Copyright by Daniel M. Holland 195? THE TAXATION OF CORPORATE EARNINGS VIEKED AS PERSONAL INCOME, 1935-1947 Submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy, in the Faculty of Political Science, Columbia University Daniel Mark Holland \ Acknowledgment My biggest debt is to Professor Robert M. Haic^of Columbia University, who suggested and supervised the investigation, and to Professor Lawrence H. Seltzer, of wayne University, who, in charge of the fiscal studies at the National Bureau of Economic Research during the period of my research associateship there, exercised a continuing interest in and influence on its development. At the National Bureau, particularly helpful advice and encourage­ ment was obtained from Professors A. F. Burns and Leo Wolman, Thor Hultgren, George Lent, and Geoffrey Moore. Bella Shapiro placed her exhaustive knowledge of the data at my disposal and was a significant source of help in all phases of the study. My thanks to Gloria Mos cov.it z and Doris Eissman, who performed a substantial part of the computations. I have benefitted from the thorough and incisive criticisms of Professors Carl S. Shoup jfiad Yfilliam Vickrey, of Columbia University. Valuable advice was also received from Professors A. G. Hart and C. Lowell Harris s' of that same institution. Professor W. L. Crum, of the University of California, very graciously furnished me an advance copy of a stimulating article embodying some of the results of the work of many years in this samp general area. Thomas R. Atkinson, of the National Bureau, responded tq an idle question with a series of sxjecial tabulations from the”1Visconsin income tax data with which he was working; I am deeply grateful for both the data and his illuminating comments thereon. Ail of the names cited above can be involved only with the merits of this study (whatever they may be); all else is strictly of my own doing. Daniel M. Holland ii The Taxation of Corporation Earnings Viewed as Personal Income^ 1935-1947 Table of Contents Page Chapter I - Introduction 1 Chapter II - A Summary of the Method 19 Chapter III - The Findings for Each Year General Introduction to the Annual Data 30 A "Warning 49 1935 ' 53 1936 66 1937 72 1938 77 1939 82 1940 88 1941 93 Note on 1942 and 1943 99 1942 100 1943 - 109 Note on the Basis of Classification 116 1944 118 1945 124 1946 Usual Method 130 Alternative Assumption 142 1947 149 Chapter IV - The Picture Over Time 159 The Differential Against Net Corporate Earnings 160 The Differential Against Fully Imputed Adjusted Gross Income 183 Effect of Corporation Income Taxes on Progressivity of the Income Tax System 202 Chapter V - The Revenue Implication of the "Partnership" Method 212 1932 213 1946 218 Chapter VI - The Results for 1935-1947 Considered as a Single Period 239 Chapter VII - A Summary of Findings 246 iii Appendix A A Description of the Methods Page I Where the Data Came From 1 II Fitting Corporate Earnings into a PersonalI ncome Framework 5 1 The Basic Assumption 5 2 The Allocation of Net Corporate Earnings to Net Income Classes 10 a Determination of the Total to be Allocated 10 b The Income Class Breakdown £9 III The Tax Load on Net Corporate Earnings 30 1 The Tax Liability Due to Corporation Income Taxes 30 2 The Tax Liability Due to the Personal Income Tax 32 a Marginal or Average? 33 . b Total Income on Adjusted Gross Income? 34 3 A Specific Description of the Method 43 4 The Bias Due to Lamping 52 5 The Bias Due to Combining the Data for Individual and Fiduciary Returns 58 IV The Hybrid Nature of Our Measure 68 V Estimates of the Number of Dividend Returns 73 1 Estimating the Number of Personal Returns Reporting Dividends as Such 74 2 Estimating the Number of Fiduciary Returns Reporting Dividends 94 3 Estimating the Number of Fiduciary Income Receivers Who Constituted Net Accretions to the Number of Dividend Returns 96 ~ 'VI How the Effective Rate Schedules Were Obtained 106 A The Effective Rate Schedules for tile Type Components 107 1 The Effective Rates on the Property Income Components of Stockholders* Statutory Adjusted Gross Income 107 2 The Effective Rates on Net Corporate Earnings 112 B The Effective Rates on Adjusted Gross Income, Alter­ nately Defined 114 1 The Effective Rates on Stockholders’ Statutory Adjusted C-ross Income 115 2 The Effective Rates on Fully Imputed Stock­ holders* Adjusted Gross Income 116 VII Measuring Progressive.ty 118 VIII An Alternative Assumption as to the Distribution of Cor­ porate Earnings in 1S46 122 IX The Revenue Loss Estimates for 1932 and 1946 128 1 1932 128 2 1946 131 IV Appendix B Some Tests Page A Test of the Assumption by which Net Corporate Earnings were Allocated to Personal Income Taxpayers 1 A Test of the Assumption Used in Allocating Corporation Income Taxes to Personal Income Taxpayers 12 Additional Information - Dividends and Stockholdings; Wisconsin Income Tax Payers, 1949 16 Appendix C Some Problems A Problems Connected with Capital Gains 1 1 The Varying Inclusion of Capital Gainsi n Taxable Income 1 2 The Income Composition of "Stockholders" 2 3 Capital Gains or Corporate Earnings to Measure Personal Income from Corporate Activity? 3 4 The Relation between Reinvested Earnings and Capital Gains 4 3 Incidence of Corporation Income Taxes 12 C The Meaning of an Average 17 Chapter I Introduction Chapter I Introduction Corporate Earnings and Personal Income It is a truism that the claims on the income that results from economic activity during a specified period are all ultimately traceable to the persons who participated in this economic activity. Real people under­ take the actions and real people secure the claims to the income that results from the things they do. Real people sell their services and permit the use of resources which they own, and the income generated is the net sum (money, real, or a combination thereof) of these transactions. But while all claims to income are ultimately traceable to persons, one particular component of the total income generated presents a theoretical and statistical hurdle in the process of attributing the income of a period to individuals. The interposition of a legal person between the activities from which this form of income is derived and the individuals who own the legal person is the source of those difficulties. Almost all prepared income class distributions fail to impute one source of income completely to persons or family units— corporate earnings.^ Similarly, in the very practical matter of the definition of taxable personal income, most taxing jurisdictions do not undertake complete ^"The undistributed portion of the income of estates and trusts is also not imputed. imputation of corporate profits or losses. Net dividends disbursed by corporations to individuals are, of course, considered personal income for both purposes.^- But the handling of the undistributed part of annual corporate earnings (or losses) is more varied and controversial. In national income accounting, net undistributed profits (or net losses) of corporations are usually listed as a total sum, and not related to personal income class distributions• But when these data are put to specific uses it frequently becomes necessary to relate undistributed corporate earnings to the individual owners of corporate enterprise both in the aggregate and by income class breakdowns. For instance, Lawrence Klein, in his econometric model, considers undistributed profits (cor­ porate earnings) to be a relevant variable in the "explanation” of con­ sumer behavior and he includes "corporate savings in the profit variable of both the consumption function and the investment function."^ Recent work in the analysis of the effect of taxation on the equali­ zation of income, likewise has stressed the importance of undistributed profits and the necessity for taking account of them on a personal basis.^ In an article on the distributions of income in 1938 and 1947, Dudley Seers says the following:4 •^But,many taxing jurisdictions treat a given amount of dividends differently from a similar amount of other income, either in terms of how much is to be reported or the amounts of tax levied thereon. ^"Capitalists will be assumed here to consider income as their own whether it is in their personal possession or in the possession of their corporations." Lawrence R. Klein, Economic Fluctuations in the United States. 1921-1941. (New Yorks John Wiley and Sons, Inc., 1950), p. '65." " *ELbor Bama, Redistribution of Incomes in 1937 (Oxford University Press, 1947), 70j Ursula K. Hicks. Public Finance (New York: Pitman Publishing Company, 1948); Dudley Seers, 'Income Distribution in 1938 and 1947,' Bulletin of the Oxford University Institute of Statistics 11 (1949),p.r2S6. 4Ibid. This comparison of distributed personal imcomes has certain uses. It is, however, rather misleading if it is taken as a measure of the levelling of incomes, because undistributed profits, which largely belong to the upper brackets, have increased from L 257 millions in 1938 to £> 940 millions in 1947. We shall therefore attempt to allocate imputed income, knovdng that by so doing we shall get less accurate estimates, but estimates of something more meaningful. . . . Similarly in an investigation of the shares of the upper income groups in total income, Professor Simon Kuznets found it of interest to measure the effect of including corporate savings in his income concept as compared with results based on a concept of personal income which excluded them. Professor Seymour Harris has criticized the Steel Industry Board (the fact­ finding board of 1949 in the Steel Wage dispute) for taking dividends rather than corporate profits as the share of personal income from corporate earn­ ings. ^ Corporate Earnings and Taxable Personal Income » The preceding instances indicate that for many purposes the more useful con­ cept of personal income from corporate earnings is that which embraces the total of such earnings rather than merely the distributed portion. Now, since they are designed for different purposes, the national income concepts ^In a manuscript prepared at the National Bureau of Economic Research and not yet published. ^Seymour E. Harris, 'Comments on the Steel Report - Introductory Remarks,' Review of Economics and Statistics. XXXI (1949), 282. (I have substituted underlining for the author's italics.) "Indeed the Board made some telling points. But the question may be raised whether the appropriate comparison is of wages and dividends or wages and profits. The fact is, however, that reinvested profits belong to the stockholders: and therefore issues of equity relate to profits not' dividends.11*