Search

ONE question that arises following recent I attempts to introduce economic theory in explaining income distribution is to what extent schooling can be used as a policy instrument to promote equality in our society. Although the link between the size of individual earnings and education has been rather well established, the change in the distribution of income among individuals resulting from a change in the level of schooling is not yet settled. In this paper we elaborate on the relationship between schooling and income distribution based on human capital theory. In order to answer the question we need an expression relating a measure of income distribution to a measure of education. It would be convenient in this respect to start from the work of what we will call the Becker-MincerChiswick (B-M-C) group,1 as perhaps the best known of those providing such an expression. In their work the level of earnings (Y) of an individual with S years of schooling is determined by an earnings generating function of the form

Economists and tax lawyers have long debated the efficacy and propriety of the income tax deduction for charitable contributions.' The effect of the deduction is to lower the individual's net cost of giving if he itemizes his deductions. More specifically, the net cost to the donor per dollar received by the charitable donee is equal to one minus the individual's marginal tax rate.2 If the elasticity of total giving with respect to this price (or net cost) is absolutely greater than one, the charitable deduction causes donees to receive more in additional gifts than the Treasury forgoes in revenue. Alternatively, if the price elasticity is absolutely less than one, the deduction is less than fully efficient in this sense. In a series of recent papers, Feldstein and his collaborators (1975a; 1975b; Feldstein and Clotfelter, 1976; Feldstein and Taylor, 1976) obtained estimates of the price elasticity that cluster around 1.2 from a variety of different data sources. All but one of these studies (Feldstein and Clotfelter, 1976) are based on the gifts of only those taxpayers who itemize their deductions. Since substantially more than half of the households either do not itemize deductions or do not file any tax return, the estimated price elasticities have been obtained primarily from the top half of the income distribution. While this part of the population accounts for a disproportionate share of charitable contributions, extrapolation to the entire population may not be warranted. A variety of policy proposals that are currently being considered, e.g., a tax credit for all taxpayers for charitable gifts or extension of the charitable deduction to non-itemizers, would alter the price of giving for households that do not now itemize. An accurate estimate of the price elasticity for this income group is required to predict the effects of such policies. The results presented in the current paper indicate that households with income under $30,000 are very sensitive to taxinduced variations in the cost of giving; the estimated price elasticities generally exceed two.

While this article was in press, the Editors were shocked to learn of the death of Dr. Ernest M. Doblin on July I4. They take this opportunity to pay tribute to the important work which Dr. Doblin did in many fields of economics and especially in the area of national income. The availability of income statistics as well as their improvement owe a great deal to Dr. Doblin. The following article stands as it was printed from manuscript without the benefit of author's corrections in proof.

M ANY attempts have been made to estimate the national income of economically backward countries. The Statistical Office of the United Nations has assembled estimates of 39 countries, roughly half of which are countries with underdeveloped economies.1 And several projects are underway for making other and better estimates. But it may be asked to what extent these efforts are justified. In making estimates of Japanese national income some years ago, the writer frequently found difficulties in applying national income and product concepts to backward areas of the economy. The problem is not so much the degree of accuracy of the estimates involved, since these estimates are usually presented to convey orders of magnitude and nothing else. Rather, the problem is: Do national income and national product exist in the backward areas? (In the sense analogous to the question: Did national states exist in Western Europe before mercantilism?) One national income economist, very much puzzled by the results of the estimates for underdeveloped economies, has posed one aspect of the problem as follows: "What meaning is to be attached to per capita income of the African population, including income in kind, in the neighborhood of ?2 a year? The reviewer has never been able to grasp figures of this magnitude, which are common for countries with much of the world's population" (italics added).2 It should be noted that the difficulty in this case is not due to inadequacies of exchange rate conversion; the currency unit of the African colony happened to be the British pound. The ?2 per capita income is an average. This means that perhaps half of the population lived on incomes between zero and ?2, while a substantial portion lived on incomes between zero and Li throughout the year. These strange results point to the possible existence of certain difficulties in the concepts, if the possibility of inadequate estimates and exchange rates is ruled out. To avoid confusion, it is necessary at the outset to distinguish for present purposes two of the major uses of national income estimates. First, there is the use of the ratios of the various national income components to the total national income of a given year (or to other components); for example, such ratios as one state's income to that of the nation (or another state); of agricultural income to the national income (or to manufacturing); of wages to the national income (or to property incomes); of investment to national income (or to consumption). The second use is the measurement of changes over time (or inter-spatially) of either the total national income or the various composition or structural ratios mentioned in the first use. For brevity and convenience, these uses may be designated as the compositional or structural use, and the index number use, respectively. This paper deals with the problems of the first or compositional use. The problems of the second use are essentially no different from those encountered in the construction of index numbers of physical production and as such they are best discussed under that topic. One of the reasons for the extensive and peculiar usefulness of national income statistics stems from the rough equivalence of actual (or accrued) net income to net value product. The national income totals estimated for the second use do not have this quality. But to the extent that the thesis of this paper has relevance for the second use, some brief comments are made at the end of this paper.

R EAL disposable income is the measure generally accepted by economists as an indicator of the welfare of a household. Were we living in a world in which each individual formed a household and in which everyone required exactly the same goods to maintain himself, real disposable income might be an adequate measure. However, households are generally complex, and the income of a household must support a number of individuals whose daily requirements range from the nominal demands of a young child to the ample needs of an adolescent. Thus the welfare of a household must be measured not only by its income but also by the size of the household and the variety of needs of its members. This paper discusses a measure of welfare which relates the resources of a household to its needs. Estimates of welfare levels in the United States population based on this measure show that disposable income and per capita disposable income are extremely crude measures of welfare. Both overstate or understate the standard of living of an appreciable fraction of the United States population. The paper concludes with a few notes on the need for more accurate measures of welfare.

IN economic welfare terminology, the social benefits of education result in an outward extension of society's utility possibility function.' These can be divided into current and future benefits with some accruing directly to the family whose child is being educated, while other benefits are mainly indirect. It is likely that of all the benefits of education, one of the most important is the gain in productivity and earnings of students. Public education increases the knowledge and skills of students and, therefore, tends to increase their potential earnings. While there is general agreement that the level of education attainment has a positive effect on income, relatively little is known about the magnitude of this relationship when differences due to other factors are taken into account.2 It is to this problem that we will address ourselves in this paper. Our procedure will be to advance a hypothesis about those factors, including years of schooling, affecting a person's income. We will then test this hypothesis by submitting survey data to a multivariate analysis, thus obtaining, in a case study, estimates of the net relation between education and future income. These findings will be used to estimate incremental lifetime earnings resulting from education and their present values, so that comparisons can be made between these benefits and corresponding annual education costs. Finally, some estimates of the rates of return of education investments will be made.