Date 18 November 1942
Author unknown
Title Revenue Program for 1943
Description Staff memo, Division of Tax Research, Treasury Department
Location Box 64; Tax Reform Programs and Studies; Records of the Office of Tax Analysis/Division of Tax Research; General Records of the Department of the Treasury, Record Group 56; National Archives, College Park, MD.

                      Revenue program for 1943

                              DRAFT IV

                                             November 18, 1942


The present inflation problem arises from the attempt by individuals to spend more than the total value (at existing prices) of available goods and services. Even if wage rates and farm prices are maintained at present levels, total income payments to individuals will probably reach $125 billion in the calendar year 1943. Personal taxes now imposed by Federal, State, and local governments will take at most $15 billion of this income, leaving at least $110 billion for individuals to spend or to save. If voluntary savings were to be maintained at approximately the level reached in the second quarter of the calendar year 1942, they would absorb approximately $24 billion. This seems a maximum estimate for 1943, since the level of savings reached during the second quarter of 1942 is not likely to be much increased in the face of higher taxes, the adjustment of individuals to their new higher levels of income, and certain other factors, without substantial inducement or compulsion. But even with this level of savings, consumers would be trying to spend $86 billion, while the supply of goods and services will be at most $70 billion, at present prices. If upward pressure on prices is to be relieved, the amount that individuals want to spend must be reduced one way or another by $16 billion.

The attached Table 1 shows the estimated distribution of income and of attempted consumption for calendar 1943 by income classes. The income classes under $3,00 include less than 8 percent of all returns. They account for a very much larger part of total attempted consumption. At some time, the total attempted consumption of these groups is only $18 billion, or 20 percent of the $86 billion for all groups. It is clear that the $16 billion reduction in consumption that is necessary to bring the total attempted consumption down to the level of available goods and services cannot be achieved solely from these groups. We shall have to go farther down in the income scale. How far must we go? The group under $1,000,though including more than $40 percent of all returns, accounts for only about as much of the total attempted consumption as the 8 percent of the returns with incomes above $3,000. Even if the $18 billion attempted consumption by this group were not reduced at all, there would still remain $68 billion from which the necessary $16 billion reduction could be achieved. It is clear then that we do not have to impair minimum subsistence standards in order to achieve the desired reduction.

The attached Table 2 gives four different methods by which the necessary reduction in consumption expenditures might be achieved. These methods are arranged in the order of progressivity. The last method -- Schedule C-4 -- is designed to show the approximate effect of a flat rate sales tax sufficiently large to reduce spending by the required amount.

While, as noted above, the reduction in consumption can be achieved without impairing minimum subsistence standards, it is clear from Table 2 that this will not be easy. Even the most progressive schedule -- Schedule C-1 -- involves a 5 percent reduction in attempted consumption in the income class below $1,000. It is intended that this reduction come primarily from single persons and supplementary earners in that class, rather than from families. This schedule involves a maximum reduction in consumption of 50 percent for the net income class of $10,000 and over. In the absence of measures to reduce consumption, individuals in this class would, on the average, try to spend about $8,800. To achieve the reduction set forth under Schedule C-1, it would be necessary that their actual consumption be reduced to $4,400. The 45,000 to $10,000 net income class would attempt to spend $4,000. Under Schedule C-1, it would be permitted to spend no more than an average of $2,200.

By imposing the reduction in consumption in a less progressive way, these very large reductions at the top of the scale an be avoided by only at the cost of much larger reductions at the bottom. For example, if we take the extreme Schedule C-4, the consumption of the top group is reduced only from $8,800 to $7,800. But this requires a reduction for the very lowest group from $650 to $500, or a reduction of 22 percent from an already inadequate level.

The schedules shown in Table 2 are only a few of the possible alternative ways of eliminating the discrepancy between what people would want to spend and the supply of goods available. This discrepancy must be eliminated one way or another. No larger supply of consumer goods can be made available except by impairing the war output. If the discrepancy is not eliminated by fiscal measures, it must be eliminated by direct measures; and if not eliminated by direct measures, it will be eliminated by inflation.


There are only two methods by which the amount that people will try to spend can be reduced to the amount that will be available for them to buy at existing prices: enough of their income must be taken away to reduce attempted consumption to $70 billion, or individuals must be induced to save enough so as to reduce attempted consumption to $70 billion. Some of the direct controls such as price control and rationing help to stimulate additional saving. This memorandum is, however, concerned with fiscal rather than direct controls.

Withdrawing income reduces the amount which individuals will save as well as the amount which they will attempt to spend. To achieve any given reduction in consumer spending, it is therefore necessary to withdraw substantially larger amounts of income than the total desired reduction in consumer spending. Stated differently, withdrawal method bear on both spending and saving. They therefore bear with particular hardship on individuals who are committed to regular forms of savings, such as the purchase of insurance, the repayment of debt, and the purchase of war savings bonds. Examples of methods that rely primarily on the withdrawal of income include this income tax, various forms of compulsory lending, and to a very large measure, a flat rate sales tax.

The problem of fixed savings commitments can be avoided by using methods designed to reduce spending without, at the same time, reducing saving. Saving can be made more attractive either indirectly by penalizing extra spending, or directly by giving a premium for saving. Examples of saving-inducement plans are a progressive rate sales tax, a spendings tax, and a combination of a spending tax with an offset for savings.


An indication of the amount that would have to be withdrawn from income in order to achieve the desired reduction in spending is given in Table 3. For Schedule C-1, the most progressive of the four schedules, the table shows that over $27 billion would have to be withdrawn from income in order to reduce consumption to the desired level. An example will show how this figure was derived. For the $5,000 to $10,00 class, average total income after present taxes is $6,608. The permissible average consumption is $2,227. This is the average attempted consumption before the additional taxes of individuals with an average income of $2,984 (obtained by interpolating in Table 1 between the $2,200 to $3,000 class and the $3,000 to $4,000 class). If when individuals have their incomes reduced by additional taxes to $2,984 they spend and save in the same way was individuals who before the taxes had an income of $2,984, then these individuals would have to have their disposable income reduced to $2,984 by the additional taxes of $3,624 in order to induce them to spend no more than $2,227. This is the assumption that was made in deriving the estimate of additional taxes needed. The estimate obtained in this way seems a minimum estimate of the amount of income that would have to be withdrawn. When individuals have their incomes reduced by additional taxes, they are reluctant to surrender their former living standard. They are therefore likely to draw heavily upon savings and to consume more than individuals who formerly had that lower income.

The additional taxes required under Schedules C-2, C-3, and C-4, were derived by making the assumption indicated above. The amounts needed are progressively smaller as the reduction in consumption is distributed more regressively because individuals in lower income groups have a smaller amount of savings on which they can draw than individuals in the upper income groups. At the same time, the differences are not very large. Even for Schedule C-4, the equivalent of the flat rate sales tax, approximately $24 billion would be needed if withdrawal of income alone were relied upon. Any method that relies primarily on the effect of withdrawal of income must therefore withdraw in the neighborhood of $25 billion to $30 billion at the very least in order to achieve the necessary reduction in consumption.


The individual income tax is the fairest tax that has yet been devised. Difference in marital status and in number of dependents are recognized, exemptions are provided to protect the low income groups, and the base has been adjusted in the course of many years of development to reflect the ability of the individual to pay taxes.

In order to raise even the minimum additional amount required -- $25 billion -- from an income tax with exemptions of $500 for a single individual, $1,2000 for a married couple and $350 for each dependent. the exemptions of the Revenue Act of 1942, an effective tax rate of about 60 percent on net income in excess of exemptions would be needed in addition to existing income tax, making an aggregate in the neighborhood of 85 percent. Even if exemptions were reduced to $400 for a single individual, $800 for a married couple, and $200 for each dependent. an effective tax rate of over 40 percent additional on net income above exemption would be needed, making an aggregate effective rate on income above exemptions of about 60 percent of the large base. Such tax rates might call for a still further lowering of exemptions; and the bulk of the tax would certainly need to be collected at source. 2. COMPULSORY LENDING

In view of the heavy rates of tax and the lower exemptions that would be required to withdraw enough income, it has often been suggested that the money taken away be treated not as a tax but as a compulsory loan to be repaid after the war. In this way the levy would have a less detrimental effect upon individual incentive to work longer or harder, and certain groups would be compensated in the post-war period for war levies that are considered too heavy as a permanent burden.

There are, however, serious disadvantages to treating the money taken away as a compulsory loan. A larger sum would have to be taken in the form of a compulsory loan than in the form of taxes. Since individuals would regard to loan to the Government as an asset and a form of savings and hence would be considerably less hesitant to meet it be reducing other forms of savings or by drawing on previously accumulated savings. Consequently, if $25 billion is a minimum estimate of the amount that would be needed through taxes to accomplish a reduction in consumption of $15 billion, perhaps as much as $30 billion or $35 billion would be a minimum estimate for the amount of compulsory lending that would be needed to accomplish the same reduction.

The rate of levy needed to raise this sum would be heavy. Total net income after present personal taxes will be about $95 billion, of this, perhaps $40 billion would be accounted for by exemptions of $400 for a single person, $800 for a married couple, and $200 for each dependent, leaving some $55 billion in the tax base. The raise $30 billion would therefore mean an average effective rate of 54 percent on net income, after present taxes and after exemptions. On this basis the following table gives the schedule that would be required to raise $30 billion.

                     Compulsory lending schedules
                        to yield $30 billion

        (Lending requirement based on income, with no offsets
                     for other forms of saving)

                    Married couple, no dependents

     Net income
     brackets /1/
   (after deduction         Bracket rates      Cumulative amount
     of regular              (percent)            of lending
     income tax)

       0 - $ 800                 0                     0
   $ 800 - 1,200                40                $  160
   1,200 - 1,600                50                   360
   1,600 - 2,000                60                   600
   2,000 - 2,400                65                   860
   2,400 - 3,000                67                 1,262
   3,000 - 4,000                68                 1,942
   4,000 - 6,000                69                 3,322
   6,000 and over               70                   -


/1/ For single persons, brackets would be half the width of those shown; for families with dependents the width of the brackets would be increased by 25 percent of these shown for each dependent.


A compulsory loan is likely to be less equitable in its immediate effect for a number of reasons. In the first place, because it is a loan rather than a tax there is likely to be less opposition to imposing it in accordance with a more regressive schedule. In the second place, even with the same rate schedule, a compulsory loan will be more regressive in its current effect on consumption standards than a tax. At all income levels, there will be a tendency to meet the compulsory loan out of other savings or out of previously accumulated savings rather than out of a reduction in consumption. The lower income classes will be least able to do this. In the third place, a compulsory loan imposes considerable inequities even among individuals at the same income level. Persons with liquid assets can avoid any current reduction in consumption simply by converting their assets; persons without assets or with assets that can only be realized on at a loss are forced to pay the loan out of reduction of consumption. Consequently, the burden on two individuals who have the same income and have previously maintained the same scale of expenditure may be very difference.

Repayment of the loan in the post-war period clearly does not make the present distribution of the reduction in consumption more equitable. Its effect is simply to commit the Government to making specified payments to specified people in the post-war period, although conditions at that time may call for distribution of different amounts to different people.

Because of the discrimination under a compulsory lending plan against persons who cannot easily transfer assets or who have fixed savings commitments, there will probably be considerable pressure to provide for offsetting against any lending requirement certain forms of regular saving such as insurance premiums, repayment of debts, or purchases of war bonds. This is the pattern that has already been followed for the post-war credit of the Victory tax.

Providing for any such offsets greatly reduces the effectiveness of the tax in curtailing spendings. Under a compulsory lending plan with no offsets, there are at least frictional hindrances to converting voluntary saving into compulsory lending -- insurance policies must be lapsed, surrendered, or borrowed on, mortgages must be refinanced, assets must be sold, and the like. But if various forms of savings are permitted to be taken as a direct offset to the lending requirement, there are not even frictional hindrances. There is merely relabeling. What was formerly called voluntary saving is now called compulsory lending.

It follows that the rates of a compulsory lending plan which provides for offsets for other forms of savings would have to be heavier than the compulsory lending plan described above if the same effect on spendings is to be obtained. This difficulty can be overcome to a considerable extent by providing for only a partial offset of other forms of savings rather than for a complete offset.

One for of compulsory lending that would require less severe rates to curtail spending by any desired amount is the payroll tax under the social security system. In the first place, the loan under the social security system is in general for a considerably longer period than the loan under most other forms of compulsory lending. Repayment is scheduled not for immediately after the war, but only for such a time as the individual reaches a specified age or satisfies certain other conditions. In consequence the benefits accruing under the social security program are not likely to be regarded as an adequate substitute for voluntary savings. In the second place, payroll taxes are imposed without exemption and hence bear relatively heavier on the low income groups with no substantial voluntary savings. The regressivity of the social security taxes is at the same time a disadvantage since it implies a regressive distribution of the current reduction in consumption.


A third major type of levy that relies primarily upon the withdrawal of purchasing power is a flat rate sales tax. A flat rate sales tax would have some incentive effect in making saving more attractive relatively to spending, but its major effect would be through the withdrawal of purchasing power. Unless exemptions were provided it would be impossible for individuals to escape the tax by curtailing consumption. A considerable part of their spending power would necessarily be drained off. This would be especially true at the lower income levels where there is insufficient leeway in consumption pattern to permit the saving incentive effect to be very important.

As the figures for Schedule C-4 in table 3 show, a flat rate sales tax operating primarily through withdrawal of income would have to raise almost $25 billion. If every dollar of consumer spending -- including expenditures on domestic services, professional services, and the like -- could be subjected to the sales tax, this would require a 35 percent flat rate. In practice it is impossible to collect a sales tax from all forms of expenditure. The practical tax base with a total volume of goods and services at existing prices of $70 billion is probably not over $50 billion. This would mean that a 50 percent flat rate tax would be required on those types of expenditures that can be reached.


A reduction of about $15 billion in the amount that consumers will want to spend could at one extreme be achieved without withdrawing any income whatsoever. This could be done if individuals could be induced one way or another to save an extra $15 billion out of unchanged income. The only certain way of accomplishing this result would be to compel the required amount of savings. Such a COMPULSORY SAVING PLAN would require that each individual save a specified amount out of his income after allowing for all capital transactions. Such a plan is very different from compulsory lending which is a requirement that an individual loan to the Government a specified amount that might be derived either from income or from the sale of assets.

It seems impracticable to achieve the objective of compulsory saving directly by specifying saving requirements. No individual could know that he had complied with the compulsory savings requirement until after the end of the period to which the requirement applied, since he would know neither his income nor his spending until that time. In consequence many individuals either deliberately or through honest error, would fail to comply with the saving requirements. It would not be feasible to "punish" these individuals except by sanctions laid down in advance, which would in effect convert the compulsory saving scheme into expenditure rationing, expenditure taxation or compulsory lending. /1/