You are using an outdated browser.
Please upgrade your browser
and improve your visit to our site.
Skip Navigation

Taxation and Democratic Values

A case for redistributing income.

The use of taxing power to achieve a redistribution of income has been increasing over time in a marked way. Popular support of this doctrine has also tended to increase, though with much ebb and flow. The 1972 presidential campaign represented something of a check, though it was perhaps the first time the issue had been made a major one in that context and the ineptness of the presentation was surely a significant factor in McGovern’s defeat.

The philosophical and ethical foundations of redistribution have come under sharp debate in intellectual and academic circles as well. Economists, after much neglect, have revivified their interest in the axioms that underlie the distribution and redistribution of income. Perhaps the most important recent intellectual event in the area of distributive justice has been the publication of John Rawls’ A Theory of Justice, followed by attacks on Rawls and on the desirability of equality in general by Irving Kristol, Robert Nisbet and others.

Among other strands of the debate is the old idea of a conflict between liberty and equality. The debate occurs against the background of a presumed common acceptance of the ideals of a democratic society, and democracy clearly includes in its sayings both these ideals. I want to argue here that a commitment to democratic values strongly implies an ideal of redistribution of income and wealth.

I think we may safely agree that the notion of democracy has two components, both indispensable: 1) the securing of the freedom of the individual so that he may develop his individual potential; 2) a symmetric mutual respect of the individuals in the society for each other. These aims are, as has been frequently remarked, partly competitive; but, it must also be stressed, they are to a very considerable extent complementary. A hierarchical society marked by great inequalities in power and esteem will surely not tolerate the liberties of those most disadvantaged. Conversely a world in which individuals have their liberties tightly confined must be one in which there are large inequalities of power.

I would like to enlarge somewhat on the meaning of the second criterion, the symmetry of respect. Any society, democratic or not, must have as a root element some degree of mutual obligation and some sense of respect for every individual, however low he may wind up in the hierarchy. The strong element of truth in all social contract theories is that the Hobbesian nightmare can be avoided only if the bulk of society feels in fact incorporated and accepted in it.

What is distinctive about the democratic society is that mutual relations are conceived of as symmetric in spirit. Individuals, at least in some initial sense, have to be treated equally by the society, if it is to be worthy of the name “democratic.” One can attempt to give deeper philosophical foundations for this symmetry, perhaps Rawls’ idea of an “original position,” in which society chooses its values at a point where the members do not know what roles they will play. More prosaically, we can think of democratic decision-making procedures as a kind of insurance arrangement, in which individuals do not know what interests they will have in the long future, a view developed by James Buchanan, for example.

An immediate implication of the symmetry assumption is clearly some form of egalitarianism. Classically this has been interpreted in a political sense, that of equality of the right to vote. A long history of struggles has led to an almost universal acceptance of this democratic principle, at least on paper.

It might be objected that political democracy is in fact only nominal, that this ideal is never realized. As we look around we see rather obviously that political power is in practice exercised by a small minority. But the implications of this are perhaps less obvious. There are two reasons for the operation of “the iron law of oligarchy”: 1) a straightforward and inescapable need for specialization of function; and 2) the translation into political power of inequalities in other forms of power, especially, in the modem world, economic power. I want to return to the second point later. The first of course is simply the fact that everyone can’t do everything; politics is a vocation and you can only expect a limited number of people to engage in it, for obvious economic reasons, just as we have specialization in other occupations. But in principle, the practical importance of the specialization and of the hierarchical political structures needed for efficiency is blunted by the fact of political competition. Like economic competition it implies that the concentrations of power apparent to the eye are less significant than they seem; the politician or the entrepreneur has to satisfy the customers or voters. Of course these checks are imperfect, and perhaps more so for political than for economic competition; but they are real and important.

Assume then a roughly viable democratic system in the political sphere. Let us take the basic structure of decision-making in the economic sphere as it is today. Most economic decisions are made in the private sector, but the government has a considerable hand in the allocation of resources and in regulating the operations of the private sector. The division of responsibility can be argued in terms of both efficiency and ethical considerations and will not be discussed here. The operations of the government require, then, control of resources toward the ends that it is supposed to satisfy. These ends have been mainly the production of public goods but increasingly also redistributive measures, payments to individuals regarded as in some way disadvantaged. The latter amount to negative taxes.

Many among both the intellectual and general publics regard redistribution as a deprivation of liberty, in that property is taken away by coercion. I disagree with this viewpoint most strongly. It ignores the freedom-enhancing value of redistribution to the recipients, and it erects property unwarrantedly into a fundamental institution. Let me develop these arguments a little more fully.

In essence the ideal of redistribution is an extension of the principles of symmetry of respect and obligation from the political sphere to the economic. The symmetric regard for others is a desire to have them treated like oneself. In the absence of good reasons the presumption is always for equality of treatment, and equality of treatment will imply equality of outcome, again in the absence of special reasons to the contrary. One can indeed argue convincingly that complete equality can be achieved only at the cost of other values that we are not prepared to sacrifice. But one cannot start with the idea that a given state of inequality needs no question.

Let us start with elementals. Income and property are certainly the instruments of an individual’s freedom. Clearly the domain of choice is enhanced by increases in those dimensions. It is true not merely in the sense of expanded consumer choice but also in broader contexts of career and opportunity to pursue one’s own aims and to develop one’s own potential. Unequal distribution of property and of income is inherently an unequal distribution of freedom. Thus a redistribution of income, to the extent that it reduces the freedom of the rich, equally increases that of the poor. Their control of their lives is increased.

There is an interesting sidelight on this in one of the outcomes of the experiments on the so-called negative income tax. The one kind of labor that is withdrawn from the markets in significant amounts is that of young adolescents; the income transfers induced them to return to their education rather than take low paying jobs. This is precisely the kind of increased freedom that the most ardent libertarian should favor. A democratic position that starts from the equal importance of all individuals can only applaud a move to a more egalitarian distribution of autonomy and liberty.

A great deal of intellectual confusion comes, in my view, from thinking of property rights as basic and axiomatic. The forms of private property are indeed capable of mystification. In a world where incomes are derived in the first place from the sale of labor and other assets, the use of resources by the government takes the form of diverting income and wealth out of private hands into public ones. It therefore presents itself as an infringement on freedom. This is only a matter of forms of payment.

One could equally well think of the world’s resources as a common pool from which allocations to private and public uses have to be made. We could make some payments in the form of rewards for services rendered and some in the form of grants to meet needs. Then the redistributive payments would not appear on their face to be a deprivation, but the effect in real terms would be the same.

I do not want to be misunderstood. Property, like the government, is an extraordinarily useful institution. Property is a protector of liberty. Equally important, the rights of property afford some guarantee for the efficient use of resources. But the current distribution of income and wealth has no particular claim to priority in itself.

There is some argument in justice and in practical efficiency for predictability, therefore an argument against arbitrary, large redistributions. In fact, however, the present capitalist system by its essence has a strong degree of unpredictability; for example destruction of property through development of someone else’s innovation is a frequent occurrence and one that we regard as essential to the interests of progress.

There is one serious objection to relativizing and pragmatizing the rights of property. The most important items of property are personal. They are the skills and talents of individuals. It can certainly be argued that we do not want to put such assets into the common pool, that somehow each individual has the right to development of his own personal attributes. I think one can make a fairly sharp differentiation between the exercise of personal attributes and the enjoyment of unlimited income rights therefrom. I would be tempted to say that an individual with unusual talents, medicine or art, has something of an obligation to the world to use these talents; but even apart from that, I see no argument against saying that if he is highly rewarded for applying his talent, those rewards should in part be devoted to the less fortunately endowed.

There is a very different line of argument for redistribution, also based on the preservation of democratic values. This is the one referred to briefly above, the fact that unequal economic power is bound to be translated into political power. The mechanisms that this transformation takes are no secret. Campaign contributions, the many rewards that can be conveyed by contributions of wealth to complaisant political leaders, the ability to control communication, the greater leisure and the ability to be informed, all of these skew the distribution of political power.

The aim of achieving an equal distribution of political power requires a restriction on the inequalities of wealth and income.

The two arguments presented thus far have some what different implications for the nature of the redistribution. The expansion of the freedom of the poorer individuals leads mainly to anti-poverty measures, to operations on the lower end of the income distribution. Of course mere arithmetic requires that this have an effect on the upper tail, but it is sometimes argued that the very high incomes (say about $100,000 a year) have such a small fraction of total income in them that they play a minor role in redistribution to the lower end. On the other hand the preservation of political equality may require especially strong redistributive measures against the upper end of the distribution. Of course there is no conflict between these two goals.

There is a third, more technical, argument. It is that a more equal redistribution of income will lead to an improved efficiency in expenditures on public goods. As matters stand now, such expenditures are partly directed toward improving efficiency in the economy but also partly directed toward redistributive aims. If a more or less equal distribution of income were achieved, one could confine the role of public goods to pure efficiency considerations.

There are of course well known arguments toward limiting the extent of redistribution. The most familiar is that redistribution is bound to be accompanied by strong adverse incentive effects. Developers of new enterprises will be inhibited if their rewards are heavily taxed. Workers will not shift jobs as readily if the wage differentials are taxed away. These considerations are real and important and do set limits on the extent to which the income tax, for example, can be used for redistributive purposes. But one must not be too cautious. It is surely true that a great deal of very high personal incomes are essentially rewards for shifting incomes from one person’s hands into another’s, as in speculation or competitive practices that have little social product. Furthermore they are in good measure a reward for exercising talents that are in themselves pleasant to exercise and will therefore continue to be used even at considerably lower reward levels. It is also possible to consider taxes other than income taxes as instruments of redistribution. One that has been frequently proposed is a progressive tax on an individual’s total consumption, thus providing an incentive for saving. There are severe administrative and conceptual difficulties, but it is possible that these are by no means insuperable.

Another argument against redistribution, which I think is less serious, is that anything that enhances the role of government is bound to increase the concentration of political power. This argument presupposes that the political competition, discussed above, is ineffective or imperfectly effective. Even to the extent that this is true, a redistribution based on an automatic tax-like law, rather than on the basis of individual applications, should not result in any real increase in the effective power of the state.

I would therefore argue the scope for redistribution, as manifest by progressive taxation extending into the negative area of lower incomes, should be high on the agenda of those desiring the expansion of democratic values.

Kenneth Arrow is James Bryant Conant University Professor at Harvard and recipient of the 1972 Nobel Prize in economics.