(navigation image)
Home American Libraries | Canadian Libraries | Universal Library | Community Texts | Project Gutenberg | Children's Library | Biodiversity Heritage Library | Additional Collections
Search: Advanced Search
Anonymous User (login or join us)
Upload
See other formats

Full text of "Investigation of concentration of economic power; monograph no. 1[-43]"

76th Cr essfon / SENATE COMMITTEE PRINT 



3dS. 



INVESTIGATION OF CONCENTRATION 
OF ECONOMIC POWER 



TEMPOEAEY NATIONAL ECONOMIC 
COMMITTEE 

A STUDY MADE FOR THE TEMPORARY NATIONAL 

ECONOMIC COMMITTEE, SEVENTY-SIXTH CONrRF^<J^ 

THIRD SESSION, PURSUANT TO Pulu^RESOLUTION 

No. 113 (SEVENTY-FIFTH CONGRESS) AUTHORT7mr 

AND DIRECTING A SELECT COMMITTEE TO MAKE A 

FULL AND COMPLETE STUDY AND INVESTIGATION 

WITH RESPECT TO THE CONCENTRATION orECONOMlC 

POWER IN, AND FINANCIAL CONTROL OVER 

PRODUCTION AND DISTRIBUTION 

OF GOODS AND SERVICES 



MONOGRAPH No 22-24 
TECHNOLOGY IN OUR ECONOMY 



Printed for the use of the 
Temporary National Economic Committee 



UNITED STATES 

GOVERNMENT PRINTING OFFICE 

WASHINGTON: 1941 



i^TwrAc^irr'^^ ?!!\!!v^p<^!Ty ^rwnnfnf ! wrmF7/r!?v 



TEMPORARY NATIONAL ECONOMIC COMMITTEE 

JOSEPH C. OMAHOXEY, Senator from Wyoming. Chairman 

HATT05f W. SUMNERS. Representative from Texas. Vice Chairman 

WILLIAM H. KING, Senator from Utah 

WALLACE H. WHITE, Jr., Senator from Maine 

CLYDE WILLIAMS, Representative from Missouri 

B. CARROLL REECE, Representative from Tennessee 

THURMAN W. ARNOLD. Assistant Attorney General 

♦WENDELL BERGE, Special Assistant to the Attorney General, 

Representing the Department of Justice' 

JEROME N. FR.\NK, Chairman 

•SUMNER T. PIKE, Commissioner, 

Representing the Securities and Exchange Commission 

GARLAND S. FERGUSON, Coniipissioner 

*EWIN L. DAVIS, Chairman, 

Representing the Federal Trade Commission 

ISADOR LUBIN, Commissioner of Labor Statistics 

*A. FORD HINRICHS, Chief Economist, Bureau of Labor Statistics, 

Representing the Department of Labor 

JOSEPH J. O'CONNELL. Jr.. Special Assistant to the General Counsel 

* CHARLES L. KADES, Special Assistant to the General Counsel, 

Representing the Department of the Treasury 

, , Representing the Department of Commerce 

LEON HENDERSON. Economic Coordinator 
DEWEY ANDERSON, Executive .Secretary 
THEODORE J. KREPS, Economic Adviser 



?cn 

Monograph No. 22 CT) 

TECHNOLOGY IN OUR ECONOMY 



LEWIS L. LORWIN And JOHN M. BL.\IR 

REPRINTED 
BY 

WILLIAM S HEIN &, CO , INC. 

BUFFALO. N. Y. 
1968 



CO 

O 



ACKNOWLEDGMENT 

This monograph was written under the general supervison of 
H. DEWEY ANDERSON 

Executive Secretary 
Temporary National. Economic Committee 

BY 

LEWIS L. LORWIN 

Economic Consultant 
Temporary National Economic Committee 

AND 

JOHN M. BLAIR 

Economist 
Temporary National Economic Committee 

ASSISTED BY 

RUTH AULL 

Technical Assi-'<tant 
Temporary National Economic Committee 

Dr. Lewis L. Lorwin is the author of Part I; John M. Blair has 
written Part II with the editorial assistance of. Ruth Aull. 

The Temporary National Economic Committee is greatly indebted 
to these authors for this contribution to the literature of the subject 
under review. 

The status of the materials in this volume is precisely the same as 
that of other carefully prepared testimony ivhen, given by individual 
witnesses; it i>i infoi^nation submitted for Committee deliberation. 
No matter ichat the offidrd capacity of the witness or author may be, 
the publication of hi^ testimony^ report^ or monograph by the Com- 
mittee in no way signifies nor implies assent to, or approval of, any 
of the facts, opinions, or recommendations, nor acceptaftice thereof 
in whole or in part by the members of the Temporary National 
Economie Committee, individually or collectively. Sole and undi- 
vided responsibility for every statement in su^h testimony, reports, 
or monographs 7'ests entirely upon the respective authors. 

(Signed) Joseph C. O'ISIahoney. 

Chairman, Temporary National Economic Committee. 



TABLE OF CONTENTS 



Pae* 

Letter of transmittal XI 

Foreword XV 

PART I 

The* problem of technological unemployment — a historical survey .. 3 

Introduction 3 

CHAPTER I 

Nineteenth century backgrounds 5 

Early statements 5 

The classical debate 6 

John Stuart Mill on machinery and the stationary state 10 

The Marxian analysis 13 

Approximate effects of machinery 14 

The "strife between workman and machine" 15 

Criticism of the "theory of compensation" 15 

Capital accumulation and the "Industrial Reserve Army" 19 

Pioneer inductive studies, 1886-99 25 

CHAPTER II 

Twentieth century problems 33 

Neo-classical theory and its critics 33 

Prosperity and technological unemployment in the United States, 

1924-29 . 37 

Technological change and the depression, 1929-33 43 

The restatement of the compensatory theory 43 

CHAPTER III 

Recent studies and reports, 1934-40 , 49 

General character of recent studies 49 

Mechanization and labor displacement J 51 

Productivity, prices, and employment 57 

The measurement of reemplo vment opportunities 68 

Unemployment and increasmg productivity 69 

Production, employment, and productivity in manufacturing 

industry 74; 

Changes in productivity, output, and employment, for all 

industries combined, 1919-36 . 78 

Changes in production, employment, and productivity in 

individual industries, 1919-36 81 

General conclusions aud comments 83 

PART II 

Teohuology and economic balance. ^ 87 

In fcroduction . 87 

CHAPTER I 

The change in labor productivity 89 

The extent of the change 89 

Types of labor-saving techniques 96 

Power and energy devcloDment 99 

Materials . ". , 106 

Substitute materials 106 

Improved materials 110 



VI TABLE OF CONTENTS 

The otj^uue in labor productivity — Continued. 

Types of labor-saving techniques — Continued P"ge 

Processes 111 

Mechanical (including muftiple-f unction machinery) 111 

Konmechanical 112 

Individual, single-function machinery 114 

Elimination of operations 114 

Increasing of speed 115 

Enlargement of capacity • 117 

Management methods 118 

Scientific management and the human factor 118 

Selection and training of workers 118 

Worliing methods 119 

Reduction of hours ■. 120 

Incentive systems of remuneration 121 

Scientific management and material conditions of manufac- 
ture 122 

CHAPTER II 

The effects of labor-saving technology 125 

Technology as a cause of unemployment 125 

The general extent of technological displacement 125 

Displacement in specific industries 130 

The duration of unemployment 133 

Technology and the displacement of skill 136 

Patterns of skill-displacement 137 

Skilled workers as a proportion of all workers 142 

Present trends in occupational requirements 143 

Technology and labor's demand for goods •__ 147 

Technology and the emplo3ed worker 155 

Nervous and mental strain 155 

Physical hardships 160 

Industrial accidents 160 

Occupational diseases 161 

Technology and the older worker 163 

CHAPTER III 

Technology and the compensatory forces 167 

The reduction in hours 167 

The development of new Lndustries 172 

The economic importance of new industries 172 

Limitations of the new industry stimulus 176 

The principle of substitution 176 

Capital-saving characteristics of modern technology 179 

The pattern of present income distribution 181 

Prospective new industries . 184 

General fields of inquiry 185 

Specific new industries 188 

The reduction of prices , 190 

Theoretical importance of price reductions 190 

Industrial prices and labor productivity 191 

CHAPTER IV 

Technology and the concentration of economic power 195 

The concentration of industry 195 

Concentration and operating efl[iciency 197 

The measurement of technological efficiency 202 

Concentration and industrial research 208 

The growth of industrial research 208 

Industrial research and the competitive system 209 

The ownership of industrial research 211 

Concentration and patents 212 

Techniques of control through patents 213 

Corporate size and earning power 217 

Conclusion . 219 



TABLE OF CONTENTS VII 

APPENDIX A Page 

Measures of labor productivity 221 

APPENDIX B 

Production and man-hour indexes 223 

APPENDIX C 

The national research project indexes of labor productivity in manufactur- 
ing industries ■. 224 

APPENDIX D 

The reduction of varieties through standardization 225 

APPENDIX E 

Wages as a percent of value added by manufacture 228 

APPENDIX F 

Types of plastics and their uses, 1940 230 

APPENDIX G 

Effect on labor of specified technological changes in two tire-manufacturing 

plants 232 

APPENDIX H 

Labor productivity and industrial prices 234 

Introduction 234 

The concentrated industries 235 

Iron and steel industry 235 

Productivity 23S 

Price 240 

Productivity and price i 243 

Nonferrous metals industry 244 

Productivity , 244 

' Price 247 

Productivity and price. : 249 

Cement industry 250 

Productivity 250 

Price 253 

Productivity and price 256 

Motor vehicles industry . 256 

Productivity •_ 256 

Price 260 

^Productivity and price 261 

Cigarette industry 262 

Productivity 262 

Price 264 

Productivity and price _• 266 

Electric light and power industry. .1 266 

Productivity 266 

Price . 269 

Productivity and price 271 

The nonconcentrated industries 272 

Cotton goods industry 272 

Productivity 272 

Price , .... 275 

Productivity and price 277 



VIII TABLE OF CONTENTS 

Labor productivity and industrial prices — Continued. 

Tlie nonconcentrated industries — Continued. Page 

Woolen and worsted goods industry 278 

Productivity 278 

Price 281 

Productivity and price 282 

Furniture industry 283 

Productivity 283 

Price 286 

Productivity and price 288 

APPENDIX I 

Employment in industries classified according to the amount of electrical 

energy used per man-hour in plants of varying size, 1937 289 

APPENDIX J 

Methodology of correlating power differential and economic concentration. 292 

APPENDIX K 

Technical progress and economic welfare (an outline of topics for study) 296 



SCHEDULE OF TABLES AND CHARTS 

PART TI 

TABLES 

Pag* 

1 . Indexes of production and productivity, 1 909-39 90 

2. Percent change in production and labor productivity, 1923-29 and 

192^39 90 

3. Indexes of output per man-hour, 1929, 1936, 1939, 40 manufacturing 

industries 95 

4. Effect of variations in capacity utilization upon man-hours- required. _ 97 

5. Consumption of bituminous coal, coal equivalent of fuel oil and 

natural gas, and output of hydroelectric power reduced to coal 
equivalent, 1923-38 101 

6. Loss of employment in coal by causes, 1923-37 103 

7. Percentage change in production and man-hours in 12 industries 127 

8. Technological displacement in manufacturing, steam railroads, and 

bituminous coal mining, 1929-39 . 129 

9. Estimated changes in the productivity of hand and machine labor in 

selected plants in making glass bulbs for 25-watt electric lamps 131 

10. Actual production and volume of technological labor displacement in 

six representative tire plants, 1922 to 1931 132 

IL Duration of unemployment since last nonrehef job, Philadelphia, 

May 1937 134 

12. Length of training period required, 1931 and 1936 144 

13. Percent change in hourly earnings, output per man-hour, and unit 

labor cost in 11 manufacturing industries, 1923 to 1935 148 

14. Percent change in hourly earnings, output per man-hour, and unit 

labor cost in 13 manufacturing industries, 1935 to 1939 150 

15. Indexes of unit labor cost in 20 manufacturine industries, 1919-39 152 

16. Indexes of output per man-hour, average hourly earnings, and unit 

^ labor cost, 1923-39 154 

17. Estimated number of industrial injuries per million workers, by 

industry group and type of disability, 1937 160 

18. Percentage of workers in each age group receiving various grades given 

for quality and quantity of work pe'rf ormed : Brick and stone masons, 
carpenters, and painters employed on Works Progress Administra- 
tion projects, January 1937 165 

19. Average actual hours per week per wage-earner in selected manufac- 

turing industries, 1920-3^ 171 

20. Wage earners in 18 new manufacturing industries since 1879 172 

21. Distribution of families and single individuals, by income level, 

1935-36 182 

22. Income residual after primary expenditures by income groups, 1935-36. 183 

23. Percent of number and value of products manufactured by four largest 

concerns, 1937 .' 196 

24. Mining of nonferrous metals, differences in man-hour requirements 

per ton in plants of different sizes 201 

25. Average standard man-hour ratios for sample brick and tile plants, 

by capacity and process 205 

26. Electric energy, kilowatt hours per man-hour, by size, of plant, 21 

manufacturing industries, 1937 206 

27. Relationship between economic concentration and power differential.. 207 

28. Growth of research personnel, from 1927 to 1938, by industrial group. 208 

29. Corporate size and rate of return, 6-year average of rates of return, 

1931-36, compared with size of corporation by total assets 218 



X SCHEDULE OF TABLES AND CHARTS 

CHARTS 

Page 

I. Production and productivity in the United States, selected indus- 

tries for selected years, 1909-39 91 

II. Percent change in production and labor productivity, 1923-20 and 

1929-39 92 

III. Percentage changes in output per man-hour, 40 manufacturing 

industries . 94 

IV. Effect of variations in capacity utilization upon man-hours required 

per unit of output 98 

V. Consumption of fuels reduced to bituminous coal equivalent, bitu- 

minous coal, coal equivalent of fuel oil and natural gas, and out- 
put of hydroelectric power reduced to coal equivalent 102 

VI. Progress of efficiency in the consumption of fuel by large industrial 

consumers in the United States 104 

VII. Changes in production and man-hours, 12 industries in the United 

States : 128 

VIII. Duration of unemployment since last nonrelief job, Philadelphia, 

May 1937 135 

IX. Change in length of training period required, production workers in 

Minnesota manufacturing, 1931 and 1936 145 

X. Percent change in hourly earriings, output per man-hour and unit 

labor cost, 1923-35 • 149 

XI. Unit labor cost in 20 manufajcturing industries, selected years, 

United States, 1919-39 : ,151 

XII. Indexes of output per man-hour, average hourly earnings, and unit 

labor cost. United States, 1923-39 ". 153 

XIII. Average actual hours per week per wage earner, in selected manu- 

facturing industries. United States, 1920-39 170 

XIV. Indexes of unit labor requirements and prices. ■_ 192 

XV. Relationship between size of plant and electric energy used (kilo- 
watt hours) per man-hour, 21 manufacturing industries. United 
States, 1937 204 

XVI. The relationship between economic concentration and power dif- 
ferential, 30 manufacturing industries, ,_. 206 

XVII. Corporate size and rate of return, 6-year average of rates of return, 

1931-36, compared with size of corporation by total assets 218 

APPENDIX H 

TABLES 

1. Degree of concentration in nine major industries, 1935 234 

2. The iron and steel industry 240 

3. The nonferrous metals industry (primary smelters and refiners) 247 

4. The cement industry L. 253 

5. The motor vehicles industry 259 

6. The cigarette industry ^ 264 

7. The electric light and power industry 268 

8. Yearly production of cotton textiles and hours run per active spindle, 

1930-39 . 274 

9. The cotton goods industry 1 275 

10. The woolen and worsted goods industry 280 

11. The furniture industry 1 284 

APPENDIX H 

CHARTS 

Indexes of unit labor requirements- and prices 239 

I. Iron and steel industry group 239 

II. Nonferrous metals industry (primary smelters and refiners) 246 

III. Cement ^ 252 

IV. Motor vehicles 259 

V. Cigarettes 263 

VI. Electric light and power 269 

VII. Cotton goods 274 

VIII. Woolen and worsted goods 280 

IX. Furniture 285 



LETTER OF TRANSMITTAL 



Hon. Joseph C. O'Mahoney, 

Chairman. Teviporary National Econoinic Committee, 

Washington, D. C. 

My Dear Senator: The monograph Technologj^ in our Economy 
is a tribute to the frontier thinking and sustained effort of Dr. Dewey 
Anderson. He organized the study, selected the authors, and main- 
tained close supervision of its progress from the first-submitt«d out- 
lines until its completion. In no small measure, this study has been 
the outgrowth of his continued study of occupational and employ- 
ment trends during the past 10 years. It took concrete form as a 
result of tlie temporary National Economic Committee's hearings 
on Technology and the Concentration of Economic Power, for which 
Dr. Anderson was the Committee's counsel. 

The Temporary National Economic Committee hearings on tech- 
nology brought together representatives of management and organ- 
ized labor in industries which use great amounts of capital, are nmch 
subject to technological change, and employ large aggregates of 
labor. By so doing, a sample representative of conditions prevail- 
ing in the national economy was secured. The hearings were opened 
with an overview of the impact of technology on modern society. 
It was my privilege, as first witness before the Committee, to con- 
trast the growth of technology during the era of independent inven- 
tion and mechanical power \\\) to the Civil War with the discovery 
of the "art of invention'' by modern industry, by means of which 
great corporations conduct research, control patents, and through 
their use of mass production macliinery and electric i)ower obtain 
piaffes of dominance in their respective fields. Thus, technologj' it- 
self becomes u tool facilitating the concentration of economic power. 
Furthermore, the access to advanced technology is lodged in rela- 
tively few corporate groups, as attested by the fact that in 1938 
thirteen corporations, representing 0.8 percent of all companies con- 
ducting industrial research, employed a third of all industrial re- 
search workers. 

But while technology raises questions of monopoly advantages 
with which public policy must necessarily come to terms, it likewise 
offers hope of achieving that state of abundance in production and 
variety of goods which will free economic man from the tyranny of 
want and privation. There is much evidence in the testimony of 
social scientists, industrialists, and labor leaders who appeared before 
the Committee that the American frontiers have not been reached, 
that living standards of our people are miserably low in comparison 
with, our technological capacity to produce and distribute goods and 
services." Here lies the hope and challenge of modern technological 
society. 



XII LETTER OF TRANSMITTAL 

The T. N. E. C. hearings on technology appear in part 30 of tlie 
series of hearings. They constitute the most up-to-date and com- 
prehensive body of facts ever assembled on this important subject. 
Management was represented by such witnesses as Edsel Ford, picsi- 
dent of the Ford Motor Co.; Charles F. Kettering, vice president 
of General Motors Corporation; Thomas Watson, president of In- 
ternational Business Machines Corporation; Fowler McCormick, vice 
president of the International Hai-vester Co.; William Henry Har- 
rison, vice president of the American Telephone & Telegraph Co., 
and Charles Hook, president of American Rolling Mills. Witnesses 
for labor included William Green, president of the x^merican Fed- 
eration of Labor; Philip Murray, now president of the C. I. O, ; 
George Harrison, president of the Brotherhood of Railway Clerks, 
representing the railway brotherhoods; A. F. Whitney, president of 
the Brotherhood of Railroad Trainmen; R. J. Thomas, president of 
the United Automobile Workers; Emil Rieve, president of the Textile 
Workers Union ; and Thomas Kennedy, secretary of the United Mine 
Workers, Their testimon;^ and the material which they . submitted 
should be read in conjunction with tliis monograph in order to obtain 
a well-rounded view of technology's place in our present-day 
economy. 

Dr. Lewis Lorwin, the author of Part I : Technology in Economic 
Thought, is a former Brookings Institution economist, lately with the 
International Labour Office in Geneva, Switzerland. His eminence in 
the field of economy theory and labor problems is attested by a long 
and substantial series of periodical and book publications. He has 
brought to the present assignment the accumulation of years of experi- 
ence, and his critical faculty has developed a document of unusual 
merit. Unfortunately, limitation of time at Dr. Lorwm's disposal 
has made it necessary to leave untouched certain essential parts of the 
theoretical discussion. The sequence of topics in the material pre- 
sented in Part I is broken in two places. Had time permitted, two 
sections dealing with cyclical unemployment in 1921-28 and the 
rationalization of unemployment in Europe would have appeared 
after the item "Early statements''. Also the topic of teclmocracy 
and its implications would have been inserted following the item 
"The classical debate". Appendix K has been included to uidicate 
the scope of the outline which it was intended to cover. Dr. Lorwin 
wishes to acknowledge ]iis appreciation to Mr. Arthur W. Wubnig, 
who collaborated with him in the preparation of the section dealing 
with measurement of reemployment opportunities. 

John M. Blair has written Part II: Technology and Economic Bal- 
ance, under the supervision of Dr. Anderson. Although the author 
has devoted many years of scientific work to the study of technology 
and its economic effects. tl\e time at his disposal for writing the mono- 
graph was limited to four months between the closin,<; -f Hearings 
on teclmology and the Committee deadline for submitting mono- 
graphs. Ruth Aull, technical assistant on the Temporary National 
> Economic Committee staff, was handed the draft manuscript of 
part II, reorganized it in substantial part, and rewrote much of it, 
contributing materially to makintr it tlie smooth-reading and 
well-knit monograph which is submitted here. Elizabeth W. Breid 
rendered valuable assistance in checking data. 



LETTER OF TRANSMITTAL XIH 

Part II is an attempt to answer certain moot questions concern- 
ing the impact of technology in modern life which affect the bal- 
ance between the forces necessary to continue nn expanding capitalistic 
economy. That it does not conclusively prove the thesis of imbalance 
is due primarily to the paucity of data available for an inductive treat- 
ment of the issue. But substantial material has been assembled here^ 
some of it original with this study, some adapted from other sources, 
which should aid in the solution of many questions concerning the role 
of technology today. 

The Committee is indebted to these several authors and their assist- 
ants for the assembly and presentation of this body of information 
and conclusions. 

Respectfully submitted. 

Theodore J. Kreps, 

Economic Adviser. 

November 23, 1940. 



XVI FOREWORD 

newer approaclips in economic research cast considerable doubt on 
the comforting generality of classical economics that there can be no 
long-term imbalance in a competitive capitalistic economy. The tech- 
nology hearings of the T. N. E. C. developed testimony concerning 
industrial production and the labor supply in the decades 1870-80, 
1890-1900. and 1929-39. In everj^ important respect the depression 
of the 1930's has been far more severe and persistent than any depres- 
sion suffered in the earlier decades. It occurred in a comiti^ whose 
technologj' had reached an advanced stage. This does not necessarily 
indicate a fundamental change in the economy as compared with the 
conditions prevailing in the earlier decades. But there is unmis- 
takable evidence of a change in kind as well as severity of unemploy- 
ment in the last depression. This change is characterized by the 
widespread use of electrical power and mass production methods 
which have shown a capacity to increase industrial activity on the 
upturn of the business cycle without a corresponding ability to absorb 
unemployed labor. 

The problems of technology developed in the Committee's hearings 
on that subject have been treated at length in part 30. They have 
been subjected, to careful, critical scrutiny in this monograph, where 
John Blair has examined increasing productivity and the balancing 
forces of new industries, lowered prices, and reduced working time 
for the labor force. Th6 conclusions are not final, for many of the 
data are fragmentary and in some instances only illustrative of what 
must prove to be profitable areas of future economic research. Yet, 
they all point in the same direction, namely, that the forces which 
can be expected to swing the economic pendulum into balance, allow- 
ing full employment of the manpower and teclmological resources of 
the e^^onomy, are not sufficiently powerful to do so, barring some 
additional support in changed public policy or substantially different 
business practice. 

Tliis monograph is offered not only for its exploratory value, but 
because it contains factual data invaluable in calculating the forces 
of a technological character at work in the economy. Such an as- 
sembly of facts should aid the. Committee in its deliberations of 
numerous problems and should furnish tlumghtful students of eco- 
nomics with material leading.to further and more decisive conclusions. 

Dewey Anderson, 
Executive Secretary^ Temporary National 
EG07ioinic Gormnlttee 



PART I 

THE PROBLEM OF TECHNOLOGICAL UNEMPLOYMENT 
AN HISTORICAL SURVEY 



THE PROBLEM OF TECHNOLOGICAL UNEMPLOYMENT 
INTRODUCTION 

It is now more than a decade and a half since the phrase '"tech- 
nological unemployment" was coined in the United States. In its 
first coinage the phrase was meant to convey the idea that technical 
progress was a factor in decreasing employment and that society was 
being confronted more and more with the problem of '"machines 
versus men/' In the 15 years since the phrase became common cur- 
rency in this popular sense, an almost incessant debate — sometimes 
more and sometimes less heated — hiis been going on as to what the 
phrase really meant; whether what it meant to denote was a fact 
or a "mere figment of the imagination''; if a real fact, how serious 
was it for national welfare and what could or should be done about it. 
What has been said on both sides of the question is registered in an 
extensive literature — in the technical discussions of the American 
statistical and economic associations, in numerous articles in the more 
or less popular magazines, in editorials in the trade press, in volumi- 
nous reports and studies of private research agencies and govern- 
mental bureaus, and in the elaborate depositions and statements made 
in public hearings held by congressional connuittees and by other 
official or semiofficial bodies. 

It cannot be said tliat this rather prolonged debate has settled the 
main issue raised. Today, as a decade ago, opinion in the United 
States is still divided between the opposing views that may be held 
on the subject. On the one hand, it is still claimed by some that 
the factor which is of special significance in making the problem of 
unemployment what it is today is that of technology. Xot only the 
number of the unemployed, but their distribution by occupations, the 
duration of their unemployment, and their chances for being reem- 
ployed are presumably influenced in large measure, if not primarily. 
by the technological changes which have been taking place during tlie 
past two decades and which promise to continue in the discernible 
future. On the other hand, there are those in this country as well 
as abroad who either deny the existence of technological unemploy- 
ment entirely or regard it as of minor importance. To many of these 
the very term is a misnomer whose use merely tends to confuse the 
real issues. 

This division of opinion, which is as old as the problem itself. 
would seem to point to underlying differences in general economic 
views which cannot seemingly be bridged. It is nevertheless also true 
that the opposing views on technological unemployment have roots 
in general economic developments and that the area of disagreement 
ha"s been narrowed down as a result of closer study of the situation 
at different times, and especially in recent years. 

3 



4 COXCENTEATIOX OF ECONOMIC POWER 

It is owing to this fact that a survey of the main discussions of 
the problem and of the studies made can serve as a convenient intro- 
duction to the problem today. Such a survey is presented here with 
a view to drawing from past discussions and studies those points 
which may be of help to current thinking. 

Chapters I and II review^ some of the principal discussions of the 
subject in economic and general literature before 1933. No attempt 
has been made to make this review complete. What is given has been 
selected presumably because it has a lasting interest, and also throws 
light on currents of thought in relation to social-economic back- 
grounds. Chapter III is a more detailed review of recent studies 
and reports. It is hoped that bringing together, within one brief 
volume, the main results of recent work in this field bearing on con- 
cepts, methods, and the relation of technology to general economic 
processes will help to clear the ground for a better understanding of 
the problem in its relation to general economic analysis and to social 
policy. 



CHAPTER I 
NINETEENTH CENTURY BACKGROUNDS 

EARLY STATEMENTS 

Already in the second half of the eighteenth century, during the 
earh'^ stages of the industrial revolution, there was speculation as to 
the effects of the then new machinery upon the condition of the 
worker. By the beginning of the nineteenth century the two trends 
which have persisted in the discussion of the subject found expression 
in the writings on the one side of Jean-Baptiste Say and on the other 
of Lord Lauderdale. Say's Traite d'Economie Politique ^ presented 
the first consistent statement of the optimistic viewpoint. On the other 
hand, Lord Lauderdale in his An Inquiry Into the Nature and Origin 
of Public Wealth,' raised the question whether the unrestricted use 
of machinery was always of benefit to the laboring population. 

Say's position was that machinery benefited society in general and 
the worker in particular. His position was based on his "Law of 
Markets" according to which production creates its own demand.^ 
The introduction of machinery meant a saving in costs and lower 
prices which in turn caused an expansion of the demand for goods 
m the same or new industries and ultimately an increase in employ- 
ment. He admitted that machines displaced workers, but he thought 
that such displacement was a temporary and transitory evil rectified 
by the growth of wealth and by increased employment which fol- 
lowed as a result of lower prices due to greater productiveness.^ 

Lauderdale's "Inquiry" was one of the earliest statements of a 
sceptical attitude on the question. Lord Lauderdale's position was 
influenced by his general view that a distinction must be made by 
what he. called "wealth" and "riches" — that is the individual and the 
social viewpoint in economics. The riches of the individual depend 
in part upon scarcity and exchange value, while public or social 
wealth is determined by abundance. 

Lord LauderdaJe's main argument was that capital is productive 
and adds to national wealth only insofar as it serves to supplement labor 
or to perform such labor as could not otherwise be performed. Hence 
a country could not be benefited by a greater accumulation of capital 
than could be employed to supplement labor, in the production of 

1 First published in 1803 ; 2d ed., Antoine-August Renouard, Paris, 1814. 

* Appeared first in 1804; 2d ed. enl., Archibald Constable, & Co., Edinburgh, 1819. 
'"According to Say's Law, every product, once created, offers instantaneously to the 

full extent of its value a market for other products. This is so because that value Is 
equivalent to the sum of the incomes of the several agents — owners of natural resources, 
capitalists, entrepreneurs, and workers — who cooi^erated in creating the product. No 
general overproduction is therefore possible." (Work Projects Administration, National 
Research Project, Survey of Economic Theory on Technological Change and Employment, 
May 1940, p. 47). 

* "In the early editions of his Treatise, Say * * » suggested * » * interven- 
tion by public authority to relieve the 'momentary and local' evils," including public 
works. (Ibid., p. 48.) 

5 



Q CONCENTRATION OF ECONOMIC POWER 

those things for which tliere already exists a demand. He argued 
against capital formation through "parsimony" which meant a de- 
crease in the demand for consumers' goods, and hence a diminution 
in the demand for labor. Public wealth, as opposed to individual 
riches, could be increased not through saving, but through the "addi- 
tional exertions of industry." Hence the "baneful passion of ac- 
cumulation" had to be restrained for the good of all. Insofar as 
capital increased wealth when its formation was not at the expense of 
consumption, its growth helped to expand markets and to increase 
eAiployment. But the expansion of capital was limited by the un- 
equal distribution of wealth which affected economic development 
unfavorably. 

THE CLASSICAL DEBATE 

The discussions of Say and Lord Lauderdale w^ere resumed on a 
nlore extensive scale and more vigorously b^^ the writers whose works 
marked the first great advance of economics in England in the decade 
following the Napoleonic wars. Despite the wastes, destructions, and 
dislocations of these wars, industrial development had proceeded at 
a remarkable pace. With the end of the wars, this development took 
on an even larger amplitude, and was accompanied by the accelerated 
use and spread of machinery not only in the textile industry but 
also in the iron trades, in coal mining, and in other industries. 

By tliis time men were becoming more clearly aware of the wide 
economic and social effects of the industrial changes that had been 
going on in England (and on a smaller scale also in France, the 
United States and elsewhere) for over 50 years. This period saw 
the publication of a number of histories of the rise of manufacturing 
in England, of the changes in the position of the working class, of 
the rise of the new middle class and- of the various problems in 
economic and political life which the changes had produced. The 
beneficial effects of machinery were no longer much in doubt but 
there persisted a division of opinion as to the immediate effects and 
the temporary dislocations which machinery was having upon the 
life and welfare of the workers. 

There could be no doubt at the time that large sections of the 
artisan and laboring population were torn from their accustomed 
industrial grooves and suffered great economic privati(ms. This was 
especially true of the skilled workers in the textile industry, the 
hand-loom weavers. The injured workers brought their plight to 
public notice by staging the so-called Luddite riots in 1817 — the 
most extensive machine smasliing campaign in English history. But 
other groups of workers were beneficially affected by the growth 
of industry, especially the workers in the trades where skill was 
not completely displaced by machinery and in which the demand 
for the product was expanding. 

The deep disturbances produced by this first period of machine- 
expansion gave rise to a considerable literature which attemj^ted to 
sum up and to appraise the economic and social changes that had 
taken place. In general, the writers of this period stressed not only 
the eiiormous industrial strides which had been made owing to the 
use of macliinery and tlie great increase in wealth but also the rise 
to power of the new middle classes who represented the new spirit 



CONCENTRATION OF ECONOMIC POWER 7 

of industrial enterprise, and the consequent changes in political 
organization and social forms. 

With regard to the effects of these changes on the condition of the 
workers, many writers, more or less socialistic in attitude, painted 
the dark sides of the picture — the long hours and low wages in the 
new factories, the exploitation of women and children, the insani- 
tary and miserable housing conditions in the growing cities. But 
there was carried on an animated debate as to the effects of machin- 
ery on labor which was concerned specifically with employment and 
wages. In this debate the two main trends of thought sketched in 
the preceding section found expression mainly in the writings of 
Sismondi on one side and of the "classical school" of economics 
on the other. 

Sismondi attacked the idea that machinery was an unmixed good 
and that workers displaced by machinery were automatically reem^ 
ployed.^ On the basis of his general analysis that productive capac- 
ity and consumer's' income during any given period are not 
necessarily balanced, he argued that invention and the introduction 
of machinery could be an unmixed benefit only if preceded by an 
increase in demand for goods and for labor which would allow the 
employment elsewhere of the labor displaced by machinery. Capital 
accumulation resulting in the introduction of new machinery leads 
to expanded production. But if the new fixed capital has been 
saved at the expense of the circulating capital, it means a decrease 
in mass consumption, the threat of overproduction and unemploy- 
ment. Sismondi also pointed out that there were difficulties in trans- 
ferring fixed capital and labor skills and in bringing supply and 
demand into equilibrium, and that as a result crises are inevitable 
and the workers must frequently suffer lower wages and unemploy- 
ment until an adjustment is made. Sismondi was in ,favor of 
restricting the adoption of machinery, and of reforms for tlie de- 
velopment of small scale industry; he was also one of the first to 
advocate the limitation of working hours for children and other 
social reforms.^ 

"While Sismondi was to exercise influence in later j^ears, the far 
more influential current of economic thought at the time was repre- 
sented by the followers of Ricardo — the exponents of the "classical 
school of economics." Their reasoning on the question was a further 
development of the position taken by Jean-Baptiste Say, though de- 
veloped on the basis of Ricardian economic reasoning. But on this 
particular question, the followers were soon shocked to find them- 
selves at variance with their master. Ricardo, who in the first edition 
of his Principles of Political Economy and Taxation, published in 
1817, has said little new on the subject, was influenced by some of the 
writings of the period which stressed the difficulties created for the 
workers by the introduction of machinery. As a result, the third 
edition of his "Principles," which appeared in 1821, contained a new 
chapter "On Machinery," which was in contradiction with the unqual- 
ified optimistic views of his disciples. 

_ ^ J. C. L. Simonde de Sismondi. Nouveaux principes d'economie politinue. 2d ed., 
Delaunay, Paris. 1827 (first published in 1819). 

** For more detailed statement, see Work Projects Administration, National Research 
Project, Survey of Economic Theory on Technological Change and Employment, May 1940, 
pp. 48-52. 



g CONCENTRATION OF ECONOMIC POWER 

Ricarclo, in line with the general spirit of his economic thinking, 
was concerned with "the influence of machinery on the interests of 
the different classes of society. * * * Ever since I first turned 
my attention to questions of political economy," Ricardo tells us, "I 
have been of opinion that such an application of machinery to any 
branch of production as should have the effect of saving labour was a 
general good, accompanied only with that portion of inconvenience 
which in most cases attends the removal of capital and labour from 
one employment to another." ^ Ricardo had not changed his opinion, 
as he further tells us, "as far as regards the landlord and the capi- 
talist"; but he had become convinced that "the substitution of ma- 
chinery for human labour is often very; injurious to the interests of the 
class of labourers" ^ and that "the opinion entertained by the labouring 
class that the employment of machinery is frequently detrimental to 
their interests, is not founded on prejudice and error but is comformable 
to the correct principles of political economy." ^ 

Ricardo's treatment of the subject, which is brief and somewhat con- 
fused, centers on a distinction between the "gross produce" and "net 
produce" of a country. Landlords and capitalists derive their rents 
and profits from the "net produce" (or net income) , while the "laboring 
class" depends mainly on the "gross produce." Now, the important 
thing is that the "net produce," owing to the introduction of labor- 
saving machinery, may nicrease at the same time that the gross produce 
diminishes. That may happen, if as a result of the invention of a new 
machine, the capitalist would so organize production as to increase 
the amount of fixed capital and reduce the amount of circulating 
Capital. As the capitalist is only interested in "net produce," he 
may carry on production with the aid of machinery to a point at 
which his own income is maintained or even increased while total 
production is diminished with consequent "distress and poverty" for 
the workers. 

Thi< situation may be offset by the fact that the "net revenue" of the 
capitalist has a greater purchasing power owin^ to the lower prices 
of commodities made possible by the labor-saving machines. As a 
result, the capitalist can satisfy his wants and at the same time save 
more and increase his capital which enables him to employ more labor. 
When this takes place a portion of the people thrown out of work by 
the introduction of machines are reemployed. If the increased pro- 
duction due to machinery is great enough to allow the capitalist to 
accumulate sufficient worlring capital ("food and necessaries"), the 
entire population may again find employment. In such a case the 
condition of the working population would be considerably improved, 
owing to the increased demand for labor and to "the low price of all 
articles of consumption on which their wages will be expended." 

Ricardo warned his readers not to draw "the inference that ma- 
chinery should not be encouraged." In general, he thought, machin- 
ery had undesirable effects when "suddenly discovered and extensively 
used." But he also thought that, as a i-ule, the discoveries of ma- 
chines are gradual, and merely redirect the employment of the capital 
which is saved and accumulated. The demand for labor continues to 



^ David Ricardo, Principles of Political- Economy and Taxation, 3d ed., John Murray, 
London, 1821. pp. 466-467. 
8 Ibid., pp. 46^69. 
• Ibid., p 474. 



CONCENTOATION OF ECONOMIC POWER 9 

increase with an increase of capital, though in a diminishing ratio. 
A great deal depends on wages. If wages rise, largely as a result of a 
rise in the cost of food, more fixed capital will be used. "Machinery 
and labour are in constant competition, and the former can fre- 
quently not be employed until labour rises." ^° 

Ricardo also warned against discouraging machinery in any one 
country, on the ground that it would affect adversely foreign trade and 
cause capital to move to foreign lands, thus diminishing the demand for 
labor at home. 

Ricardo's followers were not convinced by their master's reasoning 
on this question, and in general took the optimistic view on the ques- 
tion. They elaborated the "compensatory principle" according to 
which the workers displaced in one trade or industry are soon reab- 
sorbed in the same or new industries. The most systematic statement 
of the theory at the time was given by J. R. McCuUoch in his Principles 
of Political Economy, first published in 1830. McCulloch dismisses 
Ricardo's main argument by saying that it is entirely hypothetical: 
"In the actual business of the world, machines are never introduced to 
lessen but always to augment gross produce." " He further argues 
that improvements in machinery are similar in their effects to improve- 
ments in the skill and dexterity of the worker. When improvements 
are introduced into an industry, they result in lower costs of pro- 
duction, which tend to reduce tlie price of the commodity produced. As 
prices are reduced, the demand for such commodities increases, and an 
"additional number of hands" is employed to supply the increased 
demand. If the demand for a specific commodity is inelastic, the 
reduction in its price releases income, which is used either to pur- 
chase larger quantities of other things or to increase the savings of 
"persons belonging to the upper and middle classes," which leads to 
an increase of capital. 

The introduction of machinery cannot diminish the demand for 
labor, nor does it reduce the rate of wages. This follows from the 
assumption that employment and wages depend on the amount of cir- 
culating capital which is increased by technical improvements and 
which can be easily shifted from one trade to another. All that ma- 
chinery may do is to force the laborer to change his employment — which 
is not a "material hardship." ^^ 

McCulloch further reasons that increased productiveness due to ma- 
chinery is not the cause of general overproduction or "gluts" in the 
market. That would be the case only if the demand for things were 
limited and if the worker exerted all his productive powers. But our 
wants and desires are insatiable, and "it is absolutely impossible that 
we can ever have what we should reckon a sufficient supply of all sorts 
of commodities." ^^ On the other hand, as his productive powers in- 
crease, the worker is not likely to exert his full powers, but would be 
able to devote a greater portion of his time to purposes of instruction 
and amusement. 

On the basis of the above analysis, McCulloch comes to the conclusion 
that "how much soever it may be at variance with the popular opinion," 
improvements in machinery are always more advantageous to the 

" Ibid.; p. 479. 

"J. R. McCulloch, The Principles of Political Economy, Adam and Charles Blick,. 
Edinburgh, p. 199. 
" Ibid., p. 194. 
« Ibid., p. 191. 



IQ COXCENTKATION OF ECONOMIC TOAYER 

laborer than to the capitalist. In particular cases they may reduce the 
profits of the latter and destroy a portion of his capital, but they can- 
not, in any case, diminish the wa^es of the laborer, while they must 
lower the value of connnodities and improve his condition.^* 

The issue was discussed also by Charles Babbage in his work "On 
the Economy of Manufactures" which appeared in 1832. Babbage 
explains that the introduction of machinery does not diminish the 
quantity of labor demanded by the fact that our power to enjoy new 
things increases our desire for tlieni and this psychology enables us to 
increase our habitual comforts and to expand production in order to 
meet new wants. But he stressed (a) the effects of new machinery in 
redistributing the demand for labor so that "considerable suffering 
among the working classes" results; (b) the increased competition for 
jobs which rationalization induces among workers, and (c) the relative 
difficulties which are created by rapid or gradual improvements ; "the 
suffering which arises from a quick transition is undoubtedly more 
intense; but it is also much less permanent than that which results 
from the slower process." Babbage's general conclusion ^yas incon- 
clusive; "That machines- do not, even at their first introduction, m- 
variahly throw human labor out of employment, must be admitted; 
and it has been maintained, by persons very competent to form an 
opinion on the subject, that thej^ never produce that effect. The solu- 
tion of this question depends on facts, which unfortunately have not 
yet been collected." ^' 

JOHN STUART MILL ON MACHINERY AND THE STATION- 
ARY STATE 

While the debate on the question of the "Effect of Machinery on the 
Condition of the Laboring Classes" was going on, economic develop- 
ments were making the issue of less and less practical importance. 
The rapid expansion of new industries, the opening up of new areas 
of the world with increased natural resources, the large migrations 
from 1825 onward which removed surplus populations from con- 
gested areas and declining trades in Europe, proved that for the time 
being the job-creating possibilities of the new machinery were far 
greater than its labor-displacing effects. 

During the years following the depression of 1837 through the 
"hungry forties," the distress and misery of the working population in 
the different industrial countries were ascrited not so much to the 
labor-displacing influence of the machine as to the general economic 
and social effects of the new industrialism and of private enter- 
prise. The remedies for the evils were, sought in factory legislation, 
limitation of hours of work, political reform, and social reorganiza- 
tion. The workers' movements of the period were in large part radical 
in character as evidenced by Chartism in England, the Socialist move- 
ments of France and Germany, ami by the agrarian and industrial 
reform movements in the United States. 

There is comparatively little discussion in the economic and social 
literature of these years whicli bears on the problem under discussion 

'Mbid., pp. 19G-107. 

" Thf passLKes from Rabbape are quoted by T. E. Gregory in "Gold, Unemployment and 
Capitalism," P. S. King & Son, Ltd., London, 1933, pp. 249-250. 



CONCENTRATION OF ECONOMIC POWER H 

here which is very new. John Stuart Mill restated more fully the. 
classical position though with "some modifications." Mill's argument 
is based on one of his "fundamental the>orems respecting capital," 
namely that it is the expenditure of capital, and not the demand of 
purchasers for the product, that gives employment to labor. In his 
well known phrase — which exasperated Stanley Jevons so much that 
it caused him to seek a new basis for economic theory — "demand for 
commodities is not demand for labor." What Mill meant is that 
consumers through their demand for specific products, determine 
the direction of labor. But the quantity of labor employed is deter- 
mined by the amount of circulating capital which is directly applied 
to the sustenance and renumeration of labor.^^ 

Following Ricardo, Mill stresses the ditferent effects which circu- 
lating and fixed capital have on the "gross product" of a country 
upon which the condition of the workers depends. Whether machin- 
ery and improvements will injure the interests of the workers de- 
pends on whether the increase in fixed capital takes place at the 
expense of circulating capital. According to Mill, this is theoreti- 
cally possible. This might happen in a poor country which tried to 
increase its fixed capital considerably, or it might happen in an 
advanced country in which capital improvements were undertaken on 
a very large scale and rapidly. Mill writes : 

All attempts to make out that the labouring classes as a collective body 
cannot suffer temporarily by the introduction of machinery, or by the sinking 
of capital in permanent improvements, are, I conceive, necessarily fallacious. 
That they would suffer in the particular department of industry to which the 
change applies, is generally admitted, and obvious to common sense ; but it is 
often said, that though employment is withdrawn from labour in one depart- 
ment, an exactly equivalent employment is opened for it in others, because what 
the consumers save in the increased cheapness of one particular article enables 
them to augment their consumption of others, thereby increasing the demand 
for other kinds of labour. This is plausible, but * * * involves a fallacy ; 
demand for commodities being a totally different thing from demand for labour. 
It is true, the consumers have now additional means of buying other things; 
but this will not create the other things, unless there is capital to produce 
them, and the improvement has not set at liberty any capital, even if it has 
not absorbed some from other employilients. The supposed increase of pro- 
diiction and of employment for labour in other departments therefore will not 
take place ; and the increased demand for commodities by some consumers, will 
be balanced by a cessa*^ion of demand on the part of others, namely, the 
labourers who were superseded by the improvement and who will now be 
maintaiDed, if at all, by sharing, either in the way of competition or of 
charity, in what was previously consumed by other people." 

Nevertheless, Mill did not think that this theoretical possibility was 
really significant. In actual practice, he did not believe that "im- 
provements in production are often, if ever, injurious, even tempo- 
rarily to the labouring classes in the aggregate." The reason was 
that improvements are usually introduced gradually, and seldom, if 
ever, made by withdrawing circulating capital from production. 
Improvements are made out of savings, and generally fixed and 
circulating capital increase together and simultaneously.^^ Even if 
improvements decrease aggregate produce and circulating capital 
"for a time," in the long run they tend to increase them. The higher 
return to capital in profits and the gain to consumers from lower 

" .Tohn Stuart Mill. Principles of Political Economy, New York, 1877, vol 1, p. 114. 
"Ibid., p. 134. 
" Ibid., p. 135. 



12 CONCENTRATION OF ECONOMIC POWER 

prices due to improvements lead to an accumulation of capital which 
can be used for more production and employment. 

In advanced industrial countries improvements in production 
(machinery, etc.) increase total production and employment through 
their effects on profits and wages. The general tendency of profits is 
to fall toward a minimum. But this tendency is offset by improve- 
ments insofar as the latter extend the field of new employment 
for capital. This effect is attained, however, only if the improvements 
do not raise proportionately the habits and requirements of the work- 
ers. In that case, owing to the fact that inventions cheapen the 
article consumed by the workers, money wages will in time be 
lowered, profits raised, and more capital will be accumulated. On 
the other hand, if the workers take all the gain from productive 
improvements in a higher standard of living, money wages will not 
fall, the fall of profits will not be retarded, and capital accumula- 
tion will not be stimulated. What happens, will also depend on the 
rate of increase of the labouring population.^^ 

What is of special interest, in view of recent discussions in this 
country, are Mill's observations on , the "Stationary State." ]Mill 
starts out by saying that all political economists had seen that "the 
increase of wealth was not boundless" and that at the end of the 
"progressive state" lies the "stationary state" in which there would 
be no further improvements in the productive arts and no increase 
in population or capital. The political economists "of the last two 
generations," say Mill, thought that "an unpleasing and discouraging 
prospect." But Mill takes the opposite view that such a stationary 
state "would be on the whole, a very considerable improvement on 
our present condition." 

I confess — 

he writer — 

I am not charmed with the ideal of life held out by those who think that the 
normal state of human beings is that of struggling to get on ; that the wran- 
gling, crushing, elbowing, and treading on each other's heels, which form the 
existing type of social life, are the most desirable lot of human kind, or any- 
thing but the disagreeable symptoms of one of the phases of industrial prog- 
ress. The northern and middle states of America are a specimen of this stage 
of civilization in very favorable circumstances ; having, apparently, got rid of 
all social injustices and inequalities that affect persons of Caucasian race and 
of the male sex, while the proportion of population to capital and land is such 
as to ensure abundance to every able-bodied member of the community wlio 
does not forfeit it by misconduct. They have the six points of Chartism, and 
they have no poverty, and all that these advantages seem to have yet done for 
them (notwithstanding some incipient signs of a better tendency) is that the 
life of the whole of one sex is devoted to dollar-hunting, and of the other breed- 
ing dollar-hunters." 

This is not "the social perfection" worth realizing. 

In contrast to this, Mill sketches the outlines of the stationary 
state. "It is only in the backward countries of the world," he says, 
"that increased production is still an im|)ortant object ; in those mosr 
advanced what is economically needed is a better distribution." This 
object could be obtained by the joint action of "individual prudence" 
and of "a system of legislation favoring equality of fortunes" — such 

1* Ibid., vol. II, pp. 320-321. 

"John Stuart Mill, Principles of Political Economy, Fifth London Edition, New York, 
1884, vol. 2, pp. 336-^37. 



CON'CENTRATION OF ECONOMIC POWER j[3 

as a limitation of the sum which any one person may acquire by gift 
or inheritance. "Under this twofold influence, society would ex- 
hibit these leading features ; a well-paid and affluent body of. labor- 
ers ; no enormous fortunes, except what were earned and accumulated 
during a single life-time ; but a much larger body of persons than at 
present, not only exempt from the coarser toils, but with sufficient 
leisure, both physical and mental, from mechanical details, to culti- 
vate freely the graces of life." -^ 

In Mill's vision, the stationary state becomes the main condition in 
which the benefit of machinery can be enjoyed by all. It is worth 
while to quote again his often quoted passage on the subject : 

Hitherto it is questionable — 

the passage runs — 

if all the mechanical inventions yet made have lightened the day's toil of any 
human being. They have enabled a greater population to live the same life of 
drudgery and imprisonment, and an increased number of manufacturers and 
others to make fortunes. They hive increased the comforts of the middle 
classes. But they have not yet begun to effect those great changes in human 
destiny, which it is in their nature and in their futurity to accomplish.^ 

THE MARXIAN ANALYSIS 

In the same year which saw the publication of John Stuart Mill's 
Principles of Political Economy, Karl Marx and Fredrick Engels is- 
sued their Communist Manifesto in which the economic and social 
development of the modern world is sketched as the product of "a 
series of revolutions in the modes of production and exchange" due 
to changes in industrial techniques.^^ In 1859, in his Introduction to 
Capital: A Critique of Political Economy, Marx formulated more 
clearly the doctrine of economic determinism according to which poli- 
tical, social, and cultural forms grow out of the economic formation of 
society which, in turn, is shaped by technological changes.^* These 
general ideas were applied to the analysis of the processes of capital- 
istic production and development in Das Kapital (the first volume of 
which appeared in 1867) in which Marx elaborated his distinctive 
economic doctrine on the effects of machinery on the status of the 
worker, on the tendency of capitalistic production to create a per- 
manent "industrial reserve army," and on the inherent forces which 
underlie the growth of capitalistic economy and tend to transform it 
into a socialized economic system. 

In considering the problem of "machinery and modern industry" in 
their effects on labor, Marx follows his general dialectical method ^^ of 

" Ibid., p. 338. 

"Ibid. p. 340. 

" Karl Marx and Fredrick Engels, Manifesto of the Communist Party, American edition, 
Charles H. Kerr & Co., Chicago, p. 14. 

" One of the clearest formulations of this thesis is given by Marx in a footnote in 
Capital which reads in part as follows : "A critical history of technology would show how 
little any of the inventions of the eighteenth century were the work of one single indi- 
vidual. Hitherto no such book has been puhlislied. Darwin has aroused our interest 
in the history of natural technology, that is to say in the origin of the organs of plants 
and animals, which organs serve as instruments of production for sustaining life. Does 
not the history of the productiv^e organs of man, of organs that are the material basis of 
all social organization, deserve equal attention? * * * Technology discloses man's 
mode of dealing with nature, the process of production by which he sustains his life, and 
thereby also lays the mode of formation of his social relations, and of the mental con- 
ceptions that flow from them." (Karl Marx; "Capital," translated from III German 
section, Charles H. Kerr & Co., Chicago, 1921, vol. I, p. 406.) 

^ "The central idea in dialectic is that truth and progress are realized through conflict 
of opposing elements or tendencies." (A. D. Lindsay, Karl Marx's Capital, 1925, p. 17.) 



1^ CONCENTRATION OF ECONOMIC POWER 

regarding "every historically developed form as a fluid movement.' 
His treatment of the subject is, therefore, historical and institutional. 
He traces the historical evolution of the different forms of the division 
of labor, the development of the manufacturing system and its trans- 
formation into modern industry through the application of machinery. 
He enters into an analysis of the nature of machinery and surveys in 
some detail the inventions and technical changes which transformed 
the methods of production in various industries in England and else- 
where and which imparted to them their "capitalistic" character. 

His concern is thus with the effects of the machine not in general 
but in its historical and institutional association with "capitalism," 
that is, with a system of enterprise based on private ownership and 
initiative motivated by the desire for profit. Capitalism springs into 
existence, he writes, "only when the owner of the means of production 
and subsistence meets in the market with the free laborer selling his 
labor power." 2^ The effects which the machine has on the workers 
under capitalism spring not from the nature of the machine, but from 
the way in which it is used by "capital," that is by the employer 
interested in exchange-value and in "surplus-value." (Profits, Kent, 
Interest.) 

APPROXIMATE EFFECTS OF MACHINEKY 

It is in this historical-institutional spirit and supported by an 
extraordinary wealth of testimony drawn from the reports of factory 
inspectors, parliamentary blue books and other official and nonofficial 
sources, that Marx draws one of the most stirring indictments ever 
penned in social-economic literature of modern machine-industry in 
its "approximate effect" on the workman. Insofar as machinery dis- 
penses with muscular power, he writes, and becomes a means of em- 
ploying laborers of slight nuiscular strength and those whose bodily 
development is incomplete, it enrolls, under the direct sway of capital, 
every member of the workman's family, without distinction of age 
and sex. Compulsory work for the capitalist usurps the place, not 
only of the children's play, but also of free labor at home within 
moderate limits for the support of the family. By throwing every 
member of the family on the labor market, machinery depi-eciates the 
value of the labor power of the head of the family and "revolution- 
izes out and out'' the contract between employer and worker. "Pre- 
viously, the workman sold his own labour power, which he disposed 
of nominally as a free agent. Now he sells wife and child. He has 
become a slave dealer. The demand for children's labour often re- 
sembles in form the inquiries for Negro slaves, such as were formerly 
to be read among the advertisements in American journals.'""-^ 

Marx dwells in detail, citing statistical evidence, on "the enormous 
mortality, during the first few years of their life, of the children of 
the operatives" in England ; on the "moral degradation" an^l "in- 
tellectual desolation'' of women and children, "artificially produced 
by converting immature human beings into mere machines for the 

»'Knil Marx, CiU)Ual, vol. I. CliioaKo, 1921, p. 189 — "Capitalist production only then 
really bfgius. ♦ ♦ • wi),>n eacli individual capital employs simultaneously a com- 
paratively larse number of labourers. • • • A greater number of labourers working, 
together, at the same time, in one place, in order to produce the same sort of commodity 
under the mastership of one capitalist, constitutes, both historically and logically, tlie 
starting point of capitalist production." (Ibid., p. 353.) 

" Ibid., pp. 431-433. 



CONCENTRATION OF ECONOMIC POWER 15 

fabrication of surplus-value"; on the tendency of the machine, "in 
the hands of capital," to first lengthen the working day "beyond all 
bounds set by human nature," or, when working hours are shortened 
by law, to intensify "the exploitation of labor" by "increasing the 
speed of the machinery and by giving the workman more machinery 
to tend." -* He describes the way in which the "wrong use of ma- 
chinery" in the factory, under capitalist control, transforms the 
worker into "a part of a detail-machine," a "mere living appendage" 
of a lifeless mechanism ; -^ on the dangers to life and limb of the 
worker due to unsafe and unsanitary conditions; and, in the final 
count, on the destruction by the machine of the worker's skill which 
results in subordinating the now "insignificant individual factory 
operative to the "master-capitalist in whose brain the machinery and 
his monopoly of it are inseparably united" and who proceeds to in- 
stitute in the factory "a barrack discipline" — "a factory code in which 
capital formulates, like a private legislator, and at his own good 
will, his autocracy over his own work-people, unaccompanied by tliat 
division of responsibility, in other matters so much approved of by 
the bourgeoisie, and unaccompanied by the still more approved repre- 
sentative system." ^° 

THE "strife between WORKMAN AND MACHINE" 

The "approximate effects" of machinery, destructive as they are, 
may be offset by legislation ; for example, the factory acts with their 
educational clauses, to which Marx pays high tribute, though he 
criticizes their shortcomings and stresses the fact that they are the 
result of "centuries of struggle between capitalist and laborer." But 
no such remedial action seems possible, according to the Marxian 
analysis, with regard to the effects of the machine on the employ- 
ment outlook of labor. In what Marx calls "the strife between 
workman and machine," the worker is destined to have the worst of 
it until and unless he can free the machine from its subservience to 
"capital" and make it the instrument of free producers working co- 
operatively in a socially motivated economic system." 

Marx's analysis of the effects of machinery is closely interwoven 
with his general economic doctrines — with his theory of value and the 
formation of "surplus-value" on the one hand, and with his "gen- 
eral law of capitalist accumulation" on the other. His analysis may 
be conveniently summarized around two points: (1) the criticism of 
the "principle of compensation," and (2) the theory of a "redundant 
population" or "industrial reserve army." The ideas centering about 
these two points, supply what may be regarded as the Marxian analy- 
sis of "technological unemployment." 

CRITICISM OF THE "THEORY OF COMPENSATION" 

A whole series of "bour^geois political economists" — including James 
Mill, McCulloch, Torrens, Senior, and John Stuart Mill— writes Marx, 

^'Il>i(l., pp. 434-450. 

* "The miserable routine of endless drudgery and toil in which the .<iame mechanical 
process is gone through over and over again, is like the labour of Sisyphus. The burden 
° «n tI,*^"/' "'^^ *"*^ ^^^^' keeps ever falling back on the worn-out labourer." (Ibid., p. 462.) 

'" Ibid., pp. 4G.S-464. 

""In a communistic society." writes Marx, "there would be a very different scope for 
nie emi)loyment of machinery than tliere can be, in a bourgeois society." (Ibid., p. 429; 



IQ CONCENTRATION OF ECONOMIC POWER 

"insist that all machinery that displaces workmen, simultaneously and 
necessarily sets free an amount of capital adequate to employ the same 
identical workmen." ^^ Marx vents scorn and sarcasm on this "theory of 
compensation" and on such supporters of the theory as McCulloch and 
Jean-Baptiste Say for their "pretentious cretinism" and "insipid- 
ities." ^^ This "optimism of the economists," says Marx, is a travesty 
on the facts. In reality, employers are driven to introduce machinery 
and to improve methods of production by their desire to increase the 
productiveness of labor and to obtain larger profits. The purpose of 
the machine is to save labor and cheapen the prices of commodities as 
a means toward enlarging the "mass of surplus-value." If "it costs 
as mucli labour to produce a machine as is saved by the employment of 
that machine, there is nothing but a transposition of labour; conse- 
quently, the total labour required to produce a commodity is not 
lessened or the productiveness of labour is not increased." " Marx 
indicates that there may be other reasons for introducing machinery 
such as the desire of "capital" to repress strikes, "those periodical 
revolts of the working class against the autocracy of capital." ^^ But 
the essential and primary reason for the extension of the use of ma- 
chinery is that of increasing the productiveness of labor by saving the 
total amount of labor necessary for the production of a given com- 
modity. 

The machine thus becomes a "competitor of the workman himself." 
Modern capitalistic production is based on the sale of labor-power and 
on specialization Avhich reduces the worker's skill to the handling of a 
particular tool. When the handling of this tool becomes the work 
of a machine, the worker's labor-power loses its exchange-value, and 
"the workman becomes unsaleable, like paper thrown out of currency 
by legal enactment." The workers made superfluous by machinery 
either "go to the wall," or swamp the labor-market of other industries, 
thus reducing wages.^^ The results are aggi'avated through further 
disturbances in the rate of wages by the fact that not only the quantity 
of labor needed to produce the same result is diminished but the char- 
acter of the labor required is changed by the substitution of less skilled 
for the more skilled, juveniles for adults, females for males." 

No "compensation" can take place because of the effects Avhich the 
use of machinery has on the division of total capital into its com- 
ponent parts — 'constant" capital (i. e., machinery, buildings, raw 
materials, etc.) and "variable" capital. As more machiner}' is used, 
a larger portion of the capital resources of the employer are ''locked 
up" in constant capital, and the portion of variable capital is dimin- 
ished. But it is variable capital (necessaries of life, such as food, 
clothing, etc.) which gives employment to labor. The application of 
machinery thus turns variable ca})ital into constant capital and de- 
creases the funds out of which labor is emi)loyed. True, some labor 
is required to make the new machines, but, in the first place, the total 
amount of labor is reduced, and, in the second, it is no compensation 
to workers thrown out of work by the nuichine. 

" Ibid., p. 478. 

^ Marx sinfiles out liicardo, whom lie quotes losppctfully on this as on other questions, 
There is. of course, niucli similarity in the Marxian and Uieardian analysis of the problem. 

3' Ibid., p. 420. 

'» Ibid., p. 47.'). 

3« Il.id.. p. 470. 

" Ibid., p. 473. 



CONCENTRATION OF ECONOMIC POWE)^ 17 

What is more likely to happen is this : The workers thrown out of 
work lose their purchasing power; their demand for commodities falls; 
and if this is not compensated for ''from some other quarter," the mar- 
ket price of commodities falls, too. If this condition lasts for some 
time and extends, capital is displaced in tlie industries producing the 
necessaries of life and workers in those industries are also thrown out 
of work. "Instead, therefore, of proving that, when machinery frees 
the workman from his means of subsistence, it simultaneously con- 
verts those means into capital for his further employment, our apolo- 
gists, with their cut-and-dried law of supply and demand, prove, on 
the contrary, that machinery throws worknien on the streets not only 
in that branch of production in which it is introduced, but also in the 
branches in which it is not introduced." ^® 

This effect of machinery is thus not a "compensation*' to the workers, 
but a "frightful scourge." The displaced workers find they cannot and 
ai-e not reemployed by the capital which "formerly employed them and 
which was converted' into machinery." If they do find employment, 
they can do so only through "the intermediary of a new and additional 
capital that is seeking investment." This means for the workers a 
l)rocess of painful readjustment, competition with workers in other 
industries and with new entrants into industry, lower wages, and an 
inferior status. The "economists" claim that the sufferings of the work- 
ers displaced by machinery are only temporary ("a temporary incon- 
venience"), "as transient as are the riches of this world" and that as 
machinery is applied gradually, the extent of its destructive effects is 
diminished. "The first consolation," writes Marx, "neutralizes the sec- 
ond." AVhen machinery seizes an industry by degrees, it produces 
chronic misery among the operatives who compete with it. Where the 
transition is rapid, the effect is acute and felt by great masses. History 
discloses no tragedy more horrible than the gradual extinction of the 
English hand-loom weavers, an extinction that was spread over several 
decades, and finally sealed in 1838. Many of them clied of starvation, 
many with families vegetated for a long time on 21/0 d. a day.^^ In 
general, the facts show that the "original victims" of the machine, the 
workers first displaced by it, "for the most part starve and perish." 

In the face of the obvious facts of industrial development, ^larx can- 
not deny that the extension of machinery has gone hand in hand with an 
increase of employment. He insists, however that this has "nothing in 
common with the so-called theory of compensation." His analysis of 
the phenomenon is as follows: The use oi machinery enables employers 
to produce not the same but a larger quantity of their produce with 
fewer workers; this means that there is need for more raw materials and 
instruments of production. Hence, as the use of machinery extends in 
a given industry, the immediate effect is to increase production in the 
other industries that furnish the first with means of production. The 
incr^asfd demand for raw materials also means an increased demand 
for labor in some of the extractive industries. When machinery is 
applied to the preliminary or intermediate stages in the production of 
a commoility, there is an increased demand for labor in the trades 
using these intennediate products; for example, owing to the abmi- 
dance of clothing materials produced by machineiy there Avas an in- 



3s Ibid., pp. 480-481. 
••Ibid., p. 471. 

277551— 41— No. 22 3 



Ig COXCEXTIIATIOX OF ECONOMIC POWER 

crease in the number of tailors, seamstresses, and needlewomen. Also, 
as the increase of the means of production and of subsistence is accom- 
panied by a rehitive decrease in the number of laborei-s, there is an 
increased demand for labor in auxiliary industries such as transporta- 
tion and in new industries such as gas Avorks, teleg'raphy. photography. 
Fui-ther, as "surplus-value'* is increased by machinery, there is an in- 
crease in the production of luxuries for '"the capitalists and their 
dejiendents" and an expansion of foreign trade which leads to an 
increased demand for labor in the luxury and in the carrying trades. 
Finally, the extraordinary productiveness of labor makes possible the 
"unproductive" employment of a larger and larger part of the work- 
ing class which finds expression in the growth of a "servant class," 
of the professional occupations, of the class living on rents and 
interest, and lastly of "paupers, vagabonds, and criminals." *" 

At every point in this process, there are counteracting influences. 
Thus, as the demand for machinery grows, the machine itself becomes 
a subject of machine production, a "new type of workman springs into 
life along with the machine-making industries. In the new indus- 
tries, the demand is largely "for the crudest form of manual labour." 
Also, "the place occupied by these branches in the general production 
is, even in the most developed countries, far from important." ^^ In 
certain industries, an extraordinary extension of the factory system 
may be accompanied not only by a relative, but by an absolute de- 
crease in the number of operatives employed, as happened in the 
English worsted and silk factories between 1852 and 1862. 

Nevertheless, in spite of displacement by machinery, the number 
of factory operatives in a given industry may become more numerous 
than the manufacturing workmen and handicraftsmen that have been 
displaced, through the building of more mills and the extension of 
old ones. That is, an absolute increase in the number of workers in 
the industry is consistent with a relative decrease. This happens 
during "periods of rest" after technical changes have been introduced 
into an industry and when there is a mere quantitative extension of 
the factories on the existing technical basis. ICach industry affected 
by mechanization thus goes through a feverish period which is de- 
structive of the workers and extremely profitable to the employers, 
but sooner or later enters a "period of rest" which may increase the 
number of workers and benefit their condition until such period of 
rest is broken up by a new invasion of machinery. 

The same alteration of expansion and contraction characterizes 
the entire system of capitalistic production after it reaches a certain 
stage of niaturity. 

So soon as the factory system has gained a certain breadth of footing and a 
definite degree of maturity, and especially, so soon as its technical basis, ma- 
chinery, is itself produced by machinery ; so soon as coal mining and iron mining, 
the metal industries, and the means of transport have been revolutionized, so 
soon, in short, as the general conditions requisite for the production by the mod- 
ern industrial system have been established, this mode of production acquires an 
elasticity, a capacity for sudden extension by leaps and bounds that finds no 
hindrance except in the supply of raw material and in the disposal of the 
produce.*- 



** These concepts reflect the influence of Adam Smith and the classical economists on 
Marx's thinl^inR (ibid., 487). 
" 1bU\.. v. 487. 
"II. id., p. 492. 



CONCENTRATION OF ECONOMIC POWElt IQ 

The modern "factory system" satisfies this need for raw materials 
and outlets by conqnerino; foreion markets and by establishing a new 
international division of labor to its own advantaoe. 

But this process begets another — namely, the moA-ement in cycles — 
which accentuates the effects of machinery on the condition of the 
workers. 

The enoriiKiiis power, inherent in the factory system, of expanding by jumps, 
and the dependence of that system on the markets of the world, necessarily beget 
feverish production, followed by over-filling of the markets, whereupon con- 
traction of the markets brings on crippling of production. The life of modern 
industry becomes a series of periods of moderate activity, prosperity, over- 
production, crisis, and stagnation. The uncertainty and instability to which 
machinery subjects the employment, and consequently the condition of existence 
of the operatives become normal, owing to these periodic changes of the indus- 
trial cycle. Except in the periods of prosperity, there rages between the 
capitalists the most furious combat for the share of each in the markets. This 
share is directly proportional to the cheapness of the product. Besides the 
rivalry Uiat this struggle begets in the application of improved machinery for 
replacing labour-power, and of new methods of production, there also comes 
a time in every industrial cycle when a forcible reduction of wages beneath the 
value of labour-power, is attempted for the purpose of cheapening commodities.*^ 

In biief. the expansion of employment takes place by spurts and 
jerks and at best involves occupational shifts and displacements on 
an extensive scale, in the course of which the workers are "hustled 
from ])illar to post'' and "constant changes take place in the sex, age, 
and skill of the levies." The increase of employment is neither "com- 
pensation" for the displaced workers, nor a means of improving their 
condition. There are ))eriods in the development of capitalism when 
the -growtli (if capital is more rapid than that of labour-power and 
when, as a result, wages may temporarily rise. But these are mere 
interludes. Tiie basic tendency of ca])italism is toAVard a deteriora- 
tion of the condition of the workers which finds expression in. and is 
aggravated by. the growth of a "redundant poiMdation"' or "industrial 
reserve army."* 

f APITAI. ACCI'3It'L.\TI0N AND THE '"iXDUSTRIAL RESERVE ARMv'' 

In his criticism of the "theory of compensation." Marx dealt 
largely with the short-run effects of machinery, upon the condition 
of the workers. His theory of the "industrial reserve ;irmy" sum- 
marize- his views on the long-run effects of machine-industry on 
"the lot of the labouring class." These long-run effects are deter- 
mined l)y the "law of capital accunuilation." or the growth of cap- 
ital, the law wliich underlies the entire evolution of capitalistic 
industry and which determines the inherent trend of capitalism 
toward its own transformation. A statement of the theory of -the 
industrial reserA'e army" thus calls for a sunnnary of the'^Marxian 
doctrines on the law of capital formation and accumulation. 

The starting point of these doctrines is the idea that "the directing 
motive, the end and aim of capitalist production, is to extrai-t the 
greatest possible amount of surplus-value." ** Surplus-value is the 
increment or excess of exchange-value which the employer receives 
in the market over the original value which he advances in the pro- 

*" Uiid,. pp. : 5-49C. 

" ihh\., p. :n:>;i. 



20 CONah.iNTRATIO'N OF ElCONOMIC POWER 

duction of his commodity. It is the "unpaid labour""^ Avhich the 
capitalist-employer appropriates by paying the worker only for a 
part of his Avorking-time and only for a part of exchange-value 
which is embodied in the product of his labour. Surplus-value 
splits up into various parts, such as profits, interest, merchants' 
profit, rent, etc., which the employer pays to various individuals 
who "fulfill other functions in the complex of social production."*® 
But it is the industrialist employer who must first appropriate it 
in its original form, and whose function it is to increase its amount 
and to provide for its growth. 

In performing this function, the industrial employer is limited 
by certain characteristics of the capital which he has at his com- 
mand. As indicated above, Marx divides capital into constant and 
variable.*' Constant capital, i. e., the instruments of production, 
raw materials, etc., is merely reproduced in the process of produc- 
tion and does not itself create any new value. It is only variable 
capital which is devoted to the maintenance of the worker,*^ that 
has the capacity not only to reproduce itself but to create new value 
or surplus value. 

To obtain surplus-value and to increase its amount, employers 
must, therefore, increase the amount of variable capital employed. 
They can do so either by employing more workers or by increasing 
"the rate of absolute surplus-value," *^ chiefly by prolongation of 
the working day. "Free competition" drives every individual cap- 
italist to do his utmost in this respect, and capitalism, as a result, 
surpasses all earlier systems of production in energy, disregard of 
bounds, recklessness, and efficiency, as "a producer of the activity 
of others, as a pumper-out of surplus-labour and exploiter of labour 
power." ^° 

There are limits, however, to these possibilities in the physical en- 
durance of the worker and also in the reactions of society which sooner 
or later begins to regulate hotirfs of work. But the employer realizes 
that he can increase sur^Slus-value (or, in Marxian terms, create "rela- 
tive surplus- value"") by decreasing the amount of "necessary labour," 
that is by increasing the intensity or "the productiveness of labor" so 
as to make up for the shortening of the duration of working time. 
This is done by means of machinery and improved methods of produc- 
tion, and the employer is thus impelled, by his hunger for profits and 

■•*"An surplus value, whatever particular form (profit, interest, or rent) it may subse- 
quently crystallize into, is in substance the materialization of unpaid labour." (Ibid., 
p. 585.) This follows from Marx's labour theory of value which need not be restated 
here. 

" Ibid., pp. 618-619. 

*' "That piirt of capital then, which is represented by the means of production, by the 
raw material and the instruments of labour, does not, in the process of production, undergo 
any quantitative alteration of value. I therefore call it the constant part of capital, or, 
more shortly, constant capital. On the other hand, that part of capital, represented by 
labour-power, does, in the process of production, undergo an alteration of value. It both 
reproduces the equivalent of its own value, and also produces an excess, a surplus value, 
which may itself vary » • * j shall, therefore, call it the variable part of capital, 
or shortly, variable capital." (Ibid., pp. 232-2;?3.) 

<* "Variable capital is therefore only a particular historical form of appearance of the 
fund for providing the necessaries of life, or the labour-fund which the labourer requires 
for the maintenance of himself and family and which, whatever be the system of social 
production, he must himself produce and reproduce." (Il)id., p. 622.) 

<* The rate of absolute surplus-value is the ratio of surplus-value to the total variable 
capital, or the ratio of the surplus unpaid labour to the "necessary labour" which Is 
embodied in a commodity and for which alone the worker is paid. 

wjbid., pp. 338-339. 



CON'CEXTRATION OF ECONOMIC POWER 21 

by the inexorable laws of "free competition" to convert an ever larger 
part of his capital into "constant capital." ^'^ 

In his role as organizer and master of the productive process, the 
capitalist convei-ts as large a part of his surplus-value as possible into 
capital. The conversion of surplus-value into capital is capital ac- 
cumulation. It is up to the capitalist, "the owner of the surplus- 
value" which is produced by the worker from year to year, to decide 
how much of it he will consume or spend "as revenue" and how much 
he will save or use as capital for further production. In general, the 
capitalist has "a passion for wealth as wealth" and is besides under 
compulsion to accumulate on an ever increasing scale. "The develojD- 
ment of capitalist production makes it constantly necessary to keep 
increasing the amount of the capital laid out in a given industrial 
undertaking, and competition makes the imminent laws of capitalist 
production to be felt by each individual capitalist, as external coercive 
laws. It compels him to keep constantly extending his capital, in order 
to preserve it, but extend he cannot, except by means of progressive 
accumulation." ^- 

In the early stages of industrial dcA-elopment the capitalist tends to 
be ascetic and the "capitalist of the classical type brands individual 
consumption as a sin against his function" of accumulation. But as 
the progress of capitalist production "creates a world of delights" and 
opens up, in speculation and in the credit system, "a thousand sources 
of sudden enrichment," the "modernized" and more educated capitalist 
develops habits of more luxurious living, which he tries to reconcile 
with the virtues of accumulation.^^ What happens is that the capi- 
talists become "good livers and men of the world," ^* while at the same 
time the production of surplus-value and the accumulaf'lon of capital 
proceed on an ever increasing scale. This is made possible by the 
growing "productivity of social labour." ^^ The increasing produc- 
tivity of labor increases the mass of surplus-products, and thus enables 
the capitalist to maintain, or even to increase relatively, his accumula- 
tion fund without decreasing his consumption fund. As real wages 
never rise proportionately to the productive power of labour, an in- 
crease in labor productivity enables the employer to set in motion 
with the same variable capital a larger quantity of labour-power. The 

01 The employer is in sort of a dilemma: On the one hand, lie is eager to enlarge his 
variable capital and to prolong the working day ; on the other hand, he is driven to 
mechanize and revolutionize industry and to lock up more and more of his capital in 
machinery or "constant capital" and to accumulate surplus-value by a more intensive use 
of a relatively decreasing portion of variable capital. Marx was aware of the dilemma 
involved in his reasoning and tried to meet it in the third volume of "Capital." 

"Ibid., p. 619. 

" "When a certain stage of development has been reached, a conventional degree of 
prodigality, which is also an exhibition of wealth, and ronsequently a source of credit, 
becomes a business necessity for the "unfortunate' capitalist. Luxury enters into capital's 
expenses of representation." (Ibid., p. 651.) 

"Marx pokes fun at Nassau Senior for "discovering" the doctrine of "abstinence" and 
at "that queer saint, that knight of the woeful countenance, the capitalist abstainer." 
(Ibid., pp. 654-656.) . 

" "By increase in productiveness of labour, we mean, generally, an alteration In the 
labour-process of such a kind as to shorten the labour-time socially necessary for the 
production of a commodity and to endow a given quantity of labour with the power of 
producing a greater quantity of use-value." Op. cit., p. 345. As a result of greater 
productivity, the portion of the working day which goes to pay the worker for the 
reproduction of his labour-power is shortened, and a larger portion of the working-time 
accrues to the employer as surplus-value. The productiveness of labour is promoted in 
the industries which "produce the necessaries of life as well as in the industries making 
the instruments of production and the raw materials, thus making possible a reduction 
in the cost of living and, hence, in wages. Marx held that "an increase in the produc- 
tiveness of labour in those branches of industry which supply neither the necessaries of 
life, nor the means of production for such necessaries, leaves the value of labour-power 
undisturbed." (Ibid., p. 346.) 



22 CONCENTRATION OF ECONOMIC POWER 

growtli of labor productivity and improvements in methods enhance 
the vahie of capital Avhich is renewed in more elTective form as it wears 
out."''' Science and technology ^ive capital a power of expansion which 
is independent of the magnitude of functioning capital, and the pro- 
duction of surplus-value increases more rapidly than the value of added 
capital. ''The more, therefore, capital increases by means of successive 
accumulations, the more does the sum of the value increase that is 
divided into consumption-fund and accumulation-fund. The capi- 
talist can. therefore, live a more jolly life and at the same time show 
more 'abstinence.' " ^'' 

The accumulation of capital may, under certain conditions, not only 
bring benefits to the employer, but also improve the condition of the 
worker. This happens during those "intervals of rest," when the tech- 
nical basis of industry remains unchanged and when as a result the 
growth of capital continues while tlie "organic composition" ^* of 
capital remains constant. Under such conditions a definite mass of 
means of production needs the same mass of labour-power to set it in 
motion. When under such conditions the scale of capital accumula- 
tion is extended (owing to special stimuli, such as the opening of new 
markets or of new opportunities for investment due to new social 
wants), the variable part of capital is increased, and "the requirements 
of accunnilating capital may exceed the increase of labour-power or 
the number of labourers." In such case the demand for labor exceeds 
the suj)ply, and wages may rise. This happened in England in the 
fifteenth century and again in the first half of the eighteenth century. 
Under such conditions an increase in the numbers of the working popu- 
lation takes place. 

Marx is eager to point out in this connection that "the more or less 
favorable circumstances" which thus arise for "the wage-working 
class" do not in any way "alter the fundamental character of capitalist 
production." ^^ The dependence of the workers on capital takes on 
an "endurable" form, and the workers improve their consumption and 
living conditions and "can lay by small reserve funds of money." 
But "a rise in the price of labour, as a consequence of accumulation of 
capital, only means, in fact, that the length and weight of the golden 
chain the wage- worker has already forged for himself, allow of a 
relaxation of the tension of it.""'' " Also these "intervals of rest" in 
which "accumulation works as a simple extension of production, on 
a given technical basis" become more and more shortened, owing to 
the operation of the "industrial cycle" and of the basic tendencies of 
capitalist development. 

What is more important, however, is that the growth of capital is 
accompanied by a change in its "organic composition" which consists 
in a relative increase of constant capital in proportion to and at the 
expense of the variable capital. This "law of capital accumulation" 

""Marx indicates that tlm (Ipvelopmeiit of produrtivo power is accompaniod h)v a partial 
deprcnation of fuiict ioiiin^: capital. Hut, he says that in so far as tliis depreciation 
niatics itself acutely lelt in competition, the hunlen falls on the worker, in ''the increased 
exploitation of wliom the capitalist looks for his indemnification." (Ibid., p. 604.) 

" n)id., ]). (i07. 

='The -'organic composition" of capital, in Mar.xian terminology, is determined bv the 
proportion of constant to varial)le capital (or the value of the means of production to 
tlie sum total of wages) wliich, in turn, reflects changes in the teclmical composition of 
capital, that is. in the relation of "the ma.ss of the means of production employed" to 
"the mass of labour necessary for their employment." (Ibid., p. 671.) 

<>» Ibid., pp. 072-078. V . t' 

«' Ibid., p. 077. 



CONCENTRATION OF ECONOMIC POWER 23 

is bound up ~with the fact that capital growth is stimulated, above 
everything, by an increase in the productiveness of labor, and the 
increase in the productivity of labour depends upon the application 
of an ever larger mass of machinery and other means of production. 
The. progress of accumulation does not exclude the possibility of a 
rise in the absolute magnitude of variable capital, but it inevitably 
lessens its relative magnitude. 

The change in the "organic composition" of capital is accentuated 
and accelerated by the processes of concentration and centralization 
characteristic of capitalistic development. The "smaller capitals'' are 
beaten in the game of competition not only because large scale pro- 
duction is more efficient, but also because the centralization of capital 
finds a "powerful weapon'' in the credit system. "While "the relative 
expansion and energy of the centralization movement is determined 
to a certain degree by the superiority of the economic mechanism, yet 
the progress of centralization is by no means dependent upon the 
positive growth of the volume of social capital. * * * Centrali- 
zation may take place by a mere change in the distribution of already 
existing capitals, a simple change in the quantitative arrangement of 
(he components of social capital. Capital may in that case accumulate 
in one hand in large masses by withdrawing it from many individual 
hands.'' "^^ By doing so, centralization enables the industrialists to 
expand the scale of their operations, to develop "socially combined 
and scientifically managed processes of production." and thus to hasten 
and extend the revolutions in the technical composition of capital.*'- 

The progressive qualitative change in the composition of capital 
has dire eifects on "the lot of the labouring class." Since the demand 
for labour is determined not by the amount of capital as a whole, but 
by the amount of variable capital only, that demand falls progres- 
sively with the growth of total capital. It is true that as total capital 
increases, its variable portion also increases, but it increases in a con- 
stantly diminishing proportion. In fact, the variable capital dim- 
inishes more rapidly than the total capital increases (owing to the 
effects of increasing centralization and other factors). This means 
that as total capital grows, and in the direct ratio to the energy and 
extent of its accumulation, there is produced a "relatively redundant 
population of labourers, i. e., a population of greater extent than 
suffices for the average needs of the self-expansion of capital, and 
therefore a surplus population." ^-^ This is the law of population 
jjeculiar to the capitalist mode of production.^* 

The "surplus labouring population'' produced by "the development 
of wealth on a capitalist basis" becomes in its turn a necessai-y con- 
dition of the existence of the "capitalist mode of production." It 
forms an "industrial reserve army'' which capital uses for its self- 
expansion. As with the growth of capital and with the development 
of the credit system industrialists become eager to expand at everv 
possible opportunity, they must have "the possibility of throwing 

<" Ibid., pp. 687-688. 

»=Ibid., p. 689. 

« Ibid., pp. 690-G91. 

•"Marx attacks the Malthusian "principle of population" and argues that the niovomeiits 
of population reflect the changes in the accumulation of capital. See Op. cit.. pp. 675-681. 
692-693. In his "Principles of Political Economy", Malthus. according to Maix. (iiuUly 
discovered, with the help of Sismondi. "the beautiful trinity of capitalistic production : over- 
production, overpopulation, over-consumption — three very delicate monsters, indeed." (Jbl<l., 
p. 696. footnote.) 



24 CO^'CENTRATION OF ECOXOillC POWER 

great masses of men suddenly on the decisive points without injury 
to the scale of production in other spheres. Over-population sup- 
plies these masses." ®° Also — 

the course characteristic of modern industry, viz. a decennial cycle (inter- 
rupted by smaller oscillations), of periods of average activity, production at 
high pressure, crisis and stagnation, depends on the constant formation, the 
greater or less absorption, and the reformation of the industrial reserve army 
of surplus population. In their turn, the varying phases of the industrial cycle 
recruit the surplus population and become one of the most energetic agents of 
its reproduction. * * * The whole form of the movement of modern industry 
depends, therefore, upon the constant transformation of a part of the labouring 
population into unemployed or half-employed hands.** 

The production of a relative surplus-population goes on even more 
rapidly than the technical revolution in production and more rap- 
idly than the diminution in the variable capital as compared with 
constant capital. This is due to the "absolute interest of every capi- 
talist to press a given quantity of labour out of a smaller, rather 
than a greater number of labourers, if the cost is about the same," 
because of savings in overhead costs. The employer achieves his pur- 
pose, as already indicated above, by prolonging the working day, 
by intensifying labor and by replacing skilled with unskilled work- 
ers. Thus, one part of the working class is overworked, while the 
other part is condemned to enforced idleness. 

The relative surplus-population, or industrial reserve army, is 
found in different countries in three forms^the floating, the latent, 
and the stagnant. The "floating" surplus of labor is found in the 
centers of modern industry and is composed in large part of young 
persons who enter industry as boys and who are not given a perma- 
nent place in it. The latent surplus population is found largely in 
agriculture. The stagi.int surplus population is formed by groups 
of labor living below "the average normal level" of the working 
class who are recruited from decaying branches of industry and who, 
as a rule, have large families ancl thus supply "an inexhaustible res- 
ervoir of disposible labor-power." Marx illustrates this generaliza- 
tion by reference to miserable living conditions found among some 
strata of the working population of England and Ireland. 

The three forms or elements of the industrial reserve army suffer 
in varying degrees from the effects of irregular employment — low 
wages, malnutrition, high mortality, etc. They also act as a drag 
on the working population as a whole through their effects on wage 
movements which are regulated by the expansion and contraction of 
the industrial reserve army. During periods of stagnation and 
average prosperity, the industrial reserve army weighs down the 
active labor-army; during periods of over-production and boom, it 
''holds its pretensions in check." The general effect in the long-run 
is that "in proportion as capital accumulates, the lot of the labourer, 
be his paymeiit high or low, must grow wor.se." The law of capital 
growth establishes "an accumulation of misery, corresponding with 
accumulation of capital. Accunudation of wealth at one pole is at 
the same time accunudation of misery, agony of toil, slavery, igno- 
rance, brutality, mental degradation at the opposite pole." ^^ 

«Ibid., p. 694. 
•«Iljid., pp. 094-695. 
6^ Ibid., pp. 708-709. 



CONCENTRATION OF ECONOMIC POWER 25 

As is ^A'ell known, Marx saw no solution for the problems sketched 
above except through the transformation of capitalist into socialized 
production. This ultimate solution was inherent in "the historical 
tendency of accumulation" itself and was to be realized through the 
dialectics of capitalist development, that is, the centralization of capi- 
tal, the conscious technical application of science, socialized labor, and 
the entanglements of all peoples in the net of the world-market. The 
climaxing aspects of this process may be described in Marx's oft quoted 
passage : 

Along with the constantly diminishing number of the magnates of capital who 
usurp and monopolize all advantages of this process of transformation, grows 
the mass of misery, oppression, slavery, degradation, exploitation ; but with this 
too grows the revolt of the working-class, a class always increasing in numbers, 
and disciplined, united, organized by the very mechanism of the process of 
capitalist production itself. The monopoly of capital becomes a fetter upon the 
mode of production, which has sprung up and flourished along with, and under it. 
Centralization of rhe means of production and socialization of labour at last 
reach a point where they become incompatible with their capitalist integument. 
This Integument is burst asunder. The knell of capitalist private property 
sounds. The expropriators are expropriated.** 

PIONEER INDUCTIVE STUDIES, 1886-99 

The last quarter of the nineteenth century was marked by compara- 
tive calm in the political and social life of Europe based upon steady 
economic growth, despite the recurrence of hard times. Economic 
theory turned to reexamination of its basic premises and began the 
elaboration of those doctrines of marginal utility and of general and 
partial equilibrium which were permeated with a spirit of optimism 
and which tended to see the economic world as a sphere of harmoni- 
ously adjusting supply and demand prices. Social reformers, doc- 
trinally stemming from Sismondi, took their first steps toward a prac- 
tical "solution" of the problem of unemployment which led to the first 
international conferences on the subject and to the organization of 
the International Association for the Struggle Against Unemploy- 
ment. The Socialists began filing away the sharper edges of the 
Marxian analysis, and though they continued to regard "crises" and 
an "industrial reserve army" as inherent in and ultimately destined to 
destroy "capitalism," they devoted their energies to practical measures 
for the gradual alleviation 'of unemployment and of other hardships 
caused by the introduction of machinery. 

The main theoretical contributions during this period to the discus- 
sion of the problem considered here were largely a further develop- 
ment of the views of Jean-Baptiste Say and of the classical economists. 
J. E. Cairnes, for instance, in his work on "Some Leading Principles 
of Political Economy," published in 1874, restated in his own way 
Say's law, as follows : 

Purchasing power, in the last resort, owes its existence to the production of a 
commodity, and, the conditions of industry being given, can only be increased by 
increasing the quantity of commodities offered for sale ; that is to say, demand 
can only be increased by increasing supply.** 

In restating the wages fund doctrine, Cairnes argues that the "gen- 
eral rate" of wages depends upon three factors — the total capital of 

« Ibid., 836-837. 

«9J. B. Cairnes, Some Leading Principles of Political Economy Newly Expounded, Harper 
& Bros., New York, 1878, p. 31. 



25 COXCKXTRATION OF ECONOMIC POWER 

the country, the nature of the national industries, and the supply of 
labor. The nature of the industries means largely the relative amount 
of fixed and circulating capital used in them, and Cairnes reformulates 
the general theory of classical economics as follows : 

The modifications which occur in the distribution of capital among its several 
departments as nations adAance are by no means fortuitous, but follow on the 
whole a well-defined course, and move towards a determined goal. In effect, what 
we find is a constant growth of the national capital, accompanied with a nearly 
equally constant decline in the proportion of this capital which goes to support 
productive labor. This is the inevitable consequence of the progress of the indus- 
trial arts, the effect of which is to cause a steady substitution of the agencies of 
inanimate nature for the labor of man.'" 

Cairnes does not deduce from this tendency any pessimistic conclu- 
sions. He recognizes that at certain times the progress of wealth and 
industry may be accompanied by a contraction of the wages fund. This 
happens when a country undertakes great changes in its industrial 
structure and does it by "converting circulating into fixed capital. It 
thus may happen, and, in fact, it has happened, that extensive changes 
in the character of the industry of a country, even though they be all 
in the direction of scientific progress, improved processes, and ulti- 
mately and even immediately augmented wealth, may nevertheless 
effect a reduction in the means for supporting productive labor, and 
may for a time act disastrously on its interests." '^ 

But, fundamentally, the results are positive. "It is perhaps super- 
fluous to add," Cairnes writes — 

that it is not to be inferred from the circumstance stated that the progress of those 
arts is unfavorable to the interests of labor. Even on the lowest and most mate- 
rialistic view of the interests of labor the reverse is the fact, for what industrial 
progress under the influence of the advancing arts and sciences effects is a dimi- 
nution not in the absolute amount of the wages fund but only in the proportion 
which it bears to tlie total capital of a country — a diminution which is perfectly 
compatible with a steadily progressive increase of the fund — 

and, therefore, of tlie demand for labor. "- 

Nevertheless the law of capital growth has serious tendencies. It 
tends to result in a rehitive increase of the classes not living by hired 
labor and in increased inequality in the distribution of wealth. Cairnes 
thinks, however, that these tendencies may be offset by the action of the 
laboring ])opulation to save and to share in profits. The solution of the 
labor problem, according to Cairnes, lay in "cooperative industry" on 
a basis of private ownership."-^ 

Even more ])ositive are the views of J. S. Nicholson, whose work on 
The Effects of Machinery on Wages appeared in 1878. Nicholson 
singles out and develops the idea of the earlier economists that even the 
temporary adverse effects of machinery on labor tend to lose in force 
as the speed with wliich new machines are introduced and the scale on 
which the j)rocess takes place are lessened. Nicholson claimed that 
sucli was indeed the tendency, in accordance with what he described as 
the "Law of Continuity." According to this law, "a radical change 
made in the methods of invention will be gradually and continuously 
adopted"; furthermore, "radical changes tend to give place to advances 
by small increases of invention." The slowness of introduction of new 



•^"Ibid., p. 176. 

"Ibid., p. 180. [Italics supplied.] 

"Ibid., p. 176. 

■3Ibifl., pp. :.'74-294. 



CONCENTRATION OF ECONOMIC TOWER 27 

machines may be due to the inertia of capitalists, to the possibility of 
improving existing machinery, or to the difficulties of adapting ma- 
chines to the needs of different trades and industries or to the patent 
system, but in any case it results in the fact that the rapidity of intro- 
duction is less than the mobility of labor. As a result, labor can more 
easily adjust to the change, and the destructive effects of the machine 
are greatly reduced.^* 

Nicholson also pointed out that the operation of machinery re- 
quired general intelligence and skill, and that the machine-making 
and repairing industries created new types of skill. Wliile he saw 
a tendency toward industrial concentration as a result of machinery, 
he held that it was offset by the establishment of new small in- 
dustries. Fuithermore, the adverse effects of machinery on labor 
could be remedied by factory legislation, trade unionism, and the 
"growth of higher moral principles." "^ 

It was in the United States that a new approach to the problem 
was made during this period, and the first attempt was made to 
study the effects of machinery on output and employment inductively. 
The interest in tlie problem was enhanced in America as a result of 
the revolutionary changes which were taking place in industry and 
of the rapid economic expansion which was accompanied by serious 
dislocations leading to political and social turmoil. The preoccupa- 
tions with the problem during these years were reflected particularly 
in the study by David A. Wells on Kecent Economic Changes pub- 
lished in 1889 and especially in the investigations on Industrial De- 
pressions and on Hand and Machine Labor carried out between 1885 
and 1898 by Carroll D. Wright, first United States Commissioner of 
Labor. 

Wells' book is an attempt to survey the changes in economic life 
which were causing so much social upheaval, to examine the causes 
of these changes, and to appraise the widely held pessimistic views 
on the future to which they gave rise. Wells proceeds inductively 
with regard to a wide range of topics. He lists the important dis- 
coveries and inventions from 1838 to 1889 which have affected in- 
dustrial processes and organization. He examines a large number of 
industries, trades and occupations in Avhich machinery had become 
dominant, such as cotton textiles, coal mining, making of steel, etc., 
and quotes figures on the reduction in man-hour requirements per 
unit of product, or increased output per worker, total production, 
etc., to illustrate the great strides forward in productivity and pro- 
ducing capacity. He then considers whether this great technical 
progress can be held to be the cause of the depressions from which 
the L^nited States and the whole world were suffering and whether 
it was causing such a displacement of labor as to jeopardize the 
social and economic future of the workers. 

Wells finds that the recent industrial and economic changes are 
the same in kind as had been going on throughout history. But 
there was a difference in degree. Also, the cumulative effect of the 
increasing power to produce was such "that the world * * * for 
the first time [has] become saturated * * * with the results of 

•^Quoted by Georgf E. Barnett, "Chapters on Machinery and Labor : III. Machinery and the 
Displacement of Skill," Quarterly Journal of Economics, vol. XL, Nov. 19:i5. pp. 113-117. 

''• Works Projects Administration, National Research Project, Survey of Economic Tlieory 
on Technological Change and Employment, p. 81. 



28 CONCENTKATION OF ECONOMIC POWER 

these modem improvements"."'' This has created a serious hick of. 
bahuice between j^rodiiction and consumption. The world's cai)acity 
to i)roduce liad developed in the past 15 years in a far oreater ratio 
than the increase in population or its immediate consuming-capacity. 
One of the difficulties in the situation, accoi-din^ to Wells, \v;'s the 
sloAv response of tlie masses of the people to the increased and cheap- 
ened production of commodities. Many things, he says, have been 
showered upon the masses of the people which they don't know how 
to nse and which they don't feel they need and foi- which they 
cannot pay the market price. One of the paradoxical results of tech- 
nical progress is thus "more limited possibilities of new sources of 
demand throughout the world," and a ''larger amount of capital 
seeking employment"' — a condition which is one of the specific causes 
of the economic dejn-ession. 

Wells considers in some detail the causes of "the almost universal 
discontent of labor,'' which had characterized the recent transitions 
in the world's methods of ])roduction and distribution. The most 
important causes he lists as follows: 

1. The displacement or supplanting of labor through more 

economical and effective methods of production and dis- 
tribution. 

2. Changes in the character or nature of employments conse- 

quent upon the introduction of new methods — machinery 
or processes — which in turn liave tended to lower the grade 
of labor, and impair the independence and restrict the 
mental development of the laborer. 

3. The increase in intelligence, or general information, on the 

part of the masses, in all civilized countries.'^ 

But these results of machinery are, in Wells' opinion, part of 
the process of progress, and they are. in fact, more destructive to 
capital than to labor. "It seems to be in the nature of a natural 
law," he writes, "that no advanced stage of civilizati(m can be 
attained, except at the expense of destroying in a greater or les^ degree 
the value of the instrumentalities by which all previous attainments 
have been effected. Society proifers its highest honors and rewards 
to its inventors and discoverers; but, as a matter of fact, what each 
inventor or discoverer is miconsciously trying to do is to destroy 
property, and his measure of success and reward is always pro- 
portioned to the degree to which he effects such destruction."'* And 
further: "In short, all material progi'ess is effected by a displace- 
ment of capital equally with that of labor; and nothing marks 
the rate of such progress more clearly than the rapidity with 
which such displacements occur. There is, however, this difference 
between the two factors involved: Labor displaced, as a condi- 
tion of progress, will be eventually absorbed in other occupations; 
capital displaced, when new machinery is substituted for old, is 
practically destroyed." ^® 

Considering all the machinery used in all countries, Wells thinks, 
the number of pereons displaced in recent years by new and more 

'9 David A. Wells, RecPnt Economic Changes, D. Appleton & Co., New York, 1808, pp. 62-63. 
" Ibid., pp. 364-3G5. 
'Mbid.. pp. .'569-370. 
'"Ibid., p. 373. 



rONrENTKATIOX OF ErONOMIC rOWER 29 

effective metliods of production does not appear to be so great, as 
is generally supposed. Truly, there is a feeling among the masses 
that the op])ortunities for employment are less favorable than for- 
merly, but this is due to the large immigration, the disapi^earance 
of public lands, as well as to machinery. The fact is that in the 
United States, there is little evidence thus far that labor has been 
greatly disturbed or distressed as a result of the spread of machinery. 
Neitlier is there evidence that — 

viewed from the standpoint of 20 or 25 years ago, or before wlifit may be 
termed the advent of the "machinery epoch" * * * the aggregate of 
poverty in the world is increasing.*"' 

During tlie last quarter of a century, however, the problem of poverty 
has been complicated by a new factor, namely, the displacement of common 
labor by machinery, which has been greater than ever before in one generation 
or in one country. To what extent the numbers of the helpless poor have 
been increased from this cause is not definitely known ; but the popular idea 
is doubtless a greatly exaggerated one. In fact, considering the number and 
extent of the agencies that have been operative, it is a n>atter of wonderment 
that these influences in this direction have not been greater. In the United 
States little or no evidence has yet been in*esented that there has been any 
increase in poverty from this cau.^e." 

Wells' conclusions as to the future and as to policy are somewhat 
indefinite. "With regard to the future he writes : 

There is no reason for doubting that the wonderful material evolution of recent 
years will be continued, unless man himi^elf interposes obstacles, although the 
goal to which this evolution tends cannot be predicted or possibly imagined." 

But this outlook has its dark side. 
It will have to be admitted- 
he says — 

thilt the immense changes in i-ecent years in the conditions of production and 
distribution have considerably augmented — especially from the ranks of un- 
skilled labor and from agricultural occupations- — the number of those who have 
a rightful claim on the world's help and sympathy. That this increase is 
temporary in its nature, and not permanent, and that relief will ultimately 
come, and mainly through an adjustment of affairs to the new conditions, by 
a process of industrial evolution, there is much reason to believe. But, pend- 
ing the interval or necessary period for adjustment, the problem of what to 
do to prevent a mass of adults, whose previous education has not qualified them 
for taking advantage of the new opportunities which material progress offers 
to them, from sinking into wretchedness and perhaps permanent poverty, is 
a .serious one, and one not easy to answer." 

Carroll D. Wright may be regarded as the American pioneer in the 
inductive and statistical studies both of the business cycle and of labor 
displacing effects of mechanization. Though his methods of reporting 
and analy.sis are crude, he brings together a vast amount of statistical 
data and makes an effort to classify and interpret them in the light of 
current social-economic ideas. 

In his report on Industrial Dejjressions/* Wright defines the prob- 
lem in the followinor words: 



wiliid.. p. 432. 

8' Ibid, p. 4.-?5. 

82 Ibid., p. 67. 

«Ibid., p. 437. 

^ This was the first annual report of thp Bureau of Labor established by act of Congress, 
approved June 27, 1884, which provided for the appointment of a Commissioner of I^bor 
Jby the President. The Bureau was placed in the Department of the Interior, and Carroll 
D. Wright was the first Commissioner of Labor to be appointed. His first report was pub- 
lished in 1885. 



30 CONCENTRATION OF KCONOiMIC POWER 

The present industrial depression is the first of its kind as an en- 
tirety * * *. Histox'y is full of accounts of crisis of various descrip- 
tions * * *. Of old, . stagnations, when occurring, lasted through long pe- 
riods * * * 111 modern times, we have in the place of the long reaches of 
the past, short, sharp, and freiiuent disturbances in the business world.*' 

These disturbances or depressions must be clarified and their causes 
exphiined. They are closely related to technical progress on the one 
hand and to employment of labor, on the other. The connecting link 
is the increasing power of production, due to machines, with its con- 
sequences of overproduction (in relation to "immediate demand"), 
under-consumption, loss of markets, and unemployment. 

Labor displacement is a result of technical progress and an aggravat- 
ing cause of depressions. 

The rapid development and adaptation of machinery in all the activities 
belonging to production and transportation — 

says Wright — 

have brought what is commonly called over-production. * * * That labor- 
saving machinery so-called, but what more properly should be called labor-making 
or labor-assisting machinery, displaces labor temporarily cannot successfully be 
denied. All men of sound minds admit the permanent good effects of machinery ; 
but the permanent good effects of it do not prevent the temporary displacement 
of labor, which displacement, so far as the labor displaced is concerned, assists in 
crippling the consuming power of the community. 

The report then presents statistical data indicating for different 
industries the amount of labor displaced in recent years. The indus- 
tries covered are agricultural implements, brick making, boots and 
shoes, carriages and wagons, cotton goods, metallic goods, paper, 
pottery, etc. AVright makes the following summary : 

The mechanical industries of the United States are carried on hy steam and 
water power representing, in round numbers, 3,500,000 horsepuwei'. each horse- 
power equalling the muscular labor of 6 men ; that is to say, if men were employed 
to furnish the power to carry on the industries of this country, it would require 
21,000,000 men, and 21,000,000 men represent a population, accordinj; to the ratio 
of the census of 1880, of 105,000,000. The industries are now carried on by 
4,000,000 per.sons, in round numbers, representing a populati(ui of 20,000,000 
only. * * * The figures are only interesting because a condition represented 
by them is utterly impossible. They are to a certain extent valuable to show the 
enormous benefits gained by the people at large through the tippUcation of 
improved motive power. They illustrate, too, the extreme view of the displace- 
ment of labor, which, as already remarked, has been positive, and, it may well be 
said, to some extent permanent.^" 

Wright touches upon the question of tlie distribution of the gains of 
technical progress — 

A* a result of machinery — 

he writes — 

the worker is benefited in increased wages, sliorfer working tinu', and hii-her 
standard of living, and yet if tli(> (piestion should be asked as m whether the 
worker n-ceived his eciuitable share of benefits derived frnm the introduction of 
machinery, the answer nnist be "no." His greatest benefit has come through his 
being a consiuner. It is also true that macliiiiery has brought new occupations 
especially to women, but "there does exist a positive and emphatic over-production, 
and this over-prod\u'tion could not exist without the introduction ot power ma- 
chinery at a rale greater than the consuming power of the nations involved and 
of those depending upon their demand."" 



»= Report of the Secretary of the Interior, 49tli Cong., 1st sess., House of Represienta fives. 
Ex. Doc. 1, i,t. n. 1SS.5, p. 11. 
"'Ibid., pp. 87-88. 
«• Ibid., p. 89. 



CONXIENTRATION OF EIC'OXOMIC POWER 31 

Wright took up the question of the effects of machinery upon labor 
upon a larger scale in his report on Hand and Machine Labor.^* In 
this study Wright was concerned with a comparison of the "productive 
power" of hand and machine labor, and the effects of machinery upon 
labor unit requirements, costs of production, wages, and on the supply 
of labor. 

It is not possible to summarize the contents of these two volumes. 
The data presented cover a wide range of industries and are significant 
only insofar as they bear on the practical points at issue. What may 
be noted, however, are the thoroughness with which the study was 
planned, the care with which the samples were selected, the manner in 
which the trades and occupations were broken down into processes and 
operations, the methods of timing them, and the general spirit of objec- 
tivity in which the inquiry was conducted. It is a monument to 
Wright's constructive work in the field of industrial and economic 
reporting. 

Wright summed up the results of these studies in his book on the 
Industrial Evolution of the United States.*^ He contrasts the dual 
influence of machinery on labor, the "displacement of labor," and the 
"expansion of labor." His conclusion is that the latter is far more 
important in its effects than the former. He also stresses the "ethical 
influence of machinery on labor" — the tendency of the machine to 
diversify employment, to reduce working time, to improve living 
standards, and to extend the facilities for the education of the masses. 



s" Thirteenth Annual Report of the Commissioner of Labor, 2 volumes, 1898; Washington, 
1899. 

80 Carroll D. Wright, The Industrial Evolution of the United States, 1895. 



CHAPTER II 
TWENTIETH CENTURY PROBLEMS 

NEO-CLASSICAL THEORY AND ITS CRITICS 

In the history of social-economic thought and policy, the decade 
and a half preceding the first World AVar — the years 1900-14 — may 
be regarded as a period of transition from the nineteenth to the 
twentieth century. It was during these years that the new postu- . 
lates, methods, and concepts of economic theory, associated with the 
work of Jevons, Alfred Marshall, John B. Clark, the Austrians and 
of the Lausanne School, were elaborated by numerous followers and 
were combined into a system of "neo-classical economies'' which was 
to be the dominant school of economic thought. 

Insofar as the relation of technology to employment was conceived, 
the representatives of this school during this period saw no serious 
problems: On the whole, they hauked back to the optimistic views 
of Jean-Baptiste Say and the English "classicists," wliich were now 
reinforced by the concept of "economic equilibrium." In Marshall's 
Principles of Economics the word unemployment does not occur, 
and-his references to "discontinuous" employment are few and unim- 
portant. In general, though in his own way, he restates the "theory 
of (Compensation." The general assumption is that all factors of 
production, while competitive in a degree, are primarily comple- 
mentary and "constitute the field of employment for each other." ^ 
This means that as the efficiency of capital or labor is increased, the 
earnings of the other is enhanced. Thus — 

if the supply or efficiency of business ability increases, there is likely to be some 
displacement of manual labor by new contrivances for economizing effort, and 
by new inventions of various kinds. Biit this shrinking in some directions 
of the field of employment for manual labor will be more than comi>ensated 
in others * * * in like manner, an increase of material capital causes it 
to push its way into new uses ; and though in so doing it may occasionally 
diminish the field of employment for mamial labor in a few trades, yet on the 
whole it will very much increase the demand for manual labor and all other 
agents of production. - 

This takes place through the increase in national income which, 
owing to a fall in the rate of interest, is divided more in favor of 
labor than before. The increased demand for labor will come from 
new industries and fi-om the makers of new and more expensive 
machinery in all branches of production. 

In general, Marshall thought that "the inconstancy of employment 
in modern industry is apt to be exaggerated." ^ Though the rapidity 
of invention, the fickleness of fashion, and above all the instability 

' Alfred Marshall, Principles of Econoi scs. Second Edition, 1891, p. 709. 
^Ibid., p. 710. 
»Ibid., p. 736. 

277551— 41— No. 22 4 33 



34 CONCENTRATION OF ECONOMIC POWER 

of credit, do certainly introduce disturbing elements into modern 
industry; yet * * * other influences are workino; strongly in the 
opposite direction, and there seems to be no good reason for thinking 
that inconstancy of employment is increasing on the ^hole.^ These 
other counteracting influences were of an economic and social char- 
acter which tended to raise the standard of life, which in turn tends 
to increase the national dividend for the whole population as well as 
for different groups. 

The theories of the neo-classical economists on the effects of techni- 
cal progress on the demand for labor have been ably presented in 
some detail in the Survey of Economic Theory on Technological 
Change and Employment already referred to.^ The presentation of 
the Survey" is briefly summarized and restated in the following pages. 

The neo-classical theory rests on the concept of economic equilibrium. 
An equilibrium condition is — 

one in which all factors of production, including labor, are fully employed * * . * 
and are employed in a manner that is the best paying to everyone concerned, thus 
precluding all incentive to a change. The several producers, that is, the entre- 
preneurs whose decisions determine the pattern of production and employment, 
use the several factors of production — say, capital and lal)or — in proiwrtions wTiich 
pay them best ; and they pay for them at a rate corresponding to the value of 
their respective marginal products so that there is no incentive for the entre- 
preneurs to depart from those proportions. There is likewise no incentive for those 
who supply those factors of production — investors or workers — to shift from one 
field of employment to another since the marginal product of the several factors 
of production is the same in all fields afid so is, accordingly their remuneration. ° 

It is possible to conceive that an economic society, having attained a 
condition of equilibrium, should continue in such a condition indefi- 
nitely. Such a condition of "static equilibrium" would mean that 
there would be no further change in price relationships either between 
the commodities or between the factors of production. There would 
be- 
ne change in the apportionment of resources between the several fields of produc- 
tion, as well as none in the comparative shares of the several factors either in 
production or in the distribution of the product. In such a state the prices of 
the several factors of production are equal to the value of their marginal net 
products, and the sum of the prices of all the goods and services produced is 
equal to the sum of the remunerations of all the factors.' 

In such a "static equilibrium" there can be only oscillations around 
the point at which equilibrium is maintained. "Equilibrium eco- 
nomics" is largely concerned with the analysis of the price relationships 
and price mechanisms which equate or "equilibrate" supply and de- 
maud under such abstract and liAq^othetical static conditions. In a 
dynamic society there is movement from one condition of equilibrium 
to another. The forces which disturb a given equilibrium and bring 
about a shift to another are the increase of population, increase of 
capital, improvements in methods of production, changes in industrial 
organization and the growth of consumers' wants. But the forces 
which disturb eciuilibrium also set in motion forces which tend to estab- 
lish a new e(iuiiibrium position. We thus have a "moving equilibrium" 
which means that the economy is at one and the same time "gravitat- 

'Ibid.. p. 737. 

= Work Projects AdminLstration, National Research Project, Survey of Economic Theory 
on Teclinological ("liaiige and Employment by Alexander Gourvitch, May 1940. 
«II)irl., pp. 84-85. 
'Ibid., p. 85. 



CON'OENTRATION OF ECONOMIC POWEli 35 

ing" toward a given equilibrium and readjusting itself to disloca- 
tions in a new equilibrium position. 

The processes of adjustment and readjustment are automatic and 
take place through the several price mechanisms and the principle of 
substitution. 

In any given situation there is always a theoretical set of prices for all goods 
and services which are either on the market or in process of coining on the market, 
at which supply and demand would be equal and which would be adequate to cover 
the costs of production ; and that equilibrium price structure comprises a rela- 
tionship between prices of the several factors of production — between the wage 
rate and the interest rate — such as would assure full employment of all factors.' 

There can thus be no general overproduction of commodities. 
Neither can there be. in the long run, unemployed labor any more 
than there can be unemployed capital. There is in any situation a 
rate of wages and a rate of interest at which all labor and all capital 
will be employed. If there is unemployment, it is proof that wages 
are above the equilibrium level — that is. too high in relation to 
the marginal productivity of labor.^ The unemployment results 
because, under such conditions, capital is substituted for labor either 
through changes in industrial organization or through a shift of 
investments to industries employing relatively less labor. But the 
unemployment thus created tends to lower the wage level, while as 
more capital is substituted for labor its profitability (marginal 
productivity) will decrease and its price (or interest rate) will rise. 
These developments will check the further substitution of capital 
and stinndate the reemployment of labor. Inversely, a fall of '\^ages 
belo.w the equilibrium level will bring about a substitution of labor 
for capital; such substitution will be checked in time by the result- 
ing rise in Avages, declining marginal productivity of labor and the 
decline in the interest rate. 

Considering technological change, in particular, its first eifect is 
to cause dislocations in employment. But it results at the same 
time in a reduction of production costs, which means that some of 
the physical and economic resources, previously used in production, 
are set free and may be used to increase production elsewhere. These 
resources can be used since the aggregate purchasing power of 
the comnuniity is not impaired. These resources will generally be 
used. "As long as. they are not * * * used, the best-paying 
combination is not present, and it is the quest for that combination 
that carries a tendency toward a new equilibrium position.''' ^° Thus 
technological change while disturbing the economic equilibrium 
insofar as employment is concerned, sets into motion forces which 
tend to restore the full employment of labor. 

This tendency toward readjustment through lower costs and in- 
creased production meets no obstacles either in the possibilities of 
market expansion or of capital supply. According to neo-classical 
tlieorists, increased production means increased aggregate purchasing 
power and demand. Through the movements of wage and interest 
rates, capital accumulation adjusts itself to technical changes. The 
processes of technical change, increasing wealth, capital saving, and 

6 ibid., p. 86. 

" It follows from ♦^his that unemployment can alwaj-s be diminished by a lowering of wage 
rates. 

'"Work Pio.iects Administration. National Researoh Project. Survey of Economic Theory 
01) Technologiial Change and Employment, 1940, p. 91. 



3g CONCENTRATION OF ECONOMIC POWER 

employment tend to keep in step, thus making for harmonious de- 
velopment and progress. The dislocations caused by technical 
changes — such as disproportion in production, displacement of lal)or, 
etc. — are temporary, and are easily taken care of through the auto- 
matic operation of the delicate and intricate but nonetheless elf'ective 
mechanism of prices. 

In contrast to the elaboration of neo-classical and "equilibrium" 
concepts was the decided step forward taken in the develo})ment of the 
doctrines dealing with economic fluctuations, depressions, and inductive 
methods in the "crises." Aftalion, Tugan-Baronawski, Lescure in 
Europe and especially Wesley C. Mitchell in this country gave pre- 
cision to the idea of periodic cyclical movements in economic life and 
inaugurated the descriptive and analytical study of the "business cycle" 
which was to exercise such an important part in economic thought and 
social policy after 1918. 

While neo-classical theories were predominant, especially in academic 
circles, they had their critics who exercised considerable influence on 
thought and policy. The most important among them were perhaps 
John A. Hobson in England, ancl Thorstein Veblen in the United 
States. These two Avriters may be regarded as having reinterpreted 
the doctrines of Sismondi and Marx, being thus closely related to the 
neo-liberal and neo-marxian trends of thought of the day. Hobson 
stressed the shifts in the distribution of the labor supply caused by 
machinery and the influence of machinery on industrial depressions 
through its stimulation of "oversaving" or (what is the same) the 
intensification of "underconsumption." Veblen gave brilliant expres- 
sion to the idea that technological progress must accentuate the opposi- 
tion between the "machine process" and "business enterprise" and must 
involve the latter in ever more acute conflict with the needs for the 
preservation of capital, full employment of economic resources, and 
the development cf social and economic values. 

Both neo-classicists and their critics had a part in the making of 
policies with regard to unemployment. Sir William H. Beveridge, an 
able representative of neo-classical economics, influenced policy through 
his important volume on Unemployment: A Problem of Industry.^^ 
On the other hand, representatives of the dissident schools (Sidney and 
Beatrice Webb, John R. Commons, Charles Gide, and others) played 
an active part in the movements for specific legislative measures against 
the evils of unemployment. 

Despite theoretical differences, the general trend during these years 
was to minimize the importance of "technological unemployment" as 
the term has since been interpreted. The causes of unemployment held 
to be important were those inherent in the frictions of the economic 
system — seasonal variations, cyclical fluctuations, immobility of labor, 
etc. The remedies sought dealt, therefore, with improvement of the 
internal structure of plants and industries, the elimination of seasonal 
variations in production, and with bettering the organization of the 
labor market. The outstanding legislative achievements of these 
years in this field, were the establishment of the system of Labor 
Exchanges in England and the fii-st steps towaitl unemployment 
insurance in a number of countries. 



First published in 1909 ; new edition, Longmans, Green and Co., London, 1930. 



OON'OENTRiATION OF ECONOMIC POWER 37 

PROSPERI-TY AND TECHNOLOGICAL UNEMPLOYMENT 
IN THE UNITED STATES, 1924-29 

By 1924, interest in problems of cyclical unemployment was more 
or Less on the wane. The wave of prosperity which had begun in 
1922 gave a new turn to economic discussion. Broadly, three main 
currents of ideas may be distinguished which had a direct bearing on 
industrial and labor problems. One reflected the optimism of those 
who saw a "new era" in American economic life — an era of con- 
tinuous prosperity — rising to ever higher levels, steady in its prog- 
ress owing to "the ironing out" of industrial fluctuations, and so- 
cially stable, thanks to higher wages, a wider distribution of wealth, 
and the gi'owing "good will" in the relations of employers and 
workers. In the liglit of these ideas, there could be but little con- 
cern with problems of unemployment. Interest in the effects of the 
machine was largely in terms of its effects on wages and on tho 
future of the labor supply.^^ 

Another current was. fed by the post-war interest in the influence 
of monetary and banking policies upon price relations and upon, 
economic balance and grew out of the "new art of central banking" 
stimulated by the Federal Reserve Board in the LTnited States and 
by tlie central banks abroad. The discussion centered around the 
possibility of using the new devices of central banking — the redis- 
count rate and open-market operations — to achieve a stabilized price- 
level which was presumably desirable as a means for steadying busi- 
ness activity. This discussion led to a new emphasis in economic 
thinking on the importance of the monetary and credit mechanism 
for industrial activity and prepared the gi'ound for some of the 
ideas and proposals which were to be advanced during the yeai*s 
after 1929. with regard to employment and unemployment. 

The tliird current of thought flowed directly out of the efforts of 
organized labor to adjust to the new industrial developments. The 
main element in this current of thought was the emphasis on the 
"new industrial revolution" which was taking place in the United 
States, on the unprecedented growth of productive power and labor 
productivity which the revolution was bringing in its trail, and on 
the consequences which these industrial developments were having 
on the status and economic welfare of the workers. 

The discussion aroused by these issues went through two stages. 
Between 1922 and. 1926, the problem which was debated most was 
that of the distribution of the gains from the new increase in pro- 
ductivity — essentially its effects on wages. It was at its convention 
in 1925 that the American Federation of Labor came out with its 
theory of the "soc-ial wage" according to which labor, as a contribut- 
ing factor to the increase of productivity, should share in the benefits 
through increased wages. It was partly in connection with these 
discussions (though also in large measure owing to the war experi- 
ence which stimulated the development of economic research here 
and abroad) that the problem of measuring production and pro- 
ductivity began to attract attention and that the first important steps 

12 See for example, George E. Barnett, "Chapters on Machinery and Labor, IV. The 
Introduction of Machinery and Trade-Union Policy," Quarterly Journal of Economics, 
vol. XL, February 1926, pp. 209-231. 



38 COyQENTRATION OF ECONOMIC POWER 

were made in this direction.^'' Empliasis in the studies made by the 
Bureau of Labor Statistics was on labor productivity in rehition to 
labor costs and wages.^* 

Between 1927 and 1929, the emphasis shifted to the effects of the 
"new industrial revolution'" on employment. This change was due to 
several factors — the growing concern with the condition of surplus 
labor in the ''sick" industries — coal mining, textiles, railroads; the 
disappearance of old skilled trades; the rapid growth of new oc- 
cupations, and in some measure, also to the influence of European 
conditions and experience on American thinking. Perhaps, the most 
important influence was the situation in a few industries, such as 
cigar making, coal mining, railroads, machinery, in which rapid 
mechanization in combination with such other factors as wartime 
overexpansion, high wage rates, and rigid trade union policies, was 
responsible for large-scale displacements of labor. A contributing 
factor were the new personnel policies of management which, while 
aiming at lowering labor turn-over and at greater stabilization of 
employment, were alleged to involve the disemployment of older 
workers — the "men over forty." 

Largely, owing to the influence of organized labor, the generaliza- 
tion became current that, despite so-called prosperity, there was a 
large amount of unemployment in American industry due to mechani- 
zation and to technical changes. The term "rationalization" which 
had become current in Germany as a result of similar developments. 
was at first used to designate the new problem. But the, term 
"technological unemployment" was soon coined and became common 
currency.^^ 

The study of the "new problem" took three main forms. First, 
various government departments and private agencies made attempts 
to estimate the total number of unemployed on the basis of available 
data which were very meager indeed. These efforts gave the impetus 
to the various unemployment surveys and censuses which were to 
follow. 

Second, a series of statistical investigations were begun of the 
changes in labor productivity and of the displacement of labor by 
machinery in specific industries.^*' Labor productivity was measured 
in these studies by dividing volume of output by tlie number of em- 
ployees (or an index of output by an index of "employment ). This 
was a measure of productivity per worker, but it was realized that 
it would be desirable to obtain data for computing output per man- 

" E. E. Day, W. M. Persons, and E. D. Co.tIp. "An Index of the Physical Volume of Pro- 
duction," Review of Economic Statistics, September 1920-Jantiary 192l'; Walter W. Srewavt. 
"An Index Number of Production." American Economic Review. March 1921 : fi E. Day 
and Thomas, The Growth of Manufacturers, U. S. Census Bureau, Monograph No. IX, ch. I 
and II and appendix A. 

"See U. S. Bureau of Labor Statistics, articles in Monthly Labor Review, January 1923: 
"Labor Efficiency and Productiveness in Sawmills"; November 1924: "Labor Productivity 
and Costs in Certain Building Trades" ; September 1926 : "Labor Productivity and Labor 
Costs in Cotton Manufacturing." Bulletin N-407, published in 1926 : "Lnbor Cost of Pro- 
duction and Wages and Hours of Labor in the Paper Box Board Industry," Bulletin No 360 : 
"Time and Lalior Costs in Manufacturing 100 Pairs of Shnes," 1924. 

'^ No one seems to lay claim to priority in the use of the term. 

'"See U. S. Bureau of L'lbor Statistics. Monthly L:ilior Review for April 1937: "Displace- 
ment of Labor by Machinery In the Glass Industry"; May 1927: Comparison of Emplov- 
ment and Productivity in Manufacturing Industries, lOl'.t 25; September 1927: Article bv 
J. J. Davis, "Productivity of Labor and Industry" — "The Problem of the Worker Displac< d 
by Machinery" ; March 1929 : By Ethelbert Stewart. "Ultimate Effects of Automatic Machine 
Production." Also a series of articles during 1928-29 on Stability of Employment in 
Specific Industries such as automobiles, iron and steel, men's clothing, etc. ; and several 
bulletins on labor productivity in selected industries, e. g., Bulletin N-441 by Boris Stern: 
"Labor Productivity in the Glass Industry." 



CON'OENTRATION OF EOOXOMIC POWER 39 

hour. Displacement was conceived as a process of direct discharge 
of workers as a result of improvements in existing machines or of 
the introduction of new machines. 

The third approach to the problem was made in several studies 
which aimed to find out by the sampling method, through question- 
naires and interviews, what became of clisplaced workers, how soon 
they were reabsorbed into industry, and in what way.^' These studies 
were significant in that they disclosed the fact that, in a large pro- 
portion of cases, labor-saving devices resulted in unemployment of 
considerable duration and that the process of finding a new job was 
accompanied by hardships, in many cases by a lowering of the 
worker's industrial status and earning capacity. 

It slinnid be evident from t"he material presented above — 

wrote Dr. Lubin — 

that tliere is considerable mobility of labor at the present time and that the 
newer trades are absorbing a good percentage of the workers who are being 
discharged from the "older" industries. Absorption, however, is a relatively slow 
process. Those workers who do find new jobs are in a majority of instances 
compelled to remain idle for 3 months or more before finding new employment. 
AVhen they do secure new jobs, they frequently take them at a sacrifice in 
income. An appreciable portion of them, however, have been able to secure 
higher earnings in their new jobs than in their old.'^ 

The discus.sions of these years led to the hearings on unemployment 
held by the Committee on Education and Labor of the United States 
Senate, of which Robert F. AVagner was chairman.^-' These hearings 
stimulated national interest in practical proposals for alleviating un- 
employment and gave the impulse to the legislative efforts which cul- 
minated several years later in the passage of th Social Security Act. 

The prevailing ideas of these A^ears found their most elaborate ex- 
pression in the culminating economic document of the period — the 
leport on "Recent Economic Changes in the United States." -" The 
two volumes of the report mirror both the optimism and puzzlement 
of a country overwhelmed by extraordinary industrial developments 
which brought in their trail a mixture of prosperity and hardships, of 
economic growth and decay, of stability and unbalance. The indi- 
vidual writers of the various sections of the report reflect mixed ideas 
and attitudes ranging all the way from an unqualified optimistic faith 
in American initiative, ingenuity, versatility and mastery-^ to a 
cautious acceptance of the blessings of the present with faint mis- 
givings of the dangers lurking in the f uture.-- 

The problem of unemployment is treated in the report in this gen- 
eral spirit. The basic premise on which the treatment of the problem 
proceeds is that the industrial developments of 1922-29 were not 
revolutionary in character and differed from similar changes in the 
past only ''in speed and spread." "Acceleration rather than struc- 

" I. Lubin. "The Absorption of the Unemployed by American Industry." Brookings Insti- 
tution. 19l.'0 : and "Measuring the Labor Absorbing Power of American Industry" In Journal 
of American Statistical Association, March 1929 ; R. J. Myers, "Occupational Readjustment 
of Displaced Skilled Workers" in Journal of Political Economy. August 1929. 

^^ I. Lubin, "Measuring the Labor Absorbing Power of American Industry." Proceeding. 
American Statistical Association. 1929. p. 32. 

'" U. S. Senate. Committee on Education and Labor : Unemployment in the United States ; 
hearings. December 11, 1928. February 7. 1929. etc. (70th Cong., 2d sess.). 

-" Recent Economic Changes in the United States. Report of the Committee on Recent 
Economic Changes, of the President's Conference on Unemployment, Herbert Hoover, chair- 
man. 2 volumes. McGraw Hill. New York, first edition, 1929. 

=' See Introduction by Edwin F. Gay, vol. I. pp. 1-2. 

" "A Review" by Wesley C. Mitch3ll, vol. II, pp. 841-910. 



40 CONCENTRATION OF ECONOMIC POWER 

tural change," reads the report, "is the key to an understanding of 
our recent economic developments. Gradually the fact emerged dur- 
ing the course of this survey that the distinctive character of the years 
from 1922 to 1929 owes less to fundamental change than to intensified 
activity.^^ The report makes some attempts to measure the rate of 
acceleration, but without being conclusive. 

It is difficult— 
the report reads — 

to measure tlie technical progress of 1922-27, witii the data now available. It 
is still more difficult to make reliable measurements for earlier years, when 
censuses were taken at longer intervals and fewer supplementary figures were 
published. But doubts whether the rate of improvement in the past 6 years 
is unprecedented are not of great moment. It remains clear that the industrial 
revolution is not a closed episode; we are living in the midst of it, and the 
economic problems of today are largely problems of its making.^* 

The industrial changes between 1922 and 1929, which the report 
emphasizes, are the growth of many new industries and the great 
increase in productivity or output per worker, due to various factors 
(industrial research, new processes and materials, greater use of 
power machinery, etc.). On the whole, economic relationships and 
social institutions kept step with underlying technical and industrial 
changes. Mass production was accompanied by the growth of mass 
consumption due in large measure to the acceptance of the principle 
of "high wages" by employers; price relationships and movements 
showed balance and stability due to effectiveness of the monetary 
and banking system and to a greater reliance of the American economy 
on internal forces and markets ; the development of the service indus- 
tries and the greater use of leisure due to a shortening of hours of 
work raised living standards and added grace to American life. 

There were some maladjustments which caused hardship to one 
or another of the groups of the population, and among them the 
report singles out as of special importance the effects of technical 
progress on the workers. Thei^ is not complete agreement, however, 
among the contributors to the report as to the gravity of the prob- 
lem. One of the authors, for instance, merely states the opposing 
views held on the effect of the new techniques on skill and employ- 
ment and withholds any definite conclusiop. 

The thing that is certain is that our information concerning the extent, 
nature, and reasons of unemployment in this country is inadequate. The inade- 
quacy of public information is, unfortunately, equaled by the inadequacy of 
information in individual companies. Few of them know exactly what num- 
bers have been released or transferred on account of technical improvements, 
and almost none knows anything about what has happened to the men who 
were released."' 

In the chapter on Labor,-" a step forward is made in clarifying 
the meaning of labor productivity and its measurement, and in 
defining and estimating unemployment in the United States. With 
regard to productivity of labor, thei writer of the chapter says: 

The measures of the per capita output of labor considered in this section 
are not measures of the specific productivity of labor. They are the results 
of comparisons between the total physical output of industry and the number of 

=^ Recent Economic Changes, vol. I, p. IX. 

"Ujid., vol. II, p. 863. 

^^ Henry S. Dennison In chapter on Management, vol. II, p. 514. 

^ Leo Wolman, Labor, ch. VI, vol. II, pp. 425-493. 



CONCENTRATION OF ECONOMIC POWER 4X 

wage earners employed in producing it. . . . But statistics of man-hours are 
not available until very recent times and then only for a limited number of 
industries." 

The report indicates the factors which may influence per capita 
output : 

Changes in the i>er capita output of labor, as in total output, may clearly 
be due to a variety of factors. In the long run, the levels of education and 
skill of the working population of a country, the growth of capital and the 
use of machinery, the alertness and ingenuity of management, and the state 
of science may determine both the direction and the rate of change of indus- 
trial production. During shorter periods, accidental or abnormal factors, such 
as apparently operated from 1916 to 19^21, like sudden changes in the length 
of the workweek, marked variations in the efficiency of labor, resulting either 
from the state of mind of the workers, from the carelessness of management, 
or from the replacement of experieaiced workmen, may conceivably not only 
interrupt the prevailing trend of production, but also change its direction. The 
segregation and weighing of all these factors, or even of the most important 
of them, are not possible in the present state of knowledge. The most that 
can be done is to appeal to reasonable hypotheses and to informed common 
knowledge.^ 

The problems of method involved in computing per capita output 
are stated ii\ some detail : 

Both of the elements of the formula for deriving per capita output have 
serious, if irremediable, defects. The measures of the total output of indus- 
try are better for the mining and transportation industries than they are for 
manufacturing and agriculture ; and they do not include the highly important 
construction industry, for which there are no satisfactory statistics of either 
physical output or employment. They cannot, by their very nature, take into 
account changes in the character and quality of the products of industry. 
Since' the statistics of the production of raw materials or of commodities in 
their early stages of fabrication are more numerous than statistics of highly 
fabricated goods, the measures are too heavily weighted for raw materials and, 
consequently, underestimate the rise in total output. In general, there is less 
material for the new and growing industries than for the old est<ablished ones, 
whose rate of growth has probably already slackened before the compensating 
influence of the growth of the young industries can make itself felt in the 
\measure of total output. And it is, finally, not always certain that the chang- 
ing importance of industries is adequately allowed for in the weights used 
for the computation of average changes in total production. 

The employment indexes are probablji' superior for the manufacturing and 
transportation industries than for the rest. For manufacturing they are much 
more reliable in the later than in the early years, although one important series, 
that of the United States Bureau of Labor Statistics, appears to e:xhibit a down- 
ward bias even during the latest years. All of the series, finally register the 
amount of work done by the number of people employed and not by the time 
worked, and this procedure, as it has already been indicated, may involve mis- 
leading conclusions. 

The commodities that enter into the aggregate production of a country cannot 
all be recorded in the same unit. The simplest method of reducing all goods to a 
common denominator is to express them in pecuniary units. Variations In the 
resultant aggregates, however, would then reflect changes both in physical output 
and in prices. Correction for price changes, for many well-known reasons, raises 
as many problems as it solves. Resort must then be had to the device of the index 
number, which Is an average of the measures of relative changes in the items of 
a heterogeneous series. In place of aggregates of incommensurable units, such as 
bushels, feet, tons, trucks, and so on, the index number registers the weighted 
average of all changes in the number of units of output of the commodities under 
consideration. Even the construction of index numbers involves the difficult 
problem of discovering the importance of each industry, so that the changes in 
the output of each product may be properly weighted.^ 

«Ibid., pp. 446-447. 

»Ibid., p. 447. 

» Ibid., pp. 447-448, footnote 25. 



42 COXCIvNTRATIOX OF ECONOMIC P0WP:R 

The writer does not enter into an analysis of the causes of changing 
productivity. The factors are "hard to disentangle and, when sepa- 
rated, to measure. Even such an apparently simple concept as the 
mechanization of industry," he says, "does not yet lend itself to satis- 
factory statistical analysis." ^" Neither does he enter into a considera- 
tion of the types and causes of unemployment, being content to present 
a "new estimate of unemployment'.' based on more careful statistical 
methods. 

The position of the writers of the report with regard to the question 
of technological employment (which is put in quotation marks) is 
summed up in a section of the final review," which reads in part as 
follows : 

Among all the hardships imposed by increasing efficiency, most publicity has 
been given to the decline in the number of wage earners employed by factories. 
That is a matter of the gravest concern in view of the millions of families affected 
or threatened by the change, and in view of their slender resources. To it special 
attention has been paid in this investigation. 

The new phrase coined to describe what is happening, "technological unem- 
ployment," designates nothing new in the facts, though the numbers affected may 
be large beyond precedent. Ever since Ricardo shocked his rigid disciples by 
admitting tliat the introduction of "labor-saving" machinery may cause a tempo- 
rary diminution of employment, economists have discussed this problem. Grant- 
ing Ricardo's admission, they have nevertheless held that, in the long run, changes 
in method which heighten efficiency tend to benefit wage earners. English ex- 
perience since Ricardo's day seems to bear out this contention. The power 
looms, which put an end to hand-loom weaving after tragic struggles, have not 
reduced the number of British workers employed in weaving, or cut their average 
earnings. The railways, which displaced the old mail coaches and carters, have 
not reduced the number of transport workers or made them poorer. And the new 
trades of building and caring for the elaborate modern equipment must not be 
forgotten. There doubtless are cases in which improvements in methods have 
caused what promises to be a permanent reduction in the number of persons 
employed in an industry. By defining industry narrowly, these cases can be made 
numerous. But the broad result plainly has been that the industrial triumphs 
of the nineteenth century increased the demand for labor and increased its 
rewards. "Labor-saving" machinery has turned out to be job-making machinery. 

To recall these familiar facts (the Review continues) should not diminish 
by one jot our rating of the hardships suffered by men who are thrown out of 
jobs. They and their families often undergo severe privation before new em- 
ployment can be found ; the new jobs may pay less than the old or be less suitable ; 
too often the displaced man never finds, a new opening. Technical progress is 
continually made at cost to individuals who have committed no fault and com- 
mitted no avoidable error of judgment. No organized plan has been evolved for 
preventing such hardships, aside from the schemes devised by some trade unions 
for tiding their members over mechanical revolutions in their crafts. The 
nations have left tiie remedy to "natural forces"; they have trusted that the 
expansion of production, which improvements bring about, will presently open 
new places for the displaced workers.^^ 

The extent of disj^lacement and reabsorption during the period is 
sunmied up as follows : 

All these data are estimates of the net changes in numbers of persons "attached 
to" the occupations in question. They show that American wage earners met 
"technological unemployment" in manufacturing mainly by turning to other ways 
of making a living. The decline from 1920 to 1927 in the number of persoiis 
actually at work in manufacturing enterprises is put at 825.000, l)ut the number 
of miemployed among the people who depended on factory work for a living 
increased only 240,000 between 1920 and 1927, according to tlie best figures avail- 
able. If these estimates are approximately correct, then some ."•85,000 of the 

3" Ibid., p. 461. 

3> Written b.v We.sley C. Mitchell. 

3= Ibid., pp. 876-877. 



CONCENTRATION OF ECONOMIC POWER 43 

workers laid off by factories had taken up other occupations. That is, 71 percent 
of the workers displaced had attached themselves to new trades by 1927. 

Adopting a new occupation, however, does not guarantee getting a new job. 
The surplus workers from our farms and factories who hunted for fresh openings 
increased unemployment in other fields. The expansion of business, particularly 
the expansion of miscellaneous and mercantile occupations, made places for per- 
haps four and a half million new wage earners. But the supply of new jobs has 
not been equal to the number of new workers plus the old workers displaced. 
Hence there has been a net increase of unemployment, between 1920 and 1927, 
which exceeds 650,000 people * * *. 

One may wonder at the versatility, iniLianve, and mobility of Americans, as j 
evidenced afresh by their prompt shifting of occupations on so great a scale in 
recent years. One may wonder also at the rapid expansion of the trades which 
have absorbed some 5,000,000 employees in 7 years without reducing wage rates. 
But one must not forget that these shiftings have been compulsory in large 
measure ; men have been forced out of farming and forced out of factories as well 
as pulled into automobile services, shops, and restaurants. And the employment 
balance is on the unfavorable side. While our economic progress has meant larger 
per capita earnings for all workers taken together, it has imposed severe suffering 
upon hundreds of thousands of individuals.''^ 

In tlie nature of the report, no concrete proposals for dealing with 
the problem were in order. An interesting suggestion, however, is 
made by Jlenry S. Dennison to the effect that the crucial factor in the 
problem may be the rate of inventions of new consumers' goods as 
compared with the rate of inventions of machines and new manu- 
facturing processes. 

TECHNOLOGICAL CHANGE AND THE DEPRESSION, 

1929-33 

During the "great depression*' interest again shifted to the cyclical 
aspects of the unemployment problem. An ever larger place in the dis- 
cussion of these years was occupied by theories as to the origin and 
causes of cyclical fluctuations and by proposals for dealing Avith such 
unemployment (public works, monetary reflation, price and wage pol- 
icies, economic planning, etc.). The relation of technical change to 
unemployment ^^■as examined more clearly from the point of view of the 
part which technical improvements play in tlie development of the 
business cycle. On the other hand, the long term significance of the 
technological factor was stressed by the advocates of technocracy and 
economic planning. 

The discussion of the problem during these years centered around 
the issues of the "compensatory theory." the social effects of tech- 
nology, the metliods for alleviating the hardships due to occupa- 
tional displacements, aid the need for obtaining more information 
on the subject. Whatever advances were made in the thinking on the 
subject may best be brought out by considering the writings of those 
years. 

THE RESTATEMENT OF THE COMPEXSATORY THEORY 

While the idea of technological unemployment loomed large in the 
public mind, professional economists, by and large, continued to uphold 
the "compensatory principle" according to which workers displaced by 
technical improvements sooner or later were reemployed owing to the 
industrial expansion due to these improvements. However, in the light 

"Ibid., pp. 878-879. 



44 CONCENTRATION OF ECONOMIC PuWER 

of the facts disclosed by recent studies, economists were ready to admit 
(as some of the earlier economists had done) that the transfers of work- 
ers and capital made necessary by technical changes might cause hard- 
ships to the workers, and that such hardships were serious enough to 
call for special action by the community and the Government. 

Among the attempts to restate the "compensatory doctrine" of the 
relation of technical change to employment, that of Paul H. Douglas 
may be summarized here.^* Douglas starts out by summarizing the 
facts which have aroused the fear that American workers were being 
crowded out of industry permanently. It is true, he says, that between 
1919 and 1929, the output per man increased 45 percent, and that at the 
same time not only a relatively smaller proportion of the population 
was engaged in manufacturing, but the absolute number of workers in 
manufacturing fell from 9.000.000 to 8,100,000. In mining, output per 
person during the same period increased between 40 and 45 percent, 
while the number employed fell by about 7 percent. A similar trend 
was to be found in the railroad industry and in agriculture. 

Despite these facts, Douglas holds that "permar.ent technological 
unemployment is impossible" and that improvements in machinery and 
in managerial efficiency will not throw workers out of employment 
permanently. Douglas is ready to admit that the reasons given by 
economists in the past (e. g., by Say) in proof of this theory are 
"largely inconclusive to the Inodern mind" because they were conceived 
in terms of a barter economy. He then proceeds to restate the theory 
in terms of our modern economy. 

In substance, however, Douglas' restatement runs in terms of the 
same economic processes and behavior as the older theories. As tech- 
nical improvements increase output per worker, they reduce the labor 
costs of the commodities produced. Insofar as there is full and com- 
plete competition among employers, they will vie with one another in 
reducing prices, and prices will fall in proportion to the reduction in 
costs. As prices are lowered there will be a larger demand for prac- 
tically all commodities, there will be need for more workers to produce 
the increased quantity of goods demanded, and thus the displaced 
workers will be reemployed. 

The important factor in the process to which Douglas gives particu- 
lar emphasis is the varying elasticity of demand for different com- 
modities. Whether the elasticity of demand of a conmiodity is equal 
to, greater than, or less than unity will determine the degree to which 
the price will be reduced, the extent to which market demand and 
production in the same industry will expand, the mnnber of workers 
that may be employed in the same industry or may have to seek 
employment elsewhere. 

It has bt^en just this ehisticity of demaiul — 

he writes — 

which has caused the number of persons employed in the automobile industry to 
increase, despite tlie increase in output per worker. * * * An increase in 
production lowered both costs and prices, but the increase in tlie quantity de- 
manded more than made up for the increa.se in the average output, so that the 



3< See raul II. Douelas and A. Director, "The Problem of Unemployment," New York, 
1931, pp. 1L'1-15S ; also Panl II. Douglas, "Technological Unemployment,' American Ped- 
erationist, vol. 37, no. 8, Amr. 1930, pp. O^S-O.'O. 



CONCENTRATION OF ECONOMIC POWER 45 

relative number of persons employed in manufacturing increased with every 
decade.* 

If the demand for a specific commodity is inelastic, an increase in 
output in tliat industry will not be accompanied by a proportionate 
increase in demand. In such cases fewer workers will be needed in 
that industry, and there may be an appreciable displacement of labor. 
But the displaced workers will find employment in other industries. 
The process is simple: The improvements in the industry necessarily 
lower the price of the commodity produced by that industry. As a 
result of loAver prices, consumers are able to obtain the same quantity 
of goods for less money. They thus save purchasing power which they 
can now spend either on other necessities or luxuries or can invest in 
new capital goods. Thus at tlie same time that men are being squeezed 
out of one industry purchasing power formerly expended on the prod- 
ucts of that industry is transferred to other industries and builds up 
new opportunities for work. Furthermore, there is an exact mathe- 
matical relationship in the process. 

Not only are new opportunities for employment built up * * * but they 
are built up to an equal degree to that by which the older opportunities decay. 
For every man laid off a new job has been created somewhere, and the ratio 
between monetary purchases and employment is still the same^as before.^ 

Such a transfer of workers from some lines of employment to others 
took place between 1919 and 1929 when — 

young men who otherwise would have been farmers or clothing workers have in 
fact become movie ushers, saxophone players, and house-to-house canvassers. 

Differences in elasticity of demand thus determine whether there will 
be shifts to new industries, and to what extent. The gist of this rea- 
soning is summed up as follows : 

Improvements in industrial processes, like changes in demand, will produce, 
therefore, a shifting of labor and capital within the economy as workers and 
Investors transfer themselves from industries where their return in terms of 
exchange value is less than the average to industries where the return is above 
or approaching the average. These shifts are inevitable in a progressive society 
where there is a tendency toward an equalization of return. Those thrown out 
of work will not, however, be permanently unemployed, and to this degree, there- 
fore, the fears of permanent technological unemployment have been greatly 
exaggerated.*' 

The process and ultimate results are the same when an increase in 
output is obtained not by mechanical changes but by improvements in 
managerial methods. Whether employers eliminate waste through 
more scientific methods or shut doAvn inefficient plants or regularize 
production, in all cases the displaced workers will be reabsorbed owing 
to the fact that the savings of consumers, owing to lower prices or the 
increased earnings of the workers retained, will increase production 
and purchasing power. 

The general thesis that "permanent technological unemployment" 
cannot occur is not invalidated by the fact that owing to the existence 
of interest and rental payments the reduction in total costs may not 
be as great as the reduction in labor costs. Neither is it affected by 
the existence of restrictions on competition such as price agreements 
and other monopolistic practices. The only difference is that under 

l^ American Federationist, Vol. 37, No. 8, Aug. 1930, pp. 927-928. 
*«Ibld., p. 930. 
"Ibid., pp. 930-931. 



46 CONCENTRATION OF ECONOMIC POWER 

conditions of monopoly the gains from the technical improvements 
are appropriates, in larger measure by employers m the form of 
larger profits. But the employers must either spend their added 
gains for luxuries or invest them. In either case, they increase the 
demand for commodities' and draw the displaced workers mto other 
industries. The general argument may be summed up in Mr. Douglas 
own words, as follows: 

Summing up, therefore, we can say that the displacement of workers from 
their former occupations because of technological changes will be greater: 
(1) The less the quantity demanded of a commodity increases with a given 
reduction in price per unit; (2) the less is the proportion which lal>or costs 
form of the total expenditures; (3) the less is the degree to which a reduction 
in costs will reduce price; and (4) the less important is the operation to the 
whole industry. ^ ^ „ . ^^ 

In other words, therefore, the amount of displacement from former to other 
jobs will vary inverselv with the elasticity of demand, the importance of labor 
in the final "product, the degree of competition, and the relative importance 
of the operation or operations primarily affected by the technical changes. 

In any event however, employment opportunities are being built up elsewhere 
which will ultimately be adequate to provide for an added number of workers 
equal to those who under such conditions may have been eliminated from 
any given industry. 

in the long run, therefore, the improved machinery and greater etfaciency 
of management do not throw workers permanently out of employment nor 
create permanent technological unemployment. Instead, they raise the national 
income and enable the level of earnings and of individual incomes to rise. 

Wliile technical improvements, according to this theory, do not 
create any long-run problem of unemployment, they necessitate 
readjustments which take time and which may cause temporary 
unemployment. Even when owing to an increase in demand for 
the product workers are reemployed in the same industry, it takes 
some time before the "pick-up" in demand makes itself felt m such a 
way as to induce manufacturers to expand production, and m the 
meantime many workers are laid off. When workers have to shift 
from contracting to expanding industries, there is also a tinie lag. 
for the contraction takes place more quickly than the expansion and 
expanding firms may work longer hours with the same number of 
workers before entering on a program of larger production. The 
change-over in occupations is further complicated by the fact that 
workers.are reluctant to change their trade or place of residence and 
that they may be hampered in their movements either by strong local 
ties such as home oAvnership or by lack of funds to pay the expenses 
of moving to a new place. Finally, even if Avorkers are trans- 
ferred to new jobs, they often have to take jobs which are less satis- 
factory to them and which pay lower wages. "From all of these 
causes', therefore," says Douglas, "technological and business change 
creates a considerable amount of temporary unemployment Avhich 
in the short run creates havoc." --^ 

Douglas suggests a number of remedies to alleviate such hardships. 
He lays i)articular stress on "the forecasting by competent orgaiiiza- 
tions of the industries and trades in which displacement of labor is 
most likely to occur and the probable degree of displacement which 
may be expected." By predicting more or less accurately impending 
technical changes and* the elasticities of demand of the commodities, 



M Ib'd., p. 9.38. 
30 Ibid., p. 942. 



CONCENTRATION OF ECONOMIC POWER 47 

it Avould be possible to adjust tlie labor supply to the production 
needs of different industries. Unemployment would also be reduced 
by better timing, namely, by introducing technical improvements dur- 
ing periods of prosperity; by better organization of the labor market 
through public employment offices; by "revamping" our system of 
vocational training, the hardship caused by temporary unemploy- 
ment could be alleviated by means of a dismissal wage for those forced 
out of an industry and by an adequate system of unemployment 
insurance. 

The same position was formulated in more concrete terms by Wil- 
fred I. King.'*° In brief, King agrees that the displacement of men 
by machines has been going on rapidly for two centuries. Though 
most of the men displaced find new work "rather promptly," there are 
many instances in which skilled workers are compelled to accept un- 
skilled work at lower wages. Also many older men may never regain 
their status, and even in the best of times, some displaced workers 
will be idle for days or weeks. "It is clear, therefore, that there is 
such a thing as technological unemployment." But it is "negligible" 
in volume, and it does not in the least invalidate the contention of the 
"orthodox economists" with regard to the long-run effects of technical 
progress. 

The present situation — 

writes King — 

may be summed up by saying that no facts or figures thus far discovered cast 
any doubt upon the approximate validity of the orthodox economic theory that 
the forces giving rise to technological unemployment tend, at the same time, to 
create a demand for new goods, and that the production of these new goods 
normally calls for a volume of labor roughly equaling the quantity displaced. 
From this premise it follows that since labor-saving devices increase production 
without materially decreasing the ability of workers to find jobs, such devices 
are decidedly beneficial rather than injurious to society as a whole." 

King further claims that labor tends to obtain a liberal proportion 
of the gains in national income due to technological improvements. 

^ow. I. King, "The Relative Volume of Technological Unemployment," Proceedings of 
American Statistical Assn., 1933. 
" King, loc. cit. 



CHAPTER III 

RECENT STUDIES AND REPORTS, 1934-40 

GENERAL CHARACTER OF REQENT STUDIES 

The course of economic events in the United States after 1933 did not 
alhiy the debate on technological unemployment. On the contrary. 
The persistence of a large body of unemployed even during the peak 
of recovery in 1937, the continued lag in the indexes of employment as 
compared with indexes of pruihiction in mining, manufacturing, and 
other intlustries, the apparent dilHculties of placing new entrants in 
industry and other unfavorable asj)ects of the employment situation, 
gave the issue greater urgency. The question of the etl'ects of the 
machine on the worker's condition, and especially on his opportimities 
of employment, remained in the forefront of public discussion. 

Opinion in the United States during these years, as in the preceding 
periods, continued to be dividetl. On the one extreme were those who 
hailed tlie machine as the agency of social-economic advance and as 
the creator of jobs and who refused to admit that there can be such a 
thing as ''technological unemployment." This view was defended by 
employers and engineers, and by a considerable number of econ- 
omists. In a letter addressed to the chairman of a sub-committee of 
the Committee on Labor of the House of Representatives which held 
hearings on the subjects in February and March 1930, Xoel Sargent, 
secretary of the National Association of Manufacturers, wrote: 

In a special study, the National Industrial Conference Boaid, presented a par- 
tial list of the new industries wnich have made extensive use of labor-saving 
devices, but which in turn have created new tields of direct and indirect euiploy- 
nieiit for those who had lost their positions in the particular obsolete fields. 

These industries are — 

Electrical machinery, apparatus, and supplies. 

Motor vehicles. 

Auto parts and bodies. 

Rubber tiies and inner tubes. 

Manufactiu-e of gasoline. 

Kayon and allied products. 

Manufactured ice. 

Aluminum manufactill ;s. 

Typewriters and parts. 

Mechanical refrigerators. 

Cash registers and adding and computing machines. 

Oil, cake, meal, cotton seed manufacture. 

Aircraft and parts. 

I'honogiaphs. 

Photographic apparatus and materials. 

Motion-picture apparatus. 

Asbestos products. 

Fountain pens. 

When these new industries became a part of our business structure they, with- 
out a doubt displaced workers in obsolete lields and workers on obsolete machinery 
at the time of such introduction. However, the new labor-saving devices brought 

49 
277551— 41— No. 22 5 



50 CONCENTRATION OF ECONOMIC POWER 

in their \Aake bigger industries, need for more raw materials, more wliolesale and 
retail merchants, and greater transportation facilities. 

The final results of the study mentioned above showed that because of the 
development brought about by the.se new labor-saving devices and new technologi- 
cal developments the average number of wage earners per 100,('60 of total popula- 
tion in all manufacturing industries increased from 4.944 in 1879 to 7,273 in 1929. 
This repre.sented more than a 45 percent increase in the number of manufacturing 
wage earners per 100,000 population. These figures do not, of course, iiicllide the 
vast increase in employment opportunities which developed at the same time in 
the distribution and servicing of these new and improved products.' 

At the other extreme were those who continued to blame teclmical 
progress for the large amount of unemployment in the United States. 
Among the documents placed before the sub-committee referred to 
above was a letter from the president and secretary of the National 
Organization for Taxation of Labor-Displacing Devices which read 
in part as follows : 

The economists of big business, and their journalistic hirelings have made a 
fetish of machine progress. They develop more labor-displacing devices, cut 
down pay-rolls, and try to squeeze dividends out of a surfeited market. This 
gigantic conspiracy against American laboi- has brought our total unemployed to 
over 11,000,000 people, and still they call it "progress." 

And further — 

The use of mechanical power is the main factor in the Increase of production, 
and decrease in employment opportunities. Therefore, labor-displacing devices 
must be recognized as tlie principal cause of the existing unemployment situation.' 

Similar views Avere expressed by representatives of organized labor, 
both in the A. F. of L. and in the C. I. O. 

Between these extremes were the advocates of a middle position, to 
the eifgct. namely, that there was no problem of technological unem- 
ploym(^nt "in the long run," but that it was definitely a problem ''in 
the short run." Opinions differed as to how serious the problem was, 
its exact nature, and how it should be dealt with. 

Considerable advance, however, was made during these years in 
preparing a factual and theoretical basis for a clearer analysis of the 
problem. The work done between 1934 and 1940 has thrown a great 
deal of light on various aspects of the problem — on methods and 
measurements of mechanization, on the effects of industrial improve- 
ments upon skill and efficiency, on the rehition of such improvements 
to output, on the interrelations of technical progress to general eco- 
nomic processes, etc. Wliile this work of research and analysis has 
left room for disagreement on the main question, it has undoubtedly 
narrowed the area of disagreement. It has developed more fully con- 
cepts and methods which are necessary for a more effective approach 
to the problem, both theoretically and practically. It has lielped to 
place the problem in better historical i:)erspective by bringing within 
the view of current thinking some of the antecedent discussions of the 
subject. And it has brought to light a large amount of descriptive 
and statistical data as a basis for re-evaluating the nature of the 
problem and the validity of the proposals made for its practical 
solution. 

The work done during these years in this field of inquiry is volumi- 
nous, and only a small part of it can be considered here. The studies 



' Investigation of Unemployment Caused by Labor-Savinjj Devices In Indusfry. HearinffS 
before a subcommittee of the Committee on Labor, House of Representatives, 74th Cong.. 
2d sess. on H. Res. 49 ; Washington, 1936 ; p. 87. 

'Ibid., pp. 79-82. 



CONCENTRATION OF ECONOMIC POWER 51 

selected for examination deal with the problems of meclianization, 
labor displacement, industrial growth, and capital formation. Some 
of these studies were sponsored by Government or private research 
institutions, while others Avere carried out by economic writers indi- 
vidually. 

MECHANIZATION AND LABOR DISPLACEMENT 

In 1934 the National Bureau of Economic Research published a 
study on Mechanization in Industry analyzing the results of a 
survey conducted over a period of years by Dr. Harry Jerome. In this 
study, Dr. Jerome set himself the task of giving a systematic account 
of the process of mechanization, and of providing methods for measur- 
ing its extent and some of its industrial consequences. 

The scope of Dr. Jerome's study is limited in two ways. First, 
while recognizing fully the great importance of non-mechanical 
changes in industry, such as the sub-division of labor, better use of 
equipment, standardization and simplification of products, improve- 
ments in the quality of raw materials, elimination of seasonal fluctua- 
tions, budget control, market forecasting, methods of selecting work- 
ers and determining wages, shipping methods, etc.. Dr. Jerome is 
concerned only with mechanical changes in the strict sense of the 
term. His use of the term is described as follows : 

In the broadest sense we mean by mechaniziitioii the nse of tools or equip- 
ment of any kind to aid the linman brain and muscle, and by "increasing- 
mechanization" we refer to any change in metliods or equipment that tends to 
less(5n reliance on the unaided mental and manual endowment of the worker 

* * *. Our interest centers chielly, however, in power mechanization — in the- 
increasing reliance on equipment driven by gejierated power, be it steam, elec- 
tricity, compressed air, or gasoline that furnishes the motive power.' 

Second, the study deals entirely with labor-saving mechanical 
changes — their forms, causes, and growth. Dr. Jerome draws a dis- 
tinction between "productivity-increasing'' changes (those which in- 
crease the units of output per hour of labor) and "labor-displacing" 
changes (those which reduce the number of workers required). "While 
recognizing the close relation of the two concepts. Dr. Jerome centers 
his study on the latter. The effects of labor-saving devices on em- 
ployment, skill, industrial health, etc., are considered briefly, with 
more emphasis on labor displacement than on other effects. 

The process of mechanization is made considerably clearer by con- 
sidering it in relation to specific operations and by breaking down 
labor-saving mechanical changes into a number of types. Dr. Jerome- 
points out 

four fairly obAious ways in which changes m equipment may reduce the labor 
requirement, relative to the output, on specific operations, namely, by: (1) 
eliminating one or more hand operations; (2) increasing the si^eed of tlie ma- 
chine; (3) enlarging capacity (through greater physical size) without cor- 
responding increases in the labor requirements for feeding and attention: {i)< 
substituting a different process requiring less labor, such as the substitution of 
electric welding for riveting. 

The use of better materials or a greater durability of machine parts has, by 
keeping the machine in steadier oi^eration. substantially the same effect as making-, 
the machine faster or larger.'' 

' Harry Jerome. Mechanization in Industry, National Bureau of Dconoraic R'-scnrchu 
New York. 1934, p. 41. 
Mbld., p. 42. 



52 CONCENTRATION OF ECONOMIC POWER 

A detailed liistorical account is presented in Dr. Jerome's study of 
recent changes in teclinique and equipment in American industries — 
agriculture, manufacturing, mining, construction, transportation, serv- 
ice industries, and in the handling of materials — and a comprehensive 
l)icture of the use of mechanical appliances at the time of writing 
is presented. It is the most comprehensive statement on the subject 
that had hitherto been made. In describing the mechanical changes 
in specific industries, attention is given to their effects on produc- 
tivity as measured by annual output per man or by output per man- 
hour, on the particular ways in which skills were leveled, processes 
and occupations modified or combined, and the number of workers 
required affected. Particularly illuminating are the results of the 
special surveys in particular plants, showing the number and charac- 
ter of new installations, methods and processes and their effects on 
labor used before and after the changes in equipment and methods 
of production.^ 

While the descriptive portions of Dr. Jerome's study are interesting, 
of much greater importance are the various methods he suggests for 
measuring "changes in mechanization" in time, and for comparing 
"the degree of mechanization" betAveen industries at the same time. 
While the same methods are applicable for both purposes — 

some modes of measurement may be more serviceable for the comparison of 
differences in mechanization at a given date while others are more available or 
more useful for the comparative study of the process of mechanisation over a 
period of time." 

The use of the various methods depends upon the availability of the 
necessary statistical data for different industries and periods of time, 
on the scope and limitations of these data, on their comparability, 
and on the degree to which their value is affected by complicating 
economic processes. Dr. Jerome examines the data available for the 
use of different methods and points out their validity as well as their 
defects for the purpose. 

On the basis of the nature of the process of mechanization and of 
the data at hand, Dr. Jerome distinguishes five general methods for 
measuring mechanization, which are given in appendix A. He 
classifies the factors which determine the rapidity of mechanization 
under three headings : "Technological," "Pecuniary," and "Psycholog- 
ical." These are not entirely independent or mutually exclusive. 
Dr. Jerome thinks that "It is scarcely too much to say that a machine 
can be manufactured to special order to perform almost any series 
of operations when the need becomes sufficient to justify the ex- 
pense." ' Still, mechanization is retarded by the "limited selective 
ability of the machine" and by local and temporary technical difficul- 
ties due to plant construction and other conditions. The use and 
spread of mechanical devices are also limited by considerations of 
costs and profitability. Technical progress generally outruns the ac- 
tual advance 'of mechanization owing to the fact that a fuller use of 
machinery may involve an increase in overhead costs or discarding 
existing plant which employers are reluctant to sci-ap, or it may call 
for financing which the management finds it difficult to arrange, etc. 

»Ibi(l.. chapters III-V 

« Ibid., p. 255. 

'Jerome, op. clt., p. 329. 



CONCE^'TRATION OF ECONOMIC POWER 5g 

Amoii^ the factors which stimulate mechanization may be mentioned 
th.e ability of vertical combinations to finance new inventions, a large 
volume of funds at low interest rates (as existed in the United States 
between 1922 and 1929), larc^e-scale production,® an extensive market, 
and standardized production and consumption. Considerations 
whicli tend to retard mechanization are irregularity of production, 
short operating seasons, and the availability of casual labor, a shorter 
working day, frequent style changes in the industry, small-scale pro- 
duction, diversity of procedures, and the stagnant or declining state 
of the industry. 

The rate of mechanization is also affected by the level of wage rates 
relative to other costs and by the marketing policies of the machine 
producing firms. In general, ''Other things being equal, high wages 
encourage the use of machines and stimulate efforts to perfect them,, 
while cheap labor retards mechanization.'"^ However, where wages 
are very low, labor may not be sufficiently cheap to prevent the intro- 
duction of some mechanical equipment. The trend toward mechan- 
ization is accentuated by a rising wage level. It is also hastened by 
shifts in the relative wages of different grades of labor, especially by 
a rise in the wages of unskilled relatively to those of skilled workers. 
But the effects of wages are often overshadowed by other factors, 
Mechanization in the United States between 1922 ancl 1929 was very 
rapid, though the differential between the wages of unskilled and 
skilled workers tended to widen owing to a fall in the wage rates of 
the unskilled. ^° The marketing policies of machine producers and the 
degree of competition in machine using industries may promote or 
retard mechanization, according to circumstances. 

Among the psychological factors the attitudes of employers and 
of organized labor are of special importance. Employers may be 
influenced by inertia, uncertainty as to the value of a new device, 
failure to recognize its merits, or concern for their workers. The 
attitude of labor toward machines is determined largely by the 
expected effects on jobs, skill differentials, and earnings. The" posi- 
tion taken by labor toward the introduction of the machine has 
"varied over a wide range, from vigorous attempts to prevent its 
adoption and use, through discouraged indifference to its progress 
or acceptance qualified by restrictive control measures, to reluctant 
acquiescence, and finally, to the stage of cooperation for efficiency 
where the worker not only acquiesces in but even helps to initiate 
innovations, with the hope of sharing in the resulting gains." ^^ As 
a rule, a refusal on the part of unions to permit their members to 
operate machines has rarely been effective for any length of time. 
The workers have been more successful in prescribing the conditions 
under which new machines may be used. 

In considering the effects of mechanization on the quantity of 
labor required, Dr. Jerome advances the discussion of the subject 
by a closer analysis of terms and by suggesting methods for measur- 

= Joi7^f. ''f '^!l^^^'^ ^^^^^ *°U*^' expense of mechanizer! equipment, and the farfs that in 
r?nnL^i^.H-^ ^■f'', ^^^ ""i^ be sufficient work to keep a machine busy and that machines 
cannot ordinarily, be transferred from one type of work to another without adjustments and 
nrnA„^t"i ^^Pf"^^'} malfe high mechanization dependent in considerable part upon mass 
F„^- "S*'°;.l "^"*l methods survive chiefly where variety and distinction qu.iiitv and 
"J^iyidPality are the primarj- considerations." (Ibid., p. 338.) 

^ Ibid., *p. 344. 

'"Ibid., pp. 345-348. 

"Ibid., p. 356. 



^4 CONCBNTRATIOX OF ECONOMIC POWER 

ing labor displacement. In the first place, he distinguishes clearly 
ihe various elements of labor affected by mechanization. Dr. 
Jerome says: 

In computing the effects of productivity-increasing improvements, it is essen- 
tial to note what labor is included. For this purpose we may distinguish 
operating, auxiliary, embodied, and indirectly required labor. Operating labor 
is that required directly in a particular process, such as the operation of a 
brick-molding machine. Auxiliary labor is that required in the plant for such 
operations as oiling, inspecting, adjusting, and repairing the machine — in short, 
all plant labor that is necessitated by the use of the machine but is not con- 
sidered as engaged in its direct operation. Embodied labor is the labor 
applied to the production of the machine itself and the materials of which 
it is made (prorated over the useful life of the machine), to the production 
of the materials used in machine repairs, and to the production of the power, 
oil, grease, and other materials, if any, consumed in the operation of the 
machine. Lastly, we may extend the scope of our comparison still further 
by seeking to include not only the labor required in the factory production 
of the given commodity and in the production of the equipment used in the 
factory, but also the additional labor required to put the given commodity 
in the hands of the final consumer ready for use * * *. It is the labor 
required beyond the manufacturing stage that we have designated as 'indi- 
lectly required." While studies in productivity and lab«r displacement ordi- 
narily stop short of the point of allowing for all these indirect factors, we 
should recognize that until they are included we have not ascertained the real 
<?hange in productivity or the real amount of labor saved or displaced. ^- 

The term "labor displacement" is also given a more precise mean- 
ing. "Labor displacement is a decrease in. the number of workers 
required in (1) a specified operation * * *; (2) the plant as a 
whole; (3) an occupation, such as carpentry, wherever practiced; 
■(4) an industry as a whole." ^^ In accordance with these changes, 
we may have (1) operation labor displacement; (2) plant labor dis- 
placement; (3) occupational displacement; (4) industrial displace- 
ment; or (5) complete displacement. Summarizing his analysis, 
Dr. Jerome writes : 

* * * labor saving sometimes means an increase in productivity, sometimes 
a displacement of labor ; the two concepts are not necessarily identical. Fur- 
thermore, the productivity comparison may be based upon only the direct 
operating labor or may be broadened to include auxiliary, embodied, and 
indirectly required labor. In like manner, di.scussions of labor displacement 
may refer only to the effect on the size of the crew in a given operation,, on 
the labor in the plant as a whole, on the number in a given occupation, in a 
given industry, or in employment of any kind. Finally, there may be dis- 
placement of one type of labor, balanced more or less by increase in another 
type; there may be displacement of skill as well as of number * * * 

In practice most discussion of labor saving refers to labor di.splacement 
or increasing productivity in specific operations or plants, sometimes with and 
sometimes without allowance for increa.se in output, but ordinarily not allow- 
ing for all of the offsetting increases in embodied and indirecth- required 
labor.i' 

The above distinctions lead to the following definition: Techno- 
logical unemployment is the complete displacement of workers from 
all industry either temporary or permanent, which arises from a 
technological change. ^^ Technological unemployment may occur, 
even when the total demand for labor is unchanged, as a result of 
■occupational and industrial changes. Even when operation and 
plant labor displacement is offset by increases in the employment of 

■' Ibid., pp. 28-29. 

" Iliid., p. 30. 

"Ibid., pp. 31-32. 

"■Ibid., p. 31. 



CONCENTRATION OF ECONOMIC PO^VER 55 

indirect labor, "we have no assurance that the additional men" 
employed in producing machinery or in transporting the product 
will be "the same men replaced in the plant or even men of the same 
occupation or skill. There may be occupational and industrial labor 
displacement with possibly some technological unemployment, even 
though the total demand for labor in the industry is unchanged. 
There may be numerous shifts that result in increasing employment 
for some individuals and types, and decreasing employment for 
others." ^« 

In measuring labor displacement, it is desirable to distinguish 
between (1) actual displacement, i. e., the number of workers known 
to have lost work on an operation or in a plant, etc., regardless of 
change in output; (2) constructive displacement, which expresses 
the ratio between the labor required under the new method and the 
labor which would be required under the old method to produce the 
same total output. This difference is "constructive displacement" 
for ordinarily we cannot assume that the total output would have 
been the same if the old methods had been retained; and (3) poten- 
tial labor displacement which is an estimate of the total labor dis- 
placement which might reasonably be expected to occur upon gen- 
eral adoption of a new machine; such an estimate implies informa- 
tion on the number of persons in the occupation affected, on the 
labor reduction ratio attributable to the machine, and the percentage 
of plants in which the use of the machine is feasible. 

Dr. Jerome presents a number of examples showing the effects of 
selected technical changes on the quantity of labor required. He 
emphasizes the fact that a substantial part of the total reduction of 
labor requirements in an industry takes place by a gradual process 
of nibbling away at the size of the staff required. But he is more 
concerned with '^the possibilities of estimating aggregate construc- 
tive displacement in industry for which he suggests four methods: 

"Method A, the fixed-base, current year output method, answers the ques- 
tion : How much less labor did it require to produce the current output than 
would be required at the productivity rate of the base year?"' '^ The algebraic 
formula of this method is 

T=Pi (Li— Lo)'* 

Method B, the fixed-base, base-year-output method, answers the question: 
How much less labor would be required to prodi ■ the base-year-output at the 
current productivity rate than actually was req ii'ed at the base-year produc- 
tivity rate?" ^^ The algebraic formula for this n thod is : 

T=Po (Li— Lo) 
Method C, the ijcar-to-ijear. cHrrent-year-outpu, method, answers the question: 
What is the cumulative constructive displacement when the displacement for 
each year is computed by multiplying the current-year-output by the differ- 
ential between the labor requirement ratios of the current and the immediately 
preceding year?""" The formula for Method. C is: 

T=^[Pi (L-Lo)] 



"Ibid., pp. 34-35. 

"Ibid., p. 377. 

i'"T = tlie constructive tecbnolofzical cbanae in the volume of employment (expressed in 
number of workers or manyears. man-weeks, man-days, or man-hours) * * *. 

Pi = total output in physical units in the current year: P,, in the base year. 

L, = the labor requirement ratio in the current year (labor per unit of output) ; Lj in 
the b:ise year." (Ibid., p. 376.) 

'» Ibid., p. 377. 

^"'Ibid., pp. 377-378. 



55 CO^^CENTRATION OF ECONOMIC POWER 

Method D. the year-to-yrar, prcceding-ycar-output method, answers the ques- 
tion : What is the cumulative constructive displacement when the displacement 
for each year is computed by multiplying the output of the preceding year by 
the differential between the labor requirement ratios of the current and the 
immediately preceding year?'"^' The formula for Method D is: 
T=2[Po (Li-L.,)] 

Dr. Jerome gives — 
a simple hypothetical example * * * to make clearer the differences in 
these four methods. Assume that in 1927 an industry employing on the average 
100 workers had an output of 1.000 units and hence a labor requirement ratio 
of 0.1 worker per unit. The corresponding data for 1928 are 100 workers, 
2,000 units of output, and a labor-requirement ratio of 0.05 workers iier unit ; 
and for 1929, 150 workers, 6,000 units of output and a labor-requirement ratio 
of 0.025 workers per unit. 

Under the conditions assumed, the constructive labor displacement in 1929 
is 450 if the differential in 1927 and 1929 in the labor-requirement ratios is 
applied to the 1929 output (Method A), but only 75 if it is applied to the 
1927 output (Method B). But if the year-to-year method is applied by first 
e.stimating the displacement from 1927 to 1928, and then from 1928 to 1929. and 
cumulating' the two results, the displacement is 250 if the labor-requirement 
differential is applied to the current year output in each instance (Method C), 
but only 100 if it is applied to the preceding year output (Method D).^^ 

The year to year method is regarded by Dr. Jerome as nearer to 
reality than comparisons with a distant base. 

In the first place, the year-to-year changes in employment due to techno- 
logical improvements are probably more important than the relatively long-time 
or possibly permanent displacement effects. Furtliermore, the significance of 
estimates of the constructive displacement of labor is obscured by any degree 
of interdependence between changes in total output and in productivity rates. 
If total output and productivity are independent of each other over a long 
period, then the fixed-base method may be reasonably accurate ; but if, as is 
more often true, the trend in the total output of the industry is in part both 
effect and cause of changes in the rate of productivity, then of the two methods, 
the comparisons with the preceding year as a base are the less likely to be 
distorted by the interdependence of changes in total output and in productivity 
rates.-'' 

Estimates of constructive labor displacement have serious limita- 
tions. They tell us nothing about the factors responsible for changes 
in productivity, they do not indicate whether the identical workers 
who were made unnecessary by gains in efficiency are those who have 
been absorbed by increases in output. Nor do they explain to 'what 
extent changes in employment have absorbed the increases in popu- 
lation of working age, or suggest the rate at which occupational read- 
justment takes place. In view of the general "crudity" of our indus- 
trial mechanism. Dr. Jerome concludes that at least some temporary 
technological unemployment must occur and that the tendency is "for 
technological clianges to result in a substantial period of unemploy- 
ment for the men displaced and frequently to necessitate their taking 
employment at a lowered wage." -* He does not find convincing evi- 
dence, liowever, of an inherent tendency for mechanization to create 
an ever larger permanent body of unemployed. "The element of truth 
in this charge against the machine lies in the fact that there is a lag in 
absorption ; and consequently the more rapid the displacement, the 
greater, i)robably, is the pool of at least temporarily unemployed work- 

« Ibid., p. 378. 
"^ Idem. 



» Ibid., p. 379. 

'* .Teiome, op. cit., p. 387. 



CONCENTRATION OF ECONOMIC POWER 



0/ 



ers. It may be a pool made up of ever changing individuals but even 
at that it represents in a sense a more or less permanent addition to the 
volume of unemployment." ^^ 

PRODUCTIVITY, PRICES. AND EMPLOYMENT 

Among the studies sponsored by the National Bureau of Economic 
Research which have dealt with the more general economic aspects 
of technical change are those by Frederick C. Mills. Three of these 
are of special interest, namely, Economic Tendencies in the United 
States published in 1932; Prices in Recession and Recovery which 
appeared in 1936, and Bulletin 70 of the Bureau entitled "Employ- 
ment Opportunities in Manufacturing Industries of the United 
States," dated September 25, 1938. In these three studies. Dr. Mills 
has developed the thesis that technical changes result in industrial 
displacement, has supplied an estimate of the decline in employment 
in recent years which may be attributed to technological factors, and 
has discussed (he connection of such unemployment with the incidence 
of the gains from increased productivity and with the general opera- 
tion of the price system. It may be convenient to consider the results 
of each of these studies in the order in which they were made.-' 

In his Economic Tendencies in the United States, Dr. Mills was 
concerned with changes in productivity as one of the tendencies of 
American economic development, and treated the problem of intra - 
industrial displacement only incidentally. iBut his treatment of the 
subject assumes a larger interest when viewed in relation to general 
economic tendencies. In the pre-war period, from 1901 to 1913, ac- 
cording to his computations — 

the output of manufactured goods in the United States advanced jit a rate of 
approximately 3.9 percent a year ; the volume of raw materials produced in- 
creased at an average rate of 2.2 j^ercent a year. The increasing proportion 
of fabricated goods con.sumed with rising living standards, the steady advance 
in fabrication outside the home and changes in the character of our foreign 
trade help to account for these differences.'^ 

Significant also was the unevennjess in the rates of growth of dif- 
ferent industries and of groups of industries 

The output of articles of human consumption increased between 1901 and 191.S 
at a rate of 2.6 percent a year, a rate comfortably in excess of the rate of 
growth of population (2.0 percent a year). The margin of approximately 0.6 
percent a year represents the increase in volume of consumption goods avail- 
able per capita of the population, for raising the standard of living * * * 
The growth of production of this type was relatively stable. * * * 

The output of additions to the total supply of capital equipment increased by 
5.0 percent a year. * * * Per capita of the population, the annual in- 
crements to the country's stock of capital equipment (including replacements) 
wpre increasing at a rate close to 3.0 percent a year. Current well-be- 
ing * * * was being steadily enhanced during this period. More rapid, 
however, was the flow of new goods (and replacements) into the fund of 
capital.^ 

The growth in the physical volume of manufacturing industries 
in the 15 years between 1899 and 1914 was due, ia the first place, to 

25 Ibid., p. 388. 

"Since the above was written, there has appeared Bulletin 80 of the National Bureau 
of Economic Research, September 9, 1940, which contains a study by Frederick C. Mills 
on The Anatomv of Prices. 1890-1940, which also bears on the subject. 

2« Frederick C. Mills, Economic Tendencies in the United States, National Bureau of 
Economic Research. New York, 1932, p. 12. 

^Ibid., pp. 21-22. 



58 



CONCENTRATION OF ECONOMIC POWER 



an increase of 36.1 percent in the number of wage earners (averaging 
2.2 percent a year). In the second phice, it was the result of ad- 
vancing productivity (averaging 1.7 percent a year). Dr. Mills uses 
the term productivity as equivalent to output per wage earner per year, 
and his index is based on the number of wage earners in manufacturing 
plants; no account is taken of salaried workers or of changes in hours 
of work. Furthermore, no attempt is made to explain the factors 
(human, mechanical, organizational) which made for higher output. 
Dr. Mills says : 

It is to be borne in mind that index numbers of per capita output do not 
measure changes in the specific productivity of labor. Per capita productivity 
may increase because of improvements in equipment or in industrial organiza- 
tion, increased skill on the part of personnel or enhanced productive capacity 
due to changes in any of the factors of production. Indexes of per capita 
productivity may be accepted as measures of changes in the productive efficiency 
of industrial organizations viewed as functioning units, but not as measures 
of the net contribution of any one factor to these changes.'^" 

The relative increases in nurnber of workers and in- output per 
worker were uneven when considered by 5-year periods. This is 
shown in the foil uing table : ^^ 



Census interval 


Increase in 
volume of 
manufacturing 
production 
(percent) 


Increase in 

number of 

wage earners 

(percent) 


Increase in 

output 
per capita 
(percent) 


1899-1904 - - - 


-1-20.2 
-1-28.5 
-1-14.1 


+S.I 
-1-20.2 
+4.7 


+11.2 


1904-9 . 


+6.9 


1909-14 


+9. a 







Commenting on these tendencies. Dr. Mills writes: 

The record of the 15 years from 1899 to 1914 indicates that the factors 
responsible for the great advance in production of manufactured goods were 
an increasing number of workers, larger and better equipped establishments, 
and steadily rising output per worker employed. (The growth of demand 
was, of course, essential to the realization of the advantages of large-scale 
production.) The stream of manufactured goods produced in 1914, a stream 
greater by 76 percent in volume than that of 1899, was turned out by a 
working force (of wage-earners) only ,36 percent greater, and by a number 
of establishments only 13 percent greater. There are clear signs here of 
the growing emphasis upon technical efficiency and enhanced productivity per 
unit as factors of increased production, an emphasis which has been even 
more pronounced in recent years.^^ 



3" Ibid., p. 28, footnote 1. The limitations in measuring productivity are stated by Dr. 
Mills as follows in The Anatomy of Prices. 1890-1940. p. 8. footnote 12 : 

"The term productivity is perhaps open to misinterpretation. For purposes of compari- 
son it is useful to express changes in the aggregate output of manufacturing industries in 
ratio to changes in certain standard quantities. Thus the standard of reference might be 
number of manufacturing establishments, horsepower used in production, number of dollars 
invested in capital equipment, number of men employed, or man-hours worked. Such ratios 
may be thought of as measures of productivity. However, we should not assume that 
exclusive responsibility for observed changes in aggregate output attaches to the factor 
represented by any particular standard of reference. Changes in production are resultants 
of the interplay ot organizational, managerial, equipment, and labor factors, operating 
jointly in the productive unit. 

"Even for a single establisbnicnt the productivity measurement is an average, reflecting 
the combined resultant of the application of diverse skills, utilizing various tools, to widely 
different types of materials. In default of detailed information about the effectiveness of 
effort applied to specific tasks under controlled conditions, such an average has its value, 
but it suffers from all the defects of averages in ueneral. This limitation is more marked, 
of course, wlien the average is for an entire industry, or for the heterogeneous activities 
that constitute manufacturing enterprise as a whole. Such an average provides but a 
crude measure of the play of the thousands of specific factors that affect the fruitfulness 
of productive effort in fabricational processes." 

=' Mills, Economic Tendencies in the United States, p. 28. 

32 Ibid., pp. 38-39. 



CONCENTRATION OF ECONOMIC POWER 



59 



Dr. Mills relates the changes in production with the price move- 
ments which were taking place during the same period. In the 
general trend of rising prices there was a persistent divergence of 
prices in different sectors of the economy. The "real worth" (real 
purchasing power) of raw materials was advancing while that of 
manufactured goods was declining. 

Amoug raw materials the gains in real worth, per unit of product, were 
greatest for farm crops; the purchasing power of these commodities increased 
at the notable rate of 1.5 percent a year. * * * Producers of raw min- 
erals suffered a decline in purchasing power per unit of product. * * * 
Producers' goods [in general were] cheapened slightly, [while] consumers' goods 
tgained] slightly in real value. This was due to the fact that (leaving out 
the influence of monetary factors) the prices of producers' goods were affected 
by the lower prices of fabricated mineral products, by falling production costs 
and widening markets. On the other hand, consumers' goods showed an up- 
ward tendency owing to the rising values of farm products, especially of 
food-stuffs. The divergence between these two classes of goods, however, was 
not very marked. 

The divergent price trends of the pre-war period meant a di- 
vergence in the fortunes of the different groups of the population- 
American farmers experienced a rise in their scale of material 
well-being. On the other hand, the increased industrial produc- 
tivity did not result in any substantial addition to the real income' 
of employed workers in general, while the real returns of manu- 
facturing labor actually declined (at the rate of 0.1 percent a year).^^ 
These shifts in the terms of exchange between farm and non-farm 
elements of the population and the inequalities in the gains from 
increasing productivity are accounted for by differences in produc- 
tive conditions such as the slower rate of increase in the output of 
farm products, the varying effects of monetary conditions, and the 
repercussions of improved techniques and widening markets par- 
ticularly upon costs and prices in the manufacturing industries. 

Following the same method and using a large array of statistical 
data. Dr. Mills traces in similar fashion the economic tendencies 
of the period from 1922 to 1929. The aggregate output of movable 
goods (i. e., excluding construction) increased 34 percent during 
this period; both aggregate and per capita output were higher than 
in the pre-war era, as shown below : ^•^ 





Average annual rate of increase in— 


Year 


Volume of 
production 
(percent) 


Population 
(percent) 


Production per 
capita of popu- 
lation (percent) 


1901-13— 


-1-3.1 
-f3.8 


-h2.0 
-1-1.4 


-t-1.1 


1922-29 


+2 4. 







There was a wide margain during this period (1922-29) between the 
rates of growth of raw materials and manufactured goods, farm and 
non-farm products. The output of raw materials increased at an 
average annual rate of 2.5 percent ; manufactured goods at a rate of 4.5 

33 Ibid., pp. 82-85, 133. 

5*>Ibid., p. 244. "Productive sei-vice.s rendered directly, and not embodied in goods, are 
necessarily excluded. There is reason to believe that 'production' of this non-material type 
was growing in relative importance during this period, but the degree of advance cannot be 
estimated with any accuracy." (Ibid., p. 245.) 



^Q concf:ntratiox of economic] power 

l^ercent a year, farm products advanced at a rate of 2.0 percent a year, 
while non'-farm products increased at a rate averaging 5.1 percent a 
year. Witliin these major groups there were Avide divergences in the 
rates of growth of individual industries. Total construction averaged 
$0,700,000,000 per year (in estimated value of contracts awarded) ; the 
annual average rate of growth (in volume) was 8.1 percent for com- 
mercial buildings, 9.3 percent for industrial buildings, 4.7 percent for 
public and institutional buildings, 3.7 percent for apartments and 
hotels. 5.1 percent for one and two family houses, and 11.4 percent for 
public works and utilities.^^ 

Great differences marked the production of different classes of con- 
sumption goods during these years. The output of foods increased at 
an average annual rate of 1.6 percent (barely above the rate of popula- 
tion growth which was 1.4 percent); other perishable consumption 
goods (excepting gasoline, kerosene, anthracite coal, newsj^rint and a 
few others) also lagged behind the general advance in production; so 
did the production of semi-durable goods (boots and shoes, textiles, 
etc.). It was among durable consumption goods (automobiles, furni- 
ture, electrical equipment, etc., including residential construction) that 
the great expansion of output took place (G.3 percent per year for 
durable consumption goods and 4.3 percent for residential construc- 
tion). The rapid increase of consumer expenditures occurred, in the 
main, in the markets for goods wdiich are more or less durable, and it 
was the swelling production of these goods which gave this period its 
characteristic tone.^'' 

Even greater than the rate of growth of durable consumption goods 
was that of production goods or capital equipment. Finished equip- 
ment, including non-residential construction and public works, ex- 
panded at the average annual rate of 6.4 percent. Again there were 
wide discrepancies between the industries entering into this general 
division : the output of machinery advanced 7.3 percent a year, while 
transportation equipment declined 1.1 percent a year. But in general, 
an increasing proportion of oui' total annual output of goods took the 
form of equipment designed to further the processes of round-about 
pi-oduction."' 

Among the ec(moniic tendencies of the period from 1919 to 1929. one 
of the most striking features was that the advance in productivity in 
manufacturing industries was accompanied by a decrease in the total 
labor force employed. "From 1919 to 1929 output per ^^■orker employed 
increased approximately 43 percent among the industries'" covered 
in Dr. Mills' sample."'' "The work tliat rociuired 100 men in 1919 could 
be done by 70 in 1929. Thirty out of 100 could have been dispensed 
with, '■'- '•' * if no increase in aggregate output had been desired. 
Productivity per Avorker increased between 1919 and 1929 at an aver- 
age annual rate of 3.8 percent," and between 1923-29 at a rate of 3.3 
percent a year (compared with rate of increase of 1.7 percent a year 
between 1899 and 1914.)^^ This increase was accompanied on the one 

3''Il)ifl., pp. 250-2G4. 

3"I!)k)., pp. 270-275. 

=" Ibid., pp. 277-280. Dr. Mills sharps the view that the devotion of so hirpe a propor- 
tion of the country's productive energies to the construction of capiral equipment was a 
factor in the collapse of 1929, though "we lack a criterion * ♦ * for df^terniinius; the 
optimum relations between output of consumption goods and of capital equipment, relations 
■whicli may conduce to equilibrium." (Ibid., p. 281.) 

2«The sample covers from 42 to 46.7 percent of total value of products reported in the 
Census of Manufactures for 1923, 1925. 1927, and 1929. (Ibid., p. 289.) 

''Ibid., pp. 290-291. . ' ^ 



CONCENTRATION OF ECONOMIC POAYER 



61 



hand by a tendency toward large scale production (as shown by a 
decrease in the number of establishments of 18 percent between 
1919 and 1929 and a gain in output per establishment of 68 per- 
cent during the same i^eriod), and on the other hand by a decrease in 
the number of wage earners. This may be seen from the following, 
table : " 



Census interval 


Increase in 
volume of 
manufac- 
turing pro- 
duction 
(percent) 


Change in 

number of 

wage 

earners 

(percent) 


Increase in 

output per 

wage 

earner 

(percent) 


Change in 
number of 
establish- 
ments 
(percent) 


Increase in 
output per 

establish- 
ment 

(percent) 


1923-25 


+2.4 
+1.8 
+8.4 


-4.6 
-3.3 
+0.3 


+ 7.3 
+5.2 
+8.0 


-7.0 
-4.6 

+5.7 


+10.1 


192.5-27 


+6.7 


1927-29 ---- -- - 


+2.6. 







Thus, a new tendency makes itself felt during this period. Con- 
sidering the 30 years from 1899 to 1929, Dr. Mills writes : 

We note the highly suggestive fact that not once has there been a checli tcf 
the increase in per capita productivity. The rate of advance has varied greatly, 
but the tendency toward increasing productive efficiency has persisted, in good^ 
years and bad. In general, however, the chief factor in expanding production 
prior to 1923 was an enlarged body of wage-earners. This was true during the 
great advances from 1904 to 1909, from 1914 to 1919, from 1921 to 1923. Since 
1923, however, better technical equipment, improved organization and enhanced 
skill on the part of the working force seem definitely to have supplanted num- 
bers as instruments of expanding production. The persistence of this tendency 
must compel men to consider its implications for the future." 

The price movements during 1922-29 differed sharply from those of 
the pre-war period. The years before the war Avere marked by a 
persistent upward movement of the price level. The period from 
1922 to 1929, witnessed a declining movement, world-wide in its reach, 
due to a variety of "complex and obscure reasons.'' This had various 
effects on different groups of the population — in general unfavorable 
to farmers and favorable to "agents of fabrication". The effects of 
this general movement were complicated by the price movement dur- 
ing the war and post-war boom and by the liquidation of 1920-21. 
Owing to war demands, the prices of raw^ materials had advanced 
greatly, but for various reasons the producers of these materials were 
hard hit by the collapse of 1920-21. As a result, the index number 
of purchasing power per unit of raw materials fell from 100 in 191.'? 
to 83 in 1921 or 17 percent. On the other hand, manufactured goods, 
while not experiencing as large a rise during 1914-19 were less hit 
in 1921, and their index of unit purchasing power in 1921 stood at 
108 as against 100 in 1913, that is showed a gain of 8 percent. In 
1922, therefore, there was a wide disparity between farm and non- 
farm commodities, a relationship quite the reverse of the situation 
before 1913. Between 1922 and 1929, the prices of raw materials 
advanced, while the prices of manufactured goods declined. But 
while disparities between the wholesale prices of farm and non-farm 
products were thus reduced, manufactured goods continued to enjoy 
an advantage during these years. Also, the prices received by farmers 
increased between 1922 and 1929 at a lower rate than that at which 



" Ibid., pp. 291-301. 
"Ibid., p. 291. 



g2 CONCENTRATION OF ECONOMIC POWER 

prices of raw farm products advanced, and the gain in purchasing 
power, measured in terms of commodities actually bought by farmers 
was 0.6 percent per year, per unit of product.*^ At the same time, 
prices of foods and of consumers' goods in general advanced, while 
those of producers' goods declined. 

The price movements of these years benefited the producers of manu- 
factured goods and resulted in increased profits. Manufacturers were 
favored by low prices of raw materials, increasing productivity, and 
lower labor costs per unit. 

Increasing output and rapidly increasing productivity— 
Dr. Mills writes — 

brought substantial declines in labor costs per unit of product, even with high and 
rising wages. Conditions in world markets for raw materials favored the buyer 
and gave the manufacturer the benefit of relatively low prices for such materials. 
In this situation labor costs, per unit of manufactured goods, and material costs, 
per unit, declined more rapidly than selling prices. The gap in our information 
has to do with the course of overhead costs. During this period capital invest- 
ments in manufacturing plant and equipment were increasing at a fairly rapid 
rate. These investments * * * were made at relatively high prices. It is thus 
not impossible that overhead costs per unit of product were advancing toward the 
close of this period, despite a swelling volume of manufacturing production. But 
if such advance occurred, there is no evidence that it offset other gains sufficiently 
to curtail profits per unit of product. In most manufacturing industries falling 
general prices between 1923 and 1929 brought to the producer the gains he has 
been supposed to reap under conditions of advancing prices and brought, in addi- 
tion, declining raw material prices. These fortunate conditions combined to 
maintain, and in many industries to swell, profit margins. During this period we 
appear to have had in the majority of manufacturing industries the curious and 
perhaps unprecedented condition of falling general prices and falling prices of 
manufactured goods, combined with an expanding margin between costs and sell- 
ing prices of manufactured goods, and with manufacturing profits which increased 
not only in the aggregate but per unit of product as well.". 

One of the results of the post-war price movements was the improved 
social-economic position of the industrial wage-earner. The real earn- 
ings per capita of labor in manufacturing industries between 1922 and 
1929 increased at the rate of 1.7 percent a year (as against a decline of 
0.1 percent a year during the pre-w^ar period).^' Instead of fighting to 
hold his own against a constantly rising cost of living, as he did during 
the pre-war period, the wage-earner substantially elevated his con- 
sumption standards and exerted an influence, as never before, on 
production and investments.*^ 

Offsetting the gains in the^sition^f the workers were the increased 
difficulties of adjustment which he had to face and which showed them- 
selves in unemployment and in industrial displacements. "An increas- 
ing volume of unemployment during an ei-a of economic expansion was, 
considering its magnitude, a new phenomenon in our history. Elqually 
striking are the related statistics of industrial displacement." Dr. 
Mills relates these difficulties of the wage earner to the other malad- 
justments which made themselves felt, especially to: (1) price inflexi- 
bility due to heavy investment in overhead, price regulation, 
nionopolistic and semimonopolistic control, trade agreements, changed 
distributive methods, and otlier influences; (2) increasing shai-e of pro- 

«Ibi(l., p. 348. 
«Ibid., p. 397. 

" Wage rates advanced even more sharplv. Earnings of iinskillod male workers increased 
'»o/f' rapidly than those of skilled. (Ibid., pp. 478-479.) 
*''Ibjd., p. 414. 



CONCENTRATION OF ECONOMIC POWER 



63 



ductive energies devoted to durable goods whicli have a higher elastic- 
ity and therefore a greater instability of production ; and (3) an excep- 
tionally large accumulation of capital funds which helped maintain an 
upward swing in security prices and in urban real estate values. Price 
inflexibility inpeded the adjustments made necessary by shifts in in- 
dustry, while the growth of capital facilitated mechanical improve- 
ments and was thus a major factor in technological unemployment.*^ 
Dr. Mills measures the degree of industrial displacements between 
manufacturing industries by computijig accession and separation rates, 
•dS shown in the table below : *^ 



Census period 


Number 
of indus- 
tries 


Average 

number of 

wage 

earners 


Accessions 


Separa- 
tions 


Acces- 
sion 
rate' 


Separa 
tion 
rate* 


1923-25 


320 


8,483,768 
8, 267, 736 
8, 514, 427 


■ 234. 554 
263, 539 
626,267 


646, 191 
322, 833 
261, 135 


2.8 
3.2 
7.4 


7.6 


1925-27 


3 9 


1927-29 . .- .... 


3.1 







1 Accessions as percentage of average number employed. 

2 Separations as percentage of average number employed. 

On the basis of these computations. Dr. Mills concludes: 

On the average, over each of the three biennial census periods coming be- 
tween 1923 and 1929, 49 men out of every 1,000 employed were separated from 
given manufacturing industries. Additions to the number employed averaged 
45 to every 1,000 on the pay rolls of manufacturing plants. Separations meas- 
ure the burden placed upon wage-earners by industrial change. That it was a 
heavy burden during the prosjjerous period from 1923 to 1929 is indicated by 
these figures. Not only was the rate of separation much higher than it had 
been over longer- pre-war periods; it was higher than the accession rate, 
which may be taken as an index of employment opportunities in manufacturing 
industries. Between 1923 and 1929 men were being turned out of manufacturing 
industries in greater numbers than in pre-war years, while the numbers of 
new men taken on were relatively much smaller. High productivity and 
rapidly expanding production brought instability of employment and un- 
certainty of income to many, during this era of business prosperity.** 

In other writings, Dr. Mills has continued to emphasize that the 
economic effects of a given technical change depend upon the degree 
of elasticity of demand for the products involved, upon the flexibility 
of the price system, upon the rapidity of technical changes, and upon 
the way in which the gains from increased productivity are reflected 
in the prices of different groups of commodities.*^ He combined 
these ideas in a more systematic way in his book on Prices in Keces- 
sion and Recovery ^° to give a general view of the interrelations of 
increasing productivity, prices, and employment opportunities. The 
central fact, according to Dr. Mills, is that an increase in produc- 
tivity due to technoligical change releases productive energy,^^ but 
that the gains in productivity may be variously distributed. The fol- 
lowing main distributions, with consequent effects, are possible: (a) 



«Ibid., pp. 531-540. 

♦^Ibid., p. 420, 

*" Mills, Kconomic Tendencies, pp. 531—532. 

<" See his introduction to Harry Jerome's "Mechanization in Industry", National Bureau 
of Economic Research, Npw York. 1934. 

5" National Bureau of Economic Research, New York, 1936. 

" Dr. Mills, as already indicated above, measures productivity as output per worker 
(between 1899 and 1929) and as output per man-hour since 1929. He recognizes the defects 
in such measurement, but regards it as adequate for purposes of his analysis. See "Prices 
in Recession and Recovery," pp. 435-^36 



g4 CONCENTRATION OF ECONOMIC POWER 

Reduction in ho'"-s of work, with higher time rates of pay; selling 
price unchanged; no increase in demand for goods; no change in 
distribution of purchasing power; no stimulus to larger production; 
increased leisure for all workers, (b) Reduction in number of 
workers employed, with higher time rates to those retained in em- 
ployment, no change in selling price, no change in aggregate amounts 
of purchasing power of labor and other gi-oups. Increased unem- 
ployment for some, higher per capita earnings to employed labor. 
Some change in direction in which purchasing power will be ex- 
pended, (c) Reduction in working hours with the same or a smaller 
force and the same time rates of pay. No change in selling price; 
shift in division of aggregate purchasing power, a larger proportion 
of it going to agents of production other than labor. Unemployment 
accompanies this shift. Some modification in direction of expendi- 
tures, (d) Reduction in selling prices. Initial lowering aggre- 
gate receipts and of amount disbursed to agents of production. Pos- 
sible initial unemployment. Release of buying power of consumers 
for purchase of more goods of the same type, or other goods (de- 
pending upon elasticities of demand for various products.) ^^ 

For the smooth operation of the economic system, which means also 
greater stability of employment, it is desirable that there should be as 
direct a connection as possible betAveen the purchasing power shifteil 
and the productive energy released. The most direct connection is 
established when the selling price of a commodity is reduced to tho 
full extent made possible by the increase in productivity and when 
the demand for the commodity is highly elastic. In ir-uch cases a 
large part of the purchasing power of consumers released by the 
reduction in price would find an outlet through an increased demand 
for the commodity involved. Increased production would result, with 

{)rompt reemployment of all or part of the productive energies re- 
eased by the initial increase in productive power. This is the situa- 
tion envisaged by the theorists of the "theory of compensation."' It 
presupposes "a completely frictionless economy, marked by free 

g rices, with wages and other elements of production costs completely 
exible, with labor and capital completely mobile.'' Under such a sys- 
tem, the disposition of the gains from increased productivity would be 
a matter of inditference, insofar as economic adjustments and con- 
tinuity of employment are concerned.'^ 

No such direct connection can be established today since our econ- 
omy is marked by frictions of many types — bv rigid prices, infiex- 
ible rates for service of many sorts, and innnobility of labor and capi- 
tal. As a result of such barriers to the free operation of the price system, 
purchasing power released in one segment of the economy niay exert 
a Tiegligible effect on displaced labor and idle capital in a remote 
siH'tion, within time limits which are of practical im{)ortance for 
ordinary human activities. In view of this fact, the manner in which 
the gains resulting from higher productivity are distributed is of 
utmost importance from an economic and social point of view. For 
whether these gains from higher productivity will remain merely 
potential or will result in realized benefits, that is, in larger quanti- 
ties of goods and in higher standards of living, will depend upon 

«Ibld.. pp. 437-438. 
^ Iblfl., DO *3n-44n 



CO'CENTRATIOX OF ECONOMIC POWE^ g5 

their distribution in such a way as to facilitate a more direct con- 
nection between enhanced purchasing power and productive energies 
released by new techniques. 

Under present-day conditions, the distribution of the gains from 
increased productivity takes place through a struggle — '"a three-cor- 
nered pulling and hauling contest among primary producers, agents 
of fabrication and consumers." ^* X full explanation of the reasons 
which determine the outcome of this struggle — and, thus, also the 
incidence of increasing productivity — cannot be given, but among 
the factors which have shaped the process since 1900 are the general 
rise and decline in prices before and after 1929, respectively, the 
influence of the World War, the relative strength and weakness of 
the bargaining position of primary producers and fabricators after 
the war, the stronger bargaining position of industrial labor as a 
result of immigration restriction, the creation of a seller's market 
between 1922 and 1929 as a result of new credit devices, lending 
abroad, etc. The developments since 1900 are sinnmed up as follows: 

In pie war years piimary producers ami consumers stood in positions of rela- 
tive advantage and reaped most of the benefits of rising productivity. The 
tide turned with the end of the war. Primary producers lost bargaining power; 
the trend of prices and special post-war circumstances contributed to strengthen 
the position of fabricators. Among consumers, primary producers were in a 
weak position. The buying power of other important consuming groups was 
artificially bolstered, so that competitive pressure on the demand side, toward 
lower prices, was greatly weakened.^ 

It is the combination of the particular incidence of changing 
productivity with the nature oi the economic frictions impeding 
adjustment to change that explains whether prosperity will be broad 
or spotty and whether or not luiemployment will persist. As already 
indicated, the most favorable condition for the absorption of in- 
creased productivity is the establishment of as close a connection as 
possible between enhanced purchasing power and released produc- 
tive energy. In an economy marked by frictions such as ours, it 
is important that the effects of changing productivity be dissemi- 
nated from as many points as possible. Such wide diffusion re- 
duces the influence of particular frictions and facilitates a more, 
prompt utilization of relea.>ied energies. On the contrary, when the 
gains of higher productivity are appropriated by restricted eccmomic 
groups, the connection between the larger purchasing power accru- 
ing to these groups and the released productive power becomes 
more remote, thus impeding the adjustments to the new technical 
situation. 

A situation in which the gains in productivity woukl be dispro- 
portionately reaped by particular groups, might create a prosperity 
limited to special groups. In such a situation, fabricational margins 
would be high, prices to consumers high, rewards to primary pro- 
ducers low, corporate earnings high, and security prices rising, labor 
earnings comparatively large, and yet. in view' of such distribution 
of productivity gains, volume of sales and production would remain 
low, in comparison with productive potentialities and the needs of 
consumers at large. Under such conditions, unemployment would 
pet-sist in large volume. Furthermore, under such conditions, it if 

" Ibid., p. 456. 
'5 Idem. 

277.151— 41— No. 22 6 



gg CONCENTRATION OF ECONOMIC POWER 

conceivable that portions of the increased income of the favored 
groups might never become effective in stimulating the productive 
energies released by the gain in productivity. The persistence of 
teclmological improvements might displace workers in one section of 
the economy while the frictions of the economic system would im- 
pede the diffusion of the augmented purchasing power of the favored 
group in other sections of the economy. The result would be "a 
semipermanent condition of concurrent prosperity among some eco- 
nomic groups, unemployment and persistently low returns to other 
groups." ^^ 

This is exactly the condition which characterized the post- 
war economic scene in the United States. Industrial displacement 
and technological unemployment were in evidence before 1929, and 
the maladjustments became more pronounced during the depres- 
sion from 1929 to 1933. While the recovery from 1933 to 1936 was 
fairly broad and price disparities were reduced, the gains from 
higher productivity (which increased during these years some 25 
percent) were reaped in the main by "fairly restricted groups" 
(manufacturing labor and the managers and owners of manufac- 
turing plants) ; and shared but slightly by primary producers and 
consumers. This fact combined with the growing economic fric- 
tions (inflexible prices, relatively fixed overhead charges, private 
control, governmental regulation, etc.) explains the partial character 
of the recovery since 1933, and the persistence of a large volume of 
unemployment.^^ 

Looking ahead. Dr. Mills sees a continuance of increasing produc- 
tivity and a persistence of economic frictions in the future. But 
he thinks that society as a whole could be made to benent trom 
technological change, if restrictions on the competitive forces were 
reduced as much as possible and if manufacturers would spread as 
widely as possible the gains from increased productivity through 
lower prices. 

In Bulletin 70 of the National Bureau of Economic Research,^^ Dr. 
Mills presents a statistical summary of the decline in employment 
opportunities in manufacturing industries between 1899 and 1936 and 
of the proportion of this decline which is attributable to technological 
factors. The method is essentially the same as that used in Economic 
Tendencies in the United States,^" except for minor revisions of in- 
dexes and the use of more complete data. The results of his analysis 
are summarized as follows : 

(1) Up to 1919 there was a steady increase in the proportion of the population 
of the United States engaged in manufacturing industries. From 1899 to 1919, 
though productivity per wage earner and per man-hour increased greatly, the 
total number of employees in manufacturing establishments more than doubled 
(increasing from 5,077,000 to 10,431,000). Average normal hours decreased from 
59.1 to 52.3 a week, but total employment in man-hours per week increased about 
70 percent. 

(2) From 1919 to 1929, despite an increase of 16 percent in the population 
of the United States and a notable expansion in manufacturing production, there 
was a drop in the total number of both wage-earners and employees as well as 
in total employment in man-hours per week. The total emi)loymont of wage 

"Ibid., p. 459. 

" Ibid., pp. 459-463. 

" Employment Opportunities in Manufacturing Industries of the United States, September 

"See pp. , supra. 



CONCENTRATION OF ECONOMIC POWeA 67 

earners iu manufacturing plants declined 5 percent. Dr. Mills finds the ex- 
planation of this change in part "in fundamental technical changes that have 
speeded up the process of mechanization." In less degree, he finds rhe explana- 
tion "in internal shifts in the structure of manufacturing — changes in the char- 
acter of products turned out and the relative importance of different industries 
in the part played by capital equipment in manufacturing operations." The 
technical changes may be measured roughly by over-all changes in industrial 
productivity. 

(3) There is no simple inverse relationship between changes in productivity 
and in employment. Comparing the several groups of industries classified on 
various bases, e. g., those producing goods for human consumption, those pro- 
ducing capital goods, construction materials, food and non-foods, durable, semi- 
durable, etc., Dr. Mills finds that between 1914 and 1929 the greatest increase in 
the number of wage earners employed was shown by industries producing capital 
equipment iu whicli at the same time voUime of production and estimated output 
lier worlier increased most rapidly. In other words, there was a concurrent 
gain in productivity, total output, and employment which was facilitated by 
favorable market conditions and other factors. 

(4) It was "roughly true" that between 1923 and 1929, expansions in output 
served to offset advancing productivity. The problems of readjustment during 
these years was thus largely a question of the "inter-industrial transference of 
labor, within manufacturing as a whole," and "the allocation to non-manufactur- 
ing industries of the man-power coming on the market each year as a result 
of population growth." 

(5) During the period of recession and recovery from 1929 to 1935, the number 
of salaried employees was reduced 20.9 percent and the number of wage earners 
16..5 percent. (In absolute numbers wage earners decreased by 1,460,000 while 
total employment in man-hours per week fell 35.7 percent, about 152,000,000 
]nan-hours per week). The decline in employment between 1929 and 1935 was 
due in part to the failure of manufacturing production to regain the pre-depres- 
sion level and in part to an advance in industrial productivity that enabled a 
s-maller working force to produce a given volume of output. The extent of the 
decline in employment due to advance in productivity in manufacturing indus- 
tries. Dr. Mills measures in accordance with Methods A and B suggested by Dr. 
Jerome."" These computations indicate that : 

(a) The 1929 level of output could have been produced in 1935. with man- 
hour output equal to that prevailing in 1935, with 91,292,000 man-hours less. 
With a working week of 4S.3 houi's (that of 1929), this would have meant 
the employment of 1,882,000 workers less than in 1929. This measures the 
potential loss in employment as" a result of advancing productivity, when no 
account is taken of changes in working hours or in demand. 

(h) Of the total decline of 152,425.000 man-hours in aggregate man-hours 
worked per week between 1929 and 1935, approximately 74,664,000 man-hours 
(or 49 percent) was directly associated with rising productivity. In other 
words, it took 74,664.000 man-hours per week less in 1935 than it would have 
taken in 1929, to produce the volume of manufactured goods turned out iu 1935. 

(c) With the 1935 working week of 37.2 hours and with the 1935 rate of 
man-hour output, the 1929 level of production would have required in 1935 
some 2(X),000 workers more than were actually employed in manufacturing 
industries in 1929 (9,022,000 workers as against 8,839.000). This means that 
the reduction in working hours between 1929 and 1935 more than counterbal- 
anced the loss of employment that might have resulted from rising productivity, 
if the sole effect of productivity gains had been the displacement of workers. 

(d) The record of employment changes from 1929 to 1935 shows the per- 
sistence of the tendencies noted between 1919 and 1929; that is, the wants of 
consumers of manufactured goods could be met, on the established level, with 
a steadily declining quantity of hunian 'labor. Labor time was being released 
as a restilt of steady advances in productivity. 

(6) Comparing the 15 gn^ups into which the manufac4^uring industries of 
the United States are classified by the Bureau of the Census, Dr. Mills finds 
that the heaviest decline in employment (as measured by number of workers) 
between 1929 and 1935 occurred in the non-consumption goods industries, that 
is, , in industries producing capital goods, construction goods, and producer 
supplies. Comparing changes in volume of prodtiction, productivity, number 
of wage earners and aggregate man-hours per week between 1929 and 1935, Dr. 

* See pp. -, supra. 



g§ CONCENTKATIOX OF ECONOMIC POWER 

Mills finds that employment, whether measured in man-hours or in number of 
wage earners employed, suffered smaller losses among consumption goods, in 
which productivity advanced substantially, than among capital goods and con- 
struction materials, in which the gain in productivity was smaller. 

(7) Comparing the actual decline in aggregate man-hours between 1929 and. 
1935 with the decline in number of man-hours required to produce the 1935 
volume of output. Dr. Mills computes that the loss of aggregate employment 
attributable to productivity changes constituted some 87 percent of the total 
employment loss in the industries producing human consumption goods, about 
15 and 10 percent in industries producing capital equipment and construction 
materials, respectively. In the latter two groups of industries, technological 
changes were thus a minor factor in the loss of employment. 

(8) Summing up Dr. Mills says: "Of the actual employment decline in man- 
hours in manufacturing industries between 1929 and 1935, slightly less than half 
appears to be attributable to changes in industrial productivity, the remainder 
is apparently attributable to the decline in production * * * r^i^^ pgj.. 
sistence of reduced industrial activity and the complex market conditions 
that underlie this reduction are in a causal sense more important than in- 
creased man-hour productivity as factors in the employment decline since 1929." 
He observes further : "The 8,839,000 wprkers in manufacturing industries in 
1929 could in 1935 have produced the 1929 output with a work week 10.3 hours 
shorter than in 1929. The 1929 work week averaged 48.3 hours. The reduction 
justified by advance in productivity if all the gain were to be realized in the 
form of leisure would have meant a workweek of 38 hours. The actual work- 
week in 1935 averaged 37.2 hours. The program of shorter hours represented 
in effect the allocation to leisure or enforced idleness of all the gains of 
advancing productivity plus something more." 

(9) The situation in 1938 seemed darker than in 1935. Clearly, we are faced 
with a serious problem of adaptation to changing conditions of production and 
employment in manufacturing industries made necessary by technological 

changes. "Something of a redirection of productive resources, particularly 
the annual additions to the working population, may be necessitated by the 
cumulative technical improvements that have been steadily increasing industrial 
productivity. This redirection, which may mean the reversal of a movement 
into manufacturing and mechanical occupations that has grown steadily during 
the whole cour.se of our national history, is perhaps at the heart of the 
problem. It involves issues transcending manufacturing industries alone." 

THE MEASUREMENT OF REEMPLOYMENT 
OPPORTUNITIES 

The persistence of a large volume of unemployment and the con- 
tinued discussion of technological unemployment led the Works 
Progress Administration to set up as one of its projects a large- 
scale study of the effects of technological changes on production 
and employment. Tliis study, known as the National Research 
Project, has been in progress since 1C36 to date under the direction 
of Mr. David Weintraub. 

It is impossible, within the limits of this volume, to undertake 
a detailed review of the entire research output of the N. R. P. This 
output has been very large and has made an important contribution to 
the store of factual data needed for an understanding of the industrial 
and economic effects of technological change. From the point of 
view of the present survey, however, what is of particular inter- 
est is the attempt made by the N. R. P. to measure changes in labor 
productivity resulting from technological improvements as a basis 
for estimating the possibilities of reemployment. From this point 
of view, two reports are of special interest, namely : The report on 
Unemployment and Increasing Productivity, published in June 1937; 
and the report on Production, Employment, and Productivity, pub- 
lished in May 1939. In considering these two reports, it is necessary to- 



CONCENTRATION OF ECONOMIC POWER 59 

act 
stii 
separately. 



stress the fact that they represent a part only of the N. R. P. contribu- 
tion to the study of the subject. The two reports will be considered 



UNEMPLOYMENT AND INCREASING PRODUCTIVITY 

The problem set by this study *'^ is to measure, for the period 1920 
through 1935. the over-all changes within the total economic order 
•of the quantity of labor required to produce a given amount of na- 
tional income. The authors are fully aware that a complete investi- 
gation of the effects of changing technology on the volume of 
employment and unemployment "would involve an analysis of the 
effects of changing prices of goods and services, of changing costs 
of capital and labor and * * * of changing demands for goods 
and services, and of a multitude of other factors which play an 
important part in determining the profitableness of employing work- 
ers." They are also aware that, "only such an economic analysis, 
dealing with the fundamental elements of our economic society," is 
capable of arriving at "conclusions concerning the underlying causes 
of unemployment in general and the particular type of unemploy- 
ment which might be attributed to changing industrial techniques." **' 

However, expressly disclaiming any effort to engage in such a com- 
plete analysis, the authors confine themselves to an examination of the 
available statistical information on the volume of production and em- 
ployment since 1920 in the light of the changes in output per man-year 
whicli took place during the period. The aim of .the study is thus 
limited to measuring the year-to-year changes in the ratio between the 
total national income and the total volume of employment. 

To measure annual changes in the productivity of the national 
economy as a whole, the authors proceed by devising indexes of overall 
production and oveiall employment. They make no attempt to 
measure the physical volume of production directly by the technique 
of attributing weights to individual series of index numbers represent- 
ing tlie outi:)ut of specified goods (steel, cement, coal, electric power, 
cotton cloth, etc.) and of specified services (e. g. ton-miles of freight 
hauled by railroads). Instead, the total volume of production in any 
year is taken as equal to the dollar value of national income produced — 
i. e., to the net value of goods and services for final consumption pro- 
duced through "efforts whose results appear on the market place of our 
economy." '^^ To eliminate the distortions arising from changes in the 
purchasing power of money (rising or falling price levels), the figures 
on national income" are deflated by use of an index purporting to 
represent "prices of finished goods." '^^ As the authors expressly indi- 
cate, the deflator used is one which gives "insufficient impor- 
tance * * * to consumers' services and 'nonessentials' whicli con- 



>" David Weintraub, assisted by Harold L. Posner. "Unemployment and Increasing Pro- 
ductivity." ch. V. in Technological Trends and National Policy, National Resources Com- 
mittee. June 1937, pp. 67-87. 

"= Ibid., p. 67. 

«3 Ibid., p. 68. 

** For 1920-29, the figures on national income are derived from "realized income" less 
"Imputed income" plus "business savings" as shown in Leven, Moulton and Warburton, 
America's Capacity to Consume. For 1929-35. the figures on "national income" are talsen 
from the estimates of the U. S. Department of Commerce as shown in Survey of Current 
Business. July 193G. e.\cluding income from \vorl< relief. The Broolfings series is spliced 
to the Commerce series in 1929. Ibid., p. 69. 

•= Ibid., p. 68. 



yQ CONCENTRATION OF ECONOMIC^ TOWEK 

stituted an increasing proportion of the national income produced 
over the period" '' (1920-35). 

As for the index of employment, it is restricted to the employment 
of "hired labor employed in the creation of this national product." '^• 
It would also be well to bear in mind, as the authors stress, that em- 
ployment and unemployment statistics are ^'rather fragmentary" for 
the period prior to 1930. The following main steps were taken in 
developing tlie index of employment : 

A. From the census figures on "gainfully occupied persons'' in 

1920 and in 1930 there were subtracted all enterprisers, all 
self-employed workers, and ail unpaid family workers on 
farms.*^* The results were taken to express, for 1920 and 
1930, the "national labor force" or total "supply of labor for 
hire." 

B. Taking account of the floAv of immigration, emigration, and 

farm-city movements, estimated figures on the total supply 
of labor for hire were interpolated for each of the inter- 
censai years between 1920 and 1930.^^ 

C. For the years 1931 through 1935, the total supply of labor for 

hire was estimated by adjusting the 1930 census figures for 
changes in the age distribution of the population, immigra- 
tion, and emigration, and farm-city movements. The total 
supply for these years therefore includes young persons 
"who under 'normal' conditions would have obtained their 
first employment experience, but who, during the period of 
widespread unemployment, may never have worked." '' 

D. From the figures on the total of the country's supply of labor 

for hire in each year there was subtracted 2.5 percent in 
order to deflate the total for workers "actually not available 
for work because of illness, vacations, voluntary transfers 
between jobs, labor disputes, and similar reasons for idle- 
ness." '^ The remainder is taken to represent "the country's 
total manpower available for liire," ''^ 

E. From the annual figures on total manpower available for hire 

estimates were derived of the annual number of "man-A'ears" 
of employment and unemployment.'^ These estimates make 
use of a large body of statistical data (developed by various 
governmental agencies) purporting to represent 3'ear-to-year 
fluctuations in the volume of employment in particular 
groups of industries.'* In order to adjust for Avorkers em- 

"" Idem. The index of the ""prices of finished goods"' used for adjusting price changes 
Is derived by giving weights of 1 and 9. respectively, to: (a) The co*t of capital goods; 
(b) the B. L. S. cost-of-living index. (Idem, footnote 7.) 

"' Ibid., p. 68. 

•^ Ibid., p. 69. This prcceduie is explained by stating that tlie report "conoerrs itself 
only with the effects of changing productivity on the employment opportunities of tlio^e 
who depend upon paid jobs for their livelihood and are, therefore, subject to unempli yment." 
(Idem.) 

"'■ Idem. 

"Ibid., pp. 69-70. 

On the other hand, no account was seemingly taken of possible contractions in the total 
supply of labor attributable to child labor restrictions, retirement and pension scheuies. 
limitations on the employment of women in industry, etc. 

" Ibid., p. 70. 

The deflator is explained on the ground that the 1930 census data on unemployment indi- 
cated that about 2.5 percent of all "workers" were "unusable"' for the reasons specified. 

'-P. 10, idem. 

'» Ibid., p. 71. 

''^The figurfs used are those developed by the U. S. Bureau of the Census, the U. S. 
Bureau of Minos, the U. S. Bureau of Labor Statistics the Interstate Commerce Commission, 
the Ohio State Department of Industrial Relations. 



CONCENTRATION OF ECONOMIC POWER 



71 



ployed only part-time in the day or Aveek and for workers 
employed only part of the year, the available data were re- 
duced to "full time equivalents" or "man-years" by utilizing 
the resuks of certain studies on the amount of time lost by 
part-time workers in particular years.'^ 
F. No attempt was made to translate the "man-year" of work, 
given the average hours of Avork that ^^revailed in particular 
years, into "man-hours." The purpose of the study was to 
consider the number of jobs available rather than the num- 
ber of man-hours required ; "* * * it was therefore con- 
sidered appropriate to use each year's prevailing hour con- 
tent as representing a man-year of work and to make the 
part-time adjustments with this flexible man-year concept 
in mind." '^ 

The main results of the report may be summarized as follows : 

1. In 1935 as compared with 1920, production rose 14 percent while 

employment fell 18 percent ; an increase of 39 percent in labor 
productivity or a decrease of 28 percent in unit labcn- require- 
ments. This means that the same amount of labor was 39 
percent more productive of national income in 1935 than in 
1920. Conversely, the same amount of national income re- 
quired 28 percent less labor for its production in 1935 than 
in 1920." 

2. Between 1920 and 1929. jiroduction expanded as much as 46 

percent, while employment increased only 16 percent. The 
productivity of a man-year therefore went u}) 26 percent ; 
conversely, the labor requirement for a unit of output went 
down 21 percent. All or most of the improAenient in pro- 
ductivity took place "during the depression of 1921 and the 
recovery of 1922.'* Thereafter, from 1922 through 1929, })ro- 
ductivity fluctuated Avithin narrow limits on a plateau from 
20 to 30 percent aboA'e the level of 1920, as shown below : 





Year 




1920 


= 100 




Product 


vity 


Unit labor re- 
quirement 1 


1922 


126 
120 
123 
127 
124 
123 
129 
126 


79 


1923 


84 




81 


1925 


79 


1926 


81 


1927 


81 


1928 


78 


1929 


79 







Ibid., tables, p. 72. 

3. Between 1929 and 1932 production contracted about 40 percent 
while employment decreased 37 percent, thereby result- 
ing in a slight decrease of productivity (5 percent) or 
increase of unit labor requirement (5 percent). From 1932 



" studies made in Columbus. Buffalo. Syracuse, and Louisville at various times fri>m 1920 
through 1936. (Ibid., p. 71, footnote H.) 
'"Ibid., p. 71. 
"Ibid., pp. 71-72. 
" Ibid., pp. 14, 72. 



72 CONCENTRATION OF ECONOMIC POAVER 

to 1935, however, production expanded 27 percent while em- 
ployment grew only 12 percent, thereby resulting in a 
considerable improvement of productivity (rise of 16 per- 
cent) or unit labor requirement (fall of 11 percent).''^ 
4. Comparing the net change between 1929 and 1935, the follow- 
ing results emerge : 

Percent 

The production of national income declined 21 

The employment of man-years declined 29 

The amount of national income produced, per man-year 

of employment, rose 10 

The amount of employment required, per unit of national 

income, fell 9 

Taken at face value, these figures suggest that it would not have 
been possible to return to the 1929 level of employment in 1935 until 
and unless the output of goods and services was 40 percent greater 
than the amount actually obtained; i. e. 10 percent greater than the 
output of 1929.*" Because of the growth of employable population 
over the period, it would not have been possible to return to the 1929 
level of unemployment in 1935 until and unless the output of goods 
and services was 55 percent greater than the amount actually obtained 
in 1935, or 20 percent greater than the output of 1929.^^ A great ex- 
pansion of output was therefore necessary in 1935 before employment 
could be raised or unemployment lowered to the levels of 1929. 

Recognizing the great practical difficulties of isolating the techno- 
logical factor in labor productivity, the authors of the report are 
careful to explain that the extent to which heightened labor pro- 
ductivity may have diminished employment opportunities in 1935 as 
compared with 1929 does not measure the extent of increased "techno- 
logical unemployment" over the period in question.*^ Throughout 
their report, in fact, the authors stress the practical impossibility of 
measuring the volume of technological unemployment and warn 
against the uncritical use of figures on improved labor productivity as 
expressive of technological changes to the exclusion of other factors. 
To quote but a few of their statements on this point : 

Except in very rare cases, the effect of strictly technical changes on employ- 
ment in a single industry or even a single plant cannot be isolated.*^ 

To measure the full effect of even a single technological change on displace- 
ment and absorption (of labor) would * * * necessitate the virtually im- 
possible task of tr:icing i*: through tlie innumerablo factors which bear on the 
total volume of production and employment. 

Making direct inquiry among employers and workers would not be feasible 
either, since frequently neither the worker who loses his job nor the employer 
who lays him off knows whether the lay-off is the result of technological 
improvement or not." 

" Idpm. 

«° Ibid., p. 78. 

" Id.m. 

'-■ "The (over-all) productivity ratios * • * can be regarded as indicative of the 
effects of teclmological change only in the broadest sense. These over-all productivity 
ratios (quantity output per unit of hired-labor time) reflect a variety of factors in addition 
to the mechanical imiHovements usually characterized as 'technological.' Thus productivity 
may change as a result of non-mechanical aids to labor, or managerial Improvements, or in 
response to varying degrees of utilization of productive capacity, or changes in the linurs 
of work, or any combination of these and other factors. On the other hand, technological 
improvements are frequently made without any resulting changes in the productivity ratio, 
although th( y may cause changes in the occupational requirements of the industiv directly 
concerned or of a related industry." (Ibid., p. 79.) 

" Idem. 

*"Ibid., pp. 80-81. 



OONCBMTRATION OF ECONOMIC POWER 



73 



No satisfactory method of measuring the effect of technological changes on 
employment has yet been evolved. The complexity of the interrelationships 
between industries and between productivity and production makes impossible 
an adequate quantitative description of the full effects of technological develop- 
ments.^ 

Disclaiming, theretore, any effort to measure technological unem- 
ployment, the authors present figures for the period 1920-35 on what 
they designate as "unrealized employment." The measure of unreal- 
ized employment is the difference between the total number of jobs 
available in any one year and the number which would have been re- 
quired for the production of that year's total output had the over-all 
productivity remained at the previous year's level. This procedure 
necessitates "treating the national economy as a single industry and 
measuring the net effects of changes in output and productivity." 
Fundamental to the procedure, as the authors recognize, is the assump- 
tion that changes in productivity are independent of changes in vokmie 
of production ; an assumption which they believe it safe to make when 
comparing successive years.^*^ 

Subject to this assumption, the available statistics are examined with 
a view to answering the question : "How much of any year's unemploy- 
ment may be ascribed to the difference between the number of jobs 
available that year and the number which would have been required 
for tlie ])roduction of that year's output had the productivity remained 
at the level of the year immediately preceding it?" ^' 

The results reached for the period 1920-35 may be summarized as 
follows: 





Man-years of "unrealized employ- 
ment" as percent of i— 




Man-years 
men 


of "unrealized employ- 
" as percent of '— 




Employ- 
ment in 
1920 


Each year's 
employ- 
ment 


Each year's 
unemploy- 
ment 


Employ- 
ment in 
1920 


Each year's 
employ- 
ment 


Each year's 
unemploy- 
ment 


1920 ... 


12 

(') 

3 
3 

v") 


11 

14 


(2) 

34 
49 

ID 


192H 


5 

(') 

2 

. •■' s 

3 


5 

(3) 

2 

(3) 

12 
(') 

4 


33 


1921 


1929 




1922 

1923.. 


1930 

-1931 

1932 


1924. ... 


3 1 19 
3 1 19 

(3) (3) 1 


(2) 


1925 


1933. 




1926.. 


1934.... 

1935 


5 


1927 


(3) 


b) 



' Ibid., table 9, p. 82. 

2 Not available. 

5 No "unrealized" employment; productivity declined or was unohanped. 

On the basis of these figures, the authors conclude that : "During 
the period surveyed, except for the period after 1929, when the sharp 
decline in output resulted in a tremendous increase in unemployment, 
'unrealized' employment constituted from one-fifth to one-half of 
the unemployed manpower in years when over-all productivity in- 
creased." ®® It is essential to stress that these figures on year-to-year 
movements of "unrealized" employment "do not constitute measures 
of the number of 'technologically unemployed' workers" in any 

"Ibid., p. 82. 

^ "If the periods chosen for comparison are relatively close, * * « say two succes- 
sive years, the error resulting from an assumption of independence between changes in pro- 
ductivity and the volume of production is held to a minimum." (Ibid., p. 81.) 

" Idem. 

"Ibid., p. 82. 



74 CONCENTRATION OF ECONOMIC POWER 

year.*** At most, these figures may be used to measure roughly the 
effect upon unemployment totals of heightened productivity, regard- 
less of the factors contributing to that heightened productivity. 

In concluding, the authors make it clear that their figures "can- 
not be used either to affirm or deny any particular theory" of the 
impact of technological change upon the disemployment or displace- 
ment of workers in modern industrial society. Their major, though 
limited, finding is that an "undetermined but substantial proportion 
of the unemployed in any single year (1920-35) probably consisted 
of workers who had been displaced from their jobs in one wa}- or 
another by the employment dislocations which accompany techno- 
logical progress." ^" 

As for the outlook in the immediate future of reemploying workers 
disemployed by the impact of technological improvement, the report 
is guardedly pessimistic : 

The outlook for the immediate future seems to be in the direction of further 
technological progress towuni a level of productivity substantially higher than 
that attained prior to 1929. The rate of advance, of course, differs in different 
industries, but since our economic system has not evinced an ability to make 
the necessary adjustments fast enough, it may be expected that the disloca- 
tions occasioned by technological progress will continue to present serious 
problems of industrial, economic, and social readjustments.'' 

PRODUCTION, EMPLOYMENT, AND PRODUCTIVITY IN MANUFACTURING 
INDUSTRY '*- 

This report also derives from the N. K. P.'s main purpose of examin- 
ing the relationship between changes in productivity and reemploy- 
ment opportunities. Instead, however, of considering the national 
economy as a whole, it takes up separately each of 59 manufacturing 
industries. For each of these industries the authors consider the data 
on production and emploj^ment during the post-war years with a view 
to deriving for individual industries useful indexes of labor produc- 
tivity and of unit labor requirements."^ The findings for the 59 indus- 
tries, taken together, are regarded as representative, within narrow 
margins, of the findings that would be obtained from a study of manu- 
facturing industries in their entirety. 

It is necessary to point out that the report is concerned with develop- 
ing proper measures of productivity, not with analyzing the causes 
underlying changes of productivity in manufacturing industry since 
1919. As the authors state : 

The report does not concern itself with reasons for the changes that have 
occurred since 1919 in production, employment, man-hours, and productivity for 
the .59 manufacturing industries surveyed. The movements of production and 
productivity may be due to a number of factors arising from the economics and 
techniques of an industry. For example, productivity may change as a result of 
changes in technology, plant lay-out, managerial technique, efficiency of individual 
or groups of workers, degree of utilization of capacity, raw materials, or the 
quality of the products manufactured ; productivity may also change as a conse- 



»» Idem. 

'* Iliifl., pp. sr,-.ST. 

«' Ibid., p. 87. 

»=This roport was prepared l)y Harry Magdoff. I. H. Siegel and M B. Davis; its oxact 
title is rrodnction, Kmplo.vnieiit, and I'loductivity in »',) Manufactmiim Industries. 1919- 
.SG. It consists of tliree parts pul)lislUHi in tlireo separate voluiiu's : Pt. I — Purpose, 
Methods, and Summary of Findin<j;s ; Pt. II — Indexes for individual Industries and Methods 
of Construclion ; Pt. Ill — ^Appendixes. The report was pnblislied in May 1039. 

"3 Ibid., pt. I, Preface. 



CONCENTRATION OF ECONOMIC POWER 



/tO 



►quence of the mortality of inefficient and old plants and the entrance of new and 
more efficient one.^ or as a consequence of the shift in production between plants 
producing at different productivity levels. An analysis of the effects of changes 
in productivity on employment opportunities and the relationship of changes in 
production to changes in productivity requires intensive studies of Individual 
industries, strategic sections of the economy, and the conditions and development 
■of the economy as a whole. The report * * * is * * * confined to the 
task of measurement (that is) primarily with the construction of measures which 
are pertinent to and useful in a study of production and productivity changes and 
their bearing on employment prospects." 

While limiting their aims, the authors relate their report to the 
general purposes of the N. R. P. inquiry. 

The central question concerning unemployment today is whether the current 
unemployment problem is of a different nature from the problem in the past. 
The question is, in part, one of the changing structure of the economy and the 
relationship of this factor to the prospects of future employment. Within this 
■context, the bearing of the changing prodnctivitij of labor on the total volume 
■of labor utilised is of particular interest. '' 

They go on to claim that the ratio of total output to total man-hours 
or man-years, despite admitted shortcomings, affords a useful measure 
of the impact of changing industrial techniques on employment oppor- 
tunities. They write: 

The mere computation of changes in a ratio of quantity produced to labor time 
consimied does not, of course, establish a significant economic relationship between 
production and employment. Measures of the type presented in this report do, 
however, permit the gaging of trends in employ m "Ait opportunities. Despite the 
fac.t that changes in the degree of plant capacity utilization are known to affect 
labor productivity, that the incidence of changes in industrial techniques differs 
according to the size of plants, or that computed changes in output per unit of 
labor may reflect changes in the relative quantities of different products, the con- 
■clusion which must be drawn after a careful examination of the data for 
a large number of manufacturing industries is that the observed trend in 
labor productivity reflects principally the effects of the continuous changes 
in industrial techniques on the amount of labor required per unit of out- 
put. The annual figures * * * must, of course, be approached with 
•caution ; not only the factors mentioned above, but many others besides techno- 
logical change may have influenced the level of any industry's productivity figure 
for a particular year. "* 

To measure productivity changes between 1919 and 1936, the 
authors selected 59 manufacturing industries for which adequate 
statistics of production and employment were available.^' Broadly 
speaking, an industry was defined as an aggregation of establish- 
ments producing either a single product (where comparable labor 
statistics Avere available for that product) or in most' cases, as an 
aggregation of establishments manufacturing "the smallest possible 
combination of products," for which comparable production and 
employment indexes could l>e constructed. 

The N. R. P. production indexes were built up in three steps: by 
individual manufacturing industries; by individual groups of indus- 
tries; by all industries combined. Ignoring the years for which 
interpolation was necessary and considering only the years for which 
primary data were available in the Census of Manufactures, in the 

" Ibid., p. 2. 

^Ibid., p. 3. [Italics supplied.) 

»«Ibid.. p. XIV. [Italics supplied.! 

"' Witii a few exceptions, the 59 manufacturing industries as defined in this study are 
identical with or equivalent to, corresponding industries as defined in the biennial Census 
of Manufactures. Ibid., pp. 21-24. 



75 CONCENTRATION OF ECONOMIC POWER 

Yearbook of Agriculture, in the reports of the Bureau of Mines, etc., 
the following procedure was followed: 

Individual industries: AVliere the industry produced some single 
uniform product — e. g., beet or cane sugar, cement, cigarettes, bee- 
hive or byproduct coke, etc. — production for the year was taken as 
equal to the physical quantity of output ^^ — pounds, tons, etc. — 
reported in the official statistics. In some instances moreover, several 
products Avere counted as a single product : e. g.. total gallons of ice 
cream plus other ices; total tonnage of pig-iron plus ferro-alloys. In 
the case of petroleum products, because no otlier means of computa- 
tion wasi practicable, production was taken as equal to input — i. e., 
the volume of crude petroleum refined. 

For the gi-eat majority of industries, however, the production of 
the industry could be determined only by adding together — properly 
weighted — ^the production of a great many diverse products. 

In every case where an industry manufactured a diversity of 
products, the problem arose of how to weigh the individual items so 
as to arrive at the total production for a given year. Although the 
output of each such item was always available in some quantitative 
measure — pounds, tons, gallons, yards, barrels, bales, cases, number — 
it was possible onl}' in a few cases to obtain figures on the relative 
labor requirements of the individual commodities making up the 
production composite of a given manufacturing industry.'-*^ Weights 
derived from value of output data had, therefore, to be used as the 
best available substitute for weights derived from labor requirement 
data; in most cases, the value Aveights were determined from the 
figures for 1929.^ 

The next step was to obtain a combined index of production for 
each of 13 industry groups.- With regard to every industry forming 
part of a given group, the production indexes had already been 
determined; it therefore merely remained to weight the individual 
indexes in order to obtain an overall index for each industrial group. 
By drawing upon official statistics from many sources, it was possible 
to obtain proper weights from data on wage earners employed and 
man-hours expended in 1929 and in other years for each of the 13 
groups. 

To arrive at, a production index for the 59 manufacturing indus- 
tries combined, the production indexes of 13 industry groups and of 
24 manufacturing industries not fcuniing part of a group ^ were 

'>^ Ideally, total output ; by necessity in most cases, however, output for sale. 

''> National Research Proyct, Production. Employment, and Productivity in 59 Manu- 
facturing Industries, 1919-36, Part I, p. 37. 

> Ibid., pp. 38-39. 

2(1) Bread and other bakery products group (biscuit and crackers; b;ikery products 
other than biscuit and crackers) ; (2) canning and preserving group (canned and preserved 
fruits and vegetables; canned and cured fish); (3) coke group (bethive coke; by-proluct 
coke); (4) glass group (window glass; plate glass; glass containers; pressed and blown 
ware) ; (5) iron and steel group (blast furnaces; steel works and rolling mills) ; (0) knit 
goods group (hosiery, underwear, outerwear, knit cloth) ; (7) leather group (sole and hamt-ss 
leather ; side and upl^olsterj leather ; calfskin ; kid leather ; sheep and miscell'ncou.^ leuitier) ; 
(8) lumber and timber products group, (logfiing camps) ; sawmills and saw-plane mills) ; 
(9j nonferrous metals group (primary smelters and refineries ; secon<iary smelters and 
refineries; alloyers. rolling mills and foundries) ; (10) paper and pulp group (paper; pulp) ; 

(11) rubber products group (rubber tires and inner tubes; other rubber g.tods ; (I'J) tobacco 
products group (cigars; cigarettes; chewing and smoking tobacco and snuff) ; (13) woolen 
and worsted goods group (woolen goods; worsted goods). 

^(1) Agrieultiiral implements; (2) beet sugar; (3) boots and shoes; (4) eane sugar 
refining; (5) cement; (6) chemicals; (7) clay products (other than pottery) and non-clay 
refractories; (8) confectionery; (9) cotton goods; (10) electric lamps; (11) fertilizers; 

(12) flour and other grain-mill products; (13) furniture; (14) ice cream; (15) manufac- 
tured gas; (16) manufactured ice; (17) motor vehicles; (18) newspaper and periodical 
printing and pul)li8hing ; (19) paints and varnishes ; (20) petroleiim refining; (21) planins- 
mill products; (22) rayon; (23) ellk and rayon goods; (24) slaughtering and meat packing. 



CONCEMKATIUX OF ECONOMIC POWER 77 

weighted by substantially the same procedure as that used to de- 
termine the production index for an individual group of industries. 

In computmg employment indexes which could be applied agamst 
the production indexes for particular manufacturing industries so as 
to yield productivity ratios, it was decided to restrict the count of 
employment to wage earners. All salaried employees were excluded, 
not only "because they are not so directly engaged in the processes 
of production as are wage earners" but also because "no statistical 
series reflecting their specific output were available." * 

It must be stressed that every employment index covered by the 
N. K. P. series is restricted to wage earners as distinguished from 
salaried employees. The productivity ratios do not, therefore, meas- 
ure the productivity of the entire labor force. What they do meas- 
ure, strictly speaking, is the productivity of manual w^age earners, 
i. e., the ratio of all the man-hours or man-years of manual labor in 
manufactui-ing plants to the entire output produced by the joint labor 
of wage earners and salaried employees. 

To obtain the count of wage earners, the authors of the study re- 
sorted to the Census of Manufactures, which in odd-numbered years — 
1919, 1921 * * * 1935 — gives figures, from which annual aver- 
ages may be derived on employment in each month of the year for 
particular manufacturing industries. These figures were supple- 
mented — for interpolation and adjustment purposes — by the con- 
tinuous monthly employment indexes published by the Bureau of 
Labor Statistics and by the National Industrial Conference Board as 
well as by figures obtained from the Bureau of Mines and from 
State publications.^ 

Absence of adequate direct statistics on man-hours expended made 
it necessary to estimate the annual man-hour totals in the employ- 
ment series for all but a handful of industries.'' The procedure was 
to multiply the average annual employment of w^age earners in a 
given industry by the average actual hours of work (not the "pre- 
vailing" work-week or "fulltime" hours). The necessary figures 
were obtained from the compilations of the Bureau of Labor Statis- 
tics and the National Industrial Conference Board. ^ 

The representative character of the manufacturing industries 
selected for analysis by the N. R. P. is particularly important in view 
of the conclusions that might be drawn from the overall indexes of 
production, employment, and productivity for the 59 manufacturing 
industries combined. In 1929, the 59 industries selected for anlysis 
employed more than 4.5 million wage earners, or about 51 percent of 
the total wage earners reported by the Census of Manufactures ; they 
manufactured products to a value of almost $40,000,000,000 or about 
56 percent of the total value of the products manufactured by all re- 
porting industries. Throughout the period 1919-36, the N. R. P. 
sample consistently accounted for about half (49 to 54 percent) of 
all the wage earners employed in manufacturing as reported by the 
Census of Manufactures.^ 



* National Research Project, Production, Employment, and Productivity in 59 Manufac- 
turing Industries, 1919-36, pt. I, p. 42. 

Mbid., pp. 43-^7. 

8 For three industries, however — beehive coke, by-product cokle and electric lamps — it was 
possible to compute '1,9 indexes directly from man-hour totals as oflBeially compiled. 
(Ibid., p. 49.) 

' Ibid., pp. 51-56. 

"Ibid., pp. 60-61. 



78 CONCENTRATION OF ECONOMIC POWER 

On the other hand, the N. E. P. sample suffers from an admitted 
defect in that it fails to give adequate representation to capital goods 
industries as distinct from consumer goods industries. This leads 
the authors to observe with regard to the overall indexes of produc- 
tion, employment, and productivity for the 59 manufacturing indus- 
tries combined : 

* * * such indexes are more likely to indicate the direction of the year-to-year- 
percentage changes for all manufacturing than the correct magniture of these 
changes.* 

The two census groups which are under-represented most seriously in 
the N, R. P. sample are (1) machinery other than transportation equip- 
ment, and (2) railroad repair shops. Railroad repair shops are ex- 
cluded completely from the sample; machinery other than transporta- 
tion equipment is covered to tho extent of not more than 4 percent of 
wage earners or value of products (1929). Furthermore, the import- 
ant industry of shipbuilding is excluded from the industries covered 
in the census group — transportation equipment, air, land, and water.^'* 

Changes in Productivity, Output, and Employment for All Industries 
Combined, X919-SG. 

Combining all 59 manufacturing industries, the authors, calculate 
two series of productivity indexes : (A) which describes changes in the 
amount of labor required each year in order to obtain a fixed composite 
of production — that of 1929; and (B) which describes the changes 
since 1929 in the amount of labor required each year in order to obtain 
a changing composite of production — that of the current year. These 
series are shown in the tables which follow : 

From table I (fixed composite of production), three phases in the 
movement of productivity may be distinguished : 1919-29, 1929-32,. 
and 1932^36. 

1919-28 : The index for output per wage earner rose 55 percent and 
that for output per man-hour 57 percent. 

In other words, with the unit labor requirements of 1919 approximately 55 per- 
cent more wage earners and 57 percent more man-hours would have been neces- 
sary to produce the 1929 composite of products. 

In fact, however, while production increased, employment and man- 
hours remained below the 1919 level throughout most of this period. 
Thus— 

* ♦ * a major characteristic of the manufacturing industries 'i * ♦ * ^as 
the production of an increasing amount of goods with a relatively stable or even . 
declining volume of employment * ♦ *." The average hours worked per week 
changed little during the period under consideration * * ♦ [.so that] * * ♦ 
the indexes of employment and man-hours are similar imtil 1929." 



•Ibid., pp. 61-62 [italics supplied.] 
"Ibid., pp. 61-63. 

" other than the capital goods Industries, it should be understood, 
■^i Employment of wage earners as distinct from salaried employees. 

" National Uesearch Project : Production, Employment, .uul Productivity in 50 Manutao- 
turlug Industries, 1919-36, Part I, p. 64. 



CONCENTRATION OF ECONOMIC POWER 79 

Table I. — Combined indexes of production, eiuployment, mnn-hours. and produc- 
tivity (n-ith base-year man-hour n-eights), for N. R. P. ynanufacturing indus- 
tries: 1919-36 ' 

[1929 = 100] 



Produc 
tion 2 



Emplcy- 
ment 



Man- 
hnurs 



Output 1 



Wage 
earner 



Man- 
hour 



1920 
1921 
1922 
1923 

1924 
1925 
1926 
1927 
1928 

1929 
1930 
1931 
1932 

1933 
1934 
1935 
1936 



90.2 
88.5 
93.0 

100.0 
80.8 
68.0 
53.4 

62.1 
67.1 
77.9 
80.3 



100.0 
86.3 
73.2 
64.4 

72.6 
84.0 
86.9 
90.9 



100.5 
75.9 
91.5 

100.7 

92.1 
97.5 
98.7 
95.7 



65. 2> 
51.8 

57.7 
60.5 



81.6 
87.7 
91.3 
92.6 



92.9 
82.9 

85.5 
79.9 



63. 5 
67.0 
71.5 
76.9 
81.1 

84.0 
88.3 
91.4 
92.5 



100.0 
100 9 
104.3 
103.1 

107.6 
110.9 
118.6 
121.8 



> For many industries data are lacking for some of the years of the period 1919-36. Hence the index num- 
bers were constructed by chaining links for identical industries. 
2 Eq:nvalent to an inde.x whose component indexes were weighted with changing man-hour weights. 

Source: Works Progress Administration. National Research Project, Production, Employment, and 
Productivity in 59 Manufacturing Industrie". 1919-36, Part I, table XI, p. 65. 



Between 1929 and 1932 production declined 47 percent, the number of wage 
earners 36 percent, and the number of man-hours 48 percent. In other words, 
output per wage earner declined as the averaged hours worked dropped, but 
output per man-hour remained fairly stable and even increased slightly. By 
1936, production had recovered to 89 percent of the 1929 level and employment 
to 91 percent, but the index of man-hours was still 27 units below. Output per 
wage earner was only slightly below 1929, but output per man-hour had reached 
a point 22 percent above 1929 [a divergence resulting from the decline of 
average hours per week] " 



"Ibid., pp. 64-66. 



80 



CONCENTRATION OF ECONOMIC POWER 



Table II. — Combined indexes of production,, employment, man-hours, and pro- 
ductivity (with Changing man-hour weights) for N. R. P. manufacturing 
industries. 1919-36 

[1929 = 100] 



1919. 
1920 
1921 
1922 
1923 

1921 
1925 
1926 
1927 
1928 

1929 
1930 
1931 
1932 

1933 
1934 
1935 



Produc- 
tiou 2 



72.4 
58.6 
73.3 
83.4 

78.8 
87.1 
90.7 
89.2 
93.3 

100.0 
81.0 
69.3 
55.8 



Employ- 
ment 



100.5 
78.8 
91.7 

100.7 



95,6 
96.3 

100.0 
86.3 
73.2 
64.4 

72.6 



Man- 
hours 



100.5 
75.9 
91.5 

100.7 

92.1 
97.5 

95^7 
96.0 



57.7 
60.5 
05.7 
73.3 



Output per- 



Wage Man- 

earner hour 



70.1 
72.0 
74.4 
79.9 

82.8 

83.1 
88.7 
91.8 
93.3 



100. 
93.9 
94.7 
86.6 

90.2 
81.8 
92.5 



69.1 
72.0 
77.2 
80.1 
82.8 

85.6 
89.3 
91.9 
93.2 
97.2 

100. 
101.1 
106.3 
107.7 

113.5 
113.6 
122.4 
123.9 



' See table I, footnote (a). 

' Equivalent to an index whose component indexes were weighted with base-year man-hour weights. 

Source: Works Progress Administration, National Research Project. Production, Employment and 
Productivity in 59 Manufacturing Industries, 1919-36, Part I, table XIII, p. 67. 

As may be seen by comparing tables I and II, the movement of the 
productivity index for a fixed composite of production does not fully 
correspond to the movement of the productivity index for a chanoin<x 
composite of production. The extent of this divergence may be made 
clear by putting the two indexes side by side as in table III below 
where A gives the indexes for a production composite fixed as of the 
base year 1929; and B gives the indexes for a production composite 
changing as of each year. 



Table III. 



-Indexes of productivity, 1919-36 

[1929 = 100] 



Year 


Output per 
wage earner 


Output per 
man-hour 


Year 


Output per 
wage earner 


Output per 
man-hour 




A 


B 


A 


B 


A 


B 


A 


B 


1919 - 


64.4 
67.0 
68.9 
76.8 
81.1 
81.6 
87.7 
91.3 
92.6 


70.1 
72.0 
74.4 
79.9 
82.8 
83.1 
88.7 
91.8 
93.3 


63.5 
67.0 
71.5 
76.9 
81.1 
84.0 
88.3 
91.4 
92.5 


69 1 
72.0 
77.2 
80.1 
82.8 
SX6 
89.3 
91.9 
93.2 


1928 


96.6 
100.0 
93.6 
92.9 

82.9 
85.5 
79.9 

9S;2 


96.9 
100.0 
93.9 
94.7 
86.0 
90.2 
81.8 
92.5 
99.9 


96.9 
100.0 

100. U 

101. 3 
103.1 
107. 6 
110.9: 
118.6 
121.8 


97.2 


1920 


1929 


100.0 


1921 


1930 


101 1 




1931 






1932 .- 




1924 


1933 

1934 


113.5 




113.6 


1926 


1935 

1936 -- 


122.4 


1927 


123 9 













Source: Tables I and II, supra. 



The authors explain the divergence (of magnitude, not direction) 
for 1919 as compared with 1929 and for 1932 or 1936 as compared 
with 1929, as follows: 



CONCENTRATION OF ECONOMIC POWEI^ gl 

1919 compared with 1929 : 

* * * in geueval, production advanced at higher rates in industries regis- 
tering the greater gains in productivity. In other words, those industries which 
experienced a more rapid rise in production received relatively more weight in 
the 1929 composite than in the 1919 composite; since it was these industries 
whicli had the more rapid rise in output per man hour, * * * [which] 
were weighted more heavily in the fixed-weight index with 1929 weights than in 
the chauging-weight index with 1919 weights. Hence the index with 1929 weights 
shows a higher percentage increase in output per man-hour from 1919 to 1920 
[about 59 percent] than does the index with changing weights [about 45 per- 
cent] .'' 

1932 or 1936 compared with 1929 : 

Output per man-hour for the fixed composite rose 3 percent from 1929 to 
1932 ; it rose 22 percent from 1929 to 1936. Tlie corresponding increases for the 
changing composite were 8 and 24 percent, respectively. The dilference * * * 
for the period 1929-32 is due to the association, in general, of declining output 
per man-hour in industries with the sharper declines in production and increas- 
ing output per man-hour in industries with less severe reductions in output. 
While not true in every instance, the durable-goods industries not only suffered 
sharper declines in production between 1929 and 1932 but also were characterized 
by fairly stable or even declining output per man-hour. On the other hand, pro- 
duction in the non-durable-goods industries, in general, decreased less while output 
per man-hour rose. Since the indexes of output per man-hour for the latter 
industries were weighted more heavily in the production composite following 
1929 than in 1929, the productivity measure with changing weights is higher 
than the index with fixed weights in 1932.'* 

On the other hand — 

estimates of the man-hour requirements based on either index would not differ 
greatly after 1929. * * * By 1936, the difference was only two points (or 
less than 2 percent)." 

Changes in Production, Employment^ and Productivity in Individu<d 
Industries, 1919-1936. 

The authors then proceed to consider the average annual rates of 
change in production, employment, and productivity for individual 
manufacturing industries.^* Making separate computations for vari- 
ous periods regarded as particularly significant, they present and dis- 
cuss a series of tables which indicate {a) the number of industries 
experiencing specified average annual rates of change in production, 
employment, man-hours, output per wage earner, and output per man- 
hour; and (&) the percentage of all manufacturing wage earners em- 
ployed in 1929 who were employed by industries experiencing speci- 
fied average annual rates of change in production, employment, man 
hours, output per wage earner and output per man-hour. 

With regard to the period 1919 to 1929, for example, it is found 
that the following percentages of all wage earners employed in 59 
manufacturing industries in 1929. were employed by manufacturing 
industries where output per wage earner and output per man-hour 

«H)id., pp. 66-67. 

"Ibi(t., pp. 67-68. 

"Ibid., p. 69. 

" Tlip authors again point out hero that their study is limited to a statistical analysis 
only. They state: "Study of the differences between" the individual parts of the total is 
necessary for an inquiry into the relationships between changes in productivity, in terhnolosy, 
and in economic factors ; it also serves to facilitate estimates of future eniploymenr oppor- 
tunities. An intensive study of these factors would require a detailed analysis of suategic 
classes of industries in their relation to the changing structure of the economy and their 
behavior under varying economic conditions, t^iiah studies were outside the scope of this 
report, but a statistical analysis of the movements of production, employment, and iMnduc- 
tivity of the individual industries constitutes a necessary preliminary step." Hbid.. p. 09.) 
[Italics supplied.] 

277551— 41— -No. 22 7 



82 



CONCENTRATION OF ECONOMIC POWER 



underwent specified average annual rates of increase from 1919 to 
1929: 



Productivity increasing at average annual rate of not less than— 


Percentage distribution of 192» 
employ men t— s p e c i f i e d 
increase of— 




Output per 
wage-earner 


Output per 
man-hour 




Percent 


Percent 




2.i 

15.1 

26.9 

61.5 

■99.7 


... 






5.1 percent 


36 3 




62.3 




»92.3 







' The balance of the wage earners— 0.3 percent— were employed in manufacturing industries where output 
per wage earner declined from 0.1 to 2.5 percent annually. 

2 The balance of the wage earners— 7.7 percent— were employed in manufacturing industries where output 
per man-hour declined from 0.1 to 2.5 percent annually. 

Source: Works Progress Administration, National Research Project, Production, Employment, and 
Productivity in 59 Manufacturing Industries, 1919-36, pt. I, table XV, p. 73. 

Witli regard to the period 1929-35, in contrast, it is found that the 
following percentages of all wage earners employed in 59 manufac- 
turing industries in 1929 were employed by manufacturing industries 
where output per Avage earner and output per man-hour underw^ent 
sjDecified annual average rates of increase or decrease from 1929 to 1935 : 





Percentage distribution of 1929 
employment— specified in- 
crease or decrease of— 




Output per 
wage-earner 


Output per 
man-hour 




Percent 


Percent 

0.9- 


10 1 percent 




2 3 


7 6 percent 


0.9 
2.3 
4.6 
28.1 

39.1 
58.0 
71.8 
28.1 
71.8 


5 5 






2 6 percent 


55 * 




92.1 


Productivity decreasing at annual average rate of not more than: . 

2 5 percent 


6 2 






7.5 percent ... . 


7 9 


Industries of increasing productivity 


92 1 


Industries of decreasing productivity 


7.9 






Total 


199.9 


100 0^ 







1 The balance of the wage earners— 0.1 percent— were employed in manufacturing industries where output 
per wage earner declined more than 10.0 percent annually. 

Source: Works Progress Administration, National Research Project, Production, Employment, and 
Productivity in 59 Manufacturing Industries, 1919-36, pt. I, table XVI, p. 75. 

What these figures bring out is that: (1) Output per man-hour 
was increasing steadily for practically all wage earners employed in 
manufacturing industries between 1929 and 1935 as well as between 
1919 and 1929, although at somewhat higher annual rates during the 
the earlier period than during the later period; (2) output per wage 
earner increased for virtually all wage earners employed in manufac- 
turing industries between 1919 ancT 1929 at annual rates not far 
below the rates of increase for output per man-hour. (This parallel- 
ism retlects almost unchanged weekly working hours throughout the 
period) ; (3) output per wage earner decreased between 1929 and 1935 



CONCENTRATION OF ECONOMIC POWE^ 83 

in manufacturing industries employing about 70 percent of the wage 
earners in 1929 and increased only in manufacturing industries employ- 
ing about 30 percent of the Avage earners. (This divergence from out- 
put per man-iiour reflects considerable shortening of the work- week.) 
In other words, what individual manufacturing industries lost between 
1929 and 1935 in their power to employ labor (because man-hour out- 
put rose) was entirely offset in the bulk of those industries by the wide- 
spread movement toward a shorter work-week. 

General' Conclusions and Comnients. 

The most significant bread result reached by the study is that in 1936 
as compared with 1929, postulating the 1929 composite of production in 
each year, output per man-hour in 59 manufacturing industries was 
about 22 percent greater, output per wage earner about 2 percent 
smaller.^'' Translating into unit labor requirements, this means that if 
the output of 1929 had been literally reproduced in 1936 (the same com- 
modities in the same amounts) but using the industrial techniques of 
1936, instead of those of 1929 — then {a) for each 100 man-hours ex- 
pended in 1929, it would have been necessary to expend not more than 
82 man-hours in 1936; but (b) for each 100 wage earners employed in 
1929, it would have been necessary to employ almost 102 wage earners in 
1936. In terms of the changed production composite, however, that of 
1936. it would have taken 100 man-hours in 1929 to reproduce the amount 
and kind of output that 82 man-hours were actually producing in 1936; 
but about an equal number of workers would have been employed in 
both years.-" 

If, "therefore, the findings for the 59 industries may be regarded as 
affording a true measure of how the reemployment potential of man- 
ufacturing industry was affected between 1929 and 1936, it follows that 
the shortening of the work-week added just about enough to tliat poten- 
tial to offset what increased labor productivity took aAvay — no more, no 
less. As for the innnediate future, the authors of the report are not very 
optimistic regarding the possibilities of liquidating in manufacturing 
industries unemployment on a large scale. They state: 

The probability for the future seems to be that there will be less rather than 
more employment in manufacturing- industries. Should manufacturing industries' 
reach and sustain the 1929 level of output within the next few years, the average 
number of wage earners required will not far exceed, if at all, the average uumber 
employed in 1929, despite the 20-percent reduction in the average number of Jiours 
worked per week that occurred between 1929 and 1936. In the longer run. since 
output per man-hour seems certain to increase further, manufacturing employment 
will probably be below that of 1929. Only a great increase in production or a 
marked decline in hours worked per week could bring manufacturing employment 
to levels appreciably higher than 1929. It is well to recall, however, that even the 
50-percent increase in output between 1919 and 1929 left the average number of 
wage earners in manufacturing almost stationary.-' 

As has been repeatedly indicated here, the report siunmarized above 
represents a major contribution to the statistical analysis of employ- 
ment ciiid unemployment. The indexes developed by the authors for 
59 manufacturing industries from 1919 to 1936 indicate clearly a con- 
tinuous rise in labor productivity of such magnitude as to affect the 
capacity of manufacturing industry to employ the whole of the labor 
force attached to it or that might otherAvise seek employment in it. 

" Sre table I. supra. 
^ See table li, supra. 

-1 Works Trogress Administration, National Research Project, Production, Employment, 
and Productivity in 59 Manufacturing Industries, 1919-36, Part I, p. S2. 



PART II 

TECHNOLOGY AND ECONOMIC BALANCE 

A study of technological advance and the 
compensatory forces of reduction in work- 
ing hours, the creation of new industries, 
and the reduction of prices. 



85 



TECHNOLOGY AND ECONOMIC BALANCE 
INTRODUCTION 

Part I of this report analyzed technology as to its importance in 
the history of economic thought. In this part attention is directed to 
the present development of modern technology and the economic 
problems it creates. 

Technology, Which may be defined as the application of the analyt- 
ical methocls of science to the industrial arts, creates a primary 
economic problem through its displacement of human effort. 

Three aspects of this labor-displacing problem are examined here : 

1. The measurement of the change in labor productivity; 

2. The types of labor-saving techniques which have brought about 

the change ; and 

3. The effects of the change upon labor and the economy as a 

whole. 

Presumably, there are certain compensatory forces inherent in the 
present economic order which operate more or less automatically to 
offset the labor-displacing effects of technology. Principal among 
them are the reduction of hours (without an accompanying decline 
in wages), the development of new industries, and the reduction of 
prices. 

A balance is theoretically supposed to exist between the advance 
in labor productivity and these compensatory forces. If labor-dis- 
placement should prove greater than these compensatory forces, the 
scale would tip toward greater unemployment, less use of our economic 
resources and increased social distress. When such unbalance per- 
sists, the most far-reaching consequences emerge. 

An investigation of the social and political consequences of such 
unbalance is obviously beyond the scope of a report limited, as is this, 
to economic analysis. But it is interesting to note the existence of 
this same economic phenomenon of unbalance in Germany. Mani- 
festly, technological displacement was not the sole or even "the prime 
€ause of social and political upheavals there nor do such changes inevi- 
tably follow the persistence of technological unemployment. There 
can be little doubt, however, that a large and apparently growing body 
of unemployed in pre-Hitler Germany furnished a fertile field for 
social change and that much of the unemployment was due to techno- 
logical displacement of labor.^ 

The persistence of mass unemployment in the United States, despite 
marked economic recovery, and the growing emphasis on technological 
advance should cause grave concern to all thoughtful persons who seek 
the preservation of democracy. 

' Cf. International Labour Office, The Social Aspects of Rationalisation, Geneva. 1931. 
pp. 245-50 ; and Three Sources of Unemployment, Geneva, 1035, pp. 7S-100. 

87 



CHAPTER I 
THE CHANGE IN LABOR PRODUCTIVITY 

THE EXTENT OF THE CHANGE 

A first purpose of this part of the s'^ ady is to measure the extent of 
the increase in labor productivity. That less labor is required from 
year to year to produce a given amount of goods is widely observed, 
but appraisal of the extent of the change depends upon accurate 
measurement. 

Output per man-hour is the most accurate measure of labor pro- 
ductivity because reductions in hours or changes in price do not affect 
its validity. Labor is regarded simply as the number of hours worked.^ 
Since ovtputper man-hour is a quotient of the indexes of production 
and of man-hours, the validity of the labor productivity series depends 
entirely upon the validity of the indexes used as divisor and dividend. 

It has been possible to obtain comj^arable production and man-hour 
series in manufacturing, steam railroads, bituminous coal mining, and 
anthracite mining from 1923 to date and, except in the case of steam 
railroads, for the census years 1909, 1914, and 1919.^ The trend of 
output per man-hour in each of these four segments of the economy is 
compared with production in chart I, table 1. 

Labor productivity in these four fields has made striking advances, 
reaching an all-time high in 1939. It has been relatively unaffected by 
the major cyclical downturns and, except for a few brief interruptions, 
has steadily increased. The principal variation in the productivity 
trend has been in its rate of advance, and in three of the fields the rate 
has greatly accelerated in the last decade. 

The changes in production and productivity occurring between the 
years 1923-29 and 1929-39 in the same four segments of the economy 
are compared in chart II, table 2. These periods are compared to 
show the relative changes in production and productivity in the pros- 
perous twenties (1923 being the first comparatively normal yesLV fol- 
lowing the post-war depression) and in the past decade .of depression 
and recovery. 



^ For a discussion of various statistical techniques used to indicate the change in labor 
productivity, see appendix A'. 

2 poj. their sources, see appendix B. 

89 



90 



CONCENTRATION OF ECONOMIC POWER 



Table 1. — Indexes of production and productivity, 1909-39 

[1923 = 100] 



1909 
1914 
1919 
1923 
1924 
1925 
1926 
1927 
1928 
1929 
1930 
1931 
1932 
1933 
1934 
1935 
1936 
1937 



Manufacturing 



Produc- 
tion 



56.5 
66.5 
79.3 
100.0 
94.3 
106.5 
112.5 
113.3 
121.2 
130.1 
107.6 
93.7 
71.1 
81.7 



107.7 
127.6 
134.4 
103.6 
129.5 



Output 
per man- 
hour 



66.2 
76.4 
76.4 
100,0 
105.8 
113.2 
116.9 
120.8 
129.5 
131.9 
131.6 
141.3 
137.7 
144.8 
147.9 
158.8 
161.8 
157.5 
159.8 
174.5 



Steam railroads 



Produc- 
tion 



Output 
per man- 
hour 



Bituminous coal 
mining 



Produc- 
tion 



67.2 
74.9 
82.5 

100.0 
85.6 
92.1 

101.6 
91.7 
88.7 
94.7 
82.8 
67.7 
54.8 
59.1 
63.6 
65.9 
77.8 
78.9 
60.6 



Output 
per man- 
hour 



70.1 
77.8 
85.8 
100.0 
101.7 
101.0 
100.3 
101.6 
105.3 
108.1 
112.8 
118.0 
115.9 
110.0 
111.9 
115.4 
121.5 
124.8 
130.4 
142.1 



Anthracite 
mining 



Produc- 
tion 



97.3 
94.4 
100.0 
94.2 

90!5 
85.9 
80.7 
79.1 
74.3 
63.9 
53.4 
53.0 
CI. 3 
55.9 
58.5 
55.6 
49.4 
54.4 



Output 
per man- 
hour 



100.0 
94.4 



98.1 
96.4 
94.0 
100.2 
115.0 
126.4 
118.5 
121.3 
135.1 
142.8 
106.1 
172. G 



Source: Witt Bowden, "Wages, Hours and Productivity of Industrial Labor, 1909-39", U. S. Bureau of 
Labor Statistics, Monthly Labor Review, September 1940. Production figures computed by Witt Bowden 
for derivation of output per man-hour indexes. Original data for manufacturing from the National 
Bureau of Economic Research, Federal Reserve Board, and the U. S. Bureau of Labor Statistics; for Steam 
Railroads from the Interstate Commerce Commission; and for Bituminous and Anthracite Mining from the 
U. S. Bureau of Mines and the U. S. Bureau of Labor Statistics. 



Table 2. — Percent change in production and la'bor productivity, 1923- 

1929-39 



and 



Period 


Produc- 
tion 


Man- 
hour 
output 


Period 


Produc- 
tion 


Man- 
hour 
output 


MANUFACTURING 


1923-29 


-1-30.1 


-f3L9 


1929-39 


-0.5 


+32.3 








STEAM RAILROADS 


1923-29 


+3.3 


+18.2 


1929-39 


-25.8 


+31.0 








BITUMINOUS COAL MINING 


1923-29 _ 


-5. 3 1 -f8. 1 


1929-39 


-27.0 


+31.5 








ANTHRACITE MINII'G 


1923-29 . 


-20.9 


-" 




-31.2 


+79.0 









Source: Table 1, supra. 



The advance in labor productivity during 1923-29 concided rather 
closely with the increase in production, but in the latter period a 
wide divergence developed between the trends of productivity and 
production. In the earlier period an increase of 31.9 percent in 
labor productivity in manufacturing was matched by a rise of 30.1 
percent in production. In bituminous coal mining, labor produc- 



CONCENTRATION OF ECONOMIC POWER 



91 



en 

i 

Q 
LU 

X uj 



O z 



< 



CO 



Q 
O 

a: 
a. 



= i 



1 nr-; 1 


■ — 1 


\ 1 


« 








r ' 










\ t . ■ 




1 1 ' 








\i i . 


s 






















\ ^ ' 






"■v, 




J 1 






? 




^ ' ■ 


i 










1 1 


Bjflife) 






















1 


i 








^l' 


i \^ 






i 




' \N 




1 • ' 






_l_vj:: 






i 








5 


\ 




1 








\ y 


\ 


\ 




1 




\ 




! 
1 

1. 












i 


2 


i 


I 


% 


2 


i 


2 ■ 


? 


3 ' 



iSS28gog 



S J_S 8 





S S S 8 g S ? 



g S S S 




92 



CONCENTRATION OF ECONOMIC POWER 



Chart II 



PER CENT CHANGE IN PRODUCTION AND 
LABOR PRODUCTIVITY, 1923-29 & 1929-39 



MANUFACTURING 

1923-1929 1929-1939 

PER CENT PER CENT 

-40 -20 20 40 60 80 -40 -20 20 40 60 



PRODUCTION^ 



MAN-HOUR OUTPUT 



1 

[..ff , 


1 







m 




•■1 

..J 



STEAM RAILROADS 

1923-1929 1929-1939 

PER CENT PER CENT 

-40 -20 20 40 60 80 -40 -20 20 40 



PRODUCTION 



MAN-HOUR OUTPUT 







BITUMINOUS COAL MINING 

1923-1929 1929-1939 

PER QENT PER CENT 



-40 -20 20 40 60 80 -40 -20 20 40 60 80 



PRODUCTION^ 



MAN-HOUR OUTPUT 



, 




' 









ANTHRACITE MINING 

1923-1929 1929-1939 

PER CENT PER CENT 

-40 -20 20 40 60 80 -40 -20 20 40 



PRODUCTION 



MAN-HOUR OUTPUT 





I 











1 


1 




■ I 


1 


1 
1 


■« 


^■■^^H 


,1,1,1,1 



CONCENTRATION OF ECONOMIC POWER 93 

tivity remained remarkably stable, increasing only 8,1 percent, but 
production decreased 5.3 percent. Labor productivity in anthracite 
mining, actually declined 3.6 percent, while production went down 
20.9 percent. In steam railroads, an increase of 18.2 percent in labor 
productivity was accompanied by a rise of 3.3 percent in production. 

But during 1929-39 the pattern was entirely changed. Declines 
in production were coupled with remarkable advances in labor pro- 
ductivity. In manufacturing, productivity advanced 32.3 percent 
while production went down 0.5 percent. In bitmninous coal, labor 
productivity increased 31.5 percent but production declined 27.0 per- 
cent. In anthracite mining a similar decline in production (31.2 
percent) was overbalanced by an increase in labor productivity of 
79.0 percent. In steam railroads, labor productivity went up 31.1 
percent while production declined 25.8 percent. 

Labor productivity in three of the fields increased much more 
rapidly during 1929-39 than during 1923-29. In bituminous coal 
mining, productivity advanced 8.1 percent during 1923-29 but rose 
31.5 percent during 1929-39; in anthracite mining, it actually declined 
3.6 percent in the former period but increased 79.0 percent during 
the latter; and in steam railroads, it rose 18.2 percent in the earlier 
period but gained even more (31.1 percent) in the latter. Only in 
manufacturing was the increase in labor productivity in 1923-29 
at all comparable with that of 1929-39. This long-term increase in 
labor productivity is remai'kable because increases in output per 
man-hour occurring between 1929 and 1939 were achieved despite a 
level of output lower in the latter than in the former year. 

Of equal interest is the extent of the changes in labor productivity 
in specific manufacturing industries. The National Research Project 
on Reemployment Opportunities and Recent Changes in Industrial 
Techniques, organized by the Works Progress Administration in 1935, 
has formuLated indexes of wage-hour and man-hour productivity for 
a large number of industries,^ 

The changes in productivity, indicated by these series, are shown 
for 40 manufacturing industries in chart III, table 3,* Changes are 
indicated from the base year, 1923, to 1929 and 1936 for each of the 
industries. It was possible to project the series forward for 15 of 
the industries and thus the chart indicates also the percentage change 
to 1939 in 13 of tlipse industries, to 1938 in one and to 1937 in the other. 

Productivity increased in each of these industries over the base 
year in both 1929 and 1936, It likewise increased between 1929 
and 1936 in most of the industries. In contrast, production in 1936 
was below the 1929 level in 26 of the 40 industries as reported by the 
National Research Project. Despite this generally lower level of pro- 
duction, output per man-hour in 34 industries was higher in 1936 than 
in 1929, strikingly indicating the intensity and rapidity with which 
technological improvements were introduced during that 7-year 
period, 

* Production, Employment, and Productivity in 59 Manufacturing Industries, 1919-36, by. 
Harry Magdoff, Irving H. Siegel, and Milton B. Davis, May 1939. 

* Wherever possible, data were scoured relating to specific industries rather thnn to industry 
groups. In certain cases, such as the hosiery, underwear, outerwear, and knit-cloth indus- 
tries, productivity data are available only as output per wage-earner, while output p€r man- 
hour figures are available for the industry group, knit goods. In the case of leather, the 
Industry group is utilized, though man-hour data were available for component segments 
of that group, in order not to overbalance" the table and chart by the inclusion of five 
specific leather industries. Also, such relatively minor industries as beehive coke and 
secondary smelters and refineries were omitted from the chart. For a brief discussion of 
the adeyuacy of these series, see appendix C. 



94 CONCENTRATION OF ECONOMIC POWER 

Chart III 

PERCENTAGE CHANGES IN OUTPUT 
PER MAN-HOUR 

40 MANUFACTURING INDUSTRIES 



INDFX (1923=100) 



50 200 diO 300 350 



RUBBER TIRES & TUBES-- 

CIGARETTES 

PETROLEUM REFINING 

SILK a RAYON GOODS '- 

PULP MANUFACTURING —^ 

ICE CREAM--- 

CIGARS 

GLASS 

CHEMICALS 

CANNED 6 CURED FISH '- 

BOOTS & SHOES 

CANE-SUGAR REFINING 

KNIT GOODS - 

CEMENT 

BEET SUGAR '-"^ 

MOTOR VEHICLES 

CHEWING & SMOKING TOBACCO 

WOOLEN & WORSTED GOODS 

AGRICULTURAL IMPLEMENTS-- - 

IRON a STEEL 

MANUFACTURED ICE '--~ 

LOGGING CAMPS--- 

NEWSPAPERS a PERIODICALS - 
CANNED a PRESERVED I 
FRUITS a VEGETABLES' 
SAWMILLS a SAW-PLANE MILLS - 

PAPER MANUFACTURING 

COTTON GOODS 

FERTILIZERS 

PRIMARY SMELTERS al 

REFINERIES 1 
LEATHER INDUSTRY 

PAINTS a VARNISHES 
ALLOYERS, ROLLING MILLS! i ~ 

a FOUNDRIES I 

FLOUR a OTHER GRAIN- l.l 

MILL PRODUCTS I 

OTHER RUBBER GOODS -- 

FURNITURE--' 

BYPRODUCT COKE 

SLAUGHTERING a MEAT PACKING 

PLANING MILL PRODUCTS '-- 

BREAD a OTHER 1 
fiAKERY products/' 

CLAY PRODUCTS a NON-l 

CLAY REFRACTORIES _J 

<i,t)rrr/r„ • -Tfl^p R 




ISO 200 250 300 350 



CONCENTRATION OF ECONOMIC POWER 95 

Table 3. — Indexes of output per man-hour 1929, 1936, 1939: J/O manufacturing 

industries 



[1923 = 100] 








Industry 


1929 


1^36 


1939 


Agririiltnral implprnpnts 


130.4 
127.9 
123.5 
120.7 
102.0 
103.6 
108.7 
129.0 
153.6 
109.5 
105.3 
113.6 
131.4 
130.0 
116.3 
129.9 
136.6 
114.4 
120.0 
105.8 
119.0 
113.5 
108.0 
143.3 
122.4 
152.0 
155.3 
110.1 
120.2 
145.1 
140.1 
115.2 
167.2 
104.4 
138.3 
107.0 
121.7 
182.5 
126.3 
105.7 


153.8 

1 162. 4 

^168. 8 

107. 7 

167.2 

145.6 

1 169. 2 

165.5 

181.6 

105.0 

139.9 

138.9 

132.6 

126.5 

188.8 

190.8 

153.0 

1 166. 2 

119.9 

136.0 

151.7 

142.8 

' 152. 8 

161.3 

148.1 

136.9 

1 133. 2 

133.8 

142.8 

195.2 

221.0 

1 112.0 

304.7 

1 127. 7 

1212.0 

117.8 

190.4 

248.7 

157.1 

153.7 












Bread and other bakery products 


113 8 






















Clay products, other than pottery 


117 1 


Cotton goods 


146 7 






Flour and other grain-mill products 








Glass - ---- . . 




Ice cream . - _ . 












Byproduct coke - ... 








Logging camps . .. . .. 
















Newspapers and periodicals 


140 






Alloyers, rolling mills and foundries. 




Paints and varnishes 








Pulp manufacturing . 


2 229 7 




259 9 






Rubber tires and tubes ... 




Other rubber goods 








Slaughtering and meat packing.. .... 


129 4 


Cigars 








Chewing and smoking tobacco 




Woolen and worsted goods 


166 1 







1 1935. ' 1938. 3 1937. 

Source: National Research Project, Works Progress Administration, Production, Employment and 
Productivity in 59 Manufacturing Industries, 1919-30, Part II. The 1939 figures were projected by the 
National Research Project stafT and transmitted by letter to the Temporary National Economic Committee. 
For the 15 industries with 1939 figures the 1936 figures are revised. 

Five of the six industries in which productivity fell between 1929 
and 1936 suffered extensive declines in production, with a less severe 
decline in the sixth. The 1936 production index (19^9== 100) for the 
six industries follows : ^ 

Planing-mill products (1935) ^ 45.2 

Alloyers, rolling mills, and foundries (1935) 53.9 

Clay products 54. 1 

Primary smelters and refineries 64. 3 

Furniture 66. 5 

Byproduct coke 82. 9 

From 1936 to 1939 labor productivity continued its upward trend in 
all but one of the 15 industries for which the series have been projected. 
In newspaper and periodical printing and publishing the decline in 
labor productivity may be partially due to the decrease in output, the 
index of production falling 3.0 percent between 1936 and 1939. In the 
other industries (except motor vehicles) production was at a higher 
level in 1939 (or the last year for which data were available) than 
in 1936. 



5 Works Progress Administration, National Research Project. Production, Employment 
and Productivity in 59 Manufacturing Industries, 1919-36, Part II. 



gg CONCENTRATION OF ECONOMIC POWER 

Of particular interest are the extensive increases in labor pro- 
ductivity occurring in four industries. In the rubber tires and tubes 
industry, labor productivity in 1936 was 204.7 percent above the 1923 
level ; in petroleum refining, in 1939 it was 159.9 percent above 1923 ; in 
1936 in cigarette manufacture it was 148.7 percent above the base year; 
and in 1939, in pulp manufacturing, it was 129.7 percent above the base 
year. Although the increasing use of these commodities was undoubt- 
edly a primary cause of the advance in labor productivity, it is signifi- 
cant that three of these industries use chemical methods in production. 
The increasing use of chemical processes in productive operations ® 
probably portends a rapid increase in productivity in the fields where 
such processes are applied. 

TYPES OF LABOR-SAVING TECHNIQUES 

Before describing the various types of labor-saving techniques, 
attention should be called to two circumstances, under either of which 
labor productivity would tend to increase in practically any situation. 

(1) Labor productivity tends to advance as a greater degree of 
standardization is achieved. The smaller the variety of goods made 
by a given unit of productive equipment the greater the productivity 
of the labor involved. 

Any calculation of the savings resulting from standardization neces- 
sitates a determination of the economies of mass production and the 
fraction of such economies due to standardization. Such figures are 
almost impossible of verification, but in certain cases the economies 
attendant upon standardization are known to be great.'' 

A case showing the way in which standardization increases labor 
productivity is that of the Jantzen knitting mills; this firm for- 
merly produced sweaters, coats, caps, stockings and many other varie- 
ties of knitwear but in 1927 decided to make swimming suits of only 
one quality on a large scale. Upon this transition there occurred a 
phenomenal increase in productivity. Prior to the introduction of 
the standardization program an operator turned out an average of 
9 seams per hour. After its adoption, the worker produced 45 seam? 
per hour with no greater effort.^ 

The widespread acceptance of standardization in the United States 
is indicated in a recent publication of the National Bureau of Stand- 
ards, Reductions in Varieties Effected by the Application of Simpli- 

« For further discussion, see pp. 112-113, infra. 

TThe National Industrial Conference Board in its study. Industrial Standardization, ob- 
served that, "The automobile industry may have saved $750,000,000 a year throuuh stand- 
ardization ; the lumber industry, $250,000,000 a year; and the bcick industry. $1,000,000' 
a year; the building industry may have lost $2.000.000.000. by lack of standardization, but 
there seems to be ho conceivable way of checking up on the figures." 

As an example of the economies caused by standardization in one case, the National 
Industrial Conference Board citrd an American manufacturer of motor car parts. "He was- 
practically operating on a job-shop basis, although his total output was larw enou,uh to 
warrant quantity-production economies. Careful figuring showed him that he could cut 
his price in half, improve his delivery service, and yet make more money himself. 'if he 
could induce all his eustoniet,' to accept the same pattern. He laid the facts before them. 
They accepted the proposal. One of them saved $4,000,000 a year, or $20 a part on 200 000 
parts." National Industrial Conference Board, Inc., Industrial Standardization, New York 
1929, pp. 166, 171. 

"E. C. Robbins and F. E. Folts, Introduction to Industrial Management, McGraw-Hill, 
New York. 19.33. pp. 61-62. ^ ^„„ 

In the United States, the number of types of electrical lamps was reduco<1 from So.OOO 
in 1900 to 342 in 1923 bv the standardization of voltages and bases. The Commercial 
Standards Monthly observed that, "This eliminating of types had made it possible to develop 
and adopt high-speed machinery for lamp making. These machines eliminated practically 
all costly hand operations and made better lamps in larger quantities and at lower prices."" 
("SimplificoHon and Standardization within the Electric Lamp Industry,' Commercial Stand- 
ards Mont lily. August 1930, p. 46.) 



CONCENTRATION OF ECONOMIC POWER 97 

fication Practices. The number of varieties before and after simpli- 
fication and the resulting percentage reductions are presented for a 
large number of commodities in appendix D. The savings in unit 
labor expenditures which these reductions represent are suggested by 
the specific instances cited. 

(2) Labor productivity also tends to rise with increasing produc- 
tion. The tendency for unit costs to decline as the rate of operation 
advances is a well-known characteristic of most industries and re- 
quires no elaboration at this point. The applicability of this prin- 
ciple is qualified (a) by the existence in every form of activity of a 
point of diminishing returns, after which costs cease to decline with 
further increases in production, and (b) by the existence of certain 
"increasing cost" industries, particularly in the extractive fields. 

The general tendency of labor required per unit to decline as the 
percent of capacity utilized increases is shown in chart IV, table 4, 
for three fields in which such data are available: Iron and steel, 
cement, and brick and tile.^ 

It is difficult to classify the various types of labor-saving techniques 
because of their interdependency. For example, certain chemical 
processes which have greatly reduced labor requirements could not 
have been applied without the development of certain new materials 
such as alloy steels. Similarly, few types of machinery perform 
only one labor-saving function; for example, those that increase speed 
frequently also eliminate hand operations. Because of this inter- 
dependency any classification, such as the following, is presented solely 
for purposes of convenience. It helps to clarify the labor-saving 
effects of certain technological changes and to emphasize the number 
of ways in which technology increases the proditctivity of labor. 

Table 4. — Effect of variations in capacity utilization upon man-hoitrs required 



Industry 


Percent of capacity utilized 


Man-hours re- 
quired per unit 
of output 




55 to 60 .-._ 


100 






50 to 55.. 






45 to 50 






40 to 45 


118 




35 to 40 






30 to 35 .- 






25 to 30 


131 




20 to 25 _-.. 


135 


Cement 


100 


100.0 






80 


108 6 




60 


120.8 










20 


181 4 






100.0 






60 to 69.99 






50 to 59.99 


101 5 




40 to 49 99 


109 2 




30 to 39.99 


115.4 




20 to 29.99 


130 9 




Less than 20 


145 5 









Sources: 

Iron and Steel: U. S. Bureau of Labor Statistics, Monthly Labor Review, vol. 40, May 1935, p. 1161. 

Cement: Work Projects Administration, National Research Project, Mechanization in the Cement 
Industry, 1939, p. 23. 

Brick and tile: Works Progress Administration, National Research Project, Productivity and Employ- 
ment in Selected Industries, Brick and Tile. 1939. d. 117 



' There are two basic limitations to these data. First, the Iron and steel and brick and 
tile analyses extend only to 55-60 percent and 70 and over percent of capacity utilized, 
respectively. Second, since 1935, when the Iron and steel study was ma<le. the continuous 
process has been widely substituted in rolling mills for the semi- and non-continuous process. 
Under the continuous process the decline in man-hour requirements with increasing capacity 
utilization is probably even more pronounced. 

277551— 41— No. 22 8 



98 



CONCENTRATION OF ECONOMIC. POWER 

Chart IV 



EFFECT OF VARIATIONS IN CAPACITY 

UTILIZATION UPON MAN-HOURS REQUIRED 

PER UNIT OF OUTPUT 



INDEX OF 

MAN-HOURS REQUIRED 

PER UNIT OF 

OUTPUT 




(55 


IRON AND STEEL ^ 

60% CAPACITY UTILIZED- 100) 






INDEX OF 

MAN-HOURS REQUIRED 

PER UNIT OF 

OUTPUT 


130 
120 

no 


— ^...^ 


















■ 


^-^ 




















^"-^^^ 
























100 














inn 


20 


-25 25-30 30-35 35-40 40-45 45-50 50-55 55-60 
PERCENT OF CAPACITY UTILIZED 


INDEX 

MAN-HOURS f 

PER UNI 

OUT PL 

190 


OF -, INDEX OF 
EQUIREO CEMENT ^ MAN-HOURS REQUIRED 

T (100% CAPACITY UTILIZED" 100) OUTPUT 












180 










180 


\ 








170 


\ 








170 


\ 








160 


\ 








160 


\ 








150 


\ 










\ 










140 
130 
120 


\ 


V 








140 
130 
120 




\ 








\^ 










'"*-~«..,,^^^ 




no 






^^-^^-^^ 




no 






''^ 


««,.,__^ 


100 








" ,,.,,,,...^ 


100 


2 


40 60 80 l( 
PERCENT OF CAPACITY UTILIZED 


K) 


INDEX 

MAN HOURS R 

PER UNIT 

OUTPU 

150 


EQUIREO BRICK AND TILE^ m*%h£u 

r (70% CAPACITY UTILIZED -100) OL 


)EX OF 

RS REQUIRED 

UNIT OF 

TPUT 

150 














140 


\^ 












140 














130 


^\L 










130 






\^ 














\ 










no 








1 ^s^ 






100 








* 1 


00 
ovtr 


Unde 


20 20 


-30 


P 


30 
ER 


40 
CENT 


40- 
■CAP 


50 
ACITY 


50 
UTILI 


60 
ZED 




60 


70 


70ond 



' Bureau of Labor Statistics, Monthly Labor Ktniew, vol. 40, May 1935, p. 1161. 
2 National Research Project, Mechanization in the Cement Industry, 193l», p. 23. 
' National Research Project, Productivity and Employment in Selected Industries, Brick 
and Tile, 1939, p. 117. ' 



CONCENTRATION OF ECONOMIC POWEll 99 

TYPES OF LABOR-SAVING TECHNIQUEfS 

1. Power and energy development. ^° 
^2. Materials : 

{a) Substitute materials. 
(b) Improved materials. 

3. Processes : 

(a) MechanicaP^ (including multiple-function machinery). 
{h) NoKraechanical. 

4. Individual single-function machinery (grouped according to 
primacy of function) : 

(«) Elimination of operation. 

{h) Increase of speed. 

{c) Enlargement of capacity. 

5. Management methods : 

{a) The human factor. 

{h) The material conditions of manufacture. 

A procedure followed by the United States Patent Office was 
adopted in the classification : Under each of the five major types are 
included all labor-saving techniques pertaining to it except those 
which might be included in any of the preceding groupings. Thus 
the substitution of fuel oil for coal could not be placed under the 
"Materials" headiiig because it primarily belongs under the preced- 
ing heading of "Power and Energy Development." Though the 
classification in outline form is limited in usefulness, specific ex- 
amples are given in the discussion to illustrate the various types of 
techniques. 

POWER AND ENERGY DE\-EL0PMENT 

Power is primary to all productive operations. The story of 
technological progress is in no small part the history of the de- 
velopment and use of new and improved forms of power. At first 
man used his own brute strength and increased it with wedge and 
lever, pulley and plane, cog and treadmill. He domesticated ani- 
mals and trained them to use their strength at his commnnd. The 
winds and running water provided transportation and the motive 
power for mills to grind the grains of the earth, to hammer metals 
into shape, and even to operate complex machinery. Then came the 
practical application of steam as a source of power in removing 
water from mines, and in running continuously the already de- 
veloped machines of the textile industry. And today steam itself is 
feeling the inroads of other forms of power. 

As the forms of power change, a concomitant change in the 
,- fuels from which power is derived takes place. Electricity and 
natural gas are replacing steam, and in the generation of steam, 
fuel oil is replacing coal. Of primary importance among the many 
innovations Avhich have reduced the amount of labor required to 
produce a unit of energy (a British thermal unit) has been tb.e 

lo All products, processes, and machines for converting into power the sources of enertr.v. 
"Those that achieve a change in the product without a chemical reaction or a direct use 
■ Of electric energy (not including electric energy as a source of power or heatK 



100 CONCENTRATION OF ECONOMIC POWER 

substitution of fuel oil, natural gas, and hydroelectric power for 
coal. The use of each of these ' substitute fuels has been steadily 
increasing, with marked effects upon labor productivity. It is ther 
purpose of this section to discuss substitutions of one fuel for 
another and of new processes for obtaining greater energy from an. 
existent fuel, rather than to analyze mechanical changes which have 
increased the productivity of labor engaged in extracting any par- 
ticular fuel. 

The encroachment of fuel oil upon the bituminous coal market is 
in no firaall measure due to the methods of marketing fuel oil. It is 
a residual product of petroleum distillation and is sold at very low 
prices merely to dispose of a byproduct formerly considered waste. 
Figures published annually by a petroleum authority, Joseph E, 
Pogue, indicate that fuel oils are sold at Oklahoma refineries for 
less than is paid for the crude petroleum from which the fuel oil i& 
produced.^^ 

The increasing industrial use of natural gas is likewise due in large 
part to the low rates gas producers offer industrial users to secure 
equalized outlets for pipe-lme loads. "On the average, these 'indus- 
trial' gas rates are about one-fourth or one-fifth of the rates charged 
domestic or household users, and in some localities the domestic rate 
in as much as 10 times the industrial rate." ^^ 

The development by the Federal Government of large hydroelec- 
tric power projects which set a yardstick of low rates to both domestic 
and industrial users has undoubtedly adversely affected the market 
for coal though thousands of persons have derived benefits from this 
cheaper electric power. 

The extent to which fuel oil, natural ^as, and hydroelectric power 
have been substituted for coal since 1923 is indicated in chart V, 
table 5. In 1923 these fuels encroached upon only 18.5 percent of 
the possible coal market; by 1938 the proportion had risen to 36.9* 
percent. 

This increasing use of other fuels has resulted in material declines 
in the amount of coal actually produced. It has been estimated that 
between 1923 and 1938 approximately 810,126,000 tons of coal were 
displaced by these other three fuels. ^* The replacement of coal by 
other fuels is pertinent to this study because the production of the 
substitute fuels involves much less lalbor per unit of energy than does 
coal. Each ton of bituminous coal mined in the United States costs 
in wages approximately $1.27. Four barrels of fuel oil (fuel equiva- 
lent of 1 ton o^ coal) can be produced for a labor cost of about 6S 
cents. The production of 20,000 cubic feet of natural gas (fuel equiva- 
lent of 1 ton of coal) costs about 8 cents in wages. The production 
of 2,000 kilowatt-hours of hydroelectric power (energy equivalent to 
that produced by a ton of coal) costs less than 1 cent for labor." 
In the words of Thomas Kennedy, secretary and treasurer of the 
United Mine Workers of America : 



^» Testimony of Charles O'Neill, president, United Eastern Coal Sales Corporation, in 
hearings before the Temporary National Economic Committee, Pt. 30, Technology and 
Concentration of Econ<imic Tower, p. 17519. 

As the distance from the refinery increases the chances for the maintenance of low 
fuel oil prices are lessened, either because of high transportation charges (shortly after 
the beginning of the second World War tanker rates from the Gulf to the eastern sea- 
board rose several hundred percent) or of practices pursued by distributors. 

^ Ibid., p. 17519. 

'Uhid., p. 17529. 

"Ibid., p. 17520. 



CONCENTRATION OF ECONOMIC POWER 



101 



As against competing fuels labor plays a very important part in the pro- 
•duction of coal ♦ * *. Oil, gas, etc., come from practically laborless in- 
'dustries." 

'Table 5. — Consumption of bituminous coal, coal equivalent of fuel oil and natural 
gas, and output of hydroelectric poicer reduced to coal equivalent, 1923-38 

[Thousands of net tons] 



1923- 
1924. 
1925. 
1926. 
1927. 



1932. 
1933. 
1934. 
1935. 
1936. 
1937. 



Total 6,907,710 



Bitumi- 
nous coal 
consumed 
tnet tons) 



518, i 
484, ( 
499.: 
532,; 
499,! 
498, i 

519, i 
454,! 
371, i 
306,! 
321, ( 
347, < 
360,; 
422,: 
428, < 
340,; 



Per- 
cent 
of 
total 



Fuel oil 
consumed 

in coal 
equivalent 
(net tons) 



74, 038 
85,287 
85, 319 
88, 450 
97, 519 
102, 403 
90, 532 
83, 667 

79! 086 
85, 093 
91, 681 
102, 660 
110, 589 
102, 304 



1, 422, 266 



Per- 
cent 
of 
total 



Natural 
gas con- 
sumed in 
coal equiv- 
alent (net 
tons) 



Per- 
cent 
of 

total 



Output of 
hydroelec- 
tric power 
reduced 
to coal 
enuivalent 
on basis 
current 
efficiency 
rate for 
steam gen- 
eration 
(net tons) 



23,212 
21, 666 
23.474 
25,534 
27,485 
30. 532 
29,262 
26, 747 
23, 717 
25,574 
25, 524 
25, 033 
29,177 
29,475 
3i, 028 
31.295 



428, 735 



Per- 
cent 
of 
total 



636, 544 
609, 285 
639, 181 
678, 651 
653, 321 
667, 836 
698, 783 
619, 862 
525, 127 
452, 366 
469, 8S5 
506. 198 
535, 545 
617,862 
1640, 645 
.540, 148 



, 491, 189 



Source: Hearings before the Temporary National Economic Committee, Part 30, exhibit 2737, table 5. 

Of the estimated 74,540 jobs lost by miners between 1923 and 1937 
in the coal industry because of the use of substitute fuels, only a 
very small portion has been made up by labor expended in producing 
: substitute fuels. The bulk of the labor displaced by substitute fuels 
must be regarded as a net displacement. The total loss of employ- 
ment by the coal industry from all causes between 1923 and. 1937 is 
shown in table 6; 

Steps have been taken to decrease the amount of coal required 
to perform a given amount of work in an effort to maintain its com- 
petitive position. By better grading and sorting and more careful 
analysis of the conditions of coal utilization marked improvements 
have been made in its efficiency, and the productivity of labor involved 
has thereby been increased. 

In 1920 the average number of pounds of coal per thousand gross 
ton-miles in freight services on steam railroads was 172 ; by 1938 
it was only 115, a reduction of 33.1 percent, or a reduction in use 
•of 25,998,982 tons for 1938. 

The pounds of coal per passenger train car-mile between 1920 
and 1938 were reduced from 18.8 to 14.9, a decrease of 20.7 per- 
cent, or a reduction in use of 5,546,165 tons for 1938.^^ 

^"Ibid., p. 17191. 

"Some of the factors which have contributed to this increasing fuel economy in rail- 
roads are : Longer boiler tubes in the locomotives transfer heat more efficientiv ; im- 
provements in fireboxes permTt better draft and better combustion; superheaters raise 



102 



CO^sCENTRATION OF ECONOMIC POWER 



Chart V 

CONSUMPTION OF FUELS REDUCED TO 
BITUMINOUS COAL EQUIVALENT 

BITUMINOUS COAL, COAL EQUIVALENT OF FU^L OIL AND 

NATURAL GAS. AND OUTPUT OF HYDRO-ELECTRIC 

POWER REDUCED TO COAL EQUIVALENT 

UNITED STATES, 1923-1938 




1S:3 1924 1925 1926 1927 1928 1929 1930 1931 1932 1933 1934 1935 1936 1937 1938 



IN PERCENTAGE 



HYDRO-ELECTRIC OUTPUT ' ' [1 1 




1923 IS24 1925 1926 1927 1928 1929 1930 1931 1932 1933 1934 1935 1936 1937 1938- 

Source : Table 5. 



CONCENTRATION OF ECONOMIC POWER IQS 

Table 6. — Loss of employment in coal by causes, 1923-37 



Men i Percentage 



A. Increased labor productivity (including mechanization, etc.) 

B. Increased efficiency of utilization... 

C. Te'chnology of substitute fuels 

Total - 

D. Shrinkage of market 

Actual shrinkage of employment (1937) 



78, 140 
33,950 
74,540 



28. r 

12.4 
27.3 



186,640 I 68.4 

86,040 31 6 



Source: Hearings before the Temporary National Economic Committee, Part 30, exhibit 2738, table 24. 

In public utility electric power plants 3 pounds of coal were re- 
quired to generate a kilowatt-hour of electricity in 1920; in 1938 
only 1.41 pounds were required, a decrease of 53 percent, or a re- 
duction in use of 57,047,000 tons for 1938. 

In the manufacture of iron and steel, the number of pounds of 
coking coal required per ton of pig iron in 1920 was 3,421; by 1938 
it was reduced to 2,865 — a decrease of 16.3 percent, or a reduction 
in use of 5,325,551 tons for 1938.^« 

The general tendency toward increased efficiency in fuel consump- 
tion by large industrial consumers is shown in chart VI.^'' 

In still another way has the substitution of electric energy tended 
to increase the productivity of labor. As is pointed out in another 
chapter,2o the most modern and efficient types of productive equip- 
ment are generally those which are electrically operated. If a 
producer is induced to use a larger amount of electric energy by 
low rates, it is likely that he will install the most productive types 
of machinery. On the other hand, if electric energy is derived from 
steam-generating plants which use coal as their source of power, the 
increase in productivity in the plant may be offset to some extent 
by the large amount of labor involved in the production of coal. 
Where hydroelectric power is used, or where the energy is generated 

tha temperature of the steam and ^ive it greater expansive force ; steam separators or 
so-called desaturators reduce the moisture in the steam and permit easier supercharging ; 
special heaters preheat feed water in the locomotive boiler ; the reduction in the dead 
weight of locomotives and cars makes it possible to handle more weight in trains at 
quicker speeds with the expenditure of less fuel. Hearings before the Temporary Na- 
tional Economic Committee, Part 30, p. 16602. 

18 Ibid., p. 17518. 

" National Research Project, Fuel Efficiency in Cement Manufacture, 1909-35, p. 3. 

2« See pp. 202-208, infra. 



104 



CO^'CENTRATION OF ECONOMIC POWER 



Chakt VI 



PROGRESS OF EFFICIENCY IN THE CONSUMPTION OF FUEL 
BY LARGE INDUSTRIAL CONSUMERS IN THE UNITED STATES 











1 1 










1 CENTRAL ELECTRIC p 


















...^^^ 














. 


^^^ 














"> 


\ 














\ 


^ 










;°.^ 


MDS OF COA 
-HR GCNEHA 


TEO 


\ 














X 




^_- 


02 « 


37 ig 


2 19 


7 19 


22 19 


27 19 


52 19- 



















CLASS 1 ROAOS 


- 


o' 


^ 




FREIGHT SERVICE 
















^ 






,ooo«ossn,.«.^-M.e, 




~ 


.2' ' ' '...7 ,-.^2 


,.27 » 




,U 















CLASS 1 ROADS 
PASSENGER SERVICE 


- 


















y 


-v^ 




^ 






PASSEN«R TRAIN CAR-MILE 





r032 1037 tei2 )9I7 1922 









1 1 1 








/lENT PLANTS l_ 




































POUNDS OF 


COAL ANO 


















, , , , 














902 19 




12 19 


7 19 


22 19 


27 19 


32 19 


7 




KM l»JT 



THOUSANDS C B.T.V 





= 


_ 




















^ 


I 
















X/ 


^'^ 








OF CRUDE HUN TO STILLS 




































IPETR 




















2 la 








«'■''» 


r 









, — ■ ■ III ■ 




,^/- 


N/''"^'^' 






y 








^■'^- wrsT* w*SA!''.ssssf ' 


















. , . , 




ICOKEOV^'^SI 



Source : Works Progress Administration, National Research Project, Fuel Effi- 
ciency in Cement Manufacture, 1909-35, p. 3. 



CONCENTRATION OF ECONOMIC POWER 105 

by Diesel engines — the importance of which, as a source of electrical 
energy, is rapidly increasing ^^ — the amount of labor involved is so 
small that it would not offset increased plant productivity. 

The effect of all these forces has been an actual displacement of 
workers from the coal industry. The average number of miners- 
displaced from 1923 to 1937 by the substitution of new fuels and by 
the increased efficiency of coal has been estimated as follows : -^ 

Substitute fuel oil 16, lOO 

Substitute natural gas 23, 900 

All hydro-electric power _ 34,540 

Increased efficiency of utilization 33,950 

Total 108, 490 

An additional estimated 78,140 workers were displaced through 
mechanization and the increased productivity of labor. 

A future possibility in the displacement of labor through the sub- 
stitution of an almost laborless form of energy is atomic power. A 
recent article on the possibilities of applying atomic energy to 
industrial uses states : 

There are many gates to the reservoir of atomic energy, and one of them ha.s 
been opened * * * a new substance, uranium 235, which releases atomic 
energy easily, has been isolaited; one gram of its mass contains a total energy 
of about 11,700,000 kilowatt-hours of electricity, of which about one-tenth of 
1 percent can be annihilated and released as the equivalent of 11,700 kilowatt- 
hours of electricity, or about 5,000,000 times more energy than can be obtained 
from burning an equal weight of coal.^' 

Before atomic power can be used to run the wheels of industry 
many practical problems must be solved. Principal among them are 
the scarcity of ores containing uranium compound and the great ex- 
pense involved in freeing the active component. If the demand for 
the material becomes great enough, known deposits will be exploited, 
other sources of supply will no doubt be found, and reductions in 
price will almost inevitably follow. 

Though no one can predict when this new power will emerge from 
the laboratory ready for industrial use, the possible economic effects 
of its practical application should be borne in mind in any considera- 
tion of economic trends. Any widespread use of atomic power would 
ruin the coal industry. Practically the only markets remaining to it 
would be the production of coke for byproducts and the steel indus- 
try. Railroads would lose that substantial proportion of their reve- 
nue derived today for the transportation of coal, and their savings 
from the use of atomic power would not compensate for this loss. 
The oil industry would suffer alike with coal since the motive power 
of automobiles might well be changed from internal-combustion to 
steam, with the fuel supply for the life of the car built in at the fac- 
tory. Atomic power might become the basis for a new method of 
generating electricity without requiring any apparatus for trans- 
forming heat into mechanical, and mechanical into electrical energy, 
thus eliminating boilers, engines and dynamos.^* 

^ Diesel engine installations by 1940 represented over 15,000,000 horsepower, which 
otherwise would largely have been generated by coal. (Hearings before the Temporary 
National Economic Committee, Part 30, p. 17189.) 

^ Ibid.; p. 17190. 

^ John J. O'Neill, "Enter Atomic Power," Harpers Magazine, vol. 181, June 1940. 

2* Ibid. 



106 CONCENTRATION OF ECONOMIC POWER 

Remote as these possibilities may seem at the present time, they 
would only carry to a logical conclusion tendencies already existent 
in the field of power and energy development. 

MATERIALS 

Suhstitute Materials. 

The measurement of changes in labor productivity in terms of a 
specific type of goods tends to minimize the extent of labor-saving since 
labor-displacement resulting from the substitution of one material for 
another is not taken into account. For example, between 1927 and 1937 
the production of paper bags increased by approximately $26,000,000, 
while the production of textile bags declined by $45,000,000.^^ A com- 
parison of changes in output per man-hour in textile bag manufacture 
with those in paper bag manufacture would not at all reveal the 
amount of labor displaced in the transition to the newer type of bag 
since less labor is required to manufacture paper bags than textile 
bags. In an analysis of this nature it should be borne in mind that 
the number of man-hours required to produce a quantity of a given 
type of goods is not necessarily the same, because of substitutions, as 
the number of man-hours required to produce a quantity of goods 
necessary in the performance of a given function. 

The replacement of one commodity by another is pertinent to this 
study only if the new item requires less labor per unit to perform 
a given function than the product replaced. In general, the more 
important substitutions throughout the history of industrial society 
have usually been accompanied by considerable reductions of labor re- 
quired per unit. 

When, for example, a natural product is replaced by a chemically 
produced article, a decrease in unit labor requirements almost in- 
variably takes place. Reductions in labor required occurred in the 
case of such substitutions as rayon for silk, vanillin for vanilla beans, 
sodium silicate for dextrin and glue, lacquers for varnish, synthetic 
for natural camphor, and aspirin for quinine. The production of 
ammonia, methanol, and acetic acid from chemical raw materials by 
chemical processes likewise reduced the amount of labor required 
per unit,^^ 

Unit labor requirements are reduced by such substitutions because 
the amount of labor required in chemical production is relatively 
small. A comparison between chemical and other industries as to 
the proportion which wages constitute of the value added by manu- 
facture shows that in the chemical field, labor's share is quite low.^^ 

The changes which have taken place in the painting of automobiles 
are illustrative of the way in which the substitution of a synthetic 
product reduces the amount of labor required not only in the produc- 
tion of the material but also in its application. Before the first 
World War automobiles were painted with natural enamels. Many 
coats were needed and their application required a relatively large 

^"U. S. Buroau of the Census release, Background 1940 Census, oh. 4. p. 4 (undated). 

2« Through the use of certain well-known types of chemical reactions on a few basic 
materials, principally coal, natural gas, petroleum, plants, molasses, and wood, a host 
of products are made synthetically in competition with similar products made from 
natural materials. Vast indeed is the field of synthetic organic chemicals and dve inter- 
mediates in which a few of the approximately 200,000 known carbon compounds serve 
as the starting materials for scores of such products as medicine."?, tannins, perfumes, 
dynamite, drugs, solvents, sweetening agents, rayon, antiseptics, and lacquers. 

2T See appendix E. 



OONCENTRATION OF ECONOMIC POWER 107 

labor expenditure and much handling. Since the enamels were very 
slow in drying, approximately 2 weeks were needed to paint a car. 
The slowness of the process often tied \ip production all along the 
line with a consequent decline of labor productivity in the entire 
plant. Shortly after the war nitrocellulose lacquers replaced the old 
enamels. Although about 10 coats were required, the period of dry- 
ing was greatly reduced so that the entire process was shortened from 
weeks to hours. 

During the early thirties synthetic enamels were introduced, 3 or 4 
coats of which achieved the same result as 10 coats of lacquer. Huge 
tunnels were constructed in which each coat of the synthetic enamel 
was baked on in about 2 hours, thus eliminating slow air-drying. 
Now infra-red lamps line the ceiling and walls of the tunnel and re- 
duce the time to about 10 minutes per coat. 

Chemically produced articles are by no means the only ones featur- 
ing in replacements. Alloy steels are being used increasingly in 
place of forged steel; they decrease operations and labor expenditure 
required to perform certain functions. The introduction of cam- 
shafts and crankshafts made of alloy cast in place of forged steel 
has eliminated at least 10 operations formerly done by machine and 
handwork in one process of automobile manufacture.^* 

Some replacements raise the question of durability. Where dur- 
able goods replace a perishable commodity, a decline may occur in 
the total amount of labor needed to meet a given need over a long- 
term period. When, for example, electric refrigerators replace man- 
ufactured ice, a net decline probably takes place in the amount of 
labor required over a period of time to keep foods cool. The amount 
of labor involved in the production, sale, transportation, installation, 
and occasional repair of a refrigerator over its life-span is undoubt- 
edly less than the amount of labor required for the production, sale, 
transportation, and installation of an ice-box plus the labor required 
in the continual production and delivery of the ice. 

There is also the complicating factor that the total amount of 
labor required for the production of any new durable goods, after 
an initial spurt following its introduction, usually declines as the 
market for the product becomes increasingly saturated. Perishable 
goods, on the other hand, are almost immune from this phenomenon 
of market saturation." 

Perhaps a brief analysis of some of the substitutions in the basic 
labor required for the production of any new durable goods after 
productivity may be increased due to new materials replacing old. 
Until the beginning of the last century the basicdurable material of 

** National Recovery Administration, Research and Planning Division, Preliminary 
Beport on Study of Regularization of Employment and Improvement of Labor Condi- 
tions in the Automobile Industry, 1935, appendix B, exhibit 16, p. 15. 

A somewhat similar case is the so-called "quick-setting" or high early strength ce- 
jnent. This tjl^e of cement is usually interchangeable with ordinarv Portland cement, 
except in a few cases, such as the construction of large dams. Its particular advantage 
is that it requires only about 3 days to set compared with approximately 28 days for 
•ordinary cement. The method of laying down the cement is the same for both types, 
but a considerable saving of labor occurs in the subsequent processing. Ordinary cement 
must be kept moist for at least 10 days after it is laid down, compared with 2 days for 
<iuick-setting cement, a saving of approximately 90 percent in the labor required in 
the moistening process. The quality of ordinary cement has been improved to reduce its 
period of setting. Despite these improvements, quick-setting cement has made con- 
siderably inroads on the markets of the regular tvpe. 

f For an analysis of the problem of market saturation, see T. N. E. C. Monograph No. 1, 
SPrice Behavior and Business Policy, by S. Nelson and W. Keim. 



108 OONCEISTRATION OF ECONOMIC POWER 

our economy was wood. It was the fundamental construction material- 
for implements, tools, presses, water pipes, cylinders, household fur- 
nishings, vats, barrels, gearing, pumps, and even machinery such as. 
lathes.3° 

But with all its attributes wood could not stand the competition 
of metals. By 1917 the plight of the lumber industry had become 
acute and in a study made at that .time the Department of Agricul- 
ture found that the annual substitution for wood i)i all forms 
amounted to approximately 27,715,000,000 board feet — of which 8,090,- 
000,000 was for wood in the form of lumber — and that the general 
trend indicated increasing replacement.^^ 

There has been no abatement in the replacement of wood, because 
other materials have been developed which are less expensive and 
more durable. One authority (Dr. John E. Teeple) has observed 
that wood — 

has long since ceased to be a necessity either as a fuel or a construction ma- 
terial, for furniture or for charcoal. Its chemical use as a source of methanol,, 
acetic acid, and acetone has already been replaced by synthesis and could be 
discarded entirely without the least inconvenience to chemical operations. Its 
use as raw material for making newsprint paper and near-silk garments would 
present a more serious difficulty, and these industries would be forced to use 
annual cellulose from plants instead of perennial cellulose from trees.^" 

A unit of wood undoubtedly involves much more labor in fabrica- 
tion than a unit of metal. Wood is not malleable; it cannot be 
melted and poured into a desired shape ; it is bulky and breakable ; it 
defies complete standardization; and it requires a relatively high 
degree of skill in fabrication. 

Two outstanding results of the substitution of metals for wood are : 
(1) The amount of labor required to perform a given function has 
been markedly reduced ; ^^ and (2) the use of metal has made possible 
a great expansion in industrial production. 

The next possible change in the economy's basic durable goods, 
already looms on the horizon : the substitution of plastics for metals.. 
Plastics are made by the action of acids, formaldehyde, or other 
chemicals on such materials as vegetable fibers, soybeans, dried blood, 
or camphor. Some of the natural materials used as a base were for- 
merly regarded as waste. The products consist of these chemically 
treated, moldable substances, together with coloring matter and fillers 
of clay, talc, asbestos, or mica. 

Various types of plastics, together with their typical uses, are sum- 
marized in appendix F. It is sufficient here to note that the uses 
range from adhesives to watch crystals, from automotive and airplane- 
parts to telephone 'equipment, and from bearings to transcription 
records. Despite the extent of their use, the potentialities of plas- 
tics have hardly been tapped. In the automotive field, window 

80 "As raw material, as tool, as machine-tool, as machine, as utensil and utility, as fuel, 
and as final product, wood was the dominant Industrial resource ♦ • * " (Lewis Mum- 
ford Techniques and Civilization, Harcdurt Brace, 1034, p. 120.) 

" IJ. S. Department of Agriculture Rept. No. 117, "The Substitution of Other Materials 
for Wood," Studies of the Lumber Industry, Part XI, 1917, by Rolf Thelen. 

** Williams Haynes, Men, Money, and Molecules, Doubleday Doran, 1936, p. 158. 

«*The relatively large labor expenditure Inherent in wood applies not merely to the 
production process it.self but also to maintenance and repair. In railroads, to cito but 
one instance, the introduction of all-steel box cars and the more general use of all-steel 
open-top cars ; the practical elimination of wood from passenger cars and from a large 
portion of the freiglit cars has led to greater efficiency and less frequent repairs. (Hearings- 
before the T. N. E. C, Part 30, p. 16612.) 



CONCENTRATION OF ECONOMIC POWER 109 

frames, glove compartment doors, and headlamp shells have been 
;the major contributions to date.^* 

One of the important potentialities of this new material is the de- 
velopment of plastic fenders and molded plastic bodies for auto- 
mobiles, and of plastic airplane fuselages and wings. As early as 
1933 Henry Ford said, "Bodies * * * can be made from cellulose 
-of cornstalks, with a woven-wire reinforcement inside and steel rein- 
forcement at the doors, and will be lighter, stronger, and quieter than 
metal bodies." ^^ The attainment of this goal has been blocked up to 
the present time by limitations on the size of moldings. That this 
major difficulty is nearing a solution is indicated in a recent article 
Ijy a technician in the field, Gordon M. Kline, of the National Bureau 
of Standards, who says, "Some progress has been made toward the 
development of a technique of molding which offers promise of 
eliminating the major obstacles of limitations on press sizes and 
capacities and mold expense.'-' ^^ 

The significance of this possible substitution of plastics for metals 
is that plastics require much less labor than metals in their production. 
In fact, the process of plastic manufacture is almost completely auto- 
matic and as the industry expands, unit labor requirements may be 
■expected to fall to even lower levels.^'^ 

This saving of unit labor expenditures would appear even larger 
if the productive process were carried back to the extraction of raw 
materials.^® Most plastics are derived either from vegetable matter, 
chiefly cellulose; from animal matter, chiefly sour milk; from mineral 
matter, chiefly phenol ; and from the air itself, chiefly nitrogen in the 
form of synthetic urea. These raw materials obviously involve less 
labor per unit than the production of metals, for essential to metal 
manufacture are such labor-consimiing functions as the extraction and 
transportation of coal and ore. 

Prophecy of future technological developments is not within the 
realm of this study. The potential widespread use of plastics is 

'* Among potential appUcations of plastic moldings in automotive engines and chassis 
are the following parts : tappet covers, rocker covers, wheelcases, water-pump rotor, water- 
pump bushings, oil pumps, rockers, cam wheels, chain wheels, fan blades, camshaft bear- 
ings, thrust washers, steering-joint bearings, gear-box and rear-axle covers, cable conduits, 
filler, caps, brake cross-shaft bearings, and spring interleaves. (See Journal of the 
Society of Automotive Engineers, May 1940, "Plastics and Their Uses in the Automotive 
Industry," by Gordon M. Kline.) 

35 Time, Feb. 13, 1933, p. 13. Plastics may also come to replace textiles in certain fields. 
■"These new developments (in insulation) may over a period of time have an important 
■effect on the production of cotton yarns and fabrics. Great quantities of cotton yarns are 
used for insulation purposes, and jf * * * plastic comes up to expectations in this 
field, its increased use may replace a considerable part of the business now going to cotton 
spinners. The application of * * * plastic to ordinary fabric covers and lace fabrics 
should also be of interest to the cotton trade." (American Wool and Cotton Reporter, 
Aug. 29, 1940, p. 18.) 

3° (Kline, loc. cit.) An Associated Press dispatch of June 25, 1940, reported the forma- 
tion of a new military aircraft concern. The company, to be known as the Twentieth 
Century Aircraft, Inc., will produce "pursuit ships with the fuselage and wings of 
plastic speedniold construction, with veneer sheeting impregnated with synthetic plastic 
material." The fighter planes are expected to have a speed of 340 miles an hour. (The 
Evening Star, Washington. June 25, 1940.) 

3' The first stage is the production of the resin — a chemical process performed in 
digesters or kettles. Next, the resin is mixed with the filler by means of rolls and 
grinders. It is then packaged hs a powder in large drums. The fourth stage consists 
in molding the material into desired shapes ; the material is simply fed into a hopper, 
and, by means of recently developed machinery, automatically molded. In the final stage 
of finishing and polishing, the highly efficient equipment already developed to perform 
those functions on other materials is merely adapted to plastics. 

*' Says the Department of Agriculture : "It appears that the recent rapid technical 
advances in the arts of making and using plastic compositions are likely to lead to 
continued growth of the plastics industry. The quantity of agricultural products required 
in such growth, while not inconsiderable, is still relatively small compared with the total 
. supply of the products." (IJ. S. Department of Agriculture, The Agricultural Situation, 
September 1940, p. 22.) 



]^][Q CONCEIS'TRATION OF ECONOMIC POWER 

merely another li^ik in the chain of replacement inherent in technology 
and is pertinent to this analysis because it involves the substitution of 
a material requiring less labor than the product replaced. 

As metals replaced wood, the accompanying expansion in indus- 
trial output undoubtedly more than absorbed the labor displaced hy 
the substitution. Whether there will occur a similar expansion in 
industrial output upon the possible substitution of plastics for metals, 
an expansion which would have to be tremendous to absorb the dis- 
placed labor, is a matter of conjecture. 

Improved Materials. 

Improvements are constantly being made in existent materials which 
also reduce unit labor requirements. One type of improvement is 
greater durability which reduces the number of units of an article 
required over a period of time and the amount of labor riequired in 
pi^oduction. Breakdowns also occur less frequently and the amount of 
labor involved in maintenance and repair is lowered. For example, in 
the last 20 years the life of wooden railroad ties has been extended by 
pre-treatment to about 21 years, compared with 8 years for the un- 
treated tie. Since there are about 3,000 ties to every mile of track, the 
amount of labor saved in both their production and installation is; 
tremendous. Similarly, the use of rust-resisting steel on railroad car? 
has eliminated much servicing and painting and has also prolonged 
the life of the car.^^ Another widely cited instance of increased dura- 
bility is the automobile tire. Between 1920 and 1937 the life of the 
average automobile tire more than doubled, rising from about one and 
one-quarter years to nearly 3 years.*° 

The constant improvement of cables and wires in the electric power 
and light industry likewise has reduced the amount of maintenance- 
and repair labor required. Overhead transmission lines have been 
made more durable with improved insulators, and cables for under- 
ground high voltage transmission have been materially improved 
through the introduction of new insulating mediums. In the distri- 
bution of power, new and improved types of cables and wires have 
also reduced materially the possibility of breakage or interruption.*^ 

Innumerable instances of both substitutions and improvements in. 
quality can be found in almost any segment of the industrial world. 
All illustrate the ways in which substitute materials and improved 
materials may bring about reductions in unit labor requirements: (1) 
By replacing a product with a new article which in its production 
requires per unit a smaller amount of labor and (2) by increasing the 
durability of an existent material, so that over a period of years fewer 
units are required and the amount of maintenance and repair labor is 
reduced. The examples also indicate the vast extent of the terrain to 
be covered in an analysis of the ways in which technology operates to 
increase the productivity of labor. 

^ Hearings befoi-e the Temporary National Economic Commtttee, Part 30, p. 16612. 

*<• IT. S. Bureau of Foreign and Domestic Commerce, Rubber News Letter, vol. 13, No. 19,. 
October 15, 19.S9. 

« National Resources Committee, Technological Trends and National Policy, June 1937, 
pp. 279. 280, 286. 

The Bureau of Labor Statistics has observed that one of the principal factors responsible- 
for increased labor productivity in this industry is the improvement in the wearing qualities 
of electrical equipment, with corresponding reductions in maititenance labor. (U. S. Bu- 
reau of Labor Statistics, "Labor Productivity and Displacement in the Electric Light and; 
Power Industry," Monthly Labor Review, August 1932, pp. 249-259.), 



CONCENTRATION OF ECONOMIC POWER HI 

PROCESSES 

New processes which reduce unit labor requirements are constantly- 
being introduced. These may be classified as (1) mechanical, which 
achieve changes in products without chemical reaction or direct use of 
electrical energy ,*2 and (2) nonmechanical, which bring about changes 
either by chemical reaction or direct use of electrical energy or through 
electro-chemistry. 

Mechanical {Inclvdrng Midtiple-Fivnction Machinery). 

An entirely new mechanical process or a new type of multiple- 
function machinery usually represents labor-saving in its most spec- 
tacular form. Perhaps the most outstanding example of this type of 
technique is the replacement of the old hand mill by the continuous 
strip mill in iron -and steel manufacture. Before the introduction of 
the continuous strip mill it was necessary to cool billets or slabs upon 
leaving the open hearth, reheat them, and then to push and pull them 
by manual force through slowly operating steam-powered rollers until 
the desired steel shapes were formed. 

In a typical continuous strip mill the hot slab first passes through 
a scale breaker which removes the scales due to oxidation. Traveling 
along on rolls the slab then passes through a number of "roughing" 
mills in succession, each over 20 feet in height. Top and Jb-ottom 
rollers weigh approximately 100,000 pounds apiece and the center 
rollers 22,000 pounds apiece. At this stage the steel has again ac- 
cumulated scale, so it passes through a second scale breaker. Beyond 
the second scale breaker, rolls carry the slab — by this time elongated 
into a strip — into a series of six finishing stands, each of which has 
two sets of rollers arranged horizontally. These finishing stands 
successively reduce so effectively the thickness of the steel that a single 
strip may be going through the rollers of all six stands simultaneously. 
Each stand operates slightly faster than the one ahead, and the greater 
the degree of reduction to be accomplished, the faster each successive 
stand operates. 

The six stands in a typical mill have a combined weight of approxi- 
mately 5,000,000 pounds and are so large that, were the rollers re- 
moved, an automobile could be driven through the middle. The 
entire continuous strip process is electrically controlled and operated 
by a switchboard from a pulpit. As the steel comes out from the last 
finishing stand in a long strip reduced to the desired thickness, hot 
flying shears, synchronized with the last finishing stand, cut the strip 
to tlie desired lengths. 

The labor-saving effected by this new process has been tremendous. 
It is estimated that 126 men in the automatic steel mills can produce 
the same tonnage as 4,512 men in hand mills, a 97 percent reduction 
in man-hours. Actually, human labor is practically eliminated in 
the rolling process with electric power substituted in its place. *3 

Another relatively new metallurgical process is the pouring of 
molten materials into molds which requires little labor expenditure 
compared with. methods involving beating, hammering, pounding, 

" The "direct use of electrical energy" does not include electrical energy as a source of 
power or heat. 

"Hearings before the Temporary National Economic Committee. Part 30, p. 16459. 
For further discussion of increased labor productivity in iron and steel, se epp. 235-240, 
infra. i .7 , vv ^- 



^12 CONCENTRATION OF ECONOMIC POWER 

pressing, etc. The electrical equipment industry has recently ap- 
plied this process to the manufacture of magnets, and thus maintains 
the same output at a reduction of approximately 75 percent in labor 
expenditure. Magnets are now poured into molds instead of being 
formed, eliminating such operations as cutting, forming, hardening 
(now accomplished by the use of cobalt), aging, gap grinding, shoe- 
ing, single magnetizing, and measuring the steel. Incidentally, the 
magnet thus produced is of a better grade than that manufactured 
under the old method."** 

Of fundamental importance to the automobile and related indus- 
tries is the one piece stamping process. Its use in only one segment 
of the automobile, the so-called underbody, has eliminated 18 dif- 
ferent parts which no longer have to be built and assembled. Stamp- 
ing is an excellent illustration of the transition from process to mul- 
tiple-function machinery. In many automobile plants where pre- 
viously four or five stamping presses were necessary, a multiple press 
has been introduced. Dies are now constructed so that a whole series 
of stamping operations may be done on one machine.*^ 

A familiar though striking illustration of a new process is the dial 
telephone. As a result of its use, the American Telephone and Tele- 
graph Company has ceased hiring approximately 25,000 girls a year 
and instead is displacing many whom it has employed for years.*^ 
Improvements in the dial process frequently result in an almost com- 
plete elimination of operators. The only functions of operators in 
a modern dial office are (1) to run the cordless B board, on which 
terminal calls from manual offices come in, and (2) to handle ticket 
calls, those outside the contiguous or free area. Formerly ticket calls 
had to be put through by an operator, but with zone dialing a cus- 
tomer now dials practically any number within a metropolitan area. 
Zone dialing transfers the prospect of completely automatic phone 
service from a remote possibility to a probability. 

Nonmech anical. 

In some cases nonmechanical processes are replaced by tliose of 
greater efficiency ; in others, mechanical processes are replaced by non- 
mechanical methods. Just as chemically produced materials reduce 
unit labor requirements by replacing natural commodities, the re- 
placement of labor-consuming mechanical processes by automatic 
chemical methods also reduces the amount of labor required. 

An outstanding instance of this type of technological advance is 
the so-called selective flotation process of concentration in the non- 
ferrous metals industries and related fields.*' The conversion of ores 
of non-ferrous metal content into pigs or slabs of the refined metal 
involves three separate operations: Concentration, smelting, and re- 
fining. The concentration operation was formerly performed by the 
interaction of gravity and direct oscillation. Through the action of 
jigs, tables, vanners,'and rolls the metal content was separated from 

*' Heailntis before the Temporary National Economic rommittee, Part 30, p. 17422. 

''- Ibid., p. 1G380. 

<« The extent of the labor reduction effected by the dial telephone is illustrated by the 
Boston metropolitan area, which now employs only 3,500 to 4,000 operators. Rased on 
the number of stations and messages handled by an operator in 1025, at least 12.000 
operators woul(i now be required to operate this system manually, a job opportunity loss 
of 7.5 percent. Hearings before the Temporary National Economic Committee, Part 30, 
p. 1G()72. 

■" For further discussion of increased labor productivity in non-ferrous metals, see pp. 
244-247, infra. 



CONCENTRATION OF ECONOMIC POWEI^ 113 

the ore. The efficiency of this method in the recovery of mineral 
content was comparatively low. , , . ,, -, fi, ^ ^4= 

The process which changed completely this rather crude method of 
concentradon wa. introduced in this country ni 1911, has had a phe- 
iZenal growth and predominates in the nonferrons metals indus- 
irTs today- Tiie selective flotation process utihzes cheinical a trac- 
on to separate desired mineral particles from waste. Chemicals are 
combined With a mixture of finely ground oi;e a^d water, tlien an is 
forJed in and the mineral particles stick to the air bubbles and float 
to ?he top, there to be collected, while the waste remains at the 

^"^This process has tended to reduce unit labor requirements, not only 
in the concentration stage but also in the ensuing smelting and re- 
finiit opera ions through raising the mineral content of the feed, 
i^^hfs a^si stimulated the development of improved grinding proc- 
es4s and of the reverberatory furnace, which replaced the blast 
furnace ncreasing labor productivity in the smelting operation. 

"similar procels for ferrous metals has been developed, the mag- 
ne ic ep^^ of ores. Here magnetic force rather than chemical 

attraction performs the function of separation. This method has 
reSaced he crude method of selection m ferrous metals ]ust as selec- 
Lve Nation has in nonferrous metals, with similar savings in unit 

''LTbo'rtvrgtafalso resulted from process changes in the rubber 
M Inrhi^m Bv the use of organic accelerators the time required 
or J^i r-bb?;t!^s ha's been reduced f rom one-half to two-thirds 
wTth incidentally, onlv small attendant capital outlays.- 

An electrical process has replaced a mechanical process m the 
substitution of welding for rivlting. Rivets and their mstallation 
nvo !i S^^^^^^ labor than welding, and when the welding process 
becomes automatic-that is. when spot or flash welding is doii.e in a 
so Xd welding machine-the disparity between the two processes m 
labor requirements becomes enormous.^° , ^ ■ u^ ^^.r.A ,...l,^ 

A recent advance in paper makmg-that of securing bleached p^^^^^ 
from southern pine by the sulphate process-ilhis rates the indiiect 
waTin w ich I change in processing may result in increased labor 
mcxUctiTky Tlie piincipal product of plants in the Sou h using 
The Ipl ate process formirlv was an unbleached pulp, but there is a 
giowing^^^^^^^^^ made possible by advances m cl-^^^^ca processes 

fo manufacture and sell bleached pulp. This greater diversity of 
proXct en aicjes their market and makes possible a more sustained 
?ate of operation, which results in greater efficiency and lower unit 
labor requirements. 

.sBv V.^ns i>:on ore !?-ide the u.a.netic^opara^^^^^^^ 
ately attracted and thus '^'•^P^/^^^f /. 5"^^,;^*^ 'wU ferrous content are removed fion. the 

is really a jig with lo^^V-^^In'l^ifnnf or fl-ish welded * * * Welding of the bade and 
positions they are 'n'toniatically si^^ot or fla^^^^^ ^^ operate, and a helger for 

Muafter panels ^f q"";*%^'«;.:;^,f '^('^rni" f^^Vratton Research and Planning Division, Prelim- 
finishins.- /N^t.ona Reco^e^ Ad nm^^^^^^^ d Iiuprovement of Labor 

izTi^itSrin the /ut"3n\obflen\kus"ry. 1935. Appendix B. Exhibit 16. p. 4. 



277-151 — 11— No. 22 9 



J 24 CONCENTRATION OF ECONOMIC POWER 

Labor productivity and general efficiency will undoubtedly con- 
tinue to be increased materially by new and improved chemical and 
electrical processes. Industry has only begun to utilize their immense 
potentialities while major mechanical improvements are becoming 
increasingly difficult because of the high level of efficiency already 
attained by many mechanical techniques. 

INDI\nDUAL,, SINGLE-FUNCTION MACHINERY 

Improvements in individual single-function machinery may be 
grouped according to the three ways in which they reduce unit labor 
requirements: (1) Elimination of one or more hand operations; 
(2) increasing the speed of the machine; and (3) enlargment of capa- 
city without a corresponding increase in labor requirements for feed- 
ing and attending." 

The principal advantage of this classification is that it does not 
become hopelessly detailed due to the infinite number of variations 
in structural characteristics of most types of machinery. Certain 
improvements in individual machines reduce unit labor requirements 
in all three ways, but this does not invalidate the classification. The 
new wristpin inspection machine, for example, eliminates hand opera- 
tions, increases the speed of operation, and enlarges capacity. A dis- 
cernible difference exists, however, between those types of equipment 
designed primarily to eliminate hand operations and those designed 
to increase either the speed or the capacity of the equipment. 

Many improvements in textile manufacture, for example, have had 
as their primary objective the increase in speed of operation, but in 
certain cases capacity has also been enlarged. Consequently, in group- 
ing labor-saving machines, the primary way in which they save labor 
will be the basis of classification. It should be borne in mind that 
if a particular technique increases labor productivity in several ways, 
its difficulty of classification in no way lessens its effect upon labor. 

Elimmation of operations. 

The installation of new types of machinery or the improvement of 
existing equipment is constantly eliminating hand operations in the 
industrial world. Seldom spectacular, these changes are primary 
causes of increased labor productivity. 

The automobile industry presents some interesting examples of the 
introduction of equipment which has eliminated a considerable 
amount of labor expenditure."- One such apparatus is a completely 
automatic inspection machine, which, with a photo-electric eye, grades 
wristpins as to size in thousandths of an inch and as to shape and 
smoothness of. surface. The machine eliminates from 10 to 20 inspec- 
tion men formerly required to grade the same number of pins.^^ It 
proved so satisfactory on wristpin inspection that a similar machine 
was designed for camshaft inspection and now 4 men perform the 

"Harry Jerome, Mechanization in Industry, National Bureau of Economic Research, 
New York, 1934, p. 42. This classification of labor-saving machinery is somewhat simllai 
to Professor Jerome's, except that he restricts the third type to enlargements of capacity 
merely through greater physical' size. Recent technological developments indicate that 
capacity of certain types of equipment may be increased without either an increase in 
physical size or a corresponding increase in labor requirements for feeding and attention. 

" See pp. 256-260, infra, for further discussion. 

"National Recovery Administration, Research and Planning Division, Preliminary 
Report on Study of Regularization of Employment and Improvement of Labor Conditions 
in the Automobilf Industry, appendix B, exhibit 16, D. 11. 



CONCENTRATION OF ECONOMIC POWEli 115 

work formerly done by 18. Another robot, designed along similar 
lines to inspect crankshafts weighing 75 pounds each, enables 2 men 
to perform work formerly done by 90.^* , . . , > , _„., 

Sne automobile factory formerly placed ring inserts ior valve seats 
in cvlinder blocks by hand. Machines now insert all valve seats m 
he block at once. Three men operate the machine with one mechanic 
as repairman, with a resultant labor saving of over 60 percent m this 

""^Tlie^automatic tube molding process in the rubber tire industry is 
another instance of the elimination of hand operations By means 
of this equipment the manufacturing costs of one rubber company 
were reduced at least 40 percent, and incidentally the quality ot the 
product was improved. Similarly, new types of equipment elmimated 
entirely the use of press cloths and effected a reduction in labor re- 
quirements of over 60 percent in the extraction of oil from linseed 
One large company was able to effect savings of 4 cents per busliel 
of linseed crushed by this process.^^ , •„ . . i x .i i f 

The elimination of operations is strikingly illustrated by the almost 
uncanny devices used to regulate and control the movements ot rail- 
road trains. Chief among the innovations m this field are the auto- 
matic block signal systems, automatic train control, interlocking plant 
and remote control, which make it possible to manipulate all main 
line track switches and signals in a given district or entire division 
by a centralized traffic control board.^" . . 

A similar development in the field of electric power transmission 
is the replacement of manually operated substations by automatic 
substations. During the 1920"s the installation of improved relay and 
protection equipment made possible an extremely rapid growth m 
unattended automatic substations. "Control relays are used to start 
the machines as load conditions require, to connect the unit to the 
line, and to shut it down when no longer needed. Protective relays 
furnish the protection desired against abnormal conditions such as 
short circuits, etc." Between 1920 and 1930 the number of unattended 
substations in five representative systems increased by 320 percent, 
while the number of attended substations rose by only 41 percent. 
By 1931 in one large system 72 percent of all substations were entirely 

Automatic. , • .i j • 

The electric light and power industry was also m the vanguard m 
the installation of automatic stoking and of powdered fuel equipment. 
By 1928 over 97 percent of the coal burned for the production ot elec- 
tric power was fired mechanically.'^ „,, . 

These instances illustrate the many fields m which operation- 
eliminating equipment has been developed, thus destroying ]ob 
opportunities. 
Increasing of Speed. 

The textile industries afford some of the most outstanding examples 
of speed of productive equipment being increased without a corre- 
sponding advance in the amount of labor required.^^ Each depart- 

".Ilnd. 

"National Recovery Administration, op cit., p. 14. , _ _ . .... iQr>Q r^rv 14_T^ 

w Recent Economic Changes in the United States, vol. I, Afst edition 1929 pp. 14-15. 

"Hearings before the Temporary National Economic Committee Part 30 p. --. 

wu. S. Bureau of Labor Statistics Monthly Labor Review August 1932 Labor Pro- 
ductivity and Displacement in the Electric Light and Power Industry, p. -o8. See 
also pp. 266-269. infra, for further discussion. 

» For further discussion see pp. 272-275, infra. 



IIQ CONCENTRATION OF PX'ONOMIC POWER 

ment in the cotton-textile industry has .-hown a marked increase 
in the speed of operation during the last 30 years. In the spin- 
ning department several makes of long-draft devices permit the 
use of a roving about twice as coarse as that used a few decades 
ago, "The coarser roving can be made on a large-size bobbin, which 
decreases the frequency of replenishing the supply packages on the 
spinning frame." The speed of operation is thus obviously in- 
creased. A similar result has been achieved in the spooling process 
through the use of an "automatic machine on which all the knots 
are tied by the machine itself," and of "a high-speed tube or cone 
winder * * * designed to permit a reduction in the time re- 
quired to tie a knot. Through the use of the first type of equip- 
ment the yarn is removed from the bobbin at a rate of approximately 
1,200 yards per minute as compared with an average speed of about 
200 yards per minute in 1910." High-speed warpers are in use 
which operate at 350 to 900 yards per minute, depending upon the 
type and size of the yarn, compared with an average speed of only 
50 yards per minute in 1910. In addition, these warpers have 
magazine creels which permit continuous operation, as well as tension 
devices which reduce end breakage.^" 

The manufacture of cigarettes affords two interesting examples 
of the way in which labor productivity may be raised by increasing 
the speed of operation.^^ The first is a new type of cutting 
machine which makes use of several knives set in a rotating arbor 
like the blades of a fan. These knife blades are automatically 
cleaned and sharpened as the machine operates, eliminating the 
necessity of stopping the machine frequently to change blades. The 
old type of machine, consisting of a reciprocating knife which 
operated in guillotine fashion, had to be stopped often for re- 
sharpening. 

Similarly, the cigarette-making machine has been improved. For- 
merly machines operated at the rate of about 700 to 800 cigarettes 
per minute; the newer type of apparatus throws out cigarettes at 
a speed of 1,200 to 1,500 per minute.*^^ 

In metallurgy the introduction of the sand cast method in cen- 
trifugal cast pipe manufacture has not only resulted in a marked 
reduction in equipment but in great savings of time reauired in the 
productive process.^^ 

All through the processes involved in the manufacture of automobiles 
and related equipment, changes have steadily taken place which have 
made it possible in numerous instances for a worker to operate a larger 
number of machines or to run machines at higher rates of speed.^* 



«• Boris Stern, "Mechanical Changes in the Cotton Textile Industry, lyiO to 1936," 
U. S. Bureau of Labor Statistics, Monthly Labor Review, August 1937, pp. 327-332. 

The weaving department has seen clianges of a like nature. From 1910 to 1936 the 
actual speed of an automatic 40-inch loom was increased from about 160 picks per minute 
to about 192. "This increase of approximately 20 percent is more than realized in the 
loom output, as the weaving efficiency of looms in 1936 averages from 2 to 5 iiercent 
greater tlian could have been obtained in 1910 on the same fabric." (Ibid, pp. 334-335.) 

«i For further discussion see pp. 262-264, infra. 

8» Chemical and Metallurgical Engineering, "Cigarette Industry Rules Out Rule of 
Thumb," vol. 43, No. 3, March 1936, pp. 128-131. 

«» Recent Economic Changes, pp. 14-15. 

«' A machine typical of many new types of apparatus in manufacturing is the so-called 
special piston machine used to finish-turn pistons in automobile manufacture. Eight 
spindles with a turning speed of 1,700 feet per minute revolve about the center axis of 
the machine. As the spindle containing pistons moves to position in front of an operator, 
the piston stops turning and the top fixture backs away from the piece. The operator 
removes the completed piston and places an unfinished one in its place. One man oper- 



CONCENTRATION OF ECONOMIC POWER 117 

EnlargemeTit-of Cd'pacity. 

An illustration of the enlargement of capacity is the growth in the 
size of rotary kilns in portland cement plants.®^ Between 1900 and 
1935 the average length of rotary kilns increased from 70 to 146 feet. 
In the decade, 1925-35, alone, the increase amounted to 23 feet. Today 
there are rotary kilns over 400 feet long in operation. Between 1902 
and 1935 the estimated annual capacity per kiln rose from 34,000 bar- 
rels to 299,000 barrels. The average kiln in 1935 could produce about 
60 }3ercent more clinker per yard than a kiln in 1920 and over twice as 
much as a typical kiln in 1910, and the larger units have not required 
a correspondingly larger labor force.^'^ 

In railway shops where labor is engaged in maintaining and build- 
ing rolling stock, technological innovations have resulted in material 
enlargements of capacity. A case in point is the multiple planer by 
which it is now possible to dress 15 or 20 locomotive side rods at one 
operation. Twenty years ago it was possible to machine only one side 
lod at a time. This same principle has been applied to drill presses, 
tapping machines, and treading machines, and the capacity of each has 
been greatly enlarged.^^ 

Enlargement of capacity has been of considerable importance m the 
textile industry. For example, it has been possible to wind a larger 
quantity of yarn by increasing the size of section beams in warping 
equipment.®^ 

In industrial-type gasoline locomotives, the average size of indi- 
vidual units sold in 1932-36 was 11.4 tons compared with 7.4 tons in 
1924—27. The average dipper capacity of power shovels sold to mining 
industries by a representative group of companies in 1932-36 was 3.28 
cubic yards compared with an average capacity of 1.73 cubic yards in 
1920-23.^^ 

The instances cited of the ways in which changes in individual single- 
function machinery can increase tiie productivity of labor are by no 
means inclusive. They are indicative of the variety of forms labor- 
saving mechanical changes take. 

A significant feature common to many of these innovations ife their 
ease of application. Few of them represent revolutionary changes in 
production techniques. They are more in the nature of improvements 
and refinements which today are constantly taking place throughout 
the industrial world. In the words of the Assistant Commissioner of 
the Work Projects Administration : 

Revolutionary changes which result in the direct displacements of workers are 
by no means a thing of the past. * * * Today, however, such changes rarely 
occur. The typical changes in industrial processes at the present time are the 
day-to-day improvements of all existing equipment. They are usually not spec- 
tacular and many of them require relatively small capital outlays.™ 

ates this machine, while heretofore one man was required to start and stop the machine 
as well as to set pieces in the jig. Compared with the speed of the newer apparatus of 
1,700 feet per minute, the cutting speed of the old machine was only 600 feet per minute. 
The operator of the new machine works continuously in removing and replacing work, 
and there is a finished piston turned out in less than 10 seconds, reducing materially the 
time from the old type of machine. (National Recovery Administration, op. cit., exhibit 
16, p. 13.) 

«= For further discussion, see pp. 250-253, infra. 

*«Work Projects Administration, National Research Project, Mechanization in the 
Cement Industry. 1939, p. 354. 

<" Hearinss before the Temporary National Economic Committee, Part 30, p. 16611. 

85 Boris Stern, "Mechanical Changes in the Cotton-Textile Industry, 1910 to 1936," U. S. 
Bureau of Labor Statistics, Monthly Labor Review, August 1937, p. 320. 

«* Works Progress Administration, National Research Project, Effects of Current and 
Prospective Technological Developments Upon Capital Formation, bv David Weintraub, 
1939. p. 3. footnote 4. 

'0 Hearings before the Temporary National Economic Committee, part 30, p. 17S25. 



118 CaNCENTRATION OF ECONOMIC POWER 

MANAGEMENT METHODS 

The displacement of rule-of-thumb by scientific methods in manage- 
ment has likewise increased labor productivity. Innovations in man- 
agement are of two general types : Those which relate to (1) the human 
factor and (2) the material conditions, of manufacture.'^ 

One of the problems in determining the effect on labor productivity 
of a given innovation relating to the human factor is the difficulty of 
finding instances in which the effect of only one innovation has been 
tested. For example, Taylor's classic example of increasing produc- 
tivity in the handling of pig iron from I214 to 47i/2 tons per day per 
worker illustrates several fundamentals of scientific management: (1) 
Selection and training of workers, (2) rest pauses, (3) incentive re- 
muneration and (4) study of the science of handling pig iron.'^^ Yet 
one constantly finds citations ascribing the entire increased produc- 
tivity in this case either to scientific selection or to resting." Accord- 
ingly, in selecting the following examples the attempt has been made 
to use only those which illustrate the effect of a given technique. The 
variety of management techniques which have increased the efficiency 
and productivity of labor is so great that only a few types are dis- 
cussed. This procedure can only serve to indicate the potentialities 
of management iimovations and their various combinations in increas- 
ing labor productivity. 

Scientific Management and the Human Factor. 

Selection and training of workers. — The selection of workers ac- 
cording to ability to perform a specific type of work is usually the 
starting point for every attempt at scientific management. Many 
types of examinations to determine an applicant's fitness have been 
devised ranging all the way from intelligence quotient tests to attempts 
to ascertain the applicant's attitude toward unionism. While the mere 
use of some type of selective system does not result in increased pro- 
ductivity, where a real relationship exists between the type of selec- 
tive test given and the work to be performed output per worker has 
risen. In the above example of pig iron handling, it would be diffi- 
cult to say with certainty what proportion of the increased produc- 
tivity was due to scientific selection. Though it was found that only 
one out of eight men was physically capable of handling 4Ti/2 tons a 
day, their selection was followed by training in the scientific handling 
of pig iron. Selection was important but not the sole cause of 
increased productivity. 

Some instances where training of workers appears to be the motivat- 
ing factor in increased productivity are cited by Burtt. Five girls who 
were inefficient in packing chocolates were given special training by 
an expert and improved their output by 27 percent. Two groups of 
novices were started in this work about the same time, the one gi-oup 
receiving training for C weeks. At the end of the period the un- 
trained group required about 25 percent more time to perform the 

'1 This classification of management methods is modeled in general after that advanced 
by the International Labour Office in its publication The Social Aspects of Rationalisation, 
Geneva, 1931, pp. 12-50. 

"Frederick W. Taylor, The Principles of Scientific Management, Harper & Bros., New 
York, 1929, pp. 40-04. 

"For example, International Labour Office, op. cit., p. 12; .Journal of the Society for 
the Advancement of Management, vol. 4, 1939, p. 43 ; Harold E. Burtt, Psychology and 
Industrial Efficiency, D. Appleton & Co., 1929, p. 174. 



CONCENTRATION OF ECONOMIC POWER 119 

same operations as the trained group. Similarly, novices in polishing 
silverware reduced their time per spoon a little more than 61 percent 
after only 1 day's training. Even workers with 9 years' experience 
reduced their working time per spoon from 23.2 to 39.5 percent."^ 

Working methods. — Working methods may be changed to increase 
productivity through: (1) The analysis and division of an individual 
work process into component motions, (2) the extension of this prin- 
ciple to a group of workers, i. e. chain work, and (3) the granting of 
rest pauses to eliminate declines in output because of fatigue. 

Examples of the effectiveness of motion studies in eliminating excess 
movements and increasing output are legion, but one or two will suffice 
to indicate their potentialities. Burtt reports that — 

111 a candy factory the process of dipping Involved three changes of direction and 
stopping the hand three times. A new method was devised which made the 
motion rhythmical and circular although the path covered was somewhat longer. 
The result was an increase of 27 percent in output." 

Economy of motion in applying gummed stickers to packages (plus 
the simple expedient of having the sticker manufacturer powder them 
slightly so they did not stick together) has increased output 30 to 50 
percent on this particular operation for packagers of confectionery 
and hosiery and other large users of gummed labels.'*^ 

Another device, related to rhythm and motion studies, is the use of 
recorded music, amplified throughout the plant, to induce greater 
output both by developing greater rhythm and decreasing the boredom 
of work. In a recent British study it was found that during the time 
music was played the increase in output varied from 6.2 to 11.3 percent, 
while the total daily output increased by 2.6 to 6.0 percent. The great- 
est increase in output was obtained when music was played for 75 
minutes about the middle of the work-spell, but the most popular 
arrangement was the introduction of music during alternate half- 
hours throughout the work-spell." 

One significant application of the chain work technique is the 
straight-line system of production in the cotton garment industry. 
Introduced in 1932, it replaced the system in existence since the begin- 
ning of faclory manufacture in which the bundle, containing scores of 
layers of cloth cut according to pattern in one operation, was the unit of 
work. Under the bundle system each worker performed certain opera- 
tions on all garments in the bundle, which was then turned over to 
others for subsequent operations until all garments were completed. 
Under the straight-line system the machines are arranged in the order 
of operations and as each worker completes her operation on an indi- 
vidual garment or part of a garment she places it within reach of the 
next operator. This eliminates the carrying of bundles from one end 
of the shop to the other and the lifting of tons of garments by the 
operators throughout the course of a day. Since the straight-line 

'* Burtt, op. cit.„pp.'63-64. 

"Ibid., p. 123. 

'8 C. A. Barnes, "Motion Economy on Labeling Operations," Factory Management and 
Maintenance, vol. 97, No. 12, December 1939, pp. 59-60. 

" "Different types of music were found to have slightly different effects on output. 
The average difference was 5.6 percent, but in individual cases it approximated to 10 
percent. Although most of the workers responded to the music by an increased output, 
-which in some cases was very marked, there were Isolated instances in which the music 
seemed to have little or no effect on production. The effect in general seemed to be 
directly related to the amount of boredom experienced by the different individuals." 
(Medical Research Council, Industrial Health Research Board, Fatigue and Boredom 
in Repetitive Work, Repor' No. 77, by S. Wyatt and J. N. Langdon, H. M. Stationery 
Office, London, 1937, p. 73.) 



120 



CO>TENTRATI0N OF ECONOMIC POWER 



system aims at the minutest possible subdivision of operations, the 
operators develop great speed because of the small number of work 
elements/^ 

Lack of suflScient records makes it difficult to evaluate the effects of 
the straight-line system on productivity. One engineering firm licensed 
to install the system claims that a certain garment company increased 
production 50 percent, cut manufacturing costs 25 percent, reduced 
clerical vrork 33i/^ percent, and inspection costs 60 percent, saved 25 
percent of the floor space, and vastly reduced inventory of goods in 
process. Since it is not known precisely how efficient or inefficient the 
plant was prior to the use of the straight-line system, these percentages 
cannot be taken at their face value. In one plant included in the 
Bureau of Labor Statistics study, however, when the entire assembly 
department was put on the line a reduction of 34.2 percent in time 
required under the bundle system took place. ^^ 

Rest periods are a third expedient of increasing productivity simply 
by changes in working methods. Burtt reports some experiments with 
steel riveters whose average production was 600 rivets a day. When 
they were given a 2 minutes' rest after every 10 rivets, thus spending 
over 5 hours of the 10-hour day in rest, their production went up to 
1,600 rivets.^° Numerous instances in which rest pauses have contrib- 
uted to productivity could be cited. The following table gives con- 
cisely the effects of various types of rest pauses in a number of simple 
tasks. 

Effect of rest pauses on output for women ^ 



Nature of job 


Type of pause 


Percent of change in 
output 


Labeling 


10 minutes in a. m 

do 


13 
5 
8 
11 
13 
4 
4 

-3 
2 
2 

t 








Tying small packages 


do 






. do .-- 






5 minutes per hour 

7 minutes in a. m 

do- .. 


(Average; 6.2 after 


Folding handkerchiefs 


Some months. 








do 




Hemstitching 


do 




Folding handkerchiefs _... ■... 


10 minutes in a. m 

do 








Stamping discs 


10 minutes each spell . . 
5 minutes per hour — 
7 minutes each spell 


-Average: 2.8 im- 






Light work .. 











'H. M. and M. D. Vernon, "Five-hour Spells for Women with Reference to Rest Pauses," Industrial 
Fatigue Research Board, Report No. 47, H. M. Stationery Office, London, 1928, p. 16. 

Rediiction of hours. — Output per man-hour and even total output 
have frequently been increased by a reduction in the number of 
hours worked per day. An increase in hourly performance naturally 
follows such reductions since monotony, fatigue, and exhaustion not 
only slow up tlie workers' productivity but also cause considerable 
loss due to breakage of machinery or tools. 

However, there undoubtedly exists a point of diminisliing returns 
for successive reductions in working time. It is quite likely, for 
example, that a reduction from an 8- to a 6-hour day would have 

■'s U. S. Bureau of Labor Statistics, Bulletin No. 662, Productivity of Labor in the 
Cotton Garment Industry, by N. I. Stone, Alfred Cahen, and Saul Nelson, 1938, pp. 38-39. 
™Ibid., pp. 50^51. 
»" Burtt, op cit., p. 174. 



CONCENTRATION OF ECONOMIC POWER 



121 



less effect in, decreasing fatigue and stimulating productivity than a 
reduction from a 10- to an 8-hour day. Also, the effects of reductions 
in hours are by no means immediate ; in some cases improvement in 
output comes only after several months have elapsed on the new 
schedule. 

A specific example of the productivity-increasing effects of a shorter 
working day is afforded by the iron and steel industry. The United. 
States Bureau of Labor Statistics made this observation in its analy- 
sis of the change from the 10- and 12-hour day to the 8-hour day in 
the steel industry : 

Before this change took place it was confidently expected by many that 
there would be a considerable increase in labor cost because of the increase 
in the number of men required to operate the furnace. * * * 

Theoretically, the substitution of the 8-hour day for the 12-hour day would 
have no effect on productivity ; that is, each position requiring two men at 
12 hours each would require three men at 8 hours each and the output per 
man-hour of labor would remain the same. In actual practice, of course, it 
would be expected that the output per man-hour would be somewhat higher 
in the latter case, for it is evident that a man can work at higher speed- for 8! 
hours than he can for 12 hours. But the actual results in the blast-furnace 
industry following 1923 far exceeded anything that might have been expected. 
There are-numerous cases of plants in which, within a year after the change 
was made, the total labor force was back again at same number of men that 
had been employed under the 12-hour system." 

The effect of shorter hours on output is shown below for three 
types of work. 

Output in munitions factories ^ 



Work 


Average 
weekly 
hours 


Relative 
output 
per hour 


Relative 

total daily 

output 




f 52.2 
\ 56.5 
51.2 
( 66.2 
\ 54.8 
I 45.6 
( 64.9 
\ 54.8 
I 48.1 


100 
122 
139 
100 
134 
158 
100 
121 
133 


100 
106 




122 
100 
111 


40 women milling a screw thread 


109 
100 
102 




99 



1 Burtt, op. eit., p. 167. 

Incentive systems of remuneration. — ^Workers are frequently in- 
duced to take an interest in increasing output through payments 
based on piece rates or a bonus in addition to the normal wage. The 
bonus may be given for quality or quantity of goods produced, speed 
and regularity of work, savings effected in raw materials, power, 
fuel, etc. In certain cases, instead of receiving a bonus for increased 
output, the worker is forced to increase his output so that his net 
income may reach a normal figure because the base to which bonuses 
are applied is abnormally low. The stimulation of productivity 
through incentive remuneration is limited unless the worker can feel 
sure that demonstrating his ability to produce more will not result in 
lowering his basic wage. 



^ U. S. Bureau of Labor Statistics, Bulletin No. 
chant Blast Furnaces, 1928, pp. 46-47. 



174, Productivity of Labor in Mer- 



122 CONCENTRATION OF ECONOMIC POWER 

Wage incentive plans adopted by the Bethlehem Steel Corpora- 
tion prior to 1928 apparently increased both labor productivity and 
earnings. 

In relining open hearth steel ladles costs have decreased 14.4 pei-cent while 
the average earnings of the employees increased 21.3 percent. On track repair 
VFork costs vpere decreased 34 percent while the average earnings of the em- 
ployees increased 25 percent. In making repairs to soaking pits costs were 
decreased 20.4 percent while the average earnings of the employees increased 
19 percent.'^ 

A successful bonus plan has been in operation for 10 years for 
packing and shipping employees of the Detroit Steel Products Co. 
"For 18 to 181/2 units of work per man-hour, a bonus of 3 percent is 
added to each base rate, and a varying percentage is paid for 
increases above this level, till at 21 units or over 15 percent bonus is 
paid." Though the workers in this department have earned "close 
to 15 percent above base pay", "costs have been 10 percent lower 
over the period." ^^ 

Piece rates as an incentive to stimulate output have been widely, 
used in this country.^* But at the present time a shift back to ration- 
alized hourly rates appears to be taking place. This can be ascribed 
in large part to (1) the increasingly automatic nature of modern 
machinery and (2) the provisions of the Wages and Hours Act which 
discourage the payment of piece rates. 

The Minnesota study of changes in job requirements described the 
first factor ss follows : 

The reasons for this general movement are to be found in the changing 
nature of industry itself. After bonus or piece-rate systems had prevailed for 
many years, rates and standards could be set for most operations on the basis 
of careful time and motion studies. Moreover, with continuing technological 
advance and refinement, speed of performance became less and less a matter of 
choice for the individual worker. Thus, once a definite measure of a day's 
work on a particular operation had been established, management found that 
a flat hourly rate would eliminate costly bookkeeping, as well as tension on the 
part of the worker, without in any way decreasing the speed or quality of the 
work.*^ 

/Scientific Management and Material Conditions of Manufacture. 

The better arrangement of work places, tools, materials, etc., in 
manufacturing plants has frequently resulted in considerable in- 
creases in labor productivity by reducing the amount of handling 
required. Better lighting and ventilation- also contribute to produc- 
tivity. 

A Bureau of Labor Statistics study of the tire industry lists the effect 
on labor of various technological changes. The rearrangement of the 
curing room to take care of increased production saved 173 man-hours 
per day (at full capacity), or 22 men displaced. The moving of the 
preparation conveyor in the tube room and the rearrangement of the 
service conveyor and automatic soapstoning displaced 2 girls per shifty 
saving 48 man-hours per day. The consolidation and rearrangement 

«2 International Labour Office, The Social Aspects of Rationalisation, Geneva, 1931, p. 37. 

^ W. C. Owen, "Bonus Drops Packing Costs 10 Percent," Factory Management and 
Maintenance, vol. 97, No. 3, March 1939, pp. '51-52. 

"^ The National Industrial Conference Board reported that in 1928 less than 15 per- 
cent of a sampling of plants employing more than 1,500 workers paid their employees 
on a straight-time basis. (Systems of Wage Payment, 1930, pp. 6-7.) 

'^ Work Pro.iects Administration, National Research Project, and Employment Stabili- 
zation Research Institute, University of Minnesota, Changes in Machinery and Job Re- 
quirements in Minnesota Manufacturing, 1931-36, by C. A. Koepke and S. T. Woal, 1939, 
p. 24. 



CONCETs'TRATION OF ECONOMIC POWER 123 

of the cutting and rerolling departments displaced 14 girls and saved 
112 man-hours per day.^"^ 

One department, which employs 30 to 40 girls, of the Maiden Form 
Brassiere Co. is engaged solely in clipping off the tiny ends of thread 
left after each sewing operation. When a scientific study of this de- 
partment was undertaken, a total of 94 operations and 52 transporta- 
tions were performed by each worker with both hands on each garment 
with an average of 27 threads to clip. Then scientific, semi-circular 
workplaces were devised, the girls sitting on posture chairs with adjust- 
able footrests to take care of differences in height. These replaced 
ordinary tables and chairs and eliminated the hunched-up position the 
girls formerly assumed at work. The operations performed by both 
hands were reduced to 47, a decrease of exactly 50 percent, and the 
transportations by both hands were reduced by 51.9 percent. Produc- 
tion was accordingly increased with lower unit costs.*" 

The ordinary flat-top bench long used in packing one of the standard 
products of the Bristol Co. was replaced by a semi-circular bench with 
removable partitions to divide it into compartments of any desired size. 
The compartments all converge to the working space of about 30 by 12 
inches, eliminating excessive reaching. This scientifically devised 
workplace has cut direct labor costs 40 percent, the operation is much 
less fatiguing, and operators' earnings have not been reduced.** 

A program of lighting modernization by a large manufacturer of 
church, school, and theater furnishings has resulted in an average pro- 
duction increase of 5 percent, with some departments increasing output 
by 15 percent. The new equipment cost about twice as much as that 
replaced but the increased output due to better lighting has meant a 
saving of $250 per day, or about $75,000 per year.*^ 

''Prior to the installation of air-conditioning equipment, a large 
printing company was compelled to stop work on days when natural 
humidit}' and temperature exceeded certain limits. With the ability to 
maintain continuously the desired temperature and humidity condi- 
tions, production is uninterrupted, and the uniformity and quality of 
the work have been improved. Of equal or greater importance is the 
fact that the ease of operating the existing plant at full capacity elimi- 
nated the need for increasing the size about 50 percent to take care of 
production requirements." ^'' 

s« U. R. Bureau of Labor Statistics, Bulletin No. 585, "Labor Productivity in the Auto- 
mobile Tifp Industry." by Boris Stern, p. 25. 

"J. J Baer, "More for the Company. More for the Girls," Factory Management and 
Maintenance, vol. 97, No. 2, February 1939, pp. 54-55, 114-116. 

«3 Arthur R. Baldwin, "Motion-Economized Bench — Costs Off 40 Percent," Factory 
Management and Maintenance, vol. 97. No. 11, November 1939, p. 33. 

^ Factory Management and Maintenance, vol. 97 No. 11, November 1939, p. 70. 

9" Ibid., p. 74. 



CHAPTER II 
THE EFFECTS OF LABOR-SAVING TECHNOLOGY 

TECHNOLOGY AS A CAUSE OF UNEMPLOYMENT 

THE GENERAL EXTENT OF TECHNOLOGICAL. DISPLACEMENT 

The extent of technological unemployment is difficult to appraise. 
Sufficient specific instances of displacement of workers by tecl nological 
improvements are marshaled together in this chapter, it is hoped, to 
leave no doubt that displacement has occurred and to give some per- 
spective of its impact upon our economy. 

Many estimates have been made of the amount of technological 
unemployment. In February 1940, the Congress of Industrial Organi- 
zations estimated on the basis of admittedly rough data, that "Unem- 
ployment of between two and one-half million and three million can 
be attributed to technological changes affecting the various kinds of 
employment." ^ ' 

Stress has frequently been laid upon technology as a principal cause 
of unemployment prior to 1929. David Weintraub estimated that for 
manufacturing as a whole in the period 1920-29, 32 men out of every 
100 required in 1920 were made unnecessary by increases in output 
per man, but of that number increases in the total output absorbed 27. 
His figures show that improvements in efficiency in this period dis- 
placed 2,832,000 men, or 416,000 more than were reabsorbed into manu- 
facturing by the shortening of the work week and the increase in total 
output. 

Weintraub estimated that 345,000 workers were displaced by in- 
creased technological and managerial efficiency in steam railroads 
during the same period. Technological displacement in the bituminous 
coal and anthracite industries was placed at 95,000 by his calculations. 
For these four major industries — manufacturing, steam railroads, 
bituminous coal and anthracite mining — Weintraub estimated an em- 
ploymeii't decline during 1920-29 of 3,272,000, resulting from changes 
in productivity. Meanwhile, 2,269,000 men were required to handle 
increases in volume of output, leaving a net decline of employment in 
these industries of 1,003,000.^ 

A com]>arison of separation rates in nianufacturing industries from 
1899-1929 made by Frederick C. Mills shows that during each 2-year 
period between 1923 and 1929. 49 workers out of every 1,000 withdrew 
from, or were forced out of, the industry in which they were working, 
compared with 21 men out of every 1,000 during each 5-year period 
from 1899 to 1914. This increase in the separation rate at a time of 

^ The Economic Outlook, February 1940. 

» David Weintraub, "The Displacement ,of Workers Through Increase in Efficiency and 
Their Absorption by Industry. 1920-31," Journpl of the American Statistical Associa- 
tion, December 19.32, pp. 396-397 

125 



126 CXDNCENTRATION OF ECONOMIC POWER 

comparatively intense industrial activity is probably due to technologi- 
cal displacement. Mills also found that between 1923 and 1929 one 
worker out of 20 was compelled to seek employment in a new manu- 
facturing industry or in a non-manufacturing industry every 2 years, a 
turn-over which at that time could not be ascribed to any major de- 
crease in industrial activity. =* 

Labor-saving effected by technology during the last decade may be 
estimated from the indexes of production and productivity given in 
table 1. Manufacturing work which required 100 workers in 1923 
could have been performed by 76 workers in 1929 and by 57 in 1939, 
assuming no change in weekly hours. Work which required 100 men 
in 1923 could have been performed in bituminous coal mining by 93 
workers in 1929 and by 70 in 1939 ; in anthracite mining by 104 workers 
in 1929 and by 58 in 1939 ; and in steam railroads by 85 workers in 1929 
and by 65 in 1939. again assuming no change in weekly hours. 

Since each of these basic fields was characterized not only by an 
increase in labor productivity but also by a decline in production 
between 1929 and 1939, the aggregate number of man-hours was 
materially reduced. Total man-hours in manufacturing fell from 
19,888,000,000 in 1929 to 14,948,000,000 in 1939; in bituminous coal 
mining total man-hours fell from 916,250,000 to 508,300,000; in an- 
thracite mining from 271,853,680 to 104,463,420; and in steam railroads 
from 4,103,000,000 to 2,325,000,000. If no change in weekly hours had 
occurred in the last decade, these reductions in man-hours would have 
meant a displacement of 2,078,600 wage-earners in manufacturing, 
761,678 in steam railroads, 204,143 in bituminous coal mining, and 
87,762 in anthracite mining, or a total in all four fields of 3,132,183 
wage-earners. 

Since average weekly hours decreased over this period in each of 
the fields — 17.7 percent in manufacturing, 31.2 percent in steam rail- 
roads, 29.4 percent in bituminous coal mining, and 24.1 percent in 
anthracite mining — the actual decline in the number of wage-earners 
was considerably less than this potential displacement. The number 
of wage-earners fell 723,700 in manufacturing, 729,000 in steam rail- 
roads, 98,000 in bituminous coal mining, and 70,500 in anthracite 
mining, or a total decline of 1,621,200 wage-earners. 

It is impossible with existent data to determine precisely the role 
of technology in causing potential and actual reductions in employ- 
ment where a decline in production has occurred. But a comparison 
over a period of years of changes in the total number of man-hours 
required to produce a comparable output demonstrates rather clearly 
the extent of technological unemployment. This method ig best suited 
to a comparison of years of relatively high output, since a marked 
advance in productivity can usually be expected to occur upon a com- 
paratively small increase in production when production is at a rela- 
tively low level during the base period. Consequently, comparisons 
for specific industries in this analysis are made between 1929 as a 
base and a year of approximately comparable production in the late 
thirties. 



8 Frederick C. MUls, Economic Tendencies in the United States, National Bureau of 
Economic Research, New York, 1932, pp. 419-423. 



CONCENTRATION OF ECONOMIC POWER 



127 



Sufficient data on production and man-hours are available for a 
comparison of this nature for 12 industries. The percentage change 
from 1929 to the year in the late thirties during which production 
approached most closely the 1929 level strikingly reveals the influence 
of technology upon the total amount of labor required. In 4 of these 
12 industries the decline in man-hours amounted to over 30 percent; 
in 5 more the decline was from 20 to 30 percent ; in 2 it w^as from 10 
to 20 percent; and for 1 — ^bread and other bakery products — man- 
hours dropped only 0.5 percent. The percentage change in both pro- 
duction and man-hours in these industries for years of approximately 
comparable production is bhown in chart VII, table 7. 

Table 7. — Percentage change in production and man-hours in 12 industries 

[1929 = 100] 



Year of approxi- 
mate comparable 
output in the 
1930's 


Industry 


Produc- 
tion 


Man- 
hours 






99.9 

9411 
97.7 
102.3 
104.6 
105.9 
102.3 
104.3 
99.4 
102.5 
102.6 


87.9 


1935 


Pptrnlpiim refining 


66.4 


1938 -.. 


Chemicals 


77.0 




65.2 


1939 


Paints and varnishes 


78.1 


1937 


Cotton goods 


79 3 


1938 - - 




76.8 






78.3 


1937 


Newspapers and periodicals 


84.5 


1937 

1937 


Bread and other bakery products 


99 5 






64.4 






67.7 









Source: 

Production: Victor Perlo and Witt Bowden, "Unit Labor Cost in 20 Manufacturing Industries, 1919- 
1939," U. S. Bureau of Labor Statistics, Monthly Labor Review, July 1940, p. 37. 

Man-hours: National Research Project figures (Production, Employment, and Productivity in 59 Man- 
ufacturing Industries, Part II) extended by the use of U. S. Bureau of Labor Statistics data. 

The significance of these comparisons lies not only in the sharp 
decreases in the number of man-hours required to produce a generally 
comparable amount of goods, but also in the wide range of activities 
represented. The substantial decreases in man-hours which took 
place in all but one of these industries between years of relativelj^ 
high output indicate that the adoption and utilization of labor-saving 
innovations has by no means been confined to any small segment of 
the economic system. 

A comparison of to*^ iimii-hour requirements in years of approxi- 
mately companible output overcomes the difficulty of allowing for 
changes in labor productivity at varying rates of production. Un- 
fortunately, production is rarely at a level exactly comparable with 
a preceding year. 

If this method is to be used, comparisons must be extended to years 
of approximately comparable rates of output. Wliat constitutes ap- 
proximately comparable production depends upon an arbitrary deter- 
mination. But for practical purposes two given years are regarded as 
approximately comparable when their rates of production are within 
10 percent of each other. 

If production is at a generally similar level in the two years com- 
pared, but in the latter year is a few percent above the level of the 
former year, that increase in production might result in a consider- 



128 



CONCENTRATION OF ECONOMIC POWER 



Chart VII 

CHANGES IN PRODUCTION AND MAN-HOURS 

12 INDUSTRIES IN THE UNITED STATES 

(FROM 1929 TO YEAR OF APPROXIMATELY COMPARABLE 
OUTPUT IN THE THIRTIES) 



IRON 6 STEEL (1937) 



CHEMICALS.(l93e) 



FERTILIZER (1937) 



PAINTS a VARNISHES 
(1939) 



COTTON G00PS(I937) 



BOOTS 6 SHOES (1938) 



PAPER 6 PULP (1938) 



NEWSPAPERS & 
PERIODICALS 
(1937) 



BREAD a OTHER 
BAKERY PRODUCTS 
(1937) 



CONFECTIONERY 
(1936) 



ICE CREAM (1938) 



PRODUCTION 
MAN 



INDEX (1929-100) 
20 40 60 80 100 120 







PRODUCTION 
MAN 




PRODUCTION 
MAN-HOURS 



Source : Ta1>le 7 



CONCENTRATION OF ECONOMIC POWElt 



129 



able advance in labor productivity, particularl}^ if the comparison 
is made between years of relatively low output. The selection of 
sets of years for comparison in which production is lower in the 
latter than in the former year would tenet to prevent the appearance 
of exaggerated productivity increases. The difficulty then arises that 
part of any decline in the amount of labor required could be ac- 
counted for by the drop in production. Adjustments may be made 
for this factor, however, by reducing the total labor requirements 
m the former year by the percentage decline in production between 
the tAvo years.* 

In manufacturing, production in 1939 was at almost exactly the 
same level as in 1929. In steam railroads and bituminous coal mining. 
1930 and 1937 were years of approximately comparable output. 
Tiierefore, comparisons betAveen these sets of years highlight the 
extent of technological displacement in these three fields and ere set 
forth in table 8.^ 

Iablb 8. — Technological displacement in manufacturing, steam railioads, and 
bituminous coal mining, 1929-39 

MANUFACTURING 





Year 


Production 
(1923=100) 


Man-j-ears i 


Adjusted 
man-years '' 


1926 


130.1 
129.5 
-.6 
-.5 


8,368.800 

6, 290. 18f) 

-2.078.614 

-24.8 


8.326,956 


1339 .-.- 




-2,036,770 
24 5 


Pe''ceiit change 







STEAM RAILROADS 



1930 

1935...- ^.. 

Actual change.. 
Percent change. 



Actual change . 
Percent- change - 



-5.6 
-6.3 




1,464,531 

1,168,981 

-295,550 

-20.2 



BITUMINOUS COAL MINING 




440.700 , 
379,449 I 

-61,251 
-13.9 ! 

I 



419,987 
379. 446 
-40, 53S 



1 Based on average weekly hours: 45.7 in manufacturing (1929); 43.2 in steam railroads (1930); and 33.5 
in bituminous coal mining (1930). 

' Adjusted for the percentage declines in production: 0.5 percent in manufacturing, 6.3 percent in steam 
railroads, and 4.7 percent in bituminous coal mining. Table 1, supra. 



Source : Computed from U. S. Bureau of Labor Statistics, Monthly 
tember 1940, "Wages, Hours, and I'roductivitv of Industrial Labor, 
Witt Bowdeu. 



Labor Review. Sep- 
1009 to 10:);. •• b> 



Had there been no changes in weekly hours, the 1929 output in 
manufacturing could have been produced in 1939 with 2,036,770 
fewer wage-earners, a 2'1.5 percent reduction in a decade. The 1930 
output could have been produced in 1937 with 295,550 less wage- 
earners in steam railroads "^ and 40,538 fewer in bituminous coal 



<For example, if man-years are at 8,000,000 in the former and 6.000,000 in the latter 
yesr and production falls 10 percent between the years, the amount of technolopica! dis- 
plasenient would be 1.200.000 man-years (8,000,000 man-years minus 10 percent, or 
7,200,000, minus the labor engaged in the latter year, 6,000,000 man-years), 

' Derived from the production and man-hour data in table 1. p. 90. 

* Changes in total labor expenditures in steam railroads are occasionally attributed to 
fluctuation.s i,i the proiwrtiou which maintenance workers are of all railroad eaiployees, 
1 t , to changes in management policies toward the ,imonnt of maintenance veqiTiz-sd for 



1551— 41— No. 22- 



-10 



130 



CONCENTRATION OF ECONOMIC POWER 



percent, respectively, 



mining/ reductions of 20.2 percent and 
within only 7 years. 

The seriousness of a technological displacement of. over 2,000,000 
adjusted man-years in manufacturing during one decade, of nearly 
300,000 in steam railroads, and of over 40,000 in bituminous coal min- 
ing in only 7 years requires no elaboration. 



DISPLACEMENT IN SPECIFIC INDUSTRIES 

Certain industries present striking case histories of the impact of 
technology upon employment. For example, the electric lamp in- 
dustry illustrates the increase of labor productivity through the in- 
troduction of revolutionary new mechanical techniques. In making 
and assembling the parts of an electric lamp there have been two 
developments of outstanding importance. First is the group or unit 
system of manufacture, which coordinates and synchronizes such 
related parts of a production unit as the bulb-making machine, the 
hot belt conveyor and the burn -off machine. Second is a widely 
used mechanism consisting of a turret or spider rotating on a verti- 
cal axis operated by electric motor wdiich performs automatically 
many of the functions formerly done by manual labor.^ 

How technological improvements in this industry, of which only 
two have been mentioned, have reduced total man-hours needed in 
electric lamp assembly plants during the period 1920-31, in the face 
of a rapidly growing output, is shown in the following table. It is 
especially significant because the major portion of the labor in the 
electric lamp industry is employed in lamp assembly plants. 

Production and employment in elertric-lnnip-assembhi plants^. lO^O-.lt" 



Year 


Production 

(number of 

lamps)' 


Employment 
(number of 
man-hours) 


Year 


Production 

(number of 

lamps) 


Employmen 
(number of 
man-hours) 


1920 


362,140,000 
242,515,000 
311,265,000 
404, 226, 000 
435, 172. 000 
459, 275, 000 


36, -145, 000 
21,710,000 
24, 549, 000 
26,S21,000 
22,079,000 
19,753,000 


1926 


482,455,000 
544, 512, 000 
556, 953, 000 
643, 957, 000 
553,199.000 
503, 350, 000 


17,576,000 


1921 


1927 


17, 922, 000 


1922 


1928 


15, 976, 000 


1923 . . 


1929 

1930 


16,003,000 


1924 


13,424,000 


1925 


1931 


11,448,000 







» Ibid., table 5, p. 39. 

The upward trend of production until 1929 was accompanied by 
a marked decline in the number of man-hours. Only 16,003,000 man- 
hours were required to produce 643,957,000 lamps in 1929; in 1920 
over twice that number of man-hours were required to produce a little 
more than half as many lamps. It is estimated that 16,049 employees 
were displaced bv technological advances in electric-lamp assembly 
plants between 1920 and 1929.^° 



a specific year. However, this factor does not affect the comparison of 1930 and 1937 
because the proportion of maintenance workers to all railroad employees was almo.st the 
same for the two years, being 49.9 percent in 1930 and 48.6 percent in 1937. (Interstate 
Commerce Commission, Wage-Statistics of Class I Steam Railways in the United States, 
issues for the years indicated.) 

' This displacement does not include the amount of labor lost in bituminous coal from 
the production decreases due to the substitution of other fuels. 

» U. S. Bureau of Labor Statistics, Bulletin No. 593. 1933, "Technological Changes and 
Employment in the Electric Lamp Industry," by Witt Bowden. 

10 Ibid., p. 43. 



CONCENTRATION OF ECONOMIC POWE^ 



131 



The growing importance of labor-saving machinery in this indus- 
try is strikingly shown by a comparison of labor productivity from 
1916 to 1932 in the production of glass bulbs of a standardized type, 
the 25-watt bulb (table 9).^^ 



Table 9. — Estimated changes in the productivity of hand and machine 
selected plants m making glass bulbs for 25-watt electric lamps 


ahor in 




Year 


Output per 

unit-hour 

(bulbs) 


Output per man- 
hour 


Method of produqticn 


Bulbs 


Index 
(output 
in plant 

A = 100) 


Hand production: 
Plant A 


1916-19 
1923' 
1925' 
1925 1 

1925 

1925 3 

1931-32 « 

1931-32 ! 

1932 9 


118.2 

126.5 

406.2 

1, 870. 6 

2, 139. 5 
2, 336. 9 
3, 537. 7 
6, 242. 2 
20, 762. 


52.5 
56.2 
116. 1 
801.8 

1,284.2 
1. 699. 3 
2, 573. 7 
4.538.9 


100.0 


PlantB 


107.0 




221.1 


Automatic machine (Empire F) 


1, 527. 2 


Automatic: 


2.446.1 




3, 236. 8 


Do 


4, 902. 3 




8,645.5 













17 months. Ml months. s 16 months. 

' 4 months. < 14 months. ' 6 months. 

Source: U. S. Bureau of Labor Statistics, Bulletin No. 593, 1933, "Technological Changes and Employ- 
ment in the Electric Lamp Industry" by Witt Bowden, table 8, p. 45. 

Just as the electric lamp industry shows the effect of revolutionary 
mechanical techniques, the automobile tire industry is an example 
of the increase in labor productivity through the gradual but constant 
introduction of numerous, small, detailed innovations in the produc- 
tive process. Major changes in the processes of tire manufacture 
have been few and far between. Up to 1933 there had occurred only 
one major change in the manufacture of pneumatic tires: The re- 
placement of the core process by the flat-drum process, a change 
which by 1927 had taken place in all the major plants. A list of 
some of the typical minor changes, which in their cumulative effect 
have raised labor productivity extensively in tire-making, is pre- 
sented in appendix G. 

The trend of production in the tire industry up to 1929 was mark- 
edly upward, but total man-hours remained remarkably stable. The 
difference between the actual changes in total man-hours worked 
and the increases or decreases in man-hours brought about by changes 
in total output represents the reduction in total labor time caused by 
technological change or the total volume of labor displaced. A 
computation of the amount of labor displaced by technology from 
1922-31, together with output and man-hour figures, is given in 
table 10. 



" The indexes do not reveal the full extent of the change in productivity because 
comparable figures on the most recent improvement — the so-called ribbon bulb machine — 
were not available for the study. The output of a single unit for 24 liours runs above 
500,000 bulbs, but the exact number of man-hours was not known and therefore the index 
of man-hour output is omitted. (Ibid., table 8, p. 45.) 



132 



CONCENTRATION OF ECONOMIC POWER 



Table 10. — Actual production and volume of technological lalor displacement in 
6 representative tire plants', 1922 to 1931 



ACTUAL PRODUCTION 






Year 


Total output 
rounds 


Total man- 
houfs 


Teclinological 
displacement 
Id man-houis 


1922 - 


295, 222, 000 
324,544,000 
357. 863, 000 
466, 238, 000 
501,513,000 
599, 642, 000. 
752, 333, 000 
801,725,000 
684. 645, 000 
648, 648, 000 


26, 165, 000 
26,431,000 
28, 161, 000 
33, 860, 000 
30, 427, 000 
31, 867. 000 
35, 886, 000 
35, 167, 000 
26, 166, 000 
21,150.000 




1923 


2, 122, 000 


1924 


892, 000 


1925 -- 


2.171,000 


1926 


5, 573, 000 


1927 


3, 775, 000 


1928 - 

1929 


3, 265. 000 
2, 886, 000 


1930 


3,866,000 


1931 _ 


3, MO, 000 
28,189,000 











' Result obtained by subtracting total decrease from total increase. 

Source: U. S. Bureau of Labor Statistics, Bulletin No. 58";, 1933, "Labor Productivity in the Auton.obile- 
Tire Industry," by Boiis Stern, p. 17. 

In both of the two major segments of manufacturing, steel and tex- 
tiles, marked advances in labor productivity have recently taken place. 

The stee] industry between 1937 and 1989 was characterized by an 
increasing displacement of labor, due in large part to the introduc- 
tion of continuous strip mills. In 1937 the m,aximum work week 
in the steel industry was reduced from 48 to 40 hours in certain 
companies and. consequent!}" 58,690 more workers were required 
in September 193^ to produce a tonnage equivalent to that of August 
1936. This increased employment, however, was nullified during 
the next 2 years by technological advance. The number of wage 
earners in the 2-year period from September 1937 to September 1939 
decreased from 503,000 to 415,000, an elimination of more than 88,000 
woi'kers, or a decline of 17.5 percent in the number of steel workers 
needed to produce an equivalent tonnage. Compared with August 
1936, when the maximum hours were 8 per week higher, the number 
of steel workers in September 1939 was 30,000 less.^^ 

Some specific examples of the impact of new steel technology were 
presented by Philip Murray before the Temporary National Eco- 
nomic Committee in its hearings on Teclmology. Certain of his more 
striking examples are given below in summary form: 

(1) A large steel company invested $750,000 in improvements in its open 
hearth department. Sixty-two employees weve eliminated as a result of the 
improvements which effected an annual pay roll saving of $118,000. In addition, 
savings in coal consumption and in other lines amounted to $257,000 a year. 
Thus the company's annual savings in cost totaled $375,000, or enough to write 
off its $750,000 investment and interest in less than 2V2 years. The technical 
improvements enabled the men remaining on the pay roll to produce in 133 
days the same amount of steel that formerly rtxiuired 200 days. 

(2) One of the industry's largest wire companies through technical improve- 
ments increased the speed of its cold wire drawing machines from 11? revolu- 
tions to 150 j>er minute. Production rose from 2,185 pounds per 8-hour turn to 
3,0ft0 pounds per turn, or an increase of 37 percent. In another case, a wire 
company in its fine wire department increased tlie revolutions of its machines 
from .350 to 450 per minute, or 28 per<.'ent. Production rose from 1,200 to 1,800 
pounds per 8-hour (urn, or 50 percent. Here again the 12-month employment 
year was cut at least by one-fourth. 

(3; In the steel foundry of the Bethlehem Steel Co., five to six chippers 
formerly worked 8 hours each to clean the scale from a certain producr, at a 

" Hearlng.s before the Temporary National Economic Committee, Part 30, p. Ifi480.. ■ 



CONCENTRATION OF ECONOMIC POWER 133 

total labor cost for this one operation of $31.20. Now one searfer with an 
acetylene torch and one helper does this same iob at a labor cost of $10.08, or 
a reduction in the total labor cost of 70 percent. Ac least 2,500 workers have 
been eliminated in the steel industry during the past decade by the shift 
from the chipping to scarfing raethod for taking bad seams out of slabs, 
billets, etc. 

Ill describing the inclusive naaire of the many technoiogical 
changes, Mr. Murray obsei v ed : 

I can cite numerous other cases, aud the story is the same : One man here, 
three rnen there, a dozen men here, and so forth are being displaced by tech- 
nical improvements. The effects of these improvements have been that fewer 
and fewer workers and less and less man-hours are required tc produce more 
and more steel products." 

Also, technological displacement of workers has occurred in the 
textile industries. For example, one concern in the viscose filament 
rayon industry changed its method of spinning and doifing and 
thereby displaced 1,100 workers; it i> likely that many of the work- 
ers, thus displaced, will be witliout any opportunities for reemploy- 
ment in their communities. In another textile plant, "production 
per man-hour increased approximately 100 percent from June 1937 
to June 1939 as a result of the installation of faster, more continuous 
machines, and the introduction of the cake-vrash method. Employ- 
ment at this plant declined 46.3 percent and production rose 10.2 
percent m the same period. At another plant em.ployment declined 
26 percent and production increased 60 percent within 2 years.^* 

Most of the innovations m textile manufacture have been in the 
direction of continuous operation, speeding up, and larger sized 
equipment. The most impressive advance toward continuous oper- 
ation has occurred in the spinning and finishing operations developed 
by the Industrial Rayon Co. at its Painesvilie, Ohio, plant. Spin- 
ning and finishing time lias been reduced from the custom.ary 85- 
hour period in })lants wnh discontinuous procedures to less than 
5 minutes. The elimination of numerous batch processes and the 
construction of compact spinnmg equipment permits the yarn to be 
washed, desulf urized, bleached, finished, dried, and twisted successively 
on- the same machine. Larjre package bobbins have reduced the 
doffing time to once every 19.5 hours compared with the former 4- to 
9-hour doffing time.^^ 

THE DURATION OF TIN EMPLOYMENT 

A considerable proportion of those ^vithout work remain jobless 
for extended periods of time. During recent years, studies of dura- 
tion of unemployment have been made by the National Res&arch Proj- 
ect of the Work Projectfj Administration.^^ In Philadelphia an 
analysis of the duration of unemployment was made in May 1937, 
a month of considerably industrial activity. The survey v.as con- 
ducted by means of a hoi'se-to-house canvass in which 46,000 house- 
holds, including about 9 percent of the estimated employable popu- 
lation of the city, were covered. 



« Ibid., p. 16479. 

^< Ibid . T5. 16842. 

" Ibid., p. 16841. 

"A surnmary of :he pioneeiing studies on the duration of unemployment — those of 
LiiDin, Myers, Clague, and Couper, Bf.kke and Lumpkin — by David Weintraub and Harold 
Posner, it to be found in the report of the National Resources Committee, Technological 
Trends and National Policy, 1937, pp. 83-85. 



134 



C ONCE>TRATION OF ECONOMIC POWER 



Philadelphia is a center of industrial activity, particularly in the 
manufacture of specialized textile, metal, and chemical products 
and machinery and transportation equipment. 

In 1930 about half of the gainful workers in the city were attached to the 
manufacturing and mechanical industries, in which various types of metal 
and machinery manufacturi^lg, textile manufacturing, and building construc- 
tion predominated. One-fifth of the gainful workers in 1930 were employed in 
trade, and the remainder in other types of industries." 

•The duration of unemployment in this study was defined as "the 
length of time from the date of the loss of the last nonrelief job 
which lasted 1 month or more to June 1, 1937." Employment on 
Emergency Works Program projects was counted as unemployment. 
For the majority of workers the duration of unemployment repre- 
sented the length of time they had been seeking work since the loss of 
their last job, but for an undetermined number it may have included 
some periods of time out of the labor market.^^ 

The duration of unemployment for both men and women in Phila- 
delphia, as of May 1937, is summarized in Chart VIII, table 11. 



Table 11. 



-Duration of unemployment since last nonrelief job, Philadelphia, May 
1937 



ALL UNEMPLOYED 



Duration of unemploy- 


Cumulative percentages 


Duration of unemploy- 
ment in months, total 


Cumulative percentages 


ment in months, total 


Men 


Women 


Men 


Women 




100.0 
61.7 
48.0 


100.0 
48.8 
34.9 


36 to 47 


37.7 
29.4 
21.6 


26.1 


12 to 23 


48 to 59 


19.6 


24 to 35 




14.2 









Source: AVorks Progress Administration, National Research Project, Employment and Unemployment 
n Philadelphia in 1936 and 1937, Part II, May 1937, 1938, p. 26. 

Long-term unemployment characterizes a substantial proportion 
of the jobless in this labor market. Forty-eight percent of the men 
and 34.9 percent of the women had been without private jobs (which 
lasted more than 1 month) for 2 years or over; 29,4 percent of the 
men and 19.6 percent of the women had been unable to obtain jobs 
for 4 years or more. The large [jroportion of women engaged in 
the textile and clothing industries, which were not so seriously affected 
as other industries by curtailment of production, probably accounts 
for their being less seriously affected than men. 

The Federal Reserve Board reported that industrial production, 
adjusted for seasonal variation, had risen from the 1923-25 base 
of 100 to 118 by May 1937, compared with 122 for the corresponding 
month in 1929. Hence, the month studied was by no means one of 
unusually depressed conditions. Yet a fourth of the persons in this 
labor market in May 1937 Avere totally unemployed." 

The seriousness of the duration of unemployment is even more 
acute in particular fields of work. For example, men customarily 

" Works Progress Administration, National Research Project, Recent Trends in Employ- 
ment and Unemployment in Philadelphia, 1937, p. 3. 

•■s Works Progress Administration. National Research Project, Employment and Unem- 
ployment in Philadelphia in 1936 and 1937, Part H, May 1&;:7. p. 24. 

^» Works Progress Administration. National Research Project, Employment and Unem- 
ployment in Philadelphia in 1936 and 1937, Part II, p. 11. 



CONCENTRATION OF ECONOMIC POWER 



135 



Chart VIII 

DURATION OF UNEMPLOYMENT SINGE 
LAST NON-RELIEF JOB 

PHILADELPHIA. MAY 1937 



PER CENT OF ALL 

UNEMPLOYED 

100 



ALL UNEMPLOYED MEN 



PER CENT OF ALU 

UNEMPLOYED 

100 




ni . „ 










, 


LESS THAN 1 YEAR 
1 YEAR AND OVER 


2 rEAi^S 3 YEARS 4 YEARS 
AND OVLR AND OVER AND OVER 

DURATION OF UNEMPLOYMENT 


- .EAHS 
AND OVLR 


PER CENT OF ALL 
UNEMPLOYED 


ALL UNEMPLOYED WOMEN 


PER CENT OF ALL 
UNEMPLOYED 


% 










80 


'^■i- \ 




























60 
















'■ '^"^x 














4fi :--. 


^^"•>^ 








40 


"-' 


~'-^. 


^?»^ 


^ 


— 


20 


n f., .^ . . ^ 


i \'y/„(-,,-. -.. '■■'/, {.. -^^ 


<.:..d-mM 


o 


LESS THAN 1 YEAR 
1 YEAR AND OVER 


2 YEARS 3 YEARS 4 YEARS 
AND OVER AND OVER AND OVER 

DURATION OF UNEMPLOYMENT 


5 YEARS 
AND OVER 


Source : Table 11. 

















136 CQJSTENTRATION OF ECONOMIC POWER 

employed in the manufacture of transportation equipment had been 
out of v;ork in May 1937. for an average of 40.2 months; those usu- 
illy employed in the manufacture of metal products for 39.3 months. 
An average of 34.3 months of unemployment was reported by men 
normally working for public utilities, and 31.5 months by men in the 
budding and construction field. 

The role of technology in long-term unemployment cannot be pre- 
cisely determined, but it is undoubtedly great. When a labor-saving 
technique takes the place of a Avorker, the duration of the affected 
worker's unemployment is likely to be extended. Frequently he must 
seek Mork in an industry new and foreiirn to his experience. If a 
mature and skilled worker, he may prefer to await the possibility 
that new work in his field will develop rather than to seek a less 
skilled and lower paid position in anotlier field. Often, owing to the 
practice followed by many firms of filling their unskilled and semi- 
skilled job openings with young men who are new entrants to the 
labor market, the mature but displaced worker can find no employ- 
ment even in an expanding field. Furthermore, it is difficult, if not 
often impossible, for him to move readily about the country, seeking 
a field of expanding employment. 

For these and related reasons, technological unemployment is likely 
to be long-term unemployment. Concrete evidence of this is attested 
in the above-cited study: In the industries in which technological 
advances have been greatest there are large numbers of unemployed 
who have remained jobless for long periods. For example: 

The highest average duration of uneniployment for women was repor^ed by 
those usually employed in public utilities (40.0 months), especially in telephone 
and telegraph companies."" 

TECHNOLOGY AND THE DISPLACEMENT OF SKILL 

Teclmology not only reduces the amount of labor required to per- 
form a given function; it also brings about changes in xi\q type of 
labor required, which often involve a displacement of skiU.^^ This 

20 Ibid., p 28. The widespread extent of this Imij • • ■• ■ iii)iloyment is corroborated 
by the Census of Unemployment in Massachusetts, i.i - : ; i. 

■'The questionnaire used in securing inform ; li. . i.,Ke to the duration of 

■unemployment of persons who were uriemploynl -', I :.!■ .1 -w" imiuiries, ar. follows: (a) 
Duration of unemployment since last employed at any occupation, and (b) duration of 
unemployment since employed at custon)ary occupation. These two phases of che sub- 
ject are here considered separately. 

"The total number of persons who were reported as wholly unemployed since tlieir last 
employnient at any occupation was 346.021. For 341.127 of these persons the duration 
of unemployment was reported, and of this number 134,913, or 39.5 percent, had noc been 
employed at any occupation for 1 year or less ; 62,482, or 18.3 percent, for 1 year hut less 
than 2 years ; 6.5,707. or 19.3 percent, for 2 years but less than 3 years ; 42.621, or 12..5 
percent, for 3 years but less tlian 4 years ; and 35,344, or 10.4 percent, f^)r 4 years or 
over. For males and females the distribution, according to duration of niicmployment. 
'.vas quite similar, except that in the case of the females the percentages Uiiem ployed for 
periods of stiorter duration were higher than the corresponding percentages for mules. 

"The total number of persons who were reported as wliolly unemployed, temporarily 
employed, and employed part-time was 624,526. For 615,889 of these persons information 
as to the diiration of their unemployment since they were last employed at their custom- 
ary occupations was obtained. Of this number (G1.5,SS9), 190,724. or 31.0 percent, had 
not been employed at their customary occupations for 1 year or less ; 115,013, or 18.6 
percent, for 1 year but less than 2 years: 137.808, or 22.4 percent, for 2 years but less 
than 3 years; 92,985, or 15.1 percent, for 3 years but less than 4 years; and 79,359, or 
12.9 percent, for over 4 years. For males and females the duration of unemployrrent 
since they were last employed at their customary occupations was quite similar, excepr 
that in the case of females the percentage (37.3) unemployed for 1 yc.ir or less was 
higher than the corresponding -percentage (28.7) for males. (Massachusetts Department 
of Labor and Industries, Report on the Census of Unemployment in Massachusetts. 1934, 
pp. 18-19.) 

=1 '■'By displacement of skill * * ♦ is meant the loss of the opporUmity to sell 
acquired skill at the rate of remuneration which would have been received if the machine 
liad uot I)eon introduced. Displacement, therefore, does not necessarily mean loss of 



CONCENT'RATION OF ECONOMIC POWER 137 

displacement is limited by the amount of skilled workmanship m 
modern industries. In the words of Harry Jerome : 

The potential displacement of skilled labor by the further substitution of 
machine methods for hand processing is limited by the fact that, while there are 
still manv hand workers in industry, the number engaged in hand crafts that are 
of a distinctly skilled tyi^e is relatively small, esijecially if we exclude the building 
industries. * * * -^ 

PATTERNS OF SKILL-DISPLACEMENT 

Techniques which displace skill are generally of three basic pat- 
terns : (1) Those designed for one specific operation — special-purpose 
techniques; (2) those so designed that with slight modification they 
can he applied to other operations — general-purpose techniques; 
(3) those which go through successive radical changes, after being 
applied to a specific operation, culminating in entirely new and even 
more efficient techniques. 

A simple example of the first pattern is the replacement of skilled 
stonecutters by mechanical stoneplaners. As early as 1915, a single- 
platen planer "of improved type could do as much work in an hour as 
10 stonecutters. 

• It may be roughly estimated that in 1900 there v>'ere between 20,(X)0 and 25,000 
stonecutter in the United States. The labor-saving devices introduced in the 
trade, chiefly after 1800, did in 1915 an amount of work which, at the lowest esti- 
mate, would have required the labor of 10,000 hand cutters. By 1915 probably 
one-half of the stonecutters had been displaced from the trade." 

An improved wood shaper now turns out special shapes of ironing 
boards. The operator simply stacks a number of boards which are 
cut and shaped at one time by the machine; he then removes the cut 
boards and inserts others. Formerly, a moderately skilled worker 
manipulated a single ironing board around the cutting tool until the 
edges were smooth and even.'^ 

Highly skilled wood carvers in the furniture industry have been, 
displaced to a con.siderable extent by wood carving machines, which 
can produce up to 24 identical carvings from a master form. Re- 
placing the skilled carver is a semiskilled operator who is generally 
incapable of carving with hand tools. Only a few of the craftsmen 
are retained — and these only in the large furniture factories — to make 
the original master forms.^^. 

Craftsmanship is disappearing also in the pottery industry. The 
potter's wheel is rarely used. Instead, a plaster-of-paris mold is 
filled with fluid clay, and later a comparatively unskilled worker 
breaks open the mold to discharge the ware.-® 

Special-purpose techniques are much more limited as to skill- 
displacing potentialities than those which can be applied to a number 
of occupations. Industrial instruments exemplify the general-purpose 

employment. If the employment after the introduction of the machine, and as a con»o 
quence thereof, is at a lower rate of remuneration, displacement of skill has occurred." 
(George E. Barnett, Machinery and Labor. Harvard University Press, Cambridge, 1926, 
p. 117.) 

22 Mechanization in Industry, National Bureau of r':.'ouomic Research, New York, 19.34, 
p. 397. 

23 Barnett, op. cit., p. .34. 

2* Work Projects Administration, National Research Project, Changes in Machinery and. 
Job Requirements in Minnesota Manufacturing, 1931-36 by C. A. Koepke and S. T. Woal. 
1939, p. 9. 

^Ibi6., p. 44. 

2« Ibid., pp. 44—45. 



138 CONCENTRATION OF ECONOMIC POWEP 

type of skill-displacement techniques. Their application has been 
steadily extended into a larger number of fields. 

The skill-displacing potentialities of instruments %Yhich control are 
much greater than those which merely indicate or record. .Of all the 
new instruments. placed on the market during the period 1928-37 the 
number of the controlling type rose most sharply. In 1928 only 8 
percent of the new instruments were control devices; by 1935 their 
proportion had risen to 40 percent. Tlie proportion of recording in- 
struments declined during this period from 15 to 11 percent and in- 
dicators fell from 78 to 49 percent.-^ 

The widespread application of industrial instruments has resulted 
in a marked change in the composition of the labor foi-ce in numerous 
industrial operations, since the need for skilled labor is reduced when 
instruments simplify and standardize the method of work. 

An outstanding illustration (of this elimination of individual skills) is to be 
found in steel treating where, before the development and application of pyrom- 
eters, heat treating was assigned to skilled craftsmen who judged the tempera- 
ture of the metal by observing its color. (A dull red color, for example, indi- 
cated low temperature, whereas white was a sign of high temperature.) A 
worker had to have considerable skill and training to be able to gage metal 
temperatures correctly at the several stages of the heat-treating process, and 
only experienced craftsmen were able to produce high-quality products. * * * 
Pyrometers (during the 1920's) had eliminated the need for ijerceptory skill, 
and the operator now merely followed instructions prepared by the plant metal- 
lurgist. Although the worker had to possess some skill in order to follow the 
instructions, the responsibility' for quality no longer rested on his shoulders 
alone, since he relied not only on the metallurgist but on the pyrometer as well. 
When, more recently, automatic temi>erature control of heat treating was de- 
veloped, the process became highly standardized and the last elements of skill 
were eliminated.'^ 

In recent years changes of almost revolutionary character have 
affected certain basic stages of metal work. The most striking are 
stamping and welding. Each of these general purpose techniques 
has been known for years, but has been progressively extended to a 
growing number of fields. When metals were formed into shape 
with planers, lathes, broachers, shapers, and files, the skill require- 
ments per unit of output were among the highest in the industrial 
world. But today, the role of many metal workers is becoming in- 
creasingly that of an attendant to a machine which batters and ham- 
mers the metal into form with tremendous force and extreme precision 
at an almost incredible speed. This machine is the punch press. 
The punch press and closely related types of equipment are today 
capable of trimming, shearing, parting, notching, blanking, punch- 
ing, piercing, bending, beading, expanding, curling, contracting, 
burring, wiring, drawing, extruding, forging (hot and cold) swaging, 
flanging, embossing and pinching.-'-^ 

"Because of flexibility in feeding stock and discharging work, 
these presses have proved to be profitable investment in the produc- 
tion of an endless variety of metal stampings for automotive and air- 
craft parts, refrigeration and air-conditioning equipment, radios, 
electrical equipment and appliances, agricultural implements, business 
machines, stoves, hardware, toys, containers, and tinware." {Iron 
Age, January 4, 1940, p. 297, advt.) 

^ Works Progress Administration, National Research Project, Industrial Instruments 
and Changing Technology, 1938, pp. 44-46. 
2« Ibid., p. 8G. 
'^0 Ravmond F. Yates, Machines over Men, Fredericli A. Stokes Co., New York, 1939, p. 138. 



CONCENTRATION OF ECONOMIC POWER 139 

The speed of the punch press has been steadily increased. One 
punch press recently developed delivers 1,200 strokes per minute and 
operates with a continuous-feed mechanism; unit punch set-ups have 
been developed for automobile work which punch up to 52 holes at 
once. In fact, it is now possible "to speed up presses to a maximum 
production, limited largely by ability of the equipment to discharge 
nnished work.'- ^° 

The amount of skill per unit of output has been decreased not only 
by the speed of these modern battering rams ; their very development 
has comi^letely eliminated certain functions formerly performed by 
skilled workers.^^ The increasing use of the stamping process has so 
lowered labor requirements that certain types of presses are reputedly 
capable of returning initial investment in from 60 to 90 days of full- 
time operation.^- 

Not many years ago the electric arc and the acetylene torch were 
regarded merely as tools for repair. The development of the resist- 
ance welder and the spot welder (which automatically ceases func- 
tioning upon attaining the precise conditions desired in the metals 
being fused) has practically displaced the riveter, formerly a highly 
skilled worker in a number of the steel trades. 

The most striking use of spot welding is in large jigs, particularly 
in the automobile industry. In the production of 1940 automobiles 
a new method of spot welding fused a 4-door sedan together almost 
instantly at 222 points. In this type of automatic application, spot 
welding represents more than a transition from the skilled riveter 
to the welder, since the mechanism, after being set in motion, advances 
to the correct temperature, fuses numerous points, and then ceases 
operation, with hardly any human labor involved. 

Even where welding is not performed automatically in a large jig, 
miprovements have reduced considerably the amount of skill required 
on the part of the individual welder, especially through automatic 
tracing of patterns by means of magnetic attraction. The science 
of welding has been so developed that the worker only needs to 
know how to make good welds. A lengthy experience with the 
process and an intimate knowledge' of metals are no longer requisites 
of the occupation.^^ 

The skill-displacement caused by the punch-press and the spot-weld- 
ing process is reflected in tliQ occupational requirements of an industry 
in which they are widely used. It has been reported that in the plants 
of one large automobile company 43 percent of the workers require 
only 1 day to learn their jobs, 36 percent up to 8 days, 6 percent up to 
2 weeks, 14 percent from a month to a year, and only 1 percent more 
than a year.^* 

The changes which have taken place in telegraphy exemplify the 
third pattern of skill displacement. In the larger offices (classed as 
functional offices) Morse operators had been largely displaced by 

3" steel, January 2, 1939, p. 320. 

'1 As an instance, in the stamping and forming of metal in the production of valve 
rocker arms for automobile engines, the "part is composed of two halves, formed from 
identical steel sheet blanks. A hub is drawn in each half, so that when the two pieces are 
• * * brazed vogether, they form a complete arm and hub ready to receive a bushing. 
The hole is sized so accurately in the drawing operation that machining is unnecessary. 
Oil ductg are coined in each half, so that when brazing is finished, a complete oil line is 
formed, thereby eliminating a drilling operation." (Ibid., p. 320.) 

^ Ibid., p. 320. 

33 Ibid., p. 184. 

3' Hearings before the Temporary National Economic Committee, Part 30, p. 16372. 



140 CONCENTRATION OF ECONOMIC POWER 

Operators of teletypewriters by 1932. The productivity of printer 
operators was so great tiiat approximately 50 percent of ':he inimber 
of operators who would be required for Morse manual operation had 
been displaced. In all offices combined the proportion of male opera- 
tors had been reduced from about 80 percent to about iO percent of the 
total, and the skill and training characteristic of Morse operators was 
supplanted by ability simply to operate a typewriter keyboard adapted 
to telegraphic purposes.^^ 

Kecently labor-displacement has been greatly increat~cd by the in- 
troduction of improved mechanical features, particularly the Multi- 
plex-Simplex repeater and the reperforator/''^ A reperforator installed 
in 1937 b}^ one large communication comp^my in Eichmond, Ya., dis- 
l^laced 70 percent of its entire personnel.^' 

The culmination of this skill-displacement would be reached if com- 
munications companies adopt a technique with which they are now ex- 
perimenting, the facsini51e machine. The customer simply writes out 
his own message, places it in a slot, pays a fee, and a facsimile copy of 
(he message is automatically delivered at the other end.-^ 

Printing methods have likewise been successively improved. Prior 
to 1890 the process of composition liad not been greatly improved in 
400 years, but about that time machine composition began to supplant 
type-setting by hand. Between 1887 and 1903 a total of 8 ,618 linotypes 
were manufactured in the United States and Canada. A linotype 
operator can set as much in 1 hour as a hand compositor can in 4. If 
7,500 linotype machines were operated in 1904 the same number of 
hours each day as hand compositors formerly worked, potential dis- 
placement of iiand conrpositors then was 30,000. Allowing both for 
the fact that many linotypes were run two or three shifts and that a 
reduction ii'om 10 to 8 hours in the working day took place about this 
time, potential displacement of hand compositors by 1904 was about 
36,000. Such a displacement did not take place because of the increas 
ing quantity of printing done — an increase brought about largely by 
the adoption of the new technique. A skilled handcrafr -vri.-;. liow-evei . 
displaced by a machine operation requiring less skill,^^" 

35 U. S. Bureau of Labor Statistics. Monthly Labor Review, JLirch 10,^2, "DJRplacemem 
of Morse Operators in Comuiercial Telegraph Offices," pp. 501—515. 

"Each ke.v on a (telet.vpewriter) kej'board represents a character, nnrl tlie flepressing 
of the key sets up an electrical contact which automatically operates the corresponding 
key on a similar keyboard at dw receiving end. The crjutact may be est.illislu'c; by direct 
keyboard action or by means of a perforated tape which is automatically fed through a 
transmitter, each set of perforations composing a :'ode character ro- iL'sooi'ding to a 
character on the keyboard. At the receiving end the keyboard which carornaiic'illy prints 
the message may be a tape-recording printer or a page printer. In eithc; case iiie message 
is typed out, not in code but in ordinary printed characters. Several receiving machines 
may be operated on the same circuit by one transmitting machine." (Ibid., pp. 502-50H.) 

*•> The so-called Multiplex-Simplex repeater is "a device which comprises a ponable table 
on wheels which can be connected with adapters on any trunk circuit and i.ive airect service 
to any brand) having a teletype machine." The labor-saving feature of tlii-s device con 
sists in its elimination of reiays. By its use messages are routed to centers which have 
direct and continuous operation to other localities ; through this substitution of con- 
tinuous for sporadic operation, the amount of labor required to handle a gi%sn number 
of messages has been considerably reduced. 

A second improvement which has tended to reduce unit labor roOiii'^-.-iunts is the so- 
called reperforntor. "'This c'-^-ico) consists of b.nnks of relays, transmitters, and auto- 
matic enuipmenf desirincd to clhninatf' jobs. Instead of the message b'^ing printed on a 
tape and gummed by an operator us before, it is received in the form of a perforated t;ipe 
with the printed message above the perforated tape and is switched to its destination at 
\he switching center and placed in a transmitter by Ib-^ .-eceiving or switching operator. 
One operator in the switching center ran take care of tiio same number of wires that four 
operators formerly worked, thereiiy eliminating threL= operators who formerly had jobs 
Furthermore, \inder this systeni the switching center operator must now also be a route- 
clerk. ♦ * ♦ thereby eliminating the route-clei'k's job." (Hearings before the Tem- 
jiorarv National Economic Committee, Part 30, pp. 16690-£'1.) 

37 Ibid., p. IRfiOl. 

•" Ibid., p. 16692. 

»" See Barneft, op. cit, pp. 3-6. Unlike the teletype operator, the linotype operator 
is a skilled worker. The two examples are different in this respect. 



CONCENTRATION OF ECONO.AJl" POWER 141 

Improvements were tlien made which veduced materially the 
amount of labor required to operate the new technique. The linotype 
of today is immeasurably more efficient than that of 1904. In adcli- 
tion, experiments have been made directed toward replacing the man 
at the linotype kej'board with a photoelect^-ic cell scanner which 
would "read'^ from specially prepared copy produced by a typewriter 
with specifl] characters.^'' This would eliminate the printer from 
the linotype machine in the same way that the reperforator and the 
Multiplex-Simplex repeater are eliminating operators from the. 
teletype. 

The ultimate development would be the introduction of a photo- 
graphic method of printing. If use were made of paper sensitive 
to some type of cathode-ray in which the photo-chemical action set 
up would not require development of any kind, type could be replaced 
by a negative. This development awaits only the invention of an 
adequate method of preparing the negative and an inexpensive sensi- 
tizer. This method is somewhat similar to offset photography which 
has been found relatively inexpensive in making reprints of books 
when the books are printed from type and not from plates. In 
emergencies newspapers and periodicals have been composed of plates 
for each- pag? made by photo-engraving.*^ If the photo-engraving 
process were to replace printing as the facsimile machine threatens 
to replace tlie teletype system, labor in the printing trades would be 
almost completely eliminated. 

At the turn of the century window glass was produced almost 
entirely by hand craftsmanship. Five classes, of skilled Avorkers 
(gatherers, blowers, snappers, flatteners, and cutters) were necessary 
to fashion the molten gb.-^^r into a sheet. The cylinder machine, a 
mechanical window-glass blower, followed the liand method of blow- 
ing a long, cylindrical bubble of glass whieli, when cooled, was split 
open ancl flattened. It eliminated the higl^ly skilled blower and 
gatherer amd the semi-skilled snapper. From its introduction around 
1900 to its replacement by a still better technique in 1913-14, the 
c}^linder machine caused a decrease of 56 percent in the- amount of 
window glass produced by the hand process. Although the improve- 
ments made in the cylinder machine during this period did not 
increase the number of' skills eliminated, they raised the productive 
efficiency of the workers retained, so that less and less labor was 
required to produce a given amount of window glass. 

But while the facsimile machine in telegraphy and photo-engrav- 
ing in printing merely foretell almost laborless methods, the sheet 
process in window glass manufacture has practically achieved that 
result. In that process, a flat sheet of glass is drawn directly from a 
fore-hearth connected by a cooling tank with the melting tank, 
thereby eliminating the splitting and flattening of the cylinder. 
The she«t process thus displaced a fourth craft, leaving only one 
skilled, trade — cutting — as an essential process. With the adoption 
of the sheet method, the culmination of the third pattern of skill- 
displacement was definite)y achieved.*^ 

All new tecnniques, however, do not result in a reduction or elimina- 
tion of skill requirements. In certain cases new techniques cause the 

*Rajiriond F. Yetes, Machines Over Men, Frederick L, Stokes Co., New York, 1939, 
p, 221. 

" Ibid., p. 223. 

« See Jerome, op. cit., pp. 97-102, 



142 CONCENTRATION OF ECONOMIC POWER 

replacement of a large number of semi- or unskilled workers by a few 
highly skilled operators. 

For example, the warp tying-in machine useil in cotton weaving enables a 
skilled machine operator, with an assistant, to do work formerly requiring 12 to 18 
tying-in girls. Their work required considerable adeptness and experience, but 
not a degree of skill comparable with that required of the machine operator." 

A steam-shovel operator is likewise more skilled than the ditch- 
diggers he replaces. 

The tendency toward the creation of new skills is important, but even 
though a new skill is created by a new technique, unit labor require- 
ments in the process as a whole are generally reduced. Furthermore, 
producers sooner or later tend to develop new mechanisms which 
eliminate the neM' skill or standardize its operation, so that the func- 
tion can be performed by a semi- or imskilled worker. 

SKILLED WORKERS AS A PROPORTION OF ALL WORKERS 

Whether the proportion which skilled workers constitute of the 
Nation's labor force is increasing or declining appears impossible of 
absolute determination with existent data. Since 1936 the Bureau of 
Labor Statistics in its wage and hour studies of specific industries has 
classified employees as to whether they are skilled, semi-skilled or un- 
skilled. But the period 1936 to date is too short for determining 
occupational trends. 

By using the social-economic groupings made by Dr. Alba Edwards 
of the Bureau of the Census, it is possible to compute the proportion of 
all workers designated by him as skilled in 1910 and to compare that 
proportion with the percentage so designated for 1930 in manufac- 
turing and mechanical industries. Several considerations as to the 
use of this method must, however, be borne in mind: 

(1) Concerning the basic problem of the standard used in the deter- 
mination of skill, Dr. Edwards states : 

* * * The term skill, for the purposes of a grouping such as here presented. 
is considered properly applied only to those occupations for which the expenditure 
of muscular force is one of the chief characteristics. Within this field, those occu- 
pations have been considered skilled for the pursuance of which a long period of 
training or an apprenticeship usually is necessary, and which in their pursuance 
call for a degree of judgment and manual dexterity, one or both, above that re- 
quired in semiskilled occupations. Those occupations have been considered semi- 
skilled for the pursuance of which only a short period or no period of preliminary 
training is necessary, and which in their pursuance call for only a moderate degree 
of judgment or of manual dexterity.** 

The number of borderline cases between skilled and semiskilled 
classes undoubtedly bulks extremely large because Dr. Edwards under- 
took to classify all gainful workers in the United States into only 9 
groups, of which the grouping by skill was one. It is hardly necessary 
to point out the tremendous difficulties of grouping millions of gainful 
workers into so small a number of classes. As Dr. Edwards st,ates, 
"Each of the groups doubtless includes some workers who properly be- 
long in another group, and from each group doubtless are omitted some 
workers who properly belong there." *'^ 

« Jerome, op. cit., p. 398. 

*• Quarterly Publications of the American Statistical Association, vol. XV. June 1917, 
"Social-Economic Groups of the United States," by Alba M. Edwards, p. 646. 
*' Ibid., p. 6i5. 



CONCENTRATION OF ECONOMIC POWER 143 

(2) The manufacturing and mechanical industries of 1930 inchide a 
number which were non-existent in 1910. In fact, the pattern of the 
economy has been greatly changed by the various shifts in the relative 
importance of particular segments. 

(3) Designations of certain occupations may remain constant while 
actual functions performed by the workers engaged therein may be 
entirely changed. Many workers today are designated bv terms which, 
because of changing processes, no longer indicate the high degree of 
skill formerly connoted by such terms. 

(4) In the 1930 tabulation foremen were included with skilled 
workers, but in 1910 they were omitted from that group. This 
\sould obviously reduce the possibility of a decline in the proportion 
of skilled workers. 

On the other hand, both the 1910 and 1930 groupings are those 
of the same student, and thus reflect a relatively constant opinion 
concerning the standards underlying the classification. Pending 
more detailed analyses, however, it is impossible to determine with 
any degree of precision whether or not the skill-displacing tendencies 
of technology have been offset by the creation of new skills. 

According to Dr. Edwards' data, employment in manufacturing 
and mechanical industries in 1910 amounted to 10,658,881, of whom 
3,821,327, or 35.9 percent, were classed as skilled workers.'**' In 
1930 the number employed in manufacturing and mechanical in- 
dustries totaled 14,332,372, of whom 4,678,766, or 32.6 percent, were 
designated as skilled:*^ This decline between 1910 and 1930 in 
the proportion which skilled workers constitute of the total labor 
force in manufacturing and mechanical industries may represent 
the actual net change in occupational groupings. The tendency 
appears to be toward "downgrading" rather than "upgrading," 
and is substantiated by numerous specific cases in which technology 
has reduced the amount of skill required in the performance of a 
given function. 

PRESENT TRENDS IN OCCUPATIONAL REQUIREMENTS 

In 1931 a study was made of prevailing industrial conditions 
and mechanical operations performed under the auspices of the 
Employment Stabilization Research Institute of the University of 
Minnesota.*® In 1936 a similar analysis was made, its point of 
departure being likewise the strictly mechanical aspect of industrial 
processes." The 1936 study also compared the characteristics of 
the personnel employed, the work performed, and the machinery 
utilized for the processes studied in the 2 years. 

One of the best methods of ascertaining changes in skill require- 
ments is to compare the amount of time required for training at 
different periods. This method does not depend upon subjective 
evaluation of skill involved by either the analyst or the investi- 

« Ibid., p. 648. 

nf*\hA T^-i?H^1?^".°' iQ?,o^^^a"^' ^ Social-Economic Grouping of the Gainful Workers 
or the Lnited State^ 1938. The great decline in the number of unsliilled farm workers 
during this period distorts any comparison which endeavors to include workers in all 

■•« Charles A. Koenke A Job Analysis of Manufacturing Plants in Minnesota (Univ 
Minn Employment Stabilization Res. Inst. Bull., vol. II. W 8, June 1934) 

"• vNork Projects Admirjistratioxi, National Research Project Chanees in Afachinery and 
J^b Requirements in Minnesota Manufacturing, 1931-36, by C. A. Koepke and S. T. WoaL 



144 



CONCENTFCATION OF ECONOMIC POWEB 



gator. Its only qualification is that it rests npon the reasonable 
assumption that under normal conditions a decrease in the length 
of the training period generally indicates a decrease in the amount 
of skill required. 

The changes in the length of training periods required for all 
production workers in 19?.l and 1936 in five diversified industries — 
metal working, baking (hand-operated), creamery, printing (en- 
graving), and laundry^ — are shown in chart IX, table 12. The 
tendency in each of the industries, whether highly mechanized or 
hand-operated, was definitely toward a shorter training period for 
a gre;iter proportion of workers. 

Table 12. — Length of training period required, 1931 and 1936 ^ 

PRODUCTION WORKERS IN MINNESOTA MANUFACTURING 





Less than 


H to 2 


3 to 9 


10 months 


2 years 


Over 4 




1/2 month 


months 


months 


to 2 years 


to 4 years 


years 


















1931 


1936 


1931 


1936 


1931 


1930 


1931 1936 


1931 


1936 


1931 


1936 


Metalworking . . . ^ 


4 


20 


22 


30 


11 


16 


20 


23 


42 


7 


1 


4 


Baking (hand-operated) - - 


S 


45 


15 


27 


15 


16 


30 


7 


32 


5 










44 


43 


13 


)9 





20 


37 


12 


4 





2 





Printing (engraving) 2 


6 


14 


2 


11 


6 


6 


9 


6 


32 


28 


45 


35 


Laundry 


20 


39 


38 


43 


32 


9 


9 


9 


1 




















1 Based on identical plants in industries in which 85 or more workers were- surveyed in the given year. 

2 Includes high-class printing and newspaper printing. 

Source: Work Projects Administration. National Research Project, Changes in Machinery aud Job 
Requirements in Minnesota Manufacturing, 1931-36, by C. A. Koepke and S T. Woal, 1939, p. 38. 

In metalworking only 4 percent of the production workers m 1931 
required a training period of less than half a month; by 1936 that pro- 
portion had risen to 20 percent. On the other hand, the proportion of 
workers which required from 2 to 4 years of training fell from 42 per- 
cent of the total in 1931 to only 7 percent in 1936. Tliis was due to the 
use of jigs, fixtures, and other devices which reduced the degree of skill 
and therefore the amount of training required for many jobs.^*^ 

The same tendency was apparent in a less mechanized industry — 
hand-operated bakeries. The proportion of workers requiring less 
than one-lialf month's training rose from 8 to 45 percent in the same 
period, while the proportion requiring 2 to 4 years' training fell from 
32 to 5 percent. This curtailment of the training period is due largely 
to the further division of labor and to more efficient managoment. In 
some hand-operated bakeries, Avhere formerly one baker had made a 
complete product, a type of assembly line, with each worker perform- 
ing one simple operation was introduced. This is a striking illustra- 
tion of t)ie use of chain work ro save labor. 

Similarly, in printing the proportion of workers requiring a training 
])eriod of less than one-half month rose from 6 to 14 percent and tjiose 
requiring one-half to 2 months increased from 2 to 11 percent. On the 
other hand, tlie proportion rtjquiring 2 to 4 years of training fell from 
32 to 28 percent, and those requiring over 4 years dropped from 45 to 

«> During the period 10.31-36 a number of the plants studied set up industrial engi- 
neering departments composed ot men trained in production methods and factory man- 
agement who determined bow manufacturing costs might b6 lowered. In one plant sur- 
veyed they designed new equipment, improved production techniques, arranged machine 
lay-outs more etlectively, and eliminated considerable waste of materials. 



CONCENTRATION OF ECONOMIC POWElt 



145 



Chart IX 

CHANGE IN LENGTH OF TRAINING 
PERIOD REQUIRED* 

PRODUCTION WORKERS IN MINNESOTA MANUFACTURING 
1931 AND 1936 



PERCENTSGE C 
PRODUCTION 
WORKERS 



METALWORKING 



PEBCENT4GE or 
PRODUCTION 
WORKERS 









A 




^.936 




y 


\ 






^, 


/ 


\ 


/ 


N. *« 


-y^ 




\ 


/ 






■--\ 



[ MONTH MONTHS 



3 109 10 MONTHS 2 TE4RS 

MONTHS to 10 

2 YEARS 4 TEARS 
SININO PERIOD REQUIRED 



\ 



PERCENTAGE OF 
PRODUCTION 
WORKERS 













X" 


6 


I93lv^ 


— 






^.. 




/ 


\ 






"n 


/ 




\ 


^^ 








" 


\ 



PERCENTAGE Of 
PRODUCTION 
WORKERS 













«0 
30 






l93l-</^ 






v,^ 


.193« 










^ 


/- 


''N^ \ 








\ 


-.^ 





LESS THAN liIoZ 3 io9 IOMONTmS 2 YEARS OVER 

l-I MONTH MONTHS MONTHS 10 10 * YEARS 

2 YEARS 4 YEARS 
TRAINING PERIOD REQUIRED 



LESS THAN 'ilo2 3 to 9 10 MONTHS 2 YEARS OVER 

'-I MONTH MONTHS MONTHS to to 4 YEARS 

2 YEARS 4 YEARS 
TRAINING PERIOD REQUIRED 



\ 



PERCENTAGE OF 
PRODUCTION 
WORKERS 



PERCENTAGE Of 

LAUNDRY To^E^r 













y 












<^-^ 










I93I.«V, 




■'-^Z 












^^>' 


'""1 





_J^I936 








V\ 


^ 


1,3, 






\ 




- 




\ 


\ 








~ 'NJ 





LESS THAN I 2 to 2 3 to 9 10 MONTHS 2 YEAR$ OVER 

Vl MONTH MONTHS MONTHS to to 4 YEARS 

2 YEARS 4 YEARS 
TRAINING PERIOD REQUIRED 



• Based on identical plants in industries in which 85 or more workers were surveyed in 
the given year. 

Source : Table 12. 

277551— 41— No. 22 11 



146 CONCENTRATION OF ECONOMIC POWER 

35 j)ercent. This decline in the need for extensive training was due to 
the installation of automatic and semiautomatic equipment. 
The authors of the study make this observation : 

In the printing industry, operations on automatic mactiinery rose from 25 per- 
cent in 1931 to 45 percent in 1936, while those on semiautomatic machines also 
increased slightly. * * * The increase of automatic operations in the baking 
industry, from 12 percent in 1931 to 40 percent in 1936, reflects the technological 
ti'end in the mechanized bakeries. Many of the operations previously assigned to 
hand labor, such as rolling dough and forming loaves of bread, had by 1936 been 
transferred to machines. Similarly the new equipment on which operations were 
observed in the paint industry was automatic, sometimes representing a shift 
from manual operations and sometimes, as in the case of the grinder, introducing 
a striking increase in capacity. 

In the laundry industry considerable savings were effected "in labor 
and in skills by the introduction of more speedy air-driers and of semi- 
automatic shirt pressers." ^' 

It appears that the tendency of technology to reduce skill require- 
ments has not been reversed. 

In several of the plants covered * * *^ the installation of automatic machines 
in large numbers or the rendering of old machines automatic or semiautomatic by 
the application of special fixtures has resulted in a corresponding revision of the 
entire process of production. Skilled labor has been reduced and in some cases 
virtually eliminated, and training periods for many semiskilled operations have 
been cut down. Thus, although a small group of highly skilled "machine setters" 
is always required to set up and adjust the equipment, in- actual production semi- 
skilled workers tend to predominate." 

While certain skilled workers will always be required to set designs, 
etc., they are giving way in the labor force to workers of lower skill. 
But the number of completely unskilled workers also appears to be 
declining. One of the most spectacular changes between 1931-36 
noted by this study was the decline in the proportion which com- 
pletely unskilled workers were of the total number of production 
workers. In 1931 virtually all the industries covered had some pro- 
duction workers so classified ; in 1936 the proportion had dropped in 
the majority of industries surveyed, and in several plants this type 
of job had disappeared altogether. 

The ease with which techniques of production can be devised to 
replace unskilled workers and the importance from the cost stand- 
point of developing techniques to replace the most-skilled workers 
combine to reduce the proportion of most-skilled and least-skilled 
workers with a corresponding gain in the semi-skilled. The authors 
of the Minnesota study found that — 

Few of the plants resurveyed had many class A (the most-skilled) jobs, whereas 
in every industry the plants had a substantial proportion of jobs in classes O and 
D (the semi-skilled) ; these had increased at the expense of occupations in classes 
A or E (the least-skilled) in most cases.^^ 

In the opinion of the authors, the use of skilled workers is still 
declining. 

The data presented here tena ro support the contention put forward in the 1931 
study that advances in production technique have not^ only drastically cut down 
the time necessary for training but have also given rise to significant changes 
in the qualifications demanded of the worker. There is a strong indication, more- 



" Work Projects Administration, National Research Project, Changes in Machinery 
and .Tob Requirements in Minnesota Manufacturing, 1931-36, 1939, p. 11. 
6»Ibid., pp. 13-14. 
B Ibid., p. 43. 



CONCENTRATION OF ECONOMIC POWEli 147 

over, that the process of increasing division of labor through increased mechani- 
zation is still running its course. That course may not now be so rapid or sa 
revolutionary as before, but it is still causing important developments in industry 
and is showing a tendency to continue in that direction." 

TECHNOLOGY AND LABOURS DEMAND FOR GOODS 

If reductions in prices are not made to offset lower unit labor re- 
quirements and skill-displacement, technology will tend to reduce 
labor's share of the value of products. When the amount of labor 
required per unit of goods is reduced, the wage payment per unit w^ill 
decline unless average earnings increase to an extent greater than 
the advance in labor productivity. Furthermore, if unit labor re- 
quirements are materially decreased, a labor surplus will develop 
unless production is markedly advanced, and the mere existence of a 
labor surplus is a depressant upon wage rates. In addition, if the 
quality of work (the degree of skill) required per unit of goods de- 
clines, wage payrnents per unit will tend to decline. '^^ 

If the prices of goods remain constant while unit labor costs decline, 
labor will receive a diminished share of the value of products. Then 
recipients other than labor will receive an increasing share. The 
other recipients fall largely into the upper-income or propertied 
classes which save a much larger proportion of their income than does 
labor. Thus, technology b}' reducing unit labor costs may aggravate 
the existent disproportion between consumption and savings. 

The consumption-saving problem has been well described by I.auch- 
lin Currie. 

When * * * a p.^i-f of the wages received or of money realized for sales is 
not disbursed but is retained by the individual either in the form of cash or of 
deposits, OT is used to pay off debts, or even if it is invested in securities, there 
may be an interruption in the flow of the money stream. Whether there is or is 
not depends on whether the money thus withdrawn is kept idle, or hoarded, or 
whether it is returned to the stream through disbursement for new plant and 
equipment, or for renovation or enlargement of existing plant, or offset by the 
expenditure of an equal amount. The money thus restored continues to be a 
saving by the individual, but it is no longer a withdrawal from the income 
stream. * * * 

It is obvious that the larger the portion of a given national income that is with- 
held from consumption, the larger must be the expenditures that represent offsets 
to saving, if the national income is not to decline. To state the reverse of this 
proposition, the larger the portion of income that is spent on consumption, the 
smaller need be the volume of capital expenditures to sustain the given national 
income.^* 

The behavior of unit labor costs depends upon two factors : Average 
hourly earnings and output per man-hour. If the wage cost per hour 
(average hourly earnings) advances more than the amount of goods 
produced per hour (output per man-hour), unit labor costs will 
rise. On the other hand, if the increase in output per man-hour 
exceeds the rise in average hourly earnings, unit labor costs will 
fall.s- 

" Ibid., pp. 36-37. 

" "The creaking down of .iobs and operations into their elements often results in 
greater division of labor, which also weakens skill. Everv loss in skill, of course, means 
a weakeninc: of the bargaining power of labor." (Carroll" Daugherty, Labor Problems in 
American Industry, Houghton Mitflin, Boston, revised edition, 1938, p. 592.) 

^Hearings before the Temporary National Economic Committee, Savings and Invest- 
ment, Pt. 9, 1040, pp. 3521-3522. 

67 There are two methods of^ computing indexes of unit labor costs, both of which 
should theoretically give the same result. The index of unit labor costs may be derived 
either by dividing the index of pay rolls by the index of production or by dividing the 
ndex of average hourly earnings by that of output per maOrhotir. Since average hourly- 



148 



CONCENTRATION OF ECONOMIC POWER 



Indexes of production are available from the National Research 
Project study, Production, Emplo3'ment, and Productivity in Fifty- 
nine Manufacturing Industries, 1919-1936. For 24. industries, or 
groups of industries (which were responsible in 1929 for 71 percent 
of the total number of employees covered by the National Besearch 
Project study), it is possible to obtain comparable indexes of pay 
rolls.^^ By limiting the comparison to the Census of Manufacturers 
years, 1923 and 1935, comparability of the two series is assured. 

Average hourly earnings from 1923-35 are available from the Na- 
tional Industrial Conference Board for 11 of the 24 industries. Al- 
though in certain cases industry designations are somewhat dissimilar, 
such as "rubber products" (Bureau of Labor Statistics and National 
Research Project) and "rubber manufacturing" (National Industrial 
Conference Board), the basic production operations covered are essen- 
tially the same. 

That the advance in hourly earnings did not keep. pace during this 
period with the increase in output per man-hour is shown clearly by 
chart X, table 13. 

Table 13. — Percent change in hourly earnings, output per man-hour, and unit 
lalor cost in 11 manufacturing industries, 1923 to 1935 





Percent change from 1923 to 
1935 in- 


industry 


Average 
hourly 
earnings 


Output 
per man- 
hour 


Unit 
labor 
cost 




+9.9 
+ 19.8 
+28.0 

+5.8 

+7.7 

+ 15.2 

+14.2 

+.7 

+2.2 
+36.1 
+24.4 


+48.2 
+74.2 
+79.6 
+46.5 
+31.7 
+54.1 
+38.9 
+28.5 
+43.7 
+66.2 
+45.8 


-20.0 


Chemicals 


-29.5 


Rubber products 


-32.2 




-30.0 


Paints and varnishes 


-15.5 


Boots and shoes 


-38.8 




-20.6 


Cotton goods 


-22.4 


Woolen and worsted goods 


-26.3 




-21.9 


Newspapers and periodicals 


-20.1 







Source: Hourly Earnings, National Industrial Conference Board, Wages, Hours, and Employment in 
the United States, 1914-36; Output per Man-Hour, National Research Project, Production, Employment, 
and Productivity in Fifty-nine Manufacturing Industries, 1919-36, pt. II: Unit Labor Cost, U. S. Bureau 
of Labor Statistics, Monthly Labor Review, December 1939, "Employment and Production in Manufac- 
turing Industries, 1929 to 1936," p. 1404. 

In each of the industries the increase in output per man-hour was far 
greater than the advance in hourly earnings. Consequently, unit labor 
costs in each were materially lower in the latter than in the former year. 
Interestingly enough, those industries in which average hourly earnings 
advanced most noticeably — knit goods (36.1 percent), rubber products 



earnings are derived by dividing pay rolls by man-hours, and since the index of output per 
man-hour is derived by dividing the index of production by that of man-hours, the factor, 
.man-hours, is thus common to both average hourly earnings and output per man-hour. 
Dividing average hourly earnings by output per man per man-hour is equivalent to divid- 
ing the index of pay rolls by that of production, which is the first method cited. Actu- 
ally, however, there is a difference in results, in that average hourly earnings are com- 
puted by the National Industrial Conference Board and the Bureau frf Labor Statistics 
from man-hour and payroll samples which are not ad.justed for aggregates, whereas the 
index of pay rolls, published by the Bureau of Labor Statistics, used with the index of 
production, is computed from a larger sample by the chain index method to represent 
changes in aggregate pay rolls. Therefore this latter method is used herein. 

"8 U. S. Bureau of Labor Statistics, Monthly Labor Review, Decembeu-1939, pp. 1103- 
1404. . .— - i-i 



CONCENTRATION OF ECONOMIC POWER 



149 



(28.0 percent^ , newspapers and periodicals (24.4 percent) — were char- 
acterized by some of the most extensive increases in output per man- 
hour (66.2, 79.6, and 45.8 percent, respectively) . Regardless of whether 
increased hourly earnings made it necessary for producers to increase 
greatly the productivity of the labor force or whether the increase in 
output per man-hour made it possible for producers to pay higher 
hourly wages, the important fact is that unit labor costs did decline. 



tr 

UJ 
Q. 

H 

Q- 

h- 
ID 
O 



< 

X - — ) 

w o 



CO 



Si 



< 

X 

o 



1 1 

=5 < 



^°i- [ 





SI g s 



PJ ^1'-'-^ 




m 



-^ 



^.:x:^'jf- 


w| 


§' "'■ 1 




g S , f. 
g S ? 8 

-(''■"I" ! '■ !■■■ 


8 ? 
, ? ? 


Z 1 




^- : :i 




S 2 , 


' ' \ " : 


1 

..-..L.vJ 


o i. 1 , 


T„l,:.i.i 


g 8 S S ° 8 ? 
S S S S o ? s 


§ t 


g S S S ° 8 f 
g S S 8 o ? ? 




■Hi 


^Lkte 



lit 

o55 



111 
1 



This decrease was less than 20 percent in only one of the 11 indus- 
tries — paints and varnishes. It should be noted that this comparison 
stops short of very recent years in which advances in hourly earnings 
took place caused by legislation such as the National Industrial Recov- 
ery Act, the Wages and Hours Act, the Fair Labor Standards Act, and 
the National Labor Relations Act, and by accompanying marked 
growth in union organization. 

A comparison of changes in hourly earnings, output per man-hour, 
and unit labor costs in the short but "dynamic period of 1935-39 can be 
made for 13 diversified industries (table 14). 



150 



CQJNTENTRATION OF ECONOMIC POWER 



Table 14. — Percent change in hourly earnings, output per man-hour, and unit 
labor cost in 13 manufacturing industries, 1935 to 1939 





Percent change from 1935 to 
1939 in- 


Industry 


Average 
hourly 
earnings 


Output 
per man- 
hour 


Unit 
labor 
cost 


Iron and steel 


+27.0 
+21.6 
+23.1 
+20.1 
+25.7 
+22.2 
+3.5 
-1.8 
+17.2 
+ 12.6 
+16.3 
+10.4 
+12.0 


+27.5 
+26.0 
+11.5 
+10.5 
+58.1 
+25.2 
+20.9 

+6.9 
+17.3 

+3.9 
+13.9 

+7.2 

-3.8 


-0.4 




-3.5 




+10.4 


Paints and varnishes 


+8.8 




-20.5 




-2.4 


Ootton goods . . - 


-14.4 


Boots and shoes 


-8.1 




-. 1 




+8.4 


Bread and other bakery products 


+2.1 


Flour 


+3.0 




+ 16.4 







Source: U. S. Bureau of Labor Statistics, Monthly Labor Review, July 1940, p. 36. 

During this short span of years the marked increases in hourly 
earnings were generally matched or even exceeded by increases in 
output per man-hour. In only five of the industries was the increase 
in hourly earnings greater than in output per man-hour. And in 
two of these five, unit labor costs rose only 3.0 and 2.1 percent, each. 
In the rest (except cane-sugar refining wherein labor productivity 
declined) the advance in output per man-hour exceeded the increase 
in hourly earnings. In four of the six industries with wage in- 
creases exceeding 20 percent, still greater increases in output per 
man-hour were achieved and lower unit labor costs resulted.^'' 

It is thus apparent that even during the period of the most 
dynamic upward movements in hourly earnings, the increase in 
output per man-hour nearly matched ^""out of 13) or exceeded (7 
out of 13) the rise in hourly earnings, resulting in stability or 
decreases in unit labor costs. The foregoing comparison casts con- 
siderable light on the effect of technology upon wage costs in relation 
to output. But to obtain a more inclusive picture, it is desirable to 
extend the analysis of unit labor costs over a longer period of time. 
It has been possible to extend the production indexes for 20 indus- 
tries, using the same sources and methods as the National Research 
Project to later years than the published termination period of most 
of their indexes, 1935-36.^° 

Production indexes for 1938 were obtained for 18 and indexes for 
1939 were obtained for 13 of the 20 industries. Comparable pay-roll 
series were available for all 20. For 17 industries comparable data 
for the year 1919 were available and unit labor cost series were com- 
puted. The trend in unit labor costs for each of these industries 
evidenced by behavior in the selected years of 1919, 1923, 1929, 1933, 
1935, 1936, 1937, 1938, 1939 is shown in chart XI, table 15. 

^ While unit labor costs may be Influenced not only by the use of labor-saving tech- 
niques but also by the rate of operation, a decline in unit labor costs affects the con- 
sumption-saving problem regardless of its cause. 

*" Original use of the National Research Project inde.xes and the Bureau of Labor Statis- 
tics pay roll indexes for the purpose of estin^ating changes in unit labor cost was made 
by the U. S. Bureau of Labor Statistics, and he unit labor cost figures thus derived were 
published in the December 1939 issue of ♦he :onthly Labor Review under the title "Em- 
ployment and production in ManufactUiing .ndustries, 1919-36," pp. 1397-1404. The 
projection of the National Research Projec* indexes of production forward to 1939 and 
their use with pay-roll indexes for the years 1919-,3.3 and 1936-39 in the computation of 
unit labor cost were by Victor Perlo, U. S. Department of Commerce, and Witt Bowden, 
U. S. Bureau of Labor Statistics ; the figures thus derived were published in the July 1940 
issue of the Monthly Labor Review, p. 33, under the title "Unit Labor Cost in 20 Manu- 
facturing Industries, 1919-39." 



CONCENTRATION OF ECONOMIC POWER 



151 



Chart XI 

UNIT LABOR COST IN 20 MANUFACTURING 
INDUSTRIES 

SELECTED YEARS. UNITED STATES. 1919-1939 



















^,RO..ST.e. 














^-\ 














"^=^^><f-'~"r" 


/r^ 




^> 


^^^/^ 

























































1929 1933 193} 1939 



BUILDING MATERIALS GROUP 



\^«MICALS 














>X^ 
















J<^ 


4; 


"" "-• 






r'-\ 


1 ,1-r 




^%SE 


rii' 


PETROLEUM REFININC^"^"^ ^ 


i 


" 


I 


■f 


^H 


i 


1 




1 



































iii^ 


"^ 


n.ANlNG-M 

ku-BEP 


i TIM 


10 
BE 




^ 


5 


«oS^cr./ 


',r.iifo^,,< 


••■■••■^ 


X/\ [' 


























. 























1919 1923 



119 1923 1929 1933 1933 1939 




1919 1923 



933 1935 1939 




"CONFECTIONERY 



Source : Table 15. 



152 



CO^'CENTRATION OF ECONOMIC POWER 



Table 15. — Indexes of unit lalor cost in 20 manufacturing industries, 1919-39 

[1923=100] 



Blast furnaces, steel works, 

and rolling mills 

Nonferrous metals: Primary 

smelters and refineries 

Petroleum refining ".. 

Chemicals 

Fertilizers 

Paints and varnishes 

Rayon 

Planing-mill products 

Lumber and timber products. 
Clay products (other than 
pottery) and nonclay re- 
fractories 

Cement 

Cotton goods 

Boots and shoes _._ 

Paper and pulp 

Newspapers and periodicals... 
Bread and other bakery prod- 
ucts. ._ 

Confectionery 

Flour '. 

Ice cream 

Cane-suga'- refininc 



1919 


1923 


1929 


1933 


1935 


1936 


1937 


1938 


129.2 


100.0 


81.8 


69.3 


79. 9 


76.0 


91.5 


91.3 


119.1 


100.0 


70.5 


54.1 


69.6 


72; 6 


79.5 


83.2 


140.7 


100.0 


73.0 


57.5 


62.2 


59.7 


64.5 


64.6 


1.56. 1 


100.0 


77.7 


61.2 


70.5 


68.6 


79.8 


85.3 


140.3 


100.0 


86 5 


61.9 


76.0 


70.4 


75.9 


71.4 


108.5 


100.0 


96.7 


90.2 


84.6 


79.9 


90.0 


94.7 


0) 


100.0 


72.4 


32.8 


36.8 


34.6 


36.4 


34.0 


96.6 


100.0 


94.3 


71.5 


75.4 


(') 


80.3 


0) 


107.2 


100.0 


90.8 


64.0 


74.2 


76.7 


84.5 


80.2 


98.3 


100.0 


90.1 


63.2 


81.1 


81.3 


89.5 


91.0 


111.3 


100.0 


79.3 


60.5 


76.1 


64.9 


80.4 


79.6 


102.5 


100.0 


81.5 


64.1 


79.3 


6*<.9 


75.2 


72 8 


89.3 


100.0 


81.3 


55.3 


61.3 


56.0 


63.6 


58.7 


112.2 


100.0 


85.5 


58.9 


70.0 


68.1 


70.7 


74.6 


101.5 


100.0 


92.2 


83.8 


79.9 


79.4 


77.4 


91.3 


(') 


100.0 


100.2 


89.9 


99.0 


98.7 


108. 


104.5 


(•) 


2 100. 


85.2 


60.6 


59.7 


57.2 


63.1 


(') 


110.9 


100.0 


85.0 


67.2 


80.2 


79.2 


86.4 


83.4 


86.6 


100.0 


83.7 


67.2 


58.3 


50.2 


49.2 


50.4 


119.3 


100.0 


77.1 


68.7 


73.2 


'" 


78.2 


80.7 



79.6 

(') 

60.0 

77.8 

(0 

92.0 

29.3 

(0 

(') 



0) 
74.4 
67.8 
56.3 



101.0 

0) 
82.6 

(0 
85.2 



' Not available. 
2 1925. 

Source: Victor Perlo and Witt Bowden, "Unit Labor Cost in Twenty Manufacturing Industries, 1919 
to 1939," U. S. Bureau of Labor Statistics, Monthly Labor Review, July 1940, p. 34. 

It is evid'ent that the long-term trend in nnit labor costs has been 
definitely downward. The downward trend was very noticeable up 
to 1933 in both iron and steel and primai'y smelters and refineries. 
The depression declines in production of these industries were prob- 
ably the major cause of the increase in unit labor costs during 1933-35. 
Subsequent higher wage rates undoubtedly were the prime factor in 
the further advance. But in 1939 unit labor costs in iron and steel 
fell well below the 1937 level and even below the 1929 level upon the 
attainment of a fairly high rate of production — the production index 
being 9 percent below 1929 and 1937 levels. 

In primary smelters and refineries, unit labor cost data are not yet 
available for 1939. The gi'eat curtailment of production in this in- 
dustry since the depression, accompanied by recent advances in wage 
rates, is doubtless responsible for the upward trend in unit labor costs 
since 1933. 

In the chemical group, the most pronounced decline in unit labor 
costs took place in the rayon industry, and the least pronounced in 
paints and varnishes. Despite recent increases in hourly earnings in 
these industries, unit labor costs in 1938 were below the 1929 level in 
petroleum refining, fertilizers, paints and varnishes, and rayon. Only 
in chemicals were these costs higher in 1938 than in 1929, but by 1939 
they were again down to the 1929 level. 

Unit labor costs in the building materials gi^oup followed a similar 
pattern, with a gradual decline to 1929, a marked fall from 1929-33, 
and gradual increases thereafter — increases in large part due to the 
lack of recovery in production of these industries. 

In tlie a]:iparel and ])aper grou]), unit labor costs fell similarly 
between 1923 and 1933 in ])a]ier and pulp, cotton goods, and Ixwts 
and shoes. After 1933 these tliree showed minor increases, but boot.s 
and shoes and cotton goods turned downward in 1938. and paper and 
l^ulp in 1939. Newspapers and periodicals declined all through the 



CONCENTRATION OF ECONOMIC POWER 



153 



period up to 1938 when they increased abruptly, again turning 
downward in 1939. 

The industries in the baking and confectionery group showed diver- 
sified trends over the long-term period, with ice cream and confec- 
tionery at one extreme falling to a low level, and bread and other 
bakery products at the other extreme remaining comparatively stable. 
The trend after 1935 in flour follows the general pattern of a slight 
decrease in 193G, an increase in 1937, and a decline thereafter. But 
in cane-sugar refining the decline toward the end of the period does 
not occur; in 1939 the series was still climbing. This may be due in 
part to production being lower in 1939 than the 1935-38 levels. 

The general tendency, reflected in the behavior of unit labor costs 
•in these specific industries, for increases in output per man-hour 
to exceed those in average hourly earnings is indicated also by the 
indexes for the broad fields of manufacturing as a whole, steam 
railroads, bituminous coal, and anthracite mining. 

Chart XII 



3 
O 

X 

UJ 

<l- 
SCO 

S3 



ii 



u 
a. 



CO 



OCT 

if) 
UJ 
X 
UJ 
Q 



it' 


- 


' t 


i 




s 


ss 


N 


•■ ,=^ / 




" 




> 

1 




1 






> 1 








\ 


\ 








N 


V 


^ ( 




Q 




\ 


N 


s \ 




^ 




■ y 




f ) 




_l 




^ 


/ 




< 




1 


^ 








1 


s 




< 




s 


g 




\-r, 




s 


~\\ 


5 








1 


__L 










1 i 










Vi 












Y 






1 









? 




r 


8 


\ 








. 


^ 




! 




1 


/ 








[ 


\ 




~V^', 




\ 


Ss 




i\ 




j 


\ 






L 




\ 


"—1 




\ 




) 


i 


L'^ 


-Jr- 




1 


i 


Ah 


— |- 


i 


8 


\\ 










\ 

















\ 




. 


in 

rlr 


-/^ 


<^ 




• 




•N 


\ 




\ 








VI 








< 


vi- 












\ 










)^. 


! 






I i 

1 i 








W 




























/ 





? « 


\ 




t s 

^ II 




__\ 




Y 


^ '•> 








\ 






N 


t'4^ 






\ 






; 








{ 






K 






n 


















^^ 


i~i 



















« 


1 


- 


s 




n 






. 


i= 








11 








i,» 






s 


ii 


i 


^i 








IS 




15 








n 




s. 










? 


2S 



^i ii 



154 



CONCENTRATION OF ECONOMIC POWER 



In each, a scissors-like divergence took place; in each output per 
man-hour evidenced a greater long-term increase than average hourly 
earnings with the result that the secular trend in unit labor costs is 
downward. This is apparent over the long-term period 1923-39 in 
chart XII, table. 16.^^ 

Table 16. — Indexes of output per man-hour, average hourly earnings, and unit 
lahor cost; 1923-39 

[1923=100] 



Year 


Output Average 

per man- hourly 

hour earnings 

1 


Unit 
labor 
cost 


Output Average 

per man- hourly 

hour earnings 


Unit 
labor 
cost 




:Manufacturing 


Steam railroads 


1923 


1 100. 
105.8 
113.2 
116.9 
120.8 
129.5 
131.9 
131.6 
141.3 
137.7 
144.8 
147.9 
158.8 
161.8 
157.5 
159.8 
174.5 


1 100. 
104.7 
104.5 
105.4 
106.1 
107.6 
108.8 
.106.1 
99.4 

87^4 
104.0 
107.4 
108.5 
121.9 
122.8 
123.8 


3 100. 
98.9 
92.2 
90.0 
87.8 
83.0 
82.5 
80.7 
70.3 

59:6 
70.3 

65^4 
74.0 
73.1 
68.1 


1 100.0 
103.0 
108.5 
111.1 
110.8 
116.3 
118.2 
118.0 
118.8 
116.1 
129.3 
130.1 
135.9 
145.7 
148.5 
146.8 
154.9 


2 100. 
101.8 
103.0 
103.0 
105.0 
106.3 
108.4 
109.8 
110.8 
101.7 
101.0 
102.3 
110.7 
112.0 
115.1 
121.1 
121.6 


8 100.0 




99.0 


1925 . 


95.1 


1926 - 


92.9 




94.8 


1928 - 


91.6 


1929 


91.9 


1930 -- 


93.0 


1931 


93.3 


1932 --- 


87.8 


1933 . . 


78.2 


1934 


78.7 


1935 - --. 


81.6 


1936 


77.1 


1937 . _ 


77.6 


193S - - -- 


82.7 


1939 --- 


78 6 








Bituminous-coal mining 


Anthracite mining 


1923 - 


'100.0 
101.7 
101.0 
100.3 
101.6 
105.3 
■108. 1 
112.8 
118.0 
115.9 
110.0 
111.9 
115.4 
121.5 
124.8 
130.4 
142.1 


3 300.0 
96.2 
94.7 
93.0 
88.9 
84.8 
80.6 
81.1 
76.6 
61.6 
59.3 
79.5 

94.0 
101.4 
104.0 
104.8 


3 100. 
94.6 
93.7 
92.8 
73.2 
80.5 
75.6 
71.9 
64.8 
54.0 
55.1 
72.2 
76.2 
76.2 
80.4 
80.3 
72.5 


1100.0 
94.4 
94.9 
95.4 
98.3 
98.1 
96.4 
94.0 
100.2 
115.0 
126.4 
118.5 
121.3 
135.1 
142.8 
166.1 
172.6 


8 100. 
108.3 
107.7 
107.3 
106.6 
106.0 
105.4 
104.8 
104.2 
104.3 
103.5 
104.4 
104.1 
105.4 
110.4 
116.5 
116.8 


3 100.0 


1924 . . 


114.7 


1925 


113.7 


1926 - 


112.5 


1927 . . 


108.6 


1928 


108.2 


1929 


109.4 


1930 - 


111.1 


1931 


104.8 


1932 - 


90.8 


1933 ... 


79.8 


1934- 


84.7 


1935 - - 


80.7 


1936 - 


73.3 


1937 


72.9 


1938 - 


66.9 


1939 


62.7 







• See table 1. 

» U. S. Bureau of Labor Statistics, "Wages, Hours, and Productivity of Industrial Labor, 1909 to 1939," 
September 1940. ., , , , ^ , 

' Computed by Witt Bowden of the U. S. Bureau of Labor Statistics from data compiled by the Inter- 
state Commerce Commission, U. S. Bureau of I-abor Statistics, U. S. Bureau of the Census, and U. S. Bureau 
of Mines. 

"Oespite substantial advances in hourly earnings during the latter 
part of the thirties, labor productivity increased to such an extent 
that unit labor costs in 1939 were at an all-time low in anthracite 
mining and were almost down to the previous all-time low in steam 
railroads. In manufacturing, unit labor costs dropped in 1939 to a 
level exceeded only by the lows of 1932-33 and 1935-36 when hourly 
earnings were materially less than in 1939. 

The increase in average hourly earnings during the latter part of 
the thirties represented in each" case a striking departure from the 

« These series on unit labor costs for these four broad fields represent Indexes compiled 
by Witt Bowden, of the U. S. Bureau of Labor Statistics, from data supplied by the 
agencies cited in the table and chart. 



CONCENTRATION OF ECONOMIC POWER 155 

long-term trend. As is well-known, a combination of factors, in- 
cluding the enactment of labor legislation validated by the Supreme 
Court, were largely responsible for this sudden upturn in wages. 
Since it is unlikely that this combination of factors will recur in the 
near future, a levelling off in the rate of increase in hourly earnings 
may be expected. No such behavior in the rate of increase in labor 
productivity is anywhere indicated ; on the contrary, even more rapid 
advances in output per man-hour are likely. 

This possible widening of the spread between labor productivity 
and hourly earnings obviously means even greater declines in unit 
labor costs. If prices remain constant — and in a large segment of 
the economy they have been markedly stable — ^this decline in unit 
labor costs will further reduce labor's share of the value of products. 

TECHNOLOGY AND THE EMPLOYED WORKER 

NERVOUS AND MENTAL STRAIN 

The very subdivision of labor inherent in the factory method of 
production connotes the continual performance of standardized op- 
erations by the worker. Where technology has so transformed pro- 
ductive processes that the individual worker merely attends a semi- 
automatic machine, monotony and boredom are inevitable, and the 
worker who prefers the monotonous routine imposed by numerous 
teclmological processes is generally of low intelligence."^ 

In .recent years technology has perhaps tended to lessen the amount 
of highly repetitious, standardized work, but it has accomplished this 
only by eliminating the human element altogether. An apparently 
increasing tendency is the replacement of the individual worker with 
completely automatic processes or multi-purpose machines which 
perform the necessary functions. It might be observed, for example, 
that the muscular cramp which Morse telegraphers acquire after 
years of continual tapping has largely disappeared owing to the 
replacement of the Morse key by the teletype system. A recent 
British study of machine- feeding processes states : 

It is probable * * * that many of the operatives employed on machine- 
feeding will continue to suffer from boredom and strain. Only certain types 
of individuals are able to behave as mechanically as the machine on which 
they work and a number cerfainly find such conditions almost intolerable. 
Perhaps the chief hope of escape lies in the possibility of complete mechaniza- 
tion.*' 

Monotony, in addition to stunting personality development and 
stifling imagination, contributes to the occurrence of industrial ac- 
cidents. An example of this, one of the deleterious effects of mo- 
notony on industrial workers, is shown by the following experience 
of a worker in a printing shop. 

In "flat work" I very often used to daydream. The purely mechanical action 
of picking up a sheet of paper, and setting it against "marks" all day long, 
brings about a state of coma * * * Click, pick the sheet up. Bang, air 



«2Cf. Ethelbert Stewart, "Industrialization of the Feeble-Mlnded," U. S. Bureau of 
Labor Statistics, Monthly Labor Review, July 1928, pp. 7-14. 

«3 Medical Research Council, Industrial Health Research Board, The Machine and the 
Worker, Rept. No. 82, by S. Wyatt and J. N. Langdon, 1938, p. 43. 



156 CONCENTRATION OF ECONOMIC POWER 

it (make the air go under it by a flick of the wrist). Shush, bring the sheet 
to the "marks." Here the machine takes it. Repeat Ibis all (lay long, 8 
hours a day, 48 hours a week, and you may be able to visualize the 
effect on the mind. It was on these machines that I day-dreamed, vmtil I 
had the good luck to put my hand in instead of the paper. I say good luck, 
because it happened to be a small machine, and I was able to brake it quickly. 
In fact, by taking my hand, it stopi>ed itself. I had to have the top of one finger 
grafted on, and several of the bones in the hand reset. Needless to say, I never 
day-dream now.*^ 

Monotony also may lead to actual decreases in worker produc- 
tivity. A British inquiry into fatigue and boredom in repetitive 
work concluded : 

* * * (a) Boredom depresses the rate of working, (b) workers of in- 
ferior capacity tend to work more closely to their maximum than do workers 
of superior capacity, and (e) the rate of working when learning an industrial 
process is reduced when the worker is bored. '^'' 

The report goes on to point out that monotony also affords a 
fertile field on which the seeds of discontent may fall : 

Boredom and discontent are also' closely connected with the type of work, 
and even slight differences between one process and another may have widely 
different effects on the operative. Efficiency and contentment may be increased 
by giving the beginner a short trial on different types of work and assigning 
her to the process which she likes best."'' 

Discontent often leads to a deliberate slowing-down of Avork in 
all types of employment. Machinery over whose speed and rhythm 
the worker has no control is liable to produce strain and tension 
accompanied by sullen acquiescence or even resentment. Machine- 
feeding of a more or less automatic kind is, therefore, likely to in- 
volve human costs which are seldom fully evaluated, but which 
contribute to industrial inefficiency. Among them are the definite 
costs of unrest, absence, ill health, accidents, etc.^^ 

When technological equipment is operated at high rates of speed 
fatigue usually rises rapidly. Output per worker may fall so 
greatly as the Avorking day progresses that productivity of the 
employee over the entire day may drop below the pre-speed-up level. 
The British Medical Research Council discovered that when ma- 
chinery was operated at an abnormal rate of speed output per worker, 
after rising in the early part of the spell, began to fall at an ex- 
tremely rapid rate until the work ceased. In one process the output 
in the last hour of work was 19.5 percent below the highest level 
recorded during the day. In the words of the report : 

There can be little doubt, therefore, that fatigue was fairly severe and 
arose from the intensity of the work. The maclunes acted as pace-makers, 
and the operatives were impelled to work at a rate which they were unable 
to maintain. * * * 

The results obtained in this process accordingly suggest that the discrepancy 
between the speed of the machines and the capacity of the operators was too 
great for personal comfort and efficiency, and that a slower machine speed 
would probably result in a higher total output and Increased pleasure in work. 
This was found to be the case in a. box-wiring process where similar conditions 
prevailed and a reduction in the speed of the machine caused an increase in 

"' .\. P.nrratt Brown, The Maohino and the \\'orkor, Nicholson and Watson, London. 
1934, p. 77. 

"'Medical Research Council. Industrial Health Re.search Board. Fatigue and Boredom in 
Repetitive Worlv, Rept. No. 77, by S. Wyatt and J. N. Langdon, London, 1937. p. 73. 

«" Idem. 

•" See P. Sargant Florence, Economics of Fatigue and Unrest, Henry Holt, New York, 
1924, ch-. 5. 



CONCENTRATION OF ECONOMIC POWER 157 

output of 10 percent. THere seems to be no justification for machine speeds 
wtiich are greatly in excess of the natural rate of working of the operatives 
and it is difficult to understand why such speeds are set.*^* 

Fatigue, like monotony, often leads to accidents. In .the United 
States and Great Britain investigations have sliown that accidents 
increase as wage-earners go through the working periods, rising from 
a low point at the beginning of the morning shift to a high point just 
before the noon rest period, then falling at the beginning of the 
afternoon only to rise again toward the end of the day. A decline just 
before the end of each working period is attributed to the stimulating 
prospect of a change from work.^^ 

It is difficult to isolate the specific cause of an industrial accident. 
One authority has observed : 

As in the case of sickness, * * * the causes of accidents are sufficiently 
numerous to make one wary of saying that fatigue is the directly responsible 
factor. There undoubtedly is a causal relationship, but fatigue does not operate 
alone.^o 

H. W. Heinrich, of the Travelers Insurance Co., analyzed the causes 
of 50,000 industrial accidents and found that mechanical hazards were 
responsible for only 10 percent of the mishaps. The remaining 90 
percent were due to faulty instruction (30 percent), lack of concen- 
tration (22 percent), unsafe practices (14 percent), poor discipline (12 
percent), inability of employee (8 percent), physical unfitness (3 per- 
cent), and mental unfitness (1 percent).'^ Fatigue was undoubtedly 
an element in many of the cases involving lack of concentration and 
perhaps in some involving unsafe practices. 

Although the terms "speed-up""' and "stretch-out" have been rather 
indiscriminatel}' applied to practically every type of labor intensifi- 
cation, the former generally indicates an increase in the rapidity with 
which a worker must perform a given set of functions, while the 
latter connotes an increase in the number of functions a worker must 
perform within a given period of time. 

Labor-intensification systems have been widely applied throughout 
the industrial world. There is, for example, the speed-up inherent in 
the application of the straight-line system of production in lieu of the 
bundle system in the manufacture of cotton garments. This in- 
dustrial process has been described in chapter I,^- but it deserves 
further consideration here because of its effect upon conditions of 
work. 

The straight-lme system makes the operator keenly aware that if 
she falls behind in her work, the operator next in line is rendered idle, 
waiting for the garment on which she is working. Unless each opera- 
tion is completed promptly, the rest of the line will be delayed and 
the foreman will arrive to investigate the trouble. 

The necessity of speed under these circumstances is obvious. The worker can 
no longer work at her own natural pace. She must adjust herself to the speed of 
her fellow worki^rs, all of which undoubtedly results in increased individual 
productivity. When the workers are of a fairly uniform skill and equal tempo, 

*" Medical Research Council, Industrial Health Research Board, Rept. No. 82, op. cit., p. 6. 

•* See P. Sar^ant Florence, op. cit., p. 351 ; Josephine Goldmarlc, Fatigue and Efficiency, 
Russell Sage Foundation, New Yorl<, 1912, p. 76; and H. M. Vernon, Industrial Fatigue 
and Efficiency, Oeorge Rutledge & Sons, London, 1930, ch. 6. 

■"> Carroll R. Daugherty, Labor Problems in American Industry, Houghton Mifflin, Boston, 
1933, p. 13.5. 

" U. S. Bureau of Labor Statistics, Monthly Labor Review, "PreTention of Industrial 
Accidents," June 1931, p. 74. 

'2 See pp. 119-120, supra. 



158 CONCENTRATION OF ECONOIvITO POWER 

the uniformly high .?peed of the work, while it may result in greater fatigue at 
the end of the work period, need not necessarily affect injuriously the health of 
the worker. But where no attention is paid to that factor, or still worse where, 
as a result of deliberate design, fast workers are inter:; persed with workers of 
a naturally slower tempo, the health of the worker may be sr:i.,;,^,;, affected. 

In one of the plants studied, management deliberately resorted to such an 
arrangement which caused the slower workers to strain themselves to the ut- 
most. On the one hand, the slower worker was aware that work was piling up 
behind her from the fast worker on the preceding operation. On the other 
liand, she saw that the equally fast worker on the succeeding operation was 
-waiting for her work. The result was a large increase in productivity by the 
Individual workers, but at the cost of great physical and nervous strain on 
the part of the naturally slower workers. In the end, such an arrangement is 
bound to be harmful not only to the slower worker but to the efficient operation 
of the shop. 7 3 

The speed-up system may derive its tempo either from the inter- 
spersion of the more efficient workers throughout the labor force, as 
in the above case, or from the operation of machinery at a speed higher 
than the usual rate of the average worker. The latter method is 
particularly applicable to those machine-feeding processes in which 
the operator's failure to keep the machine fully supplied with material 
does not inj ure the machine or the product. In such cases the machine 
has often been set at a speed which equals or exceeds the highest rate 
attainable by the best operator, since it is believed that this increases 
output. An example of this procedure is cited by the British Medical 
Research Council. 

In this process the operator was engaged in feeding a rotating dial which, 
■when completely supplied with material, necessitated the repetition of a fairly 
•complicated cycle of movements 37 times a minute. Six operators were observed 
for a period of 3 weeks and their output was recorded at hourly intervals 
throughout the day. * * * 

The results show, in the first place, that the average efiiciency was low (64.3 
percent) and even in the best hour of the day it was only 68.2 percent. In this 
process enforced stoppages were practically negligible, therefore the low efficiency 
-was due almost entirely to the failure of the operators to keep the machine fully 
supplied with material. Even the most capable worker was seldom able to 
keep pace with the machine for more than 10 seconds, so that the maintenance 
of a regular rhythm was impossible. Further, it was noticed that the operators 
became somewhat embarrassed and emotionally disturbed each time they failed 
to remove and replace the articles on the dial as it moved beyond their reach. A 
little time elapsed before they were able to recover their poise, and although on 
each occasion the effect of this disturbance on output was scarcely measurable, 
the total effect was appreciable.'* 

Management has usually failed to recognize that the application of 
speed-up and stretch-out systems may actually result in a diminution 
of labor productivity. Certain industries have been characterized by 
an almost constant intensification of the amount of work demanded 
from the individual employee in the belief that the greater the intensi- 
fication, the higher the productivity of the individual worker. For 
example, in the automobile industry during the d'^pression workers 
were reportedly offered bonuses to exceed the production rates estab- 
lished by production engineers. When these levels had been surpassed, 
the new rates were accepted as normal. Then intense pressure was 
applied for even higher records of output which in turn were regarded 
as normal. 

With the coming of the depression the universal testimony of the auto workers 
is that speed-up increuoeil beyond the powers of human endurance." 

" U. S. Bureau of Labor Statistics, Bulletin No. 662, Productivity of Labor In the 
Cotton-Garment Industry, by N. I. Stone, A. Cahen and S. Nelson, November 1938, pp. 
65—56. 

T* Medical Research Council, Industrial Health Research Board, Rept. No. 82, op. "clt., p. 5. 

" Hearings before the Temporary National Economic Committee, Part 30, p. 1637. 



CONCENTRATION OF ECONOMIC POWER 159 

In 1935 the National Recovery Administration said of the speed-up 
system in the automobile industry : 

The only reason that it (speed-up) can exist as at present is because of the 
huge available supply of labor through which as one man falls by the wayside, 
another is there to take his place.'" 

This report went on to state : 

Everywhere workers indicated that they were being forced to work harder and 
harder to put out more products in the same amount of time and with less 
workers doing the job. There was a tendency to excuse the automobile manu- 
facturers for lack of steady work, "that is caused by market conditions." But 
when it comes to increasing their work loads, they (the workers) are vigorous 
in denouncing the management as slave-drivers and worse. If there is any one 
cause for conflagration in the automobile industry, it is this one." 

In some cases the production staff in charge of the time study de- 
partment received' a bonus each time they were successful in reducing 
the operating time of a given function. Instead of "taking honest time, 

* * * (they) would look and see what your record was last year 
and maybe boost a little more for the last model," according to the 
president of the United Automobile Workers in his testimony before 
the Temporary National Economic Committee. He continued : 

* * * because the workers in automobile plant had been speeded up so 
much, there was a tendency on their part, when a man came around to 
time-study them, to try to cheat on that time study if they possibly could, and I 
was the same as every automobile worker, I might say.'* 

But this trick was of little avail if the time study department took 
as a basis in computing the worker's potential productivity not his 
time-study record but that of his output over a previous period, per- 
haps increased by a definite percentage. 

The fear of unemployment resulting from the introduction of speed- 
up and stretch-out systems leads to discontent and, in some cases, to 
strikes. For example, the application of the stretch-out system in a 
cotton mill increased the number of automatic looms attended by an 
individual weaver from 35 to as many as 100 and made possible a 
reduction in labor costs by the use of less skilled, lower-paid workers. 
Even though the workers, through a fully recognized and long-estab- 
lished trade union, were given the right to cooperate with manage- 
ment in determining conditions under which the stretch-out system 
was to be applied, the attendant labor displacement became an in- 
creasingly acute source of dissatisfaction, leading finally to a bitter 
and protracted strike. 

The workers on their part were dissatisfied because 350 of their number, in- 
cluding some permanent workers, had been eliminated, because as many more 
had been demoted at reduced wages, and because they soon came to believe that 
the new job assignments had created excessive job burdens. 

The reduction in their wages "created in their minds a conviction 
that only by preventing further 'stretch-outs' could they preserve their 
security and well-being." ^* 



'« National Recovery Administration, Division of Research and Planning, Preliminary 
Report on Study of Regularization of Employment and Improvement of Labor Conditions 
In the Automobile Industry, 1935 ; quoted in Hearings before the Temporary National 
Economic Committee, Part 30, p. 16371. 

'^Ibid., p. 16371. 

" Hearings before the Temporary National Economic Committee, Part 30, p. 16370. 

" The Personnel Journal, February 1934, " 'Labor Extension' in a Cotton Mill," by R. C. 
Nyman, Institute of Human Relations, Yale University, p. 270. 



160 



CONCENTRATION OF ECONOMIC POWER 



These systems gf intensified labor— all designed to place the amount 
to be paid on a unit of production basis — have aggravated nervous 
strain, monotony and fatigue, increased industrial accidents, and in 
certain cases actually decreased productivity.^" 



PHYSICAL HARDSHIPS 

Industrial accidents. 

While industrial accidents, largely as a result of continuing efforts 
on the part of various governmental agencies, have tended to decline 
over a long-term period, their incidence is still sufficiently high to 
make them a major characteristic of modern technology. In 1937, a 
year of moderately intense activity, a total of 1,838,000 industrial 
injuries occurred throughout the United States representing 40,159 
injuries per million workers.*^ The estimated number of industrial 
injuries in 1937 per million workers is shown by industry group and 
type of disability in the following table. 

Table 17. — Estimated number of industrial injuries per million workers, hy 
industry group and type of disability, 1937 



Industry group 



All industries 

Agriculture • 

Mining and quarrying (includes petroleum and natural gas) 

Construction i 

Manufacturing '_ _ 

Public utilities 2 _ 

Trade— wholesale and retail '... 

Railroads' 

Miscellaneous transportation < 

Services and miscellaneous industries * 



Extent of disability 



40, 159 
24. 978 
115,048 
112, 230 
36, 379 
17, 591 
29,407 

3C! 730 
40, 842 



Per- 
manent 



2,702 
1,247 
3,522 

5,877 
2,772 
746 
3,895 
1,190 
1,887 
2,812 



37, 027 
23, 315 
109, 498 
105, 305 
33, 371 
16, 525 
25, 173 
32. 228 
34,088 
37, 772 



1 Based on fragmentary data. 

2 Based on comprehensive survey. 

3 Based on Interstate Commerce Commission data. 
< Based on small sample studies. 

Source: M. D. Kossoris and S. Kjaer, "Industrial Injuries in the United States During 1937," U. S. 
Bureau of Labor Statistics, Monthly Labor Review, March 1939, p. 600. 

The occurrence of industrial injuries is influenced by fluctuations in 
the business cycle. There exists a fairly close relationship between 
the trend of the frequency rate of industrial injuries, which measures 
the average number of injuries per million man-hours, and the volume 
of employment. ^^ Theoretically tlie frequency rate should be con- 
stant regardless of the volume of employment. Actually, however, 
it declines rapidly on the downswing of the business cycle for the 
following reasons: (1) Labor forces are curtailed, with those most 
recently added laid off first. Skilled or .semi-skilled workers with 
long years of service are retained by management to form a nucleus 
for subsequent expansion. Such workers are generally thoroughly 
familiar with job hazards and have developed safety habits. (2) 



*" For a discussion of the effect of these practices upon the individual worker, see Morris 
Cooko and Philip Murray, Organized Labor and Production. IIariK>rs, New York, 1940, 
ch. 8, "Some of MaiiaKonient's Controversial Practices and Attitudes." pp. n4-10G. 

"Max Kossoris and Swen Kjaer, "Industrial Injuries iu the United States During 
1937," U. S. Bureau of Labor Statistics, Monthly Labor Review, March 1930, pp. 597-615. 

^ See Max D. Kossoris, "Industrial Injuries and the Business Cycle," U. S. Bureau of 
Labor Statistics, Monthly Labor Review, March 1938, pp. 579-594.' 



CONCENTRATION OF ECONOMIC POWEll 161 

Lay-offs may lag behind operations and production in the early 
stages of a depression. (3) As the depression deepens, management 
shifts to the most efficient, modern, and also the fastest equipment. 
(4) Numerous injuries, Avhich under more normal conditions Avould 
have disabled workers from 1 to 3 days, are not reported during the 
downswing.^^ This is only logical because, when jobs are scarce and 
many employees are working only part time and at reduced wage 
rates, a worker will endeavor not to reveal the occurrence of an 
injury for fear of jeopardizing his employment. 

On the upswing, the frequency rate of inihistrial accidents increases 
markedly. This is probably due to the hiring of workers wlio are 
either unaccustomed to the hazards of their new jobs or whose safety 
habits have been dulled through lengthy lay-offs, and who, perhaps, 
are too eager to make a favorable showing. A study of a large petro- 
leum concern indicates that newly added employees, as a group, suffer 
the majority of injuries. Only 31.8 percent of the total working force 
were workers with less than 3 years' serA-ice but they were responsible 
for -IG.l percent of all disabling accidents.-* Also, when the employ- 
ment situation is easier and hours and wage rates are more normal, 
workers are more inclined to report injuries and take time off', even for 
only 1 or 2 days.^^ 

Therefore if the very existence of the business cycle causes indus- 
trial accidents to rise rapidly with the upswing and to decline during 
the dowiii^wing at a substantially under-reported rate, it is apparent 
that the close correlation between the business cycle and the reported 
frequency rate conceals an increase in the actual rate. Consequently, 
modern technology is not only the direct cause of a substantial num- 
ber of industrial accidents each year; it also, as one of the major 
causes of business cycles, tends to raise the actual frequency rate 
above the level it would have held had no business cycle occurred. 

Occupational Diseases. 

Certain technological processes contain the danger, if not the cer- 
tainty, of occupational disease. In 1933 the United States Bureau 
of Labor Statistics listed approximately 900 hazardous occupations, 
and during the period 1922 to 1933 the munber of poisonous sub- 
stances considered increased from 52 to 94.*^ 

The 10 years which have ehipsed * * * (1022-32) * * * have seen 
wide expansion and a marked increase in the field of industrial liygiene. Tliey 
have been noteworthy for the hirge nnmber of scientific investigations nnder- 
taken to determine * * * tlie effects of exposnre to specific indnstrial haz- 
ards. Complete reports have been pnblished n()t only on the effects of snch 
poisons as radio active paint, methyl bromide, and other refrigerants, and tetra- 
ethyl lead, which have become of importance only recently, but our knowledge 
of well-known health hazards has also been enriched. Tt) mention only a few, 
benzol, spray painting, and exposure to asbestos dust and to dusts containing 
free silica, have been thoroughly studied and reported upon.*' 

The question naturally arises as to what types of exposures, pos- 
sibly resulting in occupational disease, occur in industry and with 
what frequency. 

«' Ihia.. pp. 581-582. 
^^^U)!,!., p. 582. 
" Ibid., p, 504. 

*"* U. S. Bureau of Labor Statistics, Occupation Hazards and Diagnostic Signs, Bulletin 
No. 582. 1933. p. vi. 
" n)id., p. v. 

27T551— 41— No. 22 12 . 



162 



CONCENTRATION OF ECONOMIC POWER 



* * * Bluomfield, Jolinson and Sayers in 1935 surveyed 615 industi'ial 
plants in a typical industrial area, involving a total of 28,686 persons. The 
results of this survey showed that inorganic, nonmetallic dusts accounted for 
the largest number of exposures, 27.4 percent ; carbon monoxide came next with 
19.3 percent of the total number of exposure, while lead compounds accounted 
for 10.2 percent. Exposures to these three substances, therefore, totaled ap- 
proximately 57 percent. The following table shows the details with reference to 
these and other types of exposures: 



Materials 


Number 
of ex- 
posures 


Percent 
of ex- 
posures 


Materials 


Number 
of ex- 
posures 


Percent 
of ex- 
posures 


Inorganic nonmetallic dusts. - 


7.862 


27.4 


Tarbon monoxide 


5. 538 

2.926 

1.544 

1,124 

995 

881 

863 

561 

477 


19.3 




10.2 
5.4 


Emerv dust 


3,678 
2,585 
793 
403 
204 
199 


12.8 
9.0 
2.8 
1.4 

7 


Benzol 


Quartz dust 


Turpentine 


3.9 






3.5 


Other silicates 


Methanol 

A niline compounds. 


3.1 




3.0 




2.0. 




Ammonia 


1.7 















In connection with this study, the authors were careful to point out that "data 
on occupational exposures to these materials and conditions must not be inter- 
preted as signifying that workers were being .subjected to toxic amounts of 
hazardous materials, for no quantitative studies of the workroom environment 
were made. These -data merely indicate the potentialities in the plants studied.** 

The following classification, prepared by Dr. Carroll Daugherty, 
indicates the scope of occupational groupings in which occupational 
diseases are a definite hazard. 

(1) The dusty trades. 

(2) The poisonous trades, other than dusty. 

(3) Occupations producing germ diseases. 
<4) Occupations producing skin infections. 

(5) Occuptations involving extremes in temperature. 

(6) Occupations involving work in compressed or rarefied atmospheres. 

(7) Improper lighting. 

(8) Occupations requiring constant use of certain parts of the body. 
<9) Processes requiring artificial humidity.'* 

Each new industry brings its own occupational diseases which may 
or may not be new jporms of poisoning. For example, in the viscose 
rayon'industry. the most common form of poisoning is that from car- 
bon disulphide. While this type of poisoning occurred in connection 
with some of the early processes of manufacturing rubber products, 
the extent of the poisoning has gained greatly in the new industry 
of viscose rayon.^° 

Of especial pertinence is the prevalence of occupational diseases 
among numerous types of chemical processes. The tendency for 
metallurgical methods to be replaced by chemical processes (already 
examined in ch. I), apparently foreshadows an increase in occupa- 
tional disease. For example, ethylene oxide gas is one of the newer 
chemical products coming into industrial use. It is an intermediate 
in the snythesis of other compounds such as methyl, ethyl, and butyl 



^ C. O. Sappington, Industrial Healtlr, Asset or Liability, Industrial Commentaries, 
Chicago, mao, pp. 8."i-86. 

8' Carroll Dauuherty, Labor Problems in American Industry, Houghton Mifflin, Boston, 
19.^3. pp. 12G-127. 

""U. S. Department of Labor, Division of Labor Standards, Occupational Poisoning In 
tlie Viscose Rayon Indu.stry, Bulletin No. 34, 1940. 



CONCENTRATION OF ECONOiMIC POWEe} 163 

cellosolve, and is used as a fumigant either alone or mixed with car- 
bon dioxide. It has a mild sweetish odor which is not strong enough, 
however, to give warning of harmful concentrations in the air. 
Guinea pigs subjected to it suffered successively nasal irritation, eye 
irritation, bloody discharge from nostrils, unsteadiness on feet and 
staggering, inability to stand, respiratory disturbances and gasping, 
and death.^^ 

Another example is the revolution in painting, occurring a few years 
ago upon the introduction of nitrocellulose lacquers in spray-painting 
( also discussed in ch. I) ."- Such lacquers involve the use of materials 
which are highly inflammable, unstable, liable to ignite spontaneously, 
and are conducive to occupational disease,^^ 

The probable increasing use of chemical processes in tht tuture by 
the industrial world presages an increase in the incidence of occupa- 
tional disease unless stringent protective measures are taken to lessen 
the impact of these effects of modern technology upon the employed 
worker. 

TECHNOLOGY AND THE OLDER WORKER 

It is widely believed that older workers are unable to keep up with 
the pace of modern technology. But the belief that older workers are 
less efficient under modern methods of production apparently lacks 
factual foundation. Productivity of older workers has been found to 
be higher than the average for all age groups generally in skilled 
crafts, and frequently in mass-production industries. For example: 

Automobile Facts reports that in the motor industry, where wages are largely 
on a piece-work basis, and high-speed production is the rule, earnings reach their 
peak in the group between "lO and 55 years of age. The average annual earnings 
of that age group in a 1038 period of full production were $1,080. Men over 60 
showed average annual earnings of $1,595, which was approximately the figure 
for those 40 to 45 years of age. 

A study made in 1938 by the Industrial Relations Section of the ^Massachusetts 
Institute of Technology found in a group of New England plants 'no tendency for 
output and earnings to diminish materially with age except possibly over 60." 
Thus among a group of cotton textile weavers, age had little effect on earnings 
between 25 and ii'^. "They were slightly liigher for men aged 45 to 49 and 50 to 
54 than for either older or younger workers * * *." Similarly, among cotton 
spiimers, '"variations in average earnings by age groups were so small as to be 
unimportant." " 

A survey made by the California Department of Industrial Relations in 1930 
showed that most employers felt that "age had little relation to efficiency, although 
some jobs were more appropriate for older men than others." The statement was 
frequently made by these employers that "workers over 40 or over 50 in good health 
are as efficient as yotmger workers." ^■' 

Laboratory research also shows that productivity doe^ not always 
decline rapidly after the worker reaches 40. 

»i U. S. Public Health Service, Public Health Reports, vol. 45. Part 2. August 8, 1930. pp. 
1832-43. "Acute Response of Guinea Pigs to Vapors of Some New Commercial Organic Com- 
pounds," IV, Ethylene Oxide, by C. P. Waite and others. 

"= See pp. 10(>-U)7. supra. 

13 ••xiie liydro-carbons and alcoliols are diluents and * * * n^p considered the most 
harmful of "the different constituents of tlie lacquers. Until a few years ago benzol was the 
principal diluent, but with an understandin.g of its extreme toxicity it lias largely been 
replaced by others, the principal one in use now being petroleum naphtha. However, all 
lacquer vapors are toxic, or at least narcotic, if inhaled in a sutlieiently concentrated state 
for a long enough period of time." (U. S. Bureau of Labor Statistics. Handbook of Labor 
Statistics. 1936 ed„ Bull. No. 616, p. 338.) 

>" Beulah Amidon, Jobs After Forty. Public Affairs Cmimittee, Inc., New York. 1939, p. 19. 
(This publication was based on materials assembled and prepared by the staff of the U. S. 
Department of Labor for the Committee on Employment Problems of Older Workers,) 

"= Ibid., p. 18. 



154 CONX'ENTRATIOX OF ECONOMIC POWER 

Drs. Henderson, Dill, and MacFarland, of the Harvard Fatigue Laboratory, 
find that the assumption that there is a rapid decline after 40 years of age in the 
qualit_v and quantity of work is "a social myth which, though in some respects 
not misleading, is in general grossly inconsistent with the evidence." Summariz- 
ing ."-everal studies, these scientists state : "First, the rate of decline in the capaci- 
ties of the industrial worker after 45 years of age has been greatly exaggerated. 
Secondly, the evidence relative to the changes in abilities of the older worker must 
be considered in terms of a particular set of circumstances. In some instances, 
the decline is quite large ; in others it is of small magnitude, while there are many 
conditions of work which indicate that the okler man is a distinct asset. Thirdly, 
the problem is so complicated, with so many ramifications in human physiology 
and psychology, that at the present time there is little reason for taking the posi- 
tion as a ground for action that in general men over 4o years of age are less effec- 
tive than others in industrial occupations." "" 

Apparentl.y tlie driving tempo of modern mass-production fails to 
lessen the effectiveness of older workers. Tliey also hold their own 
in trades requiring skilled craftsmanship. For example, a study of 
productivity of skilled cigar-makers found that the mnnber of cigars 
per 40-hour week rolled by workers in the 41-50 year-old wage-earner 
group was 8.090, compared with an average for all age groups of 
3.030; 542 of the 1.909 workers included in the sample were in the 
41-50 age group. °^ 

A study of the efficiency of skilled workers on Works Progress 
Administration projects -'^ found that older workers made an ex- 
tremely favorable showing both in regard to quality and quantity of 
work ]:)erformed. The report observed that — 

There was. as a rule, a direct correlation between grades given for quality 
of work performed and age: the older workers were given the higher grad'^s 
and the younger workers the lower grades. An exception to this rule was 
found among the brick and stone masons, where the workers graded as in- 
ferior were, as a group, older than those graded as excellent or passable. 
For all workei-s combined, however, the average age of workers graded as 
excellent was 47.5 years, that of workers graded as passable, 44 years, and 
that of workers graded as inferior, 40.7 years. 

There was a direct relationship between the grades given for quantity of work 
performed and the ages of the workers. The workers who were given inferior 
grades tended to be the youngest men, tho.se given passable grades, somewhat 
older, and those given excellent grades, oldest. The average age of all workers 
graded as excellent was 46.6 years; the average of all those gi-aded as passable 
was 44.6 years; and the average of all those graded as inferior was 41.6 years."* 

The percenta<.e of workers performing work of various grades of 
quality and quantity is shown for each of ten age groups in table 18. 
This indicates that the strain and hardships of modern industrial life 
do not affect the performance of older workers to any greater degree 
than that of younger workers. If there are obstacles to the employ- 
ment of older workers, their existence appears to be due to factors 
other than technology. 



"" Ibid., pp. 19-20. 

"" r. S. Bunaii of Labor St.Ttistics. Monthly Labor Review, February 1940, "Individual 
rrodiiftivitv Differences." l)v W. D. Evans. 

"^ Works Progress Administration. The Skill of Brick and Stone Masons. Carpenters, and 
Piiinttrs Employed on Works Progress Administration Pro.iects in Seven Cities in January 
VXil. l>v W. R. Curtis, W. G. Keim, and Edward Herman, 1937. 

"■■ Ibid., p. 6. 



CONCENTRATION OF ECONOMIC POWER 



165 



Table 18. — Peccentage of icorkers in each age group receiving various grades 
given for quality and quantity of work performed: Brick and stone masons, 
carpenters, and painters employed on Works Progress Administration projects, 
January 1937 

7 CITIES 



Age class 



iQualityj Quantity IQualityj Quantity Quality Quantity Quality Quantity 



24 and under ! JOO.O i 

25-29 

30-34 

35-39 

40-44 

45-49 

50-54 



60-64 

65 and over. 



JOO.O j 


100.0 
100.0 


8.3 

18.8 




29.2 
45.6 


58.8 
53.3 1 


62.5 1 
35.6 




100.0 


15.6 


31.1 


100.0 


100.0 


27.6 


23.5 


44.3 


50.3 j 


28.1 


26.2 


100.0 


100.0 


29.4 


24,4 


41.3 


47.6 1 


29.3 


28.0 


100.0 ] 


100.0 


41.7 


29.8 


34.1 


48.1 


24.2 


22.1 


100.0 . 


100.0 


42.2 


35.0 


37.4 


47.5 


20.4 i 


17.5 


100.0 1 


100.0 


41.4 


26.0 


43.8 


55.4 


14.8 1 


18.6 


100.0 ! 


100.0 


50.6 


33.6 


35.4 


52.6 


14.0 


13.8 


100.0 ; 


100.0 


51.5 


33.3 


38 4 


53.1 


10.1 


1.3.6 


100.0 1 


100.0 


63.9 


48.2 


22.2 


40.7 


13.9 


11.1 


100.0 


100.0 


38.4 


28.3 


39.1 


50 4 


^"1 


21.3 



Source: Works Progress Administration, The Sijill of Brick and Stcne Masons, Carpenters, and Painters 
Employed on Works Progress Administration Projects in Seven Cities in January 1937, by W. R. Curtis, 
W. G. Keim, and Edward Berman. 1937, p. 60. 



CHAPTER III 

TECHNOLOGY AND THE COMPENSATORY FORCES 

There are certain forces presumably inherent in the present eco- 
nomic order which operate more or less automatically to offset the 
labor-displacing effects of technology. Principal among these com- 
pensatory forces are (1) the reduction of hours (without an accom- 
panying decline in wages), (2) the development of new industries, 
and (3) the reduction of prices. To ascertain the state of economic 
balance between the labor-displacing effects of technology and these 
compensatory forces it is necessary to examine each in some detail. 
Such is the purpose of this chapter. 

THE BEDUCnON IN HOURS 

Trade-union organizations have continuously urged the adoption 
of a shorter workweek to offset the labor-saving effected by modern 
technology.^ As Dr. Leo Wolman writes : 

The American Federation of Labor, regarding the increased productivity of 
industry as a cause of the displacement of labor and hence of an increasing 
rate of unemployment, saw in the reduction of the workweek the most effective 
means of returning the unemployed to employment. At the same time the view 
became more generally accepted that the expanding output of "mass-produc- 
tion" industries could be absorbed only by members of a working population 
who enjoyed more leisure and higher rates of pay. 

and further — 

Much of the advocacy of the shorter week is due to the belief that fewer i^er 
capita hours cause increased employment.^ 

Hours of work have declined extensively during the last century. 
In 1851 a union of newspaper compositors in New York City recom- 
mended a workweek of six l2-hour days or 72 hours. By 1938 com- 
positors worked 37i/2 hours a week. Blast furnace employees worked 
a full-time week of 84 hours as late as 1900; today their hours are 
down to 40. From 1890 to 193T the average workweek of factory 
employees in the United States fell from about 60 to 42 hours, in the 
building trades from 55 to 39, in steam railroads from 60 to 48, in 
anthracite and bituminous coal mining from 60 to 35. 

This decline in hours has been punctuated by two precipitous dips 
followed by periods of relative stability. The first abrupt down- 
turn took place during the AVorld AVar when the average full time 
workweek fell from 55.1 to 51.0 hours.^ This was due largelv to 



1 The mere reduction of hours does not necessarily mean an increase in pay rolls. Pur- 
chasing power may, however, he Increased through the overtime pay which may result from 
reductions in hours. Also, even though the reduction of hours may mean only a greater 
spreading of a given pay roll over a larger number of employees, the "economy may be stimu- 
lated as a result of greater expenditures on necessities at the expense of semiluxurlos and 
savings. 

2 Leo Wolman, "Hours of Work in American Industry," National Bureau of Economic 
Research Bulletin 71, Novemher 27, 1938, pp. 1, 3, 18. 

Ubid., p. 1. ff , 

167 



IQg CONCENTRATION OF ECONOMIC POWER 

shortages in labor and the resiiUant competition for workers among 
industries. The number of hours worked per week then remained 
reLatively stable throughout the twenties. 

The second abrupt dip occurred under the National Industrial 
Recovery Act when the 40-hour week was commonly established by 
the codes of fair competition, while in certain industries, principally 
clothing and coal mining, hours were reduced to 35 or 36 per week. 
For the majority of workers the 40-hour week meant a reduction of 
8 to 10 hours a week. For example, in the cotton goods industry 
only 1.2 percent of the employees were working 40 houi^ or less in 
1929, wh.ile 63 percent were working more than 54 hours. The effect 
of the act was similar in mining where the 9 and 10-hour days were 
virtually eliminated in 1933. Since April 1, 1934, about 97 percent 
of the miners have worked a 35 hour week of five T-hour days.* 
Only in the clothing, printing and publishing, shipbuilding, motor 
vehicles, and rubber tire industries was there a substantial number 
of emplo^^ees who were already on weekly schedules of less than 45 
hours by '1933. 

In most cases the decreases brought about by the N. I. R. A. have 
remained in effect due in large part to the increasing organization 
of labor and the enactment of labor legislation which has been held 
constitutional by the United States Supreme Court. To determine 
whether or not any further substantial reduction of hours will occur 
in the future, there must be evaluation of hours provisions in existent 
legislation and collective agreements, and of the trend in hours 
actually worked. 

The Fair Labor Standards Act, which became effective October 24, 
1938, limited the hours of persons employed in interstate commerce 
or in the production of goods for interstate commerce (with specified 
exceptions) to 44 per week during the year beginning October 1938, 
42 the second year, and 40 hours thereafter. All time in excess of 
these hours is considered overtime and must be paid for at one and 
one-half times the regular rate of pay. The law exempts persons 
workhig under collective agreements, certified as bona fide b}' the 
National Labor Relations Board, with the provision that such agree- 
ments either limit hours to 1,000 in a period of 26 weeks or to not 
more than 2,000 during any period of 52 consecutive weeks. The law 
also makes certain exemptions for seasonal industries and specific 
occupations such as agriculture, merchant marine, and other employ- 
ments in which Avorking conditions are specially regulated by Federal 
legislation. 

Federal legislation also limits all workers employed in establish- 
ments Avorking on Government contracts exceeding $10,000 in value 
to an 8-hour day and a 40-hour week, and workers employed on 
projects under the Work Proj'Gcts Administration and the Tennessee 
Valley Authority to 8 hours per day, 40 per week, and 140 per month. 

A large majority of existing collective agreements between labor 
and management provide for a 40-hour maximum workweek. With 
some exce])tions the 40-hour week prevails in the iron and steel, stone, 
rubber, lumber, petroleum, and aluminum industries; in the manu- 
facture of metals, except stoves; in the furniture and upholstery, 
pulp and paper })roducts, jewelry, pottei'y, light and power, gas and 

* Riitl.. pp. 9-10. 



CONCENTRATION OF ECONOMIC POWER IgQ 

coke, food products, except flour and cereals, industries; and in the 
printing trades and certain sections of the building trades. A small 
number of collective agreements provide for a shorter workweek 
but the difference between the workweek established and the 40-hour 
level is relatively small. In coal mining, glass, fur, men's and 
women's clothing, and newspaper publishing, and in sections of the 
building, radio, hat, and rubber industries, weekly hours vary be- 
tween 35 and 37i/^ per week. 

There exist, furlhermore, certain nidustries, small in number how- 
ever, in wdiich agreements provide for a workweek longer than 40 
hours. The 42-hour week is usual in the glass industry and a 44- 
hour week prevails in some of the building trades agreements. A 
48-hour week prevails in collective agreements covering railroad 
yards, flour and cereal products, the stove industry, retail trade, hotels 
and restaurants, and a few other industries."' 

Viewed as a whole, both Federal legislation and existing collective 
agreements are operating to stabilize the workweek at the 40-hour 
level. Furthermore, it is not to be expected that labor organizations 
will press for further hourly reductions as strenuously as they did 
when hours of work were so onerously long. It is extremely doubt- 
ful whether public sympathy could be aroused so readily in favor of 
reducing, the 40-hour week to 30 or 25 hours, as in 1923 when it 
provoked conferences between President Harding and leaders of the 
steel industry^ leading to the abandonment of the 72-hour workweek. 

In regard to the trend of hours worked, the National Industrial 
Conference Board reports average actual hours per week per wage 
earner for 25 manufacturing industries from 1920 to 1939.'" The 
establishments reporting data in the 25 selected industries cover most 
of the important branches of industry and include approximately 
1,500,000 employees. There are, however, certain defects in the 
samj^le: (1) The average hours Avorked may be aiiected by the com- 
paratively regular employment afforded by the reporting Arms;" 
(2) the sample is heavily weighted by large establishments; (3) it 
underrepresents industry in tlie South; and (4) plants working 
excessively long hours may be expected to understate tliat fact to 
any statistical agency, public or private.® Because of these limita- 
tions the conference Board's data may not always reflect exactly 
the level of hours at a given time, but the series at least provide a 
reasonably accurate indication of the long-term trend in hours 
worked. Average actual hours per week per wage earner from 1920- 
39 in 20 manufacturing industries are shown in chart XIII and 
table 19, as reported by the Conference Board. ^ 



= International Labour OfBce, International Labour Review, August 1939, "Hours of Work 
in the United States," pp. 234-235. 

® The average actual workweek is generally derived by dividing the total number of hours 
worked during the week by the total number of wage earners. Variations in the average 
computed in tliis way may be caused by short time, by overtime, or by shifts in the workers 
represented either because different departments are working on different time schedules, 
or because some workers are working only part of the week, or because they came to or left 
the plant during the week. In a normal period, actual hours are likely to be shorter than 
those flxed by collective agreements, regulations, or the nominal schedules of individual 
plants. In certain circumstances, however, a large amount of overtime may bring actual 
hours to or even above the nominal figures established in agreements, laws, etc. 

■^ See International Labour Office, International Labour Review, August 1939, p. 247. 

8 Leo Wolman, Op. cit., p. 7. 

» These 20 industries include all of the 25 fields surveyed by the National Industrial Con- 
fert?nce Board except 5 divisions of the foundry and machine shop industry. The overall 
figures of the industry are presented, but the divisions — foundries, machines and machine 
tools, heavy equipment, hardware and small parts, and other foundry and machine-shop 
products — were not considered of such importance that one-fifth of all industries included in 
the survey should he composed of segments of this one field. 



170 



CONCENTRATION OF ECONOMIC POWER 



Chart XIII 



AVERAGE ACTUAL HOURS PER WEEK 
PER WAGE EARNER 

IN SELECTED MANUFACTURING INDUSTRIES 
UNITED STATES. 1920-1939 



AVERAGE ACTUA. 
M' )RS PER WEEK 
nn WAGE EARNER 




AVERAGE ACTUAL 
HOURS PER WEEK 
PER WAGE EARNER 



I I 


5SSij5>=^ 


|^-i% 


1 


Bool ond Sho. 

1-tolh.r Tornin, o~) FinltMr, 



I920V22!/ 2* '26 28 30 K M M '56 ■» 



l920!/'22«/ Z* '26 28 » '32 "54 



AVERAGE ACTUAL 
«0ORS PER WEEK 
f^R WAGE EARNER 



AVERAGE ACTUAL 
HOURS PER WEEK 
PER WAGE EARNER 





j 












, 






















■^^\^\' 




iw 






/ 


? 


?i 




M 


*■ 
















^ 






















1^ 


'/ 




_.. 
































-^_^^^-^,^.N.„..p.,.on.M^..lo.. 






221/ 


\ 




■z 


e 




8 







•3 


2 


■3 


4 


■3 


6 


•38 


39 







_l 


i? 


sys*" 


% J 






^^ 








?o=F 


— Hotl.» ond Kr,il Good! 
~ Sm'^GooS"" 



19201/221/ ■2« 26 '28 



AVERAGE ACTUAL 
HOURS PER WEEK 
PER WAGE EARNER 





































/£U-'i" 


■:-i 1 1 


V 












& 


\ 


? 
















\ 


; 
















s 


■^ 






















\ 


I 


/ 




^ 






















1 










= 






EKclrlcol Monulocturinj 1 

---— »i.""Splkl"g 1 




^h.[ -Ll.l LJ^i iO-^ 





/ 




> 


■ 


s 








^ 




















N 














s 


S 


s 


= 


>■ 




\ 


< 


\^ 
































































Clxmlcoli 

Polnl ond Vornlih 


1 1 -1 . 1 1 1 1 i- 



19201/228/24 '26 '28 '30 ' '32 '34 '36 38 T9 



'26 '28 '30 '32 



Source : Table 19. 

Just as the trend of hours worked tended to become rehitively 
stable during the twenties, foilowin<y the wartime decline, so also do 
they appear to be flattenini^ out since the severe depression of 
1932-313. 

In o;eneral, average actual hours remained relatively stable after 
the 1932-33 decline. In certain industries, such as furniture manu- 
facturing, they turned slightly upward, while in others, such as the 
boot and shoe industry, they gradually declined. In none of the 
industries, however, did the post-N. I. R. A. increase take the level 



CONCENTRATION OF ECONOMIC POWER 



171 



of liours anywhere near the 1929 position. Hours tended to increase 
slightly up to 1936-37, then turned downward during 1938. From 
1938 to 1939 the trend was almost universally upw^ard — only in meat 
packing did a decline occur. Since the N. I. R. A. decline average 
hours have paralelled closely their behavior for almost a decade after 
the World War decline. Although their stability since the N. I. R. A. 
drop is not so great as during 1920-29, the relatively minor ups and 
downs appear to offset each other. 

If the' parallel does prove real, the greater part of a decade will 
elapse before any further precipitous downswing in hours can be 
expected. Certainly, the recent upward tendency would indicate that 
no immediate downturn sufficient to offset the increasing productivity 
of labor can be expected. 

Table 19. — Average actual hows per week per xcage-earner in selected manufac- 
turing industries, 1920-39 



Iron 
and 
steel 


Agricul- 
tural 
iinple- 
ments 


Automo- 
bile 


Foun- 
dries and 
machine 

shops 


Boot 
and 
shoe 


Leather 
tanning 
and fin- 
ishing 


Rubber 


Lumber 
and 
mill- 
work 


Paper 
and 
pulp 


€3.2 


48.7 


46.1 


49.1 


44.1 


46.2 


41.9 


48.5 


51.3 


54.1 


41.3 


44.7 


42.8 


45.7 


46.6 


43.8 


47.7 


49.2 


58.1 


48.5 


48.2 


49.3 


46.3 


47.7 


45.3 


49.5 


52.7 


57.7 


49.5 


47.7 


49.6 


45.6 


47.6 


44.8 


49.5 


51.8 


51.3 


48.2 


45.3 


47.5 


44.1 


46.2 


44.1 


48.9 


49.9 


53.6 


49.9 


47.3 


48.6 


45.8 


47.5 


44.5 


49.1 


51.3 


54.4 


49.6 


47.7 


49.1 


44.6 


46.5 


45.0 


18.0 


52.0 


53.1 


49.2 


46.4 


47.9 


45.2 


45.8 


45.4 


47.7 


51.5 


54.0 


49.9 


47.7 


48.2 


44.3 


45.2 


46.1 


47.7 


50.9 


54.9 


49.6 


46.8 


49.4 


44.2 


47.6 


44.8 


45.4 


52.1 


48.9 


43.5 


39.9 


42.8 


40.4 


44.9 


41.4 


44.0 


49.3 


43.4 


35.5 


36.9 


35.9 


44.5 


45.2 


38.1 


40.1 


44.1 


27.2 


32.9 


30.4 


30.1 


41.1 


40.9 


33.1 


36.4 


40.6 


34.0 


35.4 


36.0 


33.1 


39.6 


41.8 




35.4 


41.1 


29.3 


38.5 


33.5 


35.1 


37.3 


36.6 


32.3 


35.0 


37.5 


34.2 


39.7 


38.0 


37.8 


37.1 


38.1 


33.1 


39.3 


39.6 


39.8 


39.7 


42.3 


41.4 


36.8 




36.6 


40.7 


42.6 


36.6 


40.0 


39.6 


41.4 


38.3 


38.2 


33.3 


40.1 


42.1 


27.6 


35.1 


32.3 


34.3 


32.8 


35.6 


30.3 


37.8 


38.5 


34.6 


38.0 


34.9 


38. 6 


36.0 


38.6 


35.5 


39.6 


40.7 



Paper 
prod- 
ucts 



1921.. 
1922 2 
1923.. 
W24.. 
1925 . 
1926.. 
1927.. 
1928- - 
1929.. 
1930.. 
1931.. 
1932.. 
1933.. 
1934.. 
1935.. 
1936.. 
1937.. 



45.1 
40.4 
46.6 
47.6 
47.5 
47.4 
47.1 
47.4 
47.5 
49.5 
47.4 
45.3 
41.1 
40.5 
36.4 
38.2 
41.0 
41.0 



Year 


Print- 
ing- 
news- 
papcrs 
and mag- 
azines 


Hosiery 
and 
knit 
goods 


Cotton 
goods 


Silk 
goods 


Wool 
goods 


Electri- 
cal man- 
ufactur- 
ing 


furni- 
ture 


Meat 
packing 


Chemi- 
cals 


Paint 

and 

varnish 


1920 1..... 

1921 

1922 2..... 

1923 

1924 

1925. 

1926. 

1927 

1928 

1929 

1930 

1931 

1932. 

1933 

1934. 

1935 

1936. 

1937 

1938 

1939 


44.6 
44.4 
44.9 
45.0 
44.9 
45.2 
45.3 
45.7 
4.5.1 
45.7 
45.2 
43.6 
42.1 
39.3 
36.5 
36.2 
37.2 
37.9 
36.6 
37.0 


45.1 
46.8 
46.8 
46.3 
43.7 
45.1 
45.1 
47.0 
46.8 
47.6 
43.3 
42.2 
38.5 
39.2 
34.6 
34.5 
35.8 
36.6 
34.0 
36.6 


43.9 
42.3 
46.6 
47.8 
42.7 
47.0 
46.1 
47.2 
46.9 
48.2 
44.6 
45.2 
42.5 
41.8 
35.4 
30.4 
38.4 
37.9 
35.7 
37.9 


: 41.9 
44.9 
45.7 
46.5 
45.4 
46.9 
45.2 
45.9 
47.3 
47.8 
44.9 
44.3 
38.9 
37.5 
31.2 
32.1 
34.2 
35.3 
32.3 
35.2 


41.9 
45.9 
47.6 
47.5 
43.4 
44.1 
45.2 
45.1 
44.4 
46.4 
41.6 
41.8 
39.3 
39.5 

36^ 7 
36.1 
34.7 
32.4 
35.8 


47.1 
43.2 
47.4 
47.8 
45.6 
46.5 
46.2 
45.1 
46.1 
47.4 
42.8 
35.6 
29.4 
33.2 
33.9 
36.2 
39.2 
38.8 
33.8 


46.8 
44.4 
48.4 
48.2 
47.1 
47.5 
48.0 
47.7 
46.8 
46.9 
42.4 
.39.8 
33.6 
35.7 
34.8 
37.8 
41.8 
40.4 
35.3 
38.4 


48.5 
46.9 
49.2 
49.7' 
49.5 
49.6 
49.8 
50.2 
50.6 
50.6 
50.0 
49.0 
48. 2 
44.8 
40.9 
40.6 
41.9 
39.8 
40.5 
40. 1 


49.2 
50.0 
53.2 
53.2 
52.2 
52.8 
50.2 
50.0 
50.1 
50.4 
47.5 
44.5 
40.7 
39.1 
38.5 
39.3 
39.9 
39.8 

39; 3 


49.0 
47.3 
49.8 
49.8 
52.9 
52.7 
50.8 
50.5 
50.8 
51.8 
47.9 
44.4 
41.4 
40.3 
38.8 
39.9 
45.3 
41.2 
39.0 
40.7 



1 Average for last 7 months. ^ Average for last 6 months. 

Source: National Industrial Conference Board, Wages, Hours, and Employment in the United States, 
1914-36, tables 5-30, pp. 56-159; Conference Board Service Letter, Supplement to June 1938, "Wages, Hours, 
and Employment in the United States, July 1936-December 1937," tables 4-6, pp. 5-18; Conference Board 
Economic Record, Mar. 28, 1940, "Wages, Hours, and Employment in the United States, 1934-39," table 4, 
pp. 120-134. 



172 CONCENTRATION OF ECONOMIC POWER 

THE DEVELOPMENT OF NEW INDUSTRIES 

THE ECONOMIC IMPORTAXCE OF NEW INDUSTRIES 

Technology, by bringing forth new industries, causes 'the econ- 
omy to expand along three major fronts: (1) Employment oppor- 
tunities are created for large segments of the working population in 
the fabrication of the ne^\ product ; (2) the capital goods industries 
are stimulated by the placement of orders for needed productive 
equipment; (3) new industries frequently create activities in the 
fields of distribution, transportation, service, and maintenance. 

Dr. Alvin Hansen of Harvard University well summarized tlio 
importance of new industries in his testimony before the Temporary 
National Economic Committee: 

Thus, throughout the nineteenth century the emergence, development, and 
growth of giant new industries has phiyed an enormous role, not only in the 
cycle movement itself, but also in the vigor and intensity of America's booms, 
and the depth, severity, and length of depression periods. It is my view that 
the deep and prolonged depression of the nin(*ties relates to the cessation of 
growth of the railroad industry. There was a temporary lull before the electrical 
and automobile industries emerged, and similarly in the decade of the thirties, 
in which we are now living, we are having a similar exi>erience. The great 
automobile industry has risen to maturity, and no comparable new industry 
has appeared to fill the gap. It is my growing conviction that the combined 
effect of the declining population growth, together with the failure of any 
really important innovations of a magnitude sutticient to absorb large capital 
outlays, weigh very heavily as an explanation of the failure of the recent 
recovery to reach full employment.'" 

The importance of new industries is also emj^hasized by the data 
in table 20. This shows that 18 new manufacturing industries, which 
came into existence since 1879, absorbed almost one-seventh of all the 
labor employed in manufacturing in 1929. The number employed by 
these industries does not include either those engaged in the produc- 
tion of the necessary capital goods or those required in the fields of 
distribution, transportation, and service. 

It is Avhile new industries are rapidly expanding that they are most 
effective as stimulants to economic activity. New plants are con- 
structed, new workers are hired, and a ncAv demand is created for 
materials. 

Table 20. — Wage earners iti 18 newmanufacfurlng indti.'^tries sive JS79 

Average num- 
ber of ivafje 
Industry : earners. 1929 

Electrical machinery, apparatus, and supplies 328. 722 

Motor vehicles, not including motorcycles 22G. 110 

Motor vehicle Ix.dies and parts 221.332 

Rubber tires and inner tuiies 83,263 

Manufacture of gasoline' 39,411 

Rayon and allied products 39,106 

Manufactured ice 32. 184 

Aluminum manufactures 21, 210 

Typewriters and parts 16,945 

Refrigerators, mechanical 16, 883 

1 For tho reason that gasoline is chiefl.v used as a source of power in another now in- 
vention — tho intei'ual combustion motor", around which has been built up one of our 
greatest industries — the number of employees engaged in the manufacture of gasoline 
has been estimated for this list. 

w Hearings before the Temporary National Economic Committee, Part 9. Savings and 
Investment, p. 3514. 



CONCENTRATION OF ECONOMIC POWER 173 

Aiciaf/e ninn- 
bcr of wane 
Industry— Continued. eanwis, I929 

Cash registers and adding and computing machines 16, 840 

Oil. calie and meal, cottonseed 15.825 

Aircraft and parts 14,710 

Phonographs 14, 410 

Photographic apparatus and materials 12,967 

Motion picture apparatus except for projection in theaters 10, 784 

Asbestos products - 8, 092 

Fountain pens ' 4, 508 

Total 18 new industries 1,123,314 

Total, all manufacturing industries 8,838,743 

- Excluding steam packing and pipe and boiler covering. 
3 Estimated. 

Source: Machinery, Employment and Purcba-sing Power, National Industrial Conference 
Board, Inc., New York, 1935. p. 61. 

The importance of a new industry diminishes, however, as it develops 
sufficient capacity to meet a foreseeable demand, and as the capital 
goods industries supplying the necessary productive equipment expand 
to a })oint wliere they can meet any expected demand from the new 
industries. In commenting upon the importance of the growing stage 
of neAv industries, Dr. Hansen pointed out : 

In the decade of the twenties the great automobile industry gave a tremendous 
upward push to our entire economy, running through a vast range of related 
industries. While thi;^ industry stfjl qccupies an extremely important place in 
our economy, there is this highly significant difference, that it is no longer 
growing and when a revolutionary new industry like the railroad in the last 
century, or the automobile in our time, after having initiated in its youth a 
powerful upward surge of plant expansion in all the basic industries which serve 
its needs, after such an industry reaches maturity and ceases to grow, as all 
industries finally must, the whole economy must exi>erience a stagnation, unless 
indeed new developments equally far-reaching take its place. 

It is not enough that a mature industry continues its activity at a fcV.'- level 
on a horizontal plane ; the fact that new railroad mileage continued to oe built 
at a high rate throughout the seventies, eighties, and nineties was not sufficient. 
It is the cessation of growth which is disastrous, for when they have ceased to 
grow there is no further need for plant expansion and when giant new industries 
have spent iheir force, it may take a long time before something else of equal 
magnitude emerges." 

Certain new industries tend to ])ass quickly through this vital grow- 
ing stage, such as tho.se whicli produce durable goods and are affected by 
a rapid increase in market saturation. A comparison of specific prod- 
ucts in the field of consumers durable goods reveals the effect of market 
saturation upon sales and upon the change in tlie stimulus given to the 
economy by a new industry. Such an analysis has been made by the 
United States Bureau of Labor Statistics in a report to the Temporary 
National Economic Committee.'- Of particular interest here is a com- 
parison of tlie trend of market saturation to sales in two consumers dur- 
able goods — electric refrigerators and vacuum cleaners. 

For these products market saturation may be related to wired homes, 
showing the percentage of the total number of wired homes in the coun- 
try which pos.sess the commodity under consideration. In 1929 only 9.4 
percent of wired homes had electric refrigerators. By 1932 market sat- 
uration had risen to 21.6 percent. No decline whatever in the number 
of electric refrigerators sold occurred in the period 1929-32. As a mat- 
ter of fact, sales in 1932 were 2.6 percent above sales in 1929. By 1937, 

" Ibid., pp. 3513-3514. 

^2 Temporary National Economic Committee Monograph No. 1, Price Behavior and 
Business Policy, 1940, Pajrt I, ch. IV, pp. 117-131. 



^74 CONCENTRATION OF ECONOMIC POWER 

however, 51.7 pe^ent of wired homes had electric refrigerators, and 
sales in 1937-38 dropped 46.3 percent. 

Sales and market saturation in the vacuum cleaner industry are a 
variation of the same phenomenon. There occurred no marked increase 
in saturation in vacuum cleaners from 1929 to 1937, the figures being 
respectively 43.6 and 49.0 percent. But, unlike the electric refrige^'- 
ator industry which suffered no decrease of sales between 1929 and 
1932, the vacuum cleaner industry, with its high saturation in 1929, 
experienced a 60.1 percent drop in sales. Again in the downswing 
of 1937, by which time saturation had risen 5.4 points above the 1*929 
level, vacuum cleaner sales dropped noticeably, falling 24 percent. 

In both industries market saturation was at a high level in the "reces- 
sion" and in both sales dropped abruptly during that downswing. Dur- 
ing 1929-32, however, the electric refrigerator industry was character- 
ized by a relatively low degree of saturation and suffered no declines in 
sales at all, Avhile the vacuum cleaner industry with a high degree of 
market saturation suffered an extremely severe decline in sales. Thus 
the growth period of the electric refrigerator industry was less than 10 
years, and probably closer to 5, after which time its vitalizing effects 
on the economy were probably of decreasing importance. 

The development of new industries was to a considerable extent 
responsible for the prosperity of the twenties, which rested, according 
to Dr. Hansen, upon five factors : 

First, * * * there was residential building, which reached in this decade an 
all-time high ; 8,750,000 urban residential units, not including farm units, were 
built in this decade. * * * 

In the second place there was a high volume of public construction, financed 
heavily b.v State and local borrowing. State and local debt increased in this period 
at the' rate of $1,000,000,000 a year. Large capital outlays were made on roads, 
schools, and other public improvements. 

In the third place, there was the outlet for savings in the foreign loans and 
investments which in part provided foreign countries with a "purchasing power to 
buy from us an excess of exports over imports, amounting to $10,600,000,000 in the 
decade from 1920 to 1929. 

Fourth, there was the growing importance of consumer-installment credit, which 
financed the purchase of durable consumers' goods and which reached tJie quite 
extraordinary level of 11 billion by 1929. 

* * * Fifth, there was the prodigious growth of the automobile industry, 
together with all the related industi'ies which it fostered and sustained * * *." 

Along with the automobile industry should be included such new 
fields as the natural gas industry, the rayon industry, the motion picture 
industry, and certam segments of the electrical equipment industry. 
But by far the most important was the automobile industry. 

The reduction of automobile prices during the early part of the 
twenties made it possible for members of the middle-income groups to 
own automobiles and resulted in a tremendous expansion of economic 
activity in many major fields. The manufacture of automobiles on a 
mass-production basis created widespread markets for the iron and 
steel, gasoline, plate-glass, leather, rubber, lead, copper, aluminum, 
glycerine, and cotton industries. It also stimulated greatly the nia- 
chinery industry during the period when new automobile factories 
were being erected and new technological processes were being con- 
stantly introduced. When, however, the automobile industry became 
equipped with machinery sufficient to produce far more than any rea- 

^ Hearings before the Temporary National Economic Committee, Part 9, pp. 3512- 
3513. 



CONCENTRATION OF ECONOMIC POWER 175 

sonably expected demand for automobiles, its importance as a market 
for capital goods began to diminish. 

The automobile industry in its rapid development gave direct im- 
petus to other segments of our economy. The construction industry 
was stimulated as new garages and filling stations were erected. The 
great expansion in residential construction during the twenties was 
due in part to the automobile, for it enabled urban workers to commute 
from their homes in outlying areas. In addition, the automobile in- 
dustry greatly stimulated the cement and asphalt industries because of 
the great expansion in road building. Finall}', retail and distributing 
outlets were established to sell automobiles; supplies, equipment and 
finished cars were transported in increasing amounts; and the servicing 
and maintenance of automobiles became in itself an industry of con- 
siderable size. 

Interestingly enough, technology in recent 3'ears has tended to re- 
duce the effect of certain of the five bases of prosperity in the twenties 
mentioned by Dr. Hansen. For example, the expansion of housing 
during this period rested to no small degree upon the growth of popu- 
lation. Throughout our history the increase in the number of dwell- 
ings has maintained a remarkably close relation to the increase of 
population. From 1916 to 1922, the population of the highly indus- 
trialized and urbanized section of the United States (Massachusetts, 
Connecticut, Rhode Island, New York, New Jersey, Pennsylvania, 
Ohio, Michigan, and Illinois) increased 10 percent. From 1922 to 1928, 
there was a further increase of 10 percent in this area, but from 1930 to 
1936 the population of this area increased only 3 percent. For the 
population of the entire country, the percentage increase was 9.1 per- 
cent from 1922 to 1928 compared with 4.3 percent from 1930 to 1936. 
Thus the percentage rate of increase in the urbanized population is less 
than one-third of that of the twenties, and for the entire country below 
one-half. To a considerable extent this decline in population can be 
attributed to a technological development: The contraceptive. 

There is one invention that has an influence on all our businesses. It is not 
much discussed ; rather it is one of those hidden forces, inconspicuous, not im- 
posing, yet having profound influences. Methods of birth control are slowing up 
the growth of our population. As a market today the population is not expanding 
as it has done all through American history. In fact in about 25 years it will 
cease expanding altogether and begin to decline — unless something is done 
about it." 

The export of the techniques of production to nonindustrialized 
lands has played an important part in reducing the role of foreign 
trade and the outlet for savings in foreign loans and investments. 
The United States has sold great quantities of machinery and other 
capital goods to many nations, including Italy and Japan. Those 
nations have used that machinery to supply not only their domestic 
needs but also to manufacture goods in competition with older 
industrialized countries for the possession of world markets. These 

" William F. Ogburn, Machines and Tomorrow's World, Public Affairs Committee, Inc., 
New York. 19.38. p. 9 (pamphlet). 

"Thus the influence of contraceptives is, first, to reduce the birth-rate ; which, second, 
lessens the proportion of children in the population ; which, third, increases the propor- 
tion of old people ; which, fourth, leads to a large number of old couples without children 
who in former times were an important insurance for elders ; and which, fifth, promotes 
old age insurance by the State, a derivative effect not readilv foreseen." ("The Influence 
of Inventions on American Social In.stitutions in the Future," by William F Ogburn The 
American Journal of Sociology, November 1937, p. 369.) 



I'JQ CONCENTRATION Of ECONOMIC POWER 

newly industrialized nations have acquired more and more machin- 
ery and related capital floods, either by importing them or con- 
structing them with techniques already imported, and in recent years 
have been exporting capital goods to other nations not highly indus- 
trialized at the present time.'^ 

With technological developments affecting both population — and 
indirectly construction — and foreign trade, and with the automobile 
industry beyond its growth period, it appears that the five stimuli 
to prosperity in the twenties, discussed by Dr. Hansen, must be re- 
garded as an aggregate stimulus peculiar to that decade. 

LIMITATIONS OF THE NF,W INDUSTRY STIMULUS 

The possibilities of economic expansion by the development of new 
industries are limited because (1) the substitution of new products 
for old may, through lower unit labor requirements, decrease net 
employment, (2) new technological developments may lessen the 
requirements and outlays for new capital goods, and (3) present 
income distribution limits the market for new products. New indus- 
tries can offset the effects of labor-saving technology only if these 
three limitations are overcome. 

The Principle of /Substitution. 

The mere development of a new product, or perhaps a new indus- 
try, does not necessarily result in a net gain in employm.ent and pur- 
chasing power equivalent to the consumption of the new product, 
The new good may be merely a substitute for an older commodity. 
The extent to which a new, but substitute product, will bring about 
a net gain in employment is determined not only by the increase in 
its consumption over the old product, but also by the ratio of its unit 
labor requirements to those of the older commodity. "While more 
units of the new product may be sold than of the old, this gain might 
be offset by lower unit labor requirements in the production of the 
new commodity. For example, if the anmial consumption of the new 
product amounted to 1,000,000 units compared with one-half a 
million of the old, this gain would be offset if the unit labor require- 
ments of the new product were 50 or more percent lower than those 
of the old. 

The widespread occurrence of substitution has already been dis- 
cussed in the section on types of labor-saving techniques.^" The ex- 
amples of recent replacements cited there should leave no doubt that 
the trend in the development of new products is toward the cre- 
ation of those which involve lower unit labor requirements than the 
items they replace. Present tendencies indicate that an increasing 
proportion of new commodities will be produced by chemical or 
electrical processes. If these new commodities are substitutes, they 
will have to be consumed on an extremely wide scale to offset the 

" See John M. Blair, Seeds of Destruction, Covici-Friede, New Yorl<, 19;?S, ch. 13, 
"Total Export.s and Machinery Exports." 

To cite one example, processes for direct rollinff of nonferrous metals are among the 
most rcH'ently developed techniques of production. Shortly prior to 1937 "n 20-inch 
mill for direct rolling of hrass and copper was shipped from this country to Japan. It 
has a capacity of 1 ton in 4 minutes." (See National Resources Committee, Techno- 
logical Trends and National Policy, 1938, "Metallurgy," by C. C. Furnas, p. 356.) 

18 Se" pp. 106-110, supra. 



OOXCENTKATIOX OF ECONOMIC POWE^ 177 

almost laborless cliaracter of such processes and thus to prevent a net 
decline in employment. 

Many writers minimize the possible destructive effects of new 
industries. Perhaps an insight into this tendency may be gained by 
an examination of a recent work by Carl Snyder, Capitalism the 
Creator, wherein he has a chapter entitled "The Unlimited Poten- 
tials of Tectonics" in which he cites several examples of the ways 
new industries become established or may yet deivelop without harm- 
ing the commodities they are replacing. He has nothing but scorn 
for the idea "that new industries are coming in to ruin old industries 
and create unemployment." ^^ 

One of Dr. Snyder's examples follows : 

Another wouderful rejuveiiutiou of an industry: In the hist 10 years of 
stagnation we have been hearing much about the diflBculties of the coal indus- 
try, and dismal prophecies as to its future. That its growth has ceased. And 
now come new methods of mining and cleaning the coal so that where adequate 
capital can be employed, there will be cheaper and more efficient metliods. 
This is only the beginning. The rest is a vista of a dumfounding revolution : 
That coal may in large parr be no longer used for fuel '. At least save where 
it is exceptionally advantageous, and perhaps then principally in the form of 
coke or briquets. Its all important value will be the vast variety of chemicals 
which will come from coal tar. Already a perfect troop of new medicines, 
new dyes, and new plastics, and a wide variety of other products. So that 
the prediction has already been made that with this astounding advance of 
industrial chemistry, a day is near when coal will be too valuable to burn as 
of yore." 

It is evident, as previously f)ointed out in this report,^^ that coal 
is feeling the inroacls of competing fuels and that these fuels require 
considerably less labor per unit of heat than coal. But Dr. Snyder 
apparently believes that the market created by the demand of the 
chemical industry for coal tars will fully offset the losses suffered by 
coal to competing fuels. It is extremely unlikely, however, that the 
employment created by the demand for "coal tars will offset the labor 
displaced by the substitution of new fuels because a tremendous amount 
of coal-tar products can be produced from a relatively small amount 
of coal.-" 

It must be remembered also that the coal-tar industry has been in 
existence for many years, that its products are today widely used in 

"Carl Snyder, Capitalism the Creator, Macmillan, New York, 1940, p. 30o. 
" Ibid., p. 398. 
'» See 1)1). 09-106, .supra. 

^ The relative iiii import a nee of indu.strinl cliemioals as a market for coal is to be seen 
in that tlie coal used in the manufacture of the bvproducts in 1937 amounted to 69.573.000 
tons, or 14.5 percent, of the Nation's entire coal production of 479,400.000 tons. Of this 
b}-product use. coke and manufactured s;is were by far the largest components. Byproduct 
coke ovens yield several commodities durinp the cokins or distillation of coal. The value 
of the principal products as derived in 1937 was as follows : 

Value of 

bi/prodiicts 

per 1 ton 

of coke 

Gas $1.48 

(,ok"fe . ^ 4.42 

Tar . 59 

Benzol . 41 

Ammonia . 33 

Others . 25 

(United JNIine Woikers of America. Ensrineerinp Department Monograph No. 3, Manu- 
factured Gas Versus Natural Oas. by Walter N Tolakcv 1040.) 

Of these possible components, coke is uscj jiriiicipally for metallursrical purposes, leaving 
t)ut a vei'y small proportion for industrial (licnii< .lU'. which, together with tar, benzol, 
and ammonia, as possible sources of industrial choiicals. constitute a very small segment 
of the byproduct ase of coal uud an insignilii ai.t pi iivilo'i f the taraJ ■'<>&'. prwi1ii< t'l-a. 

277551- 41— No. 22 1.? 



;j^78 CONCENTRATION OF EC'ONOxMIC POWER 

industry, but that nonetheless unemployment in the coal industries 
has grown apace until today it is a social problem of the first mag- 
nitude. Another illustration cited by Dr. Snyder : 

In the same way cotton may in the near future be no longer used as cotton, 
save under special conditious, and in the higher grades. Already experiments 
are well advanced toward the planting and cultivating and harvesting of cotton 
just as if it were wheat, sowing the plants closely together so as to conserve the 
moisture and save the cost of cultivation; then to harvest with a mowing ma- 
chine, or perhaps with a harvester-combine. Then to put the entire plant with 
the cotton bolls into the digester, and rapidly transform it into high grade 
cellulose. Estimated to be a cheaper product, of a higher quality, than that 
being made from any other kind of pulp. Possibly a new future and a new 
hope for the dying agriculture of the Southern States."^ 

The possible development of this process of transforming cotton into 
high-grade cellulose and then into textiles raises the question of the 
fate of those thousands of workers now engaged in processes which 
would be entirely eliminated under this new method. 
Or again: 

There Is an exhibit of a new glass thread almost as strong as steel and flexible, 
impervious to water, and miraMle dictu, it is now being woven into glass cloth 
for a wide variety of uses. Glass dresses have been made for years. They may 
soon drive out silk and rayon as these have driven out cotton.^ 

The substitution by rayon textiles has been a mixed blessing with 
many textile workers displaced in the process. If the production of 
glass fabrics involves lower labor requirements than silk and rayon, 
their replacement by glass may well have adverse effects upon the 
economy. On the other hand, the substitution of glass fabrics for silk 
and rayon may increase the consumption of textiles to such an extent 
that the total amount of labor involved in producing textiles will be 
augmented. Perhaps the development of the glass fabrics industry 
w^ill require large amounts of capital goods and so stimulate the 
economy as a whole. It may happen that glass fabrics will sell at 
a lower price and release a larger part of the consumer's dollar for 
the purchase of other products. But these are merely favorable possi- 
bilities; they are by no means certainties. 

These uncertainties emphasize the necessity of evaluatmg the possi- 
ble good and bad effects of each technological innovation which may 
be substituted for an existent product. 

An evaluation of the great industry-creating technological innova- 
tions of the 1920's, presented before the Temporary National Economic 
Committee by Dr. Isador Lubin, is of particular interest here because 
it stresses the importance of substitution. 

Take automoble and truck transportation. There is no doubt that automobile 
and truck transportation displaced a large number of people who were engaged 
in activities which had to do with horse transportation, but, offsetting that, you 
had road-building and petroleum refining and steel and rubber and textiles and 
glass and real estate, and a thousand and one other activities that would not 
have been possible without the automobile. So there again you had a net 
increase in the number of people who were employed. 

The gross increase in employment in these new activities was, of course, offset 
by the displaced employment in vehicle industries and stables and raising mules 
and horses and water transportation and horse trading, but, by and large, we will 
admit that the net has been a gain to the country as a whole in the total number 
of jobs. 

Again, there are movies. Employment was created in book-writing and in 
making film and in the production of movies and distribution of movies, and that 



« Carl Snyder, op. cit., pp. 398-399. 
» Ibid., p. 3«». 



CONCENTRATION OF ECONOMIC POWEIt 179 

net was greater than the number displaced among vaudeville artists and 
musicians. 

Rayon is another case in point. Through improved quality and through lower 
prices, there was a large increase in employment which was offset in part by 
losses in the silk industry and in other natural fibers like wool and cotton. 

Electric power again created employment through endless possibilities for 
cheaper light and cheaper power. On the other hand, it created unemployment in 
coal and gas and oil, but again I believe there was a net increase for the country 
as a whole.^ 

When it is realized that even these tremendous new industries have 
had certain negative effects, it becomes apparent that the probable 
overall results of less pretentious innovations must be analyzed care- 
full}^ lest they bring losses instead of gains to the economy. 

Capttcd-Saving Charactenstics of Modeim Technology. 

The assumption that a new industry in developing expands em- 
ployment and increases purchasing power because it creates a large 
demand for capital goods is not always warranted, since there are 
two closel}' related characteristics of modern technology which oper- 
ate to reduce greatly the extent of this demand. The so-called cap- 
ital-saving innovations will be considered first. 

Certain types of equipment have been developed, particularly in 
recent j^ears, which achieve substantial increases in production with 
a relatively small amount of capital outlay and a comparatively small 
expenditure of labor. For example, the production of industrial 
instruments involves little labor; yet, as previously noted,-* they have 
proved most effective in increasing output. The development of con- 
trolling devices which safeguard machinery against breakdowns and 
excessive wear has lessened greatly the continual demand for capital 
goods caused by rapid obsolescence. Industrial instruments not only 
reduce the volume of replacement orders in the capital goods indus- 
tries but they also make possible careful analyses of existing indus- 
trial equipment to determine whether it is being used to its fullest 
capacity. In many cases greater productivity is then secured merely 
by operating existent equipment at a higher tempo. 

The use of new alloys for metal tools and parts is another method 
of eidarging production which exerts only a small demand on capital 
goods. 

A particularly significant development under way is the growing use of new 
material for cutting tools, first in the form of tungsten carbide, introduced in 
this country in 11)28, and now tending to be replaced by a mixture of tungsten 
and tantalum carbides. The use of this material results in a tool of greater 
hardness, able to withstand the wear of high-speed machine cutting, with a 
great resistance to high tempera tiu'es. Introduction of these carboloy tools has 
brought about increased speed of operation, increased feed, and longer life 
between shari>enings. * * * 

Thus, on a brass-plug job. the number of pieces that can be finished between 
sharpening was increased, through the substitiUion of carboloy tools, from 200 
to 15,000. * * * 111 \\-ork on phenol resins, * * ♦ a 10-inch saw fitted 
with 1-4 carboloy teeth cuts approximately 10.(W0 feet of material without re- 
fitting, «nd its introduction permitted a daily volume of cutting larger than the 
monthly volume attained by former methods."^ 

In view of the above, if a new industry were to require a saw in its 
production processes it could be expected to use, if possible, one with 
carjboloy teeth, and thus there would be exerted a much smaller 
demand for capital goods. 



» Hearings before the Temporary National Kconomie Coniinittee, Part 30, pn. 17247-48. 
^ See i)p. 137-138, supra. 

^ Works Progress Adnihiistration, National Research Project, The Effects of Techno- 
logical Developments Upon capital Formation, 1939, p. 9. 



2§() CO^•CENTRATIO^• OF ECONOMIC POWER 

Many new chemical processes are put into oi:)erati()n with extremely 
small capital expenditures. In the beet-sugar industry the mere im- 
provement of chemical processes has contributed to increased pro- 
ductivity, raised the proportion of the sugar extracted from the beets, 
and increased the capacity of the plants.-" Chemical processes are 
among the most important capital-saving techniques. 

To cite another example : 

Even ill the telephone. iiidnstiy, which has always been regarded as tlie classic 
example of an industry subject to diminishing returns where fixed capital re- 
quirements grew faster than increases in the number of telephones installed, 
recent and current technical changes such as the introduction of carrier current 
systems of the coaxial cable are increasing the capacity of the telephone plant 
with less than a proportional increase in fixed capital Investment. 2" 

The great adaptability of numerous existent techniques is similar to 
capital-saving innovations in the effect upon employment in the 
capital goods industries. Numerous techniques can readily be 
changed, after turning out one type of good, to produce a compar- 
atively unrelated and often dissimilar product. This obviates the 
necessity of acquiring new equipment. If a considerable proportion of 
our potential new products could be produced in large quantities 
merely by slightly readjusting or rearranging existing techniques, 
the demand for capital goods — and the labor involved in their pro- 
duction — would be materially lessened. 

This adaptability has recently been impressively demonstrated by 
the fulfillment of requirements for the national defense program. 
Certain techniques are so adaptable that with practically no capital 
outlay they have easily been shifted from the production of peace- 
time goods to the manufacture of unrelated wartime products. 

Three examples of the great adaptability of general-purpose tech- 
niques deserve special attention. (1) AVithout any change whatever in 
the process, welding has been applied to the large-scale manufacture 
of airplane bombs formerl}'^ produced from forgings. Gun carriages 
are now made by means of structural welding in place of the former 
method which involved forging, casting, and riveting. (2) The punch 
press has been easily adapted to the manufacture of clips for rifle 
bullets and of all types of standard quartermaster hardware — belt 
eyelets, etc. Draw presses, originally designed to produce acetylene 
gas bottles or large milk cans, are now turning out cartridge cases in 
great numbers. (3) The automatic screw machine is capable of mak- 
ing practically any small metal part in enormous quantities. Its 
value to national defense is chiefly in the production of fuses, boosters 
and primers of ammunition. As in the case of the press and of weld- 
ing, the screw machine was adapted to this new use with practically 
no capital outlays. 

Through the use of these and other techniques, certain ty[)es of 
plants equipped chiefly with general-purpose machinery have been 
able to shift with little added expense to the production of wartime 
goods. Farm implement manufacturers readily adapted their proc- 
esses to the production of small gun carriages, transmissions for 
tanks, and machine-gun tripods. Equipment for the production of 
printing presses was found readily adaptable to the manufacture of 
gun cari'iages, gun tools, and recoil mechanisms. Makers of heavy 
Diesel engines went readily into the production of cannon and mount- 

»Ibid.. p. 10. 

'"""—=""" hpfore the Temoornrv National Economic ("oinmjtt«»p Pnrt .SO. e.xhibir 24-38. 



CONCENTRATIOX OF ECONOMIC POWER 



181 



ings for camion, while crane and shovel firms shifted to the manu- 
facture of gun carriages and railway mounts. Similarly, loco- 
motive producers found that with little added expense they could 
turn out gun carriages, and the transmission, driving, and track 
mechanisms of tanks. Manufacturers of heavy electrical equipment 
shifted to the production of such diversified goods as turbines, drive 
shafts for destroyers, cannon, and gun carriages. 

Manufacturers of cash registers and adding machines changed to 
the production of bomb fuzes. Factories formerly producing ladies' 
underwear shifted to mosquito netting for troops in the field'. Mak- 
ers of vacuum, cleaners transformed their processes to the mass-pro- 
duction of gas masks. Amono; the more unusual adaptations was 
the change by manufacturers of church pipe organs to the production 
of wooden framework for arm^- saddles. Likewise, firms which had 
been making lathes to turn out wooden ducks for shooting galleries 
went readily into the production of lathes for army shoes. 

These examples illustrate an adaptability which is by no means 
confined to shifting to the production of wartime goods. New prod- 
ucts do not necessarily require new types of capital goods if they 
can be produced with the highly adaptable general-purpose tech- 
niques now installed and in operation. 

It is evident that (1) new industries can develop and greatly in- 
crease their production by using capital-saving innovations and (2) 
the production of new and different products can often be effected 
merely by minor changes in existing techniques. Therefore, the 
emergence of new industries does not necessarily create a gi'eat de- 
mand for capital goods with a resultant expansion of ax^tivity 
throughout the entire economic system. 

The Pattern of Present Income Distribution. 

The existing pattern of income distribution is a third limitation 
upon the stimulating effect of new industries. Because over two- 
thirds of the Nation's families and individuals are in the lower-in- 
come groups, the market for new products is automatically limited. 
Table 21 shows the percentage of families and single individuals re- 
ceiving incomes of various levels during 1935-36. 

About one-third (31.64 percent) of the total number of families 
and single individuals received incomes of less than $750. nearly 
one-half (46.54 percent) less than $1,000, and over two-thirds (68.68 
l^ercent) less than $1,500. At the other end of the income scale, 
about 2 percent had incomes of $5,000 and over, and less than 1 
percent, incomes of $10,000 and over. 

Table 21. — Distribution of families and single iudividuals, btj income level, 

1935-86 





Families and single individuals 


Income level 


Number 


Percent 
at each 
level 


Cumula- 
tive per- 
cent 


Under $250 


2. 123, 534 
4. 587, 377 
5,771,960 
5, 876. 078 
4. 990, 995 
3, 743, 428 
2,889,904 


5 38 


s s« 


$250 to $500 . - ... - 




$500 to $7.50 


14 63 1 31.64 


$750 to $1,000 


14 90 46 54 


$1 000 to $1 250 


12 65 ! 59 19 


$1,250 to $1,500 




$1,500 tu $1,750 _ 


7.32 1 76.00 



182 



CONCENTRATION OF ECONOMIC POWER 



Table 21. — Distribution of families and single individuals, by income level, 
1935-36— Conthmed 





FamQies and single individuals 


Income level 


Number 


Percent 
at each 
level 


Cumula- 
tive per- 
cent 


$1 750 to $2 000 ..... 


2, 296, 022 

1,704,535 

1,254,076 

1, 475, 474 

851, 919 

502, 159 

286, 053 

178, 138 

380, 266 

215, 642 

152, 682 

67,923 

39, 825 

25, 583 

17, 959 

8,340 

13,041 

4,144 

916 

240 

87 


5.82 
4.32 
3.18 
3.74 
2.16 
1.27 
.72 
.45 
.96 
.55 
.39 

!io 

.08 
.05 
.02 
.03 
.01 


81.82 




86.14 


$2 250to$2 500 --- 


89.32 


$2 500 to $3 000 - - 


93.06 




95.22 


$3,500 to $4 000 . - 


96.49 


$4 000 to $4 500 . ... 


97.21 




97.66 




98.62 


$7 500 to $10 000 


99.17 


$10,000 to $15,000 . . . .- 


99.56 




99.73 


$20 000 to $25 000 


99.83 


$25 000 to $30 000 


99.89 




99.94 


$40 000 to $50 000 


99.96 


$50 000 to $100 000 


99.99 




100.00 






$500 000 to $1 000 000 




$1 000 000 and over 








All levels 


39,458,300 


100.00 









1 Less than 0.005 percent. 

Source: National Resources Committee, Consumer Incomes in the United States, Their Distribution in 
1935-36, 1938, table 2, p. 6. 

The overwhelming number of persons included in the lower-income 
groups indicates that a mass market for any new product must be 
created principally among them. But the development of such a mass 
market depends upon making it possible for the poorer families to pur- 
chase more than the primary necessities of life. The proportion of 
income a family may spend for new products, which are not substitutes 
for existent items of food, clothing, shelter, and personal care, can be 
measured only in terms of what remains after these necessities are pur- 
chased, since members of the lower-income groups cannot be expected 
to decrease materially their already meager expenditures on essen- 
tials. Thus, the potential market for new commodities, outside the 
field of primary necessities, should be measured in terms of the con- 
suming units' income residual (income remaining after expenditures 
for food, clothing, shelter, and personal care). The Bureau of Labor 
Statistics has computed this income residual and has represented it 
as a percentage of total income for nonrelief families in cities of 
varying size throughout the United States.^^ The percentage of 
income residual to total income by various income groups is shown 
in table 22 taken from the above-cited report. The income residual 
of the members of lower-ii^come groups is extremely limited. Over 
two-thirds of all consuming units receive less than $1,500 annually 
and their incom6 residual generally amounts to less than 20 percent 
of their total income. In many cases, families in the lowest income 
brackets (under $750) actually spend more than their total incomes 
(a practice made possible in the case of nonrelief families by the use 
of credit or past savings). 



^ Temporary National Economic Committee Monograph No. 1, Price Behavior and 
Business Policy, pp. 130-133. 



CONCENTRATION OF ECONOMIC POWER 



Ph o 



II 
"I 



111 



S"' 









c5 « (N c5 M CO CO CO CO CO -^ Tt< uD 



i^^^O^cou^io^eocD^'-i 

'-1CSMCOCOCOCOCO'^TJ<-^IOCO 



5 (0*0 »OOSQ 



:?5S5Sfe?51!S 



12S^SSSSS5S5$J 



2'°S2S?38SS?SS??!IS 



S?5S;SgS^?g5^g 



'•2SSSgSS?S5^?2 



ciy)iam<oo-va<0'^cim 



5?5S 



S5"SSS§ggJ5S;S 



■ic<c4c^wcoroeo^Tf»o 



^1 



111 



woir>-i>ogooNt~i^M-* 






tONOiOOS 



S2?5^S§J§SS5§; 



:SSg38SggS?5 



>c050eotOr<Nai»T»<coc<<o>o 



,3^ 



S& 



f"2r;2Sg5?5SSS?S^§ 



g«'^2;2i:;SSSSSS?3^3 



^■0^0 a a a a a a a a a a a !3 o 

a 3 3T3'OT3T3'Ol3-C-OX3'a'OT3c 

^^T3 aaaaacaiaaflaaoj 

C3 ca OJOOQOOOOOQOQOO 

ill !-^-!-SIS°-ll°l-?-l 



^ 


«l 


111 


.^ 


ll ! 


1 


« 


Si° ° 


J 




isl 




1 
t 


dm 


1 

pq 
§ 






2 


^ 


a a t3-'H'»'o 


?. 


I5!rl!l 


2 




.^ t- o a< ?; 0) 0) 


1 


1 


liilSII 


!ii 


^^iS^s^cS 


S 






H 






03 




i 


o 


■o ."3 S 13 -o "aTj T 




a .. 




pj 


ag 




llllllli 


a 




na 


lllKwaKn^ 


W 










WKWgSggTs 




v-i? 


ssssg 





52 iii^°'^^>i 

S oi 2 2 2^«ototxi M S 
»o«aao.ooooaS 
^ 0-" • ■■■■■ ■'^ a 



183 



184 CONCENTRATION OF ECONOMIC POWER 

If a fnniily receiving $1,500 a year has an income residual of only 
20 percent, or $300, this constitutes its buying power for products 
other than absolute necessities. A new product, which is a non- 
necessity, must compete for a share of this $300 with items of trans- 
portation, pei'sonal care, medical care, recreation, tobacco, drugs and 
cosmetics, and taxes. If the new product competes with an old 
product witliin any of these fields, the substitution involved may 
well reduce sharply the stimulus of the new industry. 

The Bureau of T^abor Statistic^ examined the market for certain 
household equipment in relation to the size of the income residual 
and concluded that, due to the smallness of the income residual among 
the lower- income groups : 

The opp<->rtunities of purchasing such productjj as refrigerators, vacuum clean- 
ers, or washing-machines are correspondingly limited. 

This does not mean that the lower income groups must be dropped from 
consideration as a potential market for these goods. * * * Obviously, if 
these markets are to be tapped, prices must be low and payments must be 
extended over substantial periods of time.^° 

The smallness of the income residual among the lower-income 
groups will not necessarily prevent the development of any new 
industry. But since the bulk of the Nation's purchasing is done by 
the lower-income groups, any new consumers' good which is not a 
substitute for an existent necessity, must compete with many other 
products for a share of the lower-income groups' purchasing power 
available for the acquisition of non-necessities, purchasing power, 
which, per family, is extremely limited by virtue of the present pat- 
tern of income distribution. 

PROSPECTIVE NEW INDUSTRIES 

It is impossible to determine precisely what new industries will 
develop in the future to stimulate materially the rate of economic 
activity. As Dr. Alvin Hansen has said . 

Certainly no one can say at this moment what great new developments the 
future may hold in store, but I should like to call attention to what seems 
to me to be a fact, namely, that economic progress, even in the nineteenth 
century, came by spurts and not at a uniform rate. Such notable students 
of economic development as Spiethoff, Wicksell, Cassel, Schumpeter, and Robert- 
son stress the discontinuity, the jerkiness and' lumpiness of economic progress. 

The history of the last 200 years affords no basis for the assumption that 
the rise of new industries proceeds at a steady pace.^ 

Between 1900 and 1933, 32 inventions or improved processes of 

fH'imary importance were developed according to Lewis Mumford's 
ist of important iuA^entions, based largely, upon the compilations 
of Darmstaedter and Feldhaus.^^ These 'develoi)ments have been 
occurring at a diminishing rate in recent years. From 1900-13, 28 
of the 32 developments took place, but from 1920-33 only four oc- 
curred. It seems that as our body of technical knowledge increases, 
the rate of new industry dev^elopment declines. It is indeed possible 

.=™Ibid., I). 131. 

•™ lleaiinp.'t boforp the Temporary National Economic Committee, Part 9, Savings and 
Invpstrtent. p. 3514. 

•^ Lett-is Mumford, Technics and Civilization, Harcourt, Brace & Co., New York, 1934, 
pp 437-446. 



CONCENTRATION OF ECONOMIC POWER 185 

that the rapid pace of new industry development which has been a 
feature of the past may not characterize the future. 
General Fields of Inquiry. 

Certain fields give some promise of future developments. Watson 
Davis, Director of Science Service, has listed eight such fields, selected 
with the cooperation of a number of representative scientists and 
engineers throughout the country. 

According to Dr. Watson, "These fields are photosynthesis, atomic 
power, long-range weather forecasting, synthetic laterials, chemical 
cures of disease, genetics, human relations, and mobilization of scien- 
tific knowledge." ^- This section very briefly analyzes these fields as 
to their present state of scientific development and their possibilities 
of developing into important employment-creating industries. 

1. Photosynthesis. — Green plants convert solar energy into useful 
materials for mankind by photosynthesis, but the manner in which 
the plant accomplishes this still remains unknown. "According to Dr. 
O. L. Inman, Director of the Kettering Foundation for the Study 
of Chlorophyl and Photosynthesis at Antioch College, the best esti- 
mates are that the energy reaching the earth from solar radiation 
each year is equivalent to that received from burning 400 septillion 
tons of anthracite coal." ^^ If only a minute fraction of this enor- 
mous amount of energy were directly convertible, any potential prob- 
lem of adequate power in the future would be completely solved. 

If chemical techniques are developed to accomplish photosynthesis 
artificially, coal, fuel oil, and natural gas would be eliminated as 
sources of power, but society would gain a much cheaper and more 
adequate source of energy. Similarly, this development would un- 
doubtedly revolutionize agriculture. 

The attainment of controlled photosynthesis, however, appears to 
be so far distant that it should not be relied upon as a prospective 
employment -creating industry. 

2. Atomic po^ver. — This possible development has already been 
briefly discussed in the section on power and energy development.^* 
There are vast untapped stores of -energy within the atom which, if 
released, would furnish almost unlimited amounts of power. The 
active component of uranium has been separated and its atoms split 
with the release of an enormous amount of energy. But atom smash- 
ing apparatus can only be used at present in the laboratory. As in 
the case of photosynthesis, its attainment, unlikely in the near future, 
would probably displace a A'ast amount of labor, and this displace- 
ment would have to be offset by an enormous expansion in industrial 
activity if no net decline in employment were to occur. 

3. Long-range ^oeather forecasting. — 

If it were possible to know what the weather was to be next year or 
several years from now, whether the growing season iu various regions was 
to be satisfactory or unsatisfactory, whether the winter was to be abnormally 
cold, whether there was to be too little or too much rain, the savings to agricul- 
ture, industry, and the Nation would be very large.'" 



32 Hearings before the Temporary -National Economic Committee, Part 30, p. 16274. The 
following description of the present state of scientific ' development in all of these fields is 
adapted from the material presented by Dr. Davis to the Temporary National Economic 
Committee. 

^ Ibid., p. 16275. 

31 See pp. 105-106, supra. 

35 Ibid., p. 16283. 



186 CONCENTRATION OF ECONOMIC POWER 

Long-range forecasts for the United States are considered prob- 
able in the future by Dr. Charles F. Brooks, director of the 
Blue Hill Meteorological Observatory of Harvard University, but 
there are many research problems to be solved before they become 
practicable. 

Even if long-range weather forecasting were possible in the near 
future, it would create only very limited employment opportunities. 
Its value to society would undoubtedly be considerable, but its pos- 
sibilities of being a major stimulant to economic activity are ex- 
tremely remote. 

4. Synthetic materials. — Synthetic materials have already been 
discussed.^® Many synthetic materials are already in use : Drugs, 
dyes, and chemicals from coal tar; a multiplicity of plas^'ics or 
synthetic resins; alloys of iron and other materials; rayon; etc. 
In the last few years chemistry has brought forth several new tex- 
tiles : Nylon, a silk-like synthetic fiber made basically from coal, 
air,, and water; vinyon, another synthetic fiber; synthetic wool, 
made from casein or other protein; and glass fibers. Clay has been 
transformed into synthetic mica which makes the Nation poten- 
tially independent of foreign supplies. Substitute materials, though 
bringing enormous benefits to society, often result in a reduction 
of employme]it where, as is often the case, less labor is required in 
their production than in the commodities which they replace and 
output is not sufficiently increased to compensate for the attendant 
decline in unit labor requirements. 

5. Chemical cii-res for disease. — Remarkable success in the use of 
sulfanilamide and its related chemical compounds in treating a 
variety of diseases focuses attention upon the possibility of new 
chemical cures for disease in the future. It is unlikely, however, 
that either the discovery or production of new chemical cures would 
create many employment opportunities since chemical processes re- 
quire only a relatively small amount of labor. But the further use 
of chemical cures may counteract somewhat the present decline in 
population growth by lengthening the span of life and permitting 
individuals to engage in active work during a longer period of their 
life. As a source of new employment, however, the production of 
chemical cures for disease must be considered unimportant. 

6. Genetics. — Since the turn of the century considerable advances 
have been made in scientific breeding of both plants and animals. 
Superior* grains, fruits, and vegetables, and superior animals from 
the standpoint of meat and wool production, are indications of the 
economic possibilities of applied genetics. The economic value of 
cross-bred seeds which give increased yields because of hybrid vigor 
and of rust-proof and disease-proof crops is evident. Important 
raw materials, such as cellulose from cotton, paper pulp, and corn- 
stalk waste, as well as plants for food, are all controllable by genetics 
methods. 

Genetics is another field in which further developments may be 
of great value to society but scarcely promises the emergence of a 
vast new industry. As E. D. Merrill, administrator of the botanical 
collections of Harvard University, says : 

83 See pp. 106-110, supra. 



CONCENTRATION OF ECONOMIC POWER 137 

It is very difficult for one to indicate tlie tangibles in genetics in reference 
to plants and plant breeding. I personally doubt if any new major industries 
can be developed on the basis of research in this field. ♦ ♦ * New sig- 
nificant social developments could hardly be expected, but the increased utiliza- 
tion of knowledge in industry would have a very favorable reflex action on our 
whole social set-up." 

7. Hum-cm relations. — As Dr. Davis stated : 

Human relations in the factory, in the comraunSty, and in the home miglit not 
seem to be at first consideration a problem for research and technology. As a 
matter of fact, it seems probable that the methods of scientific research applied 
to this great problem in which the reagents are human beings will be capable 
of producing useful and fruitful results with as much assurance as they do in 
less animate fields. * * * Man, as an individual and in the group, is the 
subject of investigation by psychologists, psychiatrists, teachers, administrators, 
and others who deal with various human problems. It would seem logical that 
some of tile findings in these fields might be applied profitably to the difficult 
relations in the fields, of business, politics, and international affairs.^' 

This field, however, is still in its infancy. Its principal economic 
value in the near future will probably be the devising of new per- 
sonnel procedures. It will thus provide employment for a very small 
number of highly trained specialists but its importance as an expan- 
sionist fnreo in the economy will be negligible. 

8. Mobilization of scientific kno^ohdge. — In the word of Dr. Davis: 

With the 'accelerating pace of scientific research, invention, and development, 
the distribution, interpretation, and utilization of the knowledge already ob- 
tained becomes an increasingly important problem. * * * The great failure 
to our organization of our written knowledge lies in the inability of anyone to 
put his finger upon all the literature on a given subject with relative com- 
pleteness and at a reasonable cost. Our organized knowledge, as contained in 
the printed literature, is extraordinarily jworly indexed from the standpoint of 
its efficient utilization. 

In a few fields, such as cnemistry, there are abstract journals which do an 
invaluable job. But in many fields bibliographic resources are quite inadequate, 
resulting in investigators not being able to discover what researches have been 
made in a particular line of inquiry in the past. New mechanisms recently 
developed, or in the process of development, which may be called new tools for 
intelligence are likely to prove useful in this needed mobilization of knowledge.'® 

Of these mechanisms, the card index and microfilm are of primary 
importance. The latter can produce at low cost single-copy^ editions 
of anything a camera can see. 

The indirect benefits to society of marshaling available scientific 
knowledge would imdoubtedly be immeasurable, but the employment 
opportunities directly created thereby would scarcely exceed several 
thousand. 

Further developments in these eight fruitful fields, selected by 
Watson Davis and participating scientists and engineers, would un- 
doubtedly greatly benefit mankind, yet it is unlikely that a great 
employment -creating industry will emerge from any one of them in 
the near future. In the first place, several of the eight, such as 
photosynthesis, appear to he far distant in regard to practical useful- 
ness. Others, such as synthetic materials, would result in the re- 
placement of an existent product by a new material; since in all 
probability the new material would be produced with lower unit labor 
requirements than the old, its introduction would stimulate economic 
activity only if there resulted an offsetting expansion in production. 

S'' Hearings before the Temporary, National Economic Committee, Part 30. p. 16288. 
»« Ibid., p. 16289. 
^ Ibid., p. 16290. 



Igg CONX'ENTRATION OF ECONOMIC POWER 

A third group, such as the mobilization of scientific knowledoe, would 
serve a very useful purpose, which, however, cr.n be realized with an 
extremely small increase in employment. 

For example, it is difficult to see in these fields the promise of a 
new industry comparable in its effect upon the economic system with 
the railroad industry in the last century or the automobile industry 
in the 1920's. 
/Specific neio industries. 

Several new industries have each been described at one time or 
another as "the coming new industry," but their effect upon economic 
activity is still problematic. These industries are prefabricated hous- 
ing, air-conditioning, television, and Diesel engines, each of which 
will be briefly examined with reference to its possibilities of develop- 
ing into a great new industry. 

1. Prefabi^icated housiDg. — Though a number of the parts used in 
building a house are already prefabricated on a mass-production 
basis — for example, doors and windows (where stock sizes are used 
and glazing is allowed off the job), furnaces, lighting fixtures, etc. — 
repeated attempts have been made to put entirely prefabricated houses 
on the market in order to reduce field assemblage to a minimum.*" 
Sears, Roebuck & Co. have tried to popularize pre-cut houses, all 
lumber and parts to be fitted on site. Although they sold a large 
number of houses, little saving resulted from this method. Similarly, 
steel companies turned to the prefabricated field during the depres- 
sion but were unsuccessful in reducing the cost of houses much below 
those built by ordinai*y site assemblage.*^ 

About 50 efforts have been made to produce prefabricated houses, 
but none apparently has been successful in achieving marked reduc- 
tions in cost. This is due in part to the limitations upon the market. 
Various forms of resistance have been met, such as lack of acceptance 
by the building public, lack of cooperation on the part of labor, in- 
flexible requirements in building codes, efforts to popularize houses 
made of but one material, transportation difficulties of shipping the 
large-sized items at low cost, etc. Because of these obstacles, the 
companies in the business of producing prefabricated houses of more 
or less traditional materials have been able to accomplish economies of 
only about 10 percent so far, compared with ordinary methods ; and 
these small savings are due in large part to quantity purchases.*- 

Pending tlie overcoming of these obstacles, the future of pre- 
fabricated housing is none too bright. It has indeed shown few 
signs of developing into a great employment-creating industry. As 
one student of the industry remarks, "Little progress has been made 
in the last 10 years." *•' 

2. Air-condifiomnfig. — Considerable advajices have been made in 
air-conditioning in recent years, particularly for use in public build- 
ings, retail stores, movie theaters, restaurants, etc., and it is probable 
that the future will see an increased use in these fields. Air-condition- 
ing will become a major stimulus to the economy, however, only when 

<« Temporary National Economic Committee Monograph No. 8, Toward More Housing, 
Part I, "Some Kconomic Aspects of nousinc:," bv I'eter A. Stone, 1040, p. 108. 

«Ibid., p. 108. 

<2 Hearings before the Temporary National Economic Committee, Part 11, Construction 
Industry, pp. 53.37. (5338. 

*= Temporary National Economic Committee, Monograph No. 8, op. cit.. p. lOS. 



CONCENTItATIOX OF EfOXOMIC POWER IgQ 

it reaches the mass market of residential buyers. In the words of 
Dr. Paul Douglas of the University of Chicago : 

Perhaps new housing, television, and air-conditioning are the most promising 
possibilities, but unless great reductions are made in their costs, these will be 
out of reach of the incomes of the great masses who alone can create a suf- 
ficient demand to make large-scale production profitable." 

Certain technical problems have yet to be overcome in developing a 
product suitable for the residential market. For example, in areas of 
very low humidity air-conditioning requires a new type of window. 
Where cold temperatures prevail in the Avintertime, double windows 
are necessary. Though much progress has been made along these lines, 
wide development of residential air-conditioning remains for the 
future.*^ 

Perhaps no one. is better fitted to testify regarding the economic 
potentialities of these new industries, such as air-conditioning, than 
the noted inventor and scientist, Charles F. Kettering. 

We have heard a lot of talk, we will say, about air-conditioning. Air-conditioning 
has done a very, very good job, but yet the economics of air-conditioning perhap.s 
haven't reached the point at which it can be of general usefulness. We are 
trying to find out what we can do to make this thing more flexible, easier to 
handle, more easily installed, and things like that.^* 

3. Television. — Like air-conditioning, television faces technical dif- 
ficulties before it is ready for general use, chief among which is the 
problem of transmitting for more than 50 or 100 miles. Even^where 
transmission extends over only a few^ miles, reception on a set of 
reasonable cost still leaves much to be desired under most circum- 
stances. AVith regard to the possibilities of television. Mr. Kettering 
observed : 

Take television as a new industry. It will have to struggle along quite a long 
while before it strikes its pace, because you can't hit the middle of the road but 
very rarely, and then it is an accident." 

No matter what its future development, television will probably not 
bring about any considerable net gain in employment since its pro- 
duction and use will undoubtedly involve a large degree of substi- 
tution affecting the radio industry. 

4. Diesel engines. — When asked whether he saw on the horizon any 
new product which might develop into a great new industry within 
the next decade, Mr. Kettering replied that aside from the Diesel 
engine he knew of no new invention or development which promises 
to stimidate the economy in the next decade as did the automobile, or 
even the radio, in the twenties.*** 

Tlie Diesel engine not only is a substitute for steam locomotives but 
diminishes greatly the amount of maintenance and repair labor 
required in railroad transportation; according to Dr. Isador Lubin. 

The Diesel engine makes it possible to keep traffic moving, day after day, year 
In and year out, without having to have the expensive equipment and the labor 
force necessary to keep it serviced. I think I read of one the other day that 
had been in service a year and a month and had been to the shop once. It ran 
365 days a year, which makes it possible for one engine to do the work of four, 
because it can be used four times as long in 24 hours, which is not only labor- 

** Paul Douglas, Controlling Depressions. W. W. Norton, New York, 1935, p. 82. 

^'- ll^earings before the Temporary National Economic Committee, Part .30, p. 16302. 

*'' Ibid., p. 16:?02. 

" Il)id.. p. 16:102. 

«Ibi(l., 16:U5. 



190 OOXCENTRATION OF ECONOMIC POWER 

saving but also capital-saving. You need only one whereas formerly you needed 
four/' 

Perhaps today there is some great employment-creating industry 
on the horizon. Any prediction that new products will not be devel- 
oped, or if developed cannot be expected to overcome limitations in our 
economy might prove 'just as fallacious as the widespread belief that 
technology, because it has created great new industries in the past, 
will inevitably continue to do so in the future. 

Technological progress is always unpredictable, bnt in light of the 
foregoing it seems unlikely that we can rely heavily upon future new 
industries to offset the labor-displacing effects of technology. 

THE EEDUCTION OF PRICES 

THEORETICAL IMPORTANCE OF PRICE REDUCTIONS 

The importance of lower prices as a force to offset labor-displace- 
ment rests upon the assumption that price reductions increase demand, 
thereby expanding output and creating employment. Even the de- 
velopment of a new industry frequently depends upon prices low 
enough to command a wide market. 

The importance of this compensatory force in economic theory can 
hardly be overemphasized. References to its theoretical importance 
have already been presented in Part I of this report, "Technology 
and Economic Thought." 

This fundamental tenet of economic thought, that over a long-term 
period increases in the productivity of labor are reflected in lower 
prices of commodities, with a resultant increase in production, should 
be examined. If prices do not decline as productivity increases, this 
basic principle is either inoperative or the functioning of the present- 
day economy prevents its operation. 

it is not the purpose of this report to undertake an extensive exami- 
nation of price behavior. Such an analysis can be found in another 
report of the Temporary National Economic Committee.^^ But there 
is one element in the determination of price behavior which bears a 
direct relationship to technology. In recent years, the possible exist- 
ence of a relationship between the concentration of economic power and 
price inflexibility has been frequently examined. For example, a 
recent study of the National Resources Committee stated : 

The main conclusion to be reached from this analysis is that, while many factors 
influence price insensitivity the dominant factor in making for depression in- 
sensitivity of prices is the administrative control over prices which results from 
the relatively small number of concerns dominating particular markets.^ 



■•» Ibid., p. 17248. While it is true that Diesel engines already have definite application 
where relatively large power upits are required, they have not yet been proved practicable 
for general use in the lighter cars. In some quarters, moreover, it is believed that the 
extra cost for efficient Diesel units for small cars will more than offset the fuel, savings 
that they achieve. (U. S. Bureau of Labor Statistics. Monthly T.abor Review. .Tune "11140, 
"Employment Prospects in the Petroleum "and Natural-Oas Industry," by H. O. Rogers, 
p. 1301.) 

" Temporary National Economic Committee Monograph No. 1, Price Behavior and 
Business Policy, 1040. 

^'^ National Resources Committee, The Structure of the American Economy, Part I. p. 143, 
1939. 



CONCENTRATION OF ECONOMIC POWER 191 

Certain economists have pointed out that centralized industries do not 
reflect the declines in prices necessary to stimulate production and thus 
offset the effects of increasing productivity of labor. If a relationship 
exists between concentration and price stability and in turn between 
technology and concentration, then technology itself would be an 
obstacle to the functioning of the basic compensatory force of price 
reductions. 

The next chapter examines the relation between technology and 
concentration but of pertinence here is an analysis of the behavior of 
labor productivity and industrial prices in concentrated and non- 
concentrated industries. 

Industrial Prices and. Labor Productivity. 

To ascertain whether or not labor-displacement has been offset by 
price reductions, a study was made of the relationship between the 
behavior of labor productivity and of prices in nine major industries. 
The industries have been separated into two groups, concentrated and 
nonconcentrated, the principal standard of delineation being the per- 
cent of the industry's output produced by its four largest firms. 
Durable and nondurable goods industries are represented in each 
group, as are both producers' goods and consumers' goods industries 
in the concentrated group. A detailed analysis of productivity and 
prices in each industry will be found in appendix H. General con- 
clusions will suffice at this point. 

In that analysis the increasing productivity of labor and the rigid- 
ity of industrial prices are not regarded as separate phenomena. 
Kather, it is important to discover whether the industries which have 
been cliaracterized by inflexible prices also have benefited from exten- 
sive technological savings on labor expenditures. 

The National Research Project, in its study of 59 manufacturing 
industries, found only one which was not characterized by a marked 
long-term decline in unit labor requirements, the bakery products 
(other than biscuits and crackers) industry. ^^ The possibility that 
prices in certain industries have fallen only partially as much as 
unit labor requirements therefore needs exploration. 

A summary comparison of the productivity-price relationship of a 
concentrated industry w^ith that of a nonconcentrated industry is made 
in chart XIV, which contrasts the behavior of the unit labor require- 
ment and price indexes in the cement industry with the behavior of the 
indexes in the furniture industry. In general, these two industries may 
be regarded as typical of the two groups. 

The price indexes in the concentrated industries tend to remain w^ell 
above their pre-war position, while the price series of the nonconcen- 
trated industries closely approach the 1914 level. The unit labor re- 
quirement series tend to drop more extensively than price since 1919 in 
the concentrated industries, while the two series tend to parallel each 
other in the nonconcentrated industries, with the price series often 
evidencing, for sustained periods, a decline more extensive than that 
of the unit labor requirement index. This difference in the type of 
relationship is graphically apparent in the comparison between the 
cement and furniture industries and also characterizes the other 



^ Works Progress Administration, National Research Project. Production, Employment, 
and Productivity in 59 Manufacturinng Industries, 1919-36. (By nnit labor requirements 
is meant the number of man-hours required to produce a given amount of goods.) 



192 CONCENTRATION OF ECONOxMIC POWER 

Chart XIV 

INDEXES OF UNIT LABOR REQUIREMENTS 
AND PRICES 

UNITED STATES 



INDEX NUMBERS 

(i926 = IOO) 
200 



CEMENT 



INDEX NUMBERS 











































































































UNIT LABOR REQUIREMENTS 
































/ 






















<. 


\ 












































^/ 


'" 


\ 




^ 


^ 


V~ 




























p 


RICE 


, 




















^ 


^ 


-- 


"n' 




=£ 


^^ 






" 






/' 
































- 


- 


\ 




""• 


,' 

=; 














Z^ 

































1914 1916 1918 



1922 1924 1926 



1930 1932 1934 1936 '37 



INDEX NUMBERS 
( 1926 ■ 100) 

200 



FURNITURE 



























































PR 


CE 


Vi 
















































;\ 














































l\ 




i 


s^ 










































1 






v-^ 


V. 


<^ 


-- 


- 




JNIT LABOR REQUIREMENT 


s 














|\ 












/- 


/ 
















--^j; 


■^ 


^=*^ 


^nI 


K 


1 


, 


- 


— 
































*■• 


" 






' 

















































NOEX NUMBERS 

(I32e-I00l 

200 



1914 

Source : 



1916 1918 1920 1922 1924 1926 1928 1930 1932 1934 1936 '37 

Appendix H, Tables 4 and 11. 



CONCENTRATION OF ECONOMIC POWER X93 

concentrated industries — iron and steel, nonferrons metals, automo- 
biles, cigarettes, electric light and power — as well as the two other 
nonconcentrated industries, cotton and woolen goods. 

The tendency of unit labor requirements to decline more extensively 
than price in the concentrated industries naturally varies in degree and 
in time among them. It was most noticeable in the iron and steel 
industry during 1919-29 and again during 1933-37. In the nonferrous 
metals industry, it continued throughout the entire period, except for 
brief interruptions in 1931-32 and 1934-35.''* In the motor vehicles 
industrv, the price series declined considerably more than unit labor 
requirements from 1921 to 1923. During the' periods 1919-21, 1923- 
30, and 1933-37, however, the tendency was reversed." In the ciga- 
rette industry, tlie general decline in unit labor requirements throuoh- 
out the entire period was far great.er than any decrease in price, 
except for 1933-34."*^ The over-all decline in unit labor requirements 
in the electric light and power indus^try was materially greater than 
the decrease in price, despite the relatively large expenditures of man- 
hours during the twenties involved in the installation of new light and 
power facilities.^' 

In rather sharp contrast, the unit labor requirement index in the 
nonconcentrated industries seldom declined more extensively than the 
price series. In the cotton goods industry, the two indexes tended 
to parallel each other during 1920-30. During 1932-34, the price 
series turned upward from its dejiression low, but in the following- 
two years declined with the unit labor requirement index. The pro- 
ductivity-price relationship in the woolen and worsted goods indus- 
try })arallels almost exactly that of the cotton goods industry except 
for the last two years. In the furniture industry, the two indexes 
moved closely together until 1930, aftei- which the price series fluc- 
tuated moderately at a level well below that of the unit labor require- 
ment index. In the cement industrv. however, the unit labor re- 
quirement index, as noted, fell considerably more than the price index 
for practically the entire period, 1919-36. 

Perhaps of greatest pertinence to rhe conleinporary unemployment 
problem is the divergence of trends between unit labor requirements 
and prices since 1929. The enlargement of the spread may be attrib- 
uted to the upturn of prices from the depths of 1932-33 which took 
place from 1934 to 1937. By 1937 the price series in the concentrated 
industries (except in the electric light and power industry) had risen 
to levels only slightly below, and in some cases actualh* above, the 1929 
levels. But the unit labor requirement series turned sharply downward 
after 1933, following its rise in the worst years of the depression be- 
cause of curtailment of output, reaching an all-time low by 1936 (except 
in nonferrous metals).""' 



••'The toiulcnoy of unit l.ibor reiiiiii^ements to rise in poi^iods of srOMtly riirtMiled out- 
put. su<;Ii as l'.i;iO-o.">, should not ;:enorally be interpreted as invali<latin^r the lons-terni 
downward trend. The period of greatly inirtailed output in tlie nonferrous metals in- 
dtistiy after lii.'iO was extended partly because of the prioe-pe^jiinj; aetix^ities of the 
industry diiriiui 1020-30 and the resultant creation of enormous stocks. (For further 
discussion see pp. 2 47-2-" 0, infra.) 

^^ See p. 2.'")0. infra. 

^•'' See p. 2!i '.. infi^a. 

^'.See p. 2tiO. infra. 

■'■* It should be noted that, althoutrh the unit labor recjuirenient series extend only to 
lO.iO. the. indexes would uiuioul)tedly be at lower levels in tlie more prosperous year of 
10:'7 because of the temlency of uian-hdurs per unit to decline as output increases. 

2775r.l— 41— No. 22 14 



194 OOXCENTKATION OF ECONOMIC POWER 

It is apparent that the relative nonnse of the technique of price re- 
ductions in the concentrated industries as a means of meeting the 
unemployment problem created by declining unit labor requirements 
has become particularly acute in recent years. The extensive decreases 
in unit labor requirements in these industries have created an unem- 
ployment problem met only to a very limited extent by any increases 
in production which can be attributed to reductions in prices. In- 
dustrial prices, by remaining comparatively stable in the face of 
marked declines in unit labor requirements, cannot have played an 
effective role in meeting the unemployment problem created by tech- 
nological change, and as long as they maintain that relative stability 
their importance as a compensatory force must be regarded as 
negligible. 

The first of the two steps required to discover whether or not 
technology tends to lead to the negation of one of its own compensa- 
tory forces has thus been presented above in the summation of the 
relationship between concentration and price stability. The second 
step is presented in the following chapter, which is concerned with 
the possible existence of a relationship between technology and 
concentration. 



CHAPTER IV 

TECHNOLOGY AND THE CONCENTRATION OF ECONOMIC 

POWER 

THE CONCENTRATION OF INDUSTRY 

One of the oiitstaiidiiig- developments in American industry is the 
growth in the concentration of economic power. Economic concen- 
tration connotes the existence of large units of production, of vast cor- 
porate enterprises producing large quantities and often a striking 
variety of goods, of induh;trial empires owning and controlling ma- 
terials, equipment, and processes from the extraction of raw materials 
to the distribution of finished products, of dominance over an entire 
industry by a small number of giant concerns, and frequently of 
policies pursued by these concerns designed to minimize the existence 
of price competition. Concentration is a problem of vital importance 
to the Nation, particularly because technology steadily increases the 
power at the command of giant concerns by creating lew tools, proc- 
esses and products. 

A high degree of concentration exists in many segments of our 
economy and is rapidly growing in others. Even in agriculture con- 
centration is a growing phenomenon. The extent of concentration in 
manufacturing has been statistically measured by determining the 
proportion of the total output i)roduced by the four largest concerns 
manufacturing a given connnodity. AVherever the four largest pro- 
ducers mamifacture over half the entire output of a given commodity, 
a high degree of economic concentration, by almost any standard, 
has evidently been established. 

The degree of concentration in all manufacttiring in 1935 is shown 
in table 23 by physical mnnber of products and by value of products. 

Industries in which the four largest concerns make over 75 percent 
of the total number of munufactiu'ed products account for 39.4 percent 
of the total number and 25.9 percent of the total value of manufac- 
tured products, while industries in which the four largest concerns 
make over 50 percent of the total number of manufactured products 
account for 71.9 percent of tlie total mnnber and 54.4 percent of the 
total value of manufactured products. 

Other indications of concentration were cited by the President of the 
United States in his message to Congress April 20, 1938 : 

Statistics of the Bureau of lutenial Revemie ri^veal the following amazing 
figures for lOoo : 

Ownership of corporate assets: Of all corporations reporting from every part 
of the Nation, one-tenth of 1 percent of them owned .12 percent of the assets of 
all of them. 

And to clinch the point: <*t' all corporations reporting, less than 5 percent of 
them owned 87 percent of all the assets of all of then). 

Income and profits of corporations: Of all the corporations reporting from 
o^•er\■ part of the country, one-tenth c)f 1 percent of them earned 50 percent of 
the net income of all of them. 

And to clinch the point: Of all the nianufactiu'ing corporations reporting, ](>ss 
than 4 percent of them earned 84 percent of all the net profits of all of them.' 



' Toth Cong., :!il sess., S. Doc. ITo. Stroiiiithenins and Kiiforccmcnt of Antitrust Laws, 
p. 2. 

lOo 



196 



CONCENTRATION OF ECONOMIC POWER 



Table 23. — Pevcent of number and value of products manufactured by 4 largest 
concerns,^ 1937 







Products 




Value of products 


Perci'nt produced by 4 largest concerns 


Number 


Percent 


Cumula- 
tive 
percent 


Amount 

(thousand ! Percent 
dollars) j 


Cumula- 
tive 
percent 


1 to 5 


1 

10 
28 
44 
46 
54 
69 
91 
75 
85 
98 
100 
130 
124 
135 

101 
75 

153 
175 


0.4 
.6 
1.5 
2.4 
2.6 
3.0 
3.8 
5.0 
4.2 

11 

7.2 
6.8 
7.5 
6.5 
5.6 
4.2 
4.9 
8.5 
9.7 


0) 
0.4 
1.0 
2.5 
4.9 
7.5 
10.5 
14.3 
19.3 
23.5 
28.2 
33.6 

46^3 
53.1 
60.6 

76.9 
81.8 
90.3 
100.0 


3.53, 432 
357, 663 
477, 592 
792. 708 

1, 454, 942 
1,304,025 
1,421.494 
2, 892; 890 
1,461.146 

2, 060, 290 
893, 227 

1, 761, 698 
1, 287, 529 
1, 445, 246 

1, 689, 270 

2, 224, 582 

1, 449, 834 
648, 389 

2, 796, 032 
319,819 

1, 827, 858 
586, 027 


1.2 
1.2 
1.6 
2.7 
5.0 
4.4 
4.8 
9.8 
4.9 
7.0 
3.0 
6.0 
4.4 
4.9 
5.7 
7.5 
4.9 
2.2 
9.5 
1.1 
6.2 
2.0 


1 2 


5 1 to 10 


2 4 






15 1 to 20 


6 7 








16. 1 


30 1 to 35 


20 9 


35 1 to 40 


30 7 




35.6 


45 1 to 50 


42.6 




45.6 




51.6 


60 1 to 65 


.56.0 




60.9 




66.6 


75 1 to 80 


74. 1 




79.0 


85 1 to 90 


81.2 


90 1 to 95 


90. 7 




91.8 


(3) 


98.0 


(4) __. 


100.0 






Total 


1,807 


100.0 




29, 505, 693 


100.0 













1 The data represent appro.ximately one-third of all manufacturing industries: in nearly every case all of the 
products of the industries studied were covered. In an endeavor to make the sample representative an 
attempt was made to establish a balance between industries with a large and those with a small number 
of concerns, and between industries with a high and those with a small value of products. The sample 
included one-third of all industries within each of the census industry groups, with the e.xccption of printing and 
publishing. The last 2 frequency groups, involving possible disclosure of individual companies, represent 
concentration ratios of over 85 percent. 

2 Less than 0.1. 

3 ^Vithheld to avoid disclosing the operations of individual companies. 

< Withhf Id to avoid disclosing the operations of remaining companies. There is not necessarily a dis- 
closure among the 4 leading companies. 



Source; Temporary National Economic Committee Monograph No. 
Part V, "The Concentration of Production in Manufacturing," table 1. 



7, The Structure of Industry, 



Technology is undoiibtedl}^ one of the primary causes of this con- 
centration. For example, the United States Tariff Commission in 
1939 observed concerning the manufacture of electric lamp bulbs: 

There were five mamifacturers of bulbs in the United States in 1917, and there 
have been only two mailing large bulbs since 1920. This concentration in output 
was due in part to the fact that the most economical production was possible 
only in highly mechanized, plants of very large capacity, and in part to the 
control which owners of patents have exerted over production." 

The greater operating efficiency of large units of production stems 
from technology. Likewise, the control exerted by means of patents 
stems from technological developments which under our law may be 
patented. The relation of the operating efficiency of large units, of 
large industrial research organizations, and of patents to concentra- 
tion will be analyzed in this chapter. 



^ U. S. Tariff Commission, incandescent Klectrie T/amps. Rept. No. 13.3, Second Series. 
1030, p. 10. It should continuously '"' borne in iniiul that the cau.ses of concentration, 
both direct and indirect, are numerous nnd varied: and of them, technology is l)Ut one. 
The occurrence of nierscrs and consolidations, as one example, has been due in many 
cases to the prospect of profit inakinu tbrouKli s(>ciirity speculation in the eoneorns 
involved, rather than to any technoIoKical advanla^'o to bo derived theiofrom. 



CONCENTRATION OF ECONOMIC POWER 197 

COXC&XTRATIOX AND OPERATING EFFICIENCY 

The determination of whether or not technological development 
leads to the concentration of economic power rests first npon an 
examination of the relative efficiency of large-scale and small-scale 
operations in producing under similar conditions a certain number 
of units of a given commodity. Though such a comparison is impos- 
sible with existent data, certain types of analyses should cast con- 
siderable light upon large-scale versus small-scale operation. 

All discussions of relative efficiency of large and small units must 
not lose sight of the existence of a point of diminishing returnsin 
any productive process since there is obviously a point beyond which 
large-scale operation results in increased rather than decreased unit 
costs. Even though there is at present no adequate statistical method 
of determining where that point occurs, it is still possible to discuss 
in general terms the relative efficiency of large-scale equipment as 
against small-scale equipment.'^ 

The use of large-scale equipment is advantageous because it gen- 
erally results in (1) "smaller capital outlay per unit of capacity," 
(2) ''greater mechanical efficiency," (3) "the use of a considerably 
smaller amount of fuel and also of labor per unit of capacity or of 
output," (4) smaller floor space requirements and thus less plant con- 
struction per unit of capacity, (5) lower transportation costs per unit 
of capacity, and (6) "the use of refinements and of auxiliary devices 
Avhich result in improved efficiency of operation." * 

Tliree theoretical principles have been advanced by Prof. P. Sargant 
Florence in support of the contention that — 

larfje-srale production, espcciaUij ichen conducted in large-scale firms and plants, 
results in. maximum efflcienci/. * * * The Principle of Bulk Transac- 
tions. * * * The total monetary, physical or psychological costs of dealing 
in large quantities is sometimes no greater (and in any case less than propor- 
tionately greater) than those of dealing in small quantities: and hence the cost 
per unit hecomes smaller with large quantities. * * * The Principle of 
Massed Reserves. * « * The reserves that are economized may in fact he 
labor, liquid monetary resources, stocks of goods and materials, or any other 
factors in production, when the demand upon these factors are somewhat uncer- 
tain in their incidence. * * * This * * ♦ is merely [in economic appli- 
cation] * * * the statistical theory of probable error that the greater the 
number of items involved the more likely are deviations in their amounts to 
cancel out and to leave the actual result nearer to the expected result. The 
probable deviation in orders for similar items that a reserve guards against is 
thus proportionately less when orders are many, and the cost of reserves per 
unit of output fails correspondingly. The Principle of Multiples. ♦ * * The 
smaller the scale of operation and the fewer the total number of persons dividing 
and diffusing their labor, the less chance there is of all of them being fully made 



3 It should be clearly understood that this analysis does not relate to that centraliza- 
tion of economic power resulting from associational activities. As far as horizontal 
nsreements of any kind are concerned, there is little evidence that their e.xistence raises 
the operating efficiency of member units materially. Dr. Arthur Lucas, in his study of 
British experiments in the control of coini>etition, has observed : 

'It does, not appear that the terminable association can claim much credit for its aui.ity 
to increase the efficiency of an industry's productive organization. * * * The chief 
economies in this field arise out of the concentration of processes, the specialization of 
function, the scrapping of inefficient equipment, and the unification of management. 
Except in rare cases it is not possible to introduce these measures to an important de- 
gree without a stronger organization than most associations possess." (.\rthur F. Lucas, 
Industrial Reconstruction and the Control of Competition — The British Experiments, 
Loncmans, (ireen & Co.. London, 19.37, n. .S18.) 

■• Works Progress Administration, National Research Pro.iect, Effects of Current and 
Prospective Technological Developments Upon Capital Formation, by David Weiutraub, 
19:;9, p, 5. 



298 /CONCENTRATION OF ECONOMIC POWER 

use of as specialists. * * * Suppose that an article is being manufactured 
by subjection to 3 consecutive processes, the first a hand-process where a specialist 
can make 30 units a week, the second an automatic machine process where 
1,000 units can be made in a week, the third a semi-automatic machine process 
where 400 units can be made per week. Then to employ all the specialists and 
special machines fully a number of units must be made per week that is a multiple 
of 30, 400, and 1,000 ; otherwise some man or machine will be partly idle. In 
this case the lowest common multiple number of units that will employ specialists 
in all processes to full capacity is 6,0<X». * * * Clearly, this assumes large- 
scale production * * *.' 

■ It should not be thought, however, that large-scale operation is 
invariably the most efficient. The extent of its efficiency varies with 
the circumMances of production. Certain types of productive opera- 
tions readily lend themselves to the use of large-scale equipment ; 
others can be more efficiently carried out with small-scale equipment. 
For example, continuous proces.=; industries, such as steel, chemical, 
cement, flour, etc., are so organized that highly specialized, large- 
capacity handling equipment is generally extremely effective in reduc- 
ing costs per unit of output. Such equipment requires large capital 
outlays and its costliness in turn gives impetus to concentration in 
order to achieve maximum use. On the other hand, the efficient han- 
dling of materials in a clothing factory, knitting mill, or job machine 
shop can usually be performed by a small gravity or belt motor, stor- 
age conveyor, or industrial truck. Such handling equipment involves 
relatively small capital outlays and imparts little impetus to 
concentration. 

In some fields of production, highly flexible and adaptable tech- 
niques have been developed which place a premium upon small-scale 
rather than large-scale equipment. In the southern lumber industry, 
for example, the introduction of mobile and flexible logging equip- 
ment, such as tractors, trucks, and gasoline skidders, has made it 
possible to obtain timber from stands which could not have been 
logged profitably with the old type of larger and less flexible 
equipment. 

The operation of small mills on timber from second-growth stands or scattered 
old-growth stands has been made more profitable by the adoption of flexible 
mechanical logging equipment, particularly trucks. * * * The use of trucks 
as a substitute for or as a supplement to railroads and the use of tractors and 
gasoline donkeys in place of heavji steam cable yarders have increased the chances 
for profitable operation of a system of selection logging. 

The efficiency of these small-scale techniques has resulted in a re- 
placement of old-type mills with large-scale equipment by smaller 
mills. This replacement has proceeded farthest in pine lumber pro- 
duction, and to a lesser extent in hardwood production.** 

The use of the most efficient techniques of production in the southern 
lumber industry apparently accelerates decentralization rather than 
concentration. This is true for a limited number of resource-extract- 
ing industries in which the receding of the resources places the large 
but stationary producing unit at an obvious disadvantage. 

On the other hand, a number of diversified industries exemjilify the 
way in which the greater efficiency of large-scale equipment has con- 
tributed to the growth of economic concentration. 



'P. SnrKnnt Floronrp, The rngic of Industrial Organization. Kegan Paul, Trench, 
Tnibner & Co.. Ltd., London. lO.'^S. pp. 11-20. 

"Work Prolects Administration. National Reseaich Project. Mechanization in the Lumber 
Industry, by A. J. Van Tassel and D. W. Blucstone. 1940. pp. 105-107. 



CONCENTRATION OF ECONOMIC POWER 199 

(1) In the- iron and steel indiistiy, for example, the automatic strip 
mills (described elsewhere in this report)' represent the most efficient 
type of rolling mill technique. 

These expensive mills, which smaller companies operating the obsolete hand- 
mills cannot afford to install, are further concentrating the control of steel pro- 
ducing facilities in the hands of fewer and larger companies. In the steel Indus 
try at present there are 18 small independent companies with obsolete hand- 
mills. * * * These companies employ a total of 23,350. Their combined 
capacity for flat rolled products is 2,350,000 tons, or 15 percent of the industry's 
hand mill capacity in 1929 for plates, sheets, black plate, or tin plate. 

Daily they are losing business to the strip mill producers, and before long their 
entire business will have been gobbled up by the huge strip producers, and these 
small independent companies will have closed their doors for all time. * * * 

These companies are doomed, because their obsolete mills depend, in the main, 
upon manual power ; while the automatic strip mills derive their power primarily 
from electricity. The difference in the cost of production is fatal to the smaller 
companies. Men cannot compete against electricity. * * * 

The economic effects of the strip mills, therefore, are the elimination of the 
smaller companies and the further concentration of steel-producing facilities in 
the hands of the few large steel companies. The new steel technology is accen- 
tuating monopoly in the steel industry.^ 

The large units and elaborate equipment required in the iron and 
steel industry involve heavy capital outlays. 

A modern blast furnace of about 1.000 tons capacity * * * < ^virh ordinary) 
auxiliary equipment * * * costs four to five million dollars. The average 
investment required for a modern steel works of eflicient size is approximately 
$100,000,000. Such a mill would be capable of producing about 1,000,000 tons of 
ingots per annum and would have diversified finishing equipment of suffi-,'ient 
capacity to convert about half the output into billets and other semi-finished steel 
and the other half into sheets and strip. Such an investment would not include 
operations prior to the assembly of raw materials at the plant site. i. e., the plant 
would be integrated only from coke plant to continuous rolling mills. Operating 
units may be and sometimes are much larger ; a single continuous hot and cold roll- 
ing finishing plant alone may require an investment of $60,000,000.° 

(2) In nonferrous metals the tendency is the same; the greater effi- 
ciency of large-scale equipment has been of the utmost importance in 
the concentration of economic power. Ore extraction has shifted mark- 
edly aAvay from small-scale operations on high-grade ore to mass-pro- 
duction techniques on low-grade ore, particularly in copper. Since 
many of the largest low-grade deposits lie close to the surface, the 
highly efficient open-cut or surface method of operation has been in- 
creasingly used. "The initial advantages of the surface mines were 
increased further by continuous technical advances, such as the evolu- 
tion of the large-size electrified power shovel and associated 
machinery." ^° 

The open-cut method has considerably reduced the cost of producing 
copper. Based on the 1926-30 average, the direct mining cost by the 
open-cut method was 2.26 cents per pound of recoverable copper com- 
pared with 4.10 cents by caving methods and 4.53 cents for all under- 
ground mines.^^ The introduction and use of the highly efficient open- 
cut process requires huge capital outlays which small producers cannot 



'See pp. 111-112. supra. 

* Hearlnss before the Temporary National Economic Committee, Part -SO. p. 16470. 

9 Hearings before the Temporary National Economic Committee. Part 26, Iron and Steel 
Industry. Exhibit 1410, Some Factors in the Pricing of Steel, prepared by the U. S. Steel 
Corporation. 

'" Worlts Progress Administration, National Research Project. Mineral Technology and 
Output Per Man Studies, Grade of Ore, by A. V. Corry and O. E. Kiessling, 1938, pp. 49-50. 

" Worli Projects Administration, National Research Project, Technology, Employment and 
Output Per Man in Copper Mining, 1940, p. 23. 



200 OOXCRXTRATIOX OF ECONOMIC POWER 

afford. Furtliermore, improved mining methods have been of much 
greater benefit to large producers of low-grade ores than to small oper- 
ators. During the twentieth century, therefore, the expansion of the 
copper-mining industry has coincided with the concentration of the 
productive facilities of the industry in the hands of three large pro- 
ducers, the Anaconda Copper Mining Co., the Kennecott Copper Cor- 
poration, and the Phelps Dodge Corporation, eachof which is inte- 
grated vertically from mine to consumer. The growth of economic 
concentration in coj^per mining is shown in the following table : ^- 



Company 


Percent of total mine 
production of copper 




1922 


1929 


1936 


Anaconda Copper Mining Co 


22.3 
4.7 
7.6 


20.6 
16.6 
9.1 


20 3 




31.9 


Phelps Dodge Corporation 


21.8 




34.6 


46.3 


74.0 







The shift from selective mining of high-grade ore to mass mining 
of low-grade, widely disseminatecl deposits was in part made possible 
by technological advances in a milling process. The development of 
the selective flotation process has in turn tended to increase economic 
concentration since the milling of non-ferrous metals in large central 
plants effects reductions in labor cost of from 4 to 15 cents per ton 
and at the same time increases the efficiency of metal recovery. "Thus 
better concentration, made possible by larger and more efficient mills, 
not only increases the operator's margin from ores of existing gracle, 
but also permits the more wide-spread working of lower yielding 
metal." " 

The greater efficiency of large plants in the milling of nonferrous 
metals is apparent in table 24, which compares man-hour requirements 
per ton in plants of different size. 

(3) The cement industry affords another example of the greater 
efficiency of large-capacity equipment. The rotary kiln has grown 
in average length from 25 feet in 1890 to 146 feet "in 1935. In 1925, 
56 of the 810 kilns in the industry were 200 feet or more in length. 
By 1935 there were 119 such kilns out of a total of 823. Today some 
of the most modern kilns are over 400 feet in length.^^ 

The growth in the capacity of kilns ha.'; tended to reduce capital investment 
per barrel of capacity. This has been brought about by the elimination of multi- 
plicity of feeding, driving, and fuel-preparing equipment and the reduction in 
floor area, foundations, and other items contributing to greater installation costs 
when a large number of small kilns are used. 

The growth in kiln capacity has also given rise to a number of operating 
economies. * * * Larger units do not require additional labor for their op- 
eration. Reduction in labor requirements has resulted also from a decrease in 
maintenance costs achieved through the use of superior refractory mate- 
rials ♦ * *. 

Nor should economies in the consumption of fuel be overlooked ; the longer the 
kiln, the more the materials can absorb heat from the products of combustion. 



"Ibid., p. 20. 

" WoikR Progress Administration. National Research Project, Mineral Technology and 
Output Per Man Studios. Grade of Ore. pp. 78-79. 

" Works Progress Administration, National Research Project, Fuel Efficiency in Cement 
Manufacture. 1909-3.5, 1938, p. 39 



CONCENTRATIOX OF ECONOMIC POWER 



201 



111 this way the temperature of the stack gases is lowered and the loss of heat 
through the stack is reduced. The final result is a diminution of fuel require- 
ments per barrel of cement. For example, the 60 to 80-foot kilns which were 
in general use until 1910 consumed about 100 to 150 pounds of coal per barrel 
of cement, whereas modern 250- to 300-foot kilns consume less than 100 pomids 
of coal per barrel, and kilns over 350 feet long consume as little as 70 to 80 
pounds of coal per barrel.^^ 

(4) In the brick and tile industry certain large-scale techniques 
have been of considerable importance in bringing about concentration. 
Among these is the autobrick machine which performs "mechanically 
the work of sanding the mold, striking off the excess clay, and bumping 
the mold to loosen the formed clay, so that the labor associated with 
its operation was about 50 percent of that required by the nonauto- 
niatic equipment.^*' It is significant that these machines are installed 
principally in the larger plants. 

Table 24. — Milling of nonferrous metals,^ differences in man-nour requirements 
per ion in plants of different size according to normal output 





7 tons an 
hour 


17 tons an 
hour 


40 tons an 
hour 


208 tons an 
hour 




0. 1040 
.0995 


0. 1015 
.0768 


0. 0374 
J. 0374 
.1496 
.0748 
.0374 


0. 0235 


Grinding 

Concentration, gravity 


.0141 
.0141 




.3125 
.0994 
.0957 
.0762 
.1180 


.0768 
.0708 
.0256 
.0512 
.1536 


.0094 


Dewateriug concentrates 


.0047 


Samplings 


.0047 




.0374 
.0374 




Miscellaneous ' 


.0361 






Totailabor.. 


.9053 


' 6135 


.4114 


.1066 



1 Table refers specifically to lead and zinc, but similarity of processes relates it also to copper. 

2 Includes 0.0187 for classification and screening. 

3 Includes supervision, maintenance, warehouse, power, etc. 
< Includes 0.0512 for weighing and loading. 

Source : Adapted from U. S. Bureau of Mines, Bulletin .381, Lead and Zinc Mining and 
Milling in the United States, 1935, p. 190. 

The National Research Project has made a comparison of unit 
labor requirements for small, medjum-sized, and large plants manu- 
facturing brick and tile. "Plants with a capacity of less than 15,000,- 
000 brick per year were classified as small, while those of 30,000,000 
brick or over were classified as large." The plants in the study were 
sei^arated by type of process — stiif mud, soft and other processes. 

For all processes combined the average standard man-hour ratios are highest 
for small plants, the average being 11.48, and lowest for large plants, the average 
being 6.87 * * *. in other words, 40 percent less labor is used to produce 
1,000 brick in large plants than in small plants; the same relationship holds 
for each process ; that is, the large plants have the smaller man-hour ratios, and 
the smaller plants have the larger man-hour ratios. There is one exception in 
soft-mud plants, when small- and medium-sized plants are compared, but this is 
doubtless due to the fact that there are so few plants in these classes." 

This comparison of man-hour requirements by size of plant is 
shown in table 25. 



^^ Work Projects Administration. National Research Project, Mechanization in the Cement 
Industry, 1939, p. 54. 

^^ Works Progress Administration, National Research Project. Productivitv and Employ- 
ment in Selected Industries, Brick and Tile, by Miriam E. West, 1939, p. 91. 

" Ibid., p. 120. 



202 CONCENTRATION OF ECONOMIC POWER 

Table 25. — Average standard man-hour ratios for sample hrick and tile plants, 
ly capacity and process ^ 





All plants 


Annual capacity (millions of common-brick 
equivalents) 


Process 




Less than 15 


15-29.99 


30 or over 




Ratio 


Num- 
ber of 
plants 


Ratio 


Num- 
ber of 
plants 


Ratio 


Num- 
ber of 
plants 


Ratio 


Num- 
ber of 
plants 




9.66 
9.67 
10.33 
8.27 


55 
9 
5 


11.48 
12.40 
10.74 
9.03 


22 

14 
4 
4 


9.82 
9.66 
12.73 
5.23 


29 

25 
3 

1 


6.87 
5^91 


18 




16 


Soft-mud 


2 


Sand-lime and dry-press 






1 Includes all plants for which standard man-hour ratios could be calculated. 

Source: Works Progress Administration, National Research Project, Productivity and Employment in 
Selected Industries, Brick and Tile, 1939, p. 120. 

Many more examples of industries could be given in which the 
greater efficiency of large-scale equipment has led to concentration 
because of the large capital outlays required to purchase such equip- 
ment.^^ Tliis type of cumulative evidence, however, should be sup- 
plemented by a statistical analysis for a larger number of industries 
of the relationship between technological efficiency and concentration. 
Such an examination is made in the next section. 



THE MEASURFMENT OF TECHNOLOGICAL EFFICIENCY 

One measure of the degree of efficiency in different sized plants in 
a given industry is the amount of electrical energy (either purchased 
or generated in the plant) consumed per man-hour. This measure 
is usable because the tendency, especially in recent years, in the crea- 
tion of new and improved mechanical techniques of production has 
been in the direction of developing those that are electrically pow- 
ered. From 1919-29 the increase in power capacity of American in- 
dustry was accounted for almost entirely by the increased use of 
electricity ; according to the Census of Manufactures : 



"For example, "In the flour-milling industry the cost per unit of capacity of a roller mill 
with an hourly capacity of 50 bushels is twice that of a, largor mill with a capacity of 115 
bushels." Reflecting the greater eflSciency of this large-scale equipment, the average size of 
roller mills in tlie flour-milling industry introduced in the year.s 1930-34, was approximately 
22 percent higlier than that of those introduced in 1920-24. "The cost of spiral screw con- 
veyors used in flour mills declines from $2.83 per unit of capacity (100 bushels per hour per 
foot), on small conveyors with a capacity of 60 bushels per hour per foot, to $0.23 as the 
capacity is raised to 1,000 bushels per hour per foot, and to $0.10 for conveyors of capacity 
of 4,500 bushels per hour per foot." (WorKs Progress Administration, National Research 
Project. Effects of Current and Prospective Technological Developments upon Capital 
Formation, p. 4.) 

In the field of power, "a report of the B^'ederal Power Commission shows that steam-power 
units install (1 in central stations during the period 1931-34 averaged over 30,000 kilowatts 
in capacity as against approximately 15,000 kilowatts per unit in 1931-35. The original 
investment jier unit of capacity is a great deal smaller for large steam-electric generating 
stations than for small ones, ranging from $135 to $150 per kilowatt for stations of a 
capacity of 2,000 kilowatts to $92 to $115 per kilowatt for stations of 200,000 kilowatts. 
The cost per unit of capacity in the case of boilers capable of producing 1.000.000 pounds of 
steam per hour is less than one-half that of a boiler with one-tenth that capacity. Likewise, 
in Diesel plants investment per unit of capacity varies inversely with the capacity, declining 
from $230 to $100 as the capacity increases from 100 to 10.000 kilowatts." (Ibid., pp. .3-4.) 
And in electric motors. "The price per lior.sepowcr of 500 revolutions per niiiuite Westing- 
house motors of the squirrel cage tyi>e declines from $9.30 for a molor of 2.">0 horsepower 
capacity to .$0.20 for a motor of 1.000 horsepower; for motors of the wound-rotor type the 
price declines from !t;i2.00 to $7.80 as the capacity is- tluis increased," (Ibid!, p. 4, 
footnote 11.) 



COXCE^•TTlATION OF ECONOMIC POWER 203 

The combined horsepower of steam engines and turbines, internal-combustion 
engines, and water wheels and turbines, for all industries, has not varied 
materially from census to census since and including that for 1919, whereas the 
power of electric motors has increased rapidly. In other words, during the past 
decade the increase in the aggregate primary jxtwer is accounted for almost 
exactly by the increase in the power of electric motors driven by purchased 
energy." 

The more modern and efficient types of technological equipment 
are primarily those driven by electrical power, and the amount of 
electrical energy per man-hour used by plants of varying sizes within 
a given industry indicates roughly the extent to which they use 
electrically-powered techniques of production.-" Granted the greater 
efficiency of machinery designed for electrical motivation and con- 
trol, the amount of electrical energy consimied per man-hour in 
plants of different size within a specific industry therefore affords 
a criterion of their relative efficiency. 

It sliould be pointed out that the calculation of electrical energy 
used per man-hour does not indicate the relative efficiency of one 
industry compared with another. Nor does it indicate that any 
particular size of plant has more electrically-powered machinery 
than plants ot other sizes. Its purpose is to relate the use of elec- 
crically-powered equipment to the unit of labor expenditure, the 
man-hour, for plants of varying size in the same industry. 

The Bureau of Labor Statistics has calculated the amount of elec- 
trical energy used per man-hour for 105 manufacturing industries 
from data gathered by the 1937 Census of Manufactures.-^ Total 
employment in those 105 industries amounted to 4,161,934 wage- 
earners in 1937 compared with a total for all manufacturing indus- 
tries of 8,569,231 wage-earners. The industries were widely scat- 
tered among the major fields of food and tobacco, textiles, chemicals, 
stone, clay, and glass, and metals. For many of the industries, break- 
downs by size of plant were made. Thus it is possible to make com- 
parisons of electrical energy used per man-hour in plants of varying 
size. 

Chart XV (table 26) graphically shows that in 21 diversified in- 
dustries electric energy consumed per man-hour — wliich may be 
assumed to vary directly with the use of the more modern produc- 
tive techni(][ues — increased generally with the size of the plants. 



" U. S. Bureau of the Census, Census of Manufactures, 1929, vol. 1, pp. 8-9, 

For the years 1929-37 it is of interest to note that whereas in the latter year Industrial 
production as reported by the Federal Reserve Board was 7.6 percent below the level of 
the former year sales of electrical energy to "Commercial lar.ge light and power" users 
(excluding municipal street lighting, etc., and electric railways), as reported in the 
1940 Annual Statistical Number of Electrical World, were higher by 24.6 percent. 

20 "The * * * present-day period is characterized by the increasing fusion of the 
electric motor with the manufacturing equipment, by the creation of an electrical machine 
in the true sense of the term. Tlie electric motor is no longer an outsider as regards the 
manufacturing machine to which it supplies the motive power. It penetrates into the 
parts of the machine, becoming an inseparable part of it, constructively combined with it, 
directly affecting the character of the machine and of the production processes. Thus the 
whole of the production process becomes, as it were, the function of the electric motor." 
(Walter N. Polakov, The Power Age. Covici-Friede. New Yorl;. 19.'58. p. 101.) 

=^ V. S. Bureau of the Census. Census of Manufactures, 1937, Man-Hour Statistics for 
105 Selected Industries. All other data in this section relative to electric energy per 
man-hour are derived from this source. 



204 



CONCENTRATION OF ECONOMIC POWER 



Chart XV 



RELATIONSHIP BETWEEN SIZE OF PLANT 

AND ELECTRIC ENERGY USED (KWH) 

PER MAN-HOUR 

21 MANUFACTURING INDUSTRIES 
,„.„, UNITED STATES, 1937 



■ ' 


FLOUR a MILLING 




ll 








; 






, 1 1 








/ 






■ i 








/I 
/ 1 






-!i 




L_/ 


^ 












/ 








1 


l\ 










Ul 


J 












/ 

































GLASS 

i ; , 










, 








1 i 


1 



INSEO. . 'IDES a FUNGICIDES 



1^ I 



'-.[]] 









A 


1 




^ 1 


\ 






MACHINE TOOLS I MACHINE SHOP PRODUCTS 




MOTOR VEHICLE BODIES 

I I I 
I 



SIZE OF PLANT BY NUMBER OF WACE EARNERS 



Source : Table 26. 



COXCENTKATIO.X OF ECONOMIC rOWER 



205 



Table 26. — Electric cneryy used, kilou-att-hoiirs per man-hour, ly size of plant, 
21 nuniiifacturing industries, 1937 





Number of wage earners per plant 


Industry 


1-5 


6-20 


21-50 


51-100 


101-500 


501-2,500 


Over 
2,500 


Flour and other grain-mill product s 


3.14 
1.35 
4.08 
3.25 
.75 


5.56 
1.49 
4.44 
3.84 
2.23 
2 1.98 

.85 
2.23 

1.13 
1.41 
.94 
1.13 

.84 
.73 

1.01 
.99 
1.15 

.74 

.56 
.53 


10.82 
2.82 
8.29 
4.90 
3.94 
2.25 
1.39 

1.44 

1.23 

1.53 
1.28 
l.ll 
.99 

.90 
1.02 

1.35 
.96 
1.36 

.85 
.68 


11.77 
3.80 
7.93 
5.40 
3.84 
1.92 
1.23 

1.94 

2.25 

2.28 
2.96 
2.32 
1.37 

1.16 
1.06 

2.48 
1.00 
1.76 

1.30 

1.10 
.61 


117.11 
7.39 
13.60 
7.84 
5.92 
3.66 
2.32 

3.29 

4.64 

1.75 

4.45 

1 1.70 

1.58 

1.71 
1 1. 56 

1.86 
1.30 
2.46 

1.61 

1.66 
1.00 







Cereal preparations . 


9.53 
17.29 






20.20 


Feeds, prepared, for animals and fowls 


Lime - . _ 






Glass 


3 6.86 
3 5.16 

5.73 

4.40 




Aluminum products 


.67 
.81 
.26 

.65 
.80 
.78 
1.29 

1.05 
.62 

1.61 
.62 
.98 

1.10 

.55 
.63 




Nonferrous-metal alloys; nonforrous- 
metal products, except aluminum 

Wrought pipe, welded and heavy- 
riveted' = 


5.34 


Insecticides and fungicides, and indus- 
trial and household chemical corn- 




Soap - . 


3.77 








Tin cans and other tinware 


1.66 
2.75 




Stamped and pressed metal products; 
enameling, japanning, and lacquering. . . 




Wircwork 




Electrical machinery, apparatus, and 
supplies 


3.12 

3 1.41 

3 2.66 

3 2.27 

3.34 
1.94 


3 53 


Machine tools. - . - 








Engines, turbines, water wheels, and 
windmills 




Refrigerators and refrigerating and ice- 
making apparatus 


2 78 


Motor-vehicle bodies 


2 01 







1 From 101 to 2,500 wage earners. 

2 From 1 to 20 wage earners. 

3 More than 500 wage earners. 

* Made in plants not operated in connection with rolling mills. 

Source: U. S. Bureau of the Census, Census of Manufactures; 1937 
Industries, table 3, dd. 11-75. 



Man-Hour Statistics for 105 Selected 



In certain of the industries (for example, petroleum refining) the 
trend in the amount of electric energy used i^er man-hour is strik- 
ingly upward but in othei"S (for example motor vehicle bodies) the 
increase is more gradual. In the latter group, however, the nature 
of the productive process is such that the amount of electric energy 
used per man-hour is relatively small, and therefore the actual in- 
crease Mould inevitably be less precipitous. But in teniis of per- 
centage, the large plants within an industry in this latter group 
may have a decided advantage over the smaller plants. Thus the 
difference in the amount of electric energy used per man-hour by the 
largest and smallest plant size groups in the motor vehicle bodies 
industry is only 1.38 kilowatt-hours, but by percentage it is 219.^2 

A further indication of the advantages held by the plants which 
used the greatest amount of electric energy per man-hour is the fact 
tliat industries in which the largest plants used the most electric 
energy also employed the, bulk of the labor force. 



2= It should not hp forgotten tb.at this advantage refers to the production and not the 
distribution of goods. 



206 



CONCENTRATION OF ECONOMIC POWER 



Numier of loagc-ean^ers in industries in which electric energy used per man-hour 

is highest 



In plants employing- 



1 to 50 wage-earners 

51 to 100 wage-earners... 
101 to 500 wage-earners. 

601 and over 

Total .- 



Employ- 
ment 23 



347, 127 

80, 522 

365, 119 

3, 251, 525 



4,044,293 



Percent 
of total 



100.0 



23 The total figure does not include 117,641 wage-earners employed in industries for which comparable 
size groupings are not available. In order to include every possible industry in the group of the smallest 
plants, 1 to 50 wage earners, selections were made from the extremely narrow component groups— 1 to 5, 
6 to 20, and 21 to 50 wage-earners. The resultant inclusion of certain industries (e. g., radio, radio tubes 
and phonographs) in the 1 to 50 group, due to the smallness of the component segments and the consequent 
greater possibility of unusual situations applying thereto, still did not result in any noticeably large figure 
for the 1 to 60 wage-earner group. For the complete tabulation, see appendi.x I. 

If the amount of electric energy used per man-hour indicates 
roughly the degree of utilization of modern technological equipment 
in producing plants, a correlation within given industries of the 
percentage difference (power differential) in electrical energy used 
per man-hour between large and small plants with the degree of 
economic concentration would indicate whether or not the extensive 
use of modern technological processes is related to the concentration 
of economic ppwer. Such a correlation is shown graphically for 30 
diversified manufacturing industries in chart XVI (table 27). 

Chart XVI 



THE RELATIONSHIP BETWEEN 
ECONOMIC CONCENTRATION and POWER DIFFERENTIAL 

Thirty Monufocfuring Industries 




CONCENTRATION OF ECONOMIC POWER 



207 



The methodology involved in correlating the power differential be- 
tween large and small plants with the degree of economic concentra- 
tion in specific industries is summarized in appendix J. The co- 
efficient of correlation is 0.83, 2* Considering the number of variables 
involved in each of the factors, the relationship between the two 
is comparatively high. 

Table 27. — Relationship ieticeen economic concentration and power differential 
(large plants over small plants) 

30 MANUFACTURING INDUSTRIES 



Industry 



Eco- 
nomic 
concen- 
tration ' 



Power 
differen- 
tial 2 



Aluminum products ..- 

Soap 

Motor vehicle bodies and parts 

Cereal preparations , 

Distilled liquors. 

Wrought pipe, welded and heavy riveted 

Glass 

Electrical machinery apparatus and supplies 

Smelting and refining, nonferrous metal other than gold, silver, and platinum 

Cigars .- 

Petroleum refining _-- 

Cast-iron pipe and fittings _._ 

Steel barrels, kegs, and drums.. _ 

Cement 

Flour and other grain mill products 

Engines, turbines, water wheels, and wind mills _ 

Radios, radio tubes, and phonographs _ 

Drugs and medicines . _ 

Pulp (wood and other fiber) .._. 

Feeds, prepared, for animals and fowls 

W ire work 

Lime -._ - 

Machine tool accessories, etc _ _ 

Clay products, other than pottery 

Pottery, including porcelain ware 

Rubber goods, other than tires, tubes, and boots and shoes 

Paper 

Machine tools 

Insecticides and fungicides, etc . _. 

Malt liquors .- 



Percent 
76.0 
73.5 

67^0 
51.0 
47.0 
41. 9 
44 4 
41.5 
.38.5 
38.2 
37.6 
37.0 
29.2 
29.1 
28.9 
27.0 
23.4 
22.7 
21.7 
21.7 
21.6 
20.9 
19.2 
18.7 
18.5 
13.8 
13.3 
12.7 
11.8 



Percent 
432.0 
283.6 
227.9 
404.2 
188.9 
274.2 
223. 6 
167. 4 

176! 9 
260.7 
281.7 
80.9 
29.6 
162.8 
160.9 
29.1 
157.9 
6.5 
96.0 
97.5 
156. 3 
83.3 
50.0 
114.8 
59.1 
99.7 
64.0 



' Percent of industry's value of products produced by its 4 largest concerns, 1935. 

' Percent difference between large plants and small plants in use of electric energy per man-hour, 1937. 

Source- Economic concentration. National Resources Committee, The Structure of the American Econ- 
omy, pt. I, 1939, appendix 7, table I, pp. 240-249. Power differential, computed from U. S Bureau of the 
Census, Census of Manufactures: 1937, Man-Hour Statistics for 105 Selected Industries, table 3, pp. 11-75. 

The tendency for an increase in the power differential (or advan- 
tage) of large plants over small plants to be accompanied by an 
increase in the degree of economic concentration could mean that (1) 
the relationship is due merely to chance, or (2) an increase in con- 
centration has resulted in an increase in the power differential, or 
(3) an increase in the power differential has resulted in an increase 
in concentration. 

The first possibility may be eliminated because in certain industries 
a relationship between the relative efficiency of large over small plants 
and the degree of economic concentration is known to exist. The 
second implies that once a high degree of economic concentration in 
some way has been established, the power differential would be raised. 
But granting that this might have occurred, the power differential. 



24 The .standard error : 10.0833 percent ; the line of regression : Yc = 12.0243 -f0.1426134x. 



208 CONCENTKATIOX OF ECONOMIC POWER 

thus raised, wouid increase the advantage of the hirge phmts and 
make ])ossible the maintenance of the high degree of concentration 
established. 

Regardless of whether the establishment of a high degree of con- 
centration has increased the i)ower differential or whether an increase 
in the power differential has resulted in greater concentration, it is 
evident that a primary causal factor of concentration is the existence 
of an advantage in operating efficiency of large plants over small 
plants. 

CONCENTRATIOX AND INDUSTRIAL RESEARCH 

THE GROWTH OF INDUSTRIAL RESEARCH 

Technological development depends on continual invention, and 
invention today stems increasingly from large industrial research or- 
ganizations. This is a consequence of the widening of tlie body of 
industrial knowledge, which has made it impossible for any one nuin, 
or even any small group of men, to master the innumerable technical 
details in each of the scientific fields involved in the inventive effort. 

Groups of specialists w^orking in large inchiistrial laboratories, 
owned l3y vast corporate organizations, are replacing the individual 
isolated inventor. As Dr. Charles F. Kettering, vice president in 
charge of research. General INIotors Corporation, has said, we are 
passing through "the transition period frt)ni the individual as an 
inventor to the group as an inventor * * * group invention 
rather than individual invention." -^ This transition is anothei- mani- 
festation of a basic technological process — the specialization of 
function. 

Industrial research on an organized group basis has come to occupy 
a position of real importance in the economy aside from the tech- 
nological advances which have emanated from it. "Measured in 
terms of employment, the 'research industry' is equal in importance 
to the dyeing and finishing of cotton fabrics which in 1987 ranked 
among the 45 manufacturing industries which provided the largest 
number of jobs." "" 

Here, the significance of industrial research is its relation to the 
concentration of economic power. If industrial research were to 
accelerate the process of concentration, then technology in a second 
way would impart an impetus to the concentration of economic power. 

industrial research has grown extraordinarily during the past two 
decades. Between 1921 and 1938 research personnel in tliis field 
rose by approximately 300 percent.-^ Another indication of the 
growth of industrial research is the decrease in tlie number of re- 
search workers employed on a part-time basis. In 1927 approxi- 
mately 25 percent of the research personnel re])ortedly worked on 
a part-time basis; by 1938 the proportion had fallen to 3 percent.-® 



2^ Hearinsrs before tlie Temporary National Kcononiic Committee, I'art "0. p. 1C>2!i:V 
/ so Work Projects Administration. National Uesearcli Project. Industrial Research and 
Clif;ncing Technology, by George Perazich and I'. M. Field, 1!)40. p. 40. 

"Ibid., p. 7. In gathering the jiriniary data for this svu'vey the National Research 
Council stated that it "is not intended to list laboratories which are concerned merely 
with routine testing of raw materials and products, but it is desired to include all those 
which definitely devote some time to research looking toward the improvement and 
development of products." (Ibid., p. -.) 

^'Ihid., p. G. 



CONCENTRATION OF ECONOMIC POWER 209 

The increase in research personnel in 22 industrial groups during 
this period is shown in table 28. With but 6 exceptions research per- 
sonnel has grown by more than 100 percent; in the radio apparatus 
and phonographs and petroleum products industries the increase has 
been almost phenomenal in this 11-year period. 

INDUSTPJAL RESEARCH AND THE COMPETITIVE SYSTEM 

To determine whether or not increased industrial research has 
accelerated the concentration of economic power, it is necessary to 
analyze the role industrial research plays in the competitive system. 
Competition in a given industry may take three forms: (1) reduc- 
tions in price, (2) improvements in quality, or (3) the addition of 
new features, parts, or gadgets to the product. 

Table 28. — Growth of Reseakch Personnex, fkom--1927 to 193S, by Industrial 

Group 

Percentage in- 
crease, 1921-38 

Industrial group: 

Chemicals and allied products 175. 5 

Petroleum and its products 538.7 

Electrical communication , _ 3. 7 

Electrical machinery, apparatus, and supplies 58. 1 

Consulting and testing laboratories 127. 

All other- machinery 133. 2 

Rubber products 101. 8 

Motor vehicles, bodies, and parts 190. 2 

Agricultural implements (including tractors) 184.7 

Miscellaneous 205. 

Iron and steel and their products, not including machinery 189. -1 

Food and kindred products : 255. 1 

Stone, clay, and glass products 166. 4 

Nonferrous metals and their products 72. 7 

Radio apparatus and phonographs 1, 600. 

Utilities (gas, light, and power) 89.4 

Paper and allied products 177. 5 

Trade associations 96. 9 

Textiles and their products 252. 9 

Forest products ^ 190. 9 

All other transportation equipment 17. 

Leather and its manufactures 151. 6 

Total 133. 3 

Source : Work Projects Administration, National Research Project, Industrial 
Research and Changing Technology, 1940, p. 19. 

Industrial research plays an important, if not a dominant, role in 
each of these three forms of competition. For the producer engaged 
in intense' price competition for whom reductions in cost are of pri- 
mary importance, for the producer whose industry is characterized 
by a steady improvement in quality, and for the producer whose 
buyers have come to expect new features, the lack of a research 
organization to perform these functions might well be the cause of 
financial failure and disappearance from the industry. 

Also industrial research may lead to concentration by drawing a 
large concern into the production of entirely new and different goods. 
Frequently a large company devises a completely new commodity 
or improves materially upon some relatively unused product through 
industrial research. Then because it has adequate financial resources 

277551— 41— No. 22 15 



210 CONCENTRATION OF E<JONOMIC POWER 

and the technical equipment necessary to produce the article, the large 
concern comes to dominate a substantial proportion of the expanding 
market for the new commodity.^^ 

In the process, costs may be reduced so that the price of the new 
item may be lowered and its market thus increased; quality may be 
improved; or new features may be added. But in each case, the 
achievement of these objectives is greatly facilitated by an industrial 
research organization. If large concerns have a marked advantage 
over small finns in the possession of industrial research organizations, 
the chances are considerably increased that the production of new 
commodities will be dominated by a few large companies. 

The electric refrigerator industry is an interesting example of the 
way in which the production of a relatively new product has been 
dominated by a few large concerns, three of which were among the 
Nation's gigantic industrial enterprises long before their entry into 
the manufacture of electric refrigerators. In 1935 the percent of the 
industry's output of various sized domestic electric refrigerators 
produced by the industry's four largest concerns [presumably Gen- 
eral Motors (Frigidaire), General Electric, Westinghouse, and Kel- 
vinator-Nash (Kelvinator)] was as follows: 



Capacity 



Under 6 cubic feet : 

Number 

Value -- 

6 to under 10 cubic feet: 

Number 

Value_ 

10 cubic feet and over: 

Number 

Value - 



Production 
of first 4 



617,037 
$44,115,690 



348, 762 
$36, 904, 509 



11,068 
$2, 273, 002 



Percent 
of total 



70.5 
72.4 



65.0 
66.9 



94.7 



' Temporary National Economic Committee Monograph No. 27, The Structure of Industry, Part V, 
"The Concentration of Production in Manufacturing," appendix C. 

While it is impossible to ascertain the exact role of industrial 
research in the establishment of this high degree of concentration. 
General Motors, General Electric, and Westinghouse are noted for 
the size of their research organizations. The three major objec- 
tives of industrial research — reductions in cost, improvements in 
quality, and development of new features — have been pursued vig- 
orously by the research laboratories of these companies, and their 
constant attainment have been of primary importance in the ex- 
pansion of the market for electric refrigerators.^'^ 

This example Avould indicate that the concern with an industrial 
research organization is at a definite advantage in the competitive 
conflict as well as in the development of new fields. 

29 In the annual report of E. I. du Pont de Nemours & Co. of 1937 it is found that products 
relatively unknown in 1929 accounted for approximately 40 jwroent of their total .sales that 
year (Hearings before the Temporary National Economic Committee, Part 30, p. 1G241). 

3" The General Motors Research Corporation, the General Electric Co., and the Westing- 
house Electric & Manufacturing Co. are among the list of 45 companies reporting the largest 
research staffs in 1936. (See Work Projects Administration, National Research Project, 
Industrial Research and Changing Technology, p. 68.) 



CONCENTRATION OF ECONOMIC POWER 211 

THE OWNERSHIP OF INDUSTRIAL RESEARCH 

The possession of industrial research establishments is more char- 
acteristic of large firms than of small firms because of the com- 
plexity of organization and consequent expense of modern indus- 
trial research. For example, when he was before the Temporary 
National Economic Committee, Mr. Kettering described the way 
in which the research organization of General Motors is divided 
into several broad clepartments and tliese in turn are broken up 
into divisions according to fields of analysis. Thus there is a de- 
partment of chemistry, within which are a fuel research division, 
a rubber research division, a plating research division, etc. There 
is a department of physics within which are an X-ray division, a 
high-voltage division, an ignition division, a spectroscope division, 
etc. There is a department of design within which are divisions 
of designing, of building, and of testing innovations. Then there 
are segments within the various divisions. For example, the de- 
partment of chemistry has a materials division, of which a metal- 
lurgical unit is a component part.^^ Research pursued on such a scale 
is obviously costly. Mr. Kettering estimated that the total engineer- 
ing and research budgets of the General Motors Corporation would 
amount to approximately ten or twelve million dollars a year. 

In elaboration of this point, the Xational Research Project 
observes : 

In order that the full advantages of large-scale research may be realized, 
large capital outlays are required both for equipping the laboratory and for 
its operating expenses. When it is considered that a laboratory of 20 em- 
ployees (including scientists and their assistants) requires perhaps $75,000 
annually for operating expenses alone, *it is' evident that only corporations 
with large capital assets would be able to maintain a laboratory of this 
relatively modest size. If it is assumed that 1 percent of the total volume 
of sales is allocated to research a concern with a research budget of $75,000 
would have to produce each year goods or services valued at $7,500,000. 
Some of the largest laboratories, in which the advantages of systematically 
organized research are most fully realized, have annual budgets of several 
million dollars. In some industries, moreover, certain research projects may 
continue for several years, involving correspondingly large expenditures, before 
the results become apparent, and in many cases no fruitful results are obtained."" 

Industrial research is highly concentrated; there is probably no 
other basic function of g'eneral economic activity so dominated by 
a few enormous concerns. The National Research Project found 
tliat "13 companies with the largest research staffs, representing less 
than 1 percent of all companies reporting in the National Research 
Council survey, employed in 1938 one-third of all research workers, 
or as many as the 1,583 companies with the smallest research staffs." 
Half of the country's industrial laboratory personnel was employed 
by only 45 large research laboratories, "aU'but 9 of which are owned 
or controlled by companies which are among the Nation's 200 leading 
nonfinancial corporations." ^^ 

In individual industry groups the concentration was just as pro- 
nounced. In 1938 that quarter of the companies having the largest 
research staffs employed 59.3 percent of all research workers in blast 

^ Ile.Triiigs before the Temporary Xational Economic Committee, Part 30. p. 16294. 
3- Work I'lojects Administration. National Research Project, Industrial Research and 
Changing Technology, p. 4.5. 
33 Ibid., pp. 9-11. 



212 CONCENTRATION OF ECONOMIC POWER 

furnaces, steel works, and rolling mills; 82.2 percent of all research 
workers employed in electrical machinery, apparatus and supplies; 
88.3 percent in industrial chemicals; 89,0 percent in motor vehicles, 
bodies and parts ; 85.0 percent in petroleum ; 82.8 percent in radio ap- 
paratus and phonographs; 90.0 percent in rubber products; 56.9 
percent in textiles and their products; and 78.5 percent in utilities 
(gas, light, and power). ^ 

Perhaps the most important aspect of this concentration of in- 
dustrial research is that — 

By and large, the mass of specialized data assembled in the course of industrial 
research does not become available to anyone except the owners of the labora- 
tory. Although a few large concerns which carry on extensive studies in funda- 
mental sciences frequently publish the results of their experiments, this is not 
typical of industrial research. The proportion of total findings published by 
concerns having industrial laboratories is only a fraction of the proportion pub- 
lished by academic and governmental laboratories. Even in cases where per- 
mission is given to individual scientists working in industrial laboratories to 
publish their experimental findings, the management usually reserves the right 
to examine such technical papers, modify their contents, and approve their 
publication only when it finds that this would not jeopardize its competitive 
position.*' 

Since industrial research is much more characteristic of large than 
of small concerns, and since it gives a competitive advantage to those 
firms able to pursue it, technology in this second way imparts impetus 
lo the concentration of economic power. 

CONCENTRATION AND PATENTS 

Though technological advance does not depend upon patents, the 
granting of a monopoly by the State on a new product or process 
for a given period of time has been closely associated with this advance 
during recent times. There is no evidence that patents are a necessary 
inducement to technological development, for all during the industrial 
revolution, when widesweeping and fundamental technological 
changes took place, patent controls were of only limited importance. 
As a matter of fact, from 1793 to 1836 there existed in the United 
States only a registration system by which anyone who swore to the 
originality of his invention and paid the stipulated fee could secure 
a patent, the validity of which might later be determined by the 
courts. Not until 1836 were the fundamental features of our present 
patent system, particularly the policy of examining applications, 
established as part of the law of the land. 

Patents today are granted in the United States on a machine, a 
product of manufacture or composition of matter, an art or process of 
making it, upon the appearance or ornamental character of an article, 
and on plants that are asexually reproduced. Scientific principles, 
scientific discoveries fer se or the laws of nature cannot be patented. 



*• Ibid., p. 10. It may bo significant that in the least concentrated of these industry 
groups — t'xtiles and their products — that quarter of the companies having the largest re- 
search staffs employed a smaller percentage of the group's total number of researcli workers 
than was the case for any of the other industrial groups ; on the other hand, in the 2 
most highly concentrated groups — rubber products and motor vehicles, bodies and parts — 
that quarter of the companies having the largest research staffs employed higher per- 
centages of the total number of research workers than in any of the other indii.strial 
groups. 

^■'Ibid., p. 47, footnote 13. One of the primary reasons for this protection of findings is 
the fact that they may be patentable, a consideration to be discussed more fully in the 
following section. 



CONCENTRATION OF E<X^NOMIC POWER 213 

Patents extend for 17 years from the date of granting, except in the 
case of design patents which run 314, 7, or 14 years, as the inventor 
elects. The sole right to use "a machine, a product of manufacture 
or composition, an art or process of making it," for an extended period 
of time obviously entails excellent potentialities for the centralizing of 
economic power. 

TECHNIQUES OF CONTROL THROUGH PATENTS 

Patents have been used in a wide variety of ways to establish a 
high degree of concentration of economic power over an industry. 
But the power exerted is frequently far greater than might be in- 
ferred from the existence of a 17-year legal monopoly. 

There are two principal reasons for this: (1) Control may be 
extended well beyond the termination of the basic patent by means 
of so-called improvement or indirect patents. The operation of the 
basic equipment without these patented improvements is often im- 
practicable, and thus control may be extended additional years for 
each improvement patent. (2) The life of a patent affords the com- 
pany controlling it a considerable amount of time in which to become 
entrenched in a position of dominant economic power. During this 
period its monopoly income may enable it to engage in activities to 
secure the field to itself after the patent control is legally ended. The 
company may purchase potential rivals outright or buy their patents ; 
it may contract for the services of the leading inventive talent in the 
field ; it may bring expensive lawsuits upon so-called infringers, the 
expense of which in most instances can much more readily be borne 
by the established firm than by the accused companies ; it may pursue 
policies of predatory price cutting, designed to force out of business 
those firms with less financial staying-power. In short, it may accu- 
mulate during the period of its patent control such financial strength 
that it can vigorously pursue any of the numerous policies designed 
to control the extent of competitors' operations, or to force competi- 
tors completely out of business. 

It would be impossible in this brief survey to describe all of the 
numerous ways in which patents have been used to establish economic 
concentration. Several specific cases, illustrative of particular tech- 
niques of effecting economic control through patents, will be briefly 
analyzed. 

A partial list of these techniques includes the use of patents to 
establish command over the production of a given commodity (1) 
by compelling members of an industry to lease patented equipment; 
(2) by dividing territory and allocating geographic markets among 
several producers; (3) by determining the type and amount of goods 
to be produced with patented equipment; and (4) by forcing licensees 
of a given commodity to follow the pricing and marketing policies 
of patent-holding producers in an industry. 

The United Shoe Machinery Co. is a frequently cited example of 
control exerted over an industry by means of patents. This company 
was founded during the great era of consolidations in 1899, and was 
an amalgamation of seven concerns. It acquired all the assets of 
these firms, including the vital patents of the Consolidated & McKay 
Lasting Machine Co. on lasting machines, and the Goodyear Shoe 



214 CO^'CENTRATIO^!' OF ECONOMIC POWER 

Machinery Co, patents on welt-sewing and outsole-stitching machines. 
It subsequently acquired the shoe machinery business and assets, 
including patents, of 57 individuals, partnerships, and corporations. 

But the acquisition of concerns engaged in the manufacture of shoe 
machinery was not considered sufficient to obtain complete patent 
control over the industry because innovations were constantly being 
made by professional inventors in the field and by the United Co.'s 
own employees. Therefore, it contracted with 95 percent of the 
inventors of shoe machinery for all their inventive efforts and caused 
its employees to assign to the company all inventions relating to shoe 
machinery which they might devise while in the service of the com- 
pany. The control of these two principal sources of invention in the 
field has been extremely valuable to the United in maintaining its 
practically complete monopoly on shoe machinery patents. 

In its arrangements with shoe manufacturers, the United Shoe 
Machinery Co. has leased rather than sold its equipment, thereby 
retaining title to the machines and the power to withdraw them 
whenever it believed its patent rights were infringed. Furthermore, 
it has leased its machines subject to provisions which forced the shoe 
manufacturer to obtain either all or none of the shoe machinery 
needed in his factory from, the company. 

In order to compete effectively with the United, would-be rivals had 
to develop a new and complete line of shoe machinery. This was 
attempted by Thomas J. Plant who developed a complete line of shoe 
machines for manufacturing welt, turn, and McKay shoes, with which 
he equipped a factory in 1910. Confronted with this competition, the 
United proceeded to acquire the Plant assets, most important of which 
were the Plant patents — 206 granted by the United States and 259 by 
various foreign lands. 

Consequently, the United Shoe Machinery Co. came to control more 
than 95 percent of the entire shoe machinery business in the United 
States and to constitute the only firm which produces a full line of shoe 
machinery.^^ 

In 1940 the corporation owned and controlled patents and inventions 
covering more than 300 types of shoe machinery — the leasing of which 
constituted the larger proportion of the company's net income." 

The General Electric Co. for many years has been the largest organi- 
zation in the world engaged in the development of apparatus and 
equipment for generating and transmitting electrical power, and the 
manufacture of electrical appliances. It produces approximately 20 
percent of all electrical equipment sold in the United States, or about 
twice the amount made by its largest domestic competitor, the West- 
inghouse Electric & Manufacturing Co.^^ 

In specific items, its dominance is much greater. This is particularly 
true of the manufacture of electric lamp bulbs (i. e., large type bulbs 
for general household and commercial uses) . As recently as 1917 bulbs 
were manufactured largely by manual processes, and at that time there 
were five manufacturers of bulbs in the country. By 1939 the process 
of manufacture was almost entirely mechanized, and the number of 



3« Floyd L. Vaughan, Economics of Our Patent System, Macmillan, New York, 1925, pp. 
79-85. 

*' Standard Statistics Co., Inc.. Standard Individual CorporatioD Descriptions, 1940, 
"United Shoe Machinory Company." 

^ Ibid., "General Electric Co." 



CONCENTRATION OF ECONOMIC POWER 215 

producers reduced to two — General Electric and Westingliouse. This 
increase in concentration developed as the most economical produc- 
tion is possible only in highly mechanized plants of very large capac- 
ity and because control over production is exercised by the owners of 
patents. 

The General Electric Co. is one of the two owners of such patents and 
produces both glass bulbs and the finished lamps. The Corning Glass 
Works owns the patents on the glass bulbs fer se and supplies bulbs 
to all other American electric-lamp manufacturers except General 
Electric. 

The absence of competition in the manufacture and sale of bulbs is 
due principally to the control over patents exercised by these two firms. 
Together they own or control not only all of tlie product patents, that 
is, those relating to bulbs, tubing, etc., but also the patents on the 
machinery involved in the manufacture of bulbs. 

Their control of the equipment essential to the manufacturing proc- 
ess gives these concerns absolute dominance over the industry, regard- 
less of the patents they hold on finished products. But this control of 
the equipment through patents has not brought the two firms into con- 
flict. They have entered into cross-licensing agreements. The present 
agreements, setting forth terms and provisions under which licenses are 
granted to others, have been extended to 1951. 

Competition, in the face of this patent control could arise only from 
infringements by domestic producers or from foreign sources of 
supply. But the General Electric Co. has usually succeeded by means 
of lawsuits in making it legally impossible and financially ruinous 
for any concern to pursue an extended policy of violating General 
Electric patents. 

The possibility of foreign competition is rendered remote by the 
existence of agreements between the leading foreign producers, mem- 
bers of the international cartel, and a subsidiary* of General Electric, 
the International General Electric Co. In fact, definite interna- 
tional agreements exist concerning the manufacture and sale of bulbs 
in the principal world markets. In addition, the International Gen- 
eral Electric Co. owns stock in a number of foreign concerns en- 
gaged in the industry.^'' 

One of the most interesting examples of the extent to which con- 
trol of technological improvements through patents has been used to 
establish a high degree of concentration is the glass container in- 
dustry. This industry is dominated by two major companies which 
have established almost complete control over the production of glass 
containers. These two firms are the Hartford-Empire and the 
Owens-Illinois Glass Cos., of which the former is merely a patent- 
holding and research organization. Owens-Illinois is the largest 
glass container manufacturing company in the country. 

In 1937 approximately 29.2 percent of all glass containers were 
made on Owens machines while 67.4 percent were made on Hartford 
machines.*" Since Owens-Illinois has not licensed a newcomer in the 
industry since 1914, Hartford -controlled machines are practically the 
only equipment glass-making producers can operate unless they use 
obsolete, competitively inadequate machines on which patents have 

® U. S. Tariflf Commission, Incandescent Electric Lamps, Report No. 133, Second Series 
*o Hearings before the Temporary National Economic Committee, Part 2, Patents, p. 383 . 



216 CONCENTRATION OF ECONOMIC POWER 

expired, or machines of foreign make whose patent status in this 
country has not yet been established. 

The licenses granted by the Hartford-Empire Co. not only reserve 
ownership of the machine to the parent company, but in some cases 
also prescribe the type of commodity to be produced with the equip- 
ment, the quantity of production allowed, and the geographic areas 
in which the goods may be sold.*^ 

These two companies have, in effect, allocated the milk bottle field 
to the Owens-Illinois, Thatcher and Liberty Glass Cos., wliich, in 
1937, made 75 percent of the milk bottles produced. Fruit jars for 
domestic use have been allocated to Ball Brothers Co., which, in 
1937, produced 76 percent of the country's fruit jar output. Nu- 
merous similar provisions govern the production of practically every 
variety of glass container. 

Since the companies allocate quotas of production they exercise 
an effective control over prices in the various divisions of the in- 
dustry. Quotas apply to every producer of glass containers except 
Owens-Illinois and Hazel Atlas. Competition within geographical 
areas is limited through the companies' power to allow the sale of 
containers only in specified areas. These controls are effected by 
the power to withdraw or refuse licenses. 

Any action which could be construed by Hartford-Empire as an 
infringement upon its patents almost inevitably results in litigation. 
Hartford-Empire has engaged in expensive, long drawn-out law 
suits, which are extremely bothersome even to the large manufac- 
turers of glass containers, and are often ruinous to the small 
companies. 

The tecllnique of extending control over the industry after the 
expiration of basic patents has been carefully exploited by Hart- 
ford-Empire. One of the methods is the use of improvement patents 
which are taken out to protect an improvement on a machine; an- 
other is the use of so-called indirect patents. During 1927-29 Hart- 
ford applied for 200 patents, 112 of which were indirect or blocking 
patents.^2 

Another means of restricting competition by the use of patent con- 
trols is the refusal on the part of Hartford-Empire to license new 
capital, thus in effect freezing the economic structure of the in- 
dustry. 

Hartford-Empire does not attempt to stipulate in its written con- 
tracts the prices at which glass containers are to be sold, but it does 
engage in activities designed to prevent price-cutting and to insure 
"fair" prices to its licensees. Complaints from its licensees that cer- 
tain other producers are indulging in price-cutting are usually dealt 
with promptly by Hartford-Empire. In some cases it has even 
forced out of business certain concerns which were departing from 
the accepted price policies of the industry. 

As a result of this control, prices are uniform throughout most of 
the industry. When changes do occur they are initiated by recog- 
nized price leaders in the respective fields. Thatcher sets the price 
on milk bottles; Ball Bros, on domestic fruit jars; Hazel- Atlas on 
wide-mouth ware (largely packers' ware) ; while Owens-Illinois leads 



" Ibid., pp. 396-433. 

« Ibid., Exhibit No. 125, p. 779. 



CONOENTRATION OF ECONOMIC POWER 217 

on narrow-mouth ware (including proprietary and prescription ware 
and beverage bottles). 

This industry thus illustrates the way in which technological im- 
provements protected by patents have been the means not only of 
securing a high degree of economic concentration but also the con- 
trol of prices and marketing policies.*^ 

COKPORATE SIZE AND EARNING TOWER 

The advantages technology imparts to large-scale operation indi- 
cate that an increase in size of operation would logically be accom- 
panied by an increase in earning power. In fact, such a relationship 
has been found to exist. 

In a study of corporate earnings related to corporate size during 
1931-36, Professor William L. Crum of Harvard University found it 

strongly indicated though not fully established that, in all branches of industry 
large enough to show wide size distribution, (1) rate of return increases with 
size unmistakably in the low and moderately large size classes; (2) the advance 
in rate may continue in all size classes, or may disapi)ear or even be replaced by a 
moderate reduction in the high size classes; (3) even in cases for which rate 
declines in the highest size classes, rates for those classes remain emphatically 
above those for the lowest class. 

It is further to be noted that "these conclusions appear valid in 
separate years, in any stage of the business cycle." ^* 

In arriving at the rate of return. Professor Crum divided residual 
earnings — profits, as defined by the Treasury — by the owners' stake in 
enterprise — the book value of the equity, and used total assets, as 
reported on the balance sheets accompanying the corporation tax 
returns, as his criterion of size. 

Although such a computation admits of only general trends, the 
remarkable smoothness of the curves, and their close conformity to a 
basic pattern led Dr. Crum to conclude that "The larger the corpora- 
tion, the higher is the rate of return, on the average; and this relation 
holds, with surprising constancy, in each of the six years 1931-36." *^ 

The trend of corporate earnings in relation to the size of corpora- 
tions is shown on chart XVII, table 29, for all divisions and for spe- 
cific fields most likely to be affected by technology — manufacturing, 
mining, public utilities, and construction. 

It is impossible to ascertain the exact extent to which technology is 
responsible for this phenomenon. Since it was found by Professor 
Crum that this relationship between corporate earnings and size could 
not be accounted for by such non-technological factors as the degree 
of indebtedness of corporations of each size, or by profits arising from 
price fluctuations affecting inventory, or, to any considerable extent, 
by the compensation of principal executives, and since the data already 
presented in this study indicate generally that technological factors — 
efficiency of operation, industrial research, and patent control — confer 
economic, ad vantages on large-scale operation, it can be concluded that 
technology is undoubtedly responsible to an appreciable extent for 

*^ Ibid., pp. 547-548. The original legal basis of the patent system, the English 
Statute on Monopolies of 1623, contained a specific provision relating to prices ; it per- 
mitted the granting of fourteen-year monopolies, provided that "they be not contrary to 
law nor mischievous to the state, by raising prices of commodities at home, or hurt of 
trade, or generally inconvenient." 

« William L. Crum, Corporate Size and Earning Power ; Harvard University Press, Cam- 
bridge, 1939, p. 230. 

" Ibid., p. 32. 



218 



CONCENTRATION OF ECONOMIC POWER 



the rise in corporate earnings which accompanies increasing corporate 
size. 

Chart XVII 



cc 


-^ 


-) 


h- 


25^^ 


UJ 


«< 
<^H 


on 


ll 


o 


o 

Ixl 


2^- 


1- 
<r 


^§ 


cr 


^^ 








en o 

LlI q. 


< 


28 


LU 
N 


fefe 






0) 


^^ 




cr CO 


UJ 


t^T 


h- 


<l- 


< 


n-^ 


tr 


2s 


o 


n 


rr 


><f1' 


n 


CO ^ 


o 


8 



o I 1 1 1 1 1 1 

s-i-.— =ff~~ 

sLd-l 

i^ V 

— V g 

V - 



o « o 



« 9 £ 8 S S 



s s s 



1 














: 


S 1 


























; 














i' 
















J 1 












fMi 








/.\ ! ! 1 












t)\ ^ I ^' 








'%' - 


A 1 


















•i 


\' 1 


















=i 


. ^1 __J _ 




















2 






■V. 









!« o ^ ? ? ? ? ? ? 


1 


-> 
















7 














2 












! 


1- 














z 


' 








1 i 1 


L " 


I 








' 1 




^1- 


-\ 


^! ! 1 i M 




X' 


^ 


1 ' 










^"> ! 








1 




1 1 1 

1 ! ' 


a 








1 i 


-A 

5, i 


1 1 1 1 1 


1 


1 1 
i^ ! 


1 i 






° ' i 


fT 


1 1 


3 




>s 




\ : 


fe 


i 




^ 1 I 


? S 




3 lO 


2 « 




3 S S 






is i 



Table 29. — Corporate sise and rate of return, 6-year average of rates of return, 
1931-36, compared with size of corporation hy total assets 



0.05. 
0.10. 
0.25. 
0.50. 
1.... 
6.... 
10... 



Size class (millions of dollars) 



All divi- 


Mining 


Manufac- 


Construc- 


sions 


turing 


tion 


-16.34 


-33. 01 


-17.24 


-22. 14 


-4.73 


-6.24 


-5.22 


-8.74 


-2.68 


-2.95 


-2.36 


-4.92 


-1.40 


-1.38 


-.33 


-1.79 


-.80 


-1.52 


.80 


-1.17 


-.05 


-.57 


1.56 


3.26 


.38 


-.05 


2.54 




1.44 
2.59 




2.47 
3.58 













Public 
utilities 



-15.07 
-1.85 
.60 
-.51 
.15 
1.00 
1.14 
2.22 
1.62 



Source: William L. Cram, Corporate Size and Earning Power, Harvard University Press, 1939, appendix, 
table B. 



CONCENTRATION OF ECONOMIC POWER 219 

CONCLUSION 

In this study the labor-displacing effects of technology were 
weighed against the compensatory forces presumably inherent in our 
economic system. However, there exists no measure by which the 
balance between the labor-saving effects and the compensatory forces 
can be quantitatively determined. If such a measure were available, 
it would still be exceedingly difficult to allocate to each compensatory 
force its share of responsibility. 

Certain indirect methods can be used to determine whether balance 
exists. For example, the existence of a large amount of long-term 
unemployment would indicate lack of balance. 

For over a decade this condition has been an all too conspicuous 
characteristic of our economy. At the beginning of 1940, for ex- 
ample, the estimates of the total amount of unemployment ranged 
from 8,500,000 (by the National Industrial Conference Board) to 
nearly 12,000,000 (by the Congress of Industrial Organizations).*^ 

If it is assumed that under normal conditions the present economic 
system would provide full employment, then the very existence of 
large-scale unemployment denotes a state of unbalance. The question 
of greatest importance is whether this unbalance may be expected to 
continue. 

It seems apparent that technology will continue to increase labor 
productivity, to displace skilled occupations, and to reduce unit labor 
costs. In the absence of effective offsetting forces, economic and social 
distress may be expected to accumulate. 

Any reduction of heurs corresponding with the declines that took 
place during the World War and immediately after the enactment of 
the National Industrial Recovery Act must be regarded as distinctly 
remote because of legislation establishing the 40-hour week and a 
widespread acceptance of it as the norm of employment. 

The development of great, new industries likewise holds slight prom- 
ise of creating sufficient employment to offset the labor-displacing ef- 
fects of technology, especially since the eight general fields examined 
seem unlikely to create much employment. The limitations on the new 
industry stimulus involved in the substitution of products requiring 
less labor for those employing more labor per unit, the capital-saving 
characteristics of modern technology, and the pattern of present in- 
come distribution make it impossible to rely heavily upon their 
development as an immediate compensatory force. 

The third force which might offset labor-displacement is the re- 
duction of prices. In economic theory, price reductions are regarded 
as a primary stimulus to the expansion of economic activity. How- 
ever, concentrated industries tend to make relatively little use of this 
technique to expand output. If concentration continues to charac- 
terize a large segment of the economy, there is little basis for assum- 
ing that extensive use will be made of price reductions in the future. 
Actually it appears probable that much of the economy will con- 



*' For a discussion of prevailing, estimates of unemployment, see Harvard University, 
Review of Economic Statistics, "Estimates of Unemployment in the United States," by R. A. 
Nixon and P. A. Samuels, August 1940. This analysis concludes. ''With recognized qualifi- 
cations,' however, the series compiled by the organizations here studied (the Alexander 
Hamilton Institute, the American Federation of Labor. The Congress of Industrial Organiza- 
tions, the National Industrial Conference Board, and the President's Committee on Economic 
Security — Robert R. Nathan, consultant) provide fairly reliable Indications of the volume of 
unemployment in the United States." 



220 CONCENTRATION OF ECONOMIC POWER 

tinue to be marked by concentration, since technology through the 
greater efficiency of large-scale operation, through industrial re- 
search and through patents, contributes materially to the growth of 
concentration. Thus there is presented this fundamental contra- 
diction: while technology on the one hand creates tremendous eco- 
nomic prohlems through the displacement of lahor, on the other it 
induces concentration^ therehy impeding the operation of the com- 
pensatory force of price reductions. 

Higher wages are regarded by some as a possible stimulus. But in 
analyzing the trend of unit labor costs, it was found that even during 
periods of the greatest increase in wages, the advances in average 
hourly earnings were generally exceeded by still greater increases in 
output per man-hour, with the result that unit labor costs decline. 
This tendenc}'^ for increases in labor productivity to exceed those in 
wages limits greatly the possibility of a material stimulus emanating 
from this source. 

From where else can the stimulus be expected to come? From war? 
If the preparation for and the conduct of war constitute the only 
adequate compensatory force to the labor-displacing effects of tech- 
nology, the proposition would then be established that only through 
war can the present economic system be operated in such a way as to 
approximate full employment. 

Today a state of unbalance exists, and it seems likely that under 
present conditions unbalance will continue and perhaps become even 
more pronounced. 



APPENDIX A 
MEASURES OF LABOR PRODUCTIVITY ^ 

(1) One of the first methods introduced to indicate the use of 
machinery, and by inference the productivity of labor, was the measure- 
ment of the amount of horsepower utilized or the rated horsepower 
capacity of power equipment. Statistics on horsepower, as a measure- 
ment of labor productivity, are deficient in two respects : (a) They do 
not give a perfect record of the changes in power used or even of in- 
stalled power equipment; (b) the changes in power per worker, even 
if precisely measured, do not represent adequately the changes in 
productivity because some important mechanical improvementa 
actually decrease rather than increase power requirements. 

Use of power by no means corresponds with the rated capacity of 
power equipment. The possibility of running motors with an over- 
load, together with improvements in transmission, tends to lower the 
capacity without changing the amount of work that can be accom- 
plished. When electric motors are driven by current generated in the 
plant, their rated capacity is likely to exceed considerably the rated 
capacity of the prime movers because all motors in an establishment 
do not run at the same time or at full capacity ; consequently, the rated 
capacity may be increased without a corresponding change in work 
done. 

A somewhat better measure of labor productivity is the amount of 
power actually used. It is inadequate because certain labor-saving 
techniques decrease the amount of power used. Furthermore, indus- 
tries which use heavy raw materials not easily handled require a rela- 
tively large amount of power ; but their large consumption of power 
by no means indicates that labor productivity is higher than in indus- 
tries which use lighter and more tractable materials. 

(2) The machine-output measure is the ratio of output prepared by 
machine methods to hand methods in specific phases of productive 
operations. The chief problem in the use of this measure is the diffi- 
culty of .obtaining data which would clearly differentiate between the 
proportion of total output produced by machine methods and that pro- 
duced by hand methods. The border line cases between hand processes 
and machine processes would necessitate arbitrary determinations 'as 
to whether an operation in which a worker utilizes a simple machine 
tool is a machine or a hand method. Furthermore, except for special 
studies in fields such as the cigar, glass bottle, pig iron, and bitumi- 
nous coal industries, there are no primary data from which such a 
computation could be made. 

(3) The machine-labor ratio is somewhat similar to the machine- 
output ratio except that it represents the proportion of machine 
workers to the total labor force instead of the proportion of machine- 
made goods to the total physical output. The difficulty of classifica- 
tion, however, remains the same. If an arbitrary classification of all 



^ The first 4 are adapted from Harry Jerome. Mechanization in Industry, National Bureau 
of Economic Research, New York, 1934, pp. 205-254. 

221 



222 CONCENTRATION OF ECONOMIC POWER 

workers is made according to theii- connection with machine processes, 
it would still be defective as a measure of labor productivity. In two 
industries with exactly the same proportion of machine workers to 
hand workers, labor productivity might be much higher in one than 
in the other due to the use of more productive equipment. 

(4) The labor-expense measure is the ratio of wages to value added 
by manufacture. Its great advantage is that it can be regularly com- 
puted for particular industries from data gathered by the Census of 
Manufactures. Wages as a percent of value added indicates the 
amount in actual dollars labor receives for productive effort in pro- 
portion to the total value of product less the cost of materials, supplies, 
containers, fuel, purchased electric energy, and contract work. 

This ratio fluctuates, however, not only because of changes in labor 
productivity, but also because of price fluctuations in both finished 
products and raw materials. To the extent that prices are constant, 
variations in the ratio of wages to value added would indicate the 
trend of labor productivity. Until price indexes, designed for this 
purpose, which make possible the accurate adjustment for price 
changes have been compiled, the ratio will possess more usefulness 
in indicating changes of labor's share in the creation of dollar values 
rather than changes in labor productivity. 

(5) Physical output per wage-earner is a more widely used measure 
of changes in labor productivity. If changes in output of individual 
workers over an extended period of time are to be measured, produc- 
tion per wage-earner probably constitutes the best available measure- 
ment. Labor productivity is usually considered, however, in terms 
of the total amount of labor required rather than the total amount 
of goods produced by any individual worker in a given period. 
Changes in weekly hours may considerably distort output per wage- 
earner as a measure of labor productivity. The amount gf labor re- 
quired over a period of time to produce a given unit of goods may 
decrease, while the amount of goods produced by an individual wage- 
earner may actually decline due to a decrease in weekly hours. 

(6) Output per man-hour indicates most accurately the amount of 
labor required to produce a given unit of goods. Reductions in weekly 
hours, changes of price, or any of the factors which tend to distort 
the other measures of labor productivity do not affect output per 
man-hour. 

While it is undoubtedly the best available measure of labor produc- 
tivity, two limitations deserve attention. The reduction of all labor 
to a least common denominator, man-hours, removes the possibility of 
noting changes in types of work. Therefore, while man-hour output 
reflects accurately the amount of labor time required to produce a 
given unit of goods, it does not indicate changes in the quality of 
labor required. If changes in the quality of labor such as the replace- 
ment of skilled by unskilled workers are considered occupational 
rather than productivity phenomena, this objection to the use of 
man-hour output is automatically eliminated. 

Output per man-hour also fails to reflect substitutions of materials 
which may reduce the amount of labor required for a given function. 
The actual loss to labor from this source could be measured by man- 
hour indexes but since they are related to units of specified types of 
goods, they do not reveal, as now computed, those changes in labor 
productivity resulting from replacements of materials. 



APPENDIX B 
PRODUCTION AND MAN-HOUR INDEXES 

The index of production used for manufacturing is that computed 
by the National Bureau of Economic Research and extends to 1937. 
It is based upon all manufacturing industries reporting to the Census 
of Manufactures. In computing this index, the National Bureau of 
Economic Research obtains physical production series on all items 
for which such data are compiled by the Census ; and on those items 
for which such data are not available it derives indexes showing the 
probable trend of output by means of ratios based on such measures 
as value added by manufacture and employment. To indicate the 
trend of physical production for 1938 and 1939 the Federal Reserve 
Board index of manufacturing production was linked to' the National 
Bureau series.^ 

The indexes oi production for anthracite and bituminous coal min- 
ing a?^e derived from tonnage figures. Because they fail to take ac- 
count of cleaning, sorting, and adaptation to specialized means, the 
rise in man-hour output during recent years is minimized. With 
declining production, the elimination of high cost mines, especially 
in anthracite mining, tends to raise man-hour output. 

In steam railroads ^ the production index used is derived from rev- 
enue traffic units on the generally used basis of revenue passenger- 
miles times 2.6 plus revenue ton-miles. 

Man-hour output for each of the fields has been computed by Witt 
Bowden of the United States Bureau of Labor Statistics from data 
gathered by that agency.^ 



1 The production index of the National Bureau of Economic Research is not yet published 
but was supplied by Mr. Solomon Fabricant. This index is described in a forthcoming volume 
by Mr. Fabricant to be published by the National Bureau and entitled "The Output of Manu- 
facturing Industries, 1899-1937." 

2 Reference in this report to steam railroads is to class I steam railroads, which in 1938 
handled 99.39 percent of all freight traffic based on ton miles or 94.38 percent based on tons 
of freight carried, and 99.87 percent of all passenger traffic based on passenger miles or 99.61 
percent based on number of passengers carried. 

3 The indexes of man-hours, excluding the break-down for bituminous coal and anthracite 
mining, are published for certain years in an article by Dr. Bowden in the Monthly Labor 
Review, September 1940, "Wages, Hours, and Productivity of Industrial Labor, 1909-39." 

223 



APPENDIX C 

THE NATIONAL RESEARCH PROJECT INDEXES OF 
LABOR PRODUCTIVITY IN MANUFACTURING INDUS- 
TRIES 

The National Research Project, in the construction of the index 
numbers designed specifically to measure the trends of employment 
opportunities, developed formulas which necessitate the weighting 
of constituent series by labor time rather than value. It is the pur- 
pose of these formulas to present an index of production which, when 
divided by the relatives of man-hours and multiplied by 100, yield 
the same index of productivity which could be obtained by averaging 
with appropriate weights the relatives of the output per man-hour 
for the several products comprising the index of production. Certain 
data requisite for the development of such formulas are not available. 
The principal problem in the formulation of these indexes Avas a 
frequent lack of comparability between figures on physical produc- 
tion and those on man-hours. To attain approximate comparability, 
the original data were at times subjected to considerable adjustment. 

Since man-hour figures for individual products, required for 
weighting purposes, are almost completely non-existent, it was neces- 
sary to use unit-value weights or unit-value added weights for the 
purpose of combining industry indexes into group measures, and 
group indexes into over-all series. The method of computing man- 
hour figures for most of the industries consists in multiplying em- 
ployment by average weekly hours. 

Because the index of productivity can be obtained only as the 
quotient of a production index and an adjusted corresponding wage- 
earner or man-hour index, the validity of the productivity index 
depends entirely on the accuracy of the production and labor series. 
Concerning the validity of the indexes, the Central Statistical Board 
observes that — 

In view of the fact that innumerable adjustments had to be made to the avail- 
able data to correct for the lack of uniformity as to scope, coverage, and 
detail, and that many of the indexes are based on estimiates, it is difficult to 
determine the accuracy of these measurements. In general, it may be said 
that although the indexes may not register the correct magnitude of the year- 
to-year movements, they do probably indicate the direction of the trends 
accurately.* 



1 Report of the executive secretary to the Central Statistical Board for August and 
September 1939, p. 2. 

224 



APPENDIX D 

THE REDUCTION OF VARIETIES THROUGH 
STANDARDIZATION ' 



Some of the more obvious advantages gained from standardization 
follow. To the manufacturer: Less capital tied up in slow-m|oving 
stocks, simplified inspection requirements, longer runs with fewer 
changes, less idle equipment, less stock to handle, larger production 
units, less special machinery, more prompt delivery, less chance of 
error in shipment, and less obsolescence in material and equipment. 
To the jobber, wholesaler, and retailer : Increased turn-over, elimina- 
tion of slow-moving stock, staple lines — easy to buy and quick to sell — 
greater concentration of sales efforts on fewer items, decreased capi- 
tal invested in new stocks and repair parts, less storage space 
required, and decreased overhead and handling charges. To the 
consumer: Better values than otherwise possible, better service in 
delivery and repairs, and better quality of products. 

Reduction of varieties effected by standardization 







Varieties 


Item 


Before 
simpli- 
fication 


After 
simpli- 
fication 


Percent 
reduc- 
tion 




8,000 
102 
845 

124 
61 

188 
66 

32 
7 

33 
34 
44 
3 
18 
90 
78 

' 718 
244 

130 

2,328 

1,500 

76 

15 
15 
78 
16 
11 
13 
18 
8 
18 
18 


1,865 
11 
384 

26 
47 
28 
26 

11 
4 

1 
1 
1 
2 
3 
5 
11 

10 
96 

13 
38 
840 
41 

12 
12 

I 
3 
3 
4 
3 
2 
2 












Bags: 

Qlassine 


ci>aa 


79 


Paper, grocers' 


23 


Paper, notion and millinery 


sizes 


85 


Barrels, steel; and drums .• 

Bars: 


capacities. . 


^1 








Beds, hospital 

Blackboard, composition 

Blankets 


lengths.. 
widths.. 

heights.. 

colors... 
widths.. 

lengths.. 


97 
97 
98 
33 
83 
94 


Board: 

Binders... 




Box thit'*^ nesses of 


61 


Boilers: 

Range, and expansion tanks 


90 


Steel, horizontal firebox heating 


98 










Bottles, carbonated beverage: 




20 


Soda and imported ginger ale 


capacities 

heights 


20 
92 




weight of glass.. 

capacities 

heights 


50 
73 


Beers, export, select, or C. S 


weight of glass.. 

capacities 

heights 

weight of glass.. 


78 
62 
89 
89 



Source: National Bureau of Standards, Letter Circular LC 501, 



7551— 41— No. 22- 



September 3, 193 
225 



226 CONCENTRATION OF ECONOMIC POWER 

Reductions of varieties effected by standardization — Continued 







Varieties 


Item 


Before 
simpli- 
fication 


After 
simpli- 
fication 


Percent 
reduc- 
tion 




49 

150 
322 
262 
100 
1,084 

75 
75 
66 
45 

47 
34 
480 

200 
14 
49 
30 

348 
1,831 

4 

12 
16 

7 
10 

700 
700 
700 
700 
360 

200 
14 

65 
20 
25 
40 
25 
6 
16 
15 
66 
460 
110 
552 
2,072 
640 
62 
25 
38 
1,154 

\l 
13 
386 
242 

38 
22 
26 

272 

78 

1,154 

125 

120 
80 
10 

100 
65 

102 
65 
9 


4 

41 
75 
.59 
2 
194 

2 
2 
4 

1 

10 

8 

143 

4 

194 

.... 

5 
14 
4 
8 

309 
345 
345 
258 
l.M 

21 
5 

10 
3 
5 
9 
7 
4 
6 
6 

26 

94 

76 

71 
200 
475 

48 
7 

22 
873 

29 

5 

9 

180 

118 

22 
3 

13 
210 

38 
873 

29 
fi 
5 
2 

37 

23 
6 
6 
6 i 


92 


Boxes: 

Corrugated and solid fiber, for canned fruits and vegetables 


73 


Corrugated, for department and specialty store use.. 
Folding, for department and specialty store use 


sizes. - 

sizes.. 


77 
77 
98 


Set up 

Brick: 


sizes.. 


82 
97 


Face, rough and smooth; and common brick 


97 
94 


Sand-lime 


98 


Brushes: 

Calcimine, block sizes for 

Paint and varnish. 


/widths. 

Uhicknesses.. 


79 
76 
70 


Cans: 


90 


Tinned, steel, icecream types and sizes. . 

Caps, bottle, milk and cream; and milk and cream bottles 

Cartons, icecream; and icecream brick molds mold sizes.. 


64 
92 
97 
44 


Chain, welded 


34 


Checks, guest j restaurant: 

Cardboard checks 


grades... 

widths.. 

lengths., 
(grades... 


25 
58 
12 
43 


Chinaware: 




56 




51 




51 


Hotel 


63 




57 


Containers (cans): 

Fruit and vegetable (names and dimensions) 


90 


Tinned, steel, ice cream 


types and sizes 


64 


Containers: 




85 


Glass; for cottage cheese and sour cream. 

Glass, for mayonnaise and kindred products 

Glass, for preserves, jellies, and apple butter 

Cups, ice cream; and cup caps 

Drums, steel; and barrels.. 


..do.... 

capacities.. 

preserve jars 

jelly glasses 

apple butter jars., 
fcups.. 

leaps.. 

capacities.. 


85 
30 
77 
72 
33 
69 
60 
61 
80 


Elbows and fittings for eaves trough and conductor pipe 
Fencing, woven-wire; and woven-wire fence packages... 
F iles and rasps . . ... . 


.varieties.. 


31 

87 


■"'tpackage sizes . 


90 
26 


Fittings, wrought iron and wrought steel, and pipe and valves.. pipe sizes.. 


24 


Fungicides and insecticides 


package sizes.. 


42 




59 


Goring, shoe, elastic. 

Hammers, forged 


/qualities.. 
\widths.... 


69 
31 
53 






Insecticides and fungicides package sizes. . 

Investment, dental: plaster and artificial stone. Packaging of. {Pj^l'^'j^gm- 


42 
86 
50 


Jacks, screw, bell-bottom.. 


Vi,« 


51 


Japanned ware, tinware, and galvanized ware 


24 


Lath, metal (expanded and sheet) 


76 


Lights; sidewalk, floor, and roof 

Lining, brake, automobile 


(sizes.... 

styles. - 

shape?.. 


9S 
94 
80 
63 


Lockers, steel 


64 




(lengths 


11 
91 
33 




thicknesses.. 



CONCENTRATION OF ECONOMIC POWER 227 

Reductions of varieties effected hy standardization — Continued 





Varieties 


Item 


Before 
simpli- 
fication 


After 
simpli- 
fication 


Percent 
reduc- 
tion 


Millboard, asbestos; and asbestos paper l^^nh^'^rt" 


72 
21 

7 
30 
428 
423 
71 
22 
26 
38 
552 
2,072 

72 
21 
73 
13 
21 

62 
110 

26 

22 

26 

640 

188 

292 

42 

5,136 

360 

4,067 

5,136 

251 

1,260 

827 

5,136 

7 

8 

1,630 

22 

26 

11 

428 

423 

16 

21 

32 

130 

120 

8 

1,208 

454 

36 

1,154 

665 

74 

1,114 

110 

248 

1,304 

45 

62 

125 

50 

715,200 
91 


20 

4 

4 

1 

185 

127 

38 

3 

13 

22 

71 

200 

20 
4 

62 
3 
4 

48 
76 

15 
3 
13 

475 
25 
178 
11 
1,161 
154 
283 
1,161 

25 

309 

118 

1,161 

4 

6 

209 

3 

13 

185 
127 
8 
8 
5 
15 

13 
14 
3 

936 
26 
22 

873 

503 
6 
76 
76 

115 
1,124 
18 
48 
27 
10 
11 
254,400 
10 


72 


Mirrors, dental cone-socket __. sizes. 

Molds, brick, ice cream; and ice cream cartons - mold sizes.. 

Nails, cut, small, and cut tacks -{packages."^ 

Packages, salt packages. 

Packaging of dental plaster, investment, and artificial stone.. . . {investoent' 


43 
97 
57 
70 
46 
86 
oO 


Packaging, woven-wire fence; and woven-wire fencing (nke sizes 

Paper: 

Asbestos; and asbestos millboard (mimjoard " 

Photographic. 


87 
90 

70 
81 
15 


Tissue f roll, sizes... 

"*"^ - --- -Ishoe, sizes.. 

Pipe: 

Wrought-iron and wrought-steel; and valves and fittings... pipe sizes.. 

Conductor; and eaves trough, elbows, and fittings varieties.. 

Plaster: 

Adhesive 


81 

24 
31 

40 


Dental; investment, and artificial stone, packaging of --..(PJ^st'^,~t-- 


86 
50 
26 






Roofing, iron and steel . . 


39 


Rubber, dental, base and veneering colors.. 


74 


Screen cloth, insect, wire.. 


57 






Shovels, spades, and scoops (first revision) 


77 


Slate: 

Blackboard. 


90 


Roofing ^ ._. possible sizes.. 

Structural 


75 
81 






Spirals, steel, reinforcing sizes.. 

Spools, metal, for annealing, handling, and shipping wire 


43 

25 






Stone, artificial, dental; plaster and investment, packaging of {{jj^s^tment " 

Sweeps, floor. block lengths!! 

Tacks, cut, and small cut nails. .. {^'^*k" 


50 
36 

57 


fsizes 


50 


Tags. Shipping, paper.......... J.radesc^stock.. 

colors 


62 
29 
53 


Tanks: 

Expansion, and range boilers 


90 




88 






Ternos. roofing. 


22 


Textiles, cotton, hospital and institutional 


94 


Tile, building, hollow... 


39 




24 




24 


Towels, ten- ■.-, fast selvage sizes 


92 


Traps, lavatory and sink, brass 


93 


Trough, eaves; and conductor pipe, elbows, and fittings ..varieties.. 

Turnbucklcs . . sizes 


31 

54 


Twines, hard-fiber (ply and vard goods) 


14 




60 


Valves, wroughtiron and wrought-steel; and pipe and fittings. ..pipe sizes.. 
Wheelbarrows. . < vv . vv 


24 

78 


Wheels, brush, dental 


80 




35 


Wheels, grinding 


64 




89 







APPENDIX E 

WAGES AS A PERCENT OF VALUE ADDED BY 
MANUFACTURE 

The following table shows wages as a percent of value added by 
manufacture for 50 representative industries in 1935.^ The indus- 
tries in which chemical processes are of major importance fall among 
those in which the proportion is relatively low and have been itali- 
cized : 

Wages as a percent of value added by manufacture 
Industry : Percent 

Drugs and medicines 10. 5 

Manufactured gas 11. 1 

Cigarettes _*■_ 11.1 

Soap 15-4^ 

Paints, pigments, and varnishes '. 17. 4' 

Flour - 19. 8 

Butter ._: 22. 4 

Fertilizers 23.3 

Chemicals not elsewhere classified 23. 8 

Explosives 23. 9 

Blast furnaces 25. 6 

Cement 27.0 

Smelting and refining; copper _ — ^_ 29.8 

Petroleum refining 30. 4 

Tin cans ^ 32.2 

Sugar refining, cane 33. 9 

Cutlery and edged tools . 35. 1 

Coke-oven products : 37.1 

Motor A-ehicles ,_ 37. 6 

Steaiji and hot-water heating apparatus 38. 8 

Paper____':S i 38.9 

Paper boxes 38. 9 

Stoves . 40. 3 

Rubber goods, other than tires, inner tubes, and boots and shoes 40. 9 

Wirework-I . : 41.0 

Glass ^ 41.1 

Structural and ornamental metalwork 41.4 

Agricultural implements - 42. 4 

Rubber tires and inner tubes 43. 3 

Hardware 43. 5 

Aluminum manufactures 44. 6 

Cast-iron pipe 44. 9 

Carpets and rugs 46. 8 

Baking 47. 1 

Bolts, nuts, washers, and rivets 47.1 



1 Value added by manufacture represents the value of product less the cost of materials, 
supplies, containers, fuel, purchased electric energy, and contract works. It measures the 
net addition to the value of commodities, and is almost completely free from duplications 
existing in the total value of products. 

It does not indicate the percentage which the actual quantity of labor constitutes of the 
physical volume of production. But the fact that workers in chemical industries receive 
higher than averajre wage rates (in 1937 average hourly earnings and average weekly earn- 
ings for the chemical industry group were $0.71 and $27.67, respectively, compared with 
$0.63 and $24.0,5 for all manufacturing) suggests that the measurement actually overstates 
the role occupied by labor in chemical production. 

228 



CX)NCENTRATIOX OF ECONOMIC POWER 229 
Wages -MS a percent of value added by manufacture — Continued 

Industry— Continued. Percent 

Clay products, other than pottery 47. 2 

Clocks, watches, and time-recording devices 47.3 

Plumbers' supplies . 48. 7 

Rubber boots and shoes 49. 7 

Furniture 50.2 

Leather ^ ^ 50. 4 

Lumber 53. 3 

Steel works and rolling mills 53. 5 

Sewing machines and attachments 54. 5 

Woolen and worsted goods 54. Q 

Boots and shoes 55. 5 

Shirts 57.7 

Foundry products _ 61. 3 

Cotton woven goods 63. 1 

Knit goods 63. 4 

Source : Computed from Census of Manufactures, 1935. 



APPENDIX F 
TVPES of plastics and their uses, 1940 



TYPE OF PLASTIC 

Shellac plastic 



Bitumen plastic 

Phenolic jiiolding plastic 

Laminated phenolic plastic- 
Casein plastic 

Cellulose-acetate plastic 

Urea-formaldehyde plastic 



Cast phenolic plastic 



TYPICAL APPLICATIONS 

Adhesives, dental blanks for taking im- 
pressions, electrical insulation, grind- 
ing wheels, novelties, phonograph 
records, poker chips, protective coat- 
ings, thermal insulating board. 

Connector plugs on household electri- 
cal equipment, handles and knobs for 
cooking utensils, valve wheels, arc 
shields, battery boxes. 

Automotive and airplane parts, camera 
cases, closures, corrosion-resistant ap-> 
paratus, electrical insulation, han- 
dles, housings, telephone equipment. 

Bearings, gears, electrical apparatus, 
radio equipment, trays, table tops, 
refrigerator doors, wall coverings, 
doors, counter and cabinet paneling, 
translucent and opaque signs. 

Beads, buckles, buttons, game counters, 
novelties, trimming accessories. 

Airplane cockpit enclosures, automobile 
accessories, costume jewelry, combs, 
and toilet articles, electrical appli- 
ance i, lamp shades and lighting ac- 
cessories, pen and pencil barrels, 
radio parts, spectacle frames, trans- 
parent containers, watch crystals. 

Buttons and buckles, closures, contain- 
ers, illuminated dials, dome and side- 
wall lenses for automobiles, electri- 
cal appliance fittings, hardware trim, 
piano keys, reflectors, tableware, toys, 
and novelties. 

Advertising signs and displays, brush 
backs, costume jewelry, clock cases, 
game counters and pieces, novelties, 
radio housings, lighting fixtures, in- 
dustrial adhesives, laminating var- 
nishes. 



230 



CONCENTRATION OF ECONOMIC POWER 231 

TYPE OF PLASTIC TYPICAL, APPLICATIONS 

Vinyl resin plastic Adhesives, inks, metallic paints, plastic 

wood-filled compositions, molded 
articles, cable coverings, coated fab- 
rics, impregnated tape, molded arti- 
cles, tank linings, cement coatings, 
films, floor tiles, metal coatings, radio 
parts, sound records, storage bat- 
teries, plating tanks, wallboard coat- 
ings, laminated glass. 

Polystyrene plastic Bottle closures, radio parts, refrigera- 
tor trim, television parts, transpar- 
ent automotive accessories. 

Acryliq plastic Adhesives, airplane windshields, deco- 
rative articles, dentures, displays, il- 
luminated signs, lenses, protective 
coatings, reflectors for highway light- 
ing. 

Cellulose-acetate butyrate Exterior automobile accessories, fisher- 
plastic, men's equipment, handles and hous- 
ings for outdoor uses. 

Ethylcelluloee plastic Adhesives, cable coatings, extruded 

wire insulation, injection molded ar- 
ticles, hot-melt coatings for paper 
and cloth, pigment grinding base, 
protective coatings. 

Comnarane-indene Xesin Mastic floor tile, paper impregnation, 

protiective coatings, rubber com- 
pounding, transcription records. 

Source: Compiled from the Journal of the Society of Automotive 
Engineers; May 1^40; "Plastics and Their Uses in the Automotive 
Industry," by Gordon M. Kline. 



APPENDIX G 

EFFECT ON LABOR OF SPECIFIED TECHNOLOGICAL 
CHANGES IN TWO TIRE-MANUFACTURING PLANTS 



TECHNOLOGICAL CHANGE 

3 rubber plasticators installed. 



Liquid soapstoning devices in- 
stalled for Banbury mixers. 



Direct method of tire building 
. installed using gum-insert- 
ing machines, rotary cutters, 
compensators, Jliner stands, 
etc. 



Compensators installed on 40 
tire-building machines, and 
room rearrangement to take 
care of increased output. 

5 curing units equipped with 
overhead conveyors, tire re- 
movers, etc. 

Curing room rearranged to 
take care of increased pro- 
duction. 

Preparation conveyor in tube 
room moved and service con- 
veyor and automatic soap- 
stoning rearranged. 

6 automatic cutters installed 
on tube preparation unit. 

New tray skids purchased for 
the handling of tubes and 
flaps. 

Banbury mixers installed for 
2 tandem calenders. 

232 



EFFECT ON LABOR 

Saving in direct labor, due to in- 
creased man-hour output, of 328 
man-hours per day, equivalent to 
displacement of 41 men. 

1 man per shift, who formerly soap- 
stoned by hand, eliminated. Labor 
saving, 24 man-hours per day, or 3 
men displaced. 

Savings in normal production: (1) 
Replacement of male with female 
labor, (2) elimination of time lost 
by assemblers due to stock changes, 
(3) direct handling of stock from 
rotary cutter, (4) elimination of 
trucking assembled bands to tire 
room. Saving in direct labor, 248 
man-hours per day, or 31 men dis- 
placed. 

Savings in normal production esti- 
mated to exceed 416 man-hours per 
day, or 52 men displaced. 

5 men per shift eliminated. 



Savings in direct labor, when operat- 
ing at full capacity, 173 man-nours 
per day, or 22 men displaced. 

2 girls per shift eliminated, saving 
48 man-hours per day. 



1 girl per shift eliminated. 

2 bookers per shift eliminated. 



2 truckmen, 8 millmen, and 6 com- 
pounders per day eliminated, sav- 
ing in direct labor 128 man-hours. 



CONCENTRATION OF ECONOMIC POWER 



233 



TECHNOLOGICAL CHANGE 

Cutting and rerolling depart- 
ments consolid-ated and re- 
arranged. 

Festoons and working plat- 
forms erected for the sup- 
plying of stock to the auto- 
matic unit of tire building. 

20 modern shoulder- drum ma- 
chines installed to replace 
old flat-drum machines for 
building tires. 

Tire conveyor extended from 
building unit to painting 
machines. 

New system of sorting and as- 
sembling tubes installed. 

2 conveyor units, 1 for the pur- 
pose of assembling inner 
tube valves and the others 
for the testing of valves in- 
stalled: 



EFFECT OF LABOR 

Direct labor saving, 112 man-hours 
per day, or 14 girls displaced. 

3 supply girls per shift eliminated, 
72 man-hours per day. 



Direct labor saving 600 man-hours 
per day, or 75 men displaced. 



1 trucker and one-half a loading man 
per shift eliminated, saving 36 
man-hours per day. 

6 girls eliminated, saving 64 man- 
hours per day. 

5 men and 5 girls eliminated, saving 
80 man-hours per day. 



Source: U. S. Bureau of Labor Statistics^ Bulletin No. 585, Labor Produc- 
tivity in the Automobile Tire Industry, by Boris Stern, July 1933, table II, p. 5. 



APPENDIX H 
LABOR PRODUCTIVITY AND INDUSTRIAL PRICES 

INTRODUCTION ' 

It is the purpose of this appendix to determine the extent to which 
labor-saving has been offset by price reductions in concentrated as 
against non-concentrated industries. 

To ascertain whether or not labor-saving has been offset by price 
reductions, the relationship between the behavior of labor produc- 
tivity and of price is set forth here for nine major industries. Both, 
concentrated and non-concentrated industries are included. The de- 
gree of concentration for each is shown in table 1.- This approach 
contrasts for each industry the increase in labor productivity with 
the use which has been made in the industry of a technique of off- 
setting that increase — the teclinique of price reductions. . Price re- 
ductions are of importance as a compensatory force because (1) Any 
advances in labor productivity^ within an industry, unless balanced 
by increases in production, bring about decreases in man-hour em- 
ployment and (2) reduction of prices is in economic thought a basic 
method of increasing demand and output. 

Table 1. — Degree of concentration in selected major industries, 1935 





Degree of concentra- 
tion, 1935 


Value added by manu- 
facture per establish- 
ment 




Largest 4 
producers 
(percent 
of total) 


Largest 8 
producers 
(percent 
of total) 


1929 


1935 


CONCENTRATED INDUSTRIES 


49.3 

87.0 
89.7 

8.0 
23.1 
5.3 


63.8 
(') 
44.7 
94.2 
99.4 

14.4 
32.9 
8.8 


$2, 746, 000 

1, 957, 000 

990,000 

6, 262, 000 


$1,912,000 




1, 262, 000 


3. Cement 


506,000 


4. Motor vehicles * 


4 768 000 


6. Cigarettes...- 


5, 703, 000 


NON-CONCENTRATED INDUSTRIES 


439,000 
448,000 
138,000 


330,000 


7. Woolen and Worsted goods • 


399,000 


8. Furniture ' 


75,000 



' Concentration: Steel works and rolling mills; value added: Crude iron and steel and rolled products. 

' Value added: Smelting and refining; Copper, lead, and zinc. 

' Concentration in copper and lead is so high that it cannot be presented because of possibility of disclosure 
of individual enterprises. Concentration in the zinc smelting and refining industry, by value of product, 
64.0 percent. 

t Concentration: Motor vehicles, not including motorc'ycles. 

t Cotton manufactures. 

« Wool and hair manufactures. 

' Furniture, including store and ofllce fixtures. 

Sources: Degree of concentration— National Resources Committee, The Structure of the American Econ- 
omy, Part 1, 1939, appendix 7. Value added by manufacture— U. 8. Bureau of the Census, Census of Manu- 
actures, 1929, 1935. 



^ This appendix was condensed from parts of a dissertation, "Labor Productivity and 
Industrial Prices," submitted by the author as partial fulfillment for the degree of doctor 
of philosophy at the American University Graduate School. For a more adequate treat- 
ment, the source should be consulted. 

* It is to be noted that the first three industries of the concentrated group manufac- 
ture producers' goods while the latter two (plus the sixth of that group, the Electric 
Light and Power Industry) produce consumers goods; also each group contains producers 
of durablp and of nondurable goods. 

234 



CONCENTRATION OF ECONOMIC POWER 235 

The comparison, however, does not carry the suggestion that a 
given decrease in unit labor requirements means that costs of pro- 
duction declined by that amount and that, therefore, prices should 
have declined correspondingly. In this report the behavior of price 
is regarded not merely as an effect but also as a cause of cost be- 
havior. A decrease in price, by stimulating production, would prob- 
ably bring into play the principle of decreasing costs, thus affecting 
materially the level of costs. This interpretation of price behavior 
as cause as well as effect is in obvious conformity wit^i the way in 
which it is analyzed in this report, namely, in terms of the use made 
of price reductions as a possible means of expanding output and 
thereby offsetting the labor-displacing effects of teclinology. 

The .comparison 6f labor productivity to prices also does not carry 
the suggestion that the price series are a measure over any period 
of time of th& cost of obtaining a given amount of utilities. Price 
statistics as a whole have not been developed to a point where they 
measure adequately changes in quality. To cite the obvious example, 
the automobile of 1937 was not that of 1927. But the price indexes 
are not used in this report to measure changes in the cost of ob- 
taining a given amount of utilities. They are presented merely to 
show the extent to which one specific technique, the reduction in the 
actual price, has been used to enlarge the market. Obviously, there 
are other methods designed to stimulate sales such as increased ad- 
vertising, improvements in quality, etc. 

It should be pointed out, however, that quality improvements are 
by no means necessarily tantamount to price reductions. Improve- 
ments might be made ad infmitum in the quality of an automobile, 
but if the cash outlay required remains above the purchasing ability 
of a given income group, no amount of quality improvements could 
make it possible for members of that income group to purchase the 
automobile. 

The basic principle underlying this study may be summarized as 
follows: 

If in any segment of American industry increased productivity of 
labor is accompanied by long-term price stability instead of by price 
reductions, man-hour employment would almost inevitably decline 
unless some other stimulant to production, such as governmental 
expenditures, is applied. 

THE CONCENTRATED INDUSTRIES 

lEON AND STEEL INDUSTRY 

Productivity. 

The iron and steel industry includes a variety of manufacturing 
and mining operations up to and including the final fabrication 
of finished steel products. The definition of the industry could in- 
clude the extraction of iron ore, but since this operation is usually 
classified as mining, the iron and steel industrj^ will be regarded as 
extending back only to the making of pig iron in this study. 

Productivity has been . increased over the last 2 decades in 
practically every phase of the iron and steel industry. It has ad- 
vanced particularly in the three basic processing stages: The blast 
furnace which extracts the iron from the ore, the open hearth furnace 



236 CONCENTRATION OF ECONOMIC POWER 

which produces steel, and the rolling mill from which issue both 
semi-finished and finished products. 

Productivity in the first stage has risen principally through the 
enlargement of the blast furnace. When the industry was in its 
infancy, the average blast furnace had a daily capacity of only 5 
to 10 molten tons of metal. A modern furnace can produce 1,000 
tons of iron a day.^ This greatly .increased production per furnace 
has not been accompanied by a corresponding increase in workers 
per furnace. The average mineral fuel stack required 116 men in 
1884 and 120 men in 1929.= 

Man-hour output has been increased also by the reduction of the 
number of men required to operate a blast furnace plant. This has 
been accomplished by (1) joint or integrated operation, either of 
several stacks in one plant or of a blast furnace in connection with a 
coke plant, or other manufacturing process, and (2) the introduction 
of labor-saving machinery. 

The contribution of jomt operation to productivity is apparent. If 
a plant operates more than one stack, the dire'ct labor is spread over 
more continuous operation. In handling materials much of the 
necessary equipment can serve two or more stacks as well as one; 
casting machines can be kept busy more continuously ; pumping power 
and flowing equipment can be better and more cheaply operated, 
while floating labor crews can handle maintenance and certain phases 
of operation to advantage with fewer man-hours per furnace than in 
a single stack plant. 

Three principal labor-saving devices have materially increased out- 
put per man-hour. The first is the skip hoist which covers a whole 
series of improvements in the method of charging a furnace and 
usually consists of two alternating hoists which convey the materials 
to the top of the stack and automatically dump them into the fur- 
nace. The process greatly reduces the amount of labor needed to 
charge a furnace. The pig-casting machine displaces a considerable 
number of sand cutters and iron carriers in the furnace crew and 
also affects indirectly the labor required in the iron yard, for it 
facilitates the mechanical loading of the iron pigs. A third labor- 
saving device consists of two complementary machines — the ore bridge 
and the car dumper. A car dumper with a crew of two men can 
handle all the ore for a one or two furnace plant as fast as it cai. 
be brought in. Then the ore bridge, with a crew of two operators 
and two oilers, removes the ore to the stock pile and at the same 
time keeps the bin supplied with ore for immediate use.^ 

The gains in productivity brought about by these innovations may 
be ascertained by a comparison of output per furnace and per worker 
in the years 1884 and 1929-30. In 1884 the average daily output per 
mineral fuel blast furnace in the United States was 54 tons ; in 1930 



» National Resources Committee, Teclinological Trends and Nationnl Policy, 1937, p. 331. 

* Harry Jerome, Mechanization In Industry, National Bureau of Economic Research, New 
York. 1934, p. 60. 

8 The above data, relating to productivity of labor in blast furnaces, is taken from U. S. 
Bureau of Labor Statistics, Productivity of Labor in Merchant Blast Furnaces. Bulletin 474. 
1928, pp. 24-47. This bulletin cites the reduction of the working day from 12 to 8 
hours which took place in 1923, as constituting an additional factor in the increase of output 
per man-hours. 



CONCENTRATION OF ECONOMIC POWER 237 

it was 584 tons. Similarly, the average output per wage earner in 
1884 was 170 tons; by 1929 it had risen to over 1,700 tons.* 

The growth of productivity in the second basic stage of the steel- 
making process has been somewhat retatded by the increasing demand 
for steel products of higher quality. The Bessemer process of in- 
jecting air through metal revolutionized the steel industry in the 
middle of the last century and reduced tremendously both the cost 
and the time of steel making. But the steel produced was not of a 
high quality and raw materials had to be selected with great care. 
The Siemens open hearth process, although more expensive and time- 
consuming than the Bessemer, produced a higher grade of metal and 
permitted a wider selection of ores. It has therefore practically sup- 
planted the Bessemer process. 

In 1890, 86.3 peJrcent of ingots and castings were produced by the 
Bessemer process and 12.0 percent by the. open hearth ; by 1939 the 
percentages had changed to 6.4 and 91.7, respectively.^ 

Principal among labor-saving developments has been the enlarge- 
ment of the open hearth furnace. Today the modal furnace for the 
industry has a capacity of from 75 to 100 tons per heat, though large 
producers in recent years have built open hearths with capacities 
ranging from 150 to 200 tons. "As in the case of blast furnace opera- 
tions, fewer wage earners are required to operate these larger units,'' 
one authority observes.^ Automatically uniform regulation of the 
open hearth furnace is performed by industrial instruments which 
increase fuel efficiency, reduce the time per heat (thus increasing 
capacity) and extend the life of the furnace by protecting it against 
excessive temperatures.^ In addition, much "of the handling ma- 
chinery and apparatus which have been applied to blast furnaces have 
also been extended to the open hearth furnaces. 

The continuous strip mill, introduced in 1927, is the outstanding 
improvement in the manufacture of steel. It transforms ingots into 
finished products by passing them through a series of automatic rolls, 
eliminating a large number of hand processes formerly necessary to 
roll the desired shapes.^ The importance of this development to the 
industry is obvious. A single continuous mill of the new type can 
now produce annually more than 400,000 gross tons of uniform gage 
steel sheets, equivalent to the normal output of 40 or 50 of the old- 
style mills. In 1929 approximately 1,400. old-style mills had an ag- 
gregate annual capacity of only 7,500,000 gross tons of sheets.^ 

The process has been rapidly extended during the last decade. 
Between 1926 and 1937, 27 continuous sheet and wide strip mills 
with an aggregate annual capacity of 13,119,000 gross tons were 
installed by the industry." One prominent engineer has estimated 

* Harry Jerome, op. cit., pp. 59, 60. 
, pmVin?,VS' Statistical Report of the American Iron and Steel Institute, 1933 and 1939. The 
1 emnining small proportion is compos-ed of crucible and electric processed steel. 
York T»34 "ef ' ^^<=^^'>''=*'tio° in Industry, National Bureau of Economic Research, New 
and'chalTg'inT TiS^;Vv''lP3"s^ National Research Project, Industrial Instruments 

»HanT'^IeroS^o*^^cft^'''*''6'4^^ *^ continuous strip will be found on p. Ill, supra. 
In;uS^ilfhi'#£'f^4fe^^ Committee, pt. 18, Iron and Steel 



238 



CONCENTRATION OF ECONOMIC POWER 



that 5 modern continuous process mills, put into operation during 
the early thirties, can now produce with only 130 men an amount 
of steel sheets which formerly required, under the old type process, 
over 4,000 men." 

At normal utilization of capacity, it is estimated that the continuous mill 
can effect a reduction in cost of $6 to $8 per ton, and can reduce the rolling- 
mill labor force per ton of finished steel by 96 percent ; of perhaps equal 
importance, these mills can roll the wide sheets and strips — more than 72 
inches in width — in demand for automobile body tops and other uses." 

The speed at which rolling mills operate has steadily been in- 
creased. In 1939, for example, several tin plate mills were ordered 
with a maximum speed above 2,500 feet per minute ; a year previous 
a maximum speed of 1,800 feet per minute was a practical limit.^^ 

Industrial instruments likewise increase productivity in the con- 
tinuous mill. In the annealing furnace of a continuous strip mill 
a single operator is able to survey from a distance a whole array of 
temperature instruments, and through their recordings to control the 
operations of the furnace.^* "The tensiometer provides a certain 
definite tension in the steel sheets running between the stands of a 
tandem-plate mill. The electrolimit gage * * * keeps a con- 
stant check on the sheets running through the mill." This constant 
measurement by instrument greatly adds to the efficient operation 
of a continuous strip mill and eliminates measurement by workmen. ^^ 

In analyzing the productivity of the iron and steel industry, it 
must be noted that the degree of productivity varies with the 
rate of plant operation. During times of curtailed production, 
productivity drops extensively. The necessity of preventing the 
deterioration of valuable equipment because of non-use and of 
spreading the labor over a smaller output, the desire to maintain 

Man-hours required to manufacture a gross ton of finished steel products at 
specified rates of operating capacity^ 

[Average for all finished steel products] 



Percent of total plant capacity 



Actual average man- 
hours 



Manufac- 
ture 



Adminis- 
tration 



Index of man-hours 
required 



Manufac- 
ture 



Adminis- 
tration 



55 to 60 percent 
50 to 55 percent 
45 to 50 percent 
40 to 45 percent 
35 to 40 percent 
30 to 35 percent 
25 to 30 percent 
20 to 25 percent 



34.43 
36.15 
38.22 
40.63 
42.34 
43.73 
45.10 
46.48 



1.32 
1.47 
1.73 
1.94 
2.25 
2.65 
2.92 
3.20 



1 U. S. Bureau of Labor Statistics, Man-hours of Labor per Unit of Output in Steel Manufacture, 
Monthly Labor Review, May 1935, p. 1161. 



11 Van Kleeck and Fledderus, editors, On Economic Planning, Covici-Friede, New York, 
1935, ch. 14. "Unused Productive and Technical Capacity in the United States," by 
Walter N. Polakov, p. 226. . , ,^ , . o. , r ., . 

^ Daugherty, DeCfhazeau, and Stratton. The Economics of the Iron and Steel Industry, 
McGraw-Hill, New York, lO.ST, vol. I, p. 19. 

"Steel, .January 22, 1940, "1939 Electrical Developments," by L. A. Umansky, p. 48. 

"See National Research Project, Industrial Instruments and Changing Technology, 

" Raymond F. Yates, Machines Over Men, Frederick A. Stokes Co., New York, 1939, illus- 
tration facing p. 27. 



CO^X'ENTRATION OF ECONOMIC POWER 



239 



in employment certain skilled workmen, foremen, superintendents, 
and other functionaries, and the disappearance of general economies 
resulting from large-scale operation require a comparatively large 
amount of labor per unit of output in times of restricted demand. 
The increase in man-hours required as plant operating capacity 
declines is shown in the foregoing table : 

Reflecting technological developments, the unit labor requirement 
index for the iron and steel industry (chart I and table 2) shows a 
general decline during the entire period, 1919-36, marked only by two 
minor upturns in the post-war depression of 1921-23, and again in 
1930-31. These may be ascribed to marked reductions in output, as 
shown in the table giving the relationship of man-hours required to 
rate of plant ojjeration. 

Chart I 

INDEXES OF UNIT LABOR REQUIREMENTS AND PRICES 

UNITED STATES 



INDEX NUMBERS 



IRON AND STEEL INDUSTRY GROUP 



INDEX NUMBERS 







PRIC 


E 
















































/ 




V 














































\ 


V 


A 






































i 






V 


V 


u 


<IT L 


ABOf 


REQ 


UIRE 


^ENT 


i 




















\ 


/ 










1 










\ 


V- 




^ 


^ 
























, 






I 










^^ 


v 










^ 


^ 


-;: 


< 




"V. 


V 


/'" 


*v 


./ 




— ^ 


















































-=^ 
















- 


- 
















u 




= 







1914 1915 1916 1917 I9IS 1919 1920 1921 1922 1923 1924 I92S 1926 1927 1928 1929 1930 1931 1932 1933 1934 1935 1936 1937 



In 1919, the unit labor requirement index was at the high level 
of 170.4 in terms of the 1926 base. In 1920, largely as a result of a 
material increase in production, the unit labor requirement index 
dropped to 133.9. It rose again in 1921, as has been noted, when 
production fell to the extremely low level of 42.0. As production 
began to increase in 1922, the unit labor requirement index dropped. 
It rose slightly in 1923 and then dropped steadily to 1929, when it 
reached a low of 86.9. In 1930, unit labor requirements once more 
turned upward as a result of decreased production, rising to 94.7 in 
1931. But as production picked up following the depression the index 
again declined. In 1935, the series stood at 80.1 and by 1936 it had 
dropped to its all-time low of 77.0. 



240 



CONCENTRATION OF ECONOMIC POWER 



Table 2. — The iron and steel industry 

[1926=100] 



Year 


Unit labor 
require- 
ments ' 


Price' 


Produc- 
tion' 


Year 


Unit labor 
require- 
ments ' 


Price ' 


Produc- 
tion' 


1914 




57.6 
61.7 
106.0 
168.9 
143.5 
125.7 
137.5 
101.9 
87.5 
111.7 
107.5 
101.5 




1926 

1927 

1928 

1929 


100.0 

87^5 
86.9 
92.0 
94.7 
90.8 
84.5 
85.3 
80.1 
77.0 


100.0 
95.8 
94.7 
96.2 
87.9 
83.3 
81.4 
81.8 
91.7 
92.4 
91.3 

107.5 


100.0 










1916 






1917 






118 8 


1918 






1930 




1919 


170.4 
133.9 
146.6 

. i.8 

. S 

103.. 


73.2 
90.9 
42.0 
73.7 
92.6 
77.6 
93.3 


1931 


57 1 


1920 


1932 


31.5 


1921 


1933 

1934 


51.9 


1922 . 




1923 


76.1 


1924 


1936 


105.8 


1925 


1937 














' Works Progress Administration, National Research Project— Production, Employment and Produc- 
tivity in 59 Manufcturing Industries, 1919-36, 1939, pt. II. 

2 Weighted average price of steel bars, plates, shapes, pipe, sheets, strips, wire nails, and tin plate, f. o. b. 
plant, compiled by American Metal Market. 

That the long-term decline in man-hour requirements is not due 
solely to variations in output may be seen by comparing two years of 
approximately comparable production rates. Production stood at 
73.7 in 1922 and at 76.1 in 1935; the unit labor requirements index, 
however, declined from 114.8 in the former year to 80.1 in the latter. 
In 1928 and 1936 production was at the relatively high levels of 108.1 
and 105.8, respectively. The index of labor required per unit stood 
at 87.5 in 1928 but by 1936 had dropped to 77.0. 

Productivity is rising rapidly in the iron and steel industry. The 
lowest level of the unit labor requirement index was reached in 1936 
and was 11.4 percent below that of 1929, while production was 8.3 
percent below 1929. 

Price. 

Prices may be obtained for many intermediary products in the 
fabrication of finished steel products, but the quantity of such prod- 
ucts sold on the open market is negligible in comparison with the 
quantity of finished steel products. The price index utilized in this 
study ^^ is a weighted composite of selected finished steel products — 
bars, plates, shapes, pipe, wire, nails, sheets, strips, and tin plate — 
compiled by the American Metal Market. 

The price index (1926=100) rose rapidly during the first World 
W.n- from 57.6 in 1914 to 168.9 in 1917. The post-war price decline 
sent the index to a low of 87.5 in 1922 but it recovered to 111.7 in 
1923. By 1925 the index stood approximately midway between these 
two extremes though at nearly twice the 1914 level. During the 
1920"s the index was remarkably stable, particularly if a median is 
computed between the short term low of 1922 and the short term 
high of 1923. In 1921, for example, the index stood at 101.9; in 1929 
at 96.2. This comparative stability was maintained during the 
depression. Between 1929 and 1932 the series fell only 15.4 percent. 
Bv 1934 it stood at 91.7, only 5 percent below the 1929 level, and 
by 1937 it had risen to 107.5, fully 11.8 percent above the 1929 level.^^ 

" Table 2, Chart I, .supni, p. 2.30. 

"The National Research Project presents no data for production in lO.*?? ; however, the 
Fedprnl Reserve Board index shows that steel ingot production was 5.3 percent higher in 
1037 than in 1936. Even this increase, if extrapolated, would still result in a figure of 
production lower in 1037 than in 1929, while price was, as noted, 11.8 percent higher. 



CONCENTRATION OF ECONOMIC POWER 241 

The industry has long been cited as an example of a controlled or 
monopolistic field. In 1935 the four largest firms produced 49.3 
percent of the Industrie's value of product; in terms of capacity, 
three firms (the United States Steel Corporation, the Bethlehem 
Steel Corporation, and the Republic Steel Corporation, each of 
which had a capacity of not less than 4,000,000 tons) held 60.5 per- 
cent of the total steelmaking capacity of the United States, In 
addition, seven smaller firms, with capacities ranging from 1,000,000 
to 4,000,000 tons, had 22.2 percent of the steel-making capacity. 
Thus 88.7 percent of the American steel-making capacity was in the 
hands of three large firms and seven smaller concerns. Dominating 
this picture of concentration, however, is one firm, the United States 
Steel Corporation, with an annual capacity of 27,342,000 tons, or 38.3 
percent of the industry's total.^^ 

The outstanding characteristic of the steel price structure is the 
industry's use of the basing point system. This method of arriving 
at uniform delivered prices was introduced in 1880 by the Carnegie 
Co. of Pittsburgh and was gradually adopted by the entire industry 
and extended to practically all steel products. The major change in 
the basing point system followed the Federal Trade Commission's 
order in 1924 requiring the United States Steel Corporation to cease 
and desist from the practice of quoting all prices from Pittsburgh 
plus freight to destination. It consisted in the substitution of a 
number of basing points for the single one formerly used. 

During the lifetime of the National Industrial Recovery Act the code 
for the steel industry (adopted in August 1933) embodied three prin- 
cipal characteristics: (1) The basing point practice, (2) the open price 
method of announcing quotations, generally identical, and (3) a 10- 
day waiting period, which rather effectively prevented any individual 
action moderating prices or terms in any manner. In addition, the 
code contained provisions prohibiting the construction of new capacity 
and levying fines on producers or distributors who deviated from the 
accepted price structure." 

After a short period of apparent uncertainty at the ena of the code 
period in May 1935, a system of ''open prices publicly announced" was 
reestablished in 1936 for the stated purpose of preventing price de- 
moralization through secret concessions, rebates, and discriminatory 
prices as between consumers. Iron Age described this as ''the most 
successful stabilizing movement the steel industry has experienced 
other than the Steel Code.'' -° The basing point system was continued, 
as was the open price method of announcing quotations.^! 

Among the other specific techniques used by the iron and steel in- 
dustry to prevent departures from the established price structure which 
might lead to price-cutting and price competition is the extension of 
price control to include jobbers' resale prices. During the N. R. A. 
code period, jobbers were severely fined by the enforcing agency, the 
Iron and Steel Institute, for deviating from the delivered prices set 
by the basing point system. Before and after the code, compliance 

>^ American n-on and Steel Institute, Iron and Steel Works Directory of the United 
States and Canada. 1935. pp. 316, 371. 

"•S. Doc. 159. 73d Cong., 2d sess., Practices of the Steel Industry Under the Code, 
letter from the Chairman of the Federal Trade Commission in response to S. Res. No. 166, 
1934. 

2» Iron Age, Jan. 7, 1937, p. 66. 

" U. S. Tariff Commission, I on and Steel. Rept. No. 128, second series, 1938, p. 352. 

277551—41 — No. 22 17 



242 CONCENTRATION OF ECONO.^lIC POWER 

with established prices has been enforced by a virtual blacklisting of 
those jobbers who quote prices at variance with basing point prices. 
The industry sells only to "recognized" jobbers and the withdrawal 
of "recognition" from a jobber usually spells his imminent economic 
demise. 

A second device to insure conformity with established prices is the 
use of "computed" rather than actual freight rates in setting delivered 
prices. The freight charges to be added to the base prices are taken 
from freight rate books compiled by the Iron and Steel Institute. 
Furthermore, the charges are calculated in terms of all-rail freight, 
which means that if a cheaper form of transportation is utilized the 
buyer nonetheless pays a delivered price based upon all-rail freight 
charges.-^ 

A third device is the joint determination by all firms of "extras" 
which are uniform to the cent. Delivered prices for steel products 
are made up of three elements: Base price, extras, and freight. 
Extras are those charges made by the steel producer for variations 
from a given norm in the physical specifications, chemical analysis, 
or processing of any steel product. Deductions are usually allowed 
when quantity purchases are made. Printed schedules of these extras 
and deductions are issued by ^the major producers in loose-leaf book 
form and ordinarily fluctuate even less than the base prices of steel 
commodities. Whether a change in extras accompanies or is made 
independently of a change in base prices, the net invoice price is 
necessarily affected. A reduction in base price accompanied by an 
advance in extras may produce a net increase in the invoice price. 

In the aggregate, extras amount to approxii.iately 10 percent of 
the delivered price, although for particular commodities the propor- 
tion may be much higher. According to an analysis made by the 
Department of Justice for the Temporary National Economic Com- 
mittee, the extras in a group of 10 selected conmioclities shipped in 
the month of February 1939 amounted to 9.9 percent of the total 
invoice delivered value of these products. In the case of certain 
sheets and strips, the proportions amounted to as nnich as 18.8 and 
29.7 percent, respectively. With respect to each of the products exam- 
ined, the extras and deductions annouriced by every manufacturer of 
a product were found to be identical. In some cases lags in publica- 
tion of changes in extras resulted in diff.^rences among producers for 
limited periods. Otherwi.se, extras and deductions, as announced, 
are uniform for all producers throughout the steel industry.-^ 

It is theoretically possible for the iron and steel industry to offer 
more varied, and perhaps lower, delivered prices to its buyers by 
increasing the number of basing points. There can be no doubt that 
buyers are in a more advantageous position today under the multiple 
basing point system than they were under the old Pittsburgh-plus 
system. Nevertheless, 

in certain respects the multiple-point system is a more efficient device for price 
fixing than the single-point system. Under it, a competitive insurrection may be 
quelled, with little danger to the entire price structure, just as a leak in one 
of the compartments in a modern ship may be confined to that compartment. 
Disciplinary measures, if necessary, may be focused upon any recalcitrant pro- 
ducer without undue extension of costly results. Such measures may consist of 

" Fedora 1 Trade Cominission. Heport in Response to P>pcutive Order of May 30. 1034, 
With Respect to the FasinK Point System in tlie Steel Industry. Nov. 30, 1934, pp. 21-22. 

23 Hearings Ijefore tiie Temporary National Economic Committee, Part 10, exhibit No. 
139.'), pp. 10724-10728. 



CONCENTRATION OF ECONOMIC POWER' 243 

redacting base prices (to disastrous levels) if the recalcitrant be a basing-poiut 
mill. If not located at a basing point, the recalcitrant may have his mill 
declared a basing point, whereupon base prices may be cut to any desired 
level.'' 

The significance of the basing point system in the determination 
of long-term price behavior is a matter of considerable dispute. On 
one hand, the Federal Trade Commission maintains that "the basing 
point S3^stem not only permits and encourages price-fixing, but that 
it is price-fixing." " On the other hand, the opinion of such students 
as de Chazeau and Stratton is that the basing point system is merely 
one means of exercising the type of control over prices which is in- 
evitable in any industry characterized by the type of markets and 
technological requirements as the iron and steel industry, ^"^ Finally, 
the industry maintains that the basing point system permits competi- 
tion as fully and completely as any other form of pricing. 

Although much may be said in support of each of these conten- 
tions, the fact remains that through the use of the basing point sys- 
tem — but not necessarily only because of it — the United States Steel 
Corporation formulates and establishes prices which are generally 
followed by its competitors and hence constitute the industry's price 
structure." 

Corroboration of this has been offered by Eugene G. Grace, presi- 
dent of the Bethlehem Steel Co., who testified that — 

When we put out a schedule, what we call our official prices, they usually rep- 
resent and are the same as our competitor has put in the market, and * * * 
as a general practice that pace is set * * * jjy the Steel Corporation. -^ 

The use of the basing point system, tlie success of techniques de- 
sio-ned to secure adherence to the going prices, and the price leader- 
ship exercised by the United States Steel Corporation, are the basic 
elements which determine the trend of iron and steel prices. 

Productivity and Price. 

Between 1919 and 1922, both the unit labor requirement and the 
price hidexes declined. Between 1923 and 1929 unit labor require- 
ments fell 26.9 percent (from 118 8 in 1923 to 86.9 in 1929) while 
price declined but 13.9 percent. In 1927-29 the price series varied 
only sliglitly from 95.8 hi 1927 to 96.2 in 1929. Yet, the unit labor 
requirement index declined from 98.7 in 1927 to 86.9 in 1929. For 
a short time during the depression, the indexes reversed this be- 
havior, for as production decreased, unit labor requirements rose, 
while at the same time minor do^^nward adjustments were made in 
the price structure. By 1933-34, the pre-depression relationship re- 
appeared, with the price index turning sharply upward in 1933 and 
the unit labor recjuirement series declining. This decline in the unit 
labor requirement index, interrupted slightly in 1934, continued 
downward until 1936, when it reached its all-time low of 77.0, while 
the pwce index, though relatively constant between 1934 and 1936, 
turned sharply upward in 1937, reachino- a position well above its 
1926 and 1929 levels. 



=* Federal Trade Commission, Report in Response to Executive Order of May 30, 1934, 
With Respect to the Basing Point System in the Steel Industry, 1935, p. 29. 

"Ibid., p. 35. 

^ Daiigherty, de Chazeau and- Stratton, The Economics of the Iron and Steel Industry. 
McGraw-Hill, 1937, vol. II, pp. 727-7.;2. 

^ Federal Trade Commission Decisions, vol. 8, p. 32, and hearings before the Tempo- 
rajy National Economic Committee, Part 5, testimony of Eugene W. Burr, attorney. 
Federal Trade Commission, pp. 1860-1901. 

^ Hearings before the Temporary National Economic Committee, Fart 19, Iron &tx<\ 
Steel Industry, p. 10602 



244 CONCENTRATION OF ECONOMIC POWER 

NONFERROUS METALS INDUSTRY 

Productimty. 

The most noteworthy advances in productivity in the noriferrous 
metals industry have been due to a new process of ore dressing or 
concentration, the selective flotation process.^^ Since this process 
has been applied in much the same manner to copper, lead, and zinc, 
analysis of the way in which it has affected copper is sufficient for 
present purposes. 

Prior to the introduction in this country of the selective flotation 
process in 1911, recovery of copper content in the concentration proc- 
ess was comparatively low. The new process, utilizing chemical at- 
traction instead of gravity and direct oscillation, was first used jjri- 
marily to reduce losses of copper in the residue of gravity plants. As 
improvements were made in the selective flotation process, however, 
it became the major process and by 1927 had practically supplanted 
the gravity method of concentration. Among the improvements was 
the use of more effective reagents which increased the speed of float- 
ing the minerals and resulted in greater capacities of flotation units 
and consequent reductions in costs. In general, according to one 
authority — 

Improved concentration processes employing a variety of reagents not only 
increased the output of minerals per unit of labor but made possible the utiliza- 
tion of inferior ores, increased the capacity of the mines, decreased the ratio of 
capital required per unit of mine capacity, and reduced the amount of power 
consumed per unit of product.'" 

The selective flotation process has also increased the efficiency of 
primary smelters and refineries, thus affecting directly or indirectly 
the three operations necessary to convert ore into pigs or slabs of 
refined metal : Concentration, smelting, and refining.^^ Its principal 
effects on smelting and refining are — 

^ The selective flotation process is unique because it utilizes chemical attraction rather 
than gravity concentration to separate the desired mineral particles in the ore from the 
waste. In one method, pulp — the mixture of the mineral and the chemical reagent — 
flows into a cell by gravity through a feed pipe, dropping on top of a rotating impeller. 
As the pulp cascades over the impeller blades it is thrown outward and upward by the 
centrifugal force of the impeller. Space between the rotating blades of the impeller and 
the hood permits part of the pulp to cascade over the impeller blades. This creates a 
positive suction, through the injector principle, drawing large and 'controlled quantities of 
air down through the pipe and into the heart of the cell. This produces a pulp thoroughly 
aerated with very small air bubbles. These exceedingly small, intimately diffused air 
bubbles support a 'large number of mineral particles and carry them to the surface. There 
the desired mineral is removed as froth. Though there are a number of variations from 
the flow of operation, the principal features remain basically the g,ame in every type of 
selective process flowsheet. _ ^ ^^ . ^ 

3« Worl<s Progress Administration, National Research Pro.iect, Effects of Current and 
Prospective Technological Developments Upon Capital Formation, by David Weintraub, 
1939, p. 11. 

=" Other developments In the concentration process, of less importance, include : 
"(1) Improved classification, which has resulted from the introduction of bowl-type 
classifiers; (2) simplication of crushing flow sheets by larger reduction ratios and pro- 
ductiou of finer crushing-plant products by the use of cone crushers; (3) use of -screens 
operated in closed circuit with the final stage of crushing, which has produced a uniform 
maximum size of feed 1o the grinding mills with marked benefits in grinding operations; 
(4) substitution of balls for rods in the secondary grinding mills or in both primary and 
secondary grinding mills and increasing ball and rod-mill speeds, which have resulted 
in increasing capacities of grinding circuits; (5) regrinding of flotation middling before 
the return of this product to the flotation cell, which has become common practice and 
has resulted in improved grades of concentrates; (6) separation of primary slime fol- 
lowed either bv the separate conditioning of the primary slime and ground product before 
the combined "products are treated in flotation machines or by the separate flotation 
treatments of the primary slime and ground product, which has resulted in improved 
conditioning of the flotation pulps and diMM-easod losses of copper in concentrator tail- 
ings; (7) improved conditioning of ore pulps in the treatment of massive sulphide copper 
ores and reground bulk concentrates, especially the thorough aeration of these pulps 
before flotation, which has improved the speed of flotation oporations when these products 
are treated." (Hureau of Mines, Bulletin 392, Concentration of Copper Ores in North 
America, by Thomas G. Chapman, 1936, .p. 7.) 



CONCENTRATION OF ECONOMIC POWER 245 

(1) It has increased the copper content of the concentrate thus 
reducing unit labor requirements in smelting and refining. For ex- 
ample, the amount of copper pig produced per man-hour from 100 
tons of concentrate with a copper content of 35 percent would of 
course be greater than that produced from concentrate of 30 percent 
copper content. 

(2) It has encouraged the development of grinding mills, pri- 
marily the ball mill, which not only grind at lower cost but more 
finely than was possible with the equipment utilized prior to the 
introduction of the flotation process. 

(3) It has encouraged the introduction and development of the 
i-everberatory furnace, which generally has replaced the blast furnace 
in smelting operations, with resultant reductions in the costs of 
smelting. 

Though the technological changes in smelting and refining have 
been overshadowed in importance by the new process of concentra- 
tion, one change which has greatly decreased unit labor requirements 
in these latter stages of the industrial process is the widespread adop- 
tion of conveyor systems which transport materials from the ore bin 
to the roaster and from the roaster to the converter. Trucks have 
been greatly improved and electrified and are used principally in 
moving finished pigs. They not only have mechanical arms which 
slide under and hoist the pigs but also can unload automatically by 
tilting to almost any desired angle. Modern trucks can handle 
metal in stacks of over 40 pigs. Thus the amount of labor involved 
in handling, lifting, and general conveying has been greatly reduced. 

A new type of continuous process, insuring continuity of operation 
and the elimination of complete cooling and reheating between opera- 
tions has brought about great economies in fuel, time, and floor space, 
and has contributed to better control. 

Automatic charging, improved furnace construction and more eifi- 
cient means of handling materials have accompanied the replacement 
of the blast furnace by the reverberatory furnace for smelting and 
have further increased productivity. 

In the refining process, centrifugal pumps for transferring the 
molten metal from kettle to furnace, improved stirring and skimming 
machines, straight-line casting machines, pig pullers and stackers for 
use in connection with the molding wheel, have each contributed sig- 
nificantly to the reduction of unit labor requirements. 

In addition to reducing labor expenditure per unit of refined metal, 
the developments described above have resulted in an increased purity 
of product, the recovery of a larger percentage of metal from the ore, 
extraordinary savings in fuel per pound of metal produced, and the 
recovery of by-products formerly wasted.^- 

Perhaps the outstanding feature of the behavior of the unit labor 
requirement index is its sharp and almost uninterrupted decline from 

^- The above description of technological developments in the nonferrous metals indus- 
try was taken primarily from Bureau of Mines Bulletin 392, Concentration of Copper 
Ores in North America, 1936, and Bulletin 381, Lead and Zinc Minins and Milling in the 
I nited States, 193.5. Additional data were obtained from H. T. Warshow, Kepresenta!Tve 
Industries in the United States, Henry Holt & Co., New York, 1928, pp. 363-366 ; and 
National Resources Committee, Technological Trends and National Policy, 1937, p. 168. 



246 



CO>TENTRATION OF ECONOMIC POWER 



1920 io 1930 (see chart II and table 3). The index fell from 178.9 in 
1920 to 85.3 in 1929, rising slightly to 89.4 in 1930. Its downward 
sweep was otherwise continuous during the decade except for an in- 
crease of 10 points in 1922 over 1921. With output during the depres- 
sion greatly curtailed, the unit labor requirement index ceased its rapid 
descent and rose slightly in 1930 and 1932, but by 1933 it had fallen 
almost to its 1929 position. Betw^een 1933 and 1935 the index moved 
upward only to turn dow^nward again in 1936. 

Chart II 
INDEXES OF UNIT LABOR REQUIREMENTS AND PRICES 

UNITED STATES 



3EX NUMBERS 



INDEX N'JMSERS 





PRl 

— ^ 


d. 
































III 






V 


\ 


/ 
\ 1 


A 


























1 
j 


1 1 : 
1 1 I 






' 




\! 




\: 


NIT LABOR REQUIREMEN1 


' 












j 


1 I ■ 




' 




r\ 


y 










! 
j 








\ 






\ 




















i : ; 




1 














\ 


\ 


V 




» 


' 










/ 














V'' 


/ 


r 




\ 


>> 


"^ 


^"^ 


V 1 


^>-KT ••■ 










i 




















\ 


! M./ ^ 


__ 






--^ 




1 


















\ '"■ 





ISIS 1919 1920 1921 1922 1923 1924 1925 1926 1927 V B 1929 1930 193! 1932 1933 1934 1935 1936 1937 



Although productivity in nonferrous metals varies greatly with the 
rate of operation, the decrease in unit labor requirements* in the last 
two decades is largely due to techniques wdiich, regardless of output, 
have increased labor's productivity. In 1924 and 1930 the indexes of 
production were almost identical but the unit labor requirement index 
stood at 116.8 in the former year and at 89.4 in the latter. 

Tlip long-time gain in productivity may also be observed by con- 
trasting the depj-ession year of 1933, when the production index was 
40.3 and the miit labor requirement index was 86.3, with the previous 
depression year of 1921, when production stood at 42.4 and the unit 
labor requii-ement series at 139.5. 



CONCENTRATION OF ECONOMIC POWER 



247 



Table S.—The nonferrous metals industry {primary smelters and refiners) 

[1926 = 100] 



Year 


Unit labor 
require- 
ments ' 


Price 2 


Produc- 
tion' 


Year 


Unit labor 
require- 
ments 1 


Price 2 


Produc- 
tion ' 






84.1 
121.5 
176.4 
181.6 
152.9 
120.3 
120.6 

81.4 

89.8 
100.1 

94.3 
103.2 




1926 


100.0 
94.9 
91.1 
85.3 
89.4 
79.4 
94.8 

1:1 

100.2 
94.7 


100.0 
r90.3 
96.9 
116.3 
84.2 
56.4 
40.1 
51.0 
58.0 
59.9 
66.4 
90.4 


100.0 








1927 


98.8 


1916 






1928- --- 


103.4 


1917 






1929 


112.0 






"/6."9 

69.2 
42.4 
58.3 
82.0 
89.9 


1930 


89 8 


1919.... 

1920 


167.2 
178.9 
139.5 
149.7 
129.7 
116.8 
106.8 




60.2 


1932 -... 


34.3 


1921 __- 

1922 

1923 

1994 


1933 

1934. _._ 

1935 


40.3 
46.8 
57.4 


1936 -- 


72.0 




1937 













1 Works Progress Administration, National Researcii Project— Production, Employment and Productiv- 
ity in 59 Manufacturing Industries, 1919-36, Part 11, 1939. . . , , ^ ^ . 

2 U. S. Bureau of Labor Statistics. Weighted composite of copper, lead, and zinc mgots, f. 0. b. reftnenes 

In 1936 the unit labor requirement index was at 94.7 which may be 
compared with its position of 94.9 in 1927 when production was 26.8 
points above the 1936 level. 
Pi-ice. 

The price index of non-ferrous metals is a weighted composite of 
copper, lead, and zinc ingot prices. Because of the greater importance 
of the copper industry, it tends to reflect primarily the movements of 
copper prices.^^ Nevertheless, the general trend of the index reflects 
fairly closely the movements of each of the individual series, as lead and 
zinc "prices have tended to behave in much the same manner as copper 
prices.'* 

Unlike the price index of the iron and steel industry, the non-ferrous 
metals index was not stabilized after the World "War at a level well 
above its pre-war position. Instead, it dropped precipitously between 
1919 and 1921, reaching in the latter year a low of 81.4 as compared 
with its 1914 level of 84.1. It then tunned upward, reaching a level of 
103.2 in 1925, though in 1924 it dropped slightly. In 1927 the index 
droi)ped to 90.8 but by 1929 it stood at 116.3, the highest level since 1920. 
During the first years of the depression the index dropped sharply to a 
low in 1932 of 40.1, slightly more than 50 percent under the previously 
recorded low of 1921. There followed a gradual rise in the next 5 years 
and by 1937 the index stood only 9.6 points below the 1926 base. 

A high degree of concentration exists in these industries. Accord- 
ing to a report on the copper industry by the United States Tariff Com- 
mission, "Three groups of capital, no one of which is dominant, to- 
gether control from 75 to 85 percent of the domestic smeltery capacity 
nnd output, the remaining capacity being distributed among half a 
dozen relatively smaller operators." In addition, "The three capital 
groups mentioned together control about 80 percent of domestic refin- 
ing output and capacity, and each has also financial interests in foreign 
copper production." '^ 



33 The weight of the copper series expressed as a proportion of the total value of all 
commodities in the Bureau of Labor Statistics wholesale price index amounts to 0.53, as 
compared to 0.14 for lead and 0.11 for zinc. 

"^ For a comparison- of world copper, lead, and zinc prices, see Melvin T. Copeland, A 
Kaw Commodity Revolution, Harifard University, Graduate School of Business Adminis- 
traiion, Bnsinos.s Research Studips No. 19. 1938. pp. 28-32. 

■"■= TT. S. Tariff Commission, Report to the United States Senate on Copper, Rept. No. 
29. 2d series, in'^" 



248 CONCENTRATION OF ECONOMIC POWER 

Measured in terms of the value of product, the concentration is so 
high in the copper and lead industries that no figures can be presented 
under the Census rule because of the possibility of disclosure of indi- 
vidual firms. In the zinc industry the four largest firms in 1935 pro- 
duced 64.0 percent of the industry s value of product.^^ 

The short-run or immediate determination of prices in these indus- 
tries is a result of exchanges in a simplified market ; the transactions, 
for the most part, are carried on in New York City, largely by tele- 
phone. The uniformity of quality of the product (achieved through 
the electrolytic process), the relatively small number of sellers and 
the almost equally small number of principal buyers have tended to 
bring about this simplified form of marketing. For example, in copper, 
"the dozen leading buyers probably represent easily three-fourths of 
the domestic consumption of the metal." ^^ 

The long-run behavior of prices reflects not only the immediate ex- 
change transactions but also the attempts of the producers to control 
prices either by the actual fixing of prices or by the supplemental 
technique of limiting the supply.^® 

In December 1918 the Copper Export Association was formed under 
the provisions of the Webb Export Act, for the purpose of fixing 
prices on foreign sales. Most of the principal producers took advan- 
tage of this opportunity, and agreed to sell all of their exports 
through the Association. Although some producers dropped out of 
the Association after the post-war depression, the objective of con- 
trolling the world price was brought nearer- to reality by the forma- 
tion in 1926 of a supplemental organization, the Copper Institute. 
By the latter part of the twenties, these two organizations had achieved 
virtually complete control over copper prices. One trade bulletin 
commented on the situation as follows : 

It must be remembered that a handful of men, controlling 75 percent of the 
production of copper in the United States, sit around a table and decide what 
price copper should sell at and any prognostication as to the future price of copper 
is in reality only a conjecture as to what is in the minds of those gentlemen.'" 

The maintenance of copper prices from April 1929 through March 
1930 at the arbitrary level of 17.8 cents a pound, a figure nearly 4 
cents above any level reached from 1921 to 1927, in the face of a grow- 
ing depression, has constituted an oft-repeated story of price-fixing.'^'^ 
This pegging of price has been attributed to the efforts of financial 
interests to increase the value of securities of copper companies in 
which they were speculating. One prominent financier expressed the 
opinion that a difference of a cent a pound in the price of copper meant 
a difference of roughly $1.25 a share in the value of Anaconda stock 
(or 3 percent of its $50 par)." 



3" See also Yearbook of the American Bureau of Metal Statistics, 1938, pp. 49 and 78. 

3' WarsslioWj op cit., p. 247. 

38 For details concerning price-fixing in the copper industry, see Hearings Before the 
lemporary National Kconomic Committee, Part 25; Frank A. Fetter, The Masquerade of 
Monopoly, Harcourt Brace, New York, 1931. pp. 107-108 : E. 0. Nourse and H. B. Drury, 
Industrial Price Policies and Economic Progress, Brookings, Washington, 1938, pp. 149- 
156. For a description of N. R. A. code provisions concerning limitations of production 
and allocation of sales, see National Recovery AcVninistration, Research and Planning 
Division, Preliminary Report on Codal Provisions Relating to Production and Capacity- 
Control, 19::.5, pp. 21-22 (unpublished). 

^ F. A. Fetter, The Masquerade of Monopoly, Harcourt, Brace, New York, 1931, p. 198, 
quoted from Commodity licvicw. J. S. Bache *: Co., Nov. 21, 1029. 

■""See Frank A. Fetter. Tlio Masquerade of Monopoly, Harcourt, Brace. New York, 1931, 
pp. 197-198 ; and E. G. Nourse and H. B. Drury, Industrial Price Policies and Economic 
Progress, Brookings Institution, Washington, 1938, pp. 149-156. 

" Nourse and Drury, op cit., p. 155. 



CONCENTRATION OF ECONOMIC POWER 249 

Regardless of the cause, the price established proved to be too high, 
and in April 1930 it broke and fell to an all-time low of 4.8 cents 
by December 1932. The abruptness of the decline was due not only 
to the extreme height at w^hich the price had been pegged and to the 
general demoralization of markets created by the depression, but also 
to the gradual increase which had taken place in the world's actual 
and potential supply. This increase was due, first, to a remarkable 
growth in the number of mines developed throughout the world. In 
Chile, new copper mines rich in ore deposits began producing around 
1920 and rapidly expanded their output. In Africa, mines in the 
Belgian Congo started to produce on a substantial scale in 1920 and 
within the next 5 years their output increased nearly fivefold. And 
south of the Congo; in Northern Rhodesia, several mines rich in ore 
began production in 1931 and 1932.'*2 

The second cause of the increased supply lay in the development of 
the selective flotation process which not only increased the workable 
output of existing mines but made possible the working of ores which 
previously had been considered of too low a grade. Thus the new 
process led to the operation of what is now one of the world's largest 
lead-zine- mines, the Sullivan Mine in British Columbia.^^ 

These increases in potential supply, coming at a time of seriously 
curtailed, demand, resulted in the accumulation of extremely large 
stocks. This impelled producers to seek ways of controlling, through 
a limitation of production, the prices at which copper would be sold. 
At the time the National Industrial Recovery Act was put into effect, 
stocks of copper had become extremely large. Despite the fact that 
copper production between 1931 and 1933 had been carried on at rates 
ranging from 40 to 65 percent below the 1926 level, copper stocks in 
1933 amounted to about 2 years' supply.^* 

To meet this situation, a sales pool was organized through which each 
producer was permitted to sell only the quantity allocated to him. The 
code also provided that producers should coordinate their production 
with their sales quota, so that they would both freeze their existing 
stocks and refrain from enlarging those stocks.*^ 

More recently, the increase in the' pj-ice of copper, the relatively low 
levels of stocks — despite the great expansion in potential supply and 
jDroductive facilities — and the allocation among "regular" customers 
of limited permissible purchases indicate that the corrective measures 
applied to this industry during the time of the N. I. R. A. have not been 
<^ompletely discarded. 

Productivity and Price. 

Except for a brief period after the war, the unit labor requirement 
and the price indexes have followed widely divergent paths. From 
1923 through 1928 the price index remained comparatively stable. Yet 
during this same period, the index of unit labor requirements fell from 
129.7 to 91.1, a decrease wiiich, as previously noted, can be ascribed 
only partially to a rise in production. 

This divergence was accentuated in 1929 when the price index 
reached its post-war high, while unit labor requirements fell to a new 

*2 Copeland, op. cit., p. 2. 
«Ibid., p. 3. 

** National Recovery Administration, Research and Planning Division, Preliminary Re- 
port on Codal Provisions Relating to Production and Capacity Control, 1935, p. 21. 
« Ibid., p. 22. 



d 



250 CONCENTRATION OF ECONOMIC POWER 

low. It slioiikl be noted that this upturn of price, beginning in 1927, 
aralleled the price-pegging activities of the Copper Institute. Thus, 
uring this period, direct price manipulation was accompanied on the 
one hand b}' rising prices and on the other b;/ decreasmg labor require- 
ments. 

The indexes continued to diverge during the depression, as price 
plummeted downward because of an increase in available supply and a 
drastic curtailment of demand, wliile the unit labor requirement index 
turned upward because of decreased production. 

As recovery set in, the indexes reversed their movements but main- 
tained their "inverse relationship. In 1036 unit labor requirements 
moved downward, while price began an upturn to an 8-year high in 
1937. If the indexes continue in theiv present directions, the 1927-29 
relationships may soon be reestablished. 

CEMENT INDUSTRY 

PrQcluctivity. 

The gains in. pioductivity in the cement industry reflect almost 
entirely the progress made in tlie manufacture of portland cement.'"* 
In the production of cement, certain raw materials (limes, silica, and 
alumina) are combined in specified pro])ortions, and are then mixed 
and ground very fine, either dry or in water. Then intense heat is 
applied to tlie mixtu.re in a rotary kiln and the resultant product is 
known as clinker. After the clinker has been cooled, a small amount 
of gypsum is added and this mixture is ground very fine to form 
poi'tland cement.'*^ 

Continuous technological improvements in the past 30 years have 
contributed to increased productivity in the industry. Between 1919 
and 1937 output of ih^ industry rose from 81,000.000 to 118,000,000 
barrels, an increase of 45.7 percent, while the average number of 
wage-earners rose only 3.5 percent from 25,524 to 26,-120.*^ 

Gains in productivity have resulted from the construction of new 
plants, from the enlargement of kiln capacity and the improvement 
of efficiency in old plants. During the period 1923-32, the increase 
in the manufacturing capacity of the industry from these causes was 
as follows : ^» 

Annual capacity 

Barrels 
New piants built 57,651,000 

New kiln capacity in old plants 30,032,000 

Increased efficiency in old plants 26,214,000 

Total increase 113, 897, COO 

The most important techiiological improvement in the industry 
has been the rotary kiln. Capable of large output, it created a de- 
mand for rapid mechanization in all parts' of the industry. The 



**TIiis study is concerned exclusively with portland cement, as natural masonry and 
puzzolan cements have comprised not over 2 percent of total cement production in tbe 
t'nited States in the past 20 years. (Works Progress Administration, National Research 
Project, Fuel Efficiency in Cement Manufacture, 1909-:]~y, p. 16, footnote 3.) 

*'' Work Projects Administration, National Research Project, Mechanization in the 
Cement Uidustry, 1089, p. 3. 

■•8 Bureau of the Census. Biennial Census of Manufactures, 1931 and 1937. 

•^National Recovery Administration, Division of Research and Planning, The Manu- 
facturing Capacity, Volume, and Costs of I'ortlnnd Cement in the United States, pre- 
pared by H. E. Hilts (Oct, 8, 1034), p, IC, unpublished. 



CONCENTRATION OF ECONOMIC POWER 251 

efficiency of the kiln has been steadily improved and its size greatly 
enlarged. In 1925 only 56 of the 810 kilns in the industry were 200 
feet or more in length. In 1935 there were 119 such kilns out of a 
total of 823.5° Today there are rotary kilns in operation over 400 
feet in length. That this development has increased productivity is 
shown by a comparison of kiln capacity through three decades. In 
1910 the average annual capacity of cement kilns was 134,000 barrels; 
in 1920 it was 184,000; in 1925, 225,000; in 1930, 290,000; and by 
1935 it had risen to 299,000.^^ 

To supply sufficient raAv materials for large-capacity rotary kilns 
hand quarrying and loading w^ere no longer adequate, and were re- 
placed by mechanical drills, industrial locomotives, and power shovels. 
More recent developments in drilling and blasting have been increased 
mobility of drilling equipment, more efficient application of power, 
and the use of detachable bits, mechanical sharpeners, better explo- 
sives, and improved methods of firing. Alloy materials have in- 
creased the durability of drills and are replacing all-steel rigging.^^ 

Power shovels and locomotives have advanced labor procluctivity 
by increasing the amount of material capable of being loaded and 
hauled, eliminating firemen, and by reducing labor indirectly asso- 
ciated with loading and hauling operations.^^ 

As it became possible to load large pieces of rock, the size of crushers 
had to be increased. This was effected by the use of vibrating screens 
between primary and secondary crushers, the "closed-circuit" tech- 
nique for secondary crushers, and by increasing the effective crushing 
areas of the crushers, with an accompanying reduction in power.^* 

Larger and more efficient grinding mills were also needed to prepare 
the increasing quantities of raw materials for feeding the rotary kiln 
and for grinding clinker into finished cement. In 1900 grinding was 
done in two stages hi two batteries of mills. Since then the old proc- 
ess has been completely changed. Today it is possible to grind more 
efficiently in one stage instead of two in grinding mills divided into 
compartments, each designed to grind material to progressively finer 
degrees. A modern compeb diy-^rinding mill, with air separators 
for classification and transfer of different sized particles into appro- 
priate compartments, can grind as much clinker .in the same time that 
24 mills of the type in use 30 years ago would have required. More- 
over, the clinker can be ground finer with a much smaller consumption 
of power." 

Within the cement mill proper improvements have been made* 
in handling and conveying. The outstanding innovation has been 
the mixing of finely powclered material with air and moving it in 
a pipe by the application. of pressure. This has resulted in economies 
in labor and maintenance costs because operation of the valves is 
automatic by remote control and there are no moving parts. 

Similarly, the use of cylindrical or "silo" storage tanks in place 
of square storage bins has increased the productivity of labor in 

^^ Works Progress Administration, National Research Project, Fuel Efflciency in Cement 
Manufacture, lft09-35, 1938, p. 39. 

siYV'ork Projects Administration, National Research Project, Mechanization in the 
Cement Industry, 1939, p. 53. 

53 Ibid., p. 38. 

« Ibid., p. 40. 

«Ibid., pp. 44-46. 

65 Works Progress Administration, National Research Project, Fuel Efficiency in Cement 
Manufacture, 1909-3.5, N. Yaworski, V. Spencer, G. Saeger, and O. Kiessling, J^38, pp. 



252 CONCENTRATION OF ECONOMIC POWER 

handling and conveying. Automatic bagging machines today are 
able to pack 2,000 bags of cement an hour. Formerly 5 gangs of 
men, working 10 hours, were needed to accomplish the same task.^^ 
The growth in size of equipment in this industry ha.s made it 
necessary to operate at a high degree of capacity to achieve fully 
the cost-reducing potentialities of the equipment. The tendency 
of man-hours required per unit of output to increase as the per- 
cent of capacity utilized declines — a behavior previously not«d in 
the iron and steel industry — is shown in the following table. 

Average effect of reduced capacity utilization on unit labor requirements of 
individual cement plants ' 

Man-hours 
required 
per unit 
Percentage of capacity utilized: oj oiUpnt 

100. 100. 

80. 108. 6 

60. , 120. 8 

40. - 140. 4 

20. 181. 

1 Works Projects Administration, op. cit., p. 23f. 

The unit labor requirement index for the cement industry reflects 
these improvements in productive processes by a continual, almost 
unbroken, decline (chart III and table 4). The drop in unit labor 
requirements was particularly abrupt immediately after the First 

Chart III 

INDEXES OF UNIT LABOR REQUIREMENTS AND PRICES 

UNITED STATES 



INDEX NUMBERS 



CEMENT 




1914 1919 1916 1917 1918 1919 1920 1921 1922 1923 1924 1925 1926 1927 1928 1929 1930 1931 1932 1933 1934 1935 1936 1937 



"o Works Progress Administration, National Research Project, Mechanization in the 
Cement Industry, pp. 69-71. 



CONCENTRATION OF ECONOMIC POWER 



253 



World War, the index falling from 143.7 in 1919 to 109.8 in 1922. 
The index then began a more gradual decline, which remained un- 
broken through 1929, except for a slight rise in 1923 and an abrupt 
decline in 1924—25. During the depression the series remained re- 
markably stable in view of the above noted tendency of labor 
required to increase as the percent of capacity in use declines. 
In 1935 a slight increase in the index took place. This, however, 
was more than compensated for by a sharp decline in 1936 to 67.2, 
its lowest recorded level. 

The long-time gains in productivity which have taken place, irre- 
spective of the rate of production, may be noted by comparing two 
years characterized by approximately identical rates of output. In 
1922 production stood at 70.2 and the unit labor requirement index 
at 109.8; by 1936 labor required per unit had decreased to 67.2, 
with production at 68.9. 

The rapid increases in productivity during the twenties are 
shown by contrasting 1925 when production was at 97.6 and the 
unit labor requirement index was at 100.6 with 1930, when the 
corresponding indexes were at 98.2 and 82.9. 



Table 4. — The cement industry 

[1926 = 100] 



Year 


Unit labor 
require- 
ments ' 


Price > 


Produc- 
tion' 


Year 


Unit labor 
require- 
ments ' 


Price > 


Produc- 
tion » 


1914 




55.0 
51.0 
65.4 
80.3 
94.6 
102.3 
117.2 
110.8 
103.5 
107.9 
105.7 
102.6 




1926. 

1927. 

1928 


100.0 
95.9 
91.6 
86.5 
82.9 
72.5 
74.3 
73.5 
72.0 
76.6 
67.2 


100.0 
96.7 
95.9 
91.8 
91.8 
79.4 
77.2 
86.1 
93.2 
95.3 
95.5 
95.5 


100.0 


1915 






1916 






106.8 
103.7 
98.2 
75.3 


1917 






1929 


1918 . 






1930 


1919 


143.7 
139.5 
121.2 
109.8 
111.6 
109.1 
100.6 


49.7 
60.5 
59.6 
70.2 
83.7 
90.8 
97.6 


1931.... 

1932 .. 


1920 


1921 


1933 


38 6 


1922 


1934 

1935 


47.2 
46 3 


1923 


1924 


1936 


68.9 


1925 


1937 













' Works Progress Administration, National Research Project— Production, Employment, and Produc- 
tivity in 59 Manufacturing Industries, 1919-36, Part II, 1939. 
' U. S. Bureau of Labor Statistics. 

During the thirties the decline in the two indexes was almost paral- 
lel. Production in 1931 was at 75.3 and the unit labor requirement 
index stood at 72.5 ; in 1936 production was at 68.9, while the unit 
labor requirement index had also declined to 67.2. 

Price. 

The price index utilized is the composite wholesale price series of 
cement, f. o. b. plant, computed by the United States Bureau of Labor 
Statistics. The use of a plant price in this study rather than the indus- 
try's customary delivered price makes more meaningful a comparison 
of the trends in productivity and in prices since it eliminates any 
variations which might occur in delivered prices owing to changes 
in freight rates. 

With 1926 as a base the index rose during the World War years 72 
percent (ch. Ill and table 4). The upward trend continued through 
the first 2 post-war years until, in 1920, the index stood at 117.2, or 



254 CONCENTRATION OF ECONOMIC POWER 

113 percent above the 1914 level. With the exception of 1923 there 
followed a period of marked stability in the price index during the 
twenties, the series dropping slightly to a level of 91.8 in 1929. Be- 
tween 1929 and 1932 the index declined only 15.9 percent and its 
1932 level of 77.2 was still 40.3 percent above the 1914 level. From 
1932 to 1934 the index rose rapidly to 93.2 and since 1934 it has 
remained on a plateau of 95.5. 

Concentration in the cement industry unfortunately can be meas- 
ured only in national terms. In 1935 the four largest firms produced 
only 29.9 percent of the industry's output but, as lias elsewhere been 
pointed out, this figure is probably a deceptive understatement. Com- 
petition among the major geographic areas of the country is greatly 
restricted because of the importance of freight costs; therefore, the 
degree of concentration within the geographic areas may be (and is) 
quite high.^^ 

The industry furthermore carries on in a vigorous manner certain 
group activities designed to control prices. These group activities 
are organized and directed by the industry's trade associations, prin- 
cipally the Cement Institute. 

The Cement Institute, an association of cement niiinufacturers, was or;?anized 
in New. York Citj% August 13, 1929, at a meeting of 27 individuals, representing 19 
cement-manufacturing companies, 12 of wliom signed articles of association, which 
were declared officially to be operative as of November 1, 1929. During 1930 the 
institute reached the peak of its effectiveness with a membership of 53 cement- 
manufacturing companies representing 65 percent of all cement produced during 
1930 in the United States."" 

During the first years of the depression, the membershiji of the organ- 
ization slumped sbmewliat, but at the time of the N. I. K. A. hearings, 
the institute claimed to represent 95 percent of .he members of the 
industry and 97 percent of its volume of business.^" 

The Cement Institute was formed shortly after the Cement Manu- 
facturers' Protective Association was dissolved. The Cement Manu- 
facturers' Protective Association was organized in 1916 and dissolved 
in 1924, while the appeal from the dissolution decree against it in 
the Federal Court was pending in the Supreme Court. An examina- 
tion by the Federal Trade Commisison has indicated that the Cement 
Institute is, to all intents and purposes, the same organization against 
which the Federal Government has proceeded on the basis of alleged 
violation of the antitrust law\*'° 

The industry uses a multiple basing point system to determine de- 
livered prices of its product. Today the cement industry has within 
the United States approximately 60 basing points, each with it„s base 
price. The use of this system has been traced back as far as 1902, 
at which time freight basing points were established in the Lehigh 
Valley district and at Hannibal, Mo., for use in the determination 
of cement prices. 



"National Resources Committee, The Structure of the American Ecouomy, Part 1, 1930, 
p. 115. 

B* S. Doc. 71, 73d Cong., 1st sess., Cement Industry, letter from the Chairman of the 
Federal Trade Commission, a Report Relative to Competitive Couaitions in the Cement 
Indusry, 1933, p. 98. 

"■» National Recovery Administration, Division of Review, History of the Code of Fair 
Competition for the Portland Cement Industry, by W. A. Janssen, Nelson Dickerman, and 
K. M. Richards (Sept. 16, 1935), p. 4. 

«» S. Doc. 71, 73d Cong., 1st sess., op. clt., pp. 98-101. 



CONCETsTRATIOX OF ECONOMIC POWER 255 

It has been contended that the delivered-price basing-point policy 
was adopted for the purpose of facilitating the disposal of a rapidly 
expanding production without breaking down a high mill-price level, 
because diversion of shipments from one territory to another took 
place when various levels of prices existed under a system of f. o. b. 
mill prices.*^^ That this purpose was largely accomplished, regard- 
less of whether its achievement depended solely upon the effectiveness 
of the basing-point system, is indicated by the behavior of the price 
index for cement. The series during the war increased sharply; 
during the twenties it was characterized by a marked degree of sta- 
bility near the wartime levels; and during the thirties its decline 
was moderate, the index regaining as early as 1934 a position above 
that of 1929. 

Once a multiple basing-point price policy had been generally 
adopted by the cement industry, manufacturers turned their atten- 
tion toward maintaining relatively high mill prices; efforts to attain 
this result through agreement have existed since 1904.^^ 

The principal methods used by the cement industry in effectively 
controlling prices via the basing point system are as follows :^^ 

(1) Rate books for every State in the country, showing the carload 
rate on cement from every basing point which might reasonably 
affect the price in question, have been distributed to all Cement 
Institute members. Each producer utilizes, in computing delivered 
prices, the rate given in the Institute's rate book rather than the true 
railroad or Interstate Commerce Commission rate. Thus these rate 
books are instruments for the maintenance of uniform delivered 
prices and not for the furnishing of complete and accurate 
information. 

(2) The producers have taken definite steps to discourage the de- 
livery of cement by truck, as they believe truck transportation might 
tend to create price competition. Consumers requiring delivery by 
trucks aie uniformly charged 15 cents per barrel more than customers 
using rail transportation. The producers have even gone so far as 
to forbid the loading of cement on trucks furnished by customers. 

(3) By means of a so-called "control clause" in their contracts, 
only delivered prices are quoted to the Federal Government. Under 
this clause the producers arbitrarily select the route and estimate the 
rate concession to which the United States is entitled. 

(4) The producers have combined to limit their sales only to those 
middle-men who fall within their agreed and arbitrary definition of 
a "cement dealer." With a few arbitrarily selected exceptions, no 
distributor who is not approved by the producers can purchase cement. 

(5) The producers, in order to meet the competition of a rela- 
tively small volume of cement imported from Belgium and Denmark, 
have threatened to boycott and have boycotted dealer-customers who 
trade in imported cement. They have in some cases resorted to 
espionage upon dealers, and have taken other steps, such as predatory 
price cutting, designed to minimize or prevent genuine price com- 
petition resulting from the importation of cement. 

«« D)i(l., p. 30. 
»a Ibid., i>. ;!6. 

!i3 PiHloral Trade Commission, in the Matter of tiie Cement Institute et al., Docket No. 
31C.?; M)?,l, pp. J4-17. 



256 CONCENTRATION OF ECONOMIC POWER 

By the end o^ 1931, the 5 largest cement-producing companies 
owned 59 plants strategically located throughout the country. It 
has been observed that if these 5 companies with their well-placed 
mills chose to cooperate directly, they could influence the price of 95 
percent of the entire cement output east of the Rocky Mountains.*^* 

Productivity and Price. 

Perhaps the most apparent characteristic in this comparison be- 
tween the trend of unit labor requirements and of prices is that the 
unit labor requirement index has continuously and steadily fallen 
much more extensively than has the price index. During the entire 
period 1919-36 the unit labor requirement index declined 53.2 percent 
while the price series dropped only 6.6 percent. The unit labor re- 
quirement index fell very greatly between 1919 and 1922, while the 
price series fluctuated both upward and downward with the 1922 
level slightly higher than that of 1919. During the years 1922-24 
the two indexes moved quite similarly, but in 1925 the unit labor 
requirement series fell abruptly while the price index suffered only a 
slight decrease. Between 1926 and 1929 the unit labor requirement 
index dropped 13.5 percent while the price series fell only 8.2 percent. 

An abrupt decrease in the unit labor requirement index during 
1930-31 was, however, accompanied by a decline in price; and in 
1932 as the unit labor requirement index moved up slightly, while 
price suffered a small decline. Beginning in 1933, however, the price 
series moved upward rapidly while the unit labor requirement index 
decreased, and as the price index became stabilized in 1935-36, the 
unit labor requirement series, after a short upturn in 1935, turned 
downward and reached in 1936 an all-time low of 67.2. 

Thus in 1936 the price of cement was 4.0 percent above its level 
of 1929, while the unit labor requirement index was 22.3 percent 
below its position of that year. 

MOTOR VEHICLES INDUSTRY 

Productivity. 

The development of the motor vehicles industry is almost synony- 
mous with the growth of mass i^roduction methods. In no other seg- 
ment of the economy have rationalization, specialization, integration, 
and general mechanization of the productive processes been more 
highly developed. 

As early as 1903, a multiple drill press had been introduced to work 
cylinder blocks and heads. In the same year a machine to grind the 
cylinders themselves, a lathe to turn camshafts, and a vertical turret 
lathe specially desimied to turn flywheels were also developed.*'^ 
Each year has brouglit gains in productive effectiveness, due not or\\y 
to the constant development of new machinery but to new methoc\s of 
material routing, machine layout, and assembling. 

Of more immediate importance to this study ai-e the technological 
developments which have taken place within recent years. A complete 
catalog of such developments would be entirely beyond the limits of 

** Federal Trade Commission, Price Bases Inquiry, the Baslug-Point Formula and 
Cement Prices, March 19;J2. p. 89. 
J*«*w Ralph C Epstein, The Automobile Industry, A. W. Shaw Co.. Chicago, 1928. p. 44. 



257 

this study. A few striking technological changes are listed below to 
give some conception of the reasons for the steadily declining' unit 
labor requirement index in the motor vehicle industry.^^ 

1. The complete elimination of wood parts froin automobile bodies 
has greatly reduced labor costs. For example, one manufacturer has 
entirely eliminated his wood mill which in 1928 employed 3,000 men. 

2. The one-piece stamping of the underbody has eliminated the 
building and assembling of 18 parts found in the 1929 under body, a 
labor saving of 50 hours in manufacturing and assembling. 

3. The one-piece stamping of the top has eliminated the manufac- 
ture and assembly of 15 parts and 53 hours of labor. 

4. In 1929 an automobile door was made up of 21 parts and cost $4 
to assemble. In 1935 the door consisted of one outside and one inside 
panel; machine welding the two parts together cost only 15 cents. 

6. The cost of the body for one popular-priced car was reduced $30 
between 1929 and 1935 because of economies similar to those mentioned 
and many others, notwithstanding the increased amount of steel used. 

6. Tungsten carbide tips in various cutting machines have increased 
the cutting speed of the machines from 180 to 580 feet per minute. 

7. In 1929 a well-known automobile manufacturer finished 100 eight- 
cylinder motor blocks on a given line-up with 250 men. In 1935 the 
same line-up finished 250 motor blocks with only 19 men, eacli operator 
performing about 20 percent more operations. The average pay per 
100 blocks in 1929 was $13.20; in 1935 it was $5.20. 

8. An automobile manufacturer saved 40 percent of labor in 1935 
b}' boring the wrist pin hole and the crank pin hole of a connecting rod 
in one operation. Diamond cutting tools doubled the speed of the 
machine and the holes were rough and finished bored in the same 
operation. 

9. Automatic buffing and polishing machines are rapidly eliminat- 
ing hand buffing and polishing. One double automatic buffing ma- 
chine installed in 1933, operated by 1 skilled mechanic and 4 helpers,, 
turned out 12 to 14 thousand pieces a day. Formerly 1 man, working 
by hand, turned out 600 to 700 pieces per 8-hour day. If. the machine 
were used full time, the labor of 150 men would be eliminated. A 
single automatic buffing machine, with 1 skilled operator and 2 lielpers 
likewise displaced 150 men. 

10. Early in 1934 a roller bearing manufacturer employed about 
1,100 men. By 1935 he had eliminated 150 men from his pay roll and 
increased production 15 percent by speed-up and labor-saving ma- 
chinery. 

11. The change from forging to casting made possible by new 
alloys has reduced the number of machine operations, and hence 
the amount of machinery necessary, as well as the cost of casting: 
compared with that of forging. On the new alloy cast shafts lO' 
opei;ations were eliminated with a 60 percent saving on labor. 

12. The use of conveyor systems has been constantly extended in 
the automobile industry to elimiiiate trucking and other hand labor. 



'^''' For a more detailed desorijitioii of recent technol'igioal advances in automobile pro- 
duction, see National Recovery Administration, Research and Planning Division. Prelim- 
inary Rejiort on Study of Regulniization of Employment and Improvement of Labor Con- 
ditions in the Aiitonutbilo Industry. 1935. appendix B. exhibit 3 6. 

27Tr.51~41-^No. 22 18 



258 CONCENTRATION OF ECONOMIC POWER 

The timing of parts conveyors have been considerably improved so 
that the arrival of parts at the assembling conveyor is synchronized 
to avoid storage at that point. 

Impressive as the technological advances in the automobile indus- 
try have been, new and diversified techniques are constantly being 
introduced. A number of significant nnprovements were made in 
the production of the 1940 models. Entirely new methods of con- 
vey orized assembly have been devised for a number of operations. 
For example, fenders are now handled by over-and-under chain 
conveyors, fitted with wood bucks upon which the fenders are placed 
for dinging, filing, and coarse and iine polishing as they pass a 
number of worlanen. Another improvement is a new method of 
automatic spot welding bodies which welds a 4-door sedan v-.t 222 
points simultaneously. " The under-body assembly, the right and left 
side-panel assembly and the roof assembly are clamped into an as- 
sembly fixture and transferred into the equalizing fixture of the 
automatic multi-welding machine, which spot welds the four huge 
super-assemblies together. 

The basic tool of all machine producrion, the automatic lathe, has 
recently been improved considerably in both efficiency and desigPi for 
use in the production of automobiles. The grinding of "green" rear- 
axle dry pinions has been eliminated by the installation of three 
high-production automatic lathes.^^ 

i^lthough these very recent improvements are not yet reflected in 
the unit labor requirement index, their introduction indicates that 
the rate of decrease in labor required per unit from 1932 to 1936 
has probably not abated. 

The index of unit labor requirements for the motor vehicle indus- 
try reflects closely the constant increases in productive efficiency of 
this industry (chart IV, table 5). The decline in unit labor requLre- 
ments from 1919 to 1923 was particularly precipitous, the index fall- 
ing from 183.8 to 112.5. This decrease has few parallels in any seg- 
ment of American productive enterprise. From 1923 to 192C the 
index flattened out somewhat, although there was a decrease of 12.5 
points in the period. This gradual decline continued to 1928 when 
the series stood at 93.6. An astonishing decrease in unit labor re- 
quirements took place in the next 3^ear^ the index falling to 78.5 in 
1929. When production fell to 39.8 in 1932, the unit labor require- 
ment series for the first time in its history moved upward; this in- 
crease was substantial as the series reached in 1932 a position only 
4.8 points below that of 1926. With increased output, however, the 
index again tamed downward to 78.5 hi 1933 and by 1936 reached 
an all-time low of 67.8. That there has occurred a sharp decrease 
in imit labor requirements, irrespective of the rate of production, is 
shown by a comparison of the years 1923 to 1930. In the latter year 
production was only 1 point above the former, while unit labor 
requirements had declined from 112.5 to 74.1. Production in 1935 
was 3.1 points above the 1926 level but the unit labor requirement 
index was 30.2 point? lower. 



American Macbinist. Oct. 4, J931», "Chrysler Gflros for '40." pp. .SOfl-820. 



CO^'CENTRATION OF ECONOMIC POWER 259 

Chart IV 



CO 
UJ 

on 
a. 

Q 

< 

CO 







bJ 




G:: 


(0 




UJ 


-ji 


1- 


a 


^ 


UJ 


CO 


cr 


a 




ijj 


fj^ 


H 


o 


2 


CD 


_i 



b 

CO 
Iti 

X 
LlI 
Q 

z: 




Table 5. — The motor vehicles industry 

[1926 = 100] 



Unit labor 
require- 
ments ' 



1914, ._ .1 

1915 ._..! 

1916 i 

1917 J 

1918 I 

1919 1 183.8 

1920 ' 169.2 

1<)21 -.., 138.1 

1922...- I 1-28.2 

1923 .-..! 112.6 

1924 i 110.9 

1925 '..I 10.5.8 



Produc- 
tion 1 



42.4 
48.0 
34.2 I 
54.0 1 
8r..3 ! 
76.4 ! 
94.6 I 



1929- 
1930. 
1931. 
1932- 
1933- 
1934. 
1935- 
1936- 
1937. 



Unit labor 
require- 
ments ' 



100.0 



98-8 
100. 1 
93 1 
92.7 
S8.9 

y.'-). 1 

93 8 
91 4 
94.7 
lOi.b 



Produc- 
tion > 



100.0 
86.2 
110.1 
132.6 
86.3 
63.1 
39.8 
54.2 
74.7 
103.1 
113.1 



1 Works Prog'-ess Administration, National Research Project— Production, Kmployraeni, and ProdJic- 
tivitj in .')9 Manufacturing Incu.strips, 1919-36, Part II. 1939. 
■' U £ Butcf I. ; Labor Statistics 



260 CONCENTRATIOJ^ OF ECONOMIC POWER 

Price. 

The price index for this industry is a composite of prices of th& 
standard passenger models of one of the principal low-priced makes. 
It thus measures the changes which have taken place in the prices con- 
sumers have to pay for a new low-priced automobile and reflects rather 
completely the extent to which producers have endeavored to reach 
new and wider markets by reductions in prices. 

Although tlie quality of the automobile has undoubtedly been greatly 
improved over this period, a considerable portion of the changes in 
models during recent years have affected only style and should, there- 
fore, not affect materially the adequacy of the price index."^ At all 
events price indexes, as stated in the introduction, are presented in 
this study merely to show the use made of one specific technique of 
enlarging the market, the reduction in the actual price. 

The movement of this composite index corresponds generally to the- 
observation of the Federal Trade Commission in its report on the motor 
vehicle industry, that — 

In general, prices reached the lowest point during the depression years following. 
1929, but increased somewhat in more recent years.^s 

After reaching a level in 1920-21 of more than 200, about 25 points 
above its 1914 position, the price index dropped to 106.9 in 1923. This 
post-war fluctuation marked the end of major changes in the price 
index, as from 1923 to 1937 the series was remarkably stable. From 
1925 to 1926 the index fell 4.1 percent, but for the next 3 years it varied 
only 1.2 percent. The index fell only 15.0 percent between 1929 and 
1933, and then rose from 85.1 in 1933 to 93.8 in 19:34. By 1937 the 
series was slightly above both the 1929 and 1926 levels. 

The industry is one of the most highly concentrated segments of our 
national economy. In 1935 the four largest firms produced 87.3 per- 
cent of the industry's value of product. 

The growth of concentration is also shown by the sales of the three- 
largest firms, which in 1925 were 63.6 percent of total sales, and in 
1937, 88.6 percent of total sales.;° 

Since the formulation of prices for automobiles is a complicated 
procedure involving estimates of demand, of material, labor and capital 
costs, and since the costs of various manufacturers could not be ex- 
pected to be identical because of the various methods of production^ 
similarity of prices among manufacturers would not be expected. The 
Federal Trade Commission contends that active price competition 
exists in the low-priced passenger-car field largely because "Ford has 
never been a member of the powerful motor-veliicle manufacturers' 
trade group, Automobile Manufacturers' Association, or its predeces- 
sor organizations, but followed an independent course with respect to 
production, price, and sales policies." '^ 

However, from 1934 to 1936 the differential in price for comparable 
Ford and Plymouth models was less than $10 and from 1936 to 1938 

8' As one student of the industry observes, "The industry has contrived ingenious ways 
for making easily noticed changes without incurring extraordinary expense. The shaiie- 
and color of fenders and hood, the length ,of the cowl, the jiggers on the radiator rap, the 
color of ihe upholstery, can all be altered annually with but little expense' (Waltoiv 
Hamilton and Associates, Price and I'rice I'olicies, "The Aufomobilo — A Luxury Bec.imes 
a Necessity," by Mark Adams, McGraw-Hill, Now Vork. 1938. p. 47). 

"H. Doc. 468, 76th Cong., 1st sess.. Federal Trade Co/nniission, report on Motor Vehicle- 
Industry, lO.-^O. p. 1062. 

.'"Ibid., p. 29. 

"Ibid-.l)?. 32-33. 



CONCENTRATION OF ECONOMIC POWER 



261 



it was rarely more than $3 except for 3 months during 1937 ^vhen 
there was a difference of $23 in the prices.'- This simihirity, espe- 
cially marked in recent years, is shown in the following table : 

Jf-door sedan — F. 0. B. factory price, including Federal tax 

FORD PLYMOUTH 



Date of change 


Price 


Model 


Date of change 


Price 


Model 


June 15, 1934 _... 


$591.98 
597. 09 
610. 73 
662. 18 
687. 71 
728.11 
723.08 


V-8 

V-8. 

V-8-85. 

V-8-85. 

V-8-85. 

V-8-85. 

V-8-85. 


February 1935 


$.585. 99 
606.51 
611.12 
665. 00 
730.00 
730.00 
726.00 


P J Bus. 
P 1 Bus. 


Nov. 11, 1936 


November 1936 


P 3 Bus. 


Jan 2 1937 


December 1936 


P 3 Bus. 


Aug. 2. 1937 




P 5 Bus. 


Oct. 27, 1937 


April 1938 


P 5 Road King. 


Nov. 4, 1938 













Source: Compiled from the Federal Trade Commission, Report on Motor Vehicle Industry, 1939, pp. 
S94-6. 

Chevrolet prices were practically identical with Plymouth prices on 
comparable models, which was in line with a statement made in 
December 1932 by a Chrysler Corporation executive that ''Plymouth 
will go if Chevrolet goes. We don't care about Ford. If we all 
stuck together the way we always have, we wouldn't care whether he 
came in or not, but eventually Ford woidd come in." " 

Thus the intensity of price competition in the motor vehicle indus- 
try seems no greater than in .any other concentrated industry surveyed 
in this study. 

Productivity and Price. 

Though the price index rose 12.4 percent from 1919 to 1920, it 
declined 47.7 percent from 1920 to 1923. The unit labor requirement 
index showed a steady decline from 1919 to 1923, falling 38. b percent 
in the 5 years. This followed rather closely the 5-year movement of 
the price index which registered a decrease of 41.2 percent. In the 
4 years following 1923, the unit labor requirement index fell more 
than the price series, and from 1927 to 1929 the unit labor require- 
ment index dropped from 98.9 to 78.5 Avhile the price index went 
from 99.8 to 100.1. 

During the depression the curtailment of output caused an upturn 
in the unit labor requirement series, while the price index declined 
sojj>ewhat. This reversal of trend was most conspicuous in 1932. when 
production had fallen to the extremely low level of 39.8, the ]>rice 
index stood at 88.9 and the unit labor requirement index at 95.2. 

After 1932, however, the unit labor requirement index resumed its 
downward trend, while the price series moved upward. By 193(3 the 
unit labor requirement index was at an all-time low of 67.8, \\hile 
by 1937 the price index had reached the highest level since 1925. 

It would indeed be paradoxical if the automobile industry, long 
regarded as the outstanding illustration of an industry which trans- 
lates in prices the benefits of technological advances, were to become 
characterized by a relation.ship between productivity and price such 
that the smaller the number of man-hours required in the production 
of an automobile, the higher its price. 

" Ibid., pp. 894-896. 
" Ibid., p. 34. 



262 CONCEISTRATTON OF ECONOMIC POWER 

CIGARETTE INDUSTRY 

Prodiictimty. 

The manufacture of cigarettes involves a considerable number of 
distinct processes. The almost complete mechanization of the prin- 
cipal operations in the productive process from the raw material stage 
to the tinai package has resulted in practically automatic manufac- 
ture with an extraordinary increase in labor productivity. Improve- 
ments are constantlj^ being made which increase the rapidity of manu- 
facture and decrease the unit labor requirements at the various stages 
of production. 

Some outstanding improvements are listed below: 

(1) The mechanization of the stemming process was accelerated in 
the middle thirties by the introduction of a relatively small stemming 
machine, which, although much less expensive than machines in use 
prior to that time, was fully as effective as the larger machines and 
did not depend for low unit costs upon a high volume of output. It 
has been estimated that these new machines have brought about a 
decrease of approximately 75 percent in the jiumber of workers for- 
merly employed in hand stemming.^* 

(2) Blending has been considerably speeded up by the use of a- 
wide conveyor belt on which workers place different types of tobacco,, 
each worker placing a certain amount of a single kind of tobacco at 
a definite rate of speed, instead of the old-style process of piling the 
various types of tobacco in layers in large bins or on the floor of a 
large room. 

(3) The new casing process by which tobacco is tumbled through 
a revolving, cylindrical drum and there subjected to the casing solu- 
tion sprayed from nozzles by compressed air, is measurably superior 
to the old method of dipping the tobacco into the casing solution. 

(4) A new type of cutting device whicii consists of several knives 
set in a rotating arbor and does not require re-sharpening every 10 or 
15 minutes (the knife-blades being cleaned and sharpened as the ma- 
chine operates), is almost completely automatic and has greatly in- 
creased labor productivity. 

(5) A new improved machine capable of making 1,200 to 1,500 
cigarettes per minute, while functioning on the same basic principle 
as the old type of machine that produced 700 to 800 cigarettes per 
minute, has been widely adopted by the industry in recent years and 
likewise adds to productivity. 

(6) Also, the cigarette industry, like manv other industries, has 
achieved considerable savings in labor expenditures by the extensive 
use of conveyor belts, supplanting trucking and manual handling. 

(7) A further considerable decline in unit labor requirements in 
the industry is due to an ever-increasing degree of integration and 
rationalization in the various stages of the productive process. Each 
function is timed almost to the second. Thus, in modern plants, the 
tobacco never stops from the time it is placed on the conveyor until it 
issues from the packaging machine. 

The mechanization of the manufacture of cigarettes has attained 
such a high degree of perfection that the National Research Project 
observed, "In 1937 the cigarette was manufactured and packed, the 

^♦National Recovery Administration, Division of Reviow, The Tobacco Study, 193(5, p. 
102-103 



CONCBTsTRATION OF ECONOMIC POWER 



263 



package sealed and stamped, and the stamp canceled by automatic 
machinery."'^ 

The decline in the unit labor requirement index for the cigarette 
industry from 281.7 in 19'20 to 100.6 years later is one of the mos\. 
amazing advances in productivity in the history of technolcgieai de- 
veiO{,ment (chart V, table 6). Bet\\'een 1926 and 1928 this abrupt fall 
was mterrupted by a slight nicrease, the index rising to 104.8 by 1928. 
In 1929. howerer/the index t-irned down again and declined steadily 
until 1933. The series rose from 65.9 in 1933 to 79.4 in 1934 and then 
fell to 71.6 in 1935. This interruption in the decline of unit labor re- 
quiieinents miglit be ascribed to the loss in the share of the industiy's 
business suffered during those years by the four largest manufacturing 
concerns at the hands of the smaller, then less mechanized, producers. 

Chart V 



CO 


UJ 
H 
\- 
IjJ 

1 

O 


I i s 


? S 8 s o s 




o 

0^ 


r 














II 










\ 


/ 




i 


a. 




1 






j 


/ 




S 


Q 




1 


1 




\ 


< 




1 


< 












\ 


/ 




* 


(/) 










'' 




/ 




i 










i 




/ 




i 
1 


7Z. 






. 






/ 








1 71, 


j 


\^ 


f 


J — 




1 t'^ 




r^ 










i 




\! 






III 








,- 


y 


J 






^ o 






\y/' 


^ \ 






gi 




I! 


1 
I 


■ 






• J^^^' 


1 


/] 


i 


i! k 


_j 


^"^T^ 




! 


...-'' j 


7 ^ 


lili 

!i 

% mi 
^' lilt 


!— 


t^ 


1 
i 




f' 






r? 


21 






1 


1 






1 s 

' I ■ 


^. 






i ./■■-.., 




i * 


CO 










1 « 












i 

i 1 




"'"'■- 1 








1 i 






X 


if: 






i 






1 


9 




^:_j. . 1 






'- 


I 


i i 


3 


o 


1 < 


\ ' 


I i 





" Works Progress Administration, National Research Project — Production, Employment, 
and Productivity in no Manufacturing Industries, 1919-36, Part II, 1939, p. 210. 



264 



CaxCENTRATION OF ECONOMIC POWER 



Table 



-The cigarette industry 

[1926=100] 



Year 


Unit labor 
require- 
ments > 


Priip^ 


Produc- 
tion' 


Year 


Unit labor 
require- 
ments 1 


Priced 


Produc- 
tion 1 


1917 




65.1 
91.8 
113.5 
124.0 
123.7 
106.9 
99.9 
100.0 
100.0 
100.0 
100.0 




1928 


104.8 
80.8 
76.7 
74.8 
72.7 
65.9 
79.4 

59; 2 


95.8 
9.5.4 
99.7 
103.4 
106.8 
86.6 
94.9 
9.5.1 
95.1 
97.3 


Il7. 9 


1918 




1929 


132 8 


1919 


261.8 
281.7 
209.2 
183.5 
147.5 
133.0 
117.8 
100.0, 
104.5 


57.6 
51.5 
56.6, 
60.6 
72.4 
78.9 

mo 

108.4 


1930 


134 3 


1920 


1931 


127.0 


1921 


1932, 

1933 


llo 7 


1922 


124.7 


1923 


1934 


141.2 


1924._._ 


1935. ..1 


151. 9 


1925 . . 


1936 


172.4 


1926 


1937 




1927 









1 M'orks Progress Administration, National Research Projecr, Production, Employment, and Produc- 
tivity in 59 Manufacturin?: Industries, 1919-30, pt. II, 1939. 

2 1917-26, Reavis Cox, Competition in the American Tobacco Industry, Columbia University Press, 
New York, 1933, p. 199. 1926-37, U. S. Bureau of Labor Statistics. 

During 1935-36, the index declined more sharply than during 1929-33, 
but not quite so sharply as during 1923-26. By 1936 the index reached 
an all-time low of 59.3. 

Since the years of slightly decreased production, 1931-33, were 
accompanied by a steady decrease in unit labor requirements, it may 
be assumed that the rate of operation, although it undoubtedly affects 
the degree of productivity, has not been primarily responsible for the 
tremendous decline in the unit labor requirement index. 

Price. 

The wholesale price index of the Bureau of Labor Statistics for the 
cigarette industry reflects the price movements of a number of the 
major cigarette companies, and as such represents the exchange value 
of the vast majority of cigarettes sold in this country. 

The index, however, goes back only to 1926. For the years 1917-26 
an index was computed from the list prices of one of the x^opular 
brands.'*^ 

In the first 4 years the price index increased 90.5 percent, rising 
from 65.1 in 1917 to 124.4 in 1920. It remained at this level during 1921 
but dropped during the next 2 years, reaching a level of 99.9 in 1923. 

For the ye&rs 1924-27 the index remained at 100. It dropped 4.6 
percent in the next 2 years but in 1930 began to move upward. From 
a level of 95.4 in 1929 the index rose to 106.8 in 1932. This increase 
came at a time of serious curtailments in economic activity and could 
not be maintained. The series broke in 1933, dropping to 86.6, the 



•» Reavis Cox, Competition in the American Tobacco Industry, Columbia University Press, New York 
1933, p. 199. The prices presented in this work are those of the American Tobacco Co. The propriety of 
extending the series in this manner is indicated by a comparisou of the two series for the years 1920-30. 



Year 


Bureau of 
Labor Sta- 
tistics 


Ami^rioan 

Tobacco 

Co. 


1926 


$5,659 
5.659 
5.442 
5.398 
5.645 


$ii. 689 


1927 


5.689 


192S . . 


5 413 


1929 


5 376 


1930 - 


5.645 







(Both series represent list prices less 10 percent trade discount and 2 percent for cash in 10 days, per M 
delivered,) 



CONCENTRATION OF ECONOMIC POWER 265 

lowest level reached since 1918. Following this break in the price 
structure, the index rose rapidly in 1934 to a level of 94.9. Since that 
time it has been maintained at approximately that same level, remain- 
ing unchanged for the vears 1935-36 at 95.1. and rising slightly in 193T 
to 97.3. 

The cigarette industry is one of the most highl}^ concentrated major 
industries in the country. In 1935 the four largest cigarette producers 
manufactured 89.7 percent of the industry's value of product. This 
figure of approximately 90 percent is very close to the estimate of the 
N. R. A. Division of Review that in terms of unit output the three 
largest manufacturers of cigarettes had practically 91 percent of the 
total production of small cigarettes for the years 1928-30.^^ 

The stability of price is accompanied in the cigarette industry by a 
striking degree of uniformity in the quotations of the principal pro- 
ducers. For almost six years, up to 1928, there occurred no change in 
price, but in April of that year the Reynolds Tobacco Co. reduced the 
price of Camels to $6 per thousand. The other companies followed, 
Chesterfield and Lucky Strike dropping to the same price. When the 
next price change took place in 1929, the Reynolds Co. continued its role 
of leader-and announced an increase to $6.40, which was followed by the 
other major companies. When the price changed again in June 1931, 
each of the major companies raised its price to $6.85."® 

This 45-cent increase in price took place at a time wdien leaf tobacco 
was at its lowest price since the war and wage rates were suffering from 
the impact of the depression. 

The advance proved to be injurious to the Big Three, for by the end 
of 1932 they produced only 81.4 percent of the total number of small 
cigarettes manufactured compared with 91 percent prior to the price 
increase. During this period of 18 months six smaller manufacturers 
increased their proportion of cigarette cutput from 8.2 percent to 16.8 
percent. Some of these smaller manufacturers were producers of 
lO-cent cigarettes.'^ 

The high price of $6.85 per thousand, established by the Big Three, 
was reduced in January 1933 to $6 and was followed one month later 
by a further reduction to $.'').50 per thousand. This was the lowest list 
price since the war and reflected a definite attempt by the Big Three 
and the P. Lorillard Co. to regain their former proportion of the 
business. 

The complete uniformity of price determinations by the largest 
cigarette manufacturers led the N. R, A. Division of Review to point 
out "the striking adherence to the same list price" and to observe that 
were it not for a few special allowances, the statement could be made 
'•'that competition was not on a price basis." ^'' That special discounts 
are not of suiTicient importance to invalidate the price index as a re- 
flector of the actual price trend is shown by the study of chain stores 
made by the Federal Trade Commission. The Commission found that' 
in 1930 the total allowances to chain stores by all tobacco manufacturers 
amounted to only 3,57 percent of their sales to these chains,®^ 

" National Recovery Administration, Division of Reviev, The Tobacco Study, 1936, p. 25. 

"Reavis Cox, op. cit., pp. 206-207. 

" National Recovery Administration, Division of Review, The Tobacco Study, 1936, pp. 

80 Ibid., pp. 24. 27. 

»i Federal Trade Commission, Chain Stores, Special Discount and Allowances to Chain 
and Independent Producers, Tobacco Trade. S. Doc. 86, 73d Cong., 2d sess., p. 48. 



2QQ CONCEiSTRATION OF ECONOMIC POWER 

Produ/)thnty and Pnce. 

From 1920 to 1923 the unit labor requirement and price indexes 
both declined though the drop in labor required was much more pre- 
cipitous than the drop in prices. In the next 4 years the unit labor 
requirement index declined, though its steady drop was interrupted 
by slight increases- in 1927 and 1928, while the price index rem.ained 
unchanged through 1927. In 1928- the price index dropped 4.2 per- 
cent and remained at about the same level in 1929, The unit labor 
requirement index resumed its rapid downward course viy 1929 while 
the price series began to move upward in 1930. This inverse rela- 
tionship continued until 1932. The extent of the divergence is seen 
by noting that in 1928 the unit labor requirement index was at 
104.8 and the price series at 95.8, and that by 1932 the unit labor 
requirement series had dropped to 72.7 while the price index had 
risen to 106.8. 

In 1933 the unit labor requirement index declined more sharply 
than it had in the years since 1929. But the price seii.is dropped 
even more precipitously. In 1934 the unit labor requireiuent index 
rose about 20.5 percent while the price index rose only 9.5 percent. 
This increase in the unit labor requirement index follov. cJ isx'id might 
be attributed to the above noted gain in the proportioiv of the indus- 
try's output made by the smaller, less mechanized, cigarette manu- 
facturing companies. 

By 1935, however, um unit labor requirement index was again 
moving downward, vfhile the price series increased slightly. By 
1936 the unit labor requirement index had fallen to its ull-time low 
of 59.3, while the price series, unchanged in 1935 and 1536, was, m 
1937, 2 percent above its 1929 level. 

ELEOrRlO LIGHT AND POWER INDUSTRY 

Productivity. 

This industry consists principally of commercial and rr.unicipal 
utilities engaged in the generation and distribution of electric cur- 
I'Sent for sale to public or private consumers- the generation"* of current 
for sale to other light and power establishments for distribution, or 
ill the distribution of c.;rrent generated by plants under other ov/ner- 
ship. Some of the principal changes in the industry s technology 
which have increased productivity in power generation transmission, 
and distribution are listed below. 

(1) In the generation of power, significant advance? in produc- 
tivity have been made in five basic ways. 

{a) Automatic stoking or the use of powdered fuel equipment 
for boiler furnaces is now almost universal in the industry. Of 
the coal burned for the production of electric power in 1928, 97.7 
percent was fired mechrmically. An example of the reduction in 
man-hour requirements resulting from the installation of mechani- 
cal stokers is afforded by one firm which estimated thai obout 200 
firemen, in three 8-hour shifts, would be required to i?.?x\ by hand 
the 24 boilers fed by mecrianical stokers ai tended by 14 employees.®^ 

(6) Boilers have been improved to use steam at pressures of 

^ U. ;S. Bureau of Labor Statistics, Monthly Labor Review, Aug i3t 1932, "Labor 
Productivity and Displacetatnt jr, the Electric Light and Power Industry, pp. 249 ff. 



CO^CE^^TRATION OF ECONOMIC POWER 267 

1,400 pounds per square inch. Less than two decades ago 350 pounds 
per square inch was the practical maximura.^^ 

(c) The heating surface of modern generatoi-s expressed in square 
feet is from 8 to 14 times greater than those built in 1900. The rate 
of evaporation of water per hour in the steam generator has increased 
from 10,000 ix)unds in 1880 to more than 1,250,000 pounds in 1935 
and steam generator efficiencies have increased from 65 in 1900 to 
85 and even 90 in 1935. Furthermore, in terms of unit capacity, 
modern high pressure and high temperature steam generators are less 
expensive than the old types of low pressure generators.'^* 

(d) Today steam turbines with capacities of 160,000 to 212,000 kilo- 
watts are in operation in central electrical stations instead of the 
formerly used recipl'Qcating steam engines with a maximum of 10,000 
kilowatts.®^ In the case of both generators and turbmes, the greater 
output per production unit has resulted in a decrease in the amount 
of labor recjuired and the improvement in quality and efficiency rating 
of the equipment has reduced materially the amount of labor needed 
in maintenance and servicing. 

(e) Industrial instruments have played an extremely prominent 
role in the gradual improvement and augmented control of operations 
in steam electric generating plants, contributing materially to a four- 
fold increase in fuel efficiency during the past four decades. Accurate 
measurement has made possible the utilization of higher pres- 
sures and temperatures. In addition it has served to protect the 
equipment itself, for if temperatures are excessive the equipment may 
be seriously damaged. Instruments are vitally important at eight 
different stages in the production of power in a modern steam electric- 
generating station and have contributed greatly to making it almost 
a completely automatic process. Since the industrial instrument has 
been perfected to a point where it not only records but also controls 
the functioning of equipment, the amount of labor required in super- 
visory and regulatory capacities has been >?:reatly decreased.^*^ 

(2) In the transmission of power, reductions in tne amount of 
labor have been effected principally through the establishment of un- 
attended, automatic substations and through the development of more 
durable types of equipment, thus decreasing materially the amount of 
labor needed in maintenance and repairs. 

(a) Between 1920 and 1^30 the number of unattended substations 
in 5 representative systems in the United States increased by 820 per- 
cent, while the number of attended substations rose by oniy 41 percent. 
By 1931, in one large system 72 percent of a total of 325 substations 
in operation were entirely automatic, wliiie the other 28 percent re- 
quired some attendances^ 

(h) Transformers with efficiencies approaching 100 percent have 
been built in tremendous sizes; for example, each of the seven single- 
phase transformers for Boulder Dam is 55 kilo- volt amperes, 284,000 
volts,^ the highest voltage ever used anywhere except in the laboratory. 
The oil-circuit breaker and improved insulators have reduced mate- 

83 Power, June 1934, p. 353. 

^ National Resources Committee, Technological Trends and National Policy, "Power," by 
A. A. rotter and M. M. Samuels, June 1937, p. 257. 

8» IbicL. pp. 257, 258. 

^ Works Progress Administration, National Researcli Frcject, Industrial Instruments and 
Changing Technology, 1938, pp. 60-61. 

*'' U. S, Bureau of Labor Statistics, Monthly Labor Tieview, "Labor Productivity and 
Displacement in the Electric Light and Power Indus crj.'' August 1932, p. 258. 



268 



CONCENTRATION 01" ECONOMIC POWER 



rially the labor required for repair. Cable for underground high voltr 
age transmission has likewise been materially improved through th& 
introduction of new insulating mediums.^^ 

(3) In the distribution of power, maintenance labor has been greatly 
reduced through the joint use of poles by telephone companies and elec- 
tric utilities in many parts of the country. As in the transmission of 
power, the amount of labor required in repairing distribution systems 
has been materially lowered by the development of improved and less 
expensive transformers. New and improved types of cables and wires 
have reduced the possibility of breakage or interruption. Finally, the 
amount of labor required in distributing power has been reduced and 
will probably be more significantly decreased in the future by the alter- 
nating current network system. If one or more supply circuits, or 
one or more transformers become incapacitated, the other circuits and 
transformers continue to supply energy to the network without inter- 
ruption. The development and extension of this type of distribution 
system has been termed "by far the most important feature in the 
progress of electric distribution in very recent years." ^^ 

Table 7. — The electric light and power industry 

[1926 = 100] 



Year 


Unit labor 
require- 
ments ' 


Price i 


Produc- 
tion ' 


Year 


Unit labor 
require- 
ments ' 


Price ' 


Produc- 
tion I 


1917 


131.4 


100.0 
105.5 
104.2 
107.0 
107.0 
104.2 
105.4 
104.1 
101. 1 
100.0 
97.8 


40.5 


1928 


99.2 
99.1 
103.3 
100.4 
92.9 
81.6 
73.9 

65! 4 
62.7 


95.8 
93.6 
91.9 
91.3 
91.9 
90.2 
86.4 
84.5 
82.2 
79.0 


121.9 


1918 


1929 


136.1 


1919 






1930 


135. 2 


1920 




56.7 
53.5 
62.5 
73.7 
7S.8 
88.0 
100.0 
111.8 


1931 

1932 

1933 


129.5 






114.6 




117.2 
97.5 
103.8 
104.6 
100.0 
100.4 


118.2- 


1923 


1934 


127. 5 


1924 


1935 


138.9 


1925 


1936 


160.7 


1926 


1937 


177. 1 


1927 











• Works Progress Administration, National Research Project, Production, Employment, and Produc- 
tivity in 59 Manufacturing Industries, 1919-36, Part II, 1939. 

'U.S. Bureau of Labor Statistics; 1917-22 December quotations. Retail Prices; 1023-37 annual average of 
quarterly quotations. Changes in Retail Prices of Electricy; 1923-38, Bulletin 664, 1939, p 2. 

The unit labor requirement index reflects these and many other 
changes which have reduced the amount of labor required in power 
generation, transmission, and distribution (chart VI, table T). Al- 
though no data are available for the years 1918-21, the unit labor 
requirements index fell from 131.4 in 1917 to 117.2 in 1922. In 1923 
the index fell sharply to 97.5, but from 1923 through 1t)31, the inde,M 
remained strikingly const nnt, never fluctuating more than 6.3 jDoints 
between any 2 years. The extension of electrical power facilities 
in those 9 years called for a considerable amount of labor in build- 
ing new generating systems, transmission lines, and distribution sys- 
tems. The index seems to indicate that the labor expended for these 
purposes during this interval offset considerablv the savings due to 
technical improvements inado in the actual equipment. BCit during 
the depression, when pressure for reduced rates resulted in a drive 
for lower costs and a relatively small amount of labor was required 



••National Resources Commitiee, op. cit., pp. 279, 280. 
«• Ibid., p. 286. 



ifi 

UJ 

o 

Q. 

o 
< 

I- 

z 

UJ 
UJ 

cr CO 
a < 

U CO 

UJ 

cc t 
< 



z 

CO 
LJ 
X 



CONCENTRATION OF ECONOMIC POWER 
Chart VI 

S 8 ? 8 8 



269 



if, 




in the extension of transmission and distribution systems, the labor- 
saving potentialities of improved equipment began to be reflected 
in a sharp downturn in the index of unit labor rec^uirements. Thug 
from 1931 to 1934 the index dropped precipitously from 100.4 to 69.3. 
During 1935-37 this downward trend continued but at a slightly 
less rapid rate. By 1937 the unit labor requirement index ha^ 
reached the all-time low of 62.7. That this decline during the thir- 
ties had not been due primarily to any change in production but 
rather to technological advances is indicated by the fact that in 
1935 production was only 2.1 percent above the level of 1929, while 
the unit labor requirement index in 1935 was 30.1 percent below that 
of 1929. 

Price. 

The price index used for the electric light and pc^ver industry 
is that of the_ United States Bureau of Labor Statist -cs, compiled 
"frnm rpciidpritial ruio. 'sr'hpdiilps for thp tvnical monthlv nsp of 9.^ 



270 CONCENTRATION OF ECONOMIC POWER 

kilowatt-hours. From 1917 to 1922 the series is compiled from sched- 
ules showing the price for 25 kilowatt-hours monthly as of December 
of each year in 32 cities eombined.®° The series from 1923 through 
1937 is compiled from retail prices of electricity in 51 cities, an an- 
nual average computed from the rate at the end of each quarter.^^ 
The two series are not strictly comparable but general trends may 
be noted. 

For an analysis of the price data, the typical bill and unit price for the monthly 
consumption of 25 kilowatt-hours was chosen as most representative of the 
use of current by the average customer, who does not use major appli- 
ances * * *. These data (average number of customers served at specified 
consumption levels), which were supplied by the Federcil Power Com- 
mission, reveal that an average of about 60 percent of these customers con- 
sumed 60 kilowatt-hours or less monthly. From analyses of consumption habits 
of thou>sands of residential customers, it was found that the average consump- 
tion for customers using lesc than 60 kilowatt-hours per month was about 
25 kilowatt-hours and that the great majority of these customers used current 
for lighting and small appliances only. This information supports the choice 
of the 25 kilowatt-hour service for purposes of price analysis."^ 

Consumption figures for 1937 show that 45.1 percent of residential 
consumers use 40 kilowatt-hours or less monthly.^^ Since 1937 was a 
period of relatively high domestic usage of electricity, in preceding 
years the percentage using 40 kilowatt-hours or less would b~^e even 
greater. It should be remembered, however, that in recent years, 
because of the wider use of large appliances, this series would tend 
to have a slight upward bias. (Yet the very fact that it was only 
af^ler electric refrigeration had assured an enlarged market that the 
cost for larger amounts of electricity declined illustrates the nonuse 
of price reductions by this industry as a means of expanding the 
m.arket for lighting and small appliances.) 

The price index for tlie first decade under consideration, 1917-26, 
was remarkably stable, fluctuating not more than 5.5 points between 
any 2 years. In 1927 a gi-adual decline began which took the 
index from 97.8 to 79.0 bv 1937. Throughout the 20 vears the decline 
from the high of 107.0 in 1920 and 1921 to the low of 79 in 1937 Avas 
only 26.2 percent. 

The electric light and power industry is, of course, a regulated 
monopoly, and, as such, its price behavior reflects not the force of 
competition but the conflict between the profit-seeking aims of the 
monopolists and the activities of regulatory bodies. Tlie regulation 
of public utilities has rested chiefly in the hands of State commis- 
sions. Tlierefore, any study of the price (i. e., rate) behavioi- of the 
electric light and power industry must at least note the part played 
by State commissions in rate setting. 

Among the obstacles in the path of effective rate reguLation has 
been the promulgation of the "fair return on a fair value" doctrine. 
As one authority has said : 

The interpretation of fair value has been appealed to the courts on so many 
occasions and the court holdings have been so consistent that one must say 
today that fair value is customarily interpreted in terms of reproduction cost 
new' minus depreciation, with more or less— and usually less — considenition 
of the amounts actually and prudently invested. 

The result of the above holdings is that something new under the sun has 
been .set up which is peculiar to the United States— that is, the art and science 



sou. S. Bureau of Labor StatlstKS, Retail Prices, December Issue of each year, 
w U. S. Bureau of Labor Statiancb, Changes in Retail Prices of Electricity,, 1923-38, 
Bulletin No. 664, 1939, table 1, p. 2. 
w Ibid., p. 6. 
•* Jbto.. p, 7, 



CONCENTRATION OF ECONOAilC POWER 271 

of valuation :t property in terms of reproduction cost new. It is a highly 
specialized art and science, and its refinements hrve been so developed that 
It has now become one of the chief means of inflating charges to the consumers 
of utility services. The practitioners of this art and science have been so 
successful in crystallizing assumed outlays in connection with the hypothetical 
formation and upbuilding of companies that one of the New York State com- 
missioners maintained, in ths hearings before the investigating commission 
in 1929, that it was unwist for the commission to press a rate case which 
might come before the courts, because it would probably result in a valuation 
of the property concerned that would call for an increase in rates rather than 
a decreai^e."^ 

Not only has there been established a hopelessly complex method 
of making evaluations,' but also a very rigid rate of return has been 
widely accepted. As Mosher goes on to say— 

A survey made in the twenties went to show that a large number of com- 
missions practically ■ accepted 8 percent as the fair return on the assumed 
iuve?>tment. Since the depression the rate has been reduced by at least 2 
percent and in some cases by 3 percent. 

The point at issue here is that a variable rate of return might appropriately 
be used for the purpose of rewarding companies which operate efficiently and 
penalizing those" which fail to maintain efiicieiu and progressive methods. 
The commission of Wisconsin should be cited as a notable exception, because 
it has been attempted to rate the eflSciency of the companies subject to its 
control. "^ 

Public utility rate regulation has also suffered greatly from the in- 
fluences and pressures exerted by political and party interests. The 
dependence of many commissioners upon party favor for election or 
appointment has constituted an opening through which utility inter- 
ests have been able to influence, and often dominate, the attitudes and 
activities of State commissions. 

Principal among the deficiencies in the legislation upon^hich State 
utility regulation is based has been the lack of legislative power to 
deal witli vast interstate holding companies. This limitation to intra- 
state jurisdiction has caused many students tc question the ability of 
State commi^sions ever to deal adequately Avith the problem of effec- 
tively enforcing rate regulation in the public interest. 

These and many other defects in utility regulation would tend to 
indicate that electric utility companies have been allowed considerable 
leeway in determining rates. Unhampered by the forces of competi- 
tion, except in industrial sales, the industry's price behavior reflects 
the policies of a natural monopoly whose power to impose prices at 
any levels it desires has only been partially, spasmodically, and inef- 
fectually offset by the efforts of public regulatory bodies. 

Productivity and Price. 

The productivity-price relationship in the electric light and power 
industry during the period 1917-37 has been characterized by a num- 
ber of conspicuous changes. Between 1917 and 1923, the unit labor 
requirement index fell from 131.4 to 97.5. The price series remained 
rather constant, increasing 5.4 percent between 1917 and 1923. Dur- 
ing 1923-2:") the unit hibor requirement index rose 7.1 points while 
the price series fell 4.3 pointy;. The unit labor requirement index 
turned downward again in 1926 and then remaine'1 r-'-ctically un- 
changed until 1930 when it rose 3.3 percent over t!ie r.>2G level. The 
price series, however, declined slightly from 1926 to 1930. In the 

»«W. E MoshPr, "Defeats of State Regulation of Public TJtilities in the' United States." 
The Annuls of the American Academy of Political and Sofi.ai Science, vol. 201, Januarr 
3.;j39, pu. 107-108. 

»tbld, p. 108- • 



272 CONCENTRATION OF ECONOMIC POWER 

depression, the imit labor requirement series fell considerably, from 
103.3 in 1930 to 81.6 in 1933. In the same period, the price index also 
declined but to a much lesser extent, falling from 91.9 in the former 
year to 90.2 in the latter, with the index for 1932 slightly higher than 
that for 1931. The decline in unit labor requirements continued 
through 1934 at an abrupt rate and at only a slightly decreased rate 
during 1935-37. The price index, while steadily declining during 
this period, did not decrease at all as rapidly as the unit labor re- 
quirement series. 

Thus, the tendency of the unit labor requirements index to decline 
more rapidly and extensively than the price series, first manifested 
during 1917-23, again characterized the 1930-37 movement. The sta- 
bility of the unit labor requirement index during 1923-29 may be 
ascribed in part to the wide extension of electric light and power 
facilities which took place during that period, involving a great deal 
of construction, erection, and installation of facilities, particularly 
for the transmission and distribution of electric current. By 1930, 
however, such facilities had been widely installed throughout the 
Nation, and consequently unit labor requirements began to decline. 
This fact, coupled with the technological advances which have de- 
creased the amount of labor required in servicing, maintenance, and 
current generation, may thus be regarded as the principal explanation 
for the exceedingly abrupt fall of the unit labor requirement index 
during the thirties, a decline which from 1930 to 1937 amounted to 
39.3 percent. 

THE NONCONCENTKATED INDUSTRIES 

COTTON GOODS INDUSTRY 

Productivity. 

Productivity in the cotton goods industry has been increased by 
varying degrees in its various productive stages, each of which has 
been characterized by dilferent types of technological innovations. 
Some of the important changes are indicated below. 

In the carding process productivity has been increased greatly by 
the introduction of automatic blending feeders which blend the raw 
cotton properly for the production of a uniform product. Machines 
have been introduced which open and bloom the cotton and remove a 
high percentage of dirt and leaf. These machines are arranged so 
that the cotton flows from one to the next automatically, and picking 
operations, formerly done with three separate machines requiring 
manual labor at each stage, are now combined in one continuous proc- 
ess. Although few changes have been made in the actual carding 
machine during the period studied, labor requirements in the carding 
operation have been diminished through the introduction of controlled 
humidity and cleaner air, which has resulted in a diminutioi] of 
difficult start-ups and has thus reduced the amount of manual labor 
required in the operation of a carding machine. 

The most significant techuologiGal advanc(^s in the entire productive 
process have taken place in the spooling and warping depai'tment. 
One of the most strikmg innovations has been the introduction of a 
traveling device on automatic spoolers which automatically repairs 
'^x\y breaks in the tlir":ids. In the w;\rping eqr.4meMi, three major 
improvements have been made: (1) tlie speed capacity of warpers 

Vine: Kppn inovoocprl iin tO 900 Vards per minute. «<? romnororl frv o 



273 

maximum of possibly TO yards per minute in 1919; (2) the addi- 
tion to the warper of a magazine creel which, by automatically making 
available additional material, permits continuous operation; (3) the 
enlargement of the section beams which has permitted the winding 
of a larger quantity of yarn per beam. As a result of these changes, 
practically the only labor required in the spooling process is the com- 
paratively small amount needed to supply the bobbins with yarn 
and to place them in a slide. Thus, a process which formerly re- 
quired a large number of workers has been changed to such an ex- 
tent that the process is today almost completely automatic. This 
elimination of the human element has made possible a great incresfcse 
in the speeds at which machinery in the spooling and warping depart- 
ment can be operated, and this in turn has led to a further gi-eat 
decline in unit labor requirements. 

In the weaving department, productivity has been increased pri- 
marily through improvements in loom construction and design. 
Breakage of threads has been greatly reduced through the develop- 
ment of more smoothly operating looms. The modern loom is bal- 
anced in such a way as to reduce vibrations; it has improved let-off 
and take-up motions, w^hich insure more evenly woven fabrics; a 
larger loom beam, w^iich reduces the amount of labor required for 
warp changing and loom adjusting. It has larger cloth rolls, which 
permit a reduction of labor in the removal of cloth from the loom. 
It has redesigned shuttle boxes, which permit the use of larger filling 
bobbins and thus reduce the frequency of bobbin transfers and loom 
stoppages at the time of transfer. It has improvements in warp 
stop motions for the purpose of reducing time required to repair a 
warp stop ; and i edesigned and improved shuttle tensions and shuttle 
eyes, which lessen the loom stops at the time of the bobbin transfer. 

The introduction of the automatically controlled shear and brush- 
ing mr.chine in the cloth room has increased the speed of,this depart- 
ment, as this machine automatically sews tlie cloth rolls together as 
they are received and removes loose and hanging threads, thereby 
materially reducing the work of inspectors. Productivity has also 
been advanced in this department through an increase in the capacity 
of the inspecting machines which handle larger cloth rolls direct 
from the shear, diminishing greatly the manual work incidental 
to the servicing of cloth rolls in and out of the machine.^^ 

The improvement of equipment in the spinning department dur- 
ing the past decade is indicated by the growth of textile production 
per spindle and by the increase in the hours run per spindle. These 
trends are shown in table 8. 

Although fluctuations in output affect productivity indexes, the 
trends of output per spindle, and hours run per spindle definitely in- 
dicate that during the past decade improvements in quality and opera- 
tion made in this department have been extensive. Production was 
higher in 1937 than in 1939 but both output per spindle and hours 
run per spindle were higher in the latter year than in the former. 

There are many factors in the industry such as the multiplicity 
of separate processes, and the lack of nitegration, which militated 
for- many years against any material reduction in unit labor require- 

»« The above analysis of productivity changes in tne uoiton goods industry was compiled 
from the study by Boris Stern, "Mechanical Changes in the Cotton-Textile Industry, 1910 
to 1936," U. S. Bureau of Labor Statistics. Montily Labor Review, August 1937, pp. 316- 
341, and from Harry Jerome, Mechanization in Industry, National Bureau of Economic 
Research, New York, 1934, pp. 80-87. 

277551— 41— No. 22 19 



274 



CONCENTRATION OF ECONOMIC POWER 



Table 



■Yearly production of cotton textiles, production per spindle, and hours 
run per active spindle, 1930-39 



Year 


Production In 
square yards 


Average 
number of 

active 
spindles ' 


Average 
number of 
square 
yards per 
active 
spindle 


Index 
(1-9.30= 
100.0) 


Hours 
run per 
average 

active 
spindle 


Index 
a«30= 
100.0) 


1930 ...- - 


6, 448, 392, 000 
6, 955, 391, 000 
6,278,222,000 
7, 866, 040, 000 
6, 878, 579, 000 
7,135,276,000 

8, 613, 837, 000 

9, 445, 736, 000 

7, 502, 168, 000 
9, 145, 765, 000 


27, 269, 470 
25, 674, 107 
23, 250, 757 
24, 873, 270 
25,119,435 
23, 421, 150 
23, 373, 147 
24, 079, 936 
22, 042, 442 
22, 306, 734 


236.47 
270. 91 
270. 02 
316. 24 
273. 83 
304.66 
368.54 
392. 27 
340. 35 
410.00 


100.0 
114.6 
114.2 
133.7 
115.8 
128.8 
155.9 
165.9 
143.9 
173.4 


2,813 
3,030 
3,020 
3,481 
3,014 
3,246 

3', 970 
3,444 
4.149 


100.0 




107.7 


1932 


107.4 


1933 


123.7 




107.1 


1935 


115.4 


1936 


139. & 




141.1 


1938 


122.4 


1930 


147.5 







" Based on 12 monthly reports. 

Source: Computed from data assembled by the Association of Cotton Textile Merchants, Journal of 
Commerce, February 19, 1940, p. 11. 

ments. The index exhibits striking stability between 1919 and 1927, 
reaching its high point in 1921 of 110.9 (chart VII, table 9). For 

Chart VII 



CO 



UJ 

o 


i 












«, 






1 












) 


/ 




i 


a. 












/ 


/ 




% 


Q 












'I 




i 


^ 














*»- 




i 


CO 
















> 


1 


H 










I 
^ 


/ 


.--''' 




8 


LlI 










\ ^^'' 






i 










^ 




t''\ 






% 










h 


J 






% 


3 t^ 
O < 
tU \7> 


■z. 

O 

1- 










) 






i.l 








; 


\ 


i 


ill 


UJ 








'• 


5 / 


\ 


1 


ii! 


O ^ 
CD -^ 










; 




1 


Mf 


o 

£ 

So 










i. 








in 


i 










\ 


y 






liX 


H 










-<-' 








1" 


»cr.' 






) 








3 








/ 








iH 


fe 




u, '*-■ 












i n 




r 







--.,^ 






% si 


fS 














'---.^ 




i l| 


X 

UJ 


















i 1^ 


i 
















! 


^11 


m 


I i 


\ \ 


\ \ 


\ % 


8 


I i 


I S 


I § 



CONCETs'TRATION OF ECONOMIC POWER 



275 



the next 4 years it was practically stable at the level of 95-96. In 
1926 and 1927, the index turned upward and in the latter year stood 
at 103.3. But in 1928, the series once more moved down to the level 
of 1922-25 and declined to 91.9 in 1929. This fall in the index was 
interrupted by an upward movement in 1930 and 1931. when the 
series reached 97.8. 



Table 9. — The cotton goods industry 

[1926=100] 



Year 


Unit labor 
require- 
ments ' 


Price ' 


Produc- ! 
tion > 1 


Year 


rnit labor 
require- 
ments ' 


Price ' 


Produc- 
tion' 






56.0 
52.3 
68.7 
98.7 
146.6 
147. 5 
190.7 
99.5 
104.3 
116.9 
114.7 
110.0 


1 


1926 


100.0 
103.0 
96.1 
91.9 
95.1 
97.8 
89.5 
84.2 
81.0 
75.2 
69.3 


100.0 
97.1 

100.4 
98.8 
84.7 
66.1 
54.0 
71.2 
86.5 
83.4 
80.3 
S4.3 


100.0 


1915 






1927 

1928 . .. . 


105.6 


1916 




i 


94.7 








1929 


100 7 






1 


1930... 


74.8 


1919 


108.1 
102.6 
110. 9 
96.7 
96.7 
96.4 
95.4 


85.8 1 

84.9 ; 
78.8 1 
91.4 1 

100.3 
&5.1 1 
98.0 


1931 


73.9 


1920 


1932 




1921 


1933 


85.7 




73.8 


1923 


1935 

1936 


77.2 




96.3 




1937 













1 Works Progress .Administration, National Research Project— Production, Employment, and Produc- 
tivity in 59 Manufacturing Industries, 1919-36, Part II, 1939. 
J U. S. Bureau of Labor Statistics. 

But beginning in 1932 the index started a decline unprecedented in 
its history and by 1936 unit labor requirements had reached the low 
level of 69.3. 

That no material changes in productivity had taken place prior to 
this 1932-36 decline is seen by comparing unit labor requirements for 
2 years which were characterized by nearly equal rates of production, 
1922 and 1928. In the former year production stood at 91.4 and the 
unit labor requirement index at 96.7. while in the latter year the 
indexes stood at 94.7 and 96.1, respectively. This stability in unit 
labor requirements denotes the absence of those change^ Avhich bring 
about any widespread reduction of output per man-hour. That this 
has been completely reversed in the period 1932-36 is apparent if 
1928 is compared with 1936. In 1928 the indexes, as noted, were at 
94.7 and 96.1, while by 1936 the production index was at the level of 
96.3 and the unit labor requirement index had dropped to 69.3. 

Indicative of the rapidity with which unit labor requirements de- 
clined during 1932-36 is the fact that although production in 1920 
was 4.5 percent higher than in 1936, unit labor requirements in the 
latter year were 24.5 percent below the level of the former year- 

The price index used in this analysis of price behavior in the cot- 
ton goods industry is the cotton goods subgi'oup of the United States 
Bureau of Labor Statistics (ch. VII, table 9). 

This series rose rapidly during the World War from 56.0 in 1914 
to 146.6 in 1918 and continued its precipitous rise through 1920, when 
it reached the all-time high of 190.7. The index fell abruptly in 1921 
to 99.5, or within 43.5 points of the pre-war level. The index rose 
again in the next 2 years, and by 1923 had reached 116.9, but in 1924 



276 CONCENTRATION OF ECONOMIC POWER 

it turned downward, and, with the exception of 1928, declined steadily 
through 1932. Between 1923 and 1926 the index dropped from 116.9 
to 100.0. From 1926 to 1929 the index was comparatively stable, al- 
though in the latter year it was 1.2 percent below the level of 1926. 
The most pronounced stage of this decline took place during the 
depression, when it declined 45.3 percent between 1929 and 1932. 
The upturn between 1932 and 1934 brought the index only partially 
up to the predepression level, and, beginning in 1935, another decline 
set in, interrupted by only a slight increase in 1937. Indicative of the 
extent of the long-term decline is the fact that in 1937 the series stood 
at a level 14.6 percent below the 1929 position and 27.8 percent below 
the 1923 level. 

The cotton goods industry is among the least concentrated of the 
Nation's important industries. In 1935, the four largest firms pro- 
duced only 8.4 percent of the industry's value of products. 

This lack oi^ concentration arises out of the nonintegrated char- 
acter of the industry. Few of the mills produce a finished product 
from the standpoint of the final consumer. For the most part, they 
merely supply the raw material to the next processing agency. The 
typical course of a cotton goods product is substantially as follows: 
(1) From spinning mill to weaving mill, through the agency of a 
commission house or a yarn broker; (2) from weaving mill to con- 
verter via a commission house; (3) from converter to finisher, a 
transfer of operation but not of ownership; (4) from converter to 
garment manufacturer, wholesaler or retailer, either with or with- 
out the services of a broker. As a result of this separation of func- 
tions, the industry is characterized by an extremely complex market- 
ing structure, involving numerous purchases and sales at every stage 
of the productive process.^^ 

The lack of integration among the productive processes has been 
one of the most important, if not the principal, cause of the indus- 
try's highly competitive nalure. Mill units are not only widely 
scattered geographically, but the majority are also smalf in size, 
independent in ownership and specialized as to character of output. 
As the Cabinet Committee said, "The cotton textile industry is faced 
with difficulties common to all industries characterized by severe 
competition, causing constant strains and pressures." ^^ 

The severity of this competition has be^n steadily increasing. The 
industry greatly over-expanded its plant and equipment during and 
immediately after the first World War. With the slackening of 
demand at the end of the war, the competition for remaining mar- 
kets became extremely intense. In addition, the industry has been 
confronted with a slowing down in the j^er capita consumption of 
textiles. Despite an increase in population, the total consumption 
of textiles in general, and of cotton textiles, in particular, has re- 
mained nearly stationary during the last decade and a half. Fur- 
thermore, substitute textile fabrics, such as rayon, have captured 
sizable segments of the market formerly held exclusively by the 
cotton goods industry 

^ United States Cabinet Committee to Investigate Conditions in the Cotton Textile Indus- 
try. A Report on the Conditions and Problems of the Cotton Textile Industry, Made by the 
Cabinet Committee Appointed by the President of the United States 1935. p. 119 

"sibid., p. 1. 



CO^'CE^TRATION OF ECONOMIC POWER 277 

The industry's large amount of plant and equipment has made 
possible the existence- of a practice which results in intensifying 
the severity of price competition, that of the constant purchasing 
and repurchasing of bankrupt firms. As a result of this practice 
new competitors may arise in every period of temporarily expanded 
markets, with the added competitive advantage of being placed in 
operation under conditions that involve virtually no capital charges. 

The individual nature of the various stages in both the produc- 
tive process and marketing structure militate against the existence of 
knowledge on the part of most producers as to the size of the market. 
In consequence, there is no close adjustment of supply to demand in a 
short-run period, and periodical overstocking occurs which results 
in serious pressure on prices. 

These factors all tend to cause the existence of a relatively large 
number of very weak firms. Such finns are in continually urgent 
need of volume at almost any price above lowest attainable costs. 
Thus, the existence of these weak finns constitutes in itself a source 
of intense price competition. 

Among other sources of competitive rivalry which arise out of 
the non-integrated nature of the industry is the prevalence of very 
lenient sales policies. One authority has said that 

producers and distributors * * * tend to become lax in their credit policies 
in order to increase sales, overlooking the fact that at the same time they 
enlarge their losses from bad debts. Wholesalers are accused of being too 
ready to finance retailers whom they hope to see develop into steady custom- 
ers. * * * Practically all .sellers (of every kind) feel a heavy pressure 
to accept trade abuses perpetuated by their customers, because they are fear- 
ful of "losing business," even though a substantial part of the sales * * * 
yield losses rather than profits. 

Producers, for the most part, possess a notoriously "inadequate 
knowledge of their costs. They are accused of setting their prices 
at levels which trade custom, or the demands of their customers, or 
simple hunches may dictate, without regard to whether these prices 
actually provide a profit." Also, producers in the industry have been 
forced to assume an increased degree of risk due to the practice of 
purchasing small quantities at frequent intervals, a policy which 
increases handling charges and forces producers to provide storage 
formerly provided by middlemen.^^ 

These features, which have contributed to the development of a 
high degree of competition in the industry, have also been responsible 
to a considerable extent for the long-term decline in price. 
Productivity and Price. 

An increase in the price index in 1920 was accompanied by a decline 
in the unit labor requirement series, while a sharp decrease of 47.7 
percent in the price series in 1921 was paralleled by an increase of 
7.4 percent in the unit labor requirement index. During the next 
4 years, the unit labor requirement index was practically stable, while 
the price series rose for 2 years and then declined in the next 2 years. 
In 1926 and 1927, unit labor requirements rose while the price index 
declined. At the beginning of the depression, a great decline in prices 
was accompanied at first by a slight increase in the unit labor require- 

„.^ ?^?^is Cox, The Marketing of Textiles, The Textile Foundation, Washington, 1938, pp. 



278 CONCENTRATION OF ECONOMIC POWER 

ment series but beginning in 1933 the latter index decreased steadily. 
In 1933 and 1934 the price series moved upward, then dropped again 
in 1935 and 1936, and rose slightly in 1937. 

In 1936 both indexes were at levels well below those of the twenties; 
except for the depression low of 1932, the price index had reached a 
level below any previous position since the pre-war period. 

The outstanding feature of the relationship between the two in- 
dexes during the entire period is the degree to which the general 
long-term movements have paralleled each other, particularly in their 
relative stability during the twenties and their downward movements 
of the thirties. Despite short-term inverse relationships, the indexes 
over the entire period behaved in a strikingh' uniform and parallel 
manner. 

WOOLEN AND WORSTED GOODS INDUSTRY 

Productivity. 

The fundamental processes of manufacture in the woolen and 
worsted goods industry are very similar to those of the cotton goods 
industry. By far the most important from the viewpoint of man- 
hour requirements is the weaving division. Improvements have also 
been made, however, in the spinning, the carding, and spooling and 
dressing department. 

The principal technological advances in the industry may be sum- 
marized as follows : 

(1) In the weaving department, the principal development has 
been the introduction of the automatic box loom, which replenishes 
the filling supply without stopping the loom. Thus, the amount of 
labor required in replenishing the filling supply has been greatly de- 
creased. Another labor-saving feature of the new loom is that it 
automatically stops whenever a breakage occurs and indicates the 
position of the broken warp thread. Consequently, the number of 
looms which a weaver may watch has been greatly increased. These 
economies have proved so desirable to the industry that by 1938, 
25,000 of the industry's 44,000 woolen and worsted looms were of the 
automatic type.^ 

(2) In the spinning department, the ring frame spindle has re- 
placed the old style mule and has the great advantage that a much 
larger bobbin may be utilized on it. The increased size of the bobbin 
has resulted in eliminating a very large amount of the labor required 
in doffing and in the subsequent spooling and dressing operations. 

(3) "Probably the most outstanding mechanical change affecting 
the production of yarn has been the replacement of the small bobbins 
or packages formerly used on the mule spinning frame for woolens 
and cap spinning frame for worsteds by the larger packages now used 
on the ring spinning frames. Although the yarn-making processes 
have not undergone aiiy inherent changes, the larger package resulted 
in greatly reducing the amount of doffing required in 1936 as compared 
with 1910.'' 2 



» H. E. Michl : The Textile Industries, the Textile Foundation. Washington, 19.'?8. p. 224. 
* Boris Stern. "Mech.Tnical Chanijes in the Woolen and Worsted Industries." 1910 to 1936, 
U. S. Bureau of Labor Statistics, Monthly Labor Review. January 1938. p. 63. 



CONCENTRATION OF ECONOMIC POWER 279 

(4) This utilization of larger packages has made possible a com- 
plete change in the mehod of warp dressing in the spooling and dress- 
ing department. An entire intermediary process has been eliminated ; 
formerly, warp yarns were first wound from small bobbins to jack 
spools and then to creels, preparatory to dressing and being wound on 
the "loom beam. Today, high speect cone winders transfer the yarn 
from the spinning bobbins directly to the cones, which contain more 
yarn and are more suitable for dressing on the high speed warper. 

(5) In the blending and picking department, marked advances 
have been made by a continuous process which has eliminated the 
secondary reprocessing formerly needed. Large feeders mix. weigh, 
and periodically discharge the proper amount of stock on a conveyor 
which delivers the mixture to the automatic feeder of the Fearn aught 
picker, the teeth of which separate the wool bunches and mix the fibers. 

(6) In the carding department, the principal technological de- 
velopment has been a marked increase in the capacity of the carding 
machines. This has resulted in a decrease in their number and a con- 
sequent reduction in the amount of labor required as tenders. The 
design and construction of the modern card permit easy and quick ad- 
justment requiring a minimum of maintenance labor. The capacity 
of the tape codensers which separate the web of wool into ribbons is 
more than twice that of the old ring doffers which they have generally 
replaced. 

(7) Similar advances have been made in the manufacture of 
worsted products. The scouring equipment has been greatly increased 
in size with the result that there has been a material reduction in the 
number of overseers, drier and scouring men, and soap and chemical 
men. Greater durability of machinery in the top-making depart- 
ment has resulted in increased output per man-hour because of fewer 
stops required for repairs and adjustments in the machinery. In the 
slashing department, the utilization of larger section beams with 
greater speed and efficiency has resulted in a marked reduction in the 
number of slash tenders.^ 

The comparative stability in the amount of labor required during 
the twenties is indicated by the fact that in 1920 the series was at a 
level of 107.4, while in 1929, the index stood at 104.4, a net decrease of 
only 2.7 percent (chart VIII, Table 10). After dropping to 107.4 in 
1920. the index rose to 115.1 in 1921. From 1922 to 1924 the series de- 
clined only slightly from 114.5 to 108.2. In 1925 a somewhat sharper 
decline occurred. The index however, flattened out in 1926 and for 
the next 2 years remained on a plateau of about 101. By 1929 the 
series had increased to within three points of the 1920 level. But in 
1930 it turned suddenly downward and continued to decrease to the 
end of the period, 1936. From a level of 103.6 in 1930 the unit labor 
requirements index declined to the all-time low of 74.0 in 1936, a de- 
crease of 28.5 percent. 



» The above analysis of productivity changes in the woolen and worsted goods industry was 
compiled from Boris Stern, "Mechanical Changes in the Woolen and Worsted Industries, 
1910 to 1936," U. S. Bureau of Labor Statistics, Monthly Labor Review, January 1938, 
pp. 58-93. 



280 





Pc 


to 


Ji' 


HI 


i" 


o 


g 




Q 




z 




< 




0) 




H 




Z 


CO 


LxJ 




2 


^ 


UJ 


o 


o: en 


Q 


— UJ 


III 


13 H 


h- 


UJ w 




tt: o 


$ 


UJ 


Q 


OC ^ 


2 


O z 
CD "^ 


z 


^ 


y 


-J 


^ 


1 


^ 



CONCENTRATION OF ECONOMIC POWER 

Chart VIII 

8 S 8 s a 8 o 



CO 
UJ 
X 
UJ 

o 



11 













\ 
















\ 


/ 














\i 
















1 
















\ 


.» 








i 














s 




J 


/ 










1 


i 


, 


,' 










2 


/ 












< 


- 


'' 












1 


1 














/ 
































i 
















h 
















1 


N, 














.v— 












''-'3 


) 














/ 












SL^----,. 














£[ 




~ - > 


^^^ 


















'" V 


















\ 



















8 S 



Table 10. — The woolen and icorsted goods industru 

[1926=100] 



Year 


Unit labor 
require- 
ments ' 


Price' 


Produc- 
tion' 


Year 


Unit labor 
require- 
ments 1 


Price » 


Produc- 
tion' 


1914 . . 




50.5 
55.0 
70.4 
101.7 
138.6 
124.3 
153.7 
91.9 
95.7 
107.5 
106.8 
110.2 




1926. 


100.0 
101.8 
101.2 
104.4 
103.6 
94.8 
89.8 
89.0 
82.6 
76.8 
74.0 


100.0 
97.8 

100.1 
88.3 
79.0 
- 68.2 
57.7 
69.3 
79.7 
76.1 
82.9 
91.1 


100.0 


1915 






1927. 

1928 


101.1 


1916 






95.8 


1917 






1929 


96.5 


1918 




1930 


71.0 


1919 


117.4 
107. 4 
115.1 
114.5 
110.4 
108.2 
100.4 


97.8 
86.7 
95.8 
102.3 
123.7 
103.2 
107.0 


1931 

1932 

1933.... 


78.0 


1920 


64.7 


1921 


83.5 


1922 


1934. 

1935 


69.9 


1923 


110.2 


1924 


1936 


107. 6 


1925 


1937 













' Works Progress Administration, National Research Project, Production, Employment, and Productiv- 
ity in 59 Manufacturing Industries, 1919-36, Part II, 1939. 
JU. S. Bureau of Labor Statistics. 



CONCENTRATION OF ECONOMIC POWER 281 

That this rapid diminution in unit labor requirements has taken 
place independently of the rate of production is indicated in a com- 
parison of 2 years of comparable production rates, 1925 and 1936. 
In the former year production stood at 107 and the unit labor re- 
quirements index at 100,4 ; in the latter year production was at 107.5, 
while the index of labor required per unit had dropped to 74. 

The recent decline in unit labor requirements in the woolen and 
worsted goods industry is very similar to that which took place in 
the cotton goods industiy during the same period. This niay well be 
attributed to the fact tliat both industries are characterized to a great 
extent by productive mechanisms and 'functional divisions which bear 
a very close resemblance to each other. 

Price. 

The price behavior of the woolen and worsted goods industry 
(chart VIII, table 10), as represented by the wholesale price sub- 
group index of the United States Bureau of Labor Statistics, has 
been quite similar to that of the cotton goods industry. 

The index rose greatly during the first World War from 50.5 in 
1914 to 138.6 in 1918. After declining 14.3 points in 1919 the index 
rose to the all-time high of 153.7 in 1920. The post-war depression 
took the index downward in 1921 to a level approximately halfway 
between the 1914 and 1920 positions. From 1922 to 1925 the index 
increased from 95.7 to 110.2. After 1925 the index turned down- 
ward again initiating a decline which ended in 1932 at 57.7, or 34.6 
percent below the 1929 position. This drop in the series was inter- 
rupted only by a slight increase in 1928f The upturn following 1932 
brought the series to a position of 79.7 in 1934 and, after a slight 
decline in 1935, to 91.1 in 1937. 

The industry is not characterized by a high degree of concentra- 
tion, as the 4 largest firms in 1935 produced only 24.2 percent of the 
industry's value of products. The woolen and worsted goods indus- 
try is, however, characterized by a much greater degree of integi-a- 
tion than the cotton goods industry. Most mills are completely inte- 
grated in the spinning and weaving operations, and practically all 
of them perform their own finishing. In the industry there are only 
4 job printers of woolen goods and less than 50 plants engaged ex- 
clusively in finishing woolen goods. The chief finishing process in 
woolens, known as fulling, cannot profitably 'be carried out on a 
large scale, and thus there are no plants specializing in .this process.* 

The functions of a converter are not needed in ~ the industry as 
woolens and worsteds are given their design in the weave mill and 
are consequently a finished product upon leaving there. The mills 
perform their own selling directly to the cutters-up; one study has 
indicated that approximately one-fifth of men's and women's wool 
fabrics are sold through wholesalers and jobbers, while at least 70 
percent are sold directly by the mills themselves.^ 

The degree of integration is greater in the case of woolen mills 
than of worsted mills, for in recent years worsted mills have tended to 
purchase top rather than to make it. 

To a considerable extent, this high degree of integration arises 
out of. the individual character of particular woolen weaves. In the 

* Graduate School of Business Administration, Distribution of Textiles, Bureau of Business 
Research, Harvard University, Bulletin No. 56, 1926, pp. 51-52. 
Mbid.. pp. 51-53, 



282 CONCENTRATION OF ECONOMIC POWER 

cotton goods industry, the standard product of the -weavinfr mill, gray 
goods, is highly uniform. Mills in the woolen and worsted goods 
industry produce patterns of weaves which are seldom constant, due 
to ever-changing styles. To meet these style demands a weaving 
mill in this industry must create new patterns, must, test their possi- 
bilities by sampling through sales agencies, and must be able to shift 
its production, at short notice, into, new and different weaves. 

This last factor is of great significance in the determination of the 
industry's price behavior as the need for ability to change models on 
short notice puts a premium on small mills. Although the integra- 
tion of the woolen and worsted mills results in a comparatively sm^ll 
number of market transactions, the existence of a multiplicity of 
producers leads to intense price rivalry among them. 

Competition in the industry has been intensified by a large amount 
of excess capacity. During the World War, the woolen and worsted 
goods industry, like the cotton goods industry, experienced a tre- 
mendous increase in demand. This demand did not disappear com- 
pletely at the end of the war as the demobilization of millions of 
soldiers who had to be outfitted in civilian clothes created a substan- 
tial market for the industry. Furthermore, the export trade was 
large and apparently expanding. These two facts led to excessive 
optimism in a large section of the industry and from 1919 to 1925 the 
number of spindles in place increased from 5,250,000 to over 6,000,000, 
and the number of looms from 77,338 to 80,629.'' Capacity was also 
augmented by the introduction of the previously described automatic 
loom. 

Price competition became extremely severe when the export trade 
slackened and domestic demand finally decreased after the war, due 
partially to the decline in per capita consumption of clothing. The 
trend toward lighter clothing took place at the expense of the woolen 
industry ajid to the benefit of the newly developing rayon industi*y. 
By the time the woolen industry had developed satisfactory processes 
of making sheer wool dress fabrics, a considerable portion of its 
market had been lost to the lighter textiles. 

That severe price competition still characterizes the industry is 
indicated in a recent industry bulletin : 

The delay in the beginning of a recovery movement early in 1938 meant that 
the mills, in order to get a start on operations for the new season, made desperate 
efforts to generate business and to obtain orders, particularly from men's wear 
cloths. 

The ordinary method of securing an increase in volume was to make price 
concessions. This practice, fundamentally unsound in the majority of instalices, 
was carried to such extremes as to mean that most mills, although they booked 
a reasoniable volume of business, accepted this business at prices which pre- 
cluded any prospect of profit. The mills are not necessarily to be criticized for 
this situation. The situation showed, however, the effect of pa'rtial operations 
on profits and indicated what happens when mills in an industry which is more 
than adequately supplied with machinery, compete on a price basis for a volume 
of business which is not adequate to provide full employment for that 
machiniery.' 

Productivity and Price. 

The relationship between price and productivity in the woolen and 
worsted goods industry is similar to that of the cotton goods industry, 
Except for a dip of 10 points in 1920, the unit labor requirement index 

« H. E. Michl, The Textile Industries. The Textile Foundation, Washington, 1938, p. 223. 
■' Bulletin of the National Association of Wool Manufacturers, vol. LXVIII, 1938, p. 32. 



CONCENTRATION OF ECONOMIC POWER 283 

was comparatively stable from 1919 through 1922. In 1920 the price 
index went to its all-time high of 153.7 only to drop precipitously to 
91.9 in 1921. From 1923 to 1926, the unit labor requirements index 
declined while the price series rose through 1925 but dropped 10 points 
in 1926. In 1927 and 1928. the unit labor requirement series was 
strikingly stable and the price index varied only 2.3 points. How- 
ever, the price series turned abruptly downward and went to a low of 
57.7 in 1932. In 1929-30 unit labor requirements moved upward but 
in 1931 began a decline which continued through 1936 to the low of 
74.0. The price series moved upward in 1933-34, was interrupted by 
a slight decline in 1935, but by 1937 had reached 91.1. 

Tlie long-term decline in the price series of woolen and worsted 
goods through 1936' compares closely with the behavior of the price 
series of cotton goods from 1923, and even more closely if taken from 
1925. The behavior of the two series of unit labor requirements is 
also strikingly similar, each showing a slight over-all decline up to 
1930, and an abrupt decrease thereafter. 

As in the cotton goods industry the long-term movements of the unit 
labor requirement and the price indexes in the woolen and worsted 
goods industry have generally paralleled each other. 

rURNITUEE INDUSTRY 

Productivity. 

The significant gains in productivity in the furniture industry have 
taken place principally in the manufacture of household furniture. 
Altliough extremely rapid strides have been made in the manufacture 
of metal furniture, this branch occupies such a small segment of the 
industry that the index of unit labor requirements must be taken to 
reflect largely the advances in the manufacture of wooden household 
furniture. 

The most important technological advances have come through 
plant rationalization. Functions in the manufacture of furniture 
have been broken down into separate, routinized operations and today 
the parts flow from one detailed operation to another until the finished 
commodity emerges as the product of mass production methods rather 
than that of an individual craftsman's art. The principal tech- 
nological changes which have taken place in recent years in the four 
operations of the furniture manufacturing industry may be sum- 
marized as follows : 

(1) The use of carving machines in woodworking has increased 
materially both the productivity of the individual worker and the 
speed of the operation and has also almost completely displaced the 
old hand carver. Better plant layouts and routinization have 
sjmchronized and speeded up the activities of machine operators 
and miscellaneous handlers. The introduction of the multiple spindle 
in place of the single spindle carving machine has likewise materially 
increased the productivity of the workers. 

(2) The most noticeable changes involving the elimination of the 
skilled worker and the segmentation of his many functions have 
taken place in cabinet making. Formerly, the highly skilled cabinet 
maker assembled by himself a complete unit of furniture. Today 
cabinet making is performed by a number of semi-skilled workers, 
each of whom assembles only parts of a unit, such as the front, sides, 
or back of a case. The skilled fitter, who fits drawers and doors into 



284 



CONCENTRATION OF EC'ONOillC POWER 



the assembled unit, has resisted routinization more successfully than 
the cabinet maker but has been forced to speed up his operations 
because of the rapidity with which the highly specialized assemblers 
perform their specific functions. 

(3) The three functions of the finishing department — sanding, 
rubbing, and staining and varnishing — have been transformed from 
hand to machine operations to a considerable extent. The output 
of individual operators has consequently been greatly increased since 
the sanding machine, the rubbing machine, and the spray gun are 
among the most highly productive types of machine equipment in- 
troduced into the manufacture of furniture. 

(4) The gains in labor productivity in the upholstery depart- 
ment have been made in much the same way as those in cabinet- 
making. A skilled upholsterer formerly worked on an entire piece 
of furniture. Today the functions are subdivided so that one group 
of workers upholsters arms while another group works on backs. 
This change lias eliminated some of the highly skilled workers and 
has also markedly advanced the speed at which a given unit can 
be produced. The subdivision of work has not, however, been ex- 
tended so completely in this department as in cabinet making. 

That sweeping technological changes have not taken place in fur- 
niture manufacturing is indicated by the slow and gradual decline 
in unit labor requirements. The necessity of performing many 
highly dissimilar functions, the bulkiness of the material and its lack 
of pliability, have raised obstacles to the introduction of any changes 
which would result in abrupt decreases of labor expenditure. 

Immediately after the war, the unit labor requirement index went 
to its all-time high of 152.4 in 1920 but fell to 123.8 in 1921 (chart 
IX, table 11). In the next 2 years the index decreased only slightly 
but in 1924 and 1925, it dropped sharply, reaching a level of 102.4 
in the latter year. The series continued to decline through 1929 but 
in a more gradual manner. In 1930 the series turned upward and 
with the exception of a slight decline in 1932 rose steadilv until in 
1933 it stood at 99.5, compared with 91.7 in 1929. In 1934^ however, 
the series dropped to 90.7 where it remained during the next year 
but in 1936, it rose again to 93.7. 



Table 11. — The furniture industry 

[1926=100] 



Year 


Unit labor 
require- 
ments ' 


Price ' 


Produc- 
tion' 


Year 


Unit labor 
require- 
ments ' 


Price « 


Produc- 
tion! 


1914 




70.6 
70.9 
72.8 
81.7 
93.3 
114.7 
165,6 
129.9 
114.6 
116.7 
107.9 
104.6 




1926 


100.0 
98.5 
96.2 
91.7 
93.7 
95.3 
94.7 
99.5 
90.7 
90.9 
93.7 


100.0 
97.7 
96.7 
95.0 
94.0 

75:0 
75.1 
79.0 
77.0 
78.0 
85.9 


100.0 


1915 






1927 


100 








1928 


98.5 


1917 






1929 


108.5 


1918 






1930 


76.2 


1919 


126.1 
152.4 
123.8 
120.6 
119.2 
106.3 
102.4 


■ 59.8 
52.1 
48.9 
65.0 
74.5 
80.0 
92.1 


1931 


58 4 


1920 


1932 

1933 

1934... 

1935 

1936 


39.7 


1921 

1922 


42.1 
47.2 


1923 


61.3 
72.1 




1937 













1 Works Progress Administration, National Research Project— Production, Employment, and Produc 
tivity in 59 Manufacturing Industries, 1919-36, pt II, 1939. 

2 U. S. Bureau of Labor Statistics. 



CO^X'ENTRATION OF ECONOMIC PO^YER 



285 



05 




eS 


LjJ 




i^- 


O 






a: 






Q. 






Q 






^ 






< 






(/> 






h- 






z 






Ijj 






2 






LiJ 






cr 


00 




!>- ^ 




UJ 


-P! 


^ 


% 


t- ij-i 


CO 


1- 


^ cr 


Q 


z 


< 




(T 


S Q^ 




U- 


cn 


ID 




< 






-J 






l- 






^ 






3 






Li. 






O 






0) 






U 






X 






LiJ 






O 




£ 



is 











\ 




























\ 














J 














( 














\ 


y 












1 /''■ 














F 














/) 














/ 










^ 




/ 










1 


J 


r 










8 




























S 


Y 












i 














--"'^ 


^i' 










'--. 


::;^i 














^** 


=c ,^ 














^ 


^^^x. 














'v^ 














"I 


\ 












sf 


\ 



















- II 



Productivity varies in this industry quite closely with the rate of 
production. The marked decrease in productivity during 1923-26 
was accompanied by a rapidly rising rate of output, as the production 
index rose from 74.5 in the former year to 100 in the latter. In 1929, 
when the unit labor requirement index had reached its lowest level 
up to that time, production was at the highest level yet attained. 
Nevertheless, a significant long-term decline in unit labor require- 
ments has taken place in the industry independent at the rate 
of production. This is indicated by a comparison of two years of 
comparable production rates, 1922 and 1935. In the former year, 
production stood at 65.0 and unit labor requirements at 120.6; by 
1935 labor required per unit had decreased to 90.9 though production 
was at 61.3. 

In 1934 when the unit labor requirement series reached its all-time 
low of 90.T, production was 56.5 percent below the 1929 level. This 



286 CONCENTRATION OF ECONOMIC POWER 

may be taken to indicate the widespread introduction during the 
depression of numerous changes designed to reduce labor expendi- 
tures. 

Price. 

The price index used in this study is that of the furniture subgroup 
of the United States Bureau of Labor Statistics. This composite 
index reflects primarily the price behavior of wooden household fur- 
niture, although other types of furniture, principally metal, are in- 
cluded and weighted according to their value in exchange (chart IX, 
table 11). 

A marked increase in the price of furniture took place during the 
World War, the index rising from 70.6 in 1914 to 93.3 in 1918. The 
index continued upward through 1920 when it reached the all-time 
high of 165.6. Although the series dropped abruptly in the next 2 
years to 114.6 in 1922, it still remained 34 percent above the pre-war 
level. The height to which furniture prices rose during and after 
the war and their failure to drop back immediately to their pre-war 
level brought forth a Senate resolution in 1922, directing an investi- 
gation by the Federal Trade Commission into prices, markets, and 
profits in the furniture industry.^ 

After a slight upturn in 1923, the index fell from 116.7 in 1923 to 
100 in 1926. A steady, more gradual decline in the price series con- 
tinued through 1930, In the next 2 years tlie index dropped 
abruptly and in 1932 was at 75, or 21.1 percent below the 1929 level. 
It remained at this low level during 1933, turned upward in 1934 
and fluctuated between 79 and 77 for 3 years. In 1937 the series 
reached a level of 85.9 which was, however, considerably below the 
pre-depression level. 

The household-furniture industry is divided into two classes of 
manufacturers : The integrated producers and those who specialize in 
one line and buy from other manufacturers the pieces needed to com- 
plete a suite. As early as 1922, the Federal Trade Commission re- 
ported that, "There is a tendency in the industry for companies manu- 
facturing specialized lines to integrate or to become affiliated in order 
to manufacture complete suites for bedroom, living room, and dining 



['oom. 



55 9 



The manufacture of furnitiu'e is one of the least concentrated of 
the Nation's major industries; in 1935, the four largest concerns pro- 
duced only 5.6 percent of the industry's value of product. This lack 

* The role which the level of prices played in provoking this investigation is to be noted in 
the wordins of the resolution adopted by the Senate on January 4, 1922 : 

"Whereas reported statistics show that Ihe prices of house-furnishing goods readied a 
higher peak relative to pre-war prices than any other class of commodities ; and 

"Whereas since May 1920, wliile most ottier reported clo^ses of commodities were falling 
in price, the prices of house-furnishing goods continued to increase until the latter part of 
1920, and then registered only a gradual decline ; and 

"Whereas the prices of house-furnishing goods are now relatively very much higher than 
any other class of commodities, and particularly are relatively more than twice as high as 
the prices of farm products : Therefore be it 

"Resolved, That the Federal Trade Commission be, and hereby is, authorized and directed 
promptly to investigate the causes of factory, wholesale, and retail price conditions in the 
principal branches of house-furnishing goods industry and trade, beginning with .Tanuary 
1920, and particularly to ascertain the organization and interrelations of corporations and 
firms engaged therein, and whether there have been and are unfair practices or methods of 
competition, or resti'aints of trade, combinations, or manipulations, out of harmony with the 
law of public interest ; and if so, what effect the same, have had on prices; and, serially, to 
report the facts, with its recommendations, at the earliest possible tune as different phases 
of the investigation are completed." (Federal Trade Commission, House Furnishings ludua- 
tries, vol. I, Household Furniture. 1923, p. 1.) 

• Ibid., p. 44. 



CONCE^^TRATION OF ECONOMIC POWER 287 

of concentration is accompanied by a high degree of price competition. 
Causes for keen competition lie in: 

(1) A large amount of productive equipment considerably expanded 
through 1929 to take care of a backlog in furniture demand created 
during the World War and the rapid increase in residential building 
between 1921 and 1925. This greatly increased productive capacity 
came to weigh heavily upon the industry when its markets during the 
thirties were gradually and then sharply curtailed. 

(2) Decentralization has grown in the industry in recent j'ears 
due largely to new methods of merchandising. Formerly most re- 
tailers placed large orders with manufacturers and then tried to dis- 
pose of these extensive stocks. The retailer has found that he is able 
to attract more customers by offering a wide selection of furniture as 
to material and color. In order to satisf}' this varied demand without 
carrying on hand a tremendous stock of merchandise, which would de- 
crease his stock turnover and increase his losses from shop wear and 
markdowns, the retailer has found it advantageous to stock only a few 
patterns as samples from which orders can be taken for delivery within 
a short time. This has put a premium upon the rapidity of manufac- 
ture and upon strategic locations. The development of trucking has 
facilitated this method of merchandising furniture. In order to meet 
the competition of small local manufacturers who can give quick 
service, some of the old, well-established manufacturers have set up 
branch factories at strategic points. This trend toward decentraliza- 
tion in marketing, with the advantage it has given to local manufac- 
turers, has obviously militated greatly against the development of con- 
centrated ownership and control in the industry. 

(3) Retail stores have forced sales by allowing liberal credit ar- 
rangements and have found themselves with assets frozen in uncol- 
lectible installment accounts during economic downturns. Since there 
is no satisfactory market for used furniture and the cost of storage, 
handling, etc., in many cases would increase the dealer's loss if he 
were to repossess, he has purchased with extreme caution in recent 
years, seeking by every device to gain the best possible terms of sale. 
13y playing one producer against another, retailers greatly intensify 
competition among manufacturers. Manufacturers, finding that their 
outlets are curtailed, that retailers only gradually work off their 
accumulated stocks, and that they will have no volume whatsoever 
unless they meet the buyers' rigid terms, are forced not only to 
grant liberal credit concessions, but are compelled to reduce prices to 
the lowest possible levels. This obviously contributes greatly to vig- 
orous price competition in the industry. 

(4) Finally, the furniture industrj^ has been characterized by a 
conspicuous inadequacy of accounting methods up to very recent years. 
Manufacturers who are not sure of their costs may be tempted by a 
large order to quote prices which, if not actually below the cost of 
production, offer only a ipiiiimum margin of profit. Even producers 
who are fairly confident of their accounting methods are often in- 
clined to reduce their prices below a profitable level on particular 
'Orders, because of their desire to meet the prices of any competitor ; 
yet prices of competitors are often based upon inadequate cost ac- 
counting methods. All these factors have contributed toward making 



288 CONCENTRATION OI ECONOMIC POWER 

the furniture industry one of the mostly highly competitive in the 
American economy.^^ 

Productivity and Price. 

As in the cotton goods and woolen and worsted goods industries, 
the most striking characteristic of the unit labor requirement-price 
relationship in the furniture industry is the similarity of movement 
of the two indexes. 

The decline in unit labor requirements from 1920 to 1922 was 
accompanied by a more precipitous drop in price while both indexes 
leveled off during 1922-23. Both series fell approximately the same 
extent from 1923 to 1926, and they continued to parallel each other 
to 1928. In 1929 the unit labor requirement index dropjjed slightly 
below the price series, but in 1930 it rose again to approximately the 
same level as that of the price series. Beginning in 1931, however, 
the indexes parted company. The curtailment of output brought 
about by the depression resulted, on one hand, in a marked dechne 
in prices and, on the other, in an increase in unit labor requirements. 
This divergence was greatest in 1933 with the unit labor require- 
ment index at 99.5 and the price series at 75.1. In 1934 the price 
index increased slightly while labor required per unit decreased more 
sharply but in 1935 they both began gradual upturns. 

The practical uniformity of the two series from 1920 to 1930, and 
especially from 1923 to 1928, is one of the most striking unit labor 
requirement-price relationships in this study. The trend of the in- 
dexes in recent years seems to indicate that this parallelism to a 
considerable extent has been reestablished. 



" For a more detailed treatment of some aspects of these competitive problems, see U. S. 
Bureau of Foreign and Domestic Commerce, Furniture Distribution in the Midwest, by C. R. 
Niklason, Domestic Commerce Series No. 75, 1932. 



APPENDIX I 

EMPLOYMENT IN INDUSTRIES CLASSIFIED ACCORD- 
ING TO THE AMOUNT OF ELECTRICAL ENERGY USED 
PER MAN-HOUR IN PLANTS OF VARYING SIZE, 1937 ' 

Industi'ies in which electric energy used per man-hour is highest in plants 
employing — ^ 
I. 1 to 50 wage-earner group: Employment 

Liquors, malt 39, 856 

Cotton narrow fabrics 11, 002 

Knitted outerwear : 

Contract factories.^ 2,241 

Regular factories 19,630 

Work garments (including work shirts) and sport garments, 

except leather 47, 325 

Furnishing goods, men's, not elsewhere classified — regular 

factories 7, 031 

Men's, youths', and boys' clothing, not elsewhere classified : 

Contract factories , ^ 22, 145 

Regular factories 76, 691 

Perfumes, cosmetics, other toilet preparations 9, 242 

Blackings, stains, and dressings 1, 336 

Radios, radio tubes, and phonographs 46, 337 

Machine repair shops , 3, 787 

Dyeing and finishing of cotton yarns 5, 358 

Recovered wool fiber 2,772 

Silk broad woven goods — "commission" branch 2, 179 

Trousers (semidress), wash suits, and washable service 

apparel 9, 712 

Shirts (except work shirts) collars, nightwear — contract fac- 
tories 6,391 

Furnishing goods, men's, not elsewhere classified — contract 

factories 456 

Men's, youths' and boys' clothing, not elsewhere classified — 

contract sponging, examining, etc., of material 725 

Batting, padding, wadding, upholstery filling . 4, 168 

Woolen and worsted dyeing and finishing 1, 743 

Leather and slieep-lined clothing 2, 948 

Wire drawn from purchased rods 24, 052 

Total for I 347,127 

^Employment in industries for which comparuble size groupings are not available: 

Industry: Employment 

Dyeing and finishing of rayon tand silk) yarn 1, lOO 

Processed cotton waste 2,051 

Flour and other grain-mill products 2o! 660 

Sugar, beet 9, 024 

Liquor, distilled 5, 471 

Wool combing — commission, and tops for salel 8, 252 

Rayon yarn and thread — proce.ssed for sale 4, 847 

Silk broad woven goods — "regular" branch 12, 752 

Knitted gloves and mittens 3] 182 

Men's underwear — regular factories 6! 410 

Sheet metal work, not specifically classified 17, 864 

Wirework. not elsewhere classifi'^d 25 846 

Motor-vehicle trailers 5! 179 

'Total 117,641 

= In arriving at the figure representing the hiiihest use per man-hour of electric energy, 
selections were made for the 1 to 50 wage-earner group from its three component groups : 
1 to 5, 6 to 20. and 21 to 50 ; and selections were made for the 501 and over wage-earner 
group from the component groups : 501 to 2.500, 2,501 and over, and, where given, the 501 
and over group. 

289 
277551— 11— No. 22 20 



290 CONCENTRATION OF ECONOMIC POWER 

II. 51 to 100 wage-earner group : Employment 

Tol)acco (chewing and smoking) and snuff 9,746 

Rayon throwing and spinning— commission only 1,770 

Rayon broad woven goods — "commission" branch 2, 610 

Rayon narrow fabrics 5, 001 

Silk throwing and spinning — commission only 12,072 

Silk varn and thread— made for sale 9, 452 

Pulp (wood and other fiber) 25,680 

Insecticides and fungicides 2,821 

Nonclay refractories 5, 230 

Steel barrels, kegs, and drums 6, 140 

Total for II 80,522 

IT. 101 to 500 wage-earner group : 

Feeds, prepared, for animals and fowls 11, 759 

Dyeing and finishing of cotton fabric 46,904 

Woolen yarn 2, 397 

Felt goods, except woven felts 3, 349 

Paper-fiber and grass carpets and rugs 7SS 

Silk narrow fabrics 4,259 

Men's underwear — contract factories 1, 424 

Paper 104,112 

Soap 13, 764 

Cleaning and polishing preparations 2,852 

Leather, tanned, elc. : 

Contract factories . 2, 190 

Regular factories 45,603 

Rubber goods other than tires, inner tubes, and boots and shoes- 44, 137 

Cement 24,808 

Lime > 8,797 

Copper smelting and refining 14,514 

Lead smelting and refining 4,036 

Zinc smelting and refining 11,265 

Smelting and refining, nonferrous metals other than gold, silver, 

platinum, not from the ore 4,526 

Wrought pipe, welded and heavy-riveted 13,635 

Total for III 365,119 

IV. 501 and over wage-earner group : 

Cereal preparations _ 7, 887 

Meat packing, wholesale 119,283 

Sugar refining, cane 14,024 

Cigars 50,442 

Cigarettes - - 26, 149 

Cotton yarn and thread 76,706 

Cotton woven goods (over 12 in. in width) 318.283 

Worsted yarn 15, 478 

Woolen woven goods, including woven felts 58, 109 

Worsted woven goods 68,443 

Woolen and worsted carpet yarn 2,385 

Wool carpets and rugs (other than rag) 30,496 

Rayon broad woven goods (18 inches wide and over), -Regu- 
lar" branch 50, 497 

Dyeing and finishing of rayon and silk fabric 15, 669 

Hosiery 125, 958 

Knitted cloth 9. 907 

Knitted underwear 36, 028 

Shirts (except work shirts), collars, nightwear ; regular fac- 
tories 40.905 

Rayon and allied products 55,098 

Drugs and medicines 21,336 

Petroleum refining 79. 622 

Rubber boots and shoes 18.356 

Rubber tires and inner tubes 63.290 

Clay products, other than pottery 46, 769 



CONCENTRATION OF ECONOMIC POWER 291 

IV. 501 and over wage-earner group — Continued. Employment 

Pottery, including porcelain ware 28, 264 

Glass 73, 997 

Aluminum products 23,337 

Nonferrous-metal alloys ; nonferrous-metal products, except 

aluminum, not elsewhere classified 77, 694 

Blast-furnace products 22. 580 

Steel-works and rolling-mill products 473. 215 

Cast-iron pipe and fittings 15, 909 

Plumbers' supplies, not including pipe or vitreous-china sani- 
tary ware 23, 417 

Tin cans, other tinware, not elsewhere classified 32, 248 

Stamped and pressed metal products ; enameling, japan- 
ning, and lacquering 54, 840 

Electrical machinery, apparatus and supplies 243, 513 

Machine tools 45, 875 

Machine-tool accessories and machinists' precision tools 28, 977 

Machinery not elsewhere classified 126. 663 

Machine-shop products-^ 89, 519 

Engineers, turbines, water wheels, and windmills 28, 723 

Refrigerators, refrigerating, ice-making apparatus 47, 643 

Motor-vehicle bodies 84. 597 

Motor-vehicle parts and accessories 189,071 

Motor vehicles, not including motorcycles 190, 323 

Total for IV 3,251,525 

Source : U. S. Bureau of the Census, Census of Manufactures : 1937, Man-Hour 
Statistics for 105 Selected Industries, table 1, pp. 3-5. 



APPENDIX J 

METHODOLOGY OF CORRELATING POWER DIFFEREN 
TIAL AND ECONOMIC CONCENTRATION 

(1) The electric energy used per man-hour was averaged for the 
three smallest groups of plants for which data are available, i. e., those 
employing 1 to 5, 6 to 20, and 21 to 50 wage-earners. No weighting 
factors were utilized in computing the average, for the reason that, in 
most industries, the number of wage-earners employed in these small 
plants constitutes such a slight proportion of the industry's total, the 
process of weighting by the number of wage-earners would not change 
materially the calculations. Furthermore, the purpose of the analy- 
sis — to compute the power differential between large and small 
plants — does not rest in any way upon the proportion of the industry's 
Avage-earners employed in the various groups of plants; the question 
of employment in plants of varying sizes relative to their use of power 
has already been discussed. 

(2) The electric energy used per man-hour for the large plants 
was taken for the one group of plants employing over 100 workers 
which was characterized by the largest use of electric energy per 
man-hour. Selections from 5 size classifications (not invariably 
contiguous in the primary data) could thus be made.^ These size 
groupings together with the number of industries represented by each, 
in the correlation are as follows : ^ 



> This is due to the possibility of disclosure of individual firms. 
* Industries in each optimum large size group : 

1. 101 to 500 wage-earner group : 

Feeds, prepared, for animals and fowls. 

Paper. 

Insecticides and fungicides. 

Soap. 

Rubber goods other than tires, inner tubes, and boots and shoes. 

Cement. 

Lime. 

Smelting and refining, nonferrous metals other than gold, silver, platinum, 

not from the ore. 
Wrought pipe, welded and heavy-riveted. 

2. 101 to 2,500 wage-earner group : 

Flour and other grain-mill products. 
Liquors, distilled, 
steel barrels, kegs, and drums. 
Wirework not elsewhere classified. 

3. 501 to 2,500 wage-earner group : 

Cereal preparations. 

Liquors, malt. 

Pulp (wood and other fiber). 

Drugs and medicines. 

Clay products, other than pottery. 

Pottery, including porcelain ware. 

Cast-iron pipe and fittings. 

4. 501 and over wage-earner group : 

Cigars. 

Glass. 

Aluminum products. 

Machine tools. 

Machine-tool accessories and machinists' precision tools. 

Engines, turbines, water wheels, and winamills. 

5. 2,501 and over wage-earner group : 

Petroleum refining. 

Electrical machinery, apparatus, and supplies. 
Radio.'!, radio tubes, nad phonographs. 
Bod'f^K (motor-vehiolt'.i. 



CONCENTRATION OF ECONOMIC POWER 293 

Number of industries represented 

Size of plant (by number of wage-earners) : 

101 to 500 , 9 

101 to 2,500 .— 4 

501 to 2,500 7 

501 and over 6 

2,501 and over _ 4 

Total 30 

(3) The percent difference between the two figures arrived at in 
(1) and (2) was then compute^!, with the figure representing the 
small plants serving as the base.^ 



' It is widely believed that any differentiation between large and small plants must be 
made on the basis of each specific industry involved, since a plant which appears large in 
one industry would seem small in another. To a certain extent this is true, but many 
industries, which by the very nature of their productive processes might be considered 
unfavorable to the operation of large units, are found to have plants in the two largest 
size groups. Of the 30 industries in the correlation only 4 — feeds, prepared ; insecticides 
and fungicides ; lime ; and smelting and refining of nonferrous nonprecious metals not 
from ore — rt'port the largest size groups employed as under 501 (see following table). 
This indicates that large-scale operation has been extended throughout the framework of 
-American manufacturing enterprise and is not confined to a few basic industries. 

Largest size group reported for indvstries iih the correlation 

Largest size 
. group 
Industry : reported 

Flour and other grain-mill products 101-2,500. 

Feeds, prepared, for animals and fowls 101—500. 

Cereal/ preparations 501-2,500. 

Distilled liquors 101-2,500. 

Malt liquors 501-2,500. 

Cigars 501 and over. 

Paper 501-2,500. 

Pulp (wood and other fiber) 501-2,500. 

Drugs and medicines 501—2,500. 

Insecticides and fungicides 101-500. 

Soap 501-2,500. 

Petroleum refining 2,501 and over. 

Rubber goods, other than tires, inner tubes, and boots and shoes — 501—2,500. 

Cement 501-2,500. 

Lime 101-500. 

Clay products, other than pottery ^ 501-2,500. 

Pottery, including porcelain ware 501-2,500. 

Glass 501 and over. 

Aluminum products 501 and over. 

Smelting and refining, nonferrous metals other than gold, silver^ 

and platinum, not from ore 101-500. 

Cast-iron pipe and fittings 501-2,500. 

• ■ - - - ^- - - . 501-2,500. 

101-2.500. 

101-2,500. 

2,501 and over. 

2,501 and over. 



Wrought pipe, welded and heavy riveted 

Steel barrels, kegs, and drums 

Wirework, not elsewhere class'fied 

Electrical machinery, apparatus, and supplies 

Radios, radio tubes, and phonographs 

Machine tools 501 and over 

Machine tool accessories and machinists' precision tools 501 and over. 

Engines, turbines, water wheels, and windmills 501 and over. 

Motor vehicle bodies 2,501 and over. 

This does not mean that the very largest plants are necessarily the most efficient. 
Since the purpose of the comparison was to contrast the degree of eflJciency of large-size 
groups with that of small groups, a selection was made among all groups employing 100 or 
more workers to determine the degree of efficiency of the optimum large-scale plant. 

As a matter of fact the four industries in which the largest size group reported was that 
of 101-500 could be eliminated from the correlation without affecting materially the 
result. But it was believed that plants employing 101—500 workers could not be regarded 
as small units of production, especially in cases where they were the largest units 
reported. A precedent existed for this belief. Prof. P. Sargant Florence's classification 
of size of plants, based upon somewhat different size groupings (the old census size 
groupings), was as follows: Minute (1 to 5 wage earners), very small (6 to 20), small 
(21 to 50), smallish (51 to 100), medium (101 to 250), largish (251 to 500), large (501 
to 1,000 , and very large (1,001 and over). (P. Sargant Florence, The Logic of Industrial 
Organization, Kegan, Paul. Trench, Trubner & Co., Ltd., London. 1933, p. 29.) According 
to this c asslfication, plants employing 101 to 500 workers w< represent a combination 
of medii m (101 to 250) and largish (251 to 500). Since it was desired to include as 
many inlustries as possible in the correlation, and since a distinct difference existed 
between plants of that size and the groups selected to represent the small plants they were 
included with other obviously large plants in the determination \)f the degree of optimum 
efflcieuc; of large-scale operation. 



294 CONCENTRATION OF ECONOMIC POWER 

(4) For the concentration data, recourse was made to the statistics 
denoting the percent of tlie industry's value of products produced by 
the industry's four largest concerns, as computed for 1935 by the 
National Resources Committee.* 

It will thus be noted that while the figures for the power differen- 
tial are for 1937, the concentration data are for the previous biennial 
census year of 1935. This use of different years may be explained^ 
first, by the fact that figures showing the electric energy used per 
man-hour in 1935 were available only for 59 industries, whereas in 
1937 they could be obtained for 105 industries, and second, by the 
realization that relatively little change in the degree of an industry's 
concentration, except in the case of very rapidly expanding indus- 
tries, could be expected to occur in the short space of those 2 years. 

(5) It will be observed that 30 industries are represented in the 
correlation, whereas the primary data on electric energy per man-hour 
was, in some form or another, available for 105 industries. It was 
found, however, that in the case of 8 industries, data for all, or any 
differentiated part, of the 1 to 50 wage-earner group were not avail- 
able; in these cases the smallest size classification exceeded that of 
50 wage-earners per plant. For 55 additional industries (of which 45 
were in the industry group, "Textile Mills and Apparel") no con- 
centration data were available. The concentration data compiled by 
the National Resources Committee for textiles related to broad indus- 
try groups, such as "Cotton Manufactures" whereas the electric power 
data were available only for the particular industries, such as "Cot- 
ton Woven Goods," which constituted but segments of the broad 
groups. 

For 13 additional industries the use of electric energy per man- 
hour was considered inadequate as a measure of the technological ad- 
vantage which large plants may possess over small ones. There are 
three l)asic causes, in the case or most of these industries, for this 
inadequacy. In the first place, some of the industries, such as meat- 
packing, by the very nature of their productive processes make rela- 
tively little use of those types of equipment which could be powered 
by electrical energy. In a second group of industries, such as the 
manufacture of tin cans, the process of production in a small plant 
would involve almost identically the same type of equipment that 
would be used in a large plant. And thirdly, in some of the industries, 
such as machine shop products, the industry, as defined by the census, 
covers such a multitude of products and processes that a comparison 
between large and small plants might involve processes of a very dif- 
ferent nature. 

Thus 8 industries were eliminated because of the fact that figures 
of electric energy used per man-hour in the 1 to 50 wage-earner size 
group Avere not available; 54, because concentration data were not 
available; and 13, because the use of electric energy per man-hour is 
unsuited as a measure of the technological advantage which large 

* National Resources Committee. Tiie Structure of the American Econoirv, Part I.. 
1939, appendix 7, table 1, pp. 240-249. 



CONCENTRATION OF ECONOMIC POWER 295 

plants may possess over small plants, making a total of 75 industries 
which could not be included in the correlation.^ 



^ Industries not included in the correlation : 

(I) Industries for which concentration data are not available: 
Cleaning and polishing preparations. 
Bleachings, stains, and dressings. 
Leather, tanned, etc. : 
Contract factories. 
Regular factories. 
jS'onclay refractories. 

Sheet-metal work not specifically classified. 
Machine repair snops. 
Motor vehicle parts and accessories. 
Motor vehicle trailers. 
Textile mills and apparel group (45). 

Total, 54. , ^„ 

(II) Industries for which electric energy used per man-hour in the 1-50 wage- 
earner plant size group are not available : 
Sugar, beet. 
Sugar refining, cane. 
Rayon and allied products. 
Rubber boots and shoes. 
Rubber tires and inner tubes. 
Copper smelting and refining. 
Lead smelting and refining. 
Zinc smelting and refining. 
Total, 8. 
(Ill) Industries for which the use of electric energy per man-hour constitutes an 
inadequate measure of the technological advantage which large plants may- 
possess over small plants : 
Meat packing. 
Cigarettes. 
Tobacco and snuff. 

Perfumes, cosmetics, and other toilet preparations. 
Nonferrous metal alloys and products, not elsewhere classified. 
Blast furnace products. 
Steel works anti rolling mills. 
Tin cans and other tinware. 
Stamped and pressed metal products. 
Wire drawn from purchased rods. 
Machine shop products. 

Refrigerators, refrigerating and ice-making apparatus. 
Motor vehicles, not including motorcycles. 
Total. 13. 

The 30 remaining industries, however, are widely scattered among- 
each of the basic fields of manufacture other than textiles and ap- 
parel: Food and tobacco industries; chemical industries (i. e. making 
or using chemicals, or employing chemical processes) ; stone, clay, and 
glass industries ; nonferrous metal industries ; and iron and steel and 
their products, including machinery. Employment in these indus- 
tries during 1937 amounted to 1,177,617 wage-earners, or 39.7 percent 
of the total number of wage-earners in all fields, excepting textiles, 
and apparel, included in the primary data. 



APPENDIX K 
TECHNICAL PROGRESS AND ECONOMIC WELFARE 

AN OUTLINE OF TOPICS FOR STUDY 

IThis Outline of Topics for study of technology was prepared by Dr. Lewis Lorwin. Had 
time permitted, it would have been followed in writing part 1 of the monograph. 
It is offered here for the use of students of technological developments] 

I. The Meaning of Technical Progress 

1. Is there agreement in use of term ? 
:2. Various meanings in which used. 

3. Wide scope of term. 

4. What use shall be accepted ? 

5. Problem of relation of technical change to economic welfare. 

II. Nature, Forms, and Rate or Technical Progress 

1. Motives and causes of technical progress : 

(a) Science and technology; the inventive impulse; to what ex- 

tent have technical improvements been stimulated by the 
growth of science (and vice versa, effects of empirical 
technical progress on science), or by industrial experience. 

(b) The needs of industry and the profit motive. 

(c) Organized and systematic industrial research. 

(d) The effects of the patent system. 

(e) The influence of the social environment (a society dedicated 

to rapid economic and social change). 

(/) Special methods for stimulating technical changes (exposi- 
tions, rewards, etc.). 

(g) Resistances to technical change. 
-2. Forms and types of technical chemical progress: 

(a) Elements of technology, nature of inventions and tech- 

nical improvements; basis of classification. 

(b) Types of technical progress. 

(1) Physical and biological: 

Materials and processes: Discovery of new 
materials, improved use of old by means 
of new processes : growing importance of 
application of chemicals and chemical 
processes; synthetic materials; utilization 
of waste products; improvement of seeds, 
breeds, etc. 

Power and fuel : New sources of power ; bet- 
ter utilization of coal, gas, oil, and water 
power; improved capacity for working 
coal deposits, etc. Allocation of various 
types of power to different industries. 
296 



CONCENTRATION OF ECONOMIC POWER 297 

2. Forms and types of technical progress : . ^ i ^ 

(2) Mechanical: Machniery, implements, tools, etc. 

Basic machines and machine processes: de- 
scription of machin<;ry used in different in- 
dustries; mechanical changes in different 
industries ; growth and degree of mechaniza- 
tion of specific industries. 

(3) Managerial: nationalization (its motives and 

methods) ; scientific management (concept 
and procedures) ; improved management de- 
pendent on specialized and improved equip- 
ment (interaction of managerial and me- 
chanical changes) ; industrial standardiza- 
tion, utilization of waste products ; time and 
motion studies, personnel and employment 
policies (reduction of labor turn over and 
absenteeism; adjustment to and training for 
job) ; incentives to greater efficiency (loyalty 
to firm, etc.) ; improved cost accounting. 

(4) Structural: Change in nature of product; new 

methods of production, integration of 

plants, development of new products and 

industrial areas. 

(<?) Functional classification— innovations and improvements 

(Lederer) ; labor-saving and capital-saving devices, etc. 

3. Eate of technical progress: , . 

{a) Factors of fluctuation in growth of technical improve- 
ments; rapidity of utilization (depending on character 
of technical devices, business conditions, structure of 
industry, policies of business concerns, etc.) ; resistance 
to innovations. 

(h) Measurement of rate of technical progress; methods of 
measurement. 

(c) Comparison of different periods in this respect in the 

United States (1899-1914; 1919-29; 1930-39) ; technical 
advance in specific industries at different periods. 

(d) What determines rate of technical progress? 

(e) Is there an "optimum" rate of technical change? Can 

it be secured through the processes of competition?^ 

III. Technical Changes in American Industry 

1. Technical advance in specific industries. 

2. Periods of greatest technical advance in different industries. 

3. Conditions in different industries affecting adoption of technical' 

improvements (inventions, patents, rapidity of adoption of 
new devices, policies of business concerns, banking system, etc.) 

IV. Technical Change and Agriculture 

1. Mechanization of agricuhure as a factor in farm-city population 

niovements. 

2. Effects of technical change on status of agriculture in national 

economy. 



298 CONCENTRATION OF ECONOMIC POWER 

3. Fann ownership and tenancy — as affecting use of machinery and 

the application of technical improvements. Land tenure as in- 
fluenced by size of holdings and type of farm. Technical changes 
in methods of farming as affecting costs of production. . 

4. Technical progress and agricultural depressions. 

V. Technical Change and Some Aspects of the Economic Process 

1. Primary effects : 

(a) Time-saving and increase in productivity. 

(b) Changes in plant organization and methods of production. 

(c) Changes in supply of and demand for labor. 

2. Secondary : 

(a) Changes in cost-price relationships. 

(b) Changes in distribution of purchasing power and in con- 

sumption patterns. 

(c) Changes in supply of and demand for capital. 

(d) Changes in volume and composition of national income. 

5. Hound about : 

(a) Changes in monetary policy. 

(b) Effects of technical change on cyclical movements. 

VI. Technology and Economic Concentration 

1. Technical change as a factor in the growth of large scale industry 

and busmess. To what extent does it operate through increasing 
efficiency of large business, through control of patents, control 
of investment funds, etc. Illustrations for separate industries. 

2. Effects of economic concentration on technical advance (through 

monopolistic control of patents, of funds for research and develop- 
ment, etc.). 

3. Industrial techniques and monopolistic trade practices in individual 

industries. A description and analysis of methods used in dif- 
ferent industries for the control of supply, for market controls, 
price fixing, etc. 

4. Concentration in the distributive trades as a result of technical 

improvements. Its effects on price making and price policy, and 
indirectly on consumer expenditures. , 

5. To what extent is technical progress a factor in the elimination of 

small scale enterprise. Under what technical conditions can 
small scale industry survive? Modern technique and industrial 
home \vork. 
'6. The technical basis for the deconcentration and decentralization of 
industry. 

VII. Effects of Technical Progress on Labor 

1. Changes in type and location of labor demanded : 

(a) Technical progress and structural changes in industry. 

(b) Technical change and migration of industry. (Indus- 

trialization of the South.) 

(c) Effects of mechanization on skill requirements in differ- 

ent industries. 

(d) Technical improvements and occupational shifts. 



CONCENTRATION OF ECONOMIC POWER 299 

2. Productivity, wages and the national income : 

(a) Effects of machinery and labor-saving devices on wage- 
rates and earnings; effects of changes in skill due to 
technical changes. 

(h) Effects on methods of wage payment and wage incentives. 

(<?) Effects of specialized processes on classification of work- 
ers and wages. 

(d) Increasing productivity and wages in American industry, 
192^39. 

3. Unemployment and occupational shifts : 

(a) Estimates of industrial unemployment in the United 

States; part-time employment. 

(b) Incidence and duration of unemployment. 

(c) Occupational statistics show "that changes in techniques 

and markets have produced displacement on a large 
scale." (Changes produce idleness not by increasing 
lay-off, but by decreasing the hiring rate.) 

(d) Technological change as aggravating factor in seasonal 

and cyclical unemployment. 
(s) Significance of non-mechanical factors in labor produce 

tivity and displacement. 
(/.) Explanations of short-run effects of the introduction of 

technical improvements on number, kinds, and location 
of jobs. 

(1) Technical improvements cause increase in pro- 

ductivity; marginal productivity of labor and 
capital change relatively to each other; labor's 
marginal productivity is reduced; but institu- 
tional frictions prevent corresponding fall in 
wages; hence wages are too high; hence 
unemployment. 

(2) * * * "technological changes may be so rapid 

and so far-reaching that they bring about a 
condition in which all factors of production 
may be simultaneously unemployed." 

4. Possibilities of reabsorptioni: 

(a) How to measure the labor absorbing power of American 

industry ; 

(b) Occupational readjustments of displaced workers; 

(c) Employment possibilities in new industries; 

(d) Selective factors in an expanding labor market. 

5. Personal and social aspects of workers' life : 

(a) Fatigue, monotony of work; boredom, accidents, and 

safety. 
(h) Occupational diseases and obsolescence; occupational age 

limits; life-time productivity. 

(c) Absenteism; labor turn-over; 

(d) Standards of living in declining industries; 

(e) Worker's emotional life. 

VIII. Benefits and Cost or Technical Progress 

1. What are benefits of technical progress; how measured: 

(a) Inci^ase in productivity ; 

(b) Capital accumulation; 



300 CONCENTRATION OF ECXjNOMIC POWER 

(<?) Possibilities of economic expansion; 
(d) Basis for higher standard of living. 

2, The costs of technical progress (economic and social; change may 

occur at an "imeconomically rapid rate'") : 
(a) Obsolescence of property and capital. 
(5) Obsolescence of occupations; industrial dislocations. 

(c) Destruction of workers', skill and occupational shifts. 

(d) Expenses of retraining and occupational readjustment. 

(e) Loss of income by displaced workers. 

(/) Diminution in workers' income (relative to total in- 
come) ; hence greater inequality in the distribution of 
wealth and income. 

(g) Sliift in economic structure and need for social readjust- 
ments. 

3. Distribution of gains and costs: 

(a) Who gets largest part of benefits? Initiators of change 

or community at large? 

(b) Wlio pays cost of change? 

"Both the benefits and tlie costs fall in large degree 
upon the community at large * * *." 
((?) Can costs and benefits be balanced? How? 

"Unfortunately, competition affords no metiiod by 
which the benefits of innovations can be balanced 
against the costs" * * * 

(d) Subsidy — results in fact that "the cost of change tends to 

be excessive and that many changes occur at an uneco- 
nomically rapid rate" (present system gives "enormous 
subsidy" to change). 

(e) How can costs be kept at a minimum? 

IX. Control of Technical Progress 

1. Reasons for restrictive attitudes toward technological change: 

(a) Business — ^to protect existing plants and investments; 

fear of loss through obsolescence ; reluctance to increase 

debt ; desire for monopoly profits, etc. 
(6) Labor — makes for unemployment; causes destruction of 

skills, etc. 

(c) Social (consumers) — does not decrease total costs of 

production. Must take into account costs of capital 
destruction and of supporting idle men. 

(d) Change subsidized under present system to a dangerous 

degree. 

2. Reasons for control : 

(a) More order and security in economic life. 

(b) Better guidance of direction of social change. . 

3. Possibility and extent of control : 

(a) Inventions are seldom "sudden"; most of them are slow 

in development and not too radical in character ; hence 
control is possible. 

(b) Unguided adjustments to technical changes no longer 

easy for individual or beneficial to society. 

(c) To what extent is control exercised today — and how? 

4. Methods of control. 



CONXENTRATION OF ECONOMIC POWER 301 

X. Protection of Worker Against Effects of Rapid Technical 

Change 

(a) Advocated by J, S. Mill and others. 

( b ) Elasticity of demand as a basis for predicting labor displacements. 
{c) Limitation of profits. ("If the profits of innovations possessing 

legal monopoly rights were reduced, prices conforming more 
nearly to lowered cost, consumers' purchasing power would 
be enlarged and the chain of forces which build for employ- 
ment would be set in motion.") 

(d) Dismissal wage (would cost be borne by employer or State: 

Would burden be shifted to consumer?) as a means of retarding 
introduction of labor-saving devices. ("Social responsibility 
for victims of technological change.") 

"Conversely, it will be argued that since technological pro- 
gress results in an advance in real wages even when prices are 
rising, and a greatly augmented increase in real wages when 
the general price level is falling, the worker should bear some 
of the cost incident to technological change." 

(e) Placement and vocational guidance. 
{/) Regularization of employment. 

{(/) Shorter work- week. 

(h) Unemployment insurance. 

(i) Unemployment Relief and Public Works. 

(j) Wage subsidies. 

(k) Government spending. 

{1} Decentralization of industry. 

XI. Some Major Issues Today 

1. Is "capitalism" unable to absorb further technical improvements? 

2. Effects of technical change in shifting the balance in world basic 

industries (coal, cotton, etc.) and in world economic power. 

3. Technical change and democracy. Do technical developments 

provide a basis for a new agro-industrialism based on social 
control of the strategic economic forces (electric power, trans- 
portation, banking) , decentralization of secondary industries, and 
wide diffusion of gains from technical advance — which would 
supply an economic foundation for a Renaissance of political and 
social democracy ? 



INDEX 

Page 
ADAMS, MARK: The automobile — a luxury becomes a necessity (1938) ; 

cited 260 

ALEXANDER HAMILTON INSTITUTE . ; 219 

AMERICAN ACADEMY OF POLITICAL AND SOCIAL SCIENCE: An- 
nals (Jan. 1939) : cited 271 

AMERICAN BUREAU OF METAL STATISTICS : Yearbook (1938) ; cited- 248 

American economic review (Mar. 1921) : cited 38 

Ameriean f ed era tio) list (Aug. 1930) ; cited 44,45 

AMERICAN FEDERATION OF LABOR 37, 50, 219 

AMERICAN IRON AND STEEL INSTITUTE : 

Annual statistical report (1935, 1939) ; cited 237 

Iron and steel works directory of the United States and Canada 

(1935) ; cited 241 

American journal of sociology (Nov. 1937) ; cited 175 

American machinist (Oct. 4. ]939) ; cited 258 

AMERICAN STATISTICAL ASSOCIATION: 

Journal ; cited 39. 125 

Proceedings ; cited 39, 47 

Quarterly publications (June 1917) : cited 142 

AMERICAN TELEPHONE AND TELEGRAPH CO 112 

American wool and cotton reporter (Aug. 29, 1940) ; cited 109 

AMIDON, BEULAH: Jobs after forty (1939) ; cited 163 

ANACONDA COPPER MINING CO 200,248 

ANDERSON, H. DEWEY III, xi, xil, xvi 

ASSOCIATION OF COTTON TEXTILE MERCHANTS 274 

AULL. RUTH m. xii 

Automohile facts: cited 163 

BABBAGE, CHARLES : On the economy of manufactures (1832) ; cited--_ 10 

BAER, J. J. : More for the company, more for the girls (1939) ; cited 123 

BALDWIN, ARTHUR R. : Motion-economized bench— costs off 40 percent 

(1939) ; cited 123 

BALL BROTHERS CO 216 

BARNES. C. A.: Motion economy on labeling operations (1939) ; cited 119 

BARNETT, GEORGE E. : 

Chaiiters on machinery .nd labor: III. Machinery and the displace- 
ment of skill (1925) ; cited 27 

Chapters on machinery and labor : IV. The introduction of machinery 

and trade-union policy (1926) ; cited '_ 37 

Machinery and labor (1926) ; cited__ 117,140 

BERMAN. EDWARD, joint author. See Curtis, W. R. 

BETHLEHEM STEEL CORP 122 132. 241 243 

BEVERIDGE, SIR WILLIAM H. : Unemployment : a problem of industrv 

(1909); cited _'_ ,% 

BLAIR, JOHN M m, xn. xvi 

Seeds of destruction (1938) ; cited 176 

BLUESTONE, D. W., joint author. See Van Tassel, A. J. 

BOWDEN. WITT: Wages, hours, and productivity in industrial labor. 

1909 to 19.39 (1940) ; cited. 1 129 023 

BOWDEN. WITT. See U. S. Bureau of Labor Statistics: Technological 

changes and employment in the lamp industrv. 
BOWDEN, WITT, joint author. See Perlo, Victor. 

BREID. ELIZABETH W ^_ _____ xn 

BRISTOL CO ____'" ~~~_ "_"" 123 

BROOKS, CHARLES F r__Z_ZI Z 186 

3Q3 



304 INDEX 

Page 

BROWN. A. BARRATT: The machine and the worker (1934) ; cited 156 

BURR, EUGENE W 243 

BURTT, HAROLD E. : Psychology and industries efficiency (1929); 

cited 118-121 

CAHEN, ALFRED, Joint author. See Stone, N. I. 

CAIRNES, J. E. : Some leading principles of political economy newly ex- 
pounded (1878) ; cited 25,26 

CALIFORNIA DEPARTMENT OF INDUSTRIAL RELATIONS _ 163 

CEMENT INSTITUTE 254, 255 

CEMENT MANUFACTURERS' PROTECTIVE ASSOCIATION 254 

CENSUS OF UNEMPLOYMENT IN MASSACHUSETTS, 1934 136 

CENTRAL STATISTICAL BOARD. Report of the executive secretary 

(1939) ; cited 224 

CHAPMAN, THOMAS G. See Bureau of Mines : Concentration, etc. 

Chemical and metallurgical engineering (Mar. 1936) ; cited 116 

Cigarette industry rules out rule of thumb (1936) ; cited 116 

CLARK, JOHN B 33 

Commercial standards monthly (Aug. 1930) ; cited 96 

COMMITTEE ON RECENT ECONOMIC CHANGES: Recent economic 

changes in the United States (1929) ; cited 39-43,115 

Commodity review (Nov. 21, 1929) ; cited 248 

COMMONS, JOHN R 36 

CONGRESS OF INDUSTRIAL ORGANIZATIONS 50,219 

The economic outlook (1940) ; cited 125 

CONSOLIDATED & McCAY LASTING MACHINE CO 213 

COOKE, MORRIS and MURRAY, PHILIP : Organized labor and production 

(1940) ; cited 160 

COPELAND, MELVIN T. : A raw commodity revolution (1938) : cited__- 247, 249 

COPPER EXPORT ASSOCIATION 248 

COPPER INSTITUTE , 248 

CORRY, A. V. and KIESSLING, O. E. See Work Projects Administration : 

Mineral technology, etc. 

COX, REAVIS : 

Competition in the American tobacco industry (1933) ; cited 264, 265 

The marketing of textiles (1938) ; cited 277 

COYLE, E. D.. joint author. See Day, E. E. 

CROWDER, WALTER F., joint author. See Thorp, Willard L. 

CRUM, WILLIAM L. : Corporate size and earning power (1939) ; cited___ 217, 218 

CURRIE, LAUCHLIN 147 

CURTIS. W. R.. KEIM, W. G., and BERMAN, EDWARD. See Works 
Progress Administration : The skill of brick and stone masons, etc. 

DAUGHERTY, CARROLL: Labor problenis in American industry (19.38) ; 

cited 147, 157, 162 

DAUGHERTY, C, DE CHAZEAU, M., and STRATTON. S. : The economics 

of the iron and steel industry (1937) ; cited 238, 243 

DAVIS, J. J. : Productivity of labor and industry (1927) ; cited 38 

DAVIS, M. B., joint author. See Magdoff, Harrv. 

DAVIS, WATSON .* 185. 1S7 

DAY, E. E. and THOMAS : The growth of manufacturers ; cited 38 

DAY, E. E., PERSONS, W. M., and COYLE, E. D. : An index of the phys- 
ical volume of production (1921) ; cited ^__ 38 

DENNISON, HENRY S 40.43 

DE CHAZEAU, M., joint author. See Daugherty, C. 

DETROIT STEEL PRODUCTS CO 122 

DICKERMANN, NELSON, joint author. See Janssen, W. A. 

DIRECTOR, A., joint author. See Douglas, P. H. 

DOUGLAS, PAUL H. : 

Controlling depressions (ia35) ; cited ISO 

Technological unemployment (1930) ; cited 44,46 

DOUGLAS, PAUL H., and DIRECTOR, A. : The problem of unemployment 

(1931) : cited L 44 

DRURY, H. B., joint author. See Nourse, E. G. 

du PONT DE NEMOURS. E. I. & Co__ 210 

The economio outlook (Feb. 1940) ; cited 125 



INDEX 305 

Page 
EDWARDS, ALBA M. : Social economic gi'oups in the United States 

(1917) ; cited 142,143 

Electrical tcorld, annual statistical numher (1940) ; cited 203 

ENGELS, FREDERICK, joint author. See Marx, Karl. 

EPSTEIN, RALPH C. : The automobile industry (1928) ; cited 256 

EVANS, W. D. : Individual productivity differences (1940) ; cited 164 

FABRICANT, SOLOMON : The output of manufacturing industries, 1899- 

1937 (1941); cited 223 

Factory m a/nag em ent and maintenance (Dec. 1939) ; cited 119,122,123 

Fair Labor Standards Act -168 

FEDERAL POWER COMMISSION 202 

FEDERAL TRADE COMMISSION: 

Chain stores, special discount and allowances to chain stores and inde- 
pendent producers, tobacco trade; cited 265 

Decisions ; cited 243 

Docket No. 3167, In the matter of the Cement Institute et al. (1937) ; 

cited 255 

House furnishings industries (1923) ; cited 286 

Price bases inquiry, the basing-point formula and cement prices 

(1932) ; cited 256 

Report in response to Executive Order of May 30, 1934, with respect 

• to the basing-point system in the steel industry (1934) ; cited 242 

Report on motor-vehicle industry (1939) ; cited 260,261 

FETTER, FRANK A.: The masquerade of monopoly (1931) ; cited— _^— 248 
FIELD, P. M., joint author. See Perazich, G. 
FLEDDERUS, M., joint author. See Van Kleeck, M. 
FLORENCE, P. SARGANT : 

Economics of fatigue and unrest (1924) ; cited 156,137 

The logic pf industrial organization (1933); cited ; 198,298 

FOLTS, F. E.. joint author. See Robbins, E. C. 

FORD, EDSKL xu 

FORD, HENRY .-. 109 

FURNAS, C. C. See Work Projects Administration : Technological trends, 
etc. 

GAY, EDWIN F : 39 

GENERAL ELECTRIC CO 210,214,215 

GENERAL MOTORS RESEARCH CORP 210,211 

GIDE, CHARLES 36 

GOLDMARK, JOSEPHINE: Fatigue and efficiency (191J) ; cited—'- 157 

GOODYEAR SHOE MACHINERY CO 214 

GOURVITCH. ALEXANDER. See Work Projects Administration. 

GRACE, EUGENE G 243 

GREEN, WILLIAM , xu 

GREGORY. T. E. : Gold, unemployment, and capitalism (1933) ; cited 10 

HAMILTON, WALTON, et al. : Price and price policies (1938) ; cited 260 

HANSEN. ALVIN 172-176, 184 

HARDING, WARREN G 169 

HARPER'S MAGAZINE (June 1940) ; cited 105 

HARRISON, GEORGE xii 

HARRISON, WILLIAM HENRY ^ xii 

HARTFORD EMPIRE CO 21.5, 216 

HARVARD FATIGUE LABORATORY 164 

HARV-ARD UNIVERSITY, GRADUATE SCHOOL OF BUSINESS 
ADMINISTRATION : 

Bureau of Business Research : Bulletin No. 56, Distribution of tex- 

tUes; cited — 281 

Business research studies No. 19 (1938) ; cited 247 

HAYNES, WILLIAM: Men, money, and molecules (1936) ; cited— J 108 

HAZEL-ATLAS CO 216 

HEINRICH, H. W. : Prevention oi industrial accidents (1931) ; cited 15T 

HILTS, H. E. See National Recovery Administration, etc. ; The manufac- 
• turing capjicity, etc. 

HOBSON, JOHN A 36 

HOOK, MARCUS xii 

277i551 — 41 — No. 22 21 



306 INDEX 

P8B» 

INDUSTRIAL FATIGUE llESEARi 'H BOARD: Report No. 47, Five-hour 

spells for women with reference to rest pauses (1928) ; cited 120 

INDUSTRIAL RAYON CO ^ 133 

INDUSTRIES : Passim, see especially. 

Cement 250-256 

Cigarette 262-266 

Cotton goods 1 272-278 

Electric light and power 266-272 

Furniture 283-288 

Iron and steel 235-243 

Motor vehicles 256-261 

Nonferrous metals 244-250 

Woolen and worsted goods 278-283 

INMAN, O. L 185 

INTERNATIONAL GENERAL ELECTRIC CO 215 

INTERNATIONAL LABOUR <JFFICE : 

Hours of work in the United States (1939) ; cited 160 

International labour review (Aug. 1939) ; cited 169 

The social aspects of rationalization (1931) ; cited 87, 118, 122 

Three sources of unemployment (1935) ; cited 87 

Iron Af/c; cited 138,241 

JANSSEN, W. A., DICKERMANN, NELSON, and RICHARDS, K. M. 

See National Recovery Administration, Division of Review, etc. ; cited— 254 

JANTZEN KNITTING MILLS 96 

JEROME, HARRY: Mechanization in industry (1934) ; cited 51-57, 63, 

114, 137, 141, 142, 221, 236, 237 

JEVONS, STANLEY 11. 33 

Journal of Commerce (Feb. 19. 1940) ; cited 274 

Journal of Political Economii (Aug. 1929) : cited 39 

KEIM, W. G., joint author. See Curtis, W. R.. and also NeLson, S. 

KENNECOTT COPPER CORP 200 

KENNEDY, THOMAS : xii, 100 

KETTERING, CHARLES F xii, 189, 208. 211 

KIESSLING, O. E., joint author. See Corry, A. V. 

KING, WILFRED I. : The relative volume of technological unemploy- 
ment (1933); cited 47 

KJAER, S., joint author. See Kossoris, M. D. 

KLINE, GORDON M. : Plastics and their uses in the automotive in- 
dustry (1940) ; cited 109, 231 

KOEPKE. CHARLES A. : A jo») analysis of manufacturing plants in 

Minnesota (1934).; cited 143 

kOEPKE, C. A. a-nd WOAL, S. T. .^ee Work Projects Administration 

and University of Minnesota, etc. 
KOSSORIS, MAX D. : Industrial iniuries and the business cycle (1938) ; 

cited 160, 161 

KOSSORIS, MAX D. and K.TAER, SWEN: Industrial injuries in the 

United States during 19:57 (1939) r cited 160 

KREPS, THEODORE J xiii 

LANGDON, J. N., joint author. See Wyatt, S. 

LAUDERDALE. LORD (Maitland. James) : An inquiry into the nature 

and origin of public wealth (1804) ; cited 5,6 

"LAUSANNE SCHOOL" i 33 

LEVEN. M., MOULTON. H. G., and WARBURTON, C. : America's ca- 

pacitv to con.sume; cited 69 

LINDSAY, A. D. : Karl Marx's Capital (1925) ; cited 13 

LORILLARD, P. CO 265 

LORWIN, LEWIS L ni, xn, xv, 234, 296 

Labor productivity and industrial prices ; cited 234 

LUBIN, ISADOR 78, 133 

The absorption of the unemployed by American industry (1929) ; 

cited 1 39 

Measuring the labor ab.sorblng power of American industry (1929) ; 

cited , 39 



INDEX 3Q7 

Fasje 
LUCAS, ARTHUR : Industrial reconstruction and the control of com- 
petition — the British experiments (1937) ; cited 197 

LUDDITE RIOTS •_ 6 

MoCORMICK, FOWLER xn 

McCULLOCH, J. R. : The principles of political economy (1830) ; cited- 9, 15, 16 
MAGDOFF, HARRY, SIEGEL, I. H., and DAVIS, M. B. See Work 
Projects Administration, etc. : Production, employment, and produc- 
tivity. 

MAIDEN FORM BRAISSERE CO r:>3 

MAITLAND, JAMES. " See Lauderdale, Lord. 

MALTHUS, THOMAS R. : Principles of political economy ; cited 22 

MARSHALL, ALFRED: Principles of economies (2d ed. 1891) ; cited 33,34 

MARX, KARL: 

Capital (trans. 1921) ; cited ^ 13-25 

Das Kapital (1867) ; cited 13 sA 

MARX, KARL, and ENGELS, FREDERICK: Manifesto of the Com- ' 1 

munist Party (Amer. ed.) ; cited Ijj 

MASSACHUSETTS DEPARTMENT OF LABOR AND INDUSTRIES: . ' 
Report on the census of unemployment in Massachusetts (1934) ; cited- 136! 
MASSACHUSETTS INSTITUTE OF TECHNOLOGY, INDUSTRIAL 

RELATIONS SECTION 163 

MEDICAL RESEARCH COUNCIL, INDUSTRIAL HEALTH RESEARCH 
BOARD : 

Report No. 77, Fatigue and boredom in repetitive work (1937) ; cited. 119. 

156 
Report No. 82, The machine and the worker (1938) ; cited__^— 155,157,158 

MERRILL, E. D 186 

MICHL, H. E. : The textile industries (1938) ; cited 282 

MILL, JAMES 15 

MILL, JOHN STUART 10, 1.5 

Principles of political economy (1877) ; cited 11-13 

MILLS, FREDERICK C. : 

The anatomy of prices, 1890-1940: cited l_ .57.58 

Economic tendencies in the United States (1932) ; cited 57-66,125,126 

Employment opportunities in manufacturing industries of the United 

States (1938) ; cited 57,66 

Prices, in recession and recovery (1936) ; cited 57 

MITCHELL, W^ESLEY C 36,39,42 

MOSHER, W. E. : Defects of state regulation of public utilities In the 

United States (1939) ; cited ^_ 271 

MUMFORD, LEWIS: Techniques and civiUzation (1934) ; cited 108,184 

MURRAY, PHILIP- xn, 132, 133 

MURRAY, PHILIP, joint author. See Cooke, Morris. 

MYERS. R. J. : Occupational readjustment of displaced skilled workers 

(1929) ; cited 39 

NATIONAL ASSOCIATION OF WOOL MANUFACTURERS: Bulletin 

(1938) ; cited r. 282 

NATHAN, ROBERT R 219 

NATIONAL ASSOCIATION OF MANUFACTURERS 49 

NATIONAL BUREAU OF ECONOMIC RESEARCH 51,57,223 

Bulletin: cited 57,66,167 

NATIONAL BUREAU OF STANDARDS: Letter circular (LC 501), Sep- 
tember 3, 1937; cited_- 225 

Reductions in varieties effected by the application of simplification 

practices ; cited 96 

NATIONAL INDUSTRIAL CONFERENCE BOARD 49, 77, 169, 219 

• Industrial standardization (1929) ; cited 96 

Machinery, employment, and purchasing power (1935) ; cited—, 173 

Systems of wage payment (1930) ; cited 122 

Wages, hours; and employment in the United States, 1914-36; cited- 148, 

149, 171 

National Industrial Recovery Act 168,241,249,254 

NATIONAL ORGANIZATION FOR TAXATION OF LABOR-DISPLAC- 
ING DEVICES 50 



308 INDEX 

NATIONAL RECOVERY ADMINISTRATION : 

Division of Review : Page 

History of the code of fair competition for the Portland cement 

industry (1935) ; cited 254 

The tobacco study (1936) ; cited 262,265 

Research and Planning Division : 
The manufacturing capacity, volume, and costs of Portland cement 

in the United States (1934, unpub.); cited__: 250 

Preliminary report ou codal provisions relating to production and 

capacity control (1938, unpub.) ; cited 248 

Preliminary report on study of regularization of employment and 
improvement of labor conditions in the automobile industrv 

(1935) ; cited _«_ 107,113-117,159,257 

NATIONAL RESOURCES COMMITTEE: Consumer incomes in the 

United States, their distribution in 1935-36 (1938) ; cited 182 

The structure of the American economy (1939) ; cited— 190, 207, 234, 254, 294 

Technological trends and national policy (June 1937) ; cited *110, 

133, 176, 236, 245, 267, 268 

Unemployment and increasing productivity (1937) ; cited 68-74 

NELSON, S., and KEIM, W. : 

Price behavior and business policy (1940) ; cited 107 

NELSON, SAUL, joint author. See Stone, N. I. 

NICHOLSON, J. S. : The effects of machinery on wages (1878) ; cited__ 26,27 

NIKLASON, C. R. See United States Bureau of Foreign and Domestic 

Commerce: Furniture, etc. 
NIXON, R. A., and SAMUELS, P. A.: E.stimates of unemployment in 

the United tSates (1940) ; cited 219 

NOURSE, E. G., and DRURY, H. B. : Industrial price policies and eco- 
nomic progress (1938) ; cited 248 

NYMAN, R. C. : Labor extension in a cotton mill (i§34) ; cited 159 

OGBURN, WILLIAM F. : 

The influence of inventions on American social institutions in the 

future (1937) ; cited 175 

Machines and tomorrow's world (1S38) ; cited 175 

OHIO STATE DEPARTMENT OF INDUSTRIAL RELATIONS 70 

O'MAHONEY, JOSEPH C xi.xv 

O'NEILL, CHARLES 100 

O'NEILL, JOHN J.: Enter atomic power (1940) ; cited 105 

OWEN,, W. C. : Bonus drops packing costs 10 percent (1939) ; cited 122 

OWENS-ILLINOIS GLASS CO 215,216 

PERAZICK, GEORGE, and FIELD, P. M. See Work Projects Adminis- 
- tration : Industrial research, etc. 

PERLO, VICTOR, a7id BOWDEN, WITT : Unit labor cost in 20 manu- 
facturing industries, 1919-1989 (1940); cited 127,150,152 

The personnel journal (Feb. 1934) ; cited 159 

PERSONS, W. M., joint author. See Day, E. E. 

PHELPS DODGE CORP 200 

PLANT, THOMAS J 214 

POGUE, JOSEPH E 100 

POLAKOV, WALTER N.: 

The power age (1933) ; cited. 203' 

Unused productive and technical capacity in the United States 

(1935) ; cited 238 

POLAKOV, WALTER N. See United Mine Workers of America, etc.: 

Monograph No. 3. 
POSNER, HAROLD F., joint author. See Weintraub, David. 
POTTER, A. A., and SAMUELS. M. M. See National Resources Commit- 
tee : Technological trends, etc. 

Poioer (June 1934) ; cited 267 

PRESIDENT'S COMMITTEE ON ECONOMIC SECURITY 219 

PRESIDENT'S CONFERENCE ON UNEMPLOYMENT, COMMITTEE 
ON RECENT ECONOMIC CHANGES : Report. See Committee on Re- 
cent Economic Changes. 

Quarterly journal of economics; cited 27. 37 

REFERENCES TO LITERATURE AND SOURCES. See individual au- 
thors and titles. 



INDEX 309 

r \ge 

Review of economic statistics; cited 38, . J9 

REPUBLIC STEEL CO 241.243 

REYNOLDS TOBACCO CO 265 

RICARDO, DAVID : Principles of political economy and taxation (1817) ; 

cited 7-11,16 

RICHARDS, K. M., joint author. See Janssen, W. A. 

RIEVE, EMIL XII 

ROBBINS, E. C. and FOLTS, P. E. : Introduction to industrial manage- 
ment (1933) ; cited 96 

ROGERS, H. O. : Employment prospects in the petroleum and natural- 
gas industry (1940) ; cited 190 

ROOSEVELT, FRANKLIN D. : Message to Congress April 20, 1938. 
SAEGER, G., joint author. See Taworski, N. 

SARGENT, Noel 49 

SAPPINGTON, C. O. : Industrial health, asset or liability industrial com- 
mentaries (1939) ; cited 162 

SAY, JEAN-BAPTISTE : Traite d'^conomie politique (1814); cited 5-7, 

16, 25, 33, 44 

SENIOR, NASSAU 15, 21 

SIEGEL, I. H., joint author. See Magdofif, Harryr 

SISMONDI, J. C. L. SIMONDE DE : Nouveaux prt^incipes d'^conomie poli- 
tique (1819); cited 7,25,36 

SMITH, ADAM 18 

SNYDER, CARL: Capitalism the creator (1940); cited 177,178 

SOCIETY FOR THE ADVANCEMENT OF MANAGEMENT: Journal 

(1939) ; cited 118 

SOCIETY OF AUTOMOTIVE ENGINEERS: Journal (May 1940) ; cited— 109 

SPENCER, v., joint author. See Yaworski, N. 

STANDARD STATISTICS CO., INC.: Standard individual corporation 

description (1940); cited 214 

STEEL; cited 139, 238 

STERN, BORIS : 

Labor productivity in the glass industry; cited 38 

Mechanical changes in the cotton textile industry, 1910 to 1936 

(1937) : cited 116,117,273 

Mechanical changes in the woolen and worsted industries, 1910 to 

1936 (1938) ; cited 278,279 

STEWART, ETHELBERT: 

Industrialization of the feeble-minded (1928) ; cited 155 

Ultimate effects of automatic machine production (1929) ; cited — 38 
STEWART, WALTER W. : An index number of production (1921)-; cited- 38 
STONE, N. I., CAHEN, ALFRED, and NELSON, SAUL : Productivity 

of labor in the cotton-garmeilt industrv (1938): cited 120,158 

STONE, PETER A.: Toward more housing (1940); cited 188 

STRATTON, S., joint author. See Daugherty. 

STUDY, NATURE 0F___: XT, xiii, xv, xvi, 4 

"-SULLIVAN, MINE 249 

TAYLOR, FREDERICK W. : The principles of scientific management 

(1929) ; cited 118 

TECHNOLOGY : 

Accidents, industrial diseases 160-163 

Age of worker 163-165 

Bacl'sround 5-31 

Classic 'sts V. Sismondi 6-10 

Marx, .^arl 13-25 

Mill, John Stuart - 10-13 

Pioneer studies 1886-99 2-5-31 

Changes, by industry 1 89-96 

Changes, compensatory 167-194 

Concentration 195-218,234-272 

Conclusions 219, 220 

Corporate size 217, 218 

Displacement 125-147 

Labor . ^_^.. L 125-133 

Skill ^ r.: 13a-147 



310 INDEX 

TECHNOLOGY— Continued. ^^^^ 

Economic problems ^'^"^ 

Economic thought 3-83 

Efficiency ^^197-208 

Electrical energy per man-hour 38&-291 

Fuels 99-106 

Hours 167-171 

Industrial research 208-212 

Labor-saving techniques : 9fr-123, 232,233 

Machinery 114-117 

Management — 118-123 

Materials 106-110 

Nature xi. xn, xv, x\i 

New industries 172-194 

Patents , 212-217 

Plastics ^ 230,231 

Power 99-106 

Power differential — economic concentration 292-295 

Prices 234-388 

Problems ^ 32-47 

Compensatory theory 43--47 

Neoclassicism and its critics ' 33-37 

Prosperity and unemployment, U. S., 1924-29 37-43 

Technological change and the depression, 1929-33 43-47 

Processes 111-114 

Productivity ^ 221-224,234-388 

Indexes-- 1 90-95, 223, 224 

Measures 221, 222 

Recent studies 49-83 

Changes in industries, combined, 1919-36 78-81 

Changes in industries, individual, 1919-36___ 81-83 

Conclusions 83 

Labor displacement 51-57 

Mechanization 51-57 

Nature 49-51 

Productivity, prices, and employment 57-68 

Reemployment opportunities 68-83 

W. P. A.'s "Production, employment, and productivity" analyzed — 74-88 
W. P. A.'s "Unemployment and increasing productivity" analyzed — 68-74 

Standardization j 225-227 

Strain, nervous and mental 155-160 

Topical outline ^ 296-301 

Unemployment 3^125-1615 

Unit labor costs 147-154 

Wages ^ 228-220 

TEEPLE, JOHN E 108 

TEMPORARY NATIONAL ECONOMIC COMMITTEE: 

Hearings ; cited xii, xvi, 100-108, 108-117. 132, 133, 139, 140, 147, 159, 

172-174, 179, 180, 184-189, 199, 208-211, 215-217, 237. 242, 248, 248 

Monograph No. 1; cited 107,173,182,183,190 

Monograph No. 8; cited 188 

Monograph No. 27; cited___J 1 196,210 

THATCHER AND LIBERTY GLASS COS 216 

THELEN, ROLF. See United States Department of Agriculture : Report 
No. 117, etc. 

THOMAS, R. J XII 

THOMAS, W, jcnnt author. See Day, E. E. 

THORP, WILLARD L. and CROWDER, WALTER F. : The structure of 

industry (1940) ; cited 196 

Time (Feb. 13, 1933) ; cited _- 109 

TUGAN-BARONAWSKI 36 

UMANSKY, L. A.: 1939 electrical developments (1940) ; cited 238 

UNITED AUTOMOBILE WORKERS 159 

UNITED MINE WORKERS OF AMERICA, ENGINEERING DEPART- 
MENT : Monograph No. 3, Manufactured gas versus natural gas 
(1940) ; cited , 177 



INDEX 311 

Page 

UNITED SHOE MACHINERY CO 213.214 

UNITED STATES BUREAU OF FOREIGN' AND DOMESTIC COM- 
MERCE : 

Furniture distribution in the midwest (1932) ; cited 288 

Rubber news letter (Oct. 15. 1939) ; cited 110 

UNITED STATES BUREAU OF LABOR STATISTICS : 

Bulletin ; cited 38, 120. 121 

Changes in retail prices of electricity. 1923-38 (1939) ; cited 270 

Comparison of employment and productivity in manufacturing in- 
dustries, 1919-25 (1927); cited 38 

Displacement of labor by machinery-in the glass industry (1927) ; cited 38 
Displacement of Morse operators in commercial telegraph offices 

(1932) ; cited 140 

Employment and production in manufacturing industries, 1929 to 

1936 (1939) ; cited 148-150 

Handbook of labor statistics (1936) : cited 163 

Labor cost of production and wages and hours of labor in the paper 

box board industry; cited 38 

Labor efficiency and productiveness in sawmills (1924) ; cited 38 

Labor productivity and costs in certain building trades (1926) ; cited- 38 
Labor productivity and displacement in the electric light and power 

industry (1932); cited 110,115,266,267 

Labor productivity and labor costs in cotton manufacturing (1926) ; 

cited 38 

Labor productivity in the automobile tire industry; cited 123, 132,233 

Man-hours of labor per unit of output in steel manufacture (1935) ; 

cited 238. 

Monthly labor review; cited 38. 97, 98, 110, 11.5, 

116, 127, 129, 140, 148-152,- 155, 157, 160, 164, 190, 223, 238, 278, 279 

Occupation hazards and diagnostic signs (1933) ; cited : 161 

The problem of the worker displaced by machinery (1927) ; cited .38 

Productivity of labor in the cotton-garment industry (1938) ; cited 1-58 

Productivity of labor in merchant bla.st furnaces (1928) ; cited 121,236 

Retail prices; cited 269.270 

Technological changes and employment in the electric-lamp industry 

(1933) ; cited , 130,131 

Time and labor costs in manufacturing 100 pairs of shoes (1924) ; cited- 38 
Wages, hours, and productivity of industrial labor, 1909 to 1939 (1940) ; 

cited 154 

UNITED STATES BUREAU OF MINES : 

Bulletin (1936) ; cited 244 

Concentration of copper ores in North America (1936) ; cited 244.245 

Lead and zinc mining and milling in the United States (1935) ; cited- 201. 245 
UNITED STATES BUREAU OF THE CENSUS : 

Background 1940 census: cited • 106 

Census of manufactures; cited 60.75,77,203,205,207.229.250 

Monograph No. 9. See Day, E. E. 

A social-economic grouping of the gainful workers of the United States 

(1938) ; cited 143 

UNITED STATES CABINET COMMITTEE TO INVESTIGATE CONDI- 
TIONS IN THE COTTON-TEXTILE INDUvSTRY : A report on the condi- 
tions and problems of the cotton-textile industry (1935) ; cited 276 

UNITED STATES CONGRESS: 
House of Representatives : 

Committee on Labor, subcommirtet' : Hearings; cited 40.50 

House Doc. No. 468 (76th Cong..' l.st sess. K Federal Trade Commis- 
sion: report on motor-vehicle industry (1939) : cited 260 

Senate : 

Committee on Education and Labor: Hearings; cited 39 

Senate Doc. No. 71 (73d Cong.. 1st sess.). Cement industry; cited_-_ 2.54 

Senate Doc. No. 86 (73d Cong.. 2d sess.) ; cited 265 

Senate Doc. No. 159 (73d Cong., 2d sess.). Practices of the steel in- 
dustry under the code (1934) ; cited .241 

Senate Doc. No. 173 (75th Cong.. 3d sess.). Strengthening of antitrust 
laws (1938) ; cited ^ __ 1195 



312 i^^DEx 

UNITED STATES DEPARTMENT OF AGRICULTURE : Page 

The agricultural situation (Sept. 1940) ; cited 109 

Report No. Ill, The substitution of other materials for wood (1917) ; 

cited 108 

UNITED STATES DEPARTMENT OF COMMERCE: Survey of current 

business (1936) ; cited (i9 

UNITED STATES DEPARTMENT OF LABOR, DIVISION OF LABOR 
STANDARDS : Occupational poisoning in the viscose rayon industrv 

(1940) ; cited - 162 

UNITED STATES DEPARTMENT OF THE INTERIOR, BUREAU OF 
LABOR: Annual report of the Commissioner (1885, 1899). Sec Wright, 
Carroll D. 
UNITED STATES PUBLIC HEALTH SERVICE : Public health reports 

(Aug. 8, 1930) ; cited 163 

UNITED STATES STEEL CORP 241,243 

UNITED STATES 'TARIFF COMMISSION : 

Report No. 29, second series, Report to the United States Senate on 

copper (1932) ; cited ^ 247 

Report No. 128, second series, Iron and steel (1938) ; cited 241 

Report No. 133, second series, Incandescent electric lamps (1939) ; 

cited 196,21.5 

UNIVERSITY OF MINNESOTA, EMPLOYMENT STABILIZATION RE- 
SEARCH INSTITUTE: Bulletin (June 1934) ; cited 143 

UNIVERSITY OF MINNESOTA, joint author. See Work Projects Adminis- 
tration. 
VAN KLEIOCK, M., and FLEDDERUS, eds. : On economic planning (1935) ; 

cited '• 238 

VAN TASSEL, A. J., and BLUESTONE, D. W. See Work Projects Admin- 
istration, Mechanization, etc. 

VAUGHN, FLOYD L. : Economics of our patent system (1925) ; cited 214 

VEBLEN, THORSTEliS' 36 

VERNON, H. M. : Industrial fatigue and efficiency (1930) : cited 157 

VERNON, H. M. and M. D. See Industrial Fatigue Research Board, etc. 

Wages and Hours Art 102 

WAGNER, ROBERT F 39 

WAITE, C. P., et al. : Acute response of guinea pigs to vapors of some new 

commercial organic compounds (1930) ; cited i 163 

WARSHOW, H. T. : Representative industries in the United States 

(1928) ; cited 245, 248 

Washington Evening /Siar (June 25, 1940)5 cited 109 

WATSON, THOMAS— ___: xn 

WEBB, SIDNEY and BEATRICE " 36 

Welih Export Act 248 

WEINTRAUB, DAVID 68, 117 

The displacement of workers through increase in efficiencv and their 

absorption by industrv, 1920-31 (19.32) : cited 1 125 

WEINTRAUB, DAVID and POSNER, HAROLD L. See Work Projects 
Administration, etc.: Unemployment and increasing productivity. 

WELLS, DAVID A.: Recent economic changes (18S9) ; cited I 27-29 

Wi^'^T. MIRIAM E. See Work Projects Administration: Productivity, . 
etc. 

WESTINGHOUSE ELECTRIC &, MANUFACTURING CO ^ 210, 214, 215 

WHITNEY, A. F _ __ xn 

WOAL, S. T., joint author. See Koepke, C. A. 

WOLMAN, LEO 40 

Hours of work in American industry (1938) ; cited ^ 167 

WORK PROJECTS ADMINISTRATION. See Works Progress Adminis- 
tration. ? 
WORKS PROGRESS ADMINISTRATION : Tlie skill of brick and stone 
masons, carpenters, and painters employed Works Progress Admin- 
istration projects in seven cities in January 1937 (1937) ; cited 164, 165 

"VTORKS PROGRESS ADMINISTRATION, NATIONAL RESEARCH 
PROJECT : 

Effects of current and prospective technological developments upon 

capital formation (1939) ; cited 117, 179, 197. 202, 244 

Employment and luiemplovment in Philadelphia in 1936 and 1937 

(1937); cited . 1 134 



INDEX 313 

WORKS PROGRESS ADMINISTRATION— Continued. Page 

Fuel efficiency in cement manufacture, 1909-35; cited 103, 200, 250, 251 

Industrial instruments and changing technology (1938) ; cited 138, 237, 

238, 267 

Industrial research and changing technology (1940) ; cited 20S-212 

Mechanization in the cement industry (1939) ; cited- 97, 98, 117, 201, 250-252 

Mechanization in the lumber industry (1940) ; cited 198 

Mineral technology and output per man studies, grade of ore (1938) ; 

cited 199, 200 

Production, employment, and productivity in 59 manufacturing in- 
dustries (1939) ; cited 68, 74-83, 

93, 95-98, 127, 148, 191, 247, 253, 259, 263, 264, 267, 275, 280, 284 
Productivity and employment in selected industries, brick and tile 

(1939) ; cited 201, 202 

Recent trends in employment and unemployment in Philadelphia 

(1937) ; cited 134 

Survey of economic theory on technological change and employ- 
ment (May 1940) ; cited 5, 7, 27, 34, 35 

WORK PROJECTS ADMINISTRATION, N/.TIONAL RESEARCH- 
PROJECT and UNIVERSITY OF MINNESOTA, EMPLOYMENT 
STABILIZATION RESEARCH INSTITUTE: Changes in machinery 
and job requirements in Minnesota manufacturing, 1931-36 (1939) ; 

cited 122, 137, 143, 144, 146 

WRIGHT, CARROLL D. : 

Hand and machine labor (1898) ; cited 27, 31 

Industrial depi-essions (1885); cited 27,29,30 

The industrial evolution of the United States (1895) ; cited 31 

WUBNIG, ARTHUR W xii 

WYATT, S. and LANGDON, J. N. See Medical Research CouncU, etc. 

YATES, RAYMOND F. : Machines over men (1939) ; cited 138, 141, 238 

YAWORSKI, N., SPENCER, V., SAEGER, G., and KIESSLING, O. 
See Works Progress Administration : Fuel efficiency, etc. 

o 



BOSTON 



PUBLIC UBBABV 



,.,. Ill* 

3 5S16351920 9