A series of recent judicial decisions has focused public attention on the issue of local choice in the provision of public education. In Serrano vs. Priest, Rodriguez vs. San Antonio (1971), and similar cases in other states, the lower courts confirmed that education is a responsibility of the state government and held that local expenditures on education may not be a function of the taxable wealth of the local community.1 Although the United States Supreme Court has overturned these decisions in the appeal of Rodriguez vs. San Antonio (1973), the pressure to change the current system remains strong. The Supreme Court majority indicated that its decision reflected the limits of the federal constitutional authority and was not an approval of the status quo in educational finance. Litigation is now likely to shift to challenging the current methods as unconstitutional under state constitutions which, unlike the federal constitution, do deal specifically with education.2 Moreover, fundamental changes in the financing of local education may not require further pressure from the courts; state legislatures may seek to neutralize the effects of local wealth differences even if the current systems are not held to be unconstitutional. These judicial decisions and the ensuing legislative proposals run counter to the general economic view of local government finance. The basic presumption of economic analysis is that, because local governments can select different levels of service and because individuals can choose their area of residence, decentralized finance by local governments allows the provision of public services to reflect the variety of individual preferences for public services.3 Although the level of local spending may be nonoptimal because of intercommunity externalities and because of the method of local budget determination, fiscal decentralization still remains the only alternative to the insuperable problem of determining the optimal level of expenditure on a public service provided by a central government. In effect, autonomous decentralized financing of education provides a quasi market in which households can exercise their diverse preferences by their location decisions. This paper considers the problem of * Professor of economics, Harvard University. I am grateful to Charles Clotfelter for assistance with the statistical analysis, to Stephen Weiss for providing unpublished data on school expenditures, and to the Ford Foundation and National Science Foundation for financial support. I have benefited from discussions of an earlier version in seminars at Harvard, M.I.T., and Berkeley, and from comments by Noel Edelson, Eric Toder, and David Stern. An earlier and more complete discussion of this study was distributed as Harvard Institute of Economic Research paper no. 293, May 1973 (revised July 1973). 1 In Serrano vs. Priest, the landmark case in this area, the plaintiff and the courts were very much influenced by the line of argument and suggested remedies develope(l in John Coons et al. For a further discussion of the legal precedents, see Arthur Wise. 2 Almost immediately after the United States Supreme Court decision in Rodriguez vs. San Antonio, the New Jersey Supreme Court held that the current system of local finance violated the New Jersey state constitution. See Wise for a summary of the provisions of other state constitutions. I Charles Tiebout presented a formal analysis of the full efficiency of local government provision of public services under quite special conditions. See Wallace Oates and James Buchanan and Charles Goetz for a further discussion of these issues.

A full assessment of the long-run incidence of the social security tax requires answers to four questions. 1) How much does the tax alter the quantities of and capital supplied? 2) How do changes in factor supplies affect the marginal products of and capital? 3) How is the wage rate and the return on capital related to these mlarginal products? 4) How does the tax affect the relative prices of the goods consumed disproportionately by and by the owners of capital? Brittain's theoretical discussion deals with the supply of labor. His emprical analysis is concerned onlx with questions 2) and 3) and provides no information about the effect of the tax on factor supplies. Most analyses of the incidence of the payroll tax concentrate on the tax's effect on the supply of labor. A more elastic aggregate supply generally implies that, ceteris paribus, a smaller fraction of the burden falls on labor. Brittain concludes from his theoretical discussion that the tax does not change the quantity of supplied, and that the burden of the tax therefore falls on labor. Because of the tax's affects on the capital stock and on relative prices, the second does not follow even the first is true. However, I will concentrate on examining the basis of Brittain's conclusion that the quantity of is unaffected by the tax. If the aggregate supply were completely inelastic with respect to the wage rate, it is obvious that it would also be unaffected by the tax. The important question is the effect of the tax the supply is elastic. Brittain concludes that in this case the tax will not change the quantity of supplied labor bargains in terms of total (p. 114). The assumption that bargains in terms of total compensation implies that the quantity of supplied at each gross wage is unaffected by what fraction of the gross wage is paid in taxes, and that any tax change is therefore ignored by both employer and employee. Brittain accepts this extremely implausible assumption as the basis for his strong conclusion because of a more general error in his analysis. Although the question at issue should be the incidence of the entire payroll tax, Brittain concentrates his attention on the share paid by the employer. Moreover, he implicitly assumes that the employees' share of the tax is viewed by them as equivalent to income and therefore entirely borne by labor! He then concludes that, since it would be irrational to treat the two parts of the tax differently, the employers' share must also be borne by labor. More specifically, Brittain states that the employers' share of the tax would not be fullborne by only if the supply curve of were not perfectly inelastic and the supply price excluded the employer's tax (p. 115). He then rejects the latter condition in favor of the view that is indifferent between a higher net wage and a higher employer tax contribution. He argues that to believe otherwise depends on viewing one withheld tax as part of its income but not the other. This behavior (is) difficult to rationalize . . (p. 115). That is, since income taxes and the employees' portion of the payroll tax are, by implicit assumption, treated as net income and therefore borne by * Professor of economics, Harvard University.

The paper begins with a critical examination of four mistaken propositions that characterize much of the conventional wisdom on capital income taxation. The analysis then shows the correct approach to evaluating the welfare cost of alternative tax treatments of capital income when taxes affect both the supply of labor and the timing of consumption.

The classic example of an unshiftable tax is the general tax on pure rental income. Since Ricardo, economists have believed that the annual net rental income of unimproved land falls by the amount of the annual tax and its price by the capitalized value of this tax. This paper shows that these conclusions are false, that the tax on pure land rents is at least partly shifted, and that the price of land may be increased by the imposition of a tax. Implications are suggested for the analysis of the corporate income tax and the taxation of natural resources.

The typical worker who is laid off is soon rehired by his original employer. This important and generally unnoticed fact requires a major reevaluation of our current theories of unemployment. This paper develops a theory of temporary layoffs. Specific attention is given to the question of why employment is reduced instead of hours. The role of unemployment insurance and of taxes is examined in detail.