Knowledge that Transforms

The theories of saving presented by Keynes, Duesenberry, and Friedman use three different measures of income to explain saving, and each theory has been supported by empirical evidence. The evidence consists of tests of a wide variety of hypotheses contained within or derived from the theories. The tests use many different kinds of data. The objective of this paper is to submit the basic behavioral hypotheses of each theory to a common test on constant data. The results show that the theories are equally acceptable on empirical grounds.

FOREIGN economic assistance in the form of commodities and the sale proceeds in domestic currency of those commodities have assisted post-war recovery of West European countries and economic development of many under-developed countries like India. The real significance of the accrual and disbursement of these funds, perhaps, lies in their impact on money supply. Despite its obvious importance, however, the problem of the money supply impact of counterpart funds did not receive as much attention as it deserved, although whether use of such funds is inflationary or not was discussed to some extent.1 India, which has been receiving a large and increasing commodity assistance from the United States in recent years under P.L. 480, provides an interesting example of how variations in money supply are influenced by the accrual and disbursement of counterpart funds and how they tend to influence monetary and debt management policies. The main purpose of this article is to spell out the money supply impact of accumulation and disbursement of the counterpart funds and to indicate broadly its significance for monetary and fiscal policies pursued in India.2 The practice adopted in India for holding counterpart funds has not remained uniform. Between I956, when the foreign assistance in the form of commodities from the United States under P.L. 480 first started coming in, and the middle of I960, the counterpart funds have been held with one of the commercial banks. Since then, however, they have been held with the Reserve Bank of India, which is the central bank of the country. While discussing the Indian experience, we first concentrate mainly on the money supply impact of the accumulation of counterpart funds with a commercial bank as it has acquired greater importance and raised considerable debate.3 The latter practice has been adopted only recently; the likely consequences of the changed procedure are analyzed separately in the latter part of this paper.

IT is a commonplace that inadequate and incomplete data prevent a precise quantitative statement of the secular trend of functional income shares generated by the United States economy. Nevertheless, approximations are possible, and it is on the basis of an approximation that investigators such as I. B. Kravis,' and D. G. Johnson,2 have concluded that labor's relative share in national income has been increasing since the beginning of the twentieth century.3 With this trend as a premise, the fascinating task of explanation presents itself, replete with opportunities for elegant theorizing. Efforts in this direction have already been made by Kravis and Johnson, as well as by R. Solow,4 and M. Bronfenbrenner,5 among others. This paper seeks to stem the explanatory tide, as it were, by showing that the supposed shift in income distribution well may be an illusory one, resulting primarily from procedures and methodology, rather than from real changes in structure. The methodological framework within which the shift has been demonstrated is subject to attack on two general grounds. First, national income, as defined by Commerce, is an unsuitable base upon which to compute labor's relative share. Inconsistencies among the comDonents of this denominator render changes in the ratio's value over time misleading when applied to the structure of income distribution. Second, the empirical evidence itself has been distorted by the manner in which earlier income estimates have been reconciled with Commerce figures in order to form consistent series. While any set of reconciliations inevitably contains elements of arbitrariness, those which have been employed are arbitrary than is really necessary. The first section of this paper develops these particular criticisms in detail. In the second section, alternative data are presented which, it is believed, provide an improved and more meaningful measurement of the relative labor share, and which indicate a startling lack of change in that share between I899 and I929. This finding is based upon the ratio of employee compensation paid by domestic private business enterprises to the product generated by this sector of the economy. Coupled with S. Weintraub's demonstration of the stability of the employees' share in business gross product (Commerce concept) for I929-I957,6 these results contrast sharply with the JohnsonKravis finding that employee compensation rose from 55.0 per cent of national income for the period I900-09 to 67.I per cent for I949-57.7

Hyman P. Minsky, Arthur M. Okun, Clark Warburton, Comments on Friedman's and Schwartz' Money and the Business Cycles, The Review of Economics and Statistics, Vol. 45, No. 1, Part 2, Supplement (Feb., 1963), pp. 64-78

AMONG the most elusive magnitudes to ,[_1quantify is the influence of technological change on employment, the reason being that technological change in any context has been difficult to isolate. Clearly, the traditional productivity ratios, such as output per unit of labor input, cannot be used to measure technological employment, since a productivity index embodies, in a seemingly indecomposable manner, the effects of in capital utilized, returns to scale, neutral and non-neutral technological change, and relative factor prices. Thus, in order to construct a measure of technological employment, we need to be able to quantify, at the minimum, the effects of in technology separately from the other forces. Yet, these other forces have meaning in themselves. Therefore, we should like to isolate the effect on the change in employment of in the following: (a) the scale of output, (b) the relative prices of capital and labor (assuming, for simplicity, only two factors), (c) returns to scale, and (d) neutral and non-neutral technology.' The present paper presents a method of measuring the forces (a)-(d) on employment and tests it on data for the private domestic non-farm sector of the United States for the period 1890-1958.2 It does this in such a way as to avoid the problem of the interaction among the forces (a)-(d) -at least to a first-order approximation. Stated differently, our objective is to frame a general method of measurement which permits a quantitative distinction to be drawn between structural changes and demand in terms of their effect on employment.8 Since the method is general, the forces (a)-(d) can be quantified for any subset in the total employed; for example, skilled or unskilled labor, regional unemployment, etc. The only requirement is that we be able to estimate a demand relation for the subset in question. We should like to emphasize the methodological rather than the substantive aspects of this paper for several reasons. The principal reason, though, is that the data, by virtue of their aggregative nature (inter alia), are not suitable to the method we will apply. Also, since the method derives from the micro theory of the firm, it should be applied, at most, to industry data. In what follows, the method is first presented verbally as far as possible. This is followed by a more precise statement of the method which permits a confrontation with data. The empirical measures of the private domestic non-farm sector are then set out and discussed. An appendix embodies a discussion of the data used in the paper.

Kenneth J. Arrow, Phillip Cagan, Irwin Friend, Comments on Duesenberry's The Portfolio Approach to the Demand for Money and Other Assets, The Review of Economics and Statistics, Vol. 45, No. 1, Part 2, Supplement (Feb., 1963), pp. 24-31

T HE theory of the demand for financial assets has come in for a good deal of discussion in the last few years. Undoubtedly the discussion has been fruitful and has given us many new insights into the nature of financial processes. But it cannot be said that there is any generally agreed upon view as to the way in which those processes work. It would be appropriate at a conference of this kind to review the different hypotheses and give a systematic summary of the present state of knowledge. Unfortunately, though I have read the literature assiduously I have found it rather indigestible. I do not feel prepared to give a fair summary of other people's views. I must fall back therefore on giving my own. In this paper I shall deal with the demand for liquid assets and money by households and corporations. Those two groups hold over twothirds of all liquid assets, and the same general approach though not the details can probably be applied to the demands of unincorporated businesses, farmers, state and local governments. In dealing with the demand for liquid assets we must at least implicitly deal with the demand for other types of assets, but I shall not, except incidentally, say anything in detail about the demand for stocks, bonds, or physical assets. I shall confine myself to the demand for currency, demand deposits, commercial bank time deposits, mutual savings bank deposits, savings and loan shares, savings bonds, and short-term federal securities. There are, of course, other liquid assets, but I shall have little to say about them. I have occasionally used the term money in the sense of demand deposits and currency but have usually referred to those assets specifically to avoid any confusion with other definitions of money. But though I am happy to try to avoid the semantic confusion involved in arguments about whether any particular asset should be included under the heading money, I do cling to the view that commercial bank time deposits are significantly different from demand deposits. For that matter, so is currency, and so perhaps we ought to dispense with the term money in theoretical discussions and say clearly what we mean. In the first section of the paper I have discussed very briefly the conditions under which liquid assets are supplied. There follow in section 2 a discussion of corporate motives for holding liquid assets and money and a review of some empirical evidence on the relative importance of various factors influencing corporate decisions. In section 3, this theory of household demand for liquid assets and money is discussed together with some empirical evidence.

N EW estimates of hours and earnings indicate greater declines in hours of work and greater increases in hourly earnings the past half century than the usually used, currently available historical series show. The new estimates presented in this paper are hours of work per week in the manufacturing, railroad, bituminous, and anthracite coal mining industries, as well as hourly earnings and hourly compensation in the latter three industries. The purpose in presenting the new hours and earnings series is threefold. (1) For historical data on hours and earnings, economists generally refer to the series published by the Bureau of Labor Statistics.' The conceptual basis underlying the calculation of the Bureau's series is one of total time paid for. For the researcher who is interested in problems relating to long-run changes in the hours of work, the introduction in recent years of such practices as paid vacations, paid holidays, or payment travel time gives an upward bias to an hours-per-week series when measured on a total time paid for basis. Since these practices provide additional earnings, division by the additional hours paid for, according to the total time paid for concept, tends to cancel out these gains and thus leads to understatement of the historical increase in hourly earnings. The new series attempt to correct these biases by deducting the time accounted by these practices from the measured hours both in the calculation of hours of work per week and in the calculation of hourly earnings. (2) The new series provide annual data continuously from 1900 through 1957, thus supplying information time periods since 1900 that are not covered by the Bureau's series (Tables 1 and 2 below). (3) The comprehensive review of the historical evidence on hours of work, which was undertaken in constructing the new series, suggests that the level of the Bureau's oft-used manufacturing series is in error during the 1920's. In 1929, a bench mark year the manufacturing series presented in this paper, the new estimate of hours of work per week is 48.0; the level of the Bureau's series is 44.2 hours. The construction of the hourly compensation series extends to three more industries the compensation concept developed by Rees manufacturing, to take account of the increasing importance to employees of employer contributions to such programs as pension plans and social insurance.2 Total hourly compensation is the sum of these wage supplement payments and the direct wage payments to workers measured by average hourly earnings. The employee coverage of the new series is identical to that of the Bureau's historical series. Except the railroad series, employee coverage is limited to production workers. The employee coverage of the railroad series is somewhat broader, including all wage and salary workers except the upper echelon of management which is classified as officials, and staff 3 Tables 1 (hours) and 2 (earnings and compensation) below present the new series together with, comparative purposes, the currently official historical series of the Bureau of * This paper is adapted from Chapter II and Appendix B of my Ph.D. dissertation, Hours of Work in the United States, 1900-1957 (unpublished Ph.D. dissertation, Department of Economics, University of Chicago, 1961). My thanks are due to both H. Gregg Lewis and Albert Rees their assistance on the dissertation and their helpful suggestions on an earlier draft of this paper. 1See, e.g., U.S. Bureau of the Census, Historical Statistics of the United States, Colonial Times to 1957 (Washington, 1960), 92-3; and U.S. Bureau of Labor Statistics, Employment and Earnings Statistics the United States, 1909-60, Bull. No. 1312 (Washington, 1961), 14, 31, and 422. 2Albert Rees, New Measures of Wage-Earner Compensation in Manufacturing, 1914-57, National Bureau of Economic Research, Occasional Paper 75 (New York, 1960). 'The new series exclude only the reporting divisions (1) executives, general officers, and assistants and (2) division officers, assistants, and staff assistants. See U. S. Interstate Commerce Commission, Monthly Report of Employees, Service, and Compensation, Forms A and B.

C. E. Ferguson, Cross-Section Production Functions and the Elasticity of Substitution in American Manufacturing Industry, The Review of Economics and Statistics, Vol. 45, No. 3 (Aug., 1963), pp. 305-313