Computational principles of mobile robotics 2ed.

Edition Gregory Dudek download

https://ebookultra.com/download/computational-principles-of-
mobile-robotics-2ed-edition-gregory-dudek/

Explore and download more ebooks or textbooks

at ebookultra.com
Here are some recommended products for you. Click the link to
download, or explore more at ebookultra.com

MOBILE ROBOTICS Second Edition Jaulin

https://ebookultra.com/download/mobile-robotics-second-edition-jaulin/

Computational Physics 2ed. Edition Thijssen J.

https://ebookultra.com/download/computational-physics-2ed-edition-
thijssen-j/

Computational Fluid Dynamics 2ed Edition Chung T.J.

https://ebookultra.com/download/computational-fluid-dynamics-2ed-
edition-chung-t-j/

Principles of Macroeconomics Sixth Edition N. Gregory

Mankiw

https://ebookultra.com/download/principles-of-macroeconomics-sixth-
edition-n-gregory-mankiw/
Principles of Microeconomics 7th Edition N. Gregory Mankiw

https://ebookultra.com/download/principles-of-microeconomics-7th-
edition-n-gregory-mankiw/

Principles of Economics 5th Edition N. Gregory Mankiw

https://ebookultra.com/download/principles-of-economics-5th-edition-n-
gregory-mankiw/

Principles of Economics 8th Edition N. Gregory Mankiw

https://ebookultra.com/download/principles-of-economics-8th-edition-n-
gregory-mankiw/

Principles of Microeconomics 6th Edition N. Gregory Mankiw

https://ebookultra.com/download/principles-of-microeconomics-6th-
edition-n-gregory-mankiw/

Principles of Macroeconomics 5th Edition N. Gregory Mankiw

https://ebookultra.com/download/principles-of-macroeconomics-5th-
edition-n-gregory-mankiw/
Computational principles of mobile robotics 2ed. Edition
Gregory Dudek Digital Instant Download
Author(s): Gregory Dudek; Michael Jenkin
ISBN(s): 9780521871570, 0521871573
Edition: 2ed.
File Details: PDF, 9.84 MB
Year: 2010
Language: english
9780521692120cvr.qxd 6/23/10 4:53 PM Page 1

Jenkin
Dudek
Praise for the first edition of Computational Principles of Mobile Robotics Gregory Dudek and Michael Jenkin

Computational Principles of Mobile Robotics

“ . . . a good synthesis of robotics for computer science people, and is a
useful tool for approaching the area.”– G. Gini, Computing Reviews

“ . . . clearly presents the range of topics covered in this multidisciplinary

Computational
field.” – Industrial Robot

“Technical libraries are well advised to shelve this milestone work. Readers
Principles of
from among upper-division undergraduates, graduates, and researchers
will find a high level of legibility in the writing style, the attractive print
layout, and ample quality illustrations, equations, and photographs.”
– Choice
Mobile Robotics
SECOND EDITION
Mobile robotics is a multidisciplinary field involving both computer science and
engineering. Addressing the design of automated systems, it lies at the intersection
of artificial intelligence, computational vision, and robotics.

This textbook for advanced undergraduates and graduate students emphasizes

computation and algorithms for a range of strategies for locomotion, sensing, and
reasoning. It concentrates on wheeled and legged mobile robots but also discusses
a variety of other propulsion systems. The new edition presents advances in robotics
and intelligent machines over the last 10 years, including significant coverage of
SLAM (simultaneous localization and mapping) and multi-robot systems. It includes
additional mathematical background and an extensive list of sample problems.
Various mathematical techniques that were assumed in the first edition are now briefly
introduced in appendices at the end of the text to make the book more self-contained.

Researchers and students in the field of mobile robotics will appreciate this
comprehensive treatment of state-of-the-art methods and key technologies.

GREGORY DUDEK is James McGill Professor of Computer Science and the Director of the
School of Computer Science and of the Mobile Robotics Laboratory at McGill University.
SECOND
EDITION
MICHAEL JENKIN is a Professor of Computer Science and Engineering at York University.
He has coedited eight books on human and machine vision.

Cover design by Alice Soloway

Computational Principles of Mobile Robotics

Mobile robotics is a multidisciplinary field involving both computer science and

engineering. Addressing the design of automated systems, it lies at the intersection
of artificial intelligence, computational vision, and robotics.
This textbook for advanced undergraduates and graduate students emphasizes
computation and algorithms for a range of strategies for locomotion, sensing, and
reasoning. It concentrates on wheeled and legged mobile robots but also discusses
a variety of other propulsion systems. The new edition presents advances in
robotics and intelligent machines over the last 10 years, including significant
coverage of SLAM (simultaneous localization and mapping) and multi-robot
systems. It includes additional mathematical background and an extensive list of
sample problems. Various mathematical techniques that were assumed in the first
edition are now briefly introduced in appendices at the end of the text to make
the book more self-contained.
Researchers and students in the field of mobile robotics will appreciate this
comprehensive treatment of state-of-the-art methods and key technologies.

Gregory Dudek is James McGill Professor of Computer Science and the Director
of the School of Computer Science and of the Mobile Robotics Laboratory at
McGill University.

Michael Jenkin is a Professor of Computer Science and Engineering at York

University. He has coedited eight books on human and machine vision.
Computational Principles
of Mobile Robotics
Second Edition

Gregory Dudek Michael Jenkin

McGill University York University
CAMBRIDGE UNIVERSITY PRESS
Cambridge, New York, Melbourne, Madrid, Cape Town, Singapore,
São Paulo, Delhi, Dubai, Tokyo, Mexico City

Cambridge University Press

32 Avenue of the Americas, New York, NY 10013-2473, USA
www.cambridge.org
Information on this title: www.cambridge.org/9780521692120

© Cambridge University Press 2010

This publication is in copyright. Subject to statutory exception

and to the provisions of relevant collective licensing agreements,
no reproduction of any part may take place without the written
permission of Cambridge University Press.

First published 2010

Printed in the United States of America

A catalog record for this publication is available from the British Library.

Library of Congress Cataloging in Publication data

Dudek, Gregory, 1958–
Computational principles of mobile robotics / Gregory Dudek, Michael Jenkin. – 2nd ed.
p. cm.
Includes bibliographical references and index.
ISBN 978-0-521-87157-0
1. Mobile robots. 2. Robotics – Mathematics. I. Jenkin, Michael, 1959– II. Title.
TJ211.415.D83 2010
629.8 932 – dc22 2010020795

ISBN 978-0-521-87157-0 Hardback

ISBN 978-0-521-69212-0 Paperback

Cambridge University Press has no responsibility for the persistence or accuracy of URLs for external
or third-party Internet Web sites referred to in this publication and does not guarantee that any content
on such Web sites is, or will remain, accurate or appropriate.
For Krys and Heather
 Contents

Acknowledgments page xi
Preface to the Second Edition xiii

1 Overview and Motivation 1

1.1 From Mechanisms to Computation 4
1.2 Historical Context 5
1.3 Biological Inspiration 11
1.4 Operational Regimes 11
1.5 Operational Modes 11
1.6 A Guide to This Book 12
1.7 Further Reading 13
1.8 Problems 16

2 Fundamental Problems 18
2.1 Path Planning for a Point Robot 19
2.2 Localization for a Point Robot 21
2.3 Sensing for a Point Robot 23
2.4 Mapping for a Point Robot 25
2.5 SLAM for a Point Robot 25
2.6 Looking Forward 26
2.7 Further Reading 27
2.8 Problems 27

Part One: Locomotion and Perception 29

3 Mobile Robot Hardware 31

3.1 Locomotion 31
3.2 Off-Board Communication 71
3.3 Processing 75
3.4 Further Reading 76
3.5 Problems 77

4 Non-Visual Sensors and Algorithms 82

4.1 Basic Concepts 82
4.2 Contact Sensors: Bumpers 86

vii
viii Contents

4.3 Inertial Sensors 87

4.4 Infrared Sensors 90
4.5 Sonar 91
4.6 Radar 98
4.7 Laser Rangefinders 98
4.8 Satellite-Based Positioning 100
4.9 Data Fusion 102
4.10 Biological Sensing 118
4.11 Further Reading 120
4.12 Problems 121

5 Visual Sensors and Algorithms 123

5.1 Visual Sensors 124
5.2 Object Appearance and Shading 131
5.3 Signals and Sampling 132
5.4 Image Features and Their Combination 134
5.5 Obtaining Depth 149
5.6 Active Vision 155
5.7 Other Sensors 158
5.8 Biological Vision 162
5.9 Further Reading 163
5.10 Problems 164

Part Two: Representation and Planning 165

6 Representing and Reasoning About Space 167

6.1 Representing Space 167
6.2 Representing the Robot 176
6.3 Path Planning for Mobile Robots 179
6.4 Planning for Multiple Robots 208
6.5 Biological Mapping 209
6.6 Further Reading 210
6.7 Problems 210

7 System Control 212

7.1 Horizontal Decomposition 213
7.2 Vertical Decomposition 217
7.3 Hybrid Control Architectures 223
7.4 Middleware 226
7.5 High-Level Control 226
7.6 Alternative Control Formalisms 230
7.7 The Human–Robot Interface 235
7.8 Mobile Robot Software Development as
Experimentation 237
7.9 Standard Software Toolkits 237
 Contents ix

7.10 Further Reading 238

7.11 Problems 239

8 Pose Maintenance and Localization 240

8.1 Simple Landmark Measurement 241
8.2 Servo Control 249
8.3 Recursive Filtering 250
8.4 Non-Geometric Methods: Perceptual Structure 260
8.5 Correlation-Based Localization 267
8.6 Global Localization 267
8.7 Biological Approaches to Localization 273
8.8 Further Reading 274
8.9 Problems 274

9 Mapping and Related Tasks 276

9.1 Sensorial Maps 278
9.2 Geometric Maps 279
9.3 Topological Maps 287
9.4 Exploration 291
9.5 Further Reading 294
9.6 Problems 294

10 Robot Collectives 295

10.1 Categorizing Collectives 296
10.2 Control Architectures 296
10.3 Collective Communication 299
10.4 Sensing 300
10.5 Planning for Action 301
10.6 Formation Control 302
10.7 Localization 303
10.8 Mapping 304
10.9 Further Reading 305
10.10 Problems 306

11 Robots in Practice 307

11.1 Delivery Robots 307
11.2 Intelligent Vehicles 309
11.3 Robots for Survey and Inspection 314
11.4 Mining Automation 316
11.5 Space Robotics 317
11.6 Autonomous Aircraft 319
11.7 Military Reconnaissance 320
11.8 Bomb/Mine Disposal 320
11.9 Underwater Inspection 322
11.10 Agriculture/Forestry 323
11.11 Aids for the Disabled 325
x Contents

11.12 Entertainment 326

11.13 Domestic Robots 327
11.14 Further Reading 327
11.15 Problems 328

12 The Future of Mobile Robotics 329

12.1 Locomotion 329
12.2 Sensors 331
12.3 Control 332
12.4 System Integration 332
12.5 Standardization 333
12.6 Future Directions 333

Appendix A: Probability and Statistics 335

A.1 Probabililty 335
A.2 Some Simple Statistics 338
A.3 Further Reading 339
A.4 Problems 339

Appendix B: Linear Systems, Matrices, and Filtering 341

B.1 Linear Algebra 341
B.2 Matrix Algebra 341
B.3 Signals and Systems 343
B.4 Fourier Theory 344
B.5 Sampling and the Nyquist Theorem 344
B.6 Further Reading 345
B.7 Problems 345

Appendix C: Markov Models 346

C.1 Discrete Markov Process 346
C.2 Hidden Markov Models 348
C.3 Markov Decision Process 349
C.4 POMDP 350
C.5 Further Reading 351
C.6 Problems 351

Bibliography 353
Index 381
 Acknowledgments

This book would not have been possible without the active support of our students and
colleagues who suffered through early drafts, provided imagery and papers, and put up
with us while this volume was being put together. The list of people who helped is much too
long to include here, but some require special mention: International Joint Conferences on
Artificial Intelligence (IJCAI), for letting us do a tutorial on mobile robotics that started us
on the journey; students at McGill and York, including Eric Bourque, Andrew Hogue, and
Yiannis Rekleitis, who suffered with photocopies of early drafts; Prof. Evangelos Milios
for his helpful comments and Prof. Evangelos Papadopoulos for his help on space robots;
Rob Sim for his figures; and Louis Dudek for his proofreading and helpful comments.
Finally, we thank Lauren Cowles and Cambridge University Press for encouraging us
do this in the first place.

xi
 Preface to the Second Edition

The authors surrounded by a collection of some of their robots and sensors.

The first edition of this book arose from a tutorial that was initially presented in 1993. It
evolved over a number of years and finally appeared in 2000. Robotics is, of course, a
rapidly changing field, and even as the first edition was appearing, it was apparent that it
would become dated rather quickly. Thus in late 2002, the first steps were taken toward
a second edition.
The field of mobile robotics continues to evolve, and this book has (we hope) evolved
with the field. Topics that were in their infancy when the first edition was published – such
as SLAM and multi-robot systems – have evolved into much more mature disciplines.
The mathematical foundations of much of mobile robotics has also matured, and this too
is reflected in this volume. In addition to updating the various algorithms and methods
described in the first edition, the second edition is somewhat more self-contained. Specif-
ically, various mathematical techniques that were assumed in the first edition are now
introduced, albeit briefly, in appendices at the end of the text.

xiii
Random documents with unrelated
content Scribd suggests to you:
approximated at $260 per year per man, boy and porter, who two years ago
numbered 621,341.
In 1907, Special Agent Ames, of the Interstate Commerce Commission,
reported wages on the railways of the United Kingdom as follows:

Enginemen $9.32 per week

Firemen 5.76 " "
Conductors 6.26 " "
Brakemen 6.44 " "
Shunters 5.80 " "
Examiners 5.80 " "
Signalmen 5.66 " "
Trackmen 5.58 " "

Pay of Railway Employes in Other Countries.

The contrast between the wages of American and European railway employes
is emphasized by those paid on the continent. The official statistics of the empire
show an increase of 5% in the average yearly compensation of German railway
employes in 1908. Their number and pay for that year to December 31st in the
four main classes into which they are divided were as follows:

Number and Pay of German Railway Employes by Principal Divisions for the
Year Ending December 31, 1908.

Increase
Employes Compensation Average
Division over
Number (Total) per year
1907

General administration 31,996 $25,167,240 $787 $34

Maintenance and guarding road 177,633 42,891,753 241 5
Station service and train crews 302,343 116,219,657 384 24
Switching crews and shops 187,183 75,328,084 402 18
Total 699,155 $259,606,734 $371 $19
Increase over 1907 3,598 14,216,875 — —

Combined with a falling off in revenues and an increase in the cost of materials
this increase in the compensation of employes had the effect of raising the
operating ratio of German railways from 69.01 in 1907 to 73.56 in 1908. It also
increased the proportion of wages to gross earnings from 37.25 to 40.1% and
had the effect of reducing the net revenues from 5.60% to 4.51% on the cost of
construction.
How railway labor fares under government ownership in a republic as
compared with its pay in an empire may be judged from a comparison of the
following statement as to the number and pay of the railways of Switzerland with
the like classes in the preceding table for Germany.

Number and Pay of Swiss Railway Employes by Principal Divisions in 1907.

Employes Compensation Average

Division per Year
Number (Total)

General administration 1,631 $ 780,715 $478

Maintenance and inspection of way 10,308 1,459,977 142
Transportation and train service 17,815 6,829,426 383
Porters and laborers 12,219 3,209,810 262
Total 41,973 $12,279,928 $292

The wages paid the employes of Swiss railways in 1907 amounted to only 31.9
per cent. of the gross earnings, and yet they added enough to the cost of
operation to help increase the telltale ratio of expenses to revenues from 64.99
in 1906 to 67.29 in 1907. The result was increased operating expenses per mile
and a decrease in the amount available for interest in dividends from 3.26% in
1906 to 3.23% in 1907.
As the Swiss republic has to pay 3½% on government loans its investment in
railways does not appear to be a very profitable one.

Employes of French Railways.

The employes of the railways of France are divided into the following classes:

General administration 3,119

Transportation and traffic 128,823
Traction and material 80,732
Way and structures 81,897
Auxiliaries 82,809
Female employes 29,178
Total 406,558
 The official statistics only give the compensation of employes in the division of
traction and material, where the 80,732 men employed get an average of $187
per year.
On the state railways of Belgium, firemen receive from $15.20 to $22.80 per
month, the higher wage only after 15 years' service; enginemen begin at $22.50
per month and at the end of 24 years' service work up to $38.00 per month;
conductors earn from $15.97 per month up to a maximum of $34.70; brakemen,
beginning as shunters (switchmen) at 45 cents a day, when promoted get a
minimum of $17.10 per month, from which they are slowly advanced to a
maximum of $22.00. The average railway worker in Belgium gets 2.22 francs (43
cents) a day.
Whole classes of American railway employes get more in a month than Belgian
railway employes average in a year.

The Cost of Living.

What and how great the virtue and the art,
To live on little with a cheerful heart.—Pope.
Not because it has any legitimate place in fixing the standard of railway wages,
which should be relative to the part capacity, intelligence, industry, loyalty and
experience play in railway service, but because in recent years the steady
increase in the cost of living has been made the fulcrum on which every lever to
advance wages works, is it proper to refer to the subject in this report.
Now there is nothing in the whole wilderness of economics so utterly illusive
and misleading as this same cost of living. It is as incapable of statistical
expression as the airy imaginings of a dream and yet it broods over the domestic
happiness of nations with all the disquieting effects of a nightmare—and like
every nightmare it comes from eating too much and wanting to eat more.
In economics, beyond the barest subsistence, the cost of living is not ruled by
necessity but by individual choice. Each person and family settles it along the
lines of abstinence or indulgence. It ranges from the "dinner of herbs where love
is" and the virtues of self-denial are nourished, to the feasts of Lucullus and
Pompeian profligacy in whose indulgence whole peoples have perished.
In every discussion of the subject first consideration is given to the price of
food. This amounts to measuring the cost of living with an elastic string. The
proportion of the cost of food to the cost of living varies in every land, in every
occupation and in every household. It amounts to less than 40% in an average
American family, but each family fixes it for itself. Following certain well
recognized economic laws the percentage for subsistence increases as the
income decreases. For instance, in France families with an income of under $4.80
per week spend 63% of it for food alone, whereas those with $9.60 a week
spend 53%. In England, families averaging $5.12 a week spend 67% on food,
while those of $9.60 spend 57% or less. In Germany, a similar inquiry showed
that families with an average income of $4.23 per week spent 68.7% on food
(excluding beer), or 69.5% (with beer); whereas families with an income of
$9.60 per week spent less than 57% on food "excluding beer."
The exhaustive investigation made by Commissioner Carroll D. Wright when
head of the Bureau of Labor in 1903 anticipated for the United States these
results of more recent European inquiries, as appears from the following table
showing the per cent of total expenditure made for various purposes in normal
families according to classified incomes:

Per Cent of Expenditure for Various Purposes in 11,156 Normal Families,

by Classified Incomes, 1901.
Classified income Rent Fuel Lighting Food Clothing Sundries
Under $200 16.93 6.69 1.27 50.85 8.68 15.58
$200 or under $300 18.02 6.09 1.13 47.33 8.66 18.77
$300 or under $400 18.69 5.97 1.14 48.09 10.02 16.09
$400 or under $500 18.57 5.54 1.12 46.88 11.39 16.50
$500 or under $600 18.43 5.09 1.12 46.16 11.98 17.22
$600 or under $700 18.48 4.65 1.12 43.48 12.88 19.39
$700 or under $800 18.17 4.14 1.12 41.44 13.50 21.63
$800 or under $900 17.07 3.87 1.10 41.37 13.57 23.02
$900 or under $1000 17.58 3.85 1.11 39.90 14.35 23.21
$1000 or under $1100 17.53 3.77 1.16 38.79 15.06 23.69
$1100 or under $1200 16.59 3.63 1.08 37.68 14.89 26.13
$1200 or over 17.40 3.85 1.18 36.45 15.72 25.40
All classes 18.12 4.57 1.12 43.13 12.95 20.11

While it is scarcely believable that many American families with incomes under
$200 spent less than $100 a year on food—the European percentage in such
cases being more credible—there is no reason to question the general economic
law reflected in this table, that "the proportion of income spent on food
diminishes as the income increases." But it is governed more by individual
tendencies, character and taste than by any rule or principle. Each family works
out the problem on its own account.
According to the evidence presented at recent arbitration hearings in this city,
American switchmen, as a body, belong in the classes whose family expenditures
are $1,000 or over. Irrespective of the incomes of other members of their
families, the arbitrators found "that the actual monthly earnings of switchmen in
the Chicago district, for those who worked full time runs from about $80 to $100
per month." This means over $1,000 yearly compensation. Therefore they are in
the class which spends less than 39% of its income on food.
The average income for all railway employes engaged in train service, that is,
enginemen, firemen, conductors and other trainmen, is probably above the
highest figure in the foregoing table and therefore the proportion of their income
spent for food would be approximately 36%.
But accepting 40% as approximately the proportion of the pay of all railway
employes spent on food, it follows that it takes only two-fifths of one per cent
increase in wages to take care of an increase of one per cent in the price of food.
With this in mind it becomes instructive to follow the retail prices of the
various articles of food as selected by Mr. Wright in his inquiry into the cost of
living in 1901 and adopted by the Bureau of Labor in subsequent Bulletins. These
for thirty articles of food for the eighteen years 1890 to 1907, as given in Bulletin
No. 77 of the Bureau of Labor, and for the two years 1908-1909 as computed
from Bradstreet's index and other sources of commodity prices, are given in the
following statement relatively to the average price for 1890 to 1899 == 100:

Relative Retail Prices of the Principal Articles of Food in the United

States, 1890 to 1909. (Average price for 1890-1899 == 100.0.)
Chickens
Apples, Beef, Beef, (year or
Year Beans, Beef, Bread,
Evapo­‐ Fresh, Fresh, Butter Cheese more Coffee
Dry Salt Wheat
rated Roasts Roasts old),
dressed
1890 109.0 103.3 99.5 98.8 97.5 100.3 99.2 98.8 101.3 105.4
1891 110.3 106.2 100.0 99.4 98.3 100.3 106.4 100.3 104.0 105.2
1892 99.3 102.4 99.6 99.3 99.5 100.3 106.8 101.5 103.8 103.8
1893 107.0 105.0 99.0 99.6 100.3 100.1 109.9 101.8 104.2 104.8
1894 105.8 102.8 98.3 98.2 98.9 99.9 101.7 101.6 98.6 103.3
1895 97.4 100.5 98.6 99.1 99.6 99.7 97.0 99.2 98.4 101.7
1896 88.6 92.7 99.1 99.5 99.8 99.9 92.7 97.9 97.1 99.6
1897 87.8 91.5 100.3 100.2 100.9 100.0 93.1 99.0 94.0 94.6
1898 95.4 95.9 101.7 102.0 102.1 99.8 95.1 97.5 96.8 91.1
1899 99.5 99.7 103.7 103.9 103.2 99.6 97.7 102.4 101.8 90.5
1900 95.2 110.0 106.5 106.4 103.7 99.7 101.4 103.9 100.8 91.1
1901 96.8 113.9 110.7 111.0 106.1 99.4 103.2 103.3 103.0 90.7
1902 104.4 116.8 118.6 118.5 116.0 99.4 111.5 107.3 113.2 89.6
1903 100.8 118.1 113.1 112.9 108.8 100.2 110.8 109.4 113.5 89.3
1904 99.2 116.8 112.8 113.4 108.3 103.9 109.0 107.4 120.7 91.8
1905 106.0 116.3 112.2 112.9 107.9 104.5 112.7 110.9 123.6 93.6
1906 115.6 115.2 115.7 116.5 110.8 102.3 118.2 115.5 129.1 94.7
1907 124.6 118.8 119.1 120.6 114.1 104.5 127.6 123.2 131.4 95.0
1908 126.4 138.9 126.2 131.5 116.4 124.5 123.5 121.3 128.6 94.7
1909 128.6 141.2 132.6 134.1 128.2 124.5 134.8 142.0 150.2 108.6

Milk,
Corn Fish, Fish, Flour, Fresh, Mola­‐ Pork,
Year Eggs Lard Mutton Fresh
Meal Fresh Salt Wheat unski­‐ sses
mmed
1890 100.0 100.6 99.3 100.7 109.7 98.2 100.5 104.7 100.7 97.0
1891 109.7 106.9 99.6 101.7 112.5 99.8 100.5 101.7 100.6 98.7
1892 105.2 106.8 100.1 102.2 105.1 103.6 100.6 101.2 101.0 100.5
1893 103.1 108.1 100.1 103.4 96.1 117.9 100.4 100.6 99.9 107.0
1894 102.2 96.3 100.4 101.5 88.7 106.9 100.2 100.3 97.8 101.8
1895 100.8 99.3 99.8 98.9 89.0 100.1 100.0 99.0 98.7 99.7
1896 95.0 92.8 100.2 97.5 92.7 92.5 99.9 98.7 98.7 97.4
1897 93.7 91.4 99.8 95.2 104.3 89.8 99.7 97.7 99.6 97.6
1898 95.0 96.2 100.5 98.8 107.4 93.9 99.4 97.9 100.4 98.6
1899 95.1 101.1 100.2 100.2 94.6 97.1 98.9 98.2 102.6 101.7
1900 97.4 99.9 100.4 99.1 94.3 104.4 99.9 102.2 105.6 107.7
1901 107.1 105.7 101.4 100.9 94.4 118.1 101.1 101.3 109.0 117.9
1902 118.8 119.1 105.0 102.8 94.9 134.3 103.3 102.1 114.7 128.3
1903 120.7 125.3 107.3 108.4 101.2 126.7 105.8 103.8 112.6 127.0
1904 121.5 130.9 107.9 111.7 119.9 117.3 106.3 104.0 114.1 124.0
1905 122.2 131.6 109.9 113.8 119.9 116.6 107.0 104.4 117.8 126.6
1906 123.2 134.2 116.2 116.8 180.1 128.0 108.9 105.3 124.1 137.7
1907 131.6 137.7 120.6 121.6 117.7 134.2 116.8 107.7 130.1 142.5
1908 154.0 140.2 116.2 118.4 140.0 132.1 115.4 102.2 126.4 141.6
1909 160 142.2 120.4 122.6 154.4 153.8 141.6 106.4 134.8 168.2

Pork,
Pork, Pork, Potat­‐
Salt, Vine­‐
Year Salt, Salt, oes, Prunes Rice Sugar Tea Veal
dry or gar
Bacon Ham Irish
pickled
1890 95.8 95.3 98.7 109.3 116.8 101.3 118.6 100.0 98.8 102.9
1891 96.6 98.9 99.3 116.6 116.5 102.5 102.7 100.4 99.6 105.5
1892 99.1 100.5 101.9 95.7 113.5 101.3 96.2 100.2 100.0 102.7
1893 109.0 108.7 109.3 112.3 115.6 98.4 101.5 100.1 100.0 99.5
1894 103.6 103.4 101.9 102.6 100.9 99.0 93.8 98.7 98.7 99.8
1895 99.4 99.2 98.8 91.8 94.2 98.8 91.8 98.5 98.5 98.9
1896 96.7 95.5 97.6 77.0 86.8 96.7 96.6 98.8 99.5 97.2
 1897 97.4 97.3 98.2 93.0 84.3 97.9 95.7 98.5 99.9 97.4
1898 100.2 99.1 95.1 105.4 86.3 101.7 101.3 100.7 101.2 97.9
1899 102.9 101.8 99.2 96.1 85.1 102.4 101.7 104.4 103.7 98.3
1900 109.7 107.7 105.3 93.5 83.0 102.4 104.9 105.5 104.9 98.5
1901 121.0 117.5 110.2 116.8 82.6 103.5 103.0 106.7 108.8 98.9
1902 135.6 132.5 119.4 117.0 83.4 103.5 96.1 106.0 114.9 99.1
1904 137.9 125.8 118.4 121.3 79.6 101.6 101.9 105.8 115.5 98.9
1905 138.8 126.0 118.5 110.2 81.4 102.6 98.2 105.5 123.2 102.6
1907 157.3 141.2 130.7 120.6 88.4 108.5 99.6 105.3 125.0 104.5
1908 142.4 137.4 112.0 138.4 — 105.1 100.0 108.6 124.2 112.4
1909 180.0 151.2 145.0 120.0 — 103.3 105.0 109.0 130.2 113.0

No authority is claimed for the prices in these tables for the years 1908 and
1909. They merely represent the tendencies in those years, as found in official
and unofficial wholesale prices of the several commodities, and there are often
striking divergences between wholesale and retail prices over short periods.
Eventually they follow the same course, although not always in the same
proportion.
Now let us see how the average retail price of these 30 articles of food
compares with the average daily pay of the four representative classes of railway
employes in train service for the ten years 1899 to 1909.

Average Daily Compensation Relative

Prices of
Food,
Other 1890-
Year Enginemen Firemen Conductors 1899
Trainmen
==100

1899 $3.72 $2.10 $3.13 $1.94 99.6

1900 3.75 2.14 3.17 1.96 101.5
1901 3.78 2.16 3.17 2.00 105.5
1902 3.84 2.20 3.21 2.04 110.9
1903 4.01 2.28 3.50 2.27 111.6
1905 4.12 2.38 3.50 2.31 112.5
1906 4.12 2.42 3.51 2.35 116.2
1907 4.30 2.54 3.69 2.54 120.7
1908 4.46 2.65 3.83 2.64 117.7
1909 4.46 2.67 3.76 2.60 127.7
Per cent. increase 19.9 27.1 20.1 34.0 28.2
 Here it will be observed the percentage of increase in the average daily
compensation of "Other trainmen" exceeds the relative increase in the price of
food, that of firemen almost equals it, while that of enginemen and conductors is
below it by approximately 8 points. But, as demonstrated in the table from the
Eighteenth Annual Report of the Commissioner of Labor (1903), a smaller
percentage of the income of enginemen and conductors is spent on food than of
those employes receiving lower pay.
Moreover as only two-fifths of all expenditures is spent on food an increase of
20% in wages would take care of a 50% advance in the average price of food—
provided the increase in wages was not attended by a corresponding increase in
every other item entering into the cost of living.
And right here's the rub with any attempt to measure wages by the cost of
living. Which is the egg and which is the hen, in the matter of precedence. Does
the cost of living lay the income or does the income hatch the cost of living?
Economically and theoretically it is not up to the railways to solve this world
old conundrum. Practically they are called on to meet every advance in the cost
of living of their employes to which in twenty years they have not added a nickel,
and they are denied the privilege, enjoyed by every other employer of labor, to
add its increased cost to the price of their only commodity or service—
transportation.
Today the advances in the scale of railway wages awarded, proposed and
demanded mean an increase of from $60,000,000 to $75,000,000 in the annual
"cost of living" of the railways. The advance made in 1906-07 added
$120,000,000 to the pay roll of 1908. Combined, these two advances within
three years mean an increase of approximately $200,000,000 a year to the
operating expenses of the railways without adding a single unit to efficiency of
the labor factor in railway operation.
This is equal to an annual first charge of 5% on $4,000,000,000! Imagine the
hue and cry from the press, the immediate injunctions from Washington, the
despondent wail from Wall Street, if the railways proposed to pour that much
"water" into their own cost of living without getting a mile of track, a single
engine, car, or coach, a cubic yard of ballast, one untreated tie or any semblance
of improvement or new facility to show for the vast expenditure!
And yet the railways have their increased cost of living to meet just as the rest
of us. Nothing they need and must have can be purchased at the prices of a few
years back. When you mention steel rails you have named about the only railway
necessity that has not advanced its cost of living in recent years, and the
railways have to buy 100-pound rails where five years ago 80-pound rails
sufficed, and ten years ago 70 pounds was heavy enough for the lighter cars and
engines of the time.
But at the first suggestion of advancing rates to meet advancing prices of
commodities the Commissions were overwhelmed with protests from shippers
and the paring of freight rates down went on as the prices of the goods they
carried went up.
In ten years the price of lumber advanced nearly 50%. As a cheap bulky
commodity it had enjoyed a low rate in order to move it and it was moved at the
expense of other commodities. When it was able to pay a little more toward the
cost of getting it to market the proposal of an advance was met with indignant
protests from lumber shippers and dealers and reversed thumbs by the
sympathetic commissions.
The railways pay more for their lumber and other material today than they did
ten years ago but they will have to fight for any advance in rates to meet this
part of their cost of living. It is said to be a poor rule that will not work both
ways—but the cost of living seems to have only one way of working so far as
railway economics are concerned.
Just as a straw to indicate that high prices of food are the result and not the
basis of high wages the following table of comparative prices in London and New
York from the New York Times of March 27, 1910, is instructive:

Comparative Retail Prices of Articles of Food in London and New York in

March, 1910.
London. New York.
Cents. Cents.
Apples, 1 lb 4 to 6 10
Bread, 1 lb 4 5
Butter, 1 lb 24 to 32 30 to 35
Cheese, 1 lb 14 to 16 18 to 22
Cocoa, 1 lb 16 to 36 25 to 50
Coffee, 1 lb 16 to 30 20 to 50
Currants, 1 lb 4 to 8 8 to 12
Eggs, 12 to 16 25 6 to 12—25
Codfish, 1 lb 8 to 12 15 to 29
Fish (general), 1 lb 4 to 12 10 to 25
Flour, 3 lbs 9 to 10 12
Meats:
Bacon, 1 lb 16 to 24 25 to 30
Beef, 1 lb 16 to 20 22 to 30
Pork, 1 lb 12 to 16 20 to 24
 Milk, 1 pint 4 4 to 5
Oatmeal, 1 lb 4 to 6 5 to 10
Onions, 1 lb 2 4
Oranges, 1 doz 10 to 12 18 to 50
Potatoes, 1 lb 1 to 2 3 to 4
Prunes, 1 lb 8 to 12 10 to 18
Raisins, 1 lb 6 to 10 10 to 16
Rice, 1 lb 4 6
Syrup, 1 lb 6 10
Sugar white, 1 lb 6 6
Sugar, yellow, 1 lb 4 5
Tapioca, 1 lb 8 10
Tea, 1 lb 20 to 60 30 to 1.50
Tomatoes, 1 lb 8 12

The amazing feature of this statement is that the United States produces and
exports to the United Kingdom enormous quantities of breadstuffs, meat and
provisions, which constitute the chief articles of food in London and which are
sold there at prices from 20% to 25% lower than in New York. Clearly it is the
high scale of wages that fosters the high cost of living in the United States and
there can be little question but it breeds the high wages it feeds on.
It is humanly certain, though economically unsound, that wages will continue
to advance with the cost of living and will not recede proportionately as prices of
food fall. But both will decline together when for any considerable period there is
a surplus of efficient labor for the requirements of American industry. Even
railway labor in the most stable of all employments yielded to this influence in
1893 and 1894; and the prices of food receded to the low mark in the following
years 1895, 1896 and 1897. Not until wages took their upward turn in 1898 did
the cost of food begin to show above the index average of 1890-1899.
 IV

CAPITALIZATION
According to the Twenty-third Annual Report of the Interstate Commerce
Commission the amount of railway capital, including stocks and bonds
"outstanding in the hands of the public on June 30, 1908, was $12,840,091,462,
which, if assigned on a mileage basis, shows a capitalization of $57,230 per mile
of line."
In the face of all the fustian about over-capitalization of American railways, this
is a most remarkable admission, not only of their moderate, but of their
decreasing capitalization per mile.
In its report on the Intercorporate Relationships of Railways, dated March 10,
1908, the Commission found that as the result of its investigation the figure for
railway capital outstanding in the hands of the public, "Measuring the claim of
railway securities on railway revenues," reduced the amount "from $67,936 per
mile of line (1906) to $58,050 per mile of line."
Of course there was never any justification for using the larger sum as a true
measure of railway capitalization, for it was known to contain at least 15%
duplicated capital.
In its Statistics of Railways for the year ending June 30, 1907, the Commission
gave the net amount of railway capital outstanding in the hands of the public at
that date, "assigned on a mileage basis as $58,298 per mile of line," or $1,068
more than the figure reported for 1908.
As the computation for 1908 was made on a basis of 224,363 miles of line, this
would indicate a shrinkage of no less than $239,616,480 in the par value of
railway capital. It is needless to say there was no such shrinkage.

Net Capitalization in 1909.

Following the earlier judgment of the Official Statistician, this Bureau seeks to
arrive at a fair approximation of the capitalization of the railways of the United
States through the reports of operating roads and the capitalization of the rentals
paid for leased roads. This, in the more recent language of the Statistician,
furnishes the only capitalization that "measures the claim of railway securities on
railway revenues."
 Applied to the returns received by this Bureau from 221,132 miles of operated
line, this formula yields the following result for the year ending June 30, 1909:

Summary Showing Capitalization of 368 Companies Operating 221,132 Miles

of Line for the Year Ending June 30, 1909.

Capitalization
1909
(182,046 Miles Owned)
Capital stock $6,199,919,551
Funded debt 8,015,841,805
Receivers' certificates 20,497,447
$14,236,258,803
Rental of 39,086 miles, $120,784,982, capitalized at
5%. 2,415,699,640
Total $16,651,958,443
Deduct:(a)
Railway stocks owned (actual value) $1,889,157,214
Other stocks owned (actual value) 206,461,423
Railway bonds owned (actual value) 1,054,095,905
Other bonds owned (actual value) 140,282,728
3,289,997,270

Net capitalization, 1909 $13,361,961,173

Net capitalization per mile operated 60,425

(a) The par value of these stocks and bonds owned is given as $4,739,231,832.

An estimate of $25,000 per mile for the 11,870 miles of line not reporting to
this Bureau would add $296,750,000 to the above total. From this should be
deducted $150,000,000 for the sum assigned by the Official Statistician "to other
properties," and we arrive at the following close approximation of the true
measure of the capital employed in the transportation industry of the United
States:

Net capitalization, 233,002 miles operated line, 1909 $13,508,711,173

Net capitalization per mile of line 57,962
Net capitalization per mile of track 39,730
 In computing the average capital per mile last given, no allowance has been
made for the 8,927 miles operated under trackage rights for the sufficient reason
that the rental paid therefor is represented in the total capitalization just as fully
as if so much capital had been expended in the construction of that many miles
of line.
It is worthy of note that the net capitalization thus arrived at through a
straightforward analysis of the returns of the operating companies is in
substantial agreement with the Commission's report on the Intercorporate
Relationship of Railways in 1908. The construction of 11,000 miles of line since
1906 would undoubtedly account for the difference between $58,050 and
$57,962 per mile of line.

Summary Showing Net Capitalization of the Railways of the United States,

1909-1904.
Net Per Mile
Year Capital of Line
1909 $13,508,711,173 $57,962
1908 13,007,012,563 58,864
1907 13,064,279,303 59,600
1906 12,628,000,000 57,966
1905 11,167,105,992 53,328
1904 10,711,794,278 52,099

Owing to the intercorporate ownership of stocks and bonds and the

consequent intercorporate payments of interest and dividends, it is no easy
matter to make an entirely satisfactory estimate of the return paid to capital out
of the purely transportation revenues of the railways. But the persistent
reiteration by the Official Statistician of the fictitious aggregate of all the
dividends paid by operating and non-operating companies, covering in 1908, by
his own admission, $3,927,453,365 duplicated capital, justifies the attempt.
The operating income of the roads reporting to this Bureau for the year 1909
is arrived at thus:

Gross earnings (221,132 miles operated) $2,375,141,766

Operating expenses 1,568,008,389
Net earnings from operation $ 807,133,377
Less taxes 82,650,214
 Net operating income $ 724,483,163

This $724,483,163 is the balance in the hands of the 368 companies of the
moneys received by them from transportation, or, as the Official Statistician now
calls it, "rail operations," for the payment of interest, rent, other deductions,
dividends, additions and betterments, reserves, surplus and deficits. But before
proceeding to this distribution these companies received $200,725,696 income
from other sources, principally interest and dividends on stocks and bonds
owned and for rent of track, and a net balance of $5,410,338 from outside
operations. The total of these two sums, $206,136,034, may be arbitrarily
applied first to offset the item of rent, $120,784,982, paid for leased lines and
track, and the balance in payment of interest and dividends in proportion to the
value of bonds and stocks owned as above, viz.: 36% and 64%, respectively.
This enables us to make the following distribution of the net operating income
of the railways reporting to this Bureau, as follows:

Net operating income, as above $724,483,163

Disposition of same:
Interest on funded debt $324,181,521
Less paid from "other income" 30,843,416 $293,338,105
Interest on current liabilities 22,546,779
Other deductions 70,174,473
Dividends preferred stock 50,183,739
Dividends common stock 176,607,550
$226,791,289
Less paid from "other income" 54,832,742 171,958,547

Dividends on other securities 769,222

Additions and betterments charged to
income 24,807,546
Appropriations to reserves 16,984,447
Miscellaneous 5,602,761
Deficits of weak lines 4,996,195
Surplus available for adjustments and
improvements 113,205,088 $724,483,163

This table shows the actual disposition made of the net income from operation
of the roads reporting to this Bureau, representing 97% of the railway business
of the United States, except that $120,784,982 of the income from other sources
has been eliminated from the account and applied to offset the rental paid by the
reporting roads.
It will be observed that the gross dividends declared were only $226,791,289,
which is 3.64% on the par value of the stock of the 368 reporting companies.

Misrepresentations as to Dividends.
The discrepancy between this condition and the official statement as to
dividends declared in 1908 calls for an analysis of the latter. This reads, "The
amount of dividends declared during the year (1908) was $386,879,362, being
equivalent to 7.99% on dividend-paying stock. For the year ending June 30,
1907, the amount of dividends declared was $308,088,027."
Two income accounts—one of operating roads and the other of leased roads—
for the year ending June 30, 1908, give a clew as to how the Official Statistician
more than doubles the dividends actually paid out of transportation revenues.
The gross total is made up of these four items:

Operating roads:
Dividends declared from current income $271,328,453
Dividends declared out of surplus 57,733,808
Leased roads:
Dividends declared from current income 33,843,577
Dividends declared out of surplus 27,550,596
Total $390,456,434

As these income accounts show that the operating companies received

$280,427,460 "other income" from outside operations and sources other than
transportation, and the leased roads received $111,153,013 "income from lease
of road," the source of the major part of this fictitious dividend is revealed. The
$280,427,460 from other sources would pay the entire income of the leased
roads and leave nearly $170,000,000 to extinguish so much of the dividends
declared by the operating roads.
Modified as to details, this is what actually occurs every year. In the year 1908
the total amount paid out of transportation revenues on account of capital of the
97% of the railways of the United States reporting to this Bureau was
represented in the sums:
 Net interest on funded debt $282,354,000
Interest on current liabilities 31,835,708
Rent paid for lease of roads 113,529,261
Net dividends 104,074,006
Total $531,792,975

This total was equivalent to 4.15% on the net capitalization of the roads
represented. The rental paid the lessor roads constituted the fund from which
those roads paid their interest and dividends. Further remark on the misleading
and harmful statement of the Official Statistician as to dividends declared in 1908
is unnecessary.
 V

COST OF CONSTRUCTION
Incomplete as are the figures of the cost of the railways of the United States,
and exclusive as they are of the millions put back into the properties out of
income for additions, betterments and reconstruction in the process of operation,
yet the statistics of the cost of construction and equipment afford a complete
answer to all charges that American railways are over-capitalized.
Upon the question of the cost of road and equipment in 1909, the returns of
the 368 roads reporting to this Bureau furnish the following data:

Summary of Cost of Road and Equipment Covering 221,132 Miles of

Operated Line for 1909.

Item Amount
Cost of road (182,046 miles owned) $6,603,504,463
Cost of equipment 1,122,409,813
Undistributed cost of road and equipment 3,080,064,960
Cost of 39,086 miles leased lines rental capitalized 2,415,699,876
Total $13,220,678,876

Adding to this $290,750,000 to represent the 11,870 miles of road not

reporting to this Bureau at $25,000 per mile, we obtain

$13,417,438,876
as the cost of road and equipment of the 233,002 miles of line employed in the
transportation industry of the United States in 1909, or

$58,031 per mile of line.

This is an underestimate by reason of the failure of a few lines to furnish even
approximate figures on the accumulated cost of their properties. Averaging the
cost of locomotives at $15,000, of passenger cars at $6,000, of freight cars at
$800, and of company's cars at $500 apiece—their present cost rates much
higher—the equipment of American railways represents an investment of over
$3,000,000,000, and its bare maintenance alone involves an expenditure of
nearly $400,000,000 annually.
 Physical Valuation of the Railways.
It is worthy of passing note that just as the railway companies have shown
their indifference to a physical valuation of their property, the clamor of
regulators and agitators in its favor has subsided. The proposal lost its
attractiveness to them the moment they became convinced that such an
investigation would put a valuation on the roads so high as to take not only the
wind out of their sails but the last drop of water out of their mouths. To-day the
only insistent demand for this futile undertaking comes from quarters interested
in the distribution of the appropriation of several millions it would cost.
Credit for the reversal in the popular and political attitude on this subject is
largely due to the valuations attempted by the states of Minnesota, Washington
and Wisconsin. The results in these states may be briefly summarized as follows:

Capitalization Valuation by
Miles of Line
per Mile State, per Mile
Minnesota, 1907 7,596 $44,206 $54,201
Washington, 1908:
Great Northern 806 44,078 73,900
Northern Pacific 942 70,278 106,500
Oregon R. R. & Navigation Co 501 43,012 38,900
Wisconsin, 1906 7,135 33,424 34,630

Even Senator Albert B. Cummins of Iowa has seen such a bright light on this
subject that in his speech before the Traffic Club of Chicago last February he said
that he would not be willing to make a present valuation of railroad property a
basis for determining rates, "for the reason that it was more than probable that
the present capitalization of between fifteen and sixteen billions would be
increased to twenty billions."
In the Bureau's Statistics for 1908 it was said:
"If the valuations in Minnesota and Washington, made by none too friendly
commissions, are any criterions of what a national valuation made under
presumably unbiased federal authority would be, the present cost to reproduce
the railways of the United States would be nearer $20,000,000,000 than any sum
within the anticipations of those agitating for such valuation."
Capitalization of Foreign Railways.
 With both sides of the balance sheet testifying to a capital investment in
American railways of under $60,000, and official valuation abandoned because it
would demonstrate that they could not be reproduced for less than $80,000 per
mile, the reader is asked to compare the American figures with those of the
capitalization, or cost of construction, of the principal foreign countries set forth
below. These have been compiled from the latest available official returns.

Summary of Railway Capitalization of the Principal Foreign Railways from

Latest Data.

Capital or Cost of
Year Country Miles of Line
Construction Per Mile

Europe:
1908 United Kingdom 23,205 $6,382,296,742 $275,040
1908 Germany 35,558 3,903,848,400 109,788
Russia in Europe (exclusive
1907 80,985
of Finland) 32,900 (a)3,170,876,360
1907 France (b)24,730 3,447,366,000 139,390
1907 Austria 13,427 1,515,576,885 112,879
1907 Hungary 11,769 741,586,391 63,010
1907-
124,730
08 Italy (State roads only) 8,699 1,086,000,000
1905 Spain (13 roads) 6,840 583,632,000 85,327
1906 Sweden 7,938 257,408,450 32,427
1907 Belgium (State only) 2,537 430,800,000 169,806
1907 Switzerland 2,740 298,709,210 109,000
Other Countries:
1909 Canada 24,104 1,608,990,656 66,752
1908 British India 30,576 1,364,669,375 44,632
1907 Argentine Republic 13,690 820,433,796 59,930
1908 Japan 4,444 190,173,728 42,800
1909 United States of America 233,002 13,508,711,173 57,976

(a) Russian capitalization, including railways in Asia, covers a total of 39,277 miles,
from which the capital per mile is computed.
(b) This is exclusive of 4,259 miles of local interest.

The most striking feature in this table is the steady advance it shows in the
capital cost of German railways. In ten years this has increased from 251,597
marks per kilometer in 1898 to 283,608 in 1908, i. e. 31,731 marks per kilometer
or $12,282 per mile. This means an increase of $991,687,440 in capital cost for
an increase of only 5,525 miles of line.
 VI

OWNERSHIP OF AMERICAN RAILWAYS

Returns to this Bureau place the number of stockholders of record at the date
of the last election of directors prior to June 30, 1909, of the 368 roads reporting
at 320,696. As only 182,046 of the 221,132 miles operated by these roads was
covered by the capital stock, this would show 1¾ stockholders for each mile of
road and would indicate that there are at least 415,000 stockholders in all the
railways of the United States. Owing to the incompleteness of the returns on this
subject and the fact that large blocks of stock are held in the names of
associations and trustees, it is safe to estimate that the actual ownership of
railway stock is distributed among at least 440,000 persons.
In 1905 the Commission reported the number of stockholders of record prior
to June 30, 1904, as 327,851, but has given no later figures. It may be of
interest to compare these figures with the partial reports to this Bureau since
then.

Number Number of
Year
Reporting Stockholders

1904 1,182 roads 327,851

1906 284 " 226,986
1907 317 " 240,554
1908 315 " 315,727
1909 340 " 320,696

If the ownership of railway bonds, which is even more widely distributed than
that of stocks, could be traced, it would be found that over a million investors are
interested in the financial welfare of the railways. This would give to each an
interest of $13,000, from which the average income is not over $520 a year.
The attempt of the Commission in 1908 to secure evidence that the control of
the railways was concentrated in a few hands by calling for a statement of the
"ten largest holders of voting securities" of the reporting companies having
established that nowhere did they own a majority or an approach to a majority
of the controlling stock, inquiry along that line was dropped in 1909.
 In railways, as in any republic, the latent power is widely distributed among
the many, while the administrative responsibility is necessarily entrusted to the
few.
 VII

PUBLIC SERVICE OF THE RAILWAYS

It is the reproach of our system of government statistics of railways that their
first concern is financial results, which the government takes no thought to
improve, and the harrowing roll of accidents, and not the adequacy of the
service and the steady development of the means of transportation. Every
month, almost every week, the public is informed of the volume of traffic, and
every quarter the record of casualties is told in sensational head lines. It is left
for belated annual reports to record the public service of this great industry upon
whose progressive efficiency every other industry in the United States depends.
It is not upon what the railways earn, but upon what they DO that the whole
industrial fabric of the republic rests. It is not upon the dividends they pay but
upon the traffic they carry, the net income withheld from dividends and put into
improvements, that their success as carriers depends.

The Passenger Traffic.

In considering the public service of the railways it is customary to give first
attention to the passenger traffic. This is not because it is the most important
branch of the service but because passengers are numbered by millions, where
thousands suffice in the enumeration of the shippers, who frequently mistake
themselves for the entire American people.
In twenty years between June 1, 1889, and June 1, 1909, the population of
the United States increased from 61,289,000 to 88,806,000, or nearly 45%. In
the meantime the passenger cars provided by the railways increased from
24,586 to 46,026, or over 87%. But this does not measure the liberal provision
made by the railway for the travelling public, which is more fully and accurately
expressed by the amazing growth of the number of passengers carried one mile
from 11,553,820,445 in 1889 to approximately 29,452,000,000 in 1909, or nearly
155%.
Here is shown an increase of cars not far short of double the increase in
population and an increase in passengers carried proportionately greater than
the numerical increase in cars.
In the meantime the average receipts of the traffic have declined from 2.165
cents per passenger mile in 1889 to 1.916 in 1909—a decline of over 11%,
although every item involved in the service, locomotives, cars, track, stations,
labor, etc., cost more. The passenger service, except as precursor to the freight
service, and in certain densely populated sections, was unremunerative in 1889
and is more so now. It is maintained at the expense of the freight service by
what the Railroad Commission of Wisconsin has characterized as "a species of
piracy practiced upon the shippers of freight."
The salient features of the passenger service reported to this Bureau for the
year 1909, as compared with the final official returns for the preceding year, are
shown in the following statement:
 Bureau Figures Official Figures
Item
1909 1908

Miles of line represented 221,132 230,494

Passengers carried. 854,255,337 890,009,574
Passengers carried 1 mile 28,788,855,000 29,082,836,944
Passenger revenue $551,634,278 $566,832,746
Mileage of passenger trains 491,903,107 505,945,582
Average number of passengers in train 58 54
Average cars to a train 5.3 —
Passenger car miles 2,594,508,987 2,705,659,994
Average passenger journey (miles) 33.71 32.66
Average receipts per passenger mile (cents) 1.916 1.937

According to the monthly reports to the Interstate Commerce Commission

covering an average of 233,002 miles of line, the passenger revenues in
1909 were $564,302,580, or $1,943,077 less than the above revenues for
only 228,164 miles of line in 1908.
The average receipts per passenger mile in 1909 are the lowest ever
reported for American railways.
Taken in connection with the official returns covering the period since
1900, the above figures afford evidence of the confiscatory effect of the 2-
cent passenger laws on railway revenues, as appears from the following
statement:

Summary of Passenger Mileage, Revenue and Receipts per Passenger

Mile, 1900 to 1909.
Increase Receipts
Over per
Passengers Carried Passenger
Year Preceding Passenger
One Mile Revenue
Year (Per Mile
Cent)

1900 16,038,076,200 — $323,715,639 2.003

1901 17,353,588,444 8.2 351,356,265 2.013
1902 19,689,937,620 13.4 392,963,248 1.986
 1903 20,915,763,881 6.2 421,704,592 2.006
1904 21,923,213,536 4.8 444,326,991 2.006
1905 23,800,149,436 8.6 472,694,732 1.962
1906 25,167,240,831 5.7 510,032,583 2.003
1907 27,718,554,030 10.1 564,606,343 2.014
1908 29,082,836,944 4.9 566,245,657 1.937
1909 29,452,000,000 1.3 564,302,580 1.916
Increase, per cent 83.7 — 74.6 —

Here it is shown that the passenger service rendered has increased 12%
more than the passenger revenues. But more significant than this is the
column of yearly increases in service by percentages. This utterly explodes
the theory that passenger travel is greatly stimulated by low fares—aside
from some positive incentive to increased travel, such as periodical
expositions, the Louisiana Purchase Exposition for instance, the effect of
which is clearly traceable in the increased service in 1905, which includes
the heavy travel during the months of heavy attendance, July 1 to
December 1, 1904.
The 2-cent passenger laws were passed so as to become generally
effective July 1, 1907, and their effect on passenger receipts during the
following year was such that these receipts were actually less in 1909 than
in 1907, although the service performed by the railways was over 6%
greater. Had the railways received the same rate in 1909 that they did in
1907 their revenue from passengers would have been nearly $29,000,000
more than it was.

Passenger Traffic 1909-1888.

In the next statement the salient facts in regard to the passenger traffic
since the Commission began collecting the data is passed under review.

Average
Passengers Mileage Receipts
Passengers Average Average Passenger
Carried Passenger per
Year Carried Passengers Journey Revenue
One Mile Trains Passenger
(Millions) in Train Miles (Millions)
(Millions) (Millions) Mile
(Cents)

1909 888 29,452 507 58 33 504 1.916

1908 890 29,082 500 59 33 566 1.937
1907 873 27,718 509 51 32 564 2.014
1906 797 25,167 479 49 31 510 2.003
1905 738 23,800 459 48 32 472 1.962
1904 715 21,923 440 46 31 444 2.006
1903 694 20,915 425 46 30 421 2.006
1902 649 19,689 405 45 30 392 1.986
1901 607 17,353 385 42 29 351 2.013
1900 576 16,038 363 41 28 323 2.003
1899 523 14,591 347 41 28 291 1.978
1898 501 13,379 334 39 27 267 1.973
1897 489 12,256 335 37 25 251 2.022
1896 511 13,049 332 39 26 266 2.019
1895 507 12,188 317 38 24 252 2.040
1894 540 14,289 326 44 26 285 1.986
1893 593 14,229 335 42 24 301 2.108
1892 560 13,362 317 42 24 286 2.126
1891 531 12,844 308 42 24 281 2.142
1890 492 11,847 285 41 24 260 2.167
1889 472 11,553 277 42 25 254 2.199
1888 412 10,101 252 40 24 237 2.349
Increase 115% 191% 101% 45% 38% 138%
1888 to 1907
Decrease 18.4

The several increases shown in the first, second, third and sixth columns
of the table reflect the general advancement in passenger traffic. That of
45% in the average passengers to a train marks the progress in density of
that traffic which may eventually place it on a profitable basis. In
Massachusetts, where this density yields an average of 79 passengers to a
train there is no demand for a two-cent rate statute, for the conditions have
made a rate of 1.64 cents profitable. In the group of states consisting of
Ohio, Indiana, Michigan, Illinois, Iowa, Wisconsin and Minnesota, where the
density of traffic yields only 46 passengers by train, a statutory two-cent
fare becomes confiscatory because it costs at least one dollar to operate a
passenger train one mile and 46 times two cents is only 92 cents. Moreover
the 46 passengers per train is only an average and there are as many trains
that average less as more. The average has to be raised above 50 to yield
any margin of profit on passenger traffic. If it were not for the density of
traffic in the New England and North Atlantic group of states the average for
the entire United States would be well below 46 passengers per train.
The steady increase in the distance traveled per passenger reflects the
effect of trolley competition in diverting the short haul passenger traffic.
The most noteworthy feature of the seventh column is the decline of
98/1000ths of a cent in the average receipts per passenger mile between
1907 and 1909, making a new low record after hovering around the two
cent mark for fourteen years. As noted above, this reduction in the average
cost the railways nearly $29,000,000 on the passenger traffic of 1909.
In this connection it is interesting to recall that between 1888 and 1893
the Official Statistician, then as now Professor Adams, made the following
computation of the average cost of carrying one passenger one mile for the
whole United States:

1888 1889 1890 1891 1892 1893

Average cost of carrying a passenger
one mile, cents 2.042 1.993 1.917 1.910 1.939 1.955

It will be observed that the average receipts per passenger mile in 1909
are below the computed cost in every one of the years above named,
except 1891. When the advance in the cost of everything necessary to the
service—track, labor, equipment, conveniences, speed, terminal facilities—is
considered, the practical coincidence of average cost and receipts leaves no
margin for legitimate profits.

Receipts from Mail and Express.

Closely associated with the passenger traffic of the railways are the mail
and express services. Although principally carried by passenger trains, each
has a special service of its own by mail and express trains. But all are
included under the passenger service. The receipts from these two branches
of the service during the last decade are shown in the following statement:

Summary of Receipts from Mail and Express, 1899 to 1908.

Mail Express
 Percentage Percentage
Year Revenues of Revenues of
Earnings Earnings

1899 $35,999,011 2.74 $26,756,054 2.04

1900 37,752,474 2.54 28,416,150 1.91
1901 38,453,602 2.42 31,121,613 1.96
1902 39,963,248 2.31 34,253,459 2.07
1903 41,709,396 2.19 38,331,964 1.98
1904 44,499,732 2.25 41,875,636 2.12
1905 45,426,125 2.18 45,149,155 2.17
1906 47,371,453 2.04 51,010,930 2.19
1907 50,378,964 1.94 57,332,931 2.21
1908 48,517,563 2.03 58,602,091 2.45
1909 50,935,000 2.08 63,669,000 2.60
Increase, per cent 41.5 — 138.0 —

Aside from the striking contrast in the percentages of increase of

revenues from these two sources, the most significant feature of this table
is the reversal it shows in their respective importance from the railway
revenue point of view. Prior to 1905, carrying the mails brought larger, if not
more profitable, returns to the railways. Since then the returns from express
have increased so much more rapidly that they are now nearly 23% more
than those from mails.
If proof were needed of the absolute falsity of the charge that the
railways are receiving an exorbitant rate for carrying mail, the above table of
their receipts from the service in connection with the following statement of
mail handled and revenues in view of the finding of the Joint Commission of
Congress in 1899 would furnish it. After a thorough investigation of the
subject lasting from August, 1898, to July, 1900, the Commission came to
the following conclusion:
"Upon a careful consideration of all the evidence and the statements and
arguments submitted, and in view of all the services rendered by the
railroads, we are of the opinion that the prices now paid to the railroad
companies for the transportation of the mails are not excessive, and
recommend that no reduction thereof be made at this time."
The increase in the railroad service since this report was made is shown in
the following statement of miles of mail transportation by railroads, the
postal revenues and the number of clerks in the railway mail service since
1899:

Annual
Transportation Number of
Postal
of Mail by Railway
Revenues
Railroads Mail Clerks
(Miles)

1899 287,591,269 $95,021,384 8,388

1900 297,256,303 102,354,579 8,695
1901 302,613,325 111,631,193 9,105
1902 312,521,478 121,848,047 9,627
1903 333,491,684 134,224,443 10,418
1904 353,038,397 143,582,624 11,621
1905 362,645,731 152,826,585 12,474
1906 371,661,071 167,932,783 13,598
1907 387,557,165 183,585,006 14,357
1908 407,799,039 191,478,663 15,295
1909 — 203,562,383 15,866
Increase in 10 years, per cent 50.5 124.7 89.1

Compared with the increase of only 41.5% in the revenues from mail
received by the railways during the same period, each one of the above
percentages testifies to a positive reduction in the rate received by the
railways for the service. And if the increase in weight of mail carried in 1909
were known, the contrast between the service and the pay would be more
striking. In 1899 the total weight of all mail was reported as 635,180,362
pounds. In 1907 the estimates made from the special weighing placed the
weight of mail carried that year at 1,290,358,284 pounds, or an increase of
nearly 105% in eight years. By reference to the above table it will be seen
that the railway revenues from mail between 1899 and 1907 increased only
40%. The contrast is illuminating. In its light the charge that the railways
are in any way responsible for the postal deficit is grotesque.

Freight Traffic
According to the monthly returns to the Interstate Commerce
Commission, the proportion of revenues from freight of the railways of the
United States to total earnings from operation, for the years 1908 and 1909,
receded to the unusually low figures of 68.51% and 68.88% respectively.
The official summary for 1908, based on the annual returns, shows a
proportion of 69.17% for that year, which probably is nearer the mark.
The annual reports to this Bureau for 1909 yield a proportion of 69.18%
for last year.
Accepting this proportion taken from the annual returns as being based
on the same character of reports as those from which former ratios were
derived, the preponderance of freight traffic is shown in bold relief in the
following statement of the ratio of its revenues to total earnings from
operation, 1899 to 1909:

Proportion of Proportion of
Freight Freight
Year Year
Revenues to Revenues to
Total Earnings Total Earnings

1899 69.55% 1905 69.67%

1900 70.56% 1906 70.54%
1901 70.41% 1907 70.44%
1902 69.93% 1908 69.17%
1903 70.39% 1909 69.18%
1904 69.82%

The average proportion for the nine years preceding 1908 is seen to be
slightly above 70%, and the fact that it was almost one point below 70% in
1908 and 1909 indicates that it was the freight traffic that bore the brunt of
the business depression which curtailed railway revenues during those
years.
In no other of the leading countries of the world does the freight traffic
assume the overwhelming relative proportion that it does in the United
States. In the United Kingdom it amounts to 50.35%; in France to 53.64%;
and in Germany, including express and mail, to 65%. If these were classed
with freight in the United States, it would raise the proportion for that traffic
here to over 74%.
Welcome to our website – the ideal destination for book lovers and
knowledge seekers. With a mission to inspire endlessly, we offer a
vast collection of books, ranging from classic literary works to
specialized publications, self-development books, and children's
literature. Each book is a new journey of discovery, expanding
knowledge and enriching the soul of the reade

Our website is not just a platform for buying books, but a bridge
connecting readers to the timeless values of culture and wisdom. With
an elegant, user-friendly interface and an intelligent search system,
we are committed to providing a quick and convenient shopping
experience. Additionally, our special promotions and home delivery
services ensure that you save time and fully enjoy the joy of reading.

Let us accompany you on the journey of exploring knowledge and

personal growth!

ebookultra.com