the immediate reactions of individual prices may differ, if only because exemptions from a general sales tax narrow the range of its impact. There may be difference, too, in the timing of the initial price rise, with some effect on the proportions of consumers' spending and saving.


Unless consumers' dollar-spending increases by the full amount of the sales tax, business gross receipts after taxes will be less than they otherwise would have been. Ordinarily this would lead to contraction of supply. In a period when the limits to civilian supply are largely set by governmental allocations and priorities, when consumers' money incomes are expanding and the speculative expectations of businessmen favor inventory accumulation, reduction in supply is unlikely to be significant. Instead, the reduction in business gross income may result in greater resistance to wage- increase demands, in a less rapid bidding up of raw material prices, and in some contraction of business profits. These are reactions that would not occur if inflationary borrowing instead of a tax were the method of financing the government's expenditures. The main difference between a sales tax and inflationary borrowing is that the government collects the money corresponding to the price rise in the case of the tax but not in the case of borrowing. This difference is important not only from the revenue aspect but also with respect to the secondary effects of the price rise, for the tax withdraws purchasing power from the market.

The price rise resulting from either a sales tax or inflationary borrowing, it is true, probably would be used to justify demands for wage increases. The justification for such demands will be less and the resistance to them will be greater under a tax than under a like amount of governmental borrowing. The tax will put a brake on the possible spiral of wage demands and price increases.

The price rise would, likewise, be used to support demands for advanced parity prices for farm products and this again would be true for either a sales tax or inflationary borrowing. It should be remembered that the "parity price" concept is statutory rather than economic. Whatever revenue measures are enacted, developments in war finance will require a re-examination of the index of prices the farmer pays and its use in computing parity prices. To the extent that a rise in the index may be attributable to or absorbed by war taxes, the government legislation or the procedures should be modified to prevent their rise from affecting parity prices.


Finally, there is little ground for believing that price administration will be made more difficult by the general sales tax plus borrowing than by borrowing alone. If an absolute and rigid stabilization of all prices were otherwise attainable, the Price Administrator would have cause to oppose any tax policy that would impel him to revise price ceilings. But there will be an unavoidably large volume of inflationary borrowing. Its repercussions upon prices that now are uncontrolled will necessitate revisions of existing price ceilings in any case. In setting new ceilings, a new element, that of the sales tax, may be injected into the deliberations, but no new problems will be presented.


Not only is the general sales tax plus borrowing preferable to borrowing alone for achieving the objective of countering inflation; it also is preferable from the point of view of equity. A general sales tax is regressive; but is less regressive than severe inflation, which results not only in price rises but, in addition, shifts the effects of the price rise to the part of the population in the weakest bargaining position. Therefore, at the present time, the criterion of equity is subordinate to the the amount of anti- inflationary influence. Almost any successful anti-inflationary tax would be more equitable than inflation, as the President said in his 1943 budget Message.


The price-raising effects of increased borrowing aggravate the administrative problems of war production and finance. Procurement contracts have to be re-negotiated to allow for rising money costs. The dollar volume of government disbursement associated with any particular level of physical production is magnified, with consequent necessity for floating larger or more frequent loans. At the level of administrative costs, however, inflationary borrowing is simpler and, in an immediate sense, cheaper than an increase in taxes.

G. POSTWAR CONSIDERATIONS More important is the postwar problem of maintaining full employment and meeting interest payments on a large volume of indebtedness. It is most desirable to permit only such an increase in prices as can be maintained in the postwar period. Heavier immediate reliance on taxation reduces the magnitude of the postwar problem both by helping to restrain wartime price advances and by holding down the amount of interest payments.


Compared with inflationary borrowing, the general sales tax is an effective anti-inflationary instrument, is less inequitable, and creates fewer problems of administration. None of the major objections to a general sales tax, summarized in the first section (page 2), apply if the alternative to it is additional borrowing. III. THE GENERAL SALES TAX COMPARED WITH AN INCOME TAX COLLECTED AT SOURCE


Perhaps the argument that a general sales tax plus borrowing is preferable to borrowing alone will be generally acceptable. But some may believe that this limited choice will not be posed, that the choice which will actually arise is between borrowing alone and borrowing plus a broadened and increased income tax collected at source. Those of this view might hold that whatever revenue can be collected from a general sales tax could be collected from a further increase in or broadening of the income tax. It is questionable whether from an economic and equity point of view a clear preference can be stated. Some of the pros and cons, however, which are essential to any definitive answer will be stated.


With all the "givens" in the war program, the decision about the net effect on the price level of a) borrowing plus a general sales tax, compared with b) borrowing plus an income tax collected at source, depends upon the way business and consumers will react to the alternative methods of financing. The anti-inflationary effect depends mainly on the extent to which the taxes:

1. draw down consumer savings (and cash balances or increase consumer borrowing to meet taxes) as compared with what they otherwise would have been;

2. draw down or prevent an increase in business profits; or

3. reduce the physical volume of goods and services demanded by consumers.

Under the assumptions in our proposition, income before taxes and physical output are virtually given. There can be no advance in the price level unless consumers (a) spend for the same supply of goods an amount greater than they would otherwise have spent, or (b) get fewer goods for the same total expenditure. To the extent that a tax is paid from funds which would otherwise have been saved by consumers or held idle by business, its effects are not anti-inflationary. Or, more directly, a tax is anti-inflationary only to the extent that it absorbs purchasing power that otherwise would be spent for goods and services whose supply is limited.

We have indicated in a preceding section that the repercussions of the tax upon ratios of consumers' spending to consumers' saving determine, in the case of a general sales tax, the extent to which business gross receipts after taxes are less than they otherwise would be. A somewhat similar statement could be made for the withholding tax on incomes, but in this instance, because the tax is collected from the individual income-recipient rather than through business, it is not directly an element in price. If equal revenue from the two types of taxes represented an equal absorption of purchasing power that otherwise would be spent for goods, the difference in price level with a withholding tax and with a general sales tax would be determined by the amount of the tax. This price differential, being due merely to the form of the taxes, would be no clue to their relative anti-inflationary effects.

Actually, however, the two taxes probably would not represent equal absorption of purchasing power. There will be less tendency for a general sales tax to be paid out of consumers' savings than for a withholding tax. This difference depends on the differences in the proportions of the two taxes borne by low, middle, and upper income brackets. Lowering exemptions under an income tax results not only in taxing additional low income-receivers but usually, to a larger extent, in increasing the tax liabilities in the middle brackets. As long as an income tax is progressive, it will place some part of the additional burden on those groups of the population which are the more likely to draw on savings for part of their tax payments. A part of the sales tax, especially of the type of a value-added tax, may be borne initially out of business funds and not fully reflected in immediate price advances. To that extent a sales tax, too, will not fully and immediately curtail demand.

It is difficult to decide whether on the whole the income tax or the sales tax will be more effective in curtailing demand. The answer depends on the type of income tax (progression and exemptions) and the type of sales tax that are being compared. In general the sales tax will perhaps curtail demand more than an increase in income taxes. This statement has greater validity the more regressive the sales tax and the more progressive the income tax being compared.

To the extent that both taxes curtail demand, the same curtailment is accomplished by a different process, by the non- compensated price rise in the one case, by the curtailment of disposable income in the other. A cut in disposable income relieves the pressure on prices. Therefore, the price level will be somewhat lower under an income tax than under a sales tax.


It has been argued against the use of a general sales tax that it stimulates demands for increases in wages and for higher parity prices, implying that a corresponding low-bracket income tax would have no such indirect effects. Comparing the sales tax with borrowing, it has been said above that wages probably would advance less under a sales tax than under additional borrowing. When the comparison is made between the effects of a sales tax and those of a low-bracket income tax, there may be more validity to the argument that the sales tax leads to greater pressure for wage increase, though even here the difference is not clear-cut.

A sales tax becomes an element in prices and therefore in the price indexes, whereas the direct effect of an additional income tax is to restrain rising prices. To the extent that price indexes are used for guidance in wage negotiations, it follows that sales taxes are more conducive than income taxes to wage increases.

Although such a development certainly is a possibility, it is not inevitable. If fiscal measures are announced clearly as part of a program for the control of inflation, it should be possible to convince labor that recovering the amount of the tax taken from consumer purchasing power by wage increases would defeat the anti- inflationary purpose of the program. In this connection, Randolph Paul, the Treasury representative, stated before the Ways and Means Committee on March 3 (l.c., p. 83): "With respect to the increases in the excise taxes . . . it is suggested that in the legislation imposing such increases the Congress state the policy that the increases are not to be considered a justification for increasing wages and are not to enter into the computations of parity prices for agricultural products." This statement with regard to selective excise taxes applies with even greater force to general excise or sales taxes.

With respect to wages, there is no need for using the cost-of- living indexes in a mechanical way. The first decision of the War Labor Board shows that it refused in principle to depend entirely on the cost-of-living index. The Board has indicated that the ability of a company to pay will be an important consideration in its decisions. This principle was announced at a time when the funds from the price rise accrued to business and were not obtained by the government thorough a sales tax. It would be easier to apply the principle in the case of a price rise which is attributable to a tax and which does not lead to increased business profits. Certainly the course of profits will influence the results of the large volume of wage negotiations that will take place without government participation.

On the other hand, it is by no means certain that a low-bracket income tax also will not induce demands for higher wages. Such a tax affects the better-paid workers, who are most vocal and active in wage demands. They may try to recover the deductions from their pay by wage increases in order to maintain their standard of living. The wage earner may be even more conscious of a deduction from his pay than of a tax that may affect prices, especially if the effect on prices develops only gradually. Until the price rise becomes substantial, the wage earner pays most attention to his nominal wage, as compared with a previous period. The fact that a withholding tax may indirectly retard price advances will not be evident to him.

The reaction of the worker to any curtailment of his purchasing power is one of the crucial problems in an anti-inflationary policy and deserves the most serious consideration in planning such programs. It is, however, misleading to assume that this problem is serious in the case of a sales tax but does not exist in the case of a withholding tax.

In examining the effect of prices which the farmer pays on the parity price index, it is necessary to go back to the ultimate purpose of the scheme. The primary purpose was evidently to give the farmer parity INCOME and parity prices were only a tool for this objective. If higher industrial prices increase the receipts to business or non-agricultural workers, a compensating price increase in farm products appears logical. A price increase resulting from a tax that hits urban and rural people alike offers no such justification for a compensating price rise.

It may be argued that this question is one decided by political power and not by rational decision; in either case the farm parity price index presents a problem that must be solved for any anti- inflationary fiscal policy applying to present circumstances. The sales tax affords no special problem in this respect as the farmer could equally well argue that the low-bracket income tax which he has to pay should be included in the index of prices the farmer pays exactly as his property taxes are included now.

It is extremely difficult to assess the probability that any particular pattern of readjustments in the relationship of consumers' spending and saving will occur. The current state of flux -- the shifts in family incomes, the large volume of inflationary borrowing and the price advance that must occur anyhow, the restrictions on civilian supply and the pressure for conversion to war industries -- all these and other circumstances present too many large variables and imponderables to permit clear-cut conclusions. Our broad conclusion is that a general sales tax may be somewhat more anti- inflationary than a withholding tax. This conclusion rests upon a presumption that in all probability it would impinge less upon savings -- i.e., that the general sales tax would do more to absorb purchasing power that otherwise would be traded for goods.