¹ Grodzins, Morton, “The Federal System,” in the President's Commission on National Goals, Goals for Americans (New York, 1960), pp. 265 ffGoogle Scholar; “American Political Parties and the American System,” Western Political Quarterly, Vol. 13 (12, 1960), pp. 974–78CrossRefGoogle Scholar; “Why Decentralization by Order Won't Work,” in Banfield, Edward C., ed., Urban Government (New York, 1961), pp. 122–31Google Scholar.

² The discussion of economic sharing is restricted here to these two planes. This simplification was necessary (both for the devising and evaluation of tests of sharing) because the economic data essential to the analysis are for the most part not separately available for states and localities.

³ See Hansen, Alvin H. and Perloff, Harvey S., State and Local Finance in the National Economy (New York, 1944)Google Scholar; Newcomer, Mabel, “State and Local Financing in Relation to Economic Fluctuations,” National Tax Journal, Vol. 7 (06, 1954), and references cited in footnote 2, p. 98 Google Scholar; Brown, E. C., “Fiscal Policy in the ‘Thirties: A Reappraisal,” American Economic Review, Vol. 46 (12, 1956), pp. 857ff.Google Scholar

⁴ By “conventional,” we refer to the treatment in national income accounting. See Department of Commerce, National Income, 1954 (Washington, D. C., 1954), Table IV, p. 161 Google Scholar. Cf. Copeland, Morris A., Trends in Government Financing (Princeton University Press, 1961)CrossRefGoogle Scholar.

⁵ For an extensive discussion of this social accounting system, see Board of Governors of the Federal Reserve System, Flow of Funds in the United Slates, 1939–1953 (Washington, D. C., 1955), esp. ch. 1Google Scholar. The flow of funds accounts have since undergone substantial revision. The discussion in this paper is oriented to the first major version of the accounts.

⁶ The conversion of direct financing sources into purchases of goods and services or the conversion of indirect financing sources into direct financing uses are examples of moneyflows being moved forward towards the GNP. Moneyflows can also be moved varying distances away from the GNP, or they can be moved “sideways.” A move away is illustrated by a government using tax monies to increase its cash holdings. Here tax money (considered as potential expenditures in the GNP) is moved from the GNP. A “sideways” moneyflow is illustrated by a government which has borrowed funds and loaned them to a second government which, in turn, uses the loan to increase its cash holdings. Here the first government has contributed neither positively nor negatively to the direction of moneyflows.

⁷ We realize that the specific weights seem to be arbitrarily chosen. Any multiple of these weights will serve equally well. One advantage of the weights selected is that the GNP is given its natural weight of one. The weights fit the assumptions (1) that every positive conversion effect, described in the text, moves moneyflows an equal distance toward the GNP and (2) equal distances represent equal net contributions. Different assumptions of course would produce different weights. In one instance in particular, the weights assigned might be questioned. Using the reasoning of the balanced budget multiplier, expenditures on goods and services by the government would be given a heavier weight than tax receipts. We have not done this for reasons of simplicity.

⁸ The data sources were provided by Flow of Funds, op. cit., for the years 1939–50, and for 1950–56 by “Plow of Funds Sector and Transaction Accounts, 1950–56,” mimeographed (Washington, D. C., 1958).

In previous studies for other (non-governmental) sectors of the economy some alternative techniques were followed in estimating direct and indirect financing transactions. They were estimated on the basis of the nature of the sector receiving funds—whether it was an “ultimate” sector or a financial sector. Since it was assumed that all financial sources of nonfinancial sectors would be devoted to real uses and thus were “direct” financing sources, it was not as refined a technique as the present one. See Cohen, Jacob, “A Moneyflows Approach to Consumer Behavior,” The Southern Economic Journal, Vol. 23 (01, 1957), p. 245 CrossRefGoogle Scholar; “The Treatment of the Means of Payment in Social Accounting,” The Journal of Finance, Vol. 12 (12, 1957), p. 423 Google Scholar.

⁹ The flow-of-funds accounts, in one respect, apply different accounting procedures to the federal and state-local sectors. The federal sector is presented on a consolidated basis, the state-local sector on a combined basis. This means that intra-federal transactions (mainly financial) are cancelled out while intra state-local transactions are not. The effect on weighted deficits and surpluses is somewhat to favor state and local positive contributions.

¹⁰ In addition to the federal and state-local sectors eleven other sectors were analyzed: consumer, nonfarm noncorporate, corporate business, farm, banking, insurance, mutual savings banks, savings and loan associations, financial institutions not elsewhere classified, nonprofit organizations and the rest of the world. The last sector represents the external transactions of the United States.

¹¹ The apportionment process for identifying direct and indirect uses is a procedure of necessity because of the lack of precise information about the flow of funds among sectors. Where more precise information was available, as in the important grant-in-aid (“grants and donations”) category, in large measure it was possible to estimate direct and indirect financing uses directly from the pub¬lished data.

¹² This signal of inflation and deflation has the virtue of simplicity. We recognize its failure to consider other significant aspects of deflation or inflation. For example, it does not take account of price-level changes, nor does it define a “side-wise” relation when neither inflationary nor deflationary pressures are clearly dominant (e.g., unemployment rates between 3 and 5 per cent). Cf. the discussion in Money and Credil: Their Influence on Jobs, Prices and Growth (Report of the Commission on Money and Credit, Englewood Cliffs, 1961), ch. 9.Google Scholar

¹³ For a discussion of estimation procedures see Conrad, Alfred H., “Redistribution through Government Budgets in the United States, 1950,” in Peacock, Alan T. (ed.), Income Redistribution and Social Policy (London, 1954)Google Scholar; Adler, John H., “The Fiscal System, the Distribution of Income, and Public Welfare,” in Poole, Kenyon E. (ed.), Fiscal Policies in the American Economy (New York, 1951)Google Scholar.

¹⁴ On the Lorenz curve and the calculation of Gini coefficients, see Conrad, op. cit., pp. 241 ff.; Mendershausen, Horst, Changes in Income Distribution During the Great Depression (New York, 1946), App. C, pp. 162 ff.Google Scholar

¹⁵ The formula used by Mendershausen, op. cit., and followed here is:

where r_i equals the percentage of income receivers in a given income class, q_i equals the cumulated per cent of the distributed variable paid or received by this income class.

¹⁶ The intensive study by Conrad of distributional effects in the year 1950 provided the tax and benefit distributions used to calculate these coefficients. See Conrad, “Redistribution through Government Budgets,” op. cit., mainly Table II, p. 205, Table III, pp. 214–15. For reasons of space the various criteria (“incidence bases”) used by Conrad in distributing individual taxes and benefits among income classes have been omitted. The incidence bases for benefits can be found-in App. 1, Table I, p. 226 when used in conjunction with Table III, pp. 214–15.

Conrad provides no separate distributions of taxes for the federal and state-local governments in his Table II. These have been spelled out in correspondence. Except for social security taxes, similar percentage distributions of taxes among income classes are assumed by Conrad for like federal and state-local taxes. The incidence bases are discussed in his App. 1, pp. 224 ff.

¹⁷ It should be emphasized that our results may differ from studies based on data from other years. See for example, Musgrave, R. A., et al., “Distribution of Tax Payments by Income Groups: A Case Study for 1948,” National Tax Journal, Vol. 4 (03, 1951)Google Scholar.

¹⁸ Furthermore, while effective tax rates have greatly increased since 1922, the 1922 tax rates suggest that tax incidence by income class has not greatly changed. See The Federal Revenue System: Facts and Problems (Washington, 1956), Table 16, p. 151 Google Scholar. Some additional information on the similarities of tax distributions over time is offered in a study that concludes that the over-all burden of taxes was essentially unchanged between 1938–39 and 1946–47 despite numerous and important changes in the federal tax structure. (Adler, op. cit., p. 379.)