The debate about negative income taxation has been confusing in many respects. A variety of schemes have been proposed, some of which have the same name but appear to have little in common, while apparently similar schemes go under different titles. What exactly does negative income taxation entail, and how does it differ from a social dividend or a tax credit scheme? Why is it that in 1972 negative income taxation was espoused both by a liberal Democratic candidate for the American presidency and by a Conservative government in Britain?

The schemes could be classified in a variety of ways, but one of the most important distinctions is in terms of the extent to which they would replace the present income maintenance system. In what follows, we describe three of the principal possibilities.

The most far-reaching change would be the complete replacement of all present social security benefits, of the income tax and of social security contributions, by a single social dividend or demogrant scheme. In its simplest form, this scheme would involve the payment of a social dividend to everyone, dependent only on their family status; the revenue required to finance the scheme, and to replace the present income tax, would be raised by a proportional tax on all income. The effect of this kind of scheme is illustrated in the top part of Fig. 11.3. The heavy line shows the income after allowing for tax and the social dividend. The point where the dividend is exactly offset by the tax payable is shown as the ‘break-even’ point. Those with low incomes would be net gainers and better off than with the existing income tax.

Such a scheme was first proposed in Britain by Lady Rhys Williams in 1942, and has been taken up by a succession of authors, a recent example being Meade (1972). In the United States, the scheme was espoused by George McGovern in his presidential campaign, and he proposed a $1000 average per capita demogrant to be financed by a broad-based income tax (i.e. with fewer loopholes) at a rate of 33.3 per cent. This would have given a family of four a guaranteed income of $4000, with a break-even level of $12 000.

The attractions of this approach are clear. If the demogrant were set at a level sufficient to raise a family with no other income up to the poverty line, then it would be much more effective than the present system. There would be no gaps in the programme and the problem of take-up would be eliminated. However, the main reason why this kind of scheme has failed to make progress was brought out in the McGovern campaign—that to pay an adequate demogrant (in terms of the poverty standard) would involve a substantial increase in taxation. The levels of payment proposed by McGovern were designed to meet the poverty line needs of a family of four; but critics soon pointed out that the proposed tax of 33.3 per cent was inadequate to raise the required revenue. Moreover, even at that rate of tax, the scheme would have involved substantial increases in taxation for those in the middle income ranges. In the same way, the straightforward social dividend proposed by Meade in Britain involved a tax rate of 53 per cent (NB, new research by CISC and CORI due out April 2000 show that payment of a basic Citizen's Income this is possible with a much lower tax rate).

The high rates of tax associated with the extreme social dividend approach and the fear that they would be politically unacceptable have led to alternatives being proposed which would retain the main elements of the present income tax and social security schemes. In particular, they would retain the progressively increasing rates of tax of the present income tax, allowing greater flexibility than a simple proportional rate, and most of the present social security benefits, allowing the negative tax payment to be set at a lower level. In the United States, the case for a negative income tax as a supplement to the existing provisions has been put in particular by Tobin: ‘the merit of the negative income tax approach is that a workable and equitable system of aiding the poor can be introduced within the framework of present federal income taxation’ (1968, p. 118).

These proposals have appeared in a variety of forms, but basically involve a guaranteed payment and then the withdrawal of payment at a rate of 33.3 per cent or 50 per cent—see the dashed line in the middle part of Fig. lfl3. As shown in the diagram, this leads to an overlap with the present income tax, and the negative tax rate would continue to apply until the after tax income is the same as under the present ‘positive’ tax schedule. Some people, therefore, would be both receiving negative tax supplements and paying ordinary income tax.

Negative income tax schemes of this type have been put forward as a means of channeling help to low income families without the disadvantages associated with means-tested benefits. They would be universal, the coverage not being limited to people satisfying certain conditions, and Tobin, Pechman, and Mieszkowski (1967) have argued that an automatic system of payment would be possible, avoiding the need for people to apply and hence the problem of low take-up. It has similarly been claimed that the negative income tax would reduce the marginal rates of tax, and proponents have stressed the difference between the proposed tax rates and that implicit in welfare programmes (66.6 per cent).

The impact of the introduction of negative income tax on marginal rates of tax depends, however, on its relation to other means-tested benefits. As originally conceived, the negative income tax would have replaced public assistance and all benefits in kind, such as housing assistance and food stamps, but this would have meant setting the negative income tax payment at a considerably higher rate, to ensure that no families close to the poverty line would lose from the introduction of the scheme. (It is sometimes suggested that it is the necessary consequence of any reform that some people near the poverty line are made worse off, but, even apart from the equity considerations involved, such cases provide powerful ammunition for the opponents of reform.) The family shown in Fig. 11-2, for example, benefited from cash assistance, food stamps, housing and medicaid to the extent of $5860 in 1972, which means that the negative tax payment would have had to average nearly $1500 per person, or substantially more than currently envisaged. As a result, Tobin et al. (1967) see the housing, food stamp and medicaid programmes as continuing, with only public assistance being replaced by the negative income tax. The contribution to reducing the marginal rates of tax would therefore be smaller. If a person pays an extra 20 per cent of additional income in payroll and income taxes, while losing 25 per cent in housing assistance, and 20 per cent in medical assistance, then the marginal tax rate is only reduced from 84 per cent under the welfare system to 68 per cent under the negative income tax, with a tax rate of 33.3 per cent. The continuance of other means-tested benefits blunts the effectiveness of negative income tax in reducing marginal rates: ‘no matter how far Congress moves to build work incentives into cash assistance, the effort will fail unless in-kind assistance is also reformed’ (Aaron, 1973, p. 52).

The second difference is in the method of administration. Recognizing that the income tax system does not cover many of those in poverty, and that for these people the income tax procedure would not allow benefits to be paid until the end of the tax year, the Institute suggests that ‘the agency administering a RIT may have to use separate procedures for assessment of means from those now used for income tax’. This would represent, however, a major departure from the principle of using the income tax machinery to eliminate the stigma attached to present means-tested programmes. If a separate application has to be made, then it seems likely that at least some of those eligible would not apply for the reverse income tax. It would be a benefit designed for the poor, and it has yet to be demonstrated that it could be administered more successfully than present means-tested benefits.

The third type of scheme is that closest to what the man in the street might expect the term ‘negative taxation’ to mean: simply setting the present income tax machinery in reverse and paying benefits to those below the tax threshold. The early versions of the scheme, notably that of Friedman (1962), were of this form. The scheme would operate through the existing tax machinery, the main new departure being that everyone would be required to file a tax return. On the basis of the information provided, a person would either pay tax as now or receive a negative tax supplement. In this way, it is hoped that the payment would be automatic and that the problems of take-up would be avoided. The operation of the scheme is illustrated in the bottom part of Fig. 11.3. The level of benefit, shown by the dashed line, is determined by the income tax exemption level and the rate applied. The negative tax would be fully integrated into the positive tax, and there would be no overlap.

In the United States this variant has received less attention than the Tobin scheme discussed above; in Britain, however, an integrated negative income tax was proposed by the Conservative government under the title of a ‘tax credit’, and if they had been returned to power in the general election of February 1974, the scheme would have been in force by the end of the decade. Its main feature was the abolition of income tax allowances and their replacement by a weekly tax credit payable to all. This would have had the effect of helping those below the tax threshold who do not at present obtain full benefit from the tax allowances. At the same time, tax would have been payable at the basic rate from the first £1 of income.

The way in which the scheme would have worked may be illustrated by reference to the position of a married couple. The illustrative credit for a couple in 1972 would have been £6 a week and the basic rate of tax 30 per cent. If the couple’s income was £12 a week, the net payment to them would have been:

If their income had risen to £20, then the tax credit would have been exactly offset by the tax due; and with an income above £20 they would have been net payers of tax.

The enactment of the tax credit scheme would have provided definite help to certain low income groups, who at present derive less than full benefit from the tax allowances. This is especially true of those now eligible for, but not claiming, Supplementary Benefit and Family Income Supplement. A family with two children where the father earned only half the average earnings would have gained to the extent of nearly £4 a week; a single pensioner dependent solely on the National Insurance pension would have been better off by £2 a week. There were, however, serious problems with the scheme as proposed by the Conservative government. It excluded some 10 per cent of the population: the self-employed, those unemployed, sick or disabled with no entitlement to National Insurance, and single parent families with low earnings. The needs of these groups are likely to be particularly pressing. Even in the case of those covered by the scheme, the benefit depended a great deal on individual circumstances. While the very poorest pensioners would have gained, the benefit would have fallen with income and pensioners at or only just above the Supplementary Benefit level would have been made worse off. Those receiving National Insurance sickness or unemployment benefit would have become subject to tax and the tax credit could well have been more than offset by the increased tax. It would have made only a very limited contribution to reducing marginal rates of tax. Finally, the problem of financing the scheme was inadequately considered. If the cost were to have been met by a rise in VAT, then the real gain in terms of purchasing power would have been considerably reduced.

In the preceding sections we have described the inadequacies of existing social insurance systems in Britain and the United States, the failure of means-tested welfare programmes, and some of the problems with negative income taxation. What then should be done?

Opinion on the choice between these two approaches is often presented in polar terms, but this may well be misleading. Distinctions which appear ctearcut, such as that between people who favour negative income taxation and those who do not, become less so in the face of the wide variety of schemes coming under this general label. The schemes favoured by the Conservative government in Britain and the McGovern campaign in the United States were, in fact, very different. Moreover, it is important to remember that the case for one type of approach may rest heavily on the institutional framework of a particular country. Given the differences between the existing income maintenance systems in Britain and the United States, it is not really surprising that the same policy may be espoused by Conservatives in one country and liberal Democrats in the other.

We can nonetheless distinguish certain differences in emphasis:

Pattern of benefits Both approaches are concerned with guaranteeing adequate minimum income, but they differ in the extent to which they seek to concentrate benefit on those at the lowest income levels. Under the negative income tax proposals, benefits would be withdrawn at a rate of 33~- per cent or more as income rose; under the social insurance approach there would be less scope for such withdrawal, the main tapering of benefit being from their taxation under the income tax. The concentration of benefit on the lowest income groups is seen as a major advantage of negative income taxation, and Tobin has claimed that in the case of the alternative child allowance scheme, ‘of the net benefits, nearly 80 per cent would go to families above the poverty line.., the end result would be a modest but dubious redistribution from childless taxpayers to large families, and very little redistribution from rich to poor’ (1968, p. 110). If we felt fully confident about the adequacy of the poverty standard and that no one above this level was in need, then the sharp cut-off of benefits at the poverty line might well be acceptable. However, it may be doubted whether we can be that confident about the choice of a poverty line and we may feel that income redistribution towards those people not far above the poverty line may still be desirable. Moreover, the poverty line is defined for the purposes of analysis in terms of averages, and there may be people with incomes above the minimum who are, in fact, still in poverty because of individual variation in need.

Marginal tax rates and work incentives Associated with a particular benefit schedule is a pattern of marginal tax rates. The evidence concerning the effect of such differences in marginal rates on work incentives is limited, although more is becoming known, as the result of the negative income tax experiments in New Jersey and elsewhere. However, as Marmor has stressed, ‘the absence of information about work incentives is no bar to the issue being politically important’ (1971, p. 42). At a political level, concern about marginal rates is likely to focus on two issues: incentives for the poor not to work and the increase in taxation for the middle classes. The negative income tax with a high rate of withdrawal of benefit will have high marginal tax rates at the lower end (although lower than the welfare programme it would replace), and to this extent it is likely to be less attractive on the former count.’ On the other hand, it provides benefits to those in work as well as to the unemployed, in contrast to the social insurance approach which only pays allowances for adults when they are not working. Moreover, to the extent that the negative income tax has a lower cost, it will involve less increase in taxation for the middle income ranges.

Measures to change pre-redistribution income A third difference concerns the relative importance of income maintenance as opposed to measures designed to change pre-redistributon incomes. Since the social insurance approach can provide no solution for the working man with no children, such measures must be accompanied by changes in labour market policy. The supporters of the social insurance approach typically favour the introduction of more effective minimum wage legislation, coupled with active steps to improve the employment prospects of the poor (such as the government acting as an employer of last resort). On the other hand, many of the proponents of negative income taxation see it as a means of eliminating government interferences in the working of the market and envisage the abolition of minimum wage legislation.

Take-up and stigma The supporters of the social insurance approach attach great importance to securing 100 per cent take-up, and to the fact that ‘our insurance type programs have worked better and gained greater acceptance than either our public assistance programs or those designed to aid the working poor’ (US Department of Health, Education, and Welfare, 1969, p. 48). It seems safe to say that the social insurance approach would not encounter serious problems of incomplete take-up. On the other hand, with the negative income tax this is an open question, and it remains to be demonstrated that the scheme can be administered in a way which avoids the stigma of existing means-tested programmes.

Finally, whichever approach is adopted, it is widely agreed that new initiatives in the field of income maintenance are required. There is no reason to suppose that economic development and macro-economic policy can by themselves eliminate poverty, as was recognized by the Council of Economic Advisers: ‘we cannot leave the further [reductionl of poverty to the general progress of the economy’ (1964, p. 60). Such new initiatives are likely to involve a considerable cost. The Original Tobin proposals for a negative income tax would have cost $14 billion in 1962, or 2’4 per cent of the total gross national product. The British tax credit scheme had an estimated net cost of £1300 million, or almost the same percentage of gross national product. The cost of a social insurance alternative would be broadly similar. The elimination of poverty, as it has concerned government in recent years, requires a substantial transfer of resources.