fyxmll ^mmii^ ^ihux^ GOLDWiN Smith Hall FROM THE" FUND GIVEN BY aSof&mitt Smith 1909 arV1881 ^°"'*" ""'*'"■*''>' '-'""'T V.I 7 '"iSSIVSiim,.,,!!?.'?.,'^"''^'' States, its amount ,. 3 1924 031 715 828 olin.anx Cornell University Library The original of tliis book is in tine Cornell University Library. There are no known copyright restrictions in the United States on the use of the text. http://www.archive.org/details/cu31924031715828 Publications of the National Bureau of Economic Research, Incorporated. No. 1. INCOME IN THE UNITED STATES ITS AMOUNT AND DISTRIBUTION 1909-1919 INCOME IN THE UNITED STATES Its Amount and Distribution 1909-1919 BY The Stai-f of the Nationai. Bukeau of Economic Research^ Incorporated Wesuit C. Miicrell. WlLLFOBJ) I. KiKO Fbedekick R. Macaui^y OSWAU) W. Kn AUTH Volume I SUMMARY NEW YORK HARCOURT, BRACE AND COMPANY 1921 COPTMGHT, 1921, BT HAECOUBT, BllACE AXD COMPANY, INC. " -JJT THE U. S. A, PREFATORY NOTE The National Bureau of Economic Research, Incorporated, was chartered in 1920 to conduct quantitative investigations into subjects that af- fect public welfare. Its aim is to ascertain funda- mental facts within its field as accurately as may be, and to make its findings widely known. By so doing, the Bureau hopes to aid all thoughtful men, however divergent their views of public policy, to base their discussions on objective knowledge as distinguished from subjective opin- ion. The organization of the Bureau is designed to ensure not only scientific and impartial work on the part of its staff, but also a review of their findings by men who represent all the important viewpoints from which economic problems are regarded. Control is vested in a board of nine- teen directors. The present constitution of this board is as follows : Directors-at-large : T. S. Adams, Adviser to the U. S. Treasury Depart- ment. vi PREFATORY NOTE John R. Commons, Professor of Political Economy, University of Wisconsin. John P. Prey, Editor of the International Molders' Journal. Edwin F. Gay, President of the New York Evening Post. Harry W. Laidler, Secretary of the Intercollegiate Socialist Society. Elwood Mead, Professor of Rural Institutions, Uni- versity of California. "Wesley C. Mitchell, New School for Social Research. J. E. Sterrett, Member of the firm of Price, Water- house & Company. N. I. Stone, Labor Manager, Hickey, Freeman Com- pany. AUyn A. Young, Professor of Economics, Harvard University. Directors-by-Appointment representing organi- zations : F. P. Fish, The National Industrial Conference Board. Hugh Frayne, The American Federation of Labor. David Friday, The American Economic Association. W. R. Ingalls, The Engineering Council. J. M. Larkin, The Industrial Relations Association of America. George E. Roberts, The American Bankers' Associ- ation. Malcolm C. Rorty, The American Statistical Asso- ciation. A. W. Shaw, The Periodical Publishers' Association. Gray Silver, The American Federation of Farm Bureaus. PREFATORY NOTE vii The directors, through their Executive Commit- tee, choose the topics for investigation and ap- point the scientific staff. The by-laws provide that all reports made by the staff shall be sub- mitted to the directors for criticism before they are published, and that a director who dissents from any finding approved by the majority of the board shall have his dissenting opinion pub- lished in the report if he so desires. The present report owes much to the active cooperation of the directors, and many suggestions made by them are incorporated in the text. It is believed that this critical review of the staff's work by a group of men representing varied training, experience, and opinions safeguards the reports against bias.^ The chief financial support of the Bureau has come from two philanthropic foundations, the Commonwealth Fund and the Carnegie Corpora- tion. Several other contributions, none exceed- ing $1,000, have been made by public-spirited in- dividuals and business enterprises. The General Theological Seminary of New York has gener- ously furnished office space in one of its buildings. To all of these donors the Bureau takes this oc- casion, its first public opportunity, to express its thanks. * Because of absence in Europe Mr. Pish has not been able to read the present report. vlii PREFATORY NOTE The present report deals witli a subject of fun- damental importance in which the truth is hard to find — a subject so important that despite mani- fold difficulties, it has attracted investigators in many countries. Australia took an actual census of wealth and incomes in 1915 as a war measure. Excellent estimates have been made of the na- tional income of Great Britain and G-ermany, where well-administered income taxes with low exemption limits provide a solid foimdation to build upon. Approximations that are less ac- curate because the underlying statistics are less abundant have been published for France, Italy, Spain, Austria-Hungary, Canada, and Japan. In the United States, statisticians attacked the prob-. lem from time to time before the war — notably Charles B. Spahr, Frank H. Streightoff, Willford I. King, and Scott Nearing. They found the American data bulky but miscellaneous and hard to fit together. The war lent the problem pressing importance ; and several estimates of the national income, most of them based directly or indirectly upon Mr. King's figures for 1910, were made by men interested in the government's financial policy. These estimates were all rough approxi- mations, hastily constructed. Quite naturally, they differed considerably in their results. PREFATORY NOTE ix A desire to learn whether the National Income is adequate to provide a decent living for all per- sons, whether this income is increasing as rapidly as the population, and whether its distribution among individuals is growing more or less un- equal, and to sift the divergencies among the current estimates led the National Bureau of Economic Eesearch to choose this field for its first investigation. Its staff was directed to undertake a thorough canvass of all the available materials and to make as close an estimate as possible of the size of the National Income, its variations from year to year in dollars and in goods, and the way in which this income is divided among the people. More than a year has been spent upon this work, the results of which are summarized in the chap- ters that follow. Even with the addition of the income-tax tables which have recently become available, the Ameri- can data leave much to be desired. Moreover, the wild fluctuations of prices, wages, and profits dur- ing the war introduce new complications into a task that is difficult under the most stable condi^ tions. Nevertheless, it is believed that the results here presented, while necessarily subject to a mar- gin of error, are more reliable th.an those which earlier Americsm investigators, working with less X PREFATORY NOTE help and fewer data, have been able to obtain ; and that, in reliability, these results compare not un- favorably with the estimates available for foreign countries. The Bureau is planning to continue the work of estimating the National Income from year to year on a basis comparable with that fol- lowed in this volume. The results for later years will be announced as the necessary data become available. TABLE OF CONTENTS PAGE Prefatory Note v Chapter 1. Introduction 1 I. The Questions to be Answered .... 1 II. Material and Methods 2 III. The Two Ways of Estimating the National Income 9 Chapter 2. The Size op the National Income . 12 I. The Two Estimates of the National Income Compared 12 II. Analysis of the Estimate by Sources of Production 16 III. Analysis of the Estimate by Incomes Re- ceived 29 IV. What these Estimates Count as National 'Income 42 V. Final Estimate of the Size of the National Income 60 VI. The National Income Reduced to Pre- War Values 69 VII. Total and Per Capita Income in Different Countries 81 Chapter 3. The Distribution op the National Income 89 I. The Share of Employees in the National In- come 89 xi xii TABLE OF CONTENTS PAOB II. Personal Incomes Above and Below $2,000 Per Year 108 III. The Distribution of Income among Indi- viduals 116 Chapter 4. Conclusions 143 Index 149 INDEX OF TABLES PAGE . 1. The Two Estimates of the National Income ... 13 2. Analysis of the Estimate by Sources of Production 18-19 3. Percentages of the National Income Contributed by the Various Industries 24 4. Relative Fluctuations in the Value Products of the Various Industries 26 5. Analysis of the Estimate by Incomes Received . . 84-35 6. Relative Fluctuations in the Items Included in the Estimate by Incomes Received 38 7. Conjectural Estimate of the Money Valuo of House- wives' Services 59 8. Margins of Error in the Two Estimates of the National Income 63 9. Final Estimate of the National Income and the I/imits Within Which the True Values Probably Fall ... 64 10. Comparison of Estimates of the National Income Made by Different Investigators 66 11. Final Estimate of the National Income Per Capita and the Limits Within Which the True Values Prob- ably Fall 68 12. The National Income and Its Purchasing Power at the Price Level of 1913 According to the Estimate by Incomes Received 71 13. The National Income and Its Purchasing Power at the Price Level of 1913 . , 72 14. The Final Estimate of the National Income Per Capita and Its Purchasing Power at the Price Level of 1913 76 15. Comparison of the Fluctuations in the Final Estimate of the Purchasing Power of the National Income at the Price Level of 1913 with Four Index Numbers of the Physical Volume of Production 79 16. Summary Showing the Estimated National and Per Capita Income of Various Countries at the Outbreak of the War in 1914 and the Approximate Accuracy of the Respective Estimates 85 xiii xiv INDEX OF TABLES PAGE 17. Percentages of the Net Value Product of Various Industries Received by Employees in the Form of Payment for Services 92 18. Division of Combined Net Value Product of Mines, Factories and Land Transportation between Earn- ings of Employees and Returns for Management and the Use of Property 97 19. A Rough Comparison of the Salaries of Officials, the Pay of Manual and Clerical Employees, and the Net Value Product of Mines, Factories and Land Trans- portation 99 20. The Average Annual Earnings of Employees Nor- mally Engaged in Various Industries .... 102-103 21. A Rough Estimate for 1918 of the Income from All Sources of Salary and Wage Workers 107 22. Personal Incomes Above and Below $2,000 per Annum 112 23. A Conjectural Estimate of the Percentage of the Na- tional Income Received by the Highest Five Per Cent, of Income Receivers 116 24. The Distribution of Personal Incomes by Income Classes as Shown by the Official Compilation for the Calendar Year 1918 117 25. Distribution of Income Among Personal Income Re- cipients in 1918 132-134 26. Percentage Analysis of the Distribution of Personal Incomes in 1918 134-135 27. A Condensed Summary of the Distribution of Personal Incomes in 1918 138 28. The Percentage Analysis of the Condensed Summary of the Distribution of Incomes in 1918 137 29. The Distribution of Net Personal Incomes Reported to the Bureau of Internal Revenue in 1917 and 1918 Before and After Deduction of Federal Income Taxes 140 INDEX OF CHARTS PAGE 1. The Two Estimates of the National Income .... 14 2. Per Capita Income According to the Two Estimates . 15 3. Eelative Fluctuations in the National Income According to the Two Estimates 16 4. Contributions to the National Income made by the Various Industries 21 5. Percentages of the National Income Contributed by the Various Industries 23 6. Eelative Fluctuations in the Value Products of Agricul- ture, Mining and Manufacturing 27 7. Eelative Fluctuations in the Value 'Products of Trans- portation, Banking and Government 28 8. The Five Sections of the National Income According to the Estimate by Incomes Eeeeived 36 9. Percentage Division of the National Income According to the Estimate by Incomes Eeeeived 37 10. Eelative Fluctuations in the Aggregate Incomes of Farmers and of All Ofther Persons Subdivided Ac- cording to Whether They Eeceive More or Leas than $2,000 Per Year 39 11. Eelative Fluctuations in Aggregate Personal Income and in Corporate Surplus . 40 12. The Final Estimate of the National Income and Esti- mates by Other Investigators 65 13. The Two Estimates of the Purchasing Power of the National Income at the Price Level of 1913 . . . 73 14. The Final Estimate of the National Income and Its Purchasing Power at the Price Level of 1913 ... 74 15. Eelative Fluctuations in the Final Estimate of the National Income and Its Purchasing Power at the Price Level of 1913 75 16. The Final Estimate of the Income Per Capita and Its Purchasing Power at the Price Level of 1913 ... 77 XV xvi INDEX OF CHARTS PAGE 17. Relative riuctuations in the Final Estimate of the Income Per Capita and Its Purchasing Power at the Price Level of 1913 78 18. Comparison of the Fluctuations in the Final Estimate of the Purchasing Power of the National Income at the Price Level of 1913 with Four Index Numbers of the Physical Volume of Production .... 80 19. The Estimated National Income of Various Countries in 1914 86 20. The Estimated Income per Capita of Various Countries in 1914 87 21. Percentages of the Net Value Product of Various In- dustries Received by Employees in the Form of Pay- ment for Services 93 22. t*he Purchasing Power at the Price Level of 1913 of the Average Annual Earnings of Employees in Va- rious Industries 104 23. Relative Fluctuations in the Purchasing Power at the Price Level of 1913 of the Average Annual Earnings of Employees in Mining, Manufacturing, Transpor- tation and All Industries 105 24. Relative Fluctuations in the Purchasing Power at the Price Level of 1913 of the Average Annual Earnings of Employees in Agriculture, Government and All Industries 106 25. Percentages of Persons Receiving Incomes Above and Below $2,000 Per Annum 113 26. Percentages of Total Personal Income Received by Per- sons with Incomes Above and Below $2,000 Per Annum 114 27. The Income Tax Returns of 1913-1916 plotte'd on a Double Logarithmic Scale 120 28. The Income Tax Returns of 1916-1919 plotted on a Double Logarithmic Scale 121 29. Distribution of Income Among Personal Income Recip- ients in 1918. Natural Scale 128 30. Distribution of Income Among Personal Income Recip- ients in 1918. Double Logarithmic Scale .... 129 31. Lorenz Curve showing the Distribution of Income . . 141 CHAPTER 1 INTRODUCTION I. THE QUESTIONS TO BE ANSWERED What is the aggregate income of the American people ? How much does that income vary from year to year? What part of the changes is due to fluctuations in prices and what part to fluctuations in the pro- duction of goods? How is the aggregate income divided among individuals ? What proportion of the whole income goes to wage-earners and salaried employees? How does per capita income in the United States compare with that in other countries? These questions mark out the field in which the National Bureau of Economic Research has made its first investigation. The present volume summarizes the answers which the Bureau's staff has heen able to obtain by more than a year 's work. The second volume of the report shows in detail 2 INCOME IN THE UNITED STATES the sources from which the data for the answers have been drawn, the methods used by the staff, and the margins of doubt by which their results are bordered. On these topics some general statements are presented here, but the reader who wishes to form his own opinion of the investigation and the results must consult the second volume. He will find there much information about particular in- dustries which does not come into this Summary at all. II. MATBEIAL AND METHODS Only one country in the world — Australia — has ever taken a census of incomes. In other countries what is known about the size of the na- tional income and the mode of its distribution rests upon estimates. These estimates can only be made by dovetailing together data that are diverse in form, content, scope, date, and source; bulky at some points, scanty at others, and generally re- quiring adjustment of some sort before they are fit for use. Among the important bodies of data that can be used in estimating the income of the United States in recent years are the Internal Revenue Bureau's tables of personal and corporate incomes, many records of wages per hour, day or week, scattered reports on the salaries of teachers and clergymen INTRODUCTION 3 and the annual earnings of physicians and engi- neers, a few investigations into the incomes of farmers, the rent surveys of many towns made by the American Telephone & Telegraph Company, collections of family budgets, the Census statistics of occupations, the quinquennial Census of Manu- factures, the Department of Agriculture's annual estimates of the value of the crops, the Geological Survey's data on mineral products, the Interstate Commerce Commission's reports on transporta- tion, monographs on special industries investi- gated by the Tariff and Federal Trade Commis- sions, various state and municipal documents deal- ing with government expenditures, production, wages and the like, statistics compiled by the national associations or service bureaus of lead- ing industries, and the files of technical journals. From all of these sources and from others too numerous to list materials must be collected, com- pared, criticized and fitted together. One of the most serious difficulties in working with these data is the difficulty of definition. Pre- cisely what is the National Income ? Is it money, or commodities and services, or satisfactions? Is the National Income the sum of the incomes of in- dividuals, or may an individual have personal income which is not income to the nation? Ought 4. INCOME IN THE UNITED STATES the undistributed incomes of business enterprises to be added to the incomes of individuals to get the total? Ought taxes paid by individuals to be deducted from their incomes? What part of the selling value of an industry's output is produced by the industry itself? "What individual incomes are merely- parts of other incomes? Are there negative incomes to be deducted from the sum of positive incomes? It is hard enough to tell ac- curately what is the income of a single family or business, as most people who have struggled with income-tax blinks will testify. It is harder still to frame a definition which will cover both individ- ual and National Income. And it is hardest of all to frame a definition which will include both of these concepts and at the same time enable one to use the exceedingly miscellaneous data from which an estimate of the aggregate must be made. Fresh complications are introduced by the fac- tor of time, complications which have been magni- fied by the recent fluctuations in prices. Many of the most valuable pieces of evidence refer to some one year or to two or more years separated by intervals that vary in length. It is necessary to utilize much of this evidence and desirable to uti- lize all. Yet complete data are not to be had for any one year ; indeed, every year has its special INTRODUCTION 5 treasures and its deplorable gaps. To make an acceptable estimate of the National Income for any recent year it is therefore necessary to make an estimate for every recent year and to check these estimates against each other with due allowances for changes in the level of prices. What allow- ances are "due" on this score is itself a difficult problem. Clearly, no care taken in working up such data as are now available can guarantee precise re- sults. The task of putting a figure on the National Income is more like the task of valuing a railway system than like the task of drawing up its profit and loss statement. The work must be done in a broad style. It is estimating, rather than enumer- ating or measuring. In his recent summary of the estimates of the national incomes of the chief powers, Sir Josiah Stamp put in his highest grade those estimates "not likely to be inaccurate to a greater extent than 10 per cent."^ Judged by this standard, the American estimates for the last two or three years would belong in the highest grade if the errors were not likely to exceed some six billions of dollars. The estimates, however, need not be uncon- trolled. They can, if the work is properly planned ^ See Jowmal of the Eoyal Statistical Society, July, 1919, pp. 444-491. 6 INCOME IN THE UNITED STATES in advance, be safeguarded in considerable meas- ure by a system of cross-checking. As will pres- ently be explained, two independent estimates of the total National Income can be made by dis- tinct methods applied mainly to different collec- tions of data. These estimates can be used to test each other. More than that, most of the large items that enter into each estimate can be arrived at in two or more ways. In the present investi- gation much attention has been given to devising and applying such tests of the partial results, a branch of the work in which invaluable help has been received from correspondents who have scrutinized our tentative results with expert eyes. Such precautions, to repeat, do not ensure a high degree of accuracy; but they do guard against gross errors. We have not leaned heavily upon the statisti- cian 's fond hope that errors made by the way will cancel each other in the end. Doubtless they do so to some extent and our totals are the better for that fact ; but we have tried to make the estimate for each item considered by itself as nearly cor- rect as our data, time, and means have permitted. Of course the estimated errors of our figures vary widely from item to item with the quantity and quality of the underlying statistics. Therefore, INTRODUCTION 7 we have been careful to indicate the degree of con- fidence we feel in the various results. Some read- ers of the second volume will think that we have been meticulous in our treatment of minor factors. It is true that in very many of the items figures several times too large or too small would not appreciably affect the final aggregates, which run in tens of billions of dollars. But that is another comfort of which we prefer to make sparing use. Many of these minor items have an interest quite independent of their contribution to the total, and if mistakes are found even in the smallest of them by men who have special knowledge of the facts, we shall be grateful for their help in rectifying our estimates. It is true also that many of the important uses which an estimate of the National Income and its distribution serves, are served almost as well by a fair approximation as by an exact measurement, could such a measurement be made. We have treated that consideration, however, not as an excuse for slighting details, but as a spur to check the validity of our broad results as carefully as possible. These broad results are the matters of chief concern. It is necessary in many of our computations to adopt definite figures of two or three digits to express sub-totals and grand totals. 8 INCOME IN THE UNITED STATES But we attach no importance to the third digits, except as an aid in computation, and in the great majority of cases, including the grand aggregates of National Income, we regard the second digits as subject to a wide margin of error. The state- ments in which we have most confidence and to which we attach most importance are put in the form of ranges within which the National Income and its major constituents probably faU. Finally, we have profited much by the work done in this difficult field by our predecessors in America and in other countries, pioneers who achieved val- uable results despite their slender resources and scanty data. Happily, economic statistics is a progressive field, and the latest comers should be able to improve upon the results of earlier workers. We believe that the results presented in this report do constitute an advance beyond earlier American work. But at most they are merely the best approximations we can frame now from the current data. We do not regard the tables in this report as final. On the contrary, we hope to revise our estimates as fresh data be- come available and as better analytic methods are devised. In this process of constructive criti- cism and revision of the figures we cordially in- vite every one interested to share. INTRODUCTION 9 III. THE TWO WAYS OF ESTIMATING THE NATIONAL INCOME The collections of data listed in the preceding section as available for estimating the National Income are of two kinds. One kind shows income received — the. _ income-tax returns, reports on wages and salaries, investigations of the profits of farmers, and the like. The second kind shows in- come produced — the statistics of coal and metals mined, lumber cut, crops grown, raw materials transported or manufactured, and the like. These two kinds of sources cannot both be com- pletely utilized in making a single estimate of the National Income. For how c^n one combine, for example, the statistics of personal incomes over $2000 compiled by the Internal Revenue Bureau with the statistics of "value added by manufac- ture" reported by the Census Bureau? Then which set of sources should be used — the set that shows income received, or the set that shows in- come produced ? It is hard to say in advance which set will yield the more trustworthy results, and, in view of the margin of uncertainty to which the best estimates in this field are subject, one is ex- ceedingly reluctant to relinquish the use of any body of data from which help can be had. 10 INCOME IN THE UNITED STATES The Bureau's solution of this problem is to use both sets of data and to make two independent estimates of the National Income for each year. One estimate, called the "Estimate by Sources of Production," is derived from a study of the separate industrial fields in which the income originates. The second estimate, called the "Esti- mate by Incomes Eeceived," utilizes the data which show the income received by individuals, plus the income received by business enterprises but not distributed to their owners. To ensure their independence, these two esti- mates were made by different members of the staff. Mr. King had charge of the Estimate by Sources of Production and Mr. Knauth of the Estimate by Incomes Received. Not until the last large items in the more laborious estimate had been figured was it possible to tell whether the two parallel investigations were leading up to similar or to widely divergent results. Then it was found that the maximum discrepancy in any year between the two sets of preliminary totals was 7.0 per cent. After the few items in the two estimates which could properly be compared were set against each other, each estimate was critically revised. The object of the revision was not to force the two estimates into agreement, but to INTRODUCTION 11 make each considered by itself as perfect as the staff could make it on the basis of the available data. Some of the changes introduced in revision tended to bring the two series closer together, while others tended to force them further apart. In reporting the outcome of the Bureau's work it is best to begin with the general results and then to take up details. First, the two estimates of the aggregate National Income will be presented in their final form. Next, the more important items of which each estimate is made up will be shown. This exhibit will raise in concrete form the problem whether an estimate of the National Income by sources of production theoretically ought to yield the same results as an estimate by incomes received. Discussion of that problem will pave the way for a consideration of the margin of error in both estimates, of the most probable values of the National Income in the years cov- ered, and of how this income compares with that of other countries. Then will come an estimate of what the enormously inflated money incomes of the war years would amount to if reduced to dol- lars of constant purchasing power. Last but most interesting of all, we shall consider the way in which the National Income is distributed among individuals. CHAPTER 2 THE SIZE OF THE NATIONAL INCOME I. THE TWO ESTIMATES OF THE NATIONAL INCOME COMPARED Table 1 shows tlie final figures for the National Income given by the Estimate by Sources of Pro- duction and the Estimate by Incomes Eeceived. Certainly the agreement between the two esti- mates, made as they were independently of each other, is remarkable. The average National In- come in the nine years covered by both series works out at 40.2 billions in the Estimate by Sources of Production and at 39.7 billions in the Estimate by Incomes Received. Even the maxi- mum difference of 6.9 per cent, in 1913 is small for work in this field, and in two years, 1911 and 1917, the two estimates happen to agree to the nearest hundreds of millions. On the per capita basis, the maximum difference is but $24 per annum. Indeed, the only difference of note con- cerns the rate at which the National Income has increased. The increase from 1910 to 1918 is 90 per cent, in one case and 98 per cent, in the other. 12 SIZE OF THE NATIONAL INCOME 13 ^ o o !?; M o M S E-i OS ^ S O § 03 ^ 03 f^ ^ ^(§ PI o •S >o B o ;:! M " CQ o g •j3 2 'S S o ra 5 S Ch t-k M d S ^ I + I » .-s tj h a rt "=> ^ fn •*j 03 cs » ia " 03 '^ ja 02 » eio o + I 2_j ■» ^ b cQ ^ o ;^ • rH S m g y ^ o OS CO 1-1 ^ 00 OS CO (M «D 00 lA os W 00 CO ; OS « «; '^. q •-! oa o co' CO •^* ' I *^ <^ + +++ + I I ' O f-H Ca rH ' ,' O iH + ++++I I o t3 M to o r-J Oq ^ CO W OS U5 OS *> o 2 T-?THodcocqirD»r3«rH5d Sd •» ■S-J3 O »H 1-1 M ■2 I 00 o Oi OrHOCICOTfHlOCO t^-OO OS -2 O iHiHiHiHiHrHfHi-tr-li-H ^cS OS OSOSOSCSOSOSCSOSOSOS hHfH iH i-frHflr-liHrHrHrHrHi-l i^i 00 ea to to 05 iH rx OS ■* 14. INCOME IN THE UNITED STATES The charts which follow bring out the el^ief fea- tures of Table 1 in graphic form. Needless to say, the sudden acceleration during , Chaet 1. THE TWO ESTIMATES OF THE NATIONAL INCOME. 1909-1919. Based upon Table 1. For elimination of the increase due to the rise of prices, see Chart 14. i909 mo i9U 19JZ 19ii )9}4 i9tS i9l6 igi^ 19J8 i9l9 the war in the rate at which the National Income increased was due mainly to the rise of prices — a factor in the situation which will call for careful consideration after the two estimates have been analyzed more closely. SIZE OF THE NATIONAL INCOME 15 Chart 3, showing the percentage change in the National Income according to the two estimates, is based upon the average amount of the income Chaet 2. PER CAPITA INCOME ACCORDING TO THE TWO ESTIMTATES. 1909-1919. Based upon Table 1. For elimination of the increase due to the rise of prices, see Chart 16. 1909 1910 i9n mz 1913 1914 1915 1916 1917 1918 19J9 as shown by each of the estimates in the whole period common to both, because that base affords a fairer comparison than would percentages based upon the results for any single year. 16 INCOME IN THE UNITED STATES Chart 3. RELATIVE FLUCTUATIONS IN THE NATIONAL IN- COME, ACCOEDING TO THE TWO ESTIMATES. 1909-1919. Based upon Table 1. For elimination of the increase due to the rise of prices, see Chart 15. /80i «0/6CH joo KXUTIVE. ilZE or THE NATlOTiAL INCOME m i909-l9l9 AVEKACE FOR. 1910-1916=100. -ESTIMATE Br -ESTIMATE BY SOURCE or INCOMES PRDDUCTIOH. RrCETVED. 1909 72. 1910 Bll ... .79 —II 78 — «1- tin M «4 (OW «4 «2 (9/5 9a iH !>. lO i-J «D CO CO CO M CO p.QO -HCDTH ococoio>-i oo Oa O CQ >>; U3 00 CO oa o o "h § cd' 00 b-* in CO b^ b-' w oo' iH &b^ HiHiHiHrH rHi-lr-irHCq CO in b- ai OS O rH (N CO Tt< in CD b- 00 ^ m Ol 00 OrHrHrHiH iHiHrHtHrH 01 gnO O ^ oi a Oi d a Q> Oi a a o> > ^oi -4^ o^ tHtHiHiHiH iHrHTHiHiH o 3 S u ai - o ■J3 ^ t- *0 CO CO 00 rH r-i ■*' ■*' «5 l>^ !■§ 3 ^ll 00 tH (M_ o xi; 00 tH r-4 CO ■'SJH CQ CO CO CO CO o 1 Oi CO rH !>; O ira 00 GO OS r-i (M (M oq W ffO iq «) OS 5D CO iH ?d oi oi th oi t-^ CO CO -^ w to ^ oq o -^^ 5D Oi o Co' «3 lO OS CO rH CO eo "^ tH to :o 05 Tt< Oi i>. iq Oi OS cm' O TjH 00 TjH iHrHrHrHi— It-HiHiH Oso^Osososososo) SIZE OF THE NATIONAL INCOME 67 m Sizijcsf So 53-H;3tpl 1 -^ -^ oa I "bo _ O 1-1 >. s . o p^oi -^ J^ ^ PhiH m&h^ ^ «H fc S 3 :3 a ■ Is .2 b§ S> o o ^.2 05 ^ -a • o M C3' CO 03 ^ p-g 03 f« 005 ^ 2 _Q .l-f -v 03 ^^'^B g.s.fc a O O ©a -^ CO CD CO ^^ ^H ^^ ^^ ^r ^i" CO c:i CO l> 00 CD lO to t- o niSra3S'aeiiCagjdQt-;5ai -, - > em o 3 (S __ p<^ o " iH m » ^ 6BO .^ s g a or„ s J ^«b g .^ t.;fH«= a s!«r-ii 3 a •" S a : =^ O !3 o o. >i osa>i!ig„0 v** ij -0 03 ^ 2 'iH 9 'O « a -5 'S S"S ^3 _ 0) M g =* 3:s a-gg a SRaS"^ 3b- S OlOrHtMCO "*W W O iH iH IH iH iH 1-1 ■•— ' OS Oi Ci Oi OS Oi Ol iH T-t tH iH rH 1-1 rH CD t* 00 iH iH rH O) O) O) ■2 2 3 I d © •« SS »^"^-' S ^ • -, 2s?'OK'SfH-13ggh^^ "* "^ Q •1-4 C uj -iH ^ y r*\ ^ I MB 5 « -a 3 si 3 o .e 2 68 INCOME IN THE UNITED STATES Table 11 FINAL ESTIMATE OF THE NATIONAL INCOME PER CAPITA AND THE LIMITS WITHIN WHICH THE TRUE VALUES PROBABLY PALL 1909-1918 Tear Population Ten per Five per Final Five per Ten per on June 30 cent, less cent, less ( estimate of cent, more cent. (Millions) than the than the the Na- than the more than final final tional In- final the final estimate estimate come per capita estimate estimate (In Dollars per Annum) 1909 90.37 $287 $303 $319 $335 $351 1910 92.23 306 323 340 357 374 1911 93.81 300 316 333 350 866 1912 95.34 311 329 346 363 381 1913 97.28 319 336 354 372 389 1914 99.19 302 318 335 352 368 1915 100.43 322 340 358 376 394 1916 101.72 401 424 446 468 491 1917 103.06 471 497 523 549 575 1918 104.18 527 557 586 615 645 been made by other investigators for various years in our perio(i. All the other estimates fall within 10 per cent, of our final estimates, except Professor Friday's estimate for 1917 and Dr. Anderson's estimates for 1917 and 1918. Another interesting supplement to Table 9 is a reduction of the National Income to income per capita. The population figures used in this table are estimates for June 30th of the intercensal years made by Mr. King from census returns, vital statistics, and immigration records. Once more, the reminder may be entered that Tables 9 and 10 and the charts drawn from them show income in dollars or billions of dollars, and that most of the apparent increase of income in the SIZE OF THE NATIONAL INCOME 69 war years is a monetary illusion. The next task is to see how much of the increase is left if we reduce our estimates to hypothetical dollars of constant purchasing power. VI. THE NATIONAL INCOME EEDUCED TO PRE-WAR VALUES • To determine the best method of "deflating" our estimates of the National Income is a difficult problem. It will not do simply to divide the ag- gregate figures by such a series as the Bureau of Labor Statistics index number of prices at'whole- sale, because the great bulk of income is spent by families on retail purchases. Nor will it do to divide our aggregates by the same Bureau's index number of the cost of living, because these figures are made expressly to represent changes in this cost to families of small means, and our figures profess to represent all families. Some more elaborate method is necessary, and it is desirable to safeguard the results by using more than one method. Hence we have broken up both estimates into parts and applied appropriate index num- bers to each part separately. In making the Estimate by Sources of Pro- duction Mr. King subdivided the net product of 70 INCOME IN THE UNITED STATES each industry into (1) sums paid to employees, (2) sums paid out in interest, dividends, rents, royalties, and profits, and (3) income not paid out but kept in the business. The first of these sums he "deflated" by using the Bureau of Labor Sta- tistics index number of cost of living. For "de- flating" the second sum, he used a new index number designed to show fluctuations in the living expenses of families having expenditures for con- sumption goods of $5,000 to $25,000 per year. To the third item, he applied an index number of construction costs. Since the details of this com- putation, as given in Volume II, are rather elabo- rate and since in another section, we shall present his statistics of the share of employees in the Na- tional Income, it will suffice here to give merely his final results. (See Table 13). Mr. Knauth's method of "deflating" the Esti- mate by Incomes Eeceived was somewhat differ- ent. He made a very rough estimate of the amounts of income "saved" each year by persons having incomes less than $2,000, and a similar esti- mate for persons having more than that amount. Of course, these "savings" are really spent. The two great objects on which savings were spent in 1914-1919 were new industrial equipment of all sorts, including houses, and the war. Accordingly, SIZE OF THE NATIONAL INCOME 71 the total savings, including those made by business enterprises, were divided, again very roughly, between those two uses. Thus Mr. Knauth con- verted his estimate of the National Income into an estimate of National Expenditures subdivided under four heads: (1) personal and family ex- penditures of people having incomes less than $2,000 per year; (2) similar expenditures of people having incomes above $2,000; (3) expenditures on construction of houses and industrial equipment; and (4) expenditures on the war. An index num- ber was used for each of these headings as fol- lows: (1) for incomes over $2,000, an index num- ber computed by the Bureau; (2) for incomes un- der $2,000, the cost-of-living index number of the Bureau of Labor Statistics; (3) for construction, Table 12 THE NATIONAL INCOME AND ITS PUKCHASING POWEE AT THE PRICE LEVEL OF 1913, ACCORDING TO THE ESTIMATE BY INCOMES RECEIVED Tear National Income Weighted Index Purchasing (Billion dollars) Number of Prices Power at Price Level of 1913 (Billion dollars) 1910 $31.1 97.8 $31.8 1911 31.2 98.5 31.7 1912 32.4 99.4 32.6 1913 33.3 100.0 33.3 1914 32.5 100.6 32.3 1915 35.9 102.5 35.0 1916 45.5 113.4 40.1 1917 53.9 136.1 39.6 1918 61.7 160.8 38.4 1919 66.0 176.8 37.3 72 INCOME IN THE UNITED STATES an index number computed by the Statistical Di- vision of The American Telephone and Telegraph Company; (4) for war expenditures, an index number based on the War Industries Board's His- Tablb 13 THE NATIONAL INCOME AND ITS PURCHASING POWER AT THE PRICE LEVEL OF 1913 In Billions of Dollars Tear National Income Purchasing Power at Price Level of 1913 Estimate Estimate Final Estimate Estimate Final by Sources by Incomes Estimate by Sources by Incomes Estimate of Pro- Received of Pro- Received duction duction 1909 $28.8 $28.8 $30.1 $30.1 1910 31.8 $31.1 31.4 32.5 $31.8 32.2 1911 31.2 31.2 31.2 31.7 31.7 31.7 1912 33.6 32.4 33.0 33.7 32,6 33.2 1913 35.6 33.3 34.4 35.6 33.3 34.4 1914 33.9 32.5 33.2 38.6 32.3 33.0 1915 36.1 35.9 36.0 35.3 35.0 35.2 1916 45.4 45.5 45.4 41.3 40.1 40.7 1917 53.9 53.9 53.9 41.9 39.6 40.8 1918 60.4 61.7 61.0 39.1 38.4 38.8 1919 66.0 37.3 Relative Fluctuations : 1913 = 100 1909 81 84 85 88 1910 89 93 91 91 95 94 1911 88 94 91 89 95 92 1912 94 97 96 95 98 97 1913 100 100 100 100 100 100 1914 95 98 97 94 97 96 1915 101 108 105 99 105 102 1916 128 137 132 116 120 118 1917 151 162 157 118 119 119 1918 170 185 177 110 115 113 1919 198 112 tory of Prices 1913-1918. These four series, ap- propriately weighted, were combined to make a final index number, which was applied to the ag- gregate National Income. The results are shown in Table 12. SIZE OF THE NATIONAL INCOME 73 When these results are put beside Mr. King's, the two series are found once more to be in close agreement; how close is shown by Table 13 and Chart 13. Perhaps the most important discrep- Chakt 13. THE TWO ESTIMATES OF THE PUBCHASING POWER OF THE NATIONAL INCOME AT THE PRICE LEVEL OF 1913. 1909-1919. Based upon Table 13. ancy concerns the year when the National Income, considered not as a sum of money values, but as an aggregate of commodities and services which current money income would buy, reached its maximum. One estimate puts the maximum in 1916, the other in 1917. The final estimate, made by splitting the difference between Mr. 74 INCOME IN THE UNITED STATES King's and Mr. Knauth's figures, gives tlie palm to 1917 by a slight margin. Both estimates make it clear that all of the ex- Chart 14. THE FINAL ESTIMATE OF THE NATIONAL INCOME AND ITS PURCHASING POWER AT THE PRICE LEVEL OP 1913. 1909-1918. Based upon Table 13. 7i> ^ .f!^-^ r r / f" 1 ..^^ --1 ^ PCWEK. ■JDO f i909 19/0 1911 i9llt i913 1914 I91S iwe istr i9ia traordinary gains in money income after the United States entered the war were due to fluctua- tions in prices. For even according to the Esti- mate by Sources of Production, the gain regis- tered in 1917 over 1916 was by no means extraor- SIZE OF THE NATIONAL INCOME 75 dinary — not comparable for example with the gain made in 1912 over 1911. The conclusion to which the figures point is that large increases in Chaet 15. EELATIVE ITjUCTUATIONS IN THE FINAL ESTIMATE OF THE NATIONAL INCOME AND ITS PURCHASING POWER AT THE PRICE LEVEL CfF 1913. 1909-1918. Amounts in 1913 = 100. Based upon Table 13. real National Income, if we may use that term to mean the serviceable goods available for use by the population, are due either to a marked im- provement in the harvests, or to a marked increase in industrial activity, or to both of these changes 76 INCOME IN THE UNITED STATES occurring simultaneously. From the dull year 1911 to the busy year 1912, and still more from the exceedingly depressed year 1914 to the exceedingly active year 1916, the gain is great. But once people are nearly all employed and Table 14 THE PINAL ESTIMATE OF THE NATIONAL INCOME PER CAPITA AND ITS PURCHASING POWEE AT THE PRICE LEVEL OF 1913 1909-1918 Tear Popula- National Income Purchasing Power at tion iu Price Level of 1913 MlUiona Income in Per Capita Income in Per Capita Billion Income in Billion Income in Dollars Dollars Dollars Dollars 1909 90.37 $28.8 $319 $30.1 $333 1910 92.28 81.4 340 32.2 349 1911 93.81 31.2 333 31.7 338 1912 95.34 33.0 346 33.2 348 1913 97.28 34.4 354 34.4 354 1914 99.19 33.2 335 33.0 333 1915 100.43 36.0 358 35.2 350 1916 101.72 45.4 446 40.7 400 1917 103.06 53.9 523 40.8 396 1918 104.18 61.0 586 38.8 372 Relative Fluctuations; 1913 = :100 1909 93 84 90 88 94 1910 95 91 96 94 99 1911 96 91 94 92 95 1912 98 96 98 97 98 1913 100 100 100 100 100 1914 102 97 95 96 94 1915 103 105 101 102 99 1916 105 132 126 118 113 1917 106 157 148 119 112 1918 107 177 166 113 105 the factories and workshops, the mines and railways, the ships and shops are used at full capacity, further increases of output slow down to the rate made possible by current increase of population, development of natural resources, construction of new equipment, and SIZE OF THE NATIONAL INCOME 77 improvement in methods. And when a large number of the most effective workers are with- drawn from industry, as they were in 1918, it is difficult if not impossible to prevent production in Chart 16. THE FINAL ESTIMATE OF THE INCOME PEE CAPITA AND ITS PUECHASINa POWER AT THE PRICE LEVEL OF 1913. 1909-1918. Based upon Table 14. eo» i909 i9IO physical terms from falling off, however large a money premium is offered for intense effort. In some respects, the per capita figures of Table 14 are more significant than the national aggregates of Table 13. These per capita figures of course increase at a slower rate in good years, 78 INCOME IN THE UNITED STATES and fall off at a faster rate in bad years than do the aggregates from which they are computed. When the data are cast into this shape, the decline Chart 17. RELATIVE FLUCTUATIONS IN THE FINAL ESTIMATE OF THE INCOME PER CAPITA AND ITS PURCHAS- ING POWER AT THE PRICE LEVEL OF 1913. 1909-1918. Amounts in 1913 = 100. Based upon Table 14. in real National Income during American partici- pation in the war becomes rather marked. The economic prosperity of 1919 was an illusion so far as current production of serviceable goods is con- cerned. SIZE OF THE NATIONAL INCOME 79 It is interesting to compare these new figures for National Income in money of constant pur- chasing power with the index numbers of the physical volume of production which have re- Table 15 comparison of the fluotdations in the final estimate of the puechasing power of the national income at the price level of 1913 with four index numbers of the physical volume of production 1909-1919 Amounts in 1913 = 100 Year Relative Fuicbaslng Power Index Numbers of Physical Tolum* of the National Income at of Production, Compiled by the Price Level of 1013 Estimate Estimate Final by by Estimate E . E. Day W. W. Carl W.I. Sources Incomes (1) Stewart Snyder King of Pro- Received (2) (3) (4) duction 1909 85 1910 91 95 94 93 95 91 89 1911 89 95 92 89 92 90 88 1912 95 98 97 102 105 97 95 1913 100 100 100 100 100 100 100 1914 94 97 96 98 100 97 96 1915 99 105 102 105 111 104 106 1916 116 120 118 111 116 118 126 1917 118 119 119 114 123 125 119 1918 110 115 113 113 124 129 113 1919 112 107 119 116 110 (1) Review of Economic Statistics, Harvard University Committee on Economic Research, Vol. 3, No. 1, January, 1921, p. 20. Weighted according to values in 1909. Includes 90 Farm Products, 10 Minerals and 88 Products of Manufactures. (2) American Economic- Review, March, 1921. Includes 91 products. (3) These figures have not been published. Include 87 commodities. (4) Bankers' Statistics Corporation; Special Service, Vol. 2, No. 12, August 24, 1920. cently been made by four statisticians working independently of each other. It will be seen from Table 15 that the two estimates of the National Income in money of constant purchasing power fluctuate in closer harmony with each other than 80 INCOME IN THE UNITED STATES do any two of the index numbers of physical vol- ume of production. And from Chart 18, it ap- pears clearly that these two estimates, or rather Chakt 18. COMPAEISON OF THE FLUCTUATIONS IN THE FINAL ESTIMATE OF THE PUECHASING POfWER OF THE NATIONAL INCOME AT THE PRICE LEVEL OF 19ia, WITH FOUR INDEX NUMBERS OF THE PHYSICAL VOLUME OP PRO- DUCTION. 1910-1919. Amounts in 1913 = 100. Based upon Table 15. -mnONAL mCOME IflDac/nmmts orntomcnos- S.E.DAY. w.w.jSTEviemr. ■■■■• - cua. j/nrDix. "- K/.KWG. m /ste /s/3 ^Sf* /sla «« &/? /sS9 m the "final estimate" made from them, pursues an intermediate course through the field covered by the fluctuations of the physical-production index numbers. This comparison affords a further SIZE OF THE NATIONAL INCOME 81 indication that our results are substantially trust- worthy. VII. TOTAl, AND PEB CAPITA INCOME IN DIFFERENT COUNTRIES In 1919, Sir Josiah Stamp, one of the highest British authorities on income statistics, made a careful survey of all recent investigations into the wealth and income of the chief powers, and as- sembled his results in a summary table published in the Journal of the Royal Statistical Society.^ He used the year 1914 as base, because that year "at the outbreak of war, represents the latest date for which satisfactory statistics are generally available," and he graded the estimates for the several countries according to his estimate of their approximate accuracy. This work makes it easy to compare our final estimate of the National Income of the United States in 1914 with the best estimates for other countries. In reproducing Sir Josiah Stamp's table, we have made but three changes. (1) Pounds sterling are converted into dollars at their pre-war value. (2) The new estimate for the United States made by this Bureau is substituted for Sir Josiah 's continuation of Mr. King's 1910 ^See the issue for July, 1919, Vol. LXXXII, pp. 441-507. The table is on p. 491. 82 INCOME IN THE UNITED STATES figure. (3) This new American figure is put in Grade I instead of Grade II. Concerning Mr. King's former figure, Sir Josiah Stamp remarked, "As the estimate stands, unchecked by any taxa- tion data, it is in the second grade, but after the lapse of a few years, such statistics should be available from the recently instituted income tax as to make a much closer estimate possible."^ We think that time has come. The British estimate was made by Professor A. L. Bowley with a free use of materials drawn from Sir Josiah Stamp's British Incomes. It rests primarily upon income-tax returns, which include all incomes above £160 ($800) per year, and upon census data regarding wages and num- ber of persons following gainful occupations. This estimate Sir Josiah regards "as perhaps the most accurate available for any country." The source of the German estimate is Deutschlcmd's VolhswoUstand, 1888-1913, by Dr. Helfferich, di- rector of the Deutsche Bank. Dr. Helfferich used the Prussian income-tax data — which include incomes as low as 900 marks ($225), — supple- mented by estimates for evasion, which he puts at 10 per cent., and incomes of untaxed individuals 'See Journal of the Boyal Statistical Society for July, 1919, Vol. LXXXII, p. 462. SIZE OF THE NATIONAL INCOME 83 whom he credits with an average of 750 marks ($188), per year. These Prussian figures he ap- plies to the whole German Empire and gets a total which Sir Josiah Stamp thinks may be 6 per cent, too high or 9 per cent, too low. The French author- ity is Rene Pupin, La Richesse de la France devant la Guerre, 1916. Lacking income-tax data, M. Pupin made an estimate by sources of production. The main sources he distinguished are property in real estate, buildings, securities and banks, "the effort" of people employed in various industries, and "capital and labor" engaged in farming, busi- ness and the liberal professions. His results are supposed to be subject to an error of more than 10 but less than 20 per cent. Italian statistics of in- come are very weak. The figure used here is a cur- rent guess adopted by Professor E. L. Bogart in his book on The Direct Costs of the Present War, and is thought liable to an error which may exceed 40 per cent. The Austro-Hungarian estimate is another guess adopted in default of better figures by Professor Bogart. For Spain a rough approxi- mation has been made by Andre Barthe from such data as he could collect concerning income from property, wages, salaries, and profits. The Aus- tralian figures are taken from the War Census of 1915, when all persons over 18 years of age were 84. INCOME IN THE UNITED STATES required to report the amount of their property and incomes, — an undertaking which constitutes "perhaps the most thorough and complete attempt that has yet been made to ascertain national wealth. ' ' The Canadian figure rests on a " guess ' ' made by Sir Robert Giffen in 1903 and may well be far from the truth. Finally, Sir Josiah Stamp himself made the Japanese estimate from income- tax returns, plus a large allowance for evasion, and an average income of about $120 per year for the 8,500,000 families belonging to the "lower classes." From this review, it wiU be seen that the United States is the only country for which esti- mates have been made on the basis both of sources of production and of incomes received. It is true that the American income-tax figures are less sat- isfactory than the British or Prussian, because of their relatively high exemption limit — $3,000 for married people in 1914 as against $800 in the United Kingdom and $225 in Prussia — and be- cause the administration of the law certainly had not then and probably has not yet attained as high a degree of efiSciency as in countries where similar taxes have been long in operation. For example, in 1911 the number of persons assessed under the income tax in Great Britain was estimated (the SIZE OF THE NATIONAL INCOME 85 complicated sclieme of schedules makes impossible an accurate determination) at 5.7 per cent, of all persons having gainful occupations;^ in the United States the corresponding figures for 1913, 1914 and 1915 were none of them quite one per Table 16 summary showing the estimated national and per cap- ita income of various countries at the outbreak of war in 1914, and the approximate accuracy of the respective estimates Adapted from the Summary by Sir Josiah Stamp, Journal of the Royal Statistical Society;, July, 1919 Country Estimates based upon Approx- National Per the work of imate Income, Capita Ac- Millions Income, curacy; of Dollars Dollars Grade • United States National Bureau of Eco- nomic Research I $33,200 $335 United Kingdom Bowley, Stamp I 10,950 243 Germany Helfferich I 10,460 146 France Pupin II 7,800 185 Italy IV 3,890 112 Austria-Hungary IV 5,350 102 Spain Barthe IV 1,120 54 Australia Official, Knibbs I 1,260 263 Canada Giffen IV 1,460 195 Japan Stamp III 1,580 29 I. Estimate is not likely to be inaccurate to a greater extent than 10 per cent. * ' II. Estimate is not likely to be inaccurate to a greater extent than 20 per cent. * * III. Estimate is not likely to be inaccurate to a greater extent than 30 per cent. ' ' IV, Estimate may be inaccurate to a greater extent than 40 per cent. cent. Still the remarkable agreement between the Bureau's two American estimates made inde- pendently of each other gives one considerable confidence in their approximate accuracy even in 1914. It may be added that, since then, the Ameri- can income-tax data have become relatively more ^ Compare A. L. Bowley, The Division of the Product of Indas- try (1919), pp. 10, 11. 86 INCOME IN THE UNITED STATES inclusive than were the British data before the war. By 1918, the reduction in the exemption limit, the increase in money incomes, and improve- Chabt 19. THE ESTIMATED NATIONAL INCOME OF VARIOUS COUNTEIES IN 1914. Approximate accuracy of estimates indicated by grade numbers I to IV. Based upon Table 16. VmTED 1 ■ 1 Kwr;rY>iw ' ' 1 GEKMAnr |i 1 TRAKCE |l 1 AViJKM- 1 ■ 1 JTALY \n 1 JAPAS CANADA B AusnAUA [7] SPAm 5 to 15 20 25 INCOME IN SlUlOJtS or DOLLARJ 30 33 ments in administration had raised the number of persons reporting incomes over $2,000 to more than 7 per cent, of all persons gainfully employed.^ '■ If the returns between one and two thousand dollars be counted in, nearly 11 per cent, of all personal incomes are included in the Internal Eevenue Bureau 's tables. But these statistics for the lowest income class have little value for esti- mating National Income, because they are limited to single persons, and married people who do not live together. SIZE OF THE NATIONAL INCOME 87 Concerning the facts brought out by these inter- national comparisons, nothing need be added to Sir Josiah Stamp's brief commentary: "I may Chart 20. THE ESTIMATED INCOME PEK CAPITA OF VARIOUS COUNTRIES IN 1914. Approximate accuracy of estimates indicated by grade numbers I to IV. Based upon Table 16. WflTED STATES I AUSTRALIA I 1 VNITED KltiGDOM I 1 (MAM E 1 TKANCE I 1 GERMANY I 1 iViLY n 1 AV3nu4~ HVrfClARf w 1 iBWf 1 jAPAjr iZI < 3 S JO iS zo INOOME m DOLLARS zs 3.0 Si perhaps remark,", he said, "that the generally higher level of pre-war prices in America (which is reflected in the per capita average) cannot dis- count the immense absolute lead of the States in real wealth, or the rapidity of its increase. The 88 INCOME IN THE UNITED STATES difference between the United Kingdom and Ger- many is not so considerable as other writers have suggested, and the effects of the well-known thrift of the French nation are apparent. The Japanese are making immense strides, but over 60 per cent, of their population are engaged in agriculture, and live on an amount per head which would be impossible in Europe — indeed, a comparison with this leading Eastern nation's figures brings out" the fundamental difficulty of comparing the in- comes of peoples whose scales of value are radi- cally unlike.^ * Journal of the Boyai Statistical Society, July, 1919, p. 490. CHAPTEE 3 THE DISTRIBUTION OF THE NATIONAL INCOME The data from which Mr. King made the Esti- mate by Sources of Production enabled him to divide the value product of each industry into two parts: first, payments to employees; second, interest and rent payments to individuals, and profits. Similarly, the income-tax exemption limit led Mr. Knauth to divide the Estimate by Incomes Eeceived into two parts: incomes over, and incomes under, $2,000. Both these divisions possess interest, and together they form a good introduction to the rather technical study of the distribution of all incomes among persons which has been made by Mr. Macaulay. I. THE SHAKE OP EMPLOYEES IN THE NATIONAL INCOME The percentage of the value product of an in- dustry paid to employees for their services is not at all the same thing as what is sometimes re- ferred to as the "share of labor" in the product 89 90 INCOME IN THE UNITED STATES of that industry. For there is a great deal of work done that is paid for not in the form of agreed-upon wages or salary but rather in the form of profits (often referred to by economists as the "wages of management.") To determine the "share of labor" in the product of agriculture, for example, one would have not only to find the wages paid farm hands but also to split up the farmers ' own incomes into return for their labor and return for their land and capital. That task would in- volve some hypothetical division of a sum that is really not divisible. One can compute a farmer's "labor income" by supposing that it is the bal- ance of his income left after setting aside the aver- age rate of interest (whatever that may be) upon the value of his investment (if that can be ascer- tained). Or, one can compute what profits a farmer makes by supposing that the profit is the balance of his income left after setting aside aver- age wages (whatever they may be) for all the work he does (if one can find out how much he works). The first computation as usually carried out shows that the farmer gets very low wages. The second computation usually shows that he makes very small profits. Results equally en- lightening might be produced by applying methods equally hypothetical to the incomes of DISTRIBUTION OF NATIONAL INCOME 91 shopkeepers, repair men, and the many other oc- cupations conducted on a modest scale by men working on their own account. This task Mr. King has not essayed. But among the facts best known to most business men and easiest to estimate as a whole are the facts con- cerning the aggregate pay roll, including salaries as well as wages. There is nothing hypothetical about these figures, and their accuracy is subject to a margin of error probably no wider in the ma- jority of cases, and in many cases narrower than the margin of error in the estimate of the net value product of the industry. To the pay roll can be added pensions, compensation for accidents and any other payments made to employees — a figure that is less accurate but of minor size. The sum, to repeat, will not be the "share of labor", but only the share of hired labor, received in the form of wages, pensions, and compensation for acci- dents. Such figures, cast into the form of percentages of the net value products, are presented in Table 17 for the main industrial groups recognized in the Estimate by Sources of Production. The striking fact brought out by this table is the marked inequality of the percentages for different industries. The share of hired labor is very low 92 INCOME IN THE UNITED STATES Table 17 PEECENTAGES OF THE NET VAX,UE PRODUCT OF VARI- OUS INDXJSTEIES RECEIVED BY EMPLOYEES, IN THE FORM OP PAYMENT FOR SERVICES 1909-1918 Note: — These figures show merely the share of hired labor of all grades (received as wages, salaries, pensions, compensation for accidents and the like) in the net valvs product of the several industries. The net value product does not include raw materials, supplies or services received from other industries. These figures do not show the "share of labor" in industry or in the national income; neither do they show the total incomes of employees, many of whom have other sources of income besides their wages or salaries. Year All Agriculture ' Production Manufacturing Indiuitrles of Factories^ Hand IMinerals Trades" 1909 53.0 15.3 71.0 72.2 57.3 1910 52.2 12.5 73.7 71.6 58.9 1911 53.9 14.1 73.8 76.4 58.6 1912 54.9 14.4 71.4 74.5 59.3 1913 55.6 13.4 73.4 74.5 66.7 1914 54.7 12.7 72.7 77.8 58.9 1915 53.6 12.3 67.4 75.4 58.7 1916 51.9 11.7 60.9 68.7 57.8 1917 51.6 10.9 63.1 71.0 61.6 1918 54.0 9.9 70.6 78.1 59.6 Transportation Bank ■ Govern- XJnclassl- Railway, Street rail- Trans- Ing ment ' fied Express, way, Elec- portation Industries Sleeping-Car Switching , trio Light and Power, 1)7 Water and Terminal Tele- Companies graph and Telephone Companies 50.4 50.7 51.5 51.7 52.9 53.2 51.1 52.5 55.4 62.8 " Includes stock raising, market gardening, etc. ^Includes lumbering and shipbuilding. 'Includes building andjsonstruction other than shipbuilding. 'Includes schools anS government-operated enterprises under state and local as well as national governments. 1909 59.6 1910 60.3 1911 62.8 1912 64.2 1913 66.4 1914 66.3 1915 61.5 1916 60.9 1917 67.4 1918 78.2 83.5 26.6 93.3 60.4 75.0 24.3 92.2 6L7 81.7 26.5 91.6 " 61.9 77.7 28.6 91.7 62.6 79.1 31.6 91.7 63.2 85.6 31.9 91.6 63.3 79.2 34.5 91.3 62.0 72.2 35.5 91.4 56.8 79.1 34.8 90.8 52.6 83.2 36.7 90.5 52.5 DISTRIBUTION OF NATIONAL INCOME 93 in agriculture (about one-eighth of the value prod- uct on the average) because the farmer and his family do so much of their own work. It is low also Chart 21. PERCENTAGES OF THE NET VALUE PRODUCT OF VARI- OUS INDUSTRIES RECEIVED BY EMPLOYEES IN THE FORM OF PAYMENT FOR SERVICES. 1909-1918. Note: These percentagea show neither the "share of labor" in the value product nor the total income of employees. Based upon Table 17. iOOl ao 60V tnNMsur ,^ sir/Wjg-?^^ ^nnikl^^-^--'- AtUiNIlVSTRI.;S 1^ kmONG zo Asuev. Tints 1909 HIO m 191Z 191 i9l4 ma 1916 19J7 in banking (from a third to a half of the total) for a very different reason. Here most of the labor is hired, but the amount of work required is small in comparison with the capital invested. Then come the hand trades which are a little like farm- 94 INCOME IN THE UNITED STATES ing in the proportion of labor paid by profits to labor paid by wages, and local public utilities which are somewhat like banking in the proportion of capital invested to labor required. In mining, manufacturing, water transportation, and govern- ment work, the percentages oscillate about points not far from three-quarters of the total. For all industries combined, the proportion of the product paid to employees is kept down to slightly more than half of the total by the great importance of farming with its exceptionally low percentage, Another very interesting set of conclusions may be drawn from the year-to-year changes in these percentages. Except in banking and government work, which present obvious peculiarities, the per- centage of the net product going to employees fell between 1914 and 1916 and rose again between 1916 and 1918 (except in farming). The rapid rise of prices in the first period redounded imme- diately to the benefit of profit-makers. Wages lagged far behind prices in their rise ; but they be- gan to rise rapidly and the number of persons employed increased largely after the advance of prices had slowed down. The net result was that, by 1918, the employees in most industries were getting as large a slice of the product as before the war, and in some cases a decidedly larger slice. DISTRIBUTION OF NATIONAL INCOME 95 Their net gains were particularly noticeable in rail transportation, in local public utilities, in banking and in government work. The percentage for all industries in 1918 stands just a shade higher than in 1909, though not so high as in 1913. Table 17 shows, then, that a little more than half the total National Income is paid in the form of wages, salaries and the like to hired labor ; that this share varies widely from one industry to an- other with the elaborateness of organization and the amount of capital used per worker; and that in any given industry, the share varies from one year to another with changes in business condi- tions. But these conclusions, interesting as they are, raise more questions than they answer. (1) If we take only the highly organized, large-scale indus- tries, in which the net proceeds are most definitely allocated to wages, interest, rent and profits, what share do we find going to hired labor? (2) What part of the total payroll goes to high-salaried of- ficials, and what part to the manual workers and clerical staff? (3) What is the average per capita compensation of employees in the different indus- tries and how closely has this compensation fol- lowed changes in the cost of living? (4) How im- portant is the addition to their main incomes, 96 INCOME IN THE UNITED STATES whicli wage-earners and salaried men get from other sources? Tables 18, 19, 20 and 21 show what light our data throw upon these problems. The highly organized industries in our list that employ much labor and present satisfactory data for analysis include mining, large-scale manufac- turing, and the several branches of land transpor- tation. Roughly speaking, these industries pro- duce a third of the National Income. It is feasible to divide their net value products into two parts, compensation for hired labor, and compensation for management and the use of property. Need- less to say, management involves work, and even in these highly organized industries, this work is paid for in part by profits. It should also be noted that the available data come from "going concerns". Losses which such concerns suffer presumably are deducted from profits. But the losses of enterprises that go into bankruptcy or "fail to succeed" in any year are not likely to be reported in our sources, and such losses fall main- ly, though not exclusively, upon "management and property ' '. "We do not know how large such losses are, but they probably make an appreciable offset to the income received by active business men and investors. Even with these qualifications, the figures in DISTRIBUTION OF NATIONAL INCOME 97 Table 18 are highly significant. The share of the net value product paid in wages, salaries, pen- sions and the like varies from two-thirds to a lit- tle more than three-quarters. Conversely "man- agement and property" receive from a third to less than a quarter of the net proceeds. These variations in the respective shares are due mainly to changes in business conditions, and during the war were probably more violent than usual. Both the high percentage that went to "management and property" in 1916 and the high percentage that went to hired labor in 1918 might prove to be outside the usual limits of fluctuation if we had data of this sort for a long series of "normal" years. Table 18 division of combined net value product op mines, fac- tories, and land transportation between earnings of employees and returns for management and the use of property 1909-1918 Note: — "Wages and salaries" includes all pensions, compensation for accidents, and the like. "Management and property" includes rentals, royalties, interest, and dividends. "Net value product" does not include raw materials, supplies, and services received from other industries. Millions of Dollars Per Cent. Year Wages and Management Wages and Management Salaries and Property Salaries and Property 1909 $6,481 $2,950 68.7 31.3 1910 7,156 3,250 68.8 31.2 1911 7,287 2,791 72.3 27.7 1912 7,993 3,169 71.6 28.4 1913 8,651 , 3,359 72.0 28.0 1914 7,947 2,816 73.8 26.2 1915 8,722 3,470 71.5 28.5 1916 11,630 5,810 66.7 33.3 1917 14,375 6,502 68.9 81.1 1918 17,472 6,124 77.8 22.7 98 INCOME IN THE UNITED STATES The division of the total payments for hired labor between the salaries of officials and the vast army of manual and clerical workers can be ef- fected very roughly for this same group of highly organized industries. Table 19 gives the best figures of this sort which Mr. King has been able to compile. The results confirm and make more precise two generally accepted opinions, (1) that the salaries of officials do not bulk large in the total payroll, and (2) that salaries are distinctly more stable than wages. The indications are that in highly organized enterprises, salaries absorb not much more than 7 or 8 per cent, of the payroll, and not more than 5 or 6 per cent, of the net value product. In prosperous times, they increase less rapidly than wages, but fall little if at all in hard times. Indeed, if our data are representative, sal- aries actually increased somewhat in the face of the depression of 1914. The net increase from 1909 to 1918 was 145 per cent, in salaries of offi- cials as against 172 per cent, in wages of manual and clerical employees. Concerning the average annual earnings of wage and salary earners and the fluctuations in the purchasing power of their incomes, Mr. King has been able to collect data which cover substan- tially the whole field of industry, though not in DISTRIBUTION OF NATIONAL INCOME 99 5 ec -^^ >H «5 irt CQ oq eq oq oq Oi o> Ci oi a t-^ t> t-' t-' t-^ •^_ iq !> O 05 iH rH cq* CO oq' Od 05 OS Oi O) «DIO« O iH 00 00 b-^ l> l> OSOOSCqrH 05^'~iTH-«* CO c^ 00 Ci ^ U5 O W 05 tH b- 3 ^o^b-Tt^tJ^otT ocToTi-r'^'^b^ 00 lo w 50 tH t^oawcot* CD O 00 Oi tH CO CO OS 00 O rH iH i-i iH iHiHrHi-frH oi Oi a Ci d oiOiOioa 100 INCOME IN THE UNITED STATES sufficient detail to permit of refined analysis. His results are summarized in Table 20. The top section of this table shows the average money earnings each year of all employees who normally make their living by working in the spe- cified industries. Since the people "attached to an industry" are never all at work, average earn- ings are somewhat lower than would be the earn- ings of an employee of average ability, who was able to work full-time throughout the year. Aver- age actual earnings are affected not only by "un- employment" in the usual sense of that term, but also by loss of time through sickness, voluntary periods of rest, and seasonal shiftings from one kind of work to another. In agriculture, particu- larly, the average employee has a short working season so that yearly earnings of most "farm hands" are meager even when they are getting good wages by the day or month. The figures in the table do not show changes in wage rates or in "the price of labor", but something more signifi- cant — ^namely, the average earnings that the em- ployees in different industries have realized each year under the conditions of pay, employment, and health that actually obtained. More significant still is the middle section of the table in which the purchasing power of money DISTRIBUTION OF NATIONAL INCOME 101 earnings is expressed in terms of 1913 prices. These figures were made by applying the Bureau of Labor Statistics index number of "the cost of living" on the 1913 base, to the money earnings of each year. According to these fig:ures, the eco- nomic condition of the average employee im- proved in all the industries covered from 1909 to 1913, though the improvement was slight in the hand trades, water transportation, agriculture, and the "unclassified industries". The grand average shows a gain of 10.6 per cent, in purchas- ing power in these four years. From 1913 to 1918, on the contrary, the grand average undergoes wide fluctuations, caused by the violent changes in wage rates and living expenses, the net effect of which was a decline of about 5 per cent, of the purchasing power enjoyed in 1913. This decline, however, was confined to four industries — govern- ment, whose enlistment of millions of soldiers brought down the average compensation sharply in 1918 ; public utilities which suffered to a pecu- liar degree from inability to raise their selling prices and which largely increased the proportion of their female employees ; the unclassified indus- tries ; and banking, in which salaries did not ad- vance so steadily as the cost of living. On the other hand, notable gains were scored by em- 102 INCOME IN THE UNITED STATES ployees of mines, factories, railways, and water- transportation companies. All these fluctuations are reduced to a comparable base by the "indices of the purchasing power of annual earnings" in the third section of the table. TABLE THE AVERAGE ANNUAL EARNINGS OF EMPLOYEES 1909- Denomiuatoi Caleu- of dar Earnings Year All Agilcul- ludustiies ' tuie ' Produo- Manufacturing tlon of Facto- Hand Minerals ' ries ' Trades ^ 1909 $626 $302 $599 $571 $699 1910 656 301 642 620 681 1911 648 317 647 609 657 1912 692 319 687 655 714 Current 1913 723 328 755 705 748 Money 1914 674 321 649 616 640 1915 697 330 656 653 693 1916 831 357 814 873 840 1917 961 463 1,025 1,022 945 1918 1.078 590 1,283 1,148 1,194 1909 $656 $316 $627 $597 $732 1910 671 308 656 634 696 1911 659 322 658 619 667 Value at 1912 696 321 691 659 719 Prices 1913 723 328 765 705 748 of 1913 1914 668 317 643 610 634 1915 677 320 637 634 673 1916 755 325 740 794 763 1917 745 359 795 792 732 1918 682 373 812 726 756 1909 90.7 96.3 83.0 84.7 97.9 Indices 1910 92.8 93.9 86.9 89.9 93.0 of the 1911 91.1 98.2 87.1 87.8 89.2 Furcliasing 1912 96.3 97.9 91.5 93.5 96.1 Power of 1913 100.0 100.0 100.0 100.0 100.0 Annual 1914 92.4 96.6 85.2 86.5 84.8 Earnings. 1915 93.6 97.6 84.4 89.9 90.0 Base, 1913 1916 104.4 99.1 98.0 112.6 102.0 1917 103.0 109.5 105.3 112.3 97.9 1918 94.3 113.7 107.5 103.0 101.1 ^ Includes amounts paid for pensions and compensation for injuries. 2 Includes payments for work done by contract. ^ Includes subsistence but excludes pensions. DISTRIBUTION OF NATIONAL INCOME 103 Finally, how mucli incame do employees receive from other sources than their wages, salaries, pen- sions and the like ? Definite data on this head are scarce, though everyone knows that many wage and salary earners eke out their living by small 20 NORMALLY ENGAGED IN VARIOUS INDUSTRIES 1918 TranBporta.tlon Banking Govern- ■Un- All Rail- Street Bail Trans- ment' classified Trans- way > Ex- way,' Eleo- porta- Industries porta- press, Pull- trin Light tion by tion' man, Switching and Termi- and Power, Telegraph and Water ' nal Cob. Telephone Ooe. $657 $651 $623 $773 $770 $739 $716 688 890 638 788 797 763 743 697 705 641 773 843 778 715 731 747 652 808 887 798 772 762 782 678 825 930 823 779 721 723 683 807 921 842 768 727 728 666 880 1,017 861 777 842 849 732 1,081 1,170 891 867 1,017 1,063 790 1,306 1,238 940 972 1,286 1,394 878 1,590 1,461 895 1,054 $688 $682 $653 $810 $807 $774 $750 703 705 653 806 815 780 759 709 716 652 785 857 791 727 736 751 656 813 892 803 777 762 782 678 825 930 823 779 714 716 676 799 912 833 760 706 707 647 854 987 836 755 765 772 665 983 1,064 810 788 789 824 613 1,012 959 729 753 814 882 656 1,006 925 567 667 90.3 87.2 96.3 98.2 86.8 94.0 96.3 92.3 90.2 96.3 97.7 87.6 94.8 97.4 93.0 91.6 96.2 95.2 92.2 96.1 93.3 96.6 96.0 96.8 98.5 95.9 97.6 99.7 100.0 100.0 100.0 100.0 100.0 100.0 100.0 93.7 91.6 99.7 96.8 98.1 101.2 97:6 92.7 90.4 95.4 103.5 106.1 101.6 96.9 100.4 98.7 98.1 119.2 114.4 98.4 101.2 103.5 105.4 90.4 122.7 103.1 88.6 96.7 106.8 112.8 82.0 121.9 99.5 68.9 85.6 104 INCOME IN THE UNITED STATES business ventures, taking boarders or lodgers, raising poultry, cultivating gardens, or keeping cows, and that many salaried men have substan- Chabt 22. THE PUECHASING POWER AT THE PEICE LEVEL OF 1913 OF THE AVEEAGE ANNUAL EAENINGS OF EMPLOYEES IN VAEIOUS INDTJSTEIES. 1909-1918. Based upon Table 20. 1000, tial incomes from investments of one kind or an- other. A study of 1602 school teachers, made by a Com- mittee on Teachers ' Salaries, indicated an income from investments of 6 per cent, of the total in- come. A similar study of 12,096 families by the DISTRIBUTION OF NATIONAL INCOME 105 Bureau of Labor Statistics, showed from 4 to 5 per cent, of the total income as coming from investments, but these families were selected so as to exclude those having a large percentage from Chaet 23. RELATIVE FLUCTUATIONS IN THE PUECHASING POWEE AT THE PRICE LEVEL OF 1913, OF THE AVERAGE ANNUAL EARNINGS OF EMPLOYEES IN MINING, MANUFACTURING, TRANS- PORTATION, AND ALL INDUSTRIES. 1909-1918. Annual earnings in 1913 = 100. Based upon Table 20. these sources. Chapin's study indicated that the New York working class received about 6 per cent, of their total income from sources other than earn- ings. An investigation by the United States Public 106 INCOME IN THE UNITED STATES Health Service in South Carolina showed that in 1917 families of cotton mill workers derived about 12 per cent, of their income from miscellaneous sources. The higher percentage in South Caro- Chabt 24. EELATIVE FLUCTUATIONS IN THE PURCHASING POWER AT THE PEICE LEVEL OF 1913, OF THE AVERAGE ANNUAL EARNINGS OF EMPLOYEES IN AGRI- CULTURE, GOVERNMENT, UNCLASSIFIED, AND ALL INDUSTRIES. Annual earnings in 1913 = 100. Based upon Table 20. i&S lina is probably due to the fact that these mill workers live for the most part in villages where it is easy to raise gardens and keep cows, while the New York employees have few such opportunities. If ordinary salaried em- DISTRIBUTION OF NATIONAL INCOME 107 ployees are included with the wage earners, it ap- pears likely that 8 per cent, is not too high an allowance for income from sources other than earnings. That the higher salaried classes re- ceive a much larger proportion of their income from investments seems highly probable. If an estimate is to be made, then, of the sup- plemental incomes of wage and salary earners, it is desirable to break this class up into at least three sections. The Statistics of Income, pub- lished by the Bureau of Internal Revenue, makes possible a division of this sort. Before 1916, how- ever, no figures are available. Since the material is so fragmentary, it seems best to present only Table 21 a rough estimate for 1918 of the income from all SOURCES OF SALARY AND WAGE WORKERS Millions Per Cent. Per Cent, of of Total of Total Total Compensation for Services of Dollars ITatlonal N'ationaJ Employees liaving Incomes of Pay Boll Income Less than $5,000 $30,472 93.6 $5,000 to $20,000 » 1,378 4.2 Over $20,000* 725 2.2 All Classes $32,575 100.0 Total Income of Employees having Incomes of Less than $5,000' $32,910 54.5 $5,000 to $20,000 2 1,585 2.6 Over $20,000 ' 942 1.6 All Classes $35,437 58.7 Total Income of Non-Employees $24,929 41.3 Total Income of the Entire Population. . $60,366 100.0 ' Estimated at 1.08 times the total earnings. ' Estimated at 1.15 times the total earnings. ' Estimated at 1.30 times the total earni9gs. * Statistics of Income, 1918, p. 44. 108 INCOME IN THE UNITED STATES the division among the different classes as it ex- isted in 1918. The probabilities are that the divi- sion in the other years was somewhat similar if allowance is made for variations in the purchasing power of money. This estimate of the incidental income of the em- ployed classes is, of course, based upon an ex- tremely limited foundation, but it is believed, nevertheless, that even the crude figures presented are accurate enough to show in a very rough way the general magnitude of the quantities involved. Employees probably received in 1918, some three billions of dollars in addition to their wages and salaries — a sum representing approximately a twentieth of the National Income. II. PEBSONAL INCOMES ABOVE AND BELOW $2,000 PEE YEAB Since 1917, the income-tax law has required all single persons having incomes of over $1,000 a year and all married persons having, separately or jointly, incomes exceeding $2,000 a year to make returns to the Bureau of Internal Revenue. That provision of the law was responsible for two of the major sections of the Estimate by Incomes Eeceived. One of these sections is based primar- DISTRIBUTION OF NATIONAL INCOME 109 ily upon the income-tax data, supplemented by estimates of the amount of under-reporting and non-reporting of taxable incomes. The second section, dealing with incomes below the exemption limit, is made from census data concerning the number of persons following gainful occupations (after subtraction of the numbers included in the first section), and from estimates of the average incomes of persons in these occupations. Thus, the $2,000 line necessarily plays a prominent role in this estimate. And that division is a fortunate one, for the $2,000 line serves as well as any ar- bitrary line could to divide families enjoying at least modest comfort from families that can scarcely be called well-to-do. Hence Mr. Knauth has carried this line of division through those sections of the Estimate by Incomes Eeceived, which do not of themselves break in two at $2,000 — the sections dealing with farmers and with tax- exempt income. Further, he has rearranged his data for 1913-1916, when the family exemption limit was $3,000, on the $2,000 basis, and extended that distinction back to 1910-1912, when there was no income tax. In presenting the results of this work, corporate surplus is temporarily disregarded as an item of National Income. Beasons have already been 110 INCOME IN THE UNITED STATES given for believing that, during the years of high income-tax rates at least, no great amount of this income has been "realized" by stockholders.^ And no small part of these accumulated surpluses was probably lost in the readjustments of 1919 and the business depression of 1920-21 before the time came when they could be "realized" to ad- vantage. If the method of treating this item adopted here introduces serious inaccuracy into the figures, it doubtless reduces the amount of in- come assigned to the over-$2,000 class much more than it reduces the amount in the lower class. Drawing the $2,000 line through farmers' in- comes is a particularly delicate task. Several studies of the distribution of farmers' incomes have been made by experts in this field, so that Mr. Knauth has a statistical basis for his conclusions. But the statistical basis is narrow, and the appli- cation of ratios computed from a few hundred re- turns, no matter how carefully treated, to all the farmers in the country may involve an error that is considerable. Hence the general results of the inquiry will be presented for all incomes, for all except farmers ' incomes, and for farmers ' incomes by themselves. One final warning: The following figures for ^ See above. Chapter II, Section IV, pp. 43-45. DISTRIBUTION OF NATIONAL INCOME 111 incomes over $2,000 are not made on the same basis as the income-tax returns and are not com- parable with them. Not only does the Estimate by Incomes Received include income that evades the tax, but it also includes income that is not sub- ject to taxation, the large items of their own prod- uce consumed by farmers' families, the rental value of homes occupied by their owners, interest on tax-exempt bonds, and the minor item of sal- aries paid to state officials. In particular, the num- ber of farmers legally subject to income tax is very much smaller than an incautious reader might infer from these figures. Table 22 and the charts based upon it tell their own story. About the main facts of that story, there can be little doubt, though the details may be inaccurate. Certainly among the men, women and children gainfully employed in 1910, only a small fraction, perhaps as the table says one in twenty-five had an annual income exceeding $2,000. Certainly this ratio increased with the war-time rise of prices, perhaps it became one and a half persons out of every ten. Necessarily a much larger fraction of the total income than of income receivers belong above the $2,000 line — the table says a third of the income in 1910. Certainly, this fraction grew somewhat larger during the war, not 112 INCOME IN THE UNITED STATES merely because events pushed millions of small incomes above the $2,000 line (a condition par- ticularly characteristic of 1918 and 1919) but also because events for a time favored the increase in Table 22 PERSONAL INCOMES ABOVE AND BELOW $2,000 PER ANNUM 1910-1919 ALL INCOME RECEIVERS ACTUAL AMOUNTS RELATIVE AMOUNTS Year No. of Peisous Amount of Income No. of Persons Amount of Income Income Income Income Income Income Income Income Income less more less more less more less more than than than than than than than than $2,000 $2,000 $2,000 $2,000 $2,000 $2,000 $2,000 $2,000 Thousand persons Billion dollars Per cent. Per cent. 1910 34,352 1,411 $20.0 $9.9 96. 4. 67. 33. 1911 34,693 1,379 20.7 9.6 96, 4. 68. 32. 1912 34,969 1,411 21.6 9.9 96. 4. 69. 31. 1913 35,345 1,443 22.2 10.1 96. 4. 69. 31. 1914 35,752 1,444 22.2 9.8 96. 4. 69. 31. 1915 35,597 2,008 22.9 11.4 95. 5. 67. 33. 1916 35,366 2,748 26.0 15.6 93. 7. 62. 38. 1917 34,160 4,363 29.6 20.9 89. 11. 59. 41. 1918 35,021 5,291 36.8 23.2 87. 13. 61. 39. 1919 34,233 5,508 39.5 25.2 86. 14. 61. 39. ALL INCOME RECEIVERS EXCEPT FARMERS 1910 28,100 1,300- $16.3 $9.6 96. 4. 63. 37. 1911 28,400 1,300 17.2 9.4 96. 4. 65. 35. 1912 28,700 1,300 17.9 9.6 96. 4. 65. 35. 1913 29,100 1,300 18.3 9.8 96. 4. 65. 35. 1914 29,500 1,300 18.3 9.5 96. 4. 66. 34. 1915 29,400 1,800 18.7 10.9 94. 6. 63. 37. 1916 29,400 2,300 21.4 14.4 93. 7. 60. 40. 1917 29,050 3,000 24.7 17.0 91. 9. 59. 41. 1918 30,450 3,400 32.1 17.4 90. 10. 64, 36. 1919 29,800 3,500 34.9 18.9 89. 11. 65. 85. FARMERS 1910 6,252 111 $3.7 $ .3 98. 2. 93. 7. 1911 6,293 79 3.5 .2 99. 1. 95. 5. 1912 6,269 111 3.7 .3 98. 2. 93. 7. 1913 6,245 143 3.9 .3 98. 2. 93. 7. 1914 6,2S2 144 3.9 .3 98. 2. 93. 7. 1915 6,197 208 4.2 .5 97. 3. 89. 11. 1916 5,966 448 4.6 1.2 93. 7. 79. 21. 1917 5,110 1,313 4.9 3.9 80. 20. 56. 44. 1918 4,571 1,861 4.7 6.8 71. 29. 45. 65. 1919 4,433 2,008 4.6 6.3 69. 31. 42. 68. DISTRIBUTION OF NATIONAL INCOME 113 size of incomes already large (a condition par- ticularly characteristic of 1916 and 1917). This use of a fixed sum of money in studying the distribution of income has its advantages ; but Chaet 25. PEBCENTAGBS OF PBESONS RECEIVING INCCmES ABOVE AND BELOW $2,000 PER ANNUM. 1910-1919. Based upon Table 22. iOCt lllllllllll: IIIIIIP mm wm i^li m 1 90- Ra iii 970 (i fm S -&! ho WA iOlfS RECEl rimb '.LOW 2000., 'R~>rK. In. o i^lO m »ffi 1913 i9/* i^lS 1^16 i9f7 m 19J9 it may be misleading if it stands alone. For, from the viewpoint of economic welfare, a fixed money income was a rapidly changing quantity during the war. The division of income receivers by the $2,000 line in 1919 is very far from meaning what that division meant in 1913. Some point between 114 INCOME IN THE UNITED STATES $3,000 and $4,000 a year in the later year would be needed to give results comparable in economic significance with the pre-war division at $2,000, But the data are not in such shape that we can Chakt 26. PERCENTAGES OF TOTAL PERSONAL INCOME RECEIVED BY PERSONS WITH INCOMES ABOVE AND BELOW $2,000 PER ANNUM. 1910-1919. Based upon Table 22. ys/o i9it nrn. iSJs i9i4 i9ti im iQir /m ioia draw dividing lines through the whole body of income receivers at any point we like in successive years. Another approach to the problem, however, which supplements the preceding results in an DISTRIBUTION OF NATIONAL INCOME 115 interesting way, is feasible. We can estimate in each year for which we have income-tax statistics — estimate very roughly — the amount of income received by the highest 5 per cent, of the per- sons having incomes. Studies made by the In- ternal Revenue Bureau show that the individuals included within any such group change much from year to year ; but that fact is not disturbing. Nor is 5 per cent, of the income receivers a group limited to the wealthy; for, to include the highest 5 per cent, of all income receivers, we have to take in all incomes above $2,000 in 1913 and 1914, above $2,100 in 1915, above $2,600 in 1916, above $2,900 in 1917, above $3,300 in 1918, and above $3,400 in 1919. The conjectural element in the estimate arises from the difficulty of allocating non-taxable income among different income classes, of making proper allowances for under- reporting and non-reporting of incomes, and par- ticularly of distributing the farmers along the in- come scale. This last difficulty is especially seri- ous, so that we give the results in two forms, first including and then excluding the farmers. What the results indicate is that about a third of the National Income went to the most prosper- ous twentieth of the income receivers in 1913 to 1916. But after 1916 the money incomes of this 116 INCOME IN THE UNITED STATES class increased less rapidly than did those of the other nineteen-twentieths, so that the share of the total received by the most prosperous 5 per cent, dropped in 1919 to about a quarter of the total. From this point of view, also, the evidence indicates that the inequality in the distribution of income declined somewhat during the war. Table 23 A CONJECTURAL ESTIMATE OF THE PERCENTAGE OP THE NATIONAL INCOME RECEIVED BY THE HIGH- EST FIVE PEE CENT. OF INCOME RECEIVERS 1913-1919 Including Farmers Year Income of the Total Individual Per Cent, of Total Highest 5% of Income (excluding Income Received Income Receivers Corporate Surplus) by Highest 5% of (Billion Dollars) (Billion Dollars) Income Receivers 1913 $10.6 $32.3 33 1914 10.3 32.0 32 1915 11.1 34.3 33' 1916 14.3 41.6 34' 1917 14.7 50.5 29' 1918 15.4 60.0 26- 1919 15.5 64.7 24'' Excluding Farmers 1913 $ 9.9 $28.1 35 1914 9.6 27.8 34' 1915 10.4 29.6 35-- 1916 12.8 35.8 36' 1917 13.6 41.7 32 1918 13.9 49.5 28' 1919 14.4 53.8 27 III. THE DISTRIBUTION OP INCOME AMONG INDIVIDUALS The standard method of showing how incomes are distributed among individuals is to use "fre- DISTRIBUTION OF NATIONAL INCOME 117 quency tables. ' ' The following table, taken from the official Statistics of Income for 1918 is a good example of this device. Table 24 THE DISTRIBUTION OF PERSONAL INCOMES BY INCOME CLASSES AS SHOWN BY THE OFFICIAL , COMPILATION FOE THE CALENDAR YEAR 1918 Income Classes Number of Amount of Percentage Percentage Returns Incomes Number Amount (Millions of Returns of Income of Dollars) $ 1,000- $ 3,000 1,516,938 $2,232 34.28 14.02 2,000- 3,000 1,495,878 3,627 33.83 22.78 3,000- 5,000 932,336 3,535 21.06 22.20 5,000- 10,000 319,356 2,146 7,22 13.47 10,000- 25,000 116,569 1,737 2.63 10.90 25,000- 50,000 28,542 978 .65 6.14 50,000- 100,000 9,996 680 .23 4.27 100,000- 150,000 2,358 284 .05 1.78 150,000- 300,000 1,514 305 .035 1.92 300,000- 500,000 382 145 .009 .91 500,000-1,000,000 178 119 .004 .75 1,000,000 and over 67 137 .002 .86 4,425,114 $15,925 100.000 100.00 Such tables show certain features of the dis- tribution of income admirably, but they do not give a clear picture of many peculiarities of the distribution as a whole. To show the facts all at once in their relations to each other it is desirable to use graphic methods. But ordinary charts drawn on an arithmetic or natural scale do not serve the purpose. For ex- ample, if incomes be plotted along a horizontal line with one-tenth of an inch for each thousand dollars, the chart becomes unmanageably long — 42 feet of paper are required to reach $5,000,000, ahd one income larger than that was reported in 1918. 118 INCOME IN THE UNITED STATES Even that size is too small wlien the distribution of all incomes is to be presented; for below the $1,000 line differences of income at least as small as $100 per year become highly important. To make such intervals easily visible and keep the scale uniform so as not to distort the picture, over 400 feet of paper would be needed. Even more impractical demands for space are made by the vertical scale showing number of persons. N"or can the difficulty be met by breaking the problem into parts and drawing the several sections of the curve on different scales. For these sections with their dissimilar scales will not fuse into the single picture that is wanted. And taken singly no one of the sections can give an illuminating impres- sion of the curve as a whole. A more illuminating device than the natural- scale chart was used about 1896 by Vilfredo Pa- reto, when he plotted income-tax data on logarith- mic paper, such as engineers use for many pur- poses. The logarithmic scale (which assigns equal spaces to each step in such a series as 100, 200, 400, 800, 1600, etc.) makes it possible to plot both the small and the large incomes and the small and large number of income receivers on a single sheet of paper and to do it in such a way DISTRIBUTION OF NATIONAL INCOME 119 that the characteristic features of both ends of the curve may be observed. Pareto, indeed, made large claims for the re- sults attained by his use of the double logarithmic scale. He held that income-data distributions when plotted in this way give curves that closely approximate straight lines. Further, he held that income-tax figures from different countries and from different times, even data like house-rentals that presumably vary with incomes, all closely ap- proximate straight lines having nearly uniform slopes. In the first flush of his enthusiasm he even implied that his investigations indicated the im- possibility of altering substantially the propor- tions in which income is distributed among in- dividuals — the type of this distribution in all coun- tries at all stages of social development seemed to be immutable. Charts 27 and 28 illustrate Pareto 's device^ and show roughly in what degree the American in- come-tax returns for 1913 to 1918 conform to his "straight-line law". Anyone accustomed to use only charts drawn on a natural scale may be in- clined to say that the conformity is close. But » Pareto charted "cumulative" data while we are charting non- cumulative data. However, it may be mathematically proven that if the cumulative distribution be a straight line on the double logarithmic scale, the non-cumulative distribution will also be a straight line on that scale. 120 INCOME IN THE UNITED STATES 'oodGof DISTRIBUTION OF NATIONAL INCOME 121 oodoofff oo'o'of ooo'f Ob/ t)/ 122 INCOME IN THE UNITED STATES the ratio treatment involved in the double log- arithmic scale does so much compressing of the data, both for the incomes of large size and for the large numbers of income receivers, that in using it a very different standard of conformity should be set than is appropriate in interpreting natural-scale charts. And when one does look thus closely at the curves and especially when one actually tests their conformity to a straight line, one finds that the conformity is somewhat specious. (1) The lines are not straight. They show "bumps" and "hollows", — especially the most reliable of the set — that for 1918. Even if such surface irregularities be set aside as capable of being " smoothed out", the lines have slight but significant curvatures throughout their whole course. (2) The slope of the lines is not uniform. Nor can this lack of uniformity be attributed merely to the increase of population and the rise of prices, for such factors would simply shift the position of the curve as a whole without altering its form. Quite the contrary, the changes in slope suggest that changes in business conditions from one year to the next modified the distributions of income among people of large and of small means. In 1914-16 the slope grew less each year ^ with the * The income tas figures for 1916 are not strictly comparable DISTRIBUTION OF NATIONAL INCOME 123 increase of business activity and the enormous enhancement of profits. In 1916-18, on the con- trary, the slope grew steeper again as the in- crease of wages and salaries raised the smaller incomes and encroached upon profits. In 1919 the slope grew less again.^ Another most serious defect of "Pareto's Law", as Professor Pareto himself saw, is that it cannot be extended to include incomes below the tax-exemption limit. The extension of the logarithmic straight line involves the absurdity of an infinite number of persons having incomes just above zero. We have excellent reason to believe on the contrary that at some income-interval be- low the tax-exemption limit, but well above zero, there is a maximum number of incomes, and that once past this interval the numbers of incomes in successive intervals decline indefinitely. Considerations such as these have led Mr. Ma- with those for the other years. In 1916 a husband and wife mak- ing separate returns were tabulated as one person. The fact that the figures for 1913 report income for only ten months, while it lowers the log line, does not alter its slope. * Professor A. L. Bowley, Report from the Select Committee on Income Tax, 1906, pp. 81 and 227, and Professor A. C. Pigou, Economics of Welfare, p. 695, have followed the lead of Pareto, Cours d'Sconomie politique, p. 312, in curiously misinterpreting this matter of slope. The steeper the line (whether on a cumu- lative or non-cumulative basis), the less is the inequality of in- come. If all persons had the same income the distribution would be represented by a perpendicular line. The slopes are all technically negati/ve but the sense in which we have used the terms greater and less in the text is obvious. 124 INCOME IN THE UNITED STATES caulay, who had charge of this part of the Bu- reau's investigation, to put aside "Pareto's Law" as having at the present time little more than his- torical interest. But he has kept the double log- arithmic chart as a powerful instrument to be used in conjunction with other analytic devices in studying the nature of the distribution of in- comes. His task was to construct a curve which would represent the best approximation to the facts of income-distribution that can be made by adjusting the available data in conformity with current statistical principles. The materials which Mr. Macaulay had to use and the considerations which he had to keep in mind may be listed. 1. The income-tax data for 1918, the year for which the most complete returns were available, show the incomes of less than 3,000,000 out of more than 40,000,000 persons who had money in- comes according to the census.^ Further, these data had to be adjusted to include (1) the large number of persons, especially farmers and small business men, who failed to make any tax return whatever, (2) evasion by reporting persons, (3) non-monetary income, especially farm and garden ^The income-tax returns for the $1,000-$2,000 class are of but little use, because they do not include married people living together. DISTRIBUTION OF NATIONAL INCOME 125 produce consumed by their producers and the rental value of homes occupied by their owners, (4) income from tax-exempt securities, etc. Mr. Knauth had estimated the magnitude of these fac- tors; Mr. Macaulay had to distribute these amounts along the income curve in the most prob- able manner. 2. Mr. Knauth 's division of the Estimate by In- comes Eeceived into incomes of less and incomes of more than $2,000 was of help to Mr. Macaulay, though in the final adjustment of his curve to fit all the conditions that must be met he arrived at results slightly different from Mr. Knauth 's on this point. 3. To distribute the incomes of less than $2,- 000 Mr. Macaulay had to combine the results of many scattered pieces of evidence. His largest and most important groups of material consisted of data showing the distribution of the wages of employees in manufacturing industries, in tele- phone and telegraph companies, in several branches of transportation and the salaries of federal employees in the civil service. He also used the small samples available showing the distribution of the incomes of farmers. The curve for each of these groups was based upon the avail- able collections of data, weighting most heavily 126 INCOME IN THE UNITED STATES those collections which seemed most valuable as indices of the distribution of the particular type of income under consideration. While some of these collections of data included hundreds of thousands of persons, the total number repre- sented forms only a very small fraction of the millions of income receivers who had to be dis- tributed, and only in the case of farmers and civil service employees did the data profess to show an- nual incomes. Further, it was necessary to add estimates of income from other sources to the in- come from wages, salaries, and farm profits which the data showed. 4. In every year many, men in business lose money. The Estimates of the National Income by Sources of Production and by Incomes Received are made on a net basis, so far as possible. That is, negative income, so far as known, is de- ducted from positive income in computing the total. Mr. Macaulay had to estimate the number and aggregate amount of negative incomes before he could distribute the number and amount of pos- itive incomes. For these estimates his materials were especially scanty, 5. Statistical experience in dealing with fre- quency curves representing vast bodies of data justified "smoothing" the curve. There is a DISTRIBUTION OF NATIONAL INCOME 127 strong a priori probability that the income-curve has a single "mode" or apex, and that it has not many "bumps," or " rolls "^ when charted on a double logarithmic scale. This a priori expecta- tion is supported by the largest and best accredited collections of data that Mr. Macaulay found, such as the income tax figures, the great official investi- gations into wage rates, and (making allowance for the smallness of the sample) Mr. Arthur T. Emery's very careful investigation into the total incomes of 2,000 Chicago households. Such col- lections of data were also suggestive and enlight- ening as to many peculiarities which might be expected in the shape of the final income curve. The final distribution, of which a part charted on the natural scale is shown by Chart 29 and a much larger part charted on a double logarithmic scale is shown by Chart 30, was built up by an elaborate series of adjustments to fit as well as might be all these considerations. The resulting curve is strictly empirical. It is fitted to adjusted data and is not a mathematical construction except through a very small part of its range. How ac- ^That is, the curve has not numerous "points of inflexion" when charted on a double logarithmic scale. The above state- ment and the statement concerning "smoothness" must not be interpreted as meaning that the income distribution is statistically homogeneous or can be adequately described by any mathematical equation suitable to describing distributions of homogeneous data. 128 INCOME IN THE UNITED STATES « a- s sfM/3i Income Class Number of Persons Total Income Under Zero' 200,000 $ — 125,000,000 $ to $ 100 62,809 3,368,863 100 to 200 103,704 16,047,939 200 to 300 209,087 53,701,566 300 to 400 489,963 174,747,705 400 to 500 961,991 437,421,733 500 to 600 1,549,974 857,666,411 600 to 700 2,154,474 1,405,213,223 700 to 800 2,668,466 2,005,009,301 800 to 900 3,013,034 2,563,100,947 900 to 1,000 3,144,722 2,987,688,735 1,000 to 1,100 3,074,351 3,226,729,363 1,100 to 1,200 2,850,526 3,275,784,572 1,200 to 1,300 2,535,285 3,166,235,800 1,300 to 1,400 2,205,728 2,973,220,322 1,400 to 1,500 , 1,832,230 2,653,820,477 1,500 to 1,600 1,512,649 2,342,101,155 1,600 to 1,700 1,234,397 2,034,621,765 1,700 to 1,800 999,996 1,748,225,207 1,800 to 1,900 811,236 1,499,396,953 1,900 to 2,000 663,789 1,293,303,255 2,000 to 2,100 549,787 1,126,240,869 2,100 to 2,200 463,222 995,402,469 2,200 to 2,300 395,115 888,501,304 2,300 to 2,400 340,141 798,920,154 2,400 to 2,500 295,490 723,614,676 2,500 to 2,600 258,650 659,277,149 2,600 to 2,700 227,731 603,250,834 2,700 to 2,800 201,488 553,889,766 2,800 to 2,900 178,901 509,693,726 2,900 to 3,000 154,499 455,622,047 3,000 to 3,100 142,802 435,416,064 3,100 to 3,200 128,217 40.3,770,475 3,200 to 3,300 115,583 375,547,256 3,300 to 3,400 104,504 350,001,254 ^Excluding soldiers. ' Including soldiers. ° Negative incomes — i.e., net loss for year. DISTillBUTION OF NATIONAL INCOME 133 Table 25 (Continued) Income Class Number of Persons Total income $ 3,400 to $ 3,500 94,803 $ 326,995,740 3,500 to 3,600 86,405 306,672,255 3,600 to 3,700 79,023 288,376,342 3,700 to 3,800 72,562 272,057,360 3,800 to 3,900 66,900 257,520,712 3,900 to 4,000 61,894 244,442,121 4,000 to 5,000 430,474 1,913,291,198 5,000 to 6,000 234,721 1,280,426,762 6,000 to 7,000 143,330 926,352,841 7,000 to 8,000 94,927 708,947,016 8,000 to 9,000 66,511 563,480,394 9,000 to 10,000 48,335 457,976,300 10,000 to 11,000 36,432 381,732,274 11,000 to 12,000 28,306 324,954,833 12,000 to 13,000 22,473 280,498,570 18,000 to 14,000 18,174 245,042,041 14,000 to 15,000 14,951 216,555,666 15,000 to 20,000 46,869 805,775,269 20,000 to 25,000 24,857 553,731,410 25,000 to 30,000 15,205 415,329,030 30,000 to 40,000 17,063 589,416,333 40,000 to 50,000 8,851 394,040,324 50,000 to 60,000 5,220 285,043,633 60,000 to 70,000 3,389 219,188,048 70,000 to 80,000 2,361 176,418,311 80,000 to ^0,000 1,730 146,629,939 90,000 to 100,000 1,311 124,249,645 100,000 to 150,000 3,494 421,980,443 150,000 to 200,000 1,451 249,585,378 200,000 to 250,000 771 171,676,103 250,000 to 300,000 460 125,604,380 300,000 to 400,000 497 170,757,868 400,000 to 500,000 248 101,980,849 500,000 to 750,000 265 139,293,673 750,000 to 1,000,000 104 80,826,726 1,000,000 to 1,500,000 79 94,956,294 1,500,000 to 2,000,000 30 51,697,546 2,000,000 to 3,000,000 24 57,818,419 3,000,000 to 4,000,000 9 30,846,960 4,000,000 and over 10 81,000,000 Total 37,569,060 $57,954,722,341 134 INCOME IN THE UNITED STATES Table 25 (Contirmed) Income Class Number of Persons Total Income Under $2,000 32,278,411 $34,592,405,292 Over 2,000 5,290,649 23,362,317,049 Total (excluding 2,500,000 soldiers, sailors and ma- rines^) 37,569,060 $57,954,722,341 Soldiers, sailors and marines' 2,500,000 1,750,000,000 Grand Total 40,069,060 $59,704,722,341" ^Of the total number of soldiers, sailors and marines, 2,500,000 are taken as having an average income of $700. ' To make this figure comparable with the estimates of Mr. King and Mr. Knauth, it is necessary to add $1,700,000,000 (Mr. Knauth's estimate) for corporate surplus. When this addition is made, the three totals are, in biUions : Mr. King $60.4 Mr. Knauth 61.7 Mr. Maoaulay 61.4 Table 26 the percentage analysis of the distribution of personal incomes in 1918 (Excluding 2,500,000 soldiers, sailors and marines) (Based upon Table 25) Income Class Percentages of Total Cumulative Percentages Number Amount Over the Under the of of Class Below Class Above Persons Income Number Amount Number Amount of of of of Persons Income Persons Income Under Zero .53 — .22 100.00 100.00 .53 22 $ Oto $ 100 .17 .01 99.47 100.22 .70 — '2I 100 to 200 .28 .03 99.30 100.21 .98 —.18 200 to 300 .56 .09 99.02 100.18 1.64 —.09 300 to 400 1.30 .30 98.46 100.09 2 84 21 400 to 500 2.56 .75 97.16 99.79 5.40 .96 500 to 600 4.12 1.48 94.60 99.04 9.52 2.44 600 to 700 5.73 2.43 90.48 97.56 15.26 4.87 700 to 800 7.10 3.46 84.76 95.13 22.36 8.38 800 to 900 8.02 4.42 77.65 91.67 30.37 12.76 900 to 1,000 8.37 5.16 69.63 87.25 38.74 17.91 1,000 to 1,100 8.18 5.57 61.26 82.09 46.92 23.48 1,100 to 1,200 7.59 5.65 53.08 76.52 64.51 29.13 1,200 to 1,300 6.75 5.46 45.49 70.87 61.26 84.59 1,300 to 1,400 5.87 6.13 38.74 66.4], 67.13 39.72 1,400 to 1,500 4.88 4.58 32.87 60.28 72.01 44.30 1,500 to 1,600 4.03 4.04 27.99 55.70 76.04 48.34 1,600 to 1,700 3.29 3.51 23.96 51.66 79.33 61.85 1,700 to 1,800 2.66 3.02 20.67 48.15 81.99 64.87 1,800 to 1,900 2.16 2.59 18.01 45.13 84.16 67.46 1,900 to 2,000 1.77 2.23 15.85 42.64 85.92 69.69 2,000 to 2,100 1,46 1.94 14.08 40.31 87.38 61.83 DISTRIBUTION OF NATIONAL INCOME 135 Table 26 (Contimed) Income Class Percentages of Total Cumulative Percentages Number Amount Over the Under the of of Olass Below Class Above Persona Income Number Amount Number Amount of of of of Persons Income Persons Income $ 2,100 to $ 2,200 1.23 1.72 12.62 38.37 88.61 63.35 2,200 to 2,300 1.05 1.53 11.39 36.65 89.66 64,88 2,300 to 2,400 .90 1.38 10.34 35.12 90.66 66.26 2,400 to 2,600 .79 1.26 9.44 83.74 91.35 67.61 2,500 to 2,600 .69 1.14 8.66 32.49 92.04 68.65 2,600 to 2,700 .61 1.04 7.96 31.35 92.65 69.69 2,700 to 2,800 .54 .96 7.35 30.31 93.19 70.65 2,800 to 2,900 .48 .88 6.81 29.35 93.67 71.63 2,900 to 3,000 .41 .79 6.33 28.47 94.08 72.32 3,000 to 3,100 .38 .75 6.92 27.68 94.46 73.07 3,100 to 3,200 .34 .70 6.54 26.93 94.80 73.77 3,200 to 3,300 .31 .66 6.20 26.23 96.11 74.42 3,300 to 3,400 .28 .60 4.89 25.58 95.39 75.02 3,400 to 3,500 .25 .58 4.61 24.98 95.64 75.68 3,500 to 3,600 .23 .63 4.36 24.42 95.87 76.11 3,600 to 3,700 .21 .60 4.18 23.89 96.08 76.61 3,700 to 3,800 .19 .47 3.92 23.39 96.27 77.08 3,800 to 3,900 .18 .44 3.73 22.92 96.45 77.62 3,900 to 4,000 .16 .42 3.55 22.48 96.61 77.94 4,000 to 5,000 1.15 3.30 3.39 22.06 97.76 81.24 5,000 to 6,000 .62 2.21 2.24 18.76 98.38 83.45 6,000 to 7,000 .38 1.60 1.62 16.66 98.76 85.06 7,000 to 8,000 .26 1.22 1.24 14.95 99.01 86.27 8,000 to 9,000 .18 .97 .99 13.73 99.19 87.21 9.000 to 10,000 .13 .79 .81 12.76 99.32 88.03 10,000 to 11,000 .10 .66 .68 11.97 99.42 88.69 11,000 to 12,000 .075 .56 .58 11.31 99.495 89.25 12,000 to 13,000 .060 .48 .606 10.76 99.665 89.73 13,000 to 14,000 .048 .42 .446 10.27 99.603 90.16 14,000 to 15,000 .040 .37 .397 9.85 99.643 90.52 15,000 to 20,000 .125 1.39 .367 9.48 99.768 91.91 20,000 to 25,000 .066 .96 .232 8.09 99.834 92.87 25,000 to 30,000 .040 .72 .166 7.13 99.874 93.59 30,000 to 40,000 .045 1.02 .126 6.41 99.919 94.61 40,000 to 60,000 .024 .68 .081 6.39 99.943 95.29 50,000 to 60,000 .0139 .49 .067 4.71 99.9569 95.78 60,000 to 70,000 .0090 .38 .0431 4.22 99.9659 96.16 70,000 to 80,000 .0063 .30 .0341 3.84 99.9722 96.46 80,000 to 90,000 .0046 .25 .0278 3.54 99.9768 96.71 90,000 to 100,000 .0035 .21 .0232 3.29 99.9803 96.92 100,000 to 150,000 .0093 .73 .0197 3.08 99.9896 97.65 150,000 to 200,000 .0038 .43 .0104 2.36 99.9934 98.08 200,000 to 250,000 .0020 .30 .0066 1.92 99.9954 98.38 250,000 to 300,000 .00122 .22 .0046 1.62 99.99662 98.60 300,000 to 400,000 .00132 .30 .00338 1.40 99.99794 98.90 400,000 to 600,000 .00066 .18 .00206 1.10 99.99860 99.08 600,000 to 750,000 .00071 .24 .00140 .92 99.99931 99.32 750,000 to 1,000,000 .00028 .14 .00069 .68 99.99969 99.46 1,000,000 to 1,500,000 .00021 .16 .00041 • .54 99.99980 99.62 1,500,000 to 2,000,000 .00008 .09 .00020 .38 99.99988 99.71 2,000,000 to 3,000,000 .00006 .10 .00012 .29 99.99994 99.81 3,000,000 to 4,000,000 .00003 .06 .00006 .19 99.99997 99.86 4,000,000 and over .00003 .14 .00003 .14 100.00000 100.00 Total... 100.00000 100.00 136 INCOME IN THE UNITED STATES O CDCO o o cq O O0 5D uTcTco cq CD t- iH lO CO m o CO lO 5D CO ?0 OS CX) CD b» CO oi^o^co tH »-l O OS 00 CO^Oi^O^CO CD o ocTco O 00 rH CM CS^O^O OO iH t-TrH* CO ■^ ■* w o CD (M oq O cxTt-T CO OS i-H^CO 'uTco' (M b- cq i-l CO fO OCI iH OQ CM Cvl tH CO 00 O (M cq (M Tji b» Qtroo''Q(rTjr "^ rH CO lO oo^Th^co^Oi^ lo tn ic m W OO w W t- CD O .WTt101{MW ; o o^io^oq^b^co^cD^TH^co^io^Tti^'^^TH^in rn ' co?OTjtrHiHosoaTtiw:ocq W^CO W O lO CQ (M 00 (M b-" t-" lo'co'" O" Ift'oo' « 00 CO oa r-t Tt< CO OS o i-H od OS 00 CO OCDb-TtC0«b* gOiHCOCOOSt^f-tCDOO ^00 CO lO CO OS t- CO O^CD cT i^Too'^cr oo" oa" o" co" cT 1/5 OOOb-OS'^rHOSOOOSOO O^oa CO t^ CD ^ O 1-H 05^^ irT m"'(xr lo" i>^TtH''Tfri>roo''oo' MOOiHOSiHiHt-COOOS iH (D cq^oa^os^co^iH^os^oq^co CO fHO t* (M O iO^00_^(M^ "oTirTor -Tt<(MOSrHU:5COOl°^ OU5t-(MCD(MCD '^ o o o o o A w o w o o OO OO O O o o^o^ g: o'^o'^o o 2S2'P o o o o o o i-TiH cq CO lO O I o o o o o o o o o o o o o o o o o_^o^ so^cTo^o" 3 O O O O rHM W O DISTRIBUTION OF NATIONAL INCOME 137 O o M a Q W EH Ph O li g o a .o a q a;C0«0S0i0S010SO 'loodcawTjit^osoiosoiojososo cot>ooosos03a30sososososo OcqTtCTrHb-fHC0l0 ONOrHt-;CO«Sl>;01iHt>;C>asOSlO O O Oi cm' lO O t-^ CO r-H t^ Tji ffO iH OOOsOOWtHCQiHi-) OCOiHTH-^lOiHTHCOOCQOlb-TH;^ Ob^fOOSWCq-^tMb-CDb-OSSOrHO O«DOTj1Q0C0eow W hU >_/ v.** I-^ C~ PS t~ \W STJ l_) C" ^i ^i CIS P3 S O o -^ '5^ t^ 0° "3 ® « M l> iri N « iH i-i T-i Til '^* oi « O lO CO »H tH B " ^ O d >-l lH iH rH rH i-l M « iH -H 1, a ■< H — m5 *■ >< «Q0Ot-OTl1t-Q0«rHU5«0 Ot-NTjIOlr-HNt-OiOlt-OO 1^ is rlOOClb- t}I O) CO t- lO ^ H O 00 CO l> « M J Wl-34 +=3 OSW<*C0t-I>Tt(OJ00T)(iOt- lOOJONt-OWrHOHncOOJ Sp^ fe S'S v„ rH«COCD'tirr-J'r-r^"lflin010" riwcO tO CO oTcO ffi rH iH «®2? T-ieo»o«>oooio r-iMWt- •-< M CO N 01 t- t- 05 CO O rH -^ CO S *cB ■ M 0<0 t-tn Oi Oi QOQOeOO COT# t-MOS tHOTCD OIQO t-OSCD^ QM rtW«* ^"*.'^^.°l^-9.^^^"l^'^ ■^^■<* t-t- 00 N €»■■*■«* CO C5)0_ m&H h (0 PJ^ iH M"cOCOTjr^"oOrH NlOi-TcO -4"m"cO ''O T^Tj^t^" CD W 00 Oo' w to t.3j^ 1-lcOOOCOt- iH' d s 5] fcOfH ® co-^toHoo-^-^Mt-cocoin o>aDW^NO»00Tjtt-inio ^CO g^ tO^CO^t-COOOlCqCnCOCO iHOJeOCOMCOCOtOCOt-CDCO t^ cj) o *o ^< * N 'l! t> o CO oo CO t> vrJ th ' w ^' od co' ih co ■^' ^li od ■•^1 H rH >■ oj <0 iH iH H « M CO H « co'in'o in o oo o*o"o" o w cow" o'ln o o o"o'o o ® iHMinOiflOO 0_, iHMWOinOOO — 03 S rHrHcOvnO-g r^rHCOlOO'^ lU « »■ tHos ee- th3 DQ O OOOOOOOOOOOO O O O O O O ci CD O O O O M OOOOOOOOOOOO OOOOOOOOOOOO Q © O O O O 0.0 OOOOOO OOOOOOOOOOOO S iH Nofino »rt"o*"o"o"oo"o i-rc3coio"oirto'"oo"oo'o" W O i-tNlOOWOOO iHMWOinOOO P^ r rHWComO iHiHcOWO gs DISTRIBUTION OF NATIONAL INCOME 141 To most minds, Charts 29 and 30 will probably give the clearest impression of the complex esti- mate set forth in our tables. But it is well to sup- plement these charts with a Lorenz curve repre- senting the same set of figures. This device, used Chart 31. LORENZ CURVE SHOWING THE DISTRIBUTION OF INCOMES IN 1918. Based upon the data presented in Table 26. 100 "•' ^^ -jyj '7\j -/w \./\y fw ow yij II TEBCENTAGE OF PERSOffS BECrmi/fC WITH THE POOREST. in Chart 31, shows graphically the deviation of the actual distribution of incomes from a perfectly even distribution. By looking at the two scales of this chart, the reader will see that if 10 per cent, of the income receivers got just 10 per cent, of the 142 INCOME IN THE UNITED STATES total income, if 20 per cent, of them got just 20 per cent, of the total income, and so on, then the actual distribution would be represented by the straight diagonal line of the chart. From the "line of 1918 income" and the two scales, it is easy to see ap- proximately what per cent, of the total income was obtained by any given percentage of the in- come receivers. For example, on the horizontal line, take the point marked "70 per cent."; follow the perpendicular line through this point to where it intersects the curve marked "Line of 1918 In- come"; from this point of intersection, draw an imaginary horizontal line to the left until it inter- sects the left-hand perpendicular scale ; it will be seen to intersect that scale at about "42% per cent. ' ' This signifies, according to the chart, that the poorest 70 per cent, of income receivers had about 421/2 per cent, of the National Income. Vice versa, the richest 30 per cent, had about 57^/^ per cent, of the National Income. CHAPTER 4 CONCLUSIONS 1. The size of the National Income in the United States during recent years can be deter- mined with a margin of error that is probably less than 10 per cent. 2. The final estimate of the National Income in 1909-1918 runs as follows : (Billions) 1909 $28.8 1910 31.4 1911 31.2 1912 33.0 1913 34.4 1914. 33.2 1915 36.0 1916 45.4 1917 53.9 1918 61.0 These figures do not include any allowance for the money value of the work done by housewives for their own families, an item which would add several billions to the money total if all housewives were paid on a commercial basis. 143 144 INCOME IN THE UNITED STATES 3. The final estimate of the National Income on a per capita basis is as follows : 1909 $319 1910 340 1911 333 1912 346 1913 354 1914 335 1915 358 1916 446 1917 523 1918 586 4. Most of the huge increase in the National Income during the war was due merely to the rise of prices. If the preceding figures are reduced to terms of prices in 1913, we get the following re- sults : National Income Income per Capita (BilUons) 1909 $30.1 $333 1910 32.2 349 1911 31.7 338 1912 33.2 348 1913 34.4 354 1914 33.0 333 1915 35.2 350 1916 40.7 400 1917 40.8 396 1918 38.8 372 5. Not only the National Income but also the Per Capita Income is much larger in the United States than in any other country. The following figures show the National and Per Capita Income at the outbreak of the war in countries for which estimates with a margin of error probably not exceeding 10 per cent, have been made. CONCLUSIONS 145 1914 National Income Income per (Billions) Capita United States $33.2 $335 United Kingdom 10.9 243 Germany 10.5 146 Australia 1.3 263 6. The share of the net value product of dif- ferent industries which is paid to employees as compensation for their services (not by any means equivalent to the "share of labor" in industry) varies from about one-eighth of the total in agri- culture to about three-quarters of the total in mining, manufacturing, water transportation and government work. This share in most industries declined with the sudden rise of prices in 1914-16 and rose again with the advance of wages and salaries in 1917-18. The average for all industries was a trifle higher in 1918 than in 1909, but not so high as it had been in 1913. 7. In the highly organized industries conducted on a large scale, the pay of employees, including the salaries of officials, absorbs in most years some 69-72 per cent, of the net value product. The re- maining 31-28 per cent, is the share of "manage- ment and capital." From it are paid interest, rent, and profits. Even in these highly organized industries, part of the work of management is paid for under the form of profits, and in some cases, 146 INCOME IN THE UNITED STATES the subordinate officials and wage earners also share in the profits. From 1909 to 1918, the extreme fluctuations in the share of management and profits varied from 33 per cent, of the net value product in 1916 to 23 per cent, in 1918. 8. Of the total payments to employees in the highly organized industries, about 92 per cent, goes to the manual workers and clerical staffs, while 8 per cent, goes to officials. 9. In 1918, the year for which the best data are available, about 86 per cent, of persons gain- fully employed had incomes of less than $2,000 per annum, and about 14 per cent, had incomes exceeding that sum. In the same year, about 60 per cent, of the National Income was divided among the 86 per cent, of the gainfully employed who had incomes less than $2,000 per annum, and about 40 per cent, of the National Income was divided among the 14 per cent, of the gainfully employed who had in- comes exceeding $2,000. The net effect of our participation in the war was to diminish somewhat (at least temporarily) the inequality in the distribution of American in- comes. 10. If we consider the 5 per cent, of those CONCLUSIONS 147 gainfully employed who had each year the largest incomes, we find that their share in the aggregate of personal incomes declined from about 33 per cent, in 1913-16 to about 25 per cent, in 1918-19. 11. Data regarding the detailed distribution of personal incomes are scanty and difficult to sys- tematize ; but the best approximation this Bureau has been able to make indicates that in 1918, the most prosperous one per cent, of the income re- ceivers had nearly 14 per cent, of the total income, the most prosperous 5 per cent, of the income re- ceivers had nearly 26 per cent, of the total, the most prosperous 10 per cent, of the income re- ceivers had nearly 35 per cent, of the total, and the most prosperous 20 per cent, of the income re- ceivers had about 47 per cent, of the total income. It should be noted that when we start from the top of the income scale, we must go down to people receiving $8,000 per annum, in order to include one per cent, of the income receivers. Similarly, to include 5 per cent, of the income receivers, we have to descend to incomes of $3,200-$3,300. To include 10 per cent., we must take in part of the $2,300-$2,400 class ; and to include 20 per cent, we must include part of the $1,700-$1,800 class. INDEX Agricultural laborers, wages of, 33. Agriculture, effect of business cycles on, 26, 27; employees, 92, 93, 102, 104, 106; value product, 21, 23, 24. Anderson, B. M., estimate of National Income, 65, 66, 68. Australia, census of incomes, 2, 83, 85-87. Austria-Hungary, income of, 85- 87 Banking, employees, 92, 93, 103 ; value product, 18, 19, 21, 23, 24, 26, 28. Bankers Trust Company, esti- mate of National Income, 65, 66. Barthe, Andr§, income of Spain, 83, 85-87. Bogart, E. L., income of Aus- tria-Hungary, 83; income of Italy, 83. Bowley, A. L., income of United Kingdom, 82, 85. Building, effect of war restric- tions on, 26. Business cycles, and incomes under and over $2,000, 35-37; and size of National Income, 74-77; and value product of different industries, 26-28. Canada, income of, 84-87. Census of incomes, Australia, 2, 83, 85-87. Construction, value product, 18, 19. See Mamufaetv/rmg. Consumption goods, interest on, a part of the national income, 25. Corporate surplus, 32, 34, 35, 38, 40, 41, 46-48. Cycles, business, see Business cycles. Data, see Sources of informa- tion. Day, B. E., index of physical production, 79, 80. Earnings, of employees, 100-103. See different industries such as Agriculture, Banking, etc. Edwards, A. M., estimate of number of housewives, 58. Electrical industries, employees, 92, 103 ; value product, 18, 19. Emery, A. T., incomes of 2,000 Chicago households, 127. Employees, income other than earnings, 103-108. See Earn- ings and industries such as Agriculture, BanTcing, etc., and Income, National, Share of employees. Errors, estimated, 60-68, 143. Estimate of National Income by Incomes Eeeeived, 12-15, 27-42; amount of income, 14, 16, 34-38, 72, 73, 79; effect of inadequate reporting, 44- 46; estimated errors, 60-68; income per capita, 13, 15; treatment of taxes, 49-51. Estimate of National Income by Sources of Production, 8-15, 16-27; employees, 92, 93, 102, 103, 105, 106; estimated errors, 60-63; income per capita, 13, 15; total value product, 13, 14, 16, 72, 73, 79; treatment of taxes, 49, 51-56; 149 150 INDEX value product by industrien, 18, 19, 21, 23, 24, 26-28. Exempt income, see Income, tax- exempt. Factories, employees, 92, 93, 102-105; value product, 18, 19. Farmers' income, 34-39, 112. France, income of, 83, 85-87. Frequency tables, distribution of income among individuals, 128, 129, 132, 134; income tax data, 116, 117. Friday, David, estimate of National Income, 65, 66, 68. Germany, income of, 82-87. Giffen, Sir Eobert, income of Canada, 84, 85. Government, effect of business cycles on, 26, 28; employees, 92, 93, 102-104, 106; value product, 18, 19, 21-24, 26, 28, 52. Hand trades, employees, 92, 102, 103; value product, 18, 19. See Manufacturing. Helfferich, Dr., income of Ger- many, 82, 83, 85. Home ownership, 42. Housewives, amount of National Income including, 67 (note 1) ; excluded from National Income, 42, 143; number, 57- 59; value of services, 57-60. Income, distribution among in- dividuals, 115-142 ; amount received by richest 5 per cent., 115, 116; data used in esti- mating, 124-127; frequency distribution, 128, 129, 132- 135; over $2,000, except farmers, 29-35, 108-116, 146; over $2,000, including farm- ers, 112, 114; Pareto and the double logarithmic scale, 118, 119; soldiers, sailors and marines, 131, 134; under $2,000, except farmers, 30, 31, 34-39, 108-116, 146; under $2,000, including farmers, 112, 114. Income, farmers', 33-35, 38, 39. Income, foreign countries, see Australia, Austria-Hunga/ry, Canada, France, Germany, Italy, Japan, Spain, United Ki/ngdom. Income, National, amount, 64- 66, 71, 74, 75, 79; amounts contributed by various indus- tries, 18, 19, 24, 26; Amer- ican compared with that of Australia, Austria-Hungary, Canada, France, Germany, Italy, Japan, Spain, Uniteij Kingdom, 79-88, 144, 145; commodities and services for which a price is not generally paid, 57-60; definition, 42; definition, difficulties of, 3, 4 ; estimates by Anderson, Bank- ers Trust Company, Friday, Ingalls, King, Miller, Na- tional Bureau of Economic Research, Stamp, 65-67; esti- mated errors in, 60-68; per capita income, 13, 68, 76-78, 80, 144; pre-war values, 68- 81; share of employees in, 89-108, 145, 146. Income, negative, 132, 134, 136. See Natural resources, deple- tion of. Income, personal or individual, see Income, distribution among individuals. Income receivers, number of, 112, 113. Income, tax-exempt, 31, 34-39. Income tax figures, 120, 121, 140; comparison of American, British and Prussian, 84; intensive drive for incomes under $5,000, 130; necessary adjustments in, 29, 30, 111. Incomes received. Estimate by, INDEX 151 see Estimate of the National Income by Income Seceived. Industries, large scale, share of employees, 95-97, 99, 146; share of management and property, 97. Industries, unclassified and mis- cellaneous income, employees, 92 ; value product, 18, 19, 21, 23, 26. Ingalls, W. E., estimate of Na;tional Income, 65-67. Italy, income of, 85-87. Japan, income of, 85-87. King, W. I., estimate of Na- tional Income for 1910, 66, 71; index of physical pro- duction, 79, 80. Knibbs, census of Australian income, 85. Lorenz curve, 141, 142. Manufacturing, effect of busi- ness cycles on, 26, 27; em- ployees' share, 92; value product of factories, con- struction and other hand trades, 18, 19, 21, 23, 24, 26, 27. Miller, A. C, estimate of Na- tional Income, 65, 66. Mineral production, effect of business cycles on, 26, 27; employees, 92, 93, 102-106; value product, 18, 19, 21, 23, 24, 26, 27. Mines or mining, see Mineral ■production. Miscellaneous income, see Indus- tries, unclassified. Mortgages, farm, interest on, 33. Natural resources, depletion of, 56, 57, 137. Pareto, Vilfredo, use of the logarithmic scale, 118, 119. Pareto 's law, 119-123. Personal income, see Income, distribution among individu- als. Population, U. 8., each year, 1909 to 1919, 13 Prices, fluctuation jn commod- ity, and size of National Income, 69-81. Probable errors of estimate, see Errors, estimated. Production, Estimate by Sources of, see Estimate of the Na- tional Income by Sources of Production. Production, indices of physical, 79, 80. Pupin, Een6, income o;E Prance, 83, 85. Eailways and allied industries, employees,"-92, 93, 102-105; value product, 18, 19. Bate regulation, transportation, 26. Eent paid by tenant farmers, 33. Eental value of homes owned by their occupants, 42. Snyder, Carl, index of physical production, 79, 80. Sources of information, 2, 3. Spain, income of, 83, 85-87. Stamp, Sir Josiah, accuracy of income estimates, 5; estimate of National Income, 66, 67; income of Japan, 84, 85 ; wealth and income of the chief powers, 81, 82, 85. Stewart, W. W., index of phys- ical production, 79, 80. Street railways, see Electrical industries. Surplus, see Corporate surplus. Taxes, treatment under Estimate by Incomes Eeceived, 49-51; treatment under Estimate by Sources of Production, 49, 51, 52. 152 INDEX Tax-exempt income, see Income, tax-exe'mpt. Telegraph and telephone com- panies, see Electrical indms- tries. Transportation, effect of busi- ness cycles on, 26-28; em- ployees, 104, 105, 107; value product, 18, 19, 21, 23, 24, 26, 28. Transportation by water, em- ployees, 92, 102, 103; value product, 18, 19. Unclassified industries, see In- dustries, unclassified. United Kingdom, income tax, 84; National Income, 85-87. Value product, definition of, 17 (note) ; method of esti- mating, 20. See Estimate of the National Income by Sources of Production, and different industries such as AgruyuUure, Banking, etc.