A STUDY OF INDUSTRIAL
FLUCTUATION
A
STUDY OF INDUSTRIAL
FLUCTUATION
An Enquiry into the Character and Causes of
the so-called Cyclical Movements of Trade
? BY
DENNIS HOLME ROBERTSON, M.A.,
Fellow of Trinity College, Cambridge
pet.
Heraclitus.
Urge and urge and urge,
Always the procreant urge of the world.
Walt Whitman,
LONDON
P. S. KING & SON, LTD
ORCHARD HOUSE
WESTMINSTER
1915
vfc
In plain memcriam
J. R.
334291
PREFACE.
§ i. THIS inquiry into the character and causes
of the so-called cyclical fluctuations of trade is
based mainly on a study of the course of events in
the leading industrial countries, especially the United
Kingdom, from about 1870 till the eve of the great
war. In August, 1914, it was submitted to the
electors to Fellowships at Trinity College, Cambridge,
and was successful. An earlier draft of the work
obtained the Cobden Club Prize at Cambridge in
1913.*
In the ordinary course of events the essay would
have undergone considerable overhauling before
publication. A certain amount of technical appar-
atus and of detailed statistical material, more suitable
in a dissertation intended primarily for the eyes of
scholars and to give evidence of research than in
a work designed to impress certain clear conclusions
on a more general public, might perhaps have
disappeared. But the pressure of other duties
has prevented me from undertaking the task ; and
it became clear that unless publication was to be
indefinitely delayed, the work must be published
substantially as it stood.
§ 2. It may seem that at the present time there
is little room for a book dealing in considerable
detail with an economic world which has largely
1 Certain sections of the essay appeared in a condensed form
in the Journal of the Royal Statistical Society, March, 1914.
vii
iii PREFACE.
passed away, and with problems which may appear
at the moment of somewhat remote interest. But
1 think to have refrained from publication on this
ground would have been a mistake. In the first
place, the problems here discussed, however much
their form may have altered, are essentially those
which are already causing anxiety to thoughtful
minds and are likely to become of overwhelming
urgency when finally " peace breaks out." No
attempt indeed has been made to bring the text up
to date or to consider the light thrown on the prin-
ciples there enunciated by the new phenomena of a
world at war. It is, and must now remain, a study
of pre-war conditions. None the less, it is hoped
that the facts and conclusions here presented may
be found of some use by those leaders of thought and
action whose function it will be when the time comes
to guide trade and industry back into normal
channels, and to minimise the disastrous results
of our present colossal inroads on our accumulated
wealth.
Secondly, paradoxical as it may seem, public
opinion is perhaps likely to be in some ways more
receptive now than ever before to a searching
analysis of industrial problems, and less suspicious
of drastic change. . Necessity has destroyed many
shibboleths and torn down many veils. One of
the most formidable obstacles to currency reform
—the alleged impossibility of persuading the well-
to-do Briton to live without clinking golden sove-
reigns in his pockets — vanished in a week-end. The
sacred machine of high finance has been shown to
be at once infinitely vulnerable, and far more amen-
able than its hierophants supposed to conscious
manipulation and control. When the safety of the
nation and no longer merely the welfare of millions
of its citizens is at stake, that accurate knowledge
of the real state of our resources for which this
PREFACE. ix
essay pleads has become a plain duty instead of an
unthinkable impertinence. Even the deliberate
manipulation of stocks in the public interest is now,
hi the case of sugar, wheat and cotton a familiar
notion. Though the event demanded far other
measures, there was in the early days of the war a
readiness to apply on an unprecedented scale
the device of bringing a Government demand for
structural work to bear upon a slack labour market.
Above all, the co-existence of brisk trade and em-
ployment with a war expenditure of £3,000,000 a
day has compelled clear thinking on the real nature
of saving and investment in the most unlikely
quarters. In a hundred ways the shock of war is
awakening men to a sense of the economic realities
in a manner which, unless the nation and civilization
perish in the interval, may form the prelude to a
less thoughtless and anarchic industrial age.
That, indeed, is yet uncertain. It is not yet
clear that the net result of the emergency extension
of Government action will be to increase confidence
in such methods in any country. Still less is it
certain that human imagination will not fail once
more to divert into the abiding struggle of man
with his environment those streams of chivalry
and resourceful effort unloosed, even in purely
economic fields, by the stress of international war.
It may be that the old Natura nihil facit per saltum
will be justified by our relapse into the Heracleitan
flux. But it is at least possible that in industrial
as in other matters we are in the presence of one of
those definite mutations of the social life which it
is within the collective power of man to fix and
foster in accordance with his highest hopes.
§ 3. I have had the great advantage of assistance
and criticism from Professor Pigou and Mr. J. M.
Keynes of King's College, Cambridge ; to the latter
in particular I owe more than it would be possible
x PREFACE.
to acknowledge. I am also indebted to Mr. W. T.
Layton of Cambridge ; and to various business men
and others for the provision of documents and for
information on specific points.
I have made such unremitting use of the Econo-
mist, especially the Annual Histories, that I have
not thought it worth while to give the reference
in every case.
To my friend Mr. H. D. Henderson of the Board
of Trade I am indebted f of great kindness in helping
to see the volume through the press ; and to Miss
Pate of the Cambridge University Typewriting
Offices for patience and skill in producing order
out of the chaos of my MS.
Finally to my mother and sisters, and to my friend
Mr. I. M. Hedley, I owe much thanks for constant
companionship and encouragement in the pursuit
of a somewhat laborious task.
November, 1915.
D. H. R.
CONTENTS.
PRELIMINARY CHAPTER. PAGE
§ i. Introductory ..... i
§ 2. Definitions. ...... 2
§ 3. Justification and Method of Discussion . 7
PART I. FLUCTUATIONS OF INDIVIDUAL TRADES.
(a) PHENOMENA OF SUPPLY.
CHAP. PAGE
I. TEMPTATIONS TO OVER-INVESTMENT.
§ i. The Period of Gestation .... 13
§ 2. Accessibility to Investment . 25
II. AGGRAVATIONS OF DEPRESSION.
§ i. Imperfect Divisibility and Intractability of
the Instrument ..... 31
§ 2. Longevity of the Instrument 36
III. FLUCTUATIONS IN COST.
§ i. Costs and Consumptive Industry . . 46
§ 2. Costs and Constructional Industry . . 53
§ 3. Invention ....... 66
(b) PHENOMENA OF DEMAND.
IV. MISCELLANEOUS CHANGES IN DEMAND.
§ i. Fashions, Wars, and Tariffs ... 69
V. CROP VOLUMES, TRANSPORT, AND CONSTRUCTION.
§ i. Direct Influence of Crop Volumes . . 75
§ 2. Psychological Influence of Crop Volumes . 85
xi
xii CONTENTS
CHAP. PAGE
VI. CROP VALUES AND CONSTRUCTION.
§ i. Normal Influence of Crop Values . . 89
§ 2. Psychological Influence of Crop Values . 92
§ 3. Detailed Illustration from United States and
Great Britain ...... 94
VII. CROP VALUES AND CONSUMPTION.
§ i. Crop-producers and Consumptive Industry . 104
§ 2. Crop-consumers and Consumptive Industry —
India 106
§ 3. Crop-consumers and Consumptive Industry —
Great Britain and United States . .no
PART II. FLUCTUATIONS OF GENERAL TRADE.
I. REVIVAL.
§ i. False Methods of Approach . . . 121
§ 2. The Growth of Productivity . . . 125
§ 3. Harvests — The Elasticity of Demand . . 129
§ 4. Harvests — The Law of Compensation . . 138
§ 5. The Increased Attractiveness of Investment. 156
II. CRISIS AND DEPRESSION.
§ i. Agricultural Shortage. . . . .165
§ 2. Constructional Relapse — Depletion of Stocks 170
§ 3. Constructional Relapse — Essential Causes . 180
§ 4. The Survival of Consumption . . .187
§ 5. General Depression — The Law of Markets . 198
III. THE WAGE AND MONEY SYSTEMS.
§ i. The Wage System and the Volume of Pro-
duction ....... 206
§ 2. Money in the Boom ..... 211
§ 3. Money in the Crisis and Depression . . 217
§ 4. Gold — Medicine, Poison, and Intoxicant . 228
§ 5. The Theory of Underconsumption . . 235
IV. CONCLUSION.
§ i. Recapitulation ... . 239
§ 2. Remedies — The Boom .... 241
§ 3. Remedies — The Depression .... 249
Index 283
LIST OF TABLES.
TABLE PAGE
I Petroleum Output 255
IA Production of Pennsylvania, etc. . . . 256
II English Railways . . . . . .256
III British Tonnage 257
IV Proportion of vessels, removed for foreign transfer
in each year, built before certain years . . 258
V Brazilian Coffee Crop and Jute Imports . . 258
VI Building Plans authorized by Urban Districts . 259
VII Analysis of results of Table VI . . . . 260
VIII Employment in British Isles .... 261
IX New York State. Unemployed Percentages . 266
LIST OF CHARTS.
CHART PAGE
I U.K. Iron and Coal Trades .... 267
II U.S.A. Construction 268
III Steel Trust Contracts 269
IV Wheat . 270
V Shipping and Shipbuilding . . . .271
VI U.S.A. Crops . . - 272
VII Argentina ........ 274
VIII India 275
IX Cotton . . 276
X U.K. Food Consumption ..... 278
XI U.K. Alcohol Consumption .... 280
XII U.S.A. Consumption ...... 281
XIII U.K. Building Trade . . 282
xin
A Study of Industrial Fluctuation
PRELIMINARY CHAPTER.
§ i. INTRODUCTORY.
THE causes of crises and depressions alleged before the
various committees of Congress in the eighties amounted
to some 180 in number, and included the issue of free rail-
way passes and the withholding of the franchise from
women. This list remained undefeated until M. Bergmann
in 1895 was able to publish an exhaustive discussion in the
German tongue of 230 separate opinions, arranged in eight
categories. Indeed the problem of industrial fluctuation
has exercised the minds of business men, economic writers
and practical reformers of all schools throughout the past
century : and within the last five years alone six weighty
works, varying in length from 280 to 742 pages, have been
published upon it in England, America and France. In
these circumstances it might seem a presumptuous and
superfluous undertaking to add to an already voluminous
literature : nevertheless I conceive that no apology is
needed.
For on the one hand, in spite of the obvious futility of
many of the minor explanations that have been given, this
does appear to be a case in which, in the deathless words
of the Dodo, everybody has won and all must have prizes,
in the sense that almost all the writers who have made any
serious contribution to the study of the matter appear to
have had a considerable measure of right on their side.
But on the other hand each writer has been apt to over-
2 STUDY OF INDUSTRIAL FLUCTUATION.
estimate the importance of his own work, and to neglect or
under- value that of others ; with the result that while the
vital importance of the subject, affecting as it does the
prosperity and even the livelihood of millions, l is generally
recognised, there is as yet no single comprehensive explana-
tion which may be said to hold the field. It is the author's
conviction that the most important work which remains
to be done lies in the direction of developing and synthesis-
ing the various and often conflicting opinions which have
been already expressed : it is his hope that an approach
to such a synthetic 2 exposition, with perhaps enough of
original suggestion and illustration to justify its claim to
rank as independent work, may be found in the following
pages.
§ 2. DEFINITIONS.
It would obviously be desirable as a preliminary measure
to arrive at a clear understanding of what is meant by the
phrase " industrial fluctuation." By industrial fluctuation
then I understand the alleged alternate occurrence of
periods of industrial expansion and of industrial depression.
But these phrases themselves call for further comment.
The distinction between temporary financial crises and
prolonged industrial depressions has been so frequently
emphasised 3 that it is not necessary to dwell on it here.
The more interesting question is, in what the depression
may be said to consist : nor is the answer by any means
easy to give.
1 Little will be said in these pages about the disastrous social
reactions of industrial fluctuation ; for discussion thereof cf . Bever-
idge, Unemployment, chap, iii., and Tugan Baranowsky, Les Crises
Industrielles en Angleterre (Fr. edit. 1913), Book III. chap, i., and
passim.
2 Such a synthesis is ascribed by Philippovitch to the modern
German theoretical writers on this subject, in whose explanation
' ' werden die Elementen der f riiher erwahnten Theorien beniitzt,
sie werden aber aus ihrer Vereinzelung herausgehoben " (Grundriss,
Vol. I. p. 425). But the result seems to be both confused and incom-
plete.
3 Cf . for instance Burton, Crises and Depressions, pp. 6 ff . and the
quotations there given.
PRELIMINARY CHAPTER. 3
First of all let us consider a single trade. I think it will
be conceded on reflection that the term depression may in
fact, without violence to the linguistic conscience of the
ordinary reader, be applied to any of the following pheno-
mena : — a decline in margins (i.e., in the difference between
the exchange value of each unit of product and the exchange
value of the component raw materials), a decline in profits,
per unit of product, a decline in the aggregate net receipts
of those engaged in the trade, a decline in aggregate profits
or finally a decline in the ratio of profits to invested capital.
Furthermore, while these various quantities and ratios are
commonly reckoned in terms of money, we cannot deny
the term depression to be at least equally applicable if they
should be reckoned in terms of goods in general : indeed
for the purposes of a definition which is to be used as the
basis of anything approaching a scientific study, we must
insist that they should be so reckoned.
Let us consider then which criterion of depression to
adopt. There seems no good reason to regard a mere
decline in margins — still less in one constituent element of
margins — as constituting depression, unless it is accom-
panied by a decline either in
gate net profits, or the net rate__of__prnfi ts. Our choice
then lies between these three. The fact that an inquiry of
this character is obviously concerned with the welfare of all
classes might suggest that the first alternative is alone
admissible : but while this consideration establishes the
advantage of the first over the second, it is not decisive
against the advantage which the third derives from its
insistence that the estimates which men make of their own
prosperity do not depend solely upon their present enjoy-
ment, but upon how far that enjoyment corresponds with
their anticipations. It seems hopeless to attempt a decision.
When we are discussing individual trades, the ambiguity
will not always arise : when it does, attention will be drawn
to it.
To pass on to the meaning of the phrase " general indus-
trial depression." The alternatives of a decline in margins
4 STUDY OF INDUSTRIAL FLUCTUATION.
or profits per unit are at once ruled out, since a general
decline in money-prices clearly does not of itself constitute
depression,1 and a general simultaneous decline of ex-
change values in terms of goods is obviously meaningless :
a fall in the exchange value of some goods implies a rise in
the exchange value of others. The criterion of aggregate
receipts must now be read aggregate volume of exchanges,
or rather, I think (since some producers are also consumers
and the inevitable Crusoe himself must be regarded as
liable to alternations of prosperity and depression), " aggre-
gate volume of consumption." If, however, we intend to
adopt this criterion, we must recognise frankly that the
volume of consumption is by no means an easy conception.
It presents much the same difficulty as the conception of a
general level of prices. If more of one thing X and less
of another thing Y is being consumed at time B than at
time A, how are we to tell whether, and if so, to what
extent, depression prevails at B as compared with A ?
It would seem that what we want to compare is the aggre-
gate amount of utility derived from the joint consumption
of X and Y on the two occasions : but the utility schedules 2
of X and Y can only be known to us in terms either of
money — and the sooner we get out of thinking in terms of
money in this matter the better — or of assorted commo-
dities : and the movements along the utility-schedules of
these commodities are unknown to us just as are the move-
ments along the utility-schedules of X and Y. The diffi-
1 Some years ago indeed rising and falling prices might have
been accepted as the final criterion of industrial expansion and
depression : but the seeds of scepticism are now too widely sown.
Falling prices may or may not be a symptom or a cause of depres-
sion, they are certainly not its essence. German writers however
still use the French word " hausse " to denote the expansion (e.g.,
Spiethoff in Schmoller's Jahrbuch, 1902, i. pp. 721 fL), though some
of the French themselves, such as M. Lescure, prefer the more non-
committal word " essor."
2 By a " utility " schedule I mean a statement of the successive
additions made to the " utility " or satisfaction derived from any
commodity by the addition of successive increments of that com-
modity to the possessor's stock,
PRELIMINARY CHAPTER. 5
culty is not capable of general solution, but in certain cases
a fairly confident answer can be given.
But there is a further difficulty. In making an estimate
of the national dividend for ordinary purposes we are
rightly warned l that when we have counted in the finished
product, we must not also count in all the intermediate
goods that went to its making. But it is only in a sta-
tionary state that the value of the intermediate goods con-
sumed in any year bears any precise relation to the value
of the finished products consumed in that year. In a
discussion of inter-annual fluctuation, however, the hypothe-
sis of a stationary state must obviously be discarded from
the outset. Experience indicates (i) that the volume of
instrumental goods consumed in any given year is relevant
to the volume of finished products consumed not in that but
over a number of succeeding years : (2) that the utility con-
veyed by the acquisition of these goods does not, in fact,
bear any precise relation to the utility enjoyed from the
consumption of the finished products produced by them
even in subsequent years. We must then include, in our
estimate of the consumption of the year, the utility of the
capital goods created during the year : 2 nor need we be
afraid for our purpose of including also the utility of the
finished products created by them during the year.3 Nor
shall we follow Dr. Marshall in subtracting the utility of
the commodities devoted to making good the depreciation
of existing instruments.
While, however, we include the instrumental goods that
1 Cf. Marshall, Principles, p. 79, and the discussion in the intro-
duction to the final report of the Census of Production, 1912.
2 In this following Marshall's conception of the national dividend :
cf. Pigou, Wealth and Welfare, pp. 16 if. A failure to apprehend
the manner in which capital goods are thus commonly reckoned
by the " orthodox " economic schools among things " consumed "
seems to be at the bottom of M. Tugan Baranowsky's rather arro-
gant and unfortunate excursions into pure theory, op. cit., Book II.
chap. i. §§2-4. Cf. infra, Book II. chap. ii.
3 This Marshall's plan would not do — so far at any rate as the
instruments render their services directly, e.g., a piano : it is not
clear whether it would when they render them indirectly, e.g., a
factory.
6 STUDY OF INDUSTRIAL FLUCTUATION.
contribute to, it does not seem that we should include the
materials which are actually transformed into, the finished
product. There remain certain goods, of which the most
obvious is coal, which seem to fall between the two cate-
gories. The best solution of the difficulty seems to be to
include all those goods which do not actually enter into the
physical composition of the finished product. This is
obviously a purely arbitrary decision, the sole justification
for which is that it seems to be least far removed from
popular notions of industrial " expansion " and " depres-
sion."
Our estimate will also differ from ordinary estimates of
the dividend in adding in the utility of consumable and of
incompletely worked-up instrumental goods consumed but
not produced (i.e., extracted from nature) and in subtracting
that of those produced but not consumed in the year in
question.
There remains still the alternative definition of depression
as a period in which the consumption of investors is smaller
than any anticipated. However paradoxical it may appear
to stigmatise as depression a time in which the consumption
of all classes is quite possibly increased, this definition has
the same advantage in respect of industry in general as it
has in respect of a particular trade. But there is an objec-
tion to it which, though perhaps somewhat obscure, is best
faced at the outset. A disillusionment of investors in any
single trade may arise either because the physical pro-
ductivity of the instruments employed in that trade is less
than was expected, or much more probably because the
ratio of exchange has moved against that trade more than
was expected. But a general disillusionment arising from
the latter cause is impossible : nor is there any reason,
especially in view of the nature of the most familiar symp-
toms of depression, to suppose that the former cause is more
important with regard to industry in general than it is with
regard to any single trade. If, therefore, general depression
is to be defined as a condition of general disillusionment on
the part of investors, this disillusionment must be quite
PRELIMINARY CHAPTER. 7
different in nature from that which is predicated of a single
trade. While excessive investment in a single trade results
in an undesirably large volume of its products entering into
the consumption of other trades, excessive investment in
all trades apparently results in an undesirably small volume
of their j oint products entering into their j oint consumption.
The difficulty thus revealed is too far-reaching to be
pursued further in this preliminary inquiry. The quest
for a definition which shall be at once theoretically water-
tight and in tolerable accord with popular usage must per-
force be abandoned. But the labour spent upon it has not
been wasted if it has only brought to light the inherent
complexity of the problem. It is scarcely remarkable
that no consistent and comprehensive explanation has yet
been given of a phenomenon the precise nature of which it
seems impossible to define.
§ 3. JUSTIFICATION AND METHOD OF
DISCUSSION.
One further preliminary question remains. Granted
that something which may fairly be called an alternation of
industrial prosperity and depression really exists, is it
anything to make so great a to-do about ? Are these
alternations likely to be a profitable subject of study ?
When we consider the anarchic nature of modern industry —
how wants are satisfied and activities organised without
the conscious guidance of any single directing power, the
astonishing fact is surely not that fluctuations should occur
but that all things on the whole should work together so
smoothly and steadily for good. Since the possible sources
of error and dislocation are so numerous, what is the use
of investigating fluctuation as though there were anything
strange or mysterious about it ? This point of view, which
lends itself readily to rhetorical argument, seems to spring
from a healthy sceptical reaction against the almost
superstitious atmosphere which has grown up round the
subject : but it appears to me unsound for the following
reasons.
8 STUDY OF INDUSTRIAL FLUCTUATION.
In the first place, granted that there is nothing to be
surprised at in the occurrence of fluctuation and that its
causes may be infinitely numerous, it does not follow that
it does not deserve study, both from a purely scientific
and from a practical point of view. Exactly the same thing
may be said of bodily disease.
Secondly, we have already had occasion to hint that
variations are not caused solely by fortuitous external
occurrences, but that each period of " expansion " contains
as it were the seeds of its own dissolution : 1 so that we may
perhaps be able to lay our finger on certain special features
of modern industry which tend to give it a fluctuating
character, and to which especial importance may be attached
as compared with the miscellaneous crowd of potential
sources of dislocation.
Thirdly, it is not so much the occurrence of fluctuations
that has occasioned comment as certain peculiar features
which are attributed to them. Thus it is urged that they
tend to occur (i) simultaneously in all trades (so that niceties
as to the precise meaning of the " volume of consumption "
are superfluous), (2) simultaneously in all the leading
industrial countries, (3) at approximately regular intervals.
We shall see reasons for refusing to attach too much import-
ance to this argument, and for holding that no less may be
learnt of the causes of fluctuation from the divergence than
from the similarity in this respect between different trades,
countries, and periods. Nevertheless, the alleged pheno-
mena seem to be sufficiently real to afford additional justifi-
cation for a study of the subject.
A somewhat similar objection is urged by those who
find the causes of fluctuations in what they call the " psy-
chology of the business man " and assume without further
argument that they are therefore incalculable and unfit
1 Thus Lescure distinguishes between organic and inorganic
theories of depression (Crises generates et periodiques de surproduc-
tion, p. 450), and Philippovitch somewhat similarly between crises
which " auf aiissere Ereignissen (Kriege, Misseraten) zuruckzufuhren
sind," and those which are preceded by a " Periode des Auf-
schwunges," involving its own opposite (Gmndriss, I. 423).
PRELIMINARY CHAPTER. 9
for systematic study. Additional weight is lent to this
view by the curious fact that it appears to be widely held
among business men themselves ; for a psychology, like a
liver, is an apparatus of whose existence one is only aware
when it is out of order. Nevertheless the economist can
hardly be content with such a fatalistic view of the mental
health of the business man as the latter is disposed to take
himself. Granted that his peculiar states of mind are
immediately responsible for industrial dislocation, it does
not follow that they are spontaneously generated ; it seems
only natural, in the absence of proof, to give him the benefit
of the doubt, and assume that they are in part at least
induced, however irrationally, by external objective facts.
Hence this objection also to the search for such facts falls
to the ground.
A word remains to be said as to the method of discussion
adopted in the following pages. The ideal method of
economic exposition is perhaps to elaborate an independent
constructive theory, treating the results and suggestions
of others as material for incidental rejection or as buttresses
to afford incidental support, and introducing facts rather
as illustrations than as the formal ground- work of generalisa-
tions. While I have tried in the main to follow this method,
it has not seemed to me at every stage entirely applicable.
First, where the problem attacked is so familiar, the ground,
as already indicated, is naturally thickly strewn with
partial explanations, some of which are so complex or so
famous as to demand consideration in the shape of rather
more lengthy and formal digressions. Secondly, this
problem more than any other in economics seems to me to
have suffered from the neglect of a sufficiently wide and
precise study of fact. Many theoretical writings on the
subject assume without argument a simplicity of the actual
phenomena under discussion which a closer study shows
does not in fact obtain.1 Others refer only to one set of
1 Thus Mr. Hawtrey in a work almost innocent of appeal to fact
(Good and Bad Trade, p. 215) asserts that " the phenomena of trade
io STUDY OF INDUSTRIAL FLUCTUATION.
facts, frequently of a monetary nature : * others again
neglect large tracts even of that department of fact which
they profess to discuss. 2 Though, therefore, a good deal of
consecutive industrial narrative which formed part of a
first draft of this work has been rearranged or entirely
removed, a few such passages have been allowed to remain,
in the hope that they may prove useful to some who, while
rejecting the author's own explanations, may be glad to
have more data than are at present available in a con-
venient form upon which to build explanations of their own.
As to actual procedure, a problem so complicated can
only be approached piecemeal. I propose, therefore, in the
first part to consider the causes of fluctuations in particular
industries or groups of industries. The first group of
chapters deals with those causes which must be sought
primarily in phenomena of supply, the second with those
which must be sought in fluctuations of demand. In the
second part I propose to inquire how far these partial
fluctuations are sufficient to explain the so-called general
fluctuations of trade, and what further steps remain to
be taken towards the construction of a comprehensive
theory. The first two chapters form a sustained account
of the essential features of the two limbs of the industrial
cycle : the third discusses the peculiar features attributable
to the existence of our wage and monetary systems : while
the fourth contains a brief recapitulation of the whole
argument, and some suggestions of practical policy. But
fluctuations are so well established that economists and statis-
ticians on the one hand and business men on the other are all likely
to agree as to whether the correspondence " between his theory
and the facts " is established " : but I cannot think that his confi-
dence is well-grounded.
1 Notably Clement Juglar in his great work, Des Crises Commer-
ciales et de leur retour periodique. It is curious that Tugan Bara-
nowsky, while in his second (theoretical) part he reasons largely
in terms of actual commodities, in his first (historical) part deals
almost exclusively with monetary phenomena.
8 I think Mr. Hull's work (Industrial Depressions, or Iron, the
Barometer of Trade) is particularly open to this objection. Cf.
Part I. chap. iii. § 2, infra.
PRELIMINARY CHAPTER. n
in a discussion so complex some overlapping and a good
deal of cross-reference will be unavoidable.
There remains one more point. In some of the more
abstract portions of this essay I shall make use, without
further explanation or apology, of the processes and termino-
logy in common use among the school of economic thought
associated in this country chiefly with the name of Dr.
Marshall. My reason is that after a study of many facts
and theories I am deliberately of opinion that one cause
of the obscurity which still surrounds this problem is that
in the attack upon it full and systematic use has never
hitherto been made of the weapons supplied by this particular
intellectual armoury. If I had thought it possible, consist-
ently with brevity, clearness, and (as it seems to me) accur-
acy of thought, to express my arguments and conclusions
in language more familiar to the ordinary reader and more
universally acceptable to the trained economist, I would
gladly have done so. But it has not seemed possible : and
I conceive that the place of this school in the world of
economic thought is sufficiently assured, and that its
processes are, or ought to be, familiar enough even to those
scientific thinkers who reject them, to justify their employ-
ment in an essay like the present.1
1 I have, however, tried to concentrate into certain sections the
more rigidly abstract portions of the argument. Such passages
will be found mainly in Part II. chap. i. § 3, chap. ii. § i and § 5.
PART I.
Fluctuations in Particular Trades.
A. PHENOMENA OF SUPPLY.
CHAPTER I.
TEMPTATIONS TO OVER-!N)VESTMENT.
§ i. THE PERIOD OF GESTATION.
WE are to begin with the causes determining the occur-
rence and intensity of fluctuation in individual industries,
and first with those which are primarily concerned with
conditions of production and supply.
It appears then that there are inherent in the modern
system of large-scale capitalistic industry certain closely
allied peculiarities which may produce in any trade, even
in the absence of any fluctuation in demand, an alternation,
the phases of which vary in length and intensity between
different trades, between prosperity and depression. The
first of these to be considered is what we may call the
period of gestation, in other words the length of time
necessary to construct and prepare for use the requisite
instruments of production.
Let us suppose that for any reason the exchange value
of the products of any trade has risen. There will then be
an inducement to increased investment in that trade. But
the new instruments ordered will not be immediately
ready for use : meanwhile the high level of price x will
1 This argument does not depend on the existence of a monetary
economy : I use the word price in this book as synonymous with
exchange value in terms of goods in general.
13
I4 STUDY OF INDUSTRIAL FLUCTUATION.
continue, and since each producer (in a competitive regime)
is ignorant of the preparations which are being made by his
rivals to meet the high level of prices, the total amount of
new instruments prepared will be so great that the price of
the product eventually falls below its old level. If the
demand for the commodity is inelastic,1 the total quantity
of goods received by those who produce it will then be less
than during the " boom " period : but even if it is elastic, .
it will be less than is sufficient to compensate for the in-
creased efforts and sacrifices made in investment, since
these latter have been greater than they would have been
had the future course of prices been accurately foreseen.2
The first drop in prices will occur as soon as the first batch
of new instruments is brought into use : the longer there-
fore this period of gestation, the longer will the period of
high prices continue, the greater will be the over-investment,
and the more severe the subsequent depression.
This argument is, in fact, only a development of that
doctrine of quasi-rent long familiar to students of Dr.
Marshall's work : but the credit of its first formal applica- *
tion to the study of fluctuations seems to belong to Pro-
fessor Af tali on of Lille.3 There is ample evidence that
1 Those to whom economic jargon is distasteful may be reminded
once for all that the money-demand for a thing is said to be elastic
when a given rise (or fall) of price produces a more than propor-
tionate decrease (or increase) in the amount demanded : and in-
elastic when the proportionate change in amount demanded is less
than the proportionate change in price.
2 That is, while depression may exist in both of the senses indi-
cated in § 2 of the last chapter, it only necessarily exists in the
second.
3 Cf. his articles in the Revue d'Economie Politique, 1909, and the
more elaborate discussion in his recent Les Crises Periodiques de
sur production, Vol. II. Book VI. chaps, iii.-vii. With the aid of
a good deal of statistical evidence he establishes the proposition
that the course of production of the finished instruments of pro-
duction does not coincide with that of the other generally accepted
indices of prosperity (including the production of raw instrumental
goods and the volume of orders for finished instruments), but lags
behind it in an interval in general of about two years. The figures
furnished by the Statistique de T Indus trie Minerale of the aggregate
horse-power of the motive machinery in use in each year in various
TEMPTATIONS TO OVER-INVESTMENT. 15
the length of the period of gestation in different trades is
an important factor in determining the intensity and
duration of the various phases of their several fluctuations.
For instance, a glance at the first of the appended charts
will suggest, and the more detailed study in Part II. chap. ii.
§ 4 will confirm -the conclusion that while the coal and pig-
iron trades pursue the same general course, the price of coal
tends to reach both its maxima and minima later than that
of pig-iron. While there are other causes for this, part of
the explanation seems to lie in the longer period of gestation
necessary in the coal trade. According to Mr. Hull * it
takes " practically a year " in America to build a new blast
furnace. From an English ironmaster I gather the impres-
sion that in this country some fifteen months would be
required. But a coal mine which is begun to be sunk now
industries in France are of particular interest. It appears to me,
however, that owing to his determination to allow little or no import-
ance to the influence of demand, M. Aftalion is under a constant
temptation to exaggerate the length of the period of gestation.
Hence his grouping of years into periods of large and small pro-
duction seems often to be somewhat arbitrary, and to yield a longer
lag than the figures themselves warrant. To give only one instance,
there seems no justification for concluding from the following figures
that the lag with locomotive engines is as long as three years.
Value of Increase in number
locomotives of locomotives
contracted for. in use.
/i 897 .. 12-8 in francs
1898 . . 28-
" High "period — ;; —
(1900 . . 45-2
(1901 .. 11-3
Low " period
High" period
" Low " period j 1902 . . io'8
1 1903 • . 9-2
(Op. cit., Vol. II. pp. 59 and 99).
Here, while the concluding years of M. Aftalion 's " high " periods
are separated by an interval of three years, the actual maxima
are separated only by one.
My own investigations were chiefly made before the appearance
of M. Aftalion 's detailed evidence, but in persuance of the sugges-
tion made in his earlier essay, and of a similar suggestion — com-
prising also the principle of chap. ii. § 2 — made to me in conversa-
tion by Mr. G. Udny Yule.
1 Industrial Depression, p. 207.
16 STUDY OF INDUSTRIAL FLUCTUATION.
will not be in working order for several years. Further, it
seems likely that the period of gestation in the coal trade
should have become longer during the past half century
with the necessity of sinking deeper shafts. It is likely
perhaps, moreover, to be longer for those mines sunk
towards the end than for those sunk towards the beginning
of the boom.
This a priori reasoning is confirmed by the figures. In
the boom of the '70*8 the first considerable rise in prices
took place in 1872, and began to make its full effect on
output felt in 1875. Between the break of prices in 1873
and the new influx of 1875 there was restriction of output.1
Again, since new enterprises were apparently begun at
least as late as 1874,2 production was not restrained till
1878 ; though, indeed, before the latter year some of the
worst mines had begun to go out of use.3
In the next cycle the first considerable rise in prices was
in 1882 ; consequently (assuming the period to have in-
creased to about five years) the new influx was delayed
till 1887. High prices continued till 1884, and large pro-
duction till 1889 ; but by 1890 (five years after the slump)
we might expect to find some restriction, had not a new
demand for coal arisen in that year, so that the end of this
cycle is, as it were, telescoped into the beginning of the
next.
In the next cycle high prices began in 1889 and ended
in 1891, so that (still with a five years' period) the new
influx began in 1894 and increased till 1896. Again, a new
1 The initial break in prices must, I think (in opposition to M.
Aftalion), be referred to a relapse in demand.
2 In December, 1873, we learn that " the proposed line from Mans-
field to Worksop will aid materially in opening out a vast tract of
highly mineralised ground, so that collieries will be opened out in
a straight course of from 30 to 40 miles." — Economist History of
1874. And "during 1875 no less than in new collieries were in
the course of sinking in the West Riding of Yorkshire alone "
Id. of 1876.
3 Number of collieries working : 1873, 3,627 ; 1875, 4,501 ; 1877,
4,231. Times, January i, 1878.
TEMPTATIONS TO OVER-INVESTMENT. 17
demand for coal arose in 1897, so that this cycle and the
next are telescoped together.1
In the next cycle the analogous quinquennia are 1897-
1902 and 1901-06, leaving room for restriction of output
in 1901. Again, as on every occasion except 1878-79, the
final curtailment is rendered superfluous by rising demand.
In the next cycle it seems more reasonable to refer the new
growth of output in 1909 to failure to check permanently
the vast additions to the source of supply made in the 1900
boom than to the boom of 1906-07, which supposes an
improbably short period of gestation. Indeed, I have
heard of enterprises undertaken in the 1900 boom which,
even in 1912, were not in working order ; and, of course,
in every case the new mines of one cycle form a permanent
addition to productive capacity, ready to take advantage
of the first stirrings of demand in the next — witness the
large expansion of output under an apparently inadequate
price-stimulus in 1 880-81.
Mr. D. A. Thomas indeed, writing in 1903,2 asserted
that very few new enterprises had been started during the
1900 boom, which is puzzling in view of the subsequent
course of prices and production.3 But it seems not im-
possible that the period of gestation is now becoming so
long as to prove a tolerably effective deterrent from over-
investment in times of boom(^ This proposition is, however,
1 The stagnation of production in 1898 was due primarily to
the South Wales strike : but it is perhaps worth also pointing out
that the investment in new mines five years before (i.e., in the year
of the great midland strike, 1893) is likely to have been abnormally
small, even for a year of depression.
2 Stat. Jour., 1903.
3 The reports of companies in 1904 indicate a much severer
depression in coal than in iron.
3 Coal 1 6 Iron and
Companies. Steel Companies.
1902-3 (year ending June 30) .. ^190,881 £1,681,000
1903-4 92,528 i,537,6n
Economist, 1904, p. 1,540.
4 " At the present time," I have been told by the representative
of a large mixed iron-works, " there is nowhere we could sink a new
colliery, except a few parts of Yorkshire and the South of England ;
and even so we should not see any coal for years." If this view is
c
i8 STUDY OF INDUSTRIAL FLUCTUATION.
also disputed by Mr. Thomas, who claims that it has become
shorter with improvement in the methods and appliances of
pit-sinking, and is in South Wales to-day not above three
years.1 While unwilling to dispute so high an authority,
I corinot help thinking that his conclusions, both as to
the extent of investment in the 1900 boom and as to the
present period of gestation, while no doubt entirely correct
for his own district, may be subject to modification for
other parts of the country.
In the shipping trade the influence of the period of gesta-
tion is very much complicated by fluctuations in demand ;
but its effect upon investment policy can be clearly traced.
The actual time occupied in building a tramp steamer of
the usual size, i.e., about 7,000 tons,2 under favourable
conditions — i.e., if there is no exceptional scarcity of labour
or materials — is under a year. Hence we might expect to
find the new tonnage beginning, within a year after the
beginning of a boom, to show its effect in reducing the
orders for new construction. We may explain on these
lines the check to new contracts in the fourth quarter of
1910 (see Chart V.). As a rule, however, the flood- tide of
demand seems to carry the volume of contracts over this
first dead point. And as the tide flows the period of gesta-
tion lengthens, for (i) the capacity of the yards is limited,
so that some months may elapse after the placing of an
order before the vessel is begun ; (2) there is a growing
correct, it follows that while up to a point a lengthening of the
period of production aggregates the intensity of fluctuation, a further
lengthening may actually mitigate it. The gradual working out
of the British coalfields is an excellent illustration of the truth
that a " secular " or " trans -cyclical " change may be important not
only in itself but in its influence on the course of " cyclical " fluctua-
tion.
1 In a valuable criticism of a paper by the present writer, Stat.
Jour., Jan. 1914, p. 173, Mr. Thomas ascribes the stagnation of
production in South Wales since 1907 entirely to labour troubles.
2 The average tonnage of the ships built in 1912 was 2,676, and
this figure was exceptionally high ; but the most frequent size was
much greater. Thus of the 108 ships under construction, December
31, 1912, 69 were between 6,000 and 10,000 tons (Lloyd's Register
Annual Summary],
TEMPTATIONS TO OVER-INVESTMENT. 19
pressure on the supplies of labour and materials, so that
delays occur also after construction has begun.1 Thus the
period is prolonged to as much as eighteen months. For
instance, the flood of new ships launched early in 1906
seems to have been composed largely of those ordered
late in 1904,2 and at the end of 1912 builders were stated
to be full of work well into I9I4.3
The first severe falling-off in new contracts is thus often
found some fifteen to eighteen months after — not the first
increase — but an exceptionally large increase in the volume
of such contracts. Compare the third quarter of 1889
with the first of 1888, the first of 1899 with the last of 1897,
the third of 1906 with the first of 1905, the second of 1913
with the first of 1912. Yet even so, the flood of demand is
not exhausted. Some eighteen months after the maximum
of tonnage under construction is passed, i.e., just when the
efflux of new ships from the yards is likely to have spent its
full force, there are of ten signs of a renewal of orders, and
a second or even a third maximum is attained, sometimes,
though not always (1901), subordinate. Compare the last
quarters of 1890 and 1891 with the third and second quar-
ters of 1889 and 1890 respectively, and the second and
fourth quarters of 1900 with the first and third of 1899.
Attention may be briefly drawn to evidences of the
influence of the period of gestation in a number of other
trades. In English railway building the following figures4
suggest a period of about two years.
A comparison (see chart) of the average profits of 100
cotton-spinning firms 5 and of the net increase in the number
1 The delay may be enhanced by labour disputes, e.g., the boiler-
makers' lock-out (1910) and the coal strike (1912). .
2 Cf. Economist, 1906, p. 1,207.
3 Messrs. Moss' Circular, January i, 1913. In view of the large
part played by tank-steamers in the recent record shipbuilding
boom, it is significant that " it takes much longer to build an oil-
tank steamer than an ordinary steamer " (Statist, August 10, 1912).
4 Taken from Whitaker's Almanack.
5 Estimate of Mr. John Kidger of Oldham, quoted in Economist
Histories,
20 STUDY OF INDUSTRIAL FLUCTUATION.
of spindles at work1 suggests that at the beginning of a
boom the period of gestation is about a year. Thus in 1889,
in spite of the excellent Indian and Continental inquiry,
Mileage
Mileage
Mileage
Mileage
Authorised.
Opened.
Authorised.
Opened.
1843 .
[nconsiderable
_
1847 .
1,663
909
1844 .
797
—
1848 .
300
I,l82
1845 .
2,883
—
1849 .
—
904
1846 .
4,79°
595
1850 .
—
59°
the new crop of spindles projected in the boom year 1888
was able to produce a perceptible check in the advance of
profits. Similarly the effects of the over-investment of 1890
are clearly seen in i89i.2 When the wave of rising demand
is prolonged, however, we should expect the period to
become somewhat longer. This is confirmed by the follow-
ing figures from a series of annual articles in the Economist
for January : —
[In thousands]
Spindles in
Mills under
New
Spindles
Spindles in
Mills under
New
Spindles
construction,
started
construction,
started
January i.
in Year.
January i.
in Year.
1899 .
70O
340
1905 .
2,985
1,460 *
1900 .
880
708
1906 .
4,350
3,029
1901 .
!,025
865
1907 .
4301
2,249
1902 .
I,I4O
951
1908 .
4,363
1,660
1903 .
720
p
1909 .
3,043
?
1904 .
?
?
* Last six months.
1 Messrs. Ellison's estimate, 3rd Fiscal Blue-book, Cd. 4954, 1909,
P- 157-
2 In the woollen industry also the period seems normally not to
exceed a year. Thus in 1872 the trade was already suffering from
TEMPTATIONS TO OVER-INVESTMENT. 21
Assuming a uniform rate of contracting and construction
the typical period of gestation for any year is given by
dividing column i by column 2.1 The result is 1-24 years
for 1900, 1-13 for 1901, 1-44 for 1906, 1-91 for 1907 and
2-63 for 1908. This result suggests among other things
that the period of gestation is longer, and the subsequent
slump therefore more acute, when the cotton boom syn-
chronises closely with than when it lags behind the
constructional boom.2
The period of gestation in the cotton trade is also reflected
in the course of our trade in cotton yarn to the Continent.8
The check in 1882 seems to have been due to the rapid
development, in response to the high demand of 1881, of
the home capacity of France.4 Similarly in Germany a
maximum of yarn imports from England was reached in
1889 i °f home consumption of raw cotton not till iSgi.5
Again a maximum of imports was reached in 1898, while the
the excess capacity prompted by the flood of continental orders
on the restoration of peace in 1871. The dropping profits of 1890
are probably to be referred chiefly to over-investment during the
high tide of 1889, and those of 1907 and 1911 partly to over-invest-
ment in the revivals of 1906 and 1910 (but cf. also pp. 72 and 109).
1 For instance, of the 700,000 spindles under construction on
January i, 1899, only 340,000 came into being during the year: the
time required to complete 700,000 is, therefore, Hg year.
2 Investment policy in the recent boom seems to have been con-
siderably more moderate than on previous occasions.
3 The survival of the importance of our yarn exports over that
of our manufactured exports to the Continent is probably due to
the fact that the English advantages of climate and transmitted
skill are of more value in spinning than in weaving. Cf . Chapman,
Cotton Industry and Trade, p. 72.
4 The number of spindles increased from 4,609,020 in 1881 to
4,927,624 in 1882. (Neumann-Spallart, Uebersichten, apud Lescure,
op. cit., p. 153. Messrs. Ellison, however, only record an increase
from 3,840,000 to 3,900,000.) The revival in 1883-84 was due
mainly to Germany.
No. of Spindles. Consumption.
5i888 5-10 million 3-6 m. cwts.
1889 5-25 „ 4-5 »
1890 . . . . . . . . 5-50 ,, 4-6
1891 5-75 „ 4-8
1892 6-03 „ 4-5
Cd. 4954 of 1909, pp. 157, 159.
22 STUDY OF INDUSTRIAL FLUCTUATION.
prosperity of the home trade lasted through the following
year.1
To turn to a very different industry. " The coffee
plant," says a writer in the Economist,2- " does not produce
marketable berries for five years " — in other words, five
years is here the normal period of gestation. It is therefore
significant that the first check to the coffee glut which
marked the turn of the century occurred in 1903, five years
after the first loud complaints of over-production.3
The big crop of 1906 induced the Brazilian Government
to undertake its " valorisation " scheme, and was followed
by the immense production of 1907. It may be pointed
out that the reduced crop of 1903 was originally estimated
at the still smaller figure of 8,000,000 to 9,000,000 bags ;
that this may well have led to extended plantings, and
that the favourable weather which produced the record
crop of 1907 may well have brought some plants to maturity
before the normal time. Nor, I think, is it accidental that
the maximum crop of 1910 occurred five years after the
minimum crop of 1905, the minimum of 1911 five years
after the large crop of 1906, which brought the anxiety of
growers to a head ; the new maximum of 1912, five years
after their hopes had again been raised by the high-sounding
pretensions of the Government ; and the relapse of 1913,
fiye years after the futility of the scheme had become
obvious. We may, perhaps, expect some further reduction
in the 1914 4 crop, to be followed by an increase in 1915,
five years after the high-priced year 1910.
1 According to Lescure, who bases his statements on Sybel,
Baumwollindustrie, Cd. 4954 shows a fall in consumption from 6-8
to 6-2 m. cwts., but these figures confessedly take no account of
spinners' stocks.
2 ign, Vol. I. p. 257.
3 1896-97 were years of prosperity to the producing interests ;
and though in the latter year there was an ominous increase of
production, yet " when supplies were at their heaviest prices were
relatively high, as both dealers here and on the Continent had run
their stocks so low that they were compelled to buy."
1 Though the preliminary estimate given in Table V. does not
support this view.
TEMPTATIONS TO OVER-INVESTMENT. 23
In other cases the period of gestation is very much less
uniform. For instance, the copper market seems to have
been weakened in 1908-10 by the output not only of the
mines, such as those of Servia, opened up in the recent
boom, but of such sources asvthe great Cerro de Pasco of
Peru, founded as far back as 1898 but only just beginning
to make its full effect felt ; as well as of hundreds of small
mines founded in the boom of the nineties and closed down
after the 1901 debacle, whose period of regestation was very
brief, and who clung tenaciously to their new lease of life.
In other cases the period of gestation can be altered at will
within fairly elastic limits. Thus in Ceylon a rubber-planter
tells me that while the trees do not begin to reach full
maturity for five years,1 it is not uncommon to tap them
after two years if the conditions of demand are specially
favourable.
It only remains to point out that the period of gestation
comprises the time necessary not merely to bring the new
products physically to the birth but also to bring them
to the place where they are required. If, therefore, the
rise in exchange value proceeds from an increase in demand
to which the instruments of transport are unable imme-
diately to respond, the period may be considerably length-
ened.
For instance, the copper boom in America and the steel
boom in Germany in 1907 were prolonged, and the conse-
quent over-investment aggravated, by the shortage of fuel
and railroad cars. Thus also while a new oil-field is usually
bearing within a year of the inception of its development,2
1 The colossal investment of 1909 is only now beginning to make
its full effect felt on prices.
2 Cf. Table I. Thus the boom of 1906 led to a great wave of
investment, especially in California, Roumania and Dutch Indies ;
and in the same year the German Government awoke to the possi-
bilities of oil fuel as a substitute for coal for naval purposes, and
set itself to develop the resources of Galicia. (Cf. Daily Chronicle,
June 19, 1913.) The result is seen in the production figures of
those countries for 1907. Again the expected rise in oil prices
in February, 1910 (for which see p. 169), led to a rapid increase in
the output of the Californian fields, the opening up of the great
24 STUDY OF INDUSTRIAL FLUCTUATION.
the price of petroleum remained at a famine level through-
out 1912 owing largely to the shortage of oil-carrying
tonnage.1
Finally it may conveniently be noted here that when
the transfer of a product through space involves appreciable
delay, not only a universal but a localised rise in its exchange
value may result in over-investment : in other words what
we may call the period of transference plays in such cases
the same leading role which has been ascribed in other cases
to the period of physical gestation. The point is of import-
ance because it disproves the contention sometimes heard
that the natural tendency 2 to interlocal compensation in
the sources of demand might be expected to prevent alto-
gether the occurrence of industrial fluctuation.
The best illustration of this proposition is afforded by
the production of the service of sea-transport. In spite of
the well-known willingness of tramps to undertake long
voyages in ballast in the hope of return cargoes, distance
appears still to oppose considerable obstacles to the move-
ment of tonnage from one trade to another. For instance,
in spite of the failure of the Russian demand in the end of
1891, freights remained high in the North Atlantic trade,
and were a sore temptation to incautious owners to increase
their tonnage. Similarly the concentration of steamers
in the Far East after the Chino- Japanese War in 1896, 3
and in Australia owing to the miners' strike in 1909,* raised
freights in other trades and increased the volume of new
Maikop field in Russia, a new flood of investment in Roumania
and in the West Indies, and an attack by the new Commonwealth
Oil Corporation and Australian Oil Company upon the Australian
supplies (cf. EC., 1910, ii. 970, 1,327), all of which made their effect
felt in the production figures of the same year.
1 For further discussion, see p. 168.
2 Cf. p. 148.
3 Aggravated by the Indian famine, which curtailed the supply
of homeward cargoes.
4 EC., 1910, i. 430. Cf. the maintenance of Australian freights
in 1913, owing to the reluctance of steamers to sail from the Argen-
tine in ballast for fear of being prevented from recoaling by labour
troubles at Delagoa Bay (Times Suppt., Jan. 16, 1914, p. 6).
TEMPTATIONS TO OVER-INVESTMENT. 25
contracts, though the total supply of tonnage was fully
adequate to the demand. This consideration helps also to
explain the special intensity of fluctuation in the building
trade. A migration of population does not, of course,
increase the aggregate demand for house-room, but it
does increase the demand for new building, for the period
of transference is in this case infinitely great. Thus the
great building boom in London in the 'go's was accom-
panied by a net increase in the number of inhabited houses
of only 17,000 (as compared with 58,000 in the previous
decade), though in Wandsworth and Lewisham alone there
was an increase of 19, 837. 1 Moreover, not only the instru-
ments of production but those who produce them seem to
be afflicted with immobility : with the decentralisation of
London " each new suburb tends to get a building trade
of its own as it rises to importance ; and its chief firms
compete with those of older standing elsewhere." 2
§ 2. ACCESSIBILITY TO INVESTMENT.
The temptations to over-investment fostered by the
length of the period of gestation are fostered also by another
characteristic of modern industry — its accessibility to
investable resources. The differences between trades in
this respect are an important cause of the differences in the
intensity of their fluctuations.
While most of the causes of fluctuation which we have
to consider are more strongly operative the larger the fixed
capital required, that before us now is more serious, other
things being equal, the smaller that fixed capital. Thus
in cotton-spinning not only is the amount of floating capital
required comparatively small,3 but the multiplication of
1 Dearie, Unemployment in the London Building Trade, pp. 30-1,
* Ibid., p. 40.
3 EC., 1907, p. 2,023.
26 STUDY ,OF INDUSTRIAL FLUCTUATION.
productive capacity comparatively cheap,1 and the tempta-
tion to over-investment pro tanto more severe.
Again, accessibility is greater where joint-stock manage-
ment prevails than where private enterprise is the rule.
Thus in cotton-spinning, which is mostly carried on by
joint-stock companies, fluctuation seems to be more severe
than in cotton- weaving,2 which, owing to differences in the
requirements of management,3 is mostly in ,the hands
of private firms. Thus weavers were less severely affected
than spinners in the depression of 1902 ; and in the present
boom they seem to have profited more owing to the com-
parative shortage of looms.4
This consideration, however, is sometimes, notably in the
shipping industry, overborne by another — easy access to
the short loan market. In the depression of 1908 in parti-
cular loud complaints were heard of the system of " long
credits,3' by which speculative owners with inadequate
capital and experience had been enabled to make large
additions to their fleet by means of advances from the
money market.5 Again a considerable share of the cheap
money of the 'QO'S, when the doors of foreign investment
were closed, seems to have drifted for want of something
better to do into the hands of such owners. In the recent
boom it appears that these facilities have been considerably
curtailed, and that both bankers and private individuals
have been displaying more caution.6 It is noteworthy
1 Lescure, op. cit., p. 229.
2 Not that the weaving trade has been innocent of over-invest-
ment, particularly in 1906-7 ; cf. Messrs. Ellison's circulars.
3 Cf. Chapman, Cotton Industry and Trade, p. 44.
4 Cf. Ellison, reports of seasons 1910-11 and 1911-12.
5 Cf. Economist, 1908, i. p. 9.
6 Cf. Times, Shipping Supplement of Dec. 13, 1912, p. 14. Some-
thing of the same development seemed to be in the mind of a Liver-
pool business man of long experience, who told me that he remem-
bered three periods of business methods : one when a new enter-
prise in any industry was financed through some large house in
that industry ; one when the banks had begun to enter into com-
petition with these old-established houses, and invest the deposits
TEMPTATIONS TO OVER-INVESTMENT. 2;
that of the tonnage under construction at the end of 1912
the majority seems to have consisted of liners and tank
steamers ; l and the confidence with which a revival in
the level of freights was long anticipated2 seems to be
partly due to the disappearance of the impecunious tramp-
owner relying on cheap money.
Where the long credits are furnished not by the money-
market to the manufacturer but by the manufacturer to
the wholesaler, the effect upon investment is the same.
Thus in 1886 3 and again in 1899 3 the boom in the woollen
industry was accentuated by the speculative purchases of
wholesalers fostered by the system of six to twelve months'
credits in vogue in the trade.
Similar differences prevail not only between different
industries but between the same industry in different
countries. In this connection the different course in England,
Germany and the United States of the boom in electrical
transport and lighting which had its origin in the go's
is worthy of attention.4 In Germany, where the new
undertakings met with generous financial assistance from
the Banks, the boom proceeded without interruption from
1894 to 1900. But the over-investment and the consequent
of the public often without any real knowledge of the prospects of
the investment ; and finally within the last few years a tendency,
with rising prices and a rising rate of interest, for investors to take
more trouble for themselves to find out remunerative openings,
this tendency reacting upon the banks and making them also more
critical of applications for loans.
1 Moss, Circular, Jan. i, 1913. Cf. Statist, Aug. 10, 1912. " Of
all the new contracts reported as booked by the Scotch builders in
July there was not a single cargo steamer or tramp."
2 Cf. Times of July 8, 1913, p. 22, where the dip in freights was
attributed entirely to seasonal causes. The competition of the big
lines has however proved more formidable to the tramps than was
expected (cf. Times Review, Jan. 16, 1914, p. 6).
3 Cf. EC. Histories of these years. On the latter occasion especi-
ally there were loud complaints of the system, and some attempted
breach of contract by wholesalers when the inevitable reaction
came in 1900.
4 The following diagram of electrical history in the three leading
countries may be appended here for reference : —
28 STUDY OF INDUSTRIAL FLUCTUATION.
reaction were correspondingly severe, and in 1901 x-3 the
transport and electrical industries underwent a serious
depression. In England on the other hand progress was
considerably slower ; and in 1897 the capital invested in
the electrical plant manufacturing industry only amounted
to some four or five million pounds, as compared with ten
to twelve million in Germany.2
But if progress was less rapid 3 in England, it was longer
England 1895 6789 1900 12345
III
II
6789
10
II 12
Germany
III
II
I
America-
III
/
II
a.3
2,3
Level III = great advance, II = moderate advance, I = relapse.
The arable numbers indicate the predominant feature of the boom,
i standing for illumination and tramways, etc., 2 for manufacturing,
and 3 for railroad electrification.
1 Copper consumption declined in 1901 by 20,000 tons.
2 Messrs. Lewis, Circular of 1898. Indeed the statistics of
English copper consumption showed an actual decrease in 1897 °f
5,750 tons (as compared with an increase of 13,500 tons in Germany,
and of nearly as much in 1898. But this seems to be merely one
of the many traps lying in the road of those who would try to draw
conclusions from the published statistics of copper, and to indicate
merely a contraction of stocks in the hands of consumers. Never-
theless this is sufficient to indicate the presence of a more cautious
spirit in English manufacturers than prevailed in Germany.
8 The delay seems to have been due in part to the policy of English
municipalities, many of which owned extensive gas plants, and
TEMPTATIONS TO OVER-INVESTMENT. 29
sustained. In 1899 two of the largest foreign electrical
machinery manufacturing companies established their
works in the Midlands : in 1900, fired by the success of the
tubes, the old Metropolitan railways were electrified, by
1902 the process had been extended to many suburban
lines and to the London Municipal Tramways, and by 1903
to the tramways of Middlesex and North London.1 Thanks
to its more moderate and controlled expansion, the English
electrical industry escaped altogether the catastrophe that
befell its German neighbour. In America as in England
the boom of 1895 was a half-hearted one,2 and it was not
till 1899 that an expansion really set in.3 One is forced
to the conclusion that in Germany easy access to the money
market enabled the mere fact of invention to accomplish
that which in other countries required the co-operation of
other factors.
Finally some instances may be given of the effect of
sudden increases of accessibility in stimulating investment.
The Limited Liability Act of 1859 and the joint-stock
Acts of 1862 led to a sudden increase of accessibility in
put hindrances in the way of electrical development : just as the
check to the German industry, in 1910, when copper consumption
remained stationary, must be referred to the delay in the electrifica-
tion of railways, due partly to the obstructionist policy of the Prus-
sian Government (Ec., 1910, ii. 925).
1 English consumption of copper rose by 21,000 tons in 1900,
broke all records in 1901, and was responsible for the greater part
of the European increase of 53,000 tons in 1902 : while the published
decrease of 17,740 tons in 1903 is again held by Messrs. Lewis to
be misleading, since " much more copper has been used for elec-
trical purposes."
2 While the consumption of copper in Europe increased by 42 per
cent, in 1896, that in the United States declined by some 11,000
tons, or about n per cent. In 1897 there was a slight increase
of some 4,000 tons, and in 1898 a renewed fall of 8,000 tons ; and
the slight rise in price was only rendered possible by large exports to
Europe.
3 In spite of the exactions of the Copper Trust, American con-
sumption increased in that year by 55,500 tons, or 44 per cent.,
and the electrification of street railroads proceeded apace. In 1900
consumption increased again by 23! thousand tons, in 1901 it is
described as " phenomenal," and in 1902 it rose by another 55 1
thousand tons.
30 STUDY OF INDUSTRIAL FLUCTUATION.
numerous trades and countries, and precipitated the crisis
and depression of 1866. l In Germany the transport boom
of the 'go's was accentuated by the law of 1892 facilitating
the construction of local railways, the length of which
increased from 1,034-8 km. in 1892 to 8,454-9 km. in 1901.2
The analogous French law of 1880 seems to have had a
similar influence.3
1 Cf. Lord Goschen's famous " Seven per cent." The effect
upon railway building is discussed in chap. ii. § 3.
2 Lescure, op. cit., p. 208.
3 Ibid., p. 149.
CHAPTER II.
AGGRAVATIONS OF DEPRESSION.
§ i. IMPERFECT DIVISIBILITY AND INTRACTA-
BILITY OF THE INSTRUMENT.
THE next characteristic of modern industry which calls
for our attention is the large size of the most convenient
scale of production and consequently of the most usual
unit of investment. The individual producer is not merely
ignorant of the extent to which his rivals are enlarging their
productive capacity : that knowledge would not in any
event be sufficient to control his actions. He is in any
case either compelled, or at any rate subject to strong
temptation, to enlarge his own capacity to an extent which
is not warranted by the rise in the exchange value of his
product ; and the severity of depression is correspondingly
aggravated.
This difficulty of the imperfect divisibility of the unit of
investment is most acute in those industries which are often
loosely but conveniently lumped together as " public
utility services " —those, that is, which require the con-
struction of a large continuous distributive plant. This
is one important cause of the prominent part played by
railway enterprise of various kinds in the history of indus-
trial fluctuation. The provision of illumination belongs
to the same category.1
1 For instance, we find that in November, 1906, the Berlin Elec-
trical Works, founded many years before, had just reached the
limit of their generating capacity, and are projecting a new plant
of 18,000 h.p. (Ec., 1906, 1875). Similarly in England in 1911 " the
poor return on the capital " of electric supply companies " in some
31
32 STUDY OF INDUSTRIAL FLUCTUATION.
Closely allied with this peculiarity is another which can
best be described as the intractability of the unit of invest-
ment. Producers in face of depression are not only unable,
owing to the discontinuity which would characterise the
process of putting their plant out of operation, to attain
that level of production which they would have chosen had
they been perfectly free agents : it will often no longer be
to their interest to do so. Provided that the demand for
their products is elastic,1 the best they can do will be to
maintain production on the largest possible scale. If, how-
ever, production is so largely increased that the demand is
inelastic, their right course will depend upon the degree of
its inelasticity and upon the ratio of the special costs in-
curred in maintaining in operation a given unit of plant to
the cost involved by closing it down and by making pro-
vision for putting it back into use. Other things being
equal, the greater this ratio the less severe will be the
depression from which they suffer.
The contrast between different industries in this respect
is well illustrated by the British pig-iron and coal trades.
For technical reasons, a mine-owner is very unwilling to
shut down a colliery or any part of a colliery that has once
been opened;2 but the damping down of a blast-furnace,
though inconvenient and expensive, is a comparatively
easy matter. Indeed in this country even in the most
prosperous times a large percentage of the blast-furnaces
cases is due to the inability of the companies to employ their plant
to anything like its full capacity " (Id., 1911, i. 612). These cases of
transport and illumination show clearly the difference between
the points elaborated in this and the preceding chapter. For
though a large unit of investment usually implies a long period of
gestation, the excesses of competitive investment which make the
length of the period an important determinant of the severity of
depression are prevented in these trades by the prevalence of a
limited monopoly.
1 I.e., that depression only exists in the second of the two senses
discussed on p. 3.
" When long wall is the system of working there is a serious
risk that the roof may fall in and the working faces be closed up
if any portion of the pit is left unworked for more than a few days "
(Ec. H. of 1896, p. 20).
AGGRAVATIONS OF DEPRESSION. 33
in existence are out of use.1 Thus the divergence between
the productive power in existence and the productive power
in use is much greater with pig-iron than with coal : even,
therefore, if over-investment has been less in the latter than
the former trade, the resultant depression is less easily
mitigated.2
A similar difference may sometimes be observed between
the same industry in different countries. A glance at
Charts I. and II. will show that while in England the pro-
duction of pig-iron is normally restrained before the price
breaks, in America (in spite of the greater effectiveness of
combination) production has constantly tended to increase
after the break in prices, and so to aggravate their fall. One
cause of this must, I think, be found in the superior size and
capacity of the American furnaces,3 which renders their
damping down a more serious matter.4
It should be observed that with most minerals the diffi-
culties in the way of restriction do not seem to be so univer-
sal as with coal. Thus while in the slump of 1908-09 the
Amalgamated Copper Company was averse from restriction,
1 Furnaces.
In existence. In blast (average for year).
1880 . . 959 590
1889 .. 813 477-10
1899 .. 596 409
1900 . . . . . . 562 . . 397
1906 .. ..517 .. 36775
1907 . . . . 507 . . 366-25
A large number of the disused furnaces are no doubt antiquated
and unsuitable, but still the British ironmaster seems deliberately
to prefer to keep a large margin of power in reserve. [Figures from
Brit. Iron Trade Ass. Reports.]
2 According to figures given by Sir Robert Giffen (quoted Ibid.,
1878) between 1867 and 1875 the capital invested in ironworks in-
creased 314 per cent., in mines only 195 per cent. But it will be
seen that in the depression of the '70*3, the price of coal fell 63 per
cent., the price of pig-iron 58 per cent., below the maximum of 1872.
3 The average annual capacity in Great Britain in 1905 was
28,000 tons, in America twice (and in Pennsylvania nearly three
times) that amount. Jeans, English Iron Industry, p. 41.
4 In America in 1912 " many additional furnaces would have
been banked if they had been in a condition to stand the strain "
(Ec. H. of 1912).
34 STUDY OF INDUSTRIAL FLUCTUATION.
as its operations were most economically conducted on a
very large scale, the great Anaconda mine kept in reserve
three shafts with a capacity of over 26 million Ibs. a month,
the re-opening of which was only a matter of weeks.1 In-
deed the extractive industries as a whole would naturally
seem to be more favourably situated as regards restriction
of output than either agriculture or manufacture, since the
special costs involved in maintaining the maximum output
include a considerable allowance for depreciation of the
capital value of the property.2
In the shipping trade the influences adverse to restriction
are strengthened by special causes. In times of low freights
much tonnage continues to be actively employed which is
not really earning enough to cover special costs. The reason
is that owners accept outward cargoes at obviously un-
remunerative rates in the hope of obtaining homeward
freights which will cover the cost of the double voyage ;
and this hope is often not fulfilled.3 More especially is
this liable to occur in the numerous instances in which ships
are managed on a commission basis on behalf of the owning
company. The manager, in order to obtain his commission
on gross receipts, accepts outward charters at low rates on
the plea of the expectation of good homeward freights, even
when he knows that no such expectation can reasonably
be entertained.4
1 EC., 1912, i. 8.
a In language more plain but less in accord with the form of our
analysis, the rendering of their characteristic service is only post-
poned and not irretrievably abandoned by restriction. An excep-
tion that supports this rule is furnished by rubber. (Cf. p. 23.)
3 " Some extraordinary voyages were undertaken last year
(1907). A steamer was sent from the Tyne to Java in ballast to
load home at a very ordinary rate. When rates were very low
from the River Plate in July, steamers were sent in ballast from
the Plate to Java and Plate to Karachi to load home. A steamer
was sent in ballast from the Plate to Natal, and then, finding the
Indian market depressed, was sent on to Australia to load home." —
J. White, Circular.
* The transfer of tonnage to more prudent owners by the fore-
closure of mortgages led to some beneficial laying-up of tonnage in
1909. In the same year the compulsory " laying up " of a certain
AGGRAVATIONS OF DEPRESSION. 35
The practicability of a policy of restriction depends of
course largely on the facilities for combined action. Owing
to its greater localisation,1 the pig-iron trade is considerably
better situated than the coal trade in this respect. The
copper industry is less favourably situated than might be
supposed. After the collapse of 1907, it was not until 1910,
when the consumptive situation was already rapidly im-
proving, that an arrangement for restriction was made
between the Amalgamated and the Rio Tinto, and even then
the result seems to have been small. The tramp-shipping
trade is very badly organised : the rings and conferences
into which the great lines are organised have never been
extended with success to the tramps. Attempts at organi-
sation have indeed been made ; but even the more energetic
of them, such as that concerned with the Baltic and White
Sea trade, do not seem to have had any success in develop-
ing a common policy as to freights, still less as to building.2
On the other hand, as is well known, the Lancashire cotton
trade is exceptionally strongly organised with regard to
the adoption of a common policy of short time.
Finally it must be pointed out that the " intractability "
of the instrument not only aggravates the effect of such
over-investment as has taken place, but also actually in-
creases the temptation to over-investment. For the absence
of a margin of idle productive power to draw on permits
amount of deadweight capacity owing to the adoption of an inter-
national load-line seems to have been a sensible relief to freights,
just as the revision of the Free-board Tables, which added some i£
millions to the tonnage in existence, had aggravated the depression
in 1908.
1 The centres of coal production are too widely distributed over
the country to afford facilities for combined action (cf. Levy, Mono-
poly and Competition, p. 182) : on the other hand, about 13 per
cent, of the total output of pig-iron is produced in the South of
Scotland and about 20 per cent, in the Cleveland district, and within
these areas intercommunication is easy.
2 Cf. EC., 1908, ii. 1,228. N.B. however the proposal recently
mooted to impose a levy of id. a ton on all ships belonging to the
International Shipping Federation, which, it is calculated, would
suffice to compensate for the laying up of a million tons for a month
(Ec., Feb. 21, 1914)'
36 STUDY OF INDUSTRIAL FLUCTUATION.
a rise in exchange values to exercise its whole influence upon
new investment. Further, the relative smallness of special
costs which a high degree of intractability implies is an
inducement to investment : for people are more ready to
invest in instruments which they know they can immediately
put into use without much additional expenditure and with
a reasonable chance of covering working expenses. This
consideration seems to be of special importance in shipping :
for the depreciation of ships being due to the rapid progress
of the technique of construction and to the liability to large
falls in the " cost of reproduction " rather than to physical
wear and tear, does not force itself upon the attention of
owners, and allowance for it is apt to be relegated to the
category of general costs for an indefinitely long period.1
§ 2. LONGEVITY OF THE INSTRUMENT.
The last characteristic to be considered is the longevity
of the unit of investment. The normal length of life of
the instruments of production might be expected to be an
even more important determinant of the duration of de-
pression than the ease with which they can be consigned to
a living death. Indeed Karl Marx long ago suggested 2 that
the alleged decennial character of crises was due to the fact
that the average length of life of an instrument of production
was ten years, so that at the end of that interval the need
for replacement would automatically create a new demand
for instrumental goods. This suggestion has not received
much attention at the hands of subsequent writers, except
for M. Aftalion,3 who adopts it parenthetically, and M.
1 In the recent boom, however, the steady rise of special costs,
especially those of labour, insurance and provisions, seems to have
exercised a deterrent effect on speculative building (Times Ship-
ping Supplement, Dec. 13, 1912, p. 14).
z Capital, Vol. II. Part II. chap. ix. (Eng. Tr., p. 211). "...
This cycle is determined by the lifetime, in other words by the
period of reproduction or turnover, of the invested capital. . . .
One may assume that this life-cycle, in the essential branches of
great industry, now averages ten years. . . . This cycle furnishes
a material basis for . . . periodical commercial crises."
3 Journal d' Economic Politique, 1909, p. 206.
AGGRAVATIONS OF DEPRESSION. 37
Lescure,1 who dismisses it with contempt. In the hands
of Marx indeed it seems to have been little more than an
ingenious guess, unsupported by evidence and invented to
suit the facts : and as an explanation of general fluctuation
it is open to obvious objections. It is clearly extremely
improbable that the length of life of all the principal instru-
ments of production should be the same.2 Nor is the
matter improved by Marx's qualification that it may be
the period not of material life but of technical usefulness
that is in question — that psychical depreciation as well as
physical deterioration must be taken into account; for
the inventions and improvements by which the instruments
of production are superseded are obviously less likely to be
uniform in their occurrence between trade and trade and
period and period than the visitations of physical decay.
Nevertheless there is considerable evidence in support of a
careful application of the principle to particular trades.
The operation of the principle in its simplest form is as
follows. Granted an initial rise in the exchange value of
any product, there will follow, as we have seen, an abnor-
mally large investment in the instruments of production.
The approximately simultaneous wearing out of these
instruments will produce an appreciable shortage of the
product in question, reflected in a rise in its exchange value
and leading to a fresh outburst of investment on an abnormal
scale, followed by the usual results.
This analysis disposes at once of one objection to the
theory of the importance of the length of life of the instru-
ment. This comes from M. Lescure, who points out that
the need for decennial renewals would not necessarily imply
a special outburst of constructive activity every ten years.
' The fixed capital A created in 1880 would be worn out
and replaced in 1890, the fixed capital B created in 1881
1 Op. dt., p. 496.
2 Marx indeed only contends for a mean of ten years : but unless
the mean corresponds with the mode, there is no reason for any
exceptional activity in the tenth year : and the only instance he
offers of an actual ten year period is that of a locomotive (p. 201).
38 STUDY OF INDUSTRIAL FLUCTUATION.
would be worn out and replaced in 1891 " 1 : there is nothing
so far to prevent the process of investment from being per-
fectly uniform. But the theory does not profess to offer a
complete explanation of the phenomena of fluctuation : it
merely suggests, when taken in company with the theory
of the period of gestation, that granted some single initial
outburst of investment, the subsequent alternation of high
and low prices tends to be self-perpetuating.
A second and more plausible objection is that the theory
is irrelevant because it is quite clear that in fact the periods
of exceptional activity in investment are periods not simply
of replacement but of new construction. This objection in
its crude form is rebutted by the reflection that in a com-
petitive regime each producer feels bound to make his own
arrangements to meet a rise, from whatever cause proceed-
ing, in the exchange value of the product. But in fact it is
clear in certain important cases that the wearing out of the
instruments causes no such rise in exchange value, because
it is physically necessary that the instruments be replaced
before any shortage occurs : and further that the localisa-
tion of supply would make a rise in the exchange value of
existing services irrelevant to investment policy elsewhere.
In such cases, of which the railway industry is the most
conspicuous instance, we must have recourse to that real
but much-abused factor the " psychology of the business
man." At the one extreme in this matter stands Mr. Hull,
who regards the business man as an entirely rational and
clear-headed expert, " not impelled by ' an organic defect
of human nature/ nor by ' celestial influences/ nor by
' mental waves ' nor by ' sun spots/ but governed solely
by plain, simple, unadorned business considerations."2 At
the other extreme stand those who appear to regard him as
a wholly incalculable and irresponsible person, whose mental
processes it is entirely beyond our power to analyse. But
it is clearly possible to hold that while a business man's
actions are often not rational, they are nevertheless very
1 Op. cit., p. 497.
2 Op. cit., p. 126.
AGGRAVATIONS OF DEPRESSION. 39
seldom wholly incalculable in the sense that they are not
provoked by any external stimulus : they may be the result
of non-rational or only semi-rational inference *• from the
actions of other people or from other observed facts. Now
if a business man sees other business men making extensive
purchases for renewals, it is likely to arouse in him a not
wholly rational feeling of optimism, which will tempt him
to engage in new enterprises of a similar character. He may
perhaps have been planning such an enterprise for some
time, watching a gradual growth of demand and maturing
his designs : but in any case the rush for renewals will be
the factor which actually determines the moment at which
his designs are put into effect.2
It remains to apply this principle in detail to particular
industries. The late editor of the Reports of the British
Iron Trade Association* estimates the average life of an iron
rail at ten years. And the only instrument for which Marx
is definitely prepared to assert a ten years' life is the railway
1 Cf. Graham. Wallas, Human Nature in Politics, pp. 98 ff.
2 The suggestion may be made clearer by an analogy from other
fields. The modern method of treating many diseases is by in-
jecting a culture of the patient's own blood which has been sterilized.
The patient's white corpuscles first attack the dead germs in the
culture, and then encouraged by their easy victory advance against
the live germs in the remainder of the blood. It is not clear to the
outsider why they should not do this at once ; but the psychology
of phagocytes, like that of business men, is an obscure but no doubt
a fascinating study.
Or we may compare the manner in which the ancient Romans
introduced little runnels of fresh water into the Cloaca Maxima,
thereby stimulating the flow of the inert mass.
It should be noted that while the business man's action is not
wholly rational, neither is it wholly irrational. In the first place
the mere fact that other people think it worth while to replace
their plant on a large scale instead of letting it fall into disuse is
pro tanto an indication that the industry is in a good condition
and the moment favourable for investment. In so far as he is
uncertain, as he may well be in some cases, whether the new orders
are for renewals or for new construction, the effect of this con-
sideration is enhanced. In the second place when he sees the flood
of new orders he will anticipate a rise in costs and conclude that
if he is going to invest at all the sooner he does so the better (cf.
chap. iii. § 2).
3 Review of 1878, introduction.
40 STUDY OF INDUSTRIAL FLUCTUATION.
locomotive. It seems possible that this throws some light
on the successive outbursts of English railway investment,
which (comparing Table II with our hypothesis of a two-
years' period of gestation) we must date at 1834, 1844, 1854,
1863, the last revival being hastened by the Joint-Stock
Acts,1 just as we find an intermediate revival in 1860 pre-
cipitated by the Limited Liability Act of 1859 and telling
on the mileage of 1862. 2 In the mileage expansions of
1 Cf. chap. i. § 2.
2 If we may suppose the life of a rail to begin in the same year
as the railroad is authorised, i.e., if production follows thus quickly
upon investment, the theory stands ; but if, as seems in some ways
more likely, we should regard its life as beginning in the year between
authorisation and opening, we must postulate a nine years' life,
and revise the dates as follows : —
First large output of rails 1835 renewed in 1844.
Second „ ,, 1845 ,, 1854.
Third „ „ 1855 „ 1864.
In this case we must suppose the large renewals spread over 1853-4
and 1862-3-4 ; and suppose that this dissipation did not appreciably
affect matters in the first period, but was an additional cause of
the prematurity of the boom in the second. If this account is
correct the large rise in iron prices or production which is observ-
able in 1844 and 1854 still more clearly owes its initial origin to
renewals rather than to new construction. Cf. the following
figures : —
Iron Price (J evens, In-
vestigations in Currency
Pig-iron Production and Finance, p. 147.
(Hull, op. cit., p. 268). Index [1782 - 100]).
1843 • 1,215,350 tons . . 43
1844 . 1,999,608 „ 43
1852 . . 2,701,000 „ . . 40
1853 . . (not available) . . 53
1854 . 3,218,154 tons . . 60
The view that a large increase in the consumption for new build-
ing takes place in the year following a large increase in the con-
sumption for renewals is confirmed by the following figures, belong-
ing to a period where the life of the iron rail has ceased to be relevant.
Steel Rails Steel Rails for
for renewals. new lines.
1877 . . 240,427 . . 32,520
1878 . . 330,282 . . 51,960
1879 116,653 •• 74,640
1880 . . 231,229 . . 44,280
1881 .' . . 391,881 .. 37,440
1882 . . . 453,866 .. 48,000
Brit. Iron Trade Assoc. Report of 1882, p. 122.
AGGRAVATIONS OF DEPRESSION. 41
I868-91 and 1878-9 we have the harmonics of the series
beginning with the invention of railways, and in those of
1872-4 and 1882 2 the harmonics of the series beginning with
the extended facilities for joint-stock enterprise.3 After this
the life of the iron rail, of course, becomes irrelevant. I do
not make this suggestion with any confidence, but I do not
think it is entirely fanciful.
It is at least curious that for the machinery of cotton-
spinning, the other industry which played a leading part
in the English crises of the first part of the nineteenth
century, Professor Taussig estimates an average life of ten
years.4 In Lancashire, at the present day, according to
an expert informant, it is usual to write down the value of
engines and boilers by 10 per cent, per annum. Machinery,
however, is commonly written down 7^ per cent., and it is
possible that the thirteen-years' period thus suggested
helps to explain the net decrease of spinning capacity in
1896 (thirteen years after the prosperity of 1883), its stag-
nation in 1903 (thirteen years after the maximum profits
of 1890), and its very moderate increase, in spite of high
profits, in 1912 (thirteen years after the maximum profits
of 1899). (Cf. chart.)
I have no similar information for the woollen industry,
but it is at least curious that the great booms in that trade
1 Supposing the revival shown in the mileage of 1865-66 to
have been prematurely cut short by the " joint-stock " crisis of
1866, and to have been resumed, without need of further stimulus
from renewals, in the mileage of 1868-69.
2 Though the substitution of steel for iron rails in the later seven-
ties breaks the periodicity for ever, it would naturally in the first
instance be most considerable in the years in which the greatest
number of iron rails reached their natural term.
3 There is indeed an alternative explanation of the shortening of
the interval between successive booms — the deterioration in quality
and wearing power of the iron rail. A writer in Science in 1872
(quoted in EC. H. of that year) asserts that while the first English
iron rails lasted 15-20 years, 8 years is now considered good, and
the average is not above 3 to 6. I see no reason, however, for
preferring these estimates to that of the Brit. Iron Trade Assn.
4 Quar. Jour, of EC., 1908, p. 342.
42 STUDY OF INDUSTRIAL FLUCTUATION.
in recent years arose in 1886, 1899 and 1912, i.e., at inter-
vals of thirteen years.1
In the shipping trade also the length of life of the instru-
ment throws light on some dark places in the history of
the course of contracts for new construction.
The Economist of 1884 estimates the annual loss of ton-
nage from the British register at 5 per cent., of the total
tonnage existing at the beginning of the year. At first
sight this seems to correspond with the estimate given me
by a Liverpool shipowner of twenty years as the average
length of life of a vessel upon the British register. There
are various reasons, however, for considering this estimate
too large : — (i) The 1884 figures include only tonnage
destroyed, not that transferred to other countries ; (2)
with a steadily growing tonnage an indication of the length
of life of a vessel is given by the inverse of the fraction ob-
tained by dividing the total loss in any year not by the
tonnage in existence at the beginning of the year, but by
the average tonnage in existence during a period, dating
back from the said year, equal to the length of life in ques-
tion.2
From the figures given in Table III, taking the average
value of the formula in the footnote, we obtain an average
length of life 'of 16-8 years. This result is confirmed by
the figures given in Table IV. When we remember that
in years of high freights (such as 1898-1900 and 1912) the
foreign sales are naturally unusually large, and likely to
1 There is perhaps a trace of another period in 1877, when in spite
of the complaints of bad trade, it was generally affirmed that the
amount of machinery in the industry was no longer excessive,
though manufacturers were working it to the utmost. But the
later harmonics of this series cannot be distinguished.
2 If a is the amount on the register at the beginning of the year,
b the total loss during the year, c the average annual net increase
in tonnage (assumed to be uniform throughout the period), x the
length of life, we get the formula :
AGGRAVATIONS OF DEPRESSION. 43
contain an unusually large percentage of fairly new ships,
and conversely in years of depression, such as 1908, we may
with some confidence place the median of the 1885, 1890
and 1895 curves in 1901, 1906-7 and 1912 respectively,
which, if we assume a correspondence of the median and
the mode, corresponds admirably with our sixteen to seven-
teen years' period.
We shall then see in the large natural losses and sales
of 1890-1 the harmonics of the large output of 1874 ; in
the large natural losses of 1896-7 and sales of 1895 and
1898-9, those of the outburst of 1880-3 ; in the losses of
1903-5 the echoes of the great output of 1888-9 '» an(^ m
those of 1908-10 the echoes of 1891-2. 1 These considera-
tions help to explain the revival of new contracts in the
last quarters of 1890 and 1891, when the apex of the freight
boom was passed : the curious increase, during the preva-
lence of low freights, in the second half of 1895 and the first
half of 18962 : the similar phenomenon of 19043 and of
the last quarter of 1908 4 and the first of 1909 ; as well as
the first wave of the subsequent boom.5
Yet one or two more instances of some interest may be
given. The coffee plant ceases to yield good crops after its
1 The large foreign sales of 1900 must be attributed to high
freights. But conversely both the sale and break-up of ships is
precipitated by low freights ; hence we may legitimately connect
the large sales of 1895 with the 1880 boom, and the large " loss "
of 1903 with 1888. Cf. Board of Trade Freight Index (1900 = 100).
1893 • • • 83-6 1896 . . . 79-4 1901 . . . 75-0
1894 • • • 79'8 1897 . . . 82-9 1902 . . . 707
1895 . . . 75*3 1900 . . . 100-0 1903 . . . 72-8
The high pre-i895 percentage for 1910 and the loud and general
complaints and warnings of an over-supply of antiquated tonnage
(cf. EC., 1910, i. 430, and Moss Circular, July i, 1910), justify us
in assuming that the ships eventually broken up in that year had
had an unusually long life.
2 Cf. EC., 1897, P- 12.
3 Cf. Id., 1904, p. 2138.
4 Cf. Id., 1908, ii. 1228.
6 The expansion of orders in 1885 may also possibly be connected
with the revival of 1868, though the " losses " of 1885 do not seem
to have been unusually large.
44 STUDY OF INDUSTRIAL FLUCTUATION.
twentieth year,1 so that towards the end of this decade we
may witness a falling-off of production due to the wearing
out of the plants which caused the trouble at the end of
last century. There is, of course, no uniform length of
life for an oil-field ; but Table I, combined with other
evidence,2 suggests that its typical history may be repre-
sented as follows : —
Stage i represents the gradual increase during the period
of exploration, 2 the activity of spontaneous " gushers,"
3 the exhaustion of the latter, 4 the gradual increase of pro-
duction by pumping, 5 the natural exhaustion of the wells,
6 the desperate efforts by improved apparatus and new
sinkings to postpone the evil day, 7 the final relapse into
passivity. 1-4 are well represented by the Dutch Indies,
1901-6, and Mexico, 1907-11 ; 5-7 by the Appalachian
field.
The simultaneous attainment of 5 by the Baku and
Pennsylvanian wells [Table IA] led to the boom in the
Scottish mineral oil trade in 1903, that of stage 6 to the
relapse in 1904 ; but in 1906 their arrival at 7 and the
simultaneous arrival of the Texas field at 33 led to a shortage
which precipitated the first wave of the great oil boom.
Finally it should be observed that the almost infinite
length of life (from a cyclical point of view) of the instru-
ment of production of house room helps to explain the
chronic stagnation of the building trade since the boom
towards the end of the nineties — its failure to participate in
1 Bartholomew's Atlas of the World's Commerce, p. 85.
2 Cf. article on oil-supplies of Canada, Daily Chronicle, July 17,
3 Cf. Enc. Brit., Article " Petroleum."
AGGRAVATIONS OF DEPRESSION. 45
the boom of 1906-7 and its reluctance to follow that of
recent years. On the other hand, it is plausibly urged by
the Land Enquiry Committee (vol. ii. p. 94) that in this case
the mere fact of longevity induces an undesirable caution
in investment policy.
CHAPTER III.
FLUCTUATIONS IN COST.
§ i. COSTS AND CONSUMPTIVE INDUSTRY.
WE have been concerned so far with certain tendencies
towards fluctuation which, granted an initial rise in the
exchange value of the services rendered by any trade, are
found to be inherent in the modern system of large scale
competitive capitalistic industry. Except, however, in
discussing the longevity of the instrument, we have shed
little light upon the possible causes of such a rise : and in
fact most of our illustrations have been taken from cases
in which the rise is due to a raising of demand for the finished
product, the origin of which is not relevant to the subject
of the present group of chapters. It is clear, however, that
the rise in margins may proceed equally well from a fall in
the costs of production : and further that, apart from all
considerations of over-investment, a fall in margins and a
depression in both the senses indicated on page 3 may result
from a rise in the costs of production.
Abundant illustrations of this action of increased costs
might be given from those industries which are dependent
on annual harvests for their supplies of raw material. Thus
the level of profits in the Lancashire cotton-spinning trade
is largely affected by the size of the American crop and the
consequent price of raw cotton (see Chart). For instance,
the depression of 1910 was mainly due to the American
crop shortage of 1909 ; conversely the big yields of 1889-90
worked in conjunction with a good demand to produce the
record profits of 1890, and those of 1896-8 helped (in spite
46
FLUCTUATIONS IN COST. 47
of the temporary set-back noted below on page 50) to pro-
duce the boom culminating in 1899. Again in 1904 and 1911
large crops led to a sufficient fall in prices to permit the full
effects of a good consumptive demand to make themselves
felt in the spinning trade.1
In the jute trade also the effects of fluctuations in the crop
upon manufacturing profits are important, and are aggra-
vated by several influences. First, the fluctuations in the
price of raw jute are enhanced by the wildness of the Indian
Government's estimates of the growing crop.2 Secondly,
there is sometimes a tendency for a shortage of jute to coin-
cide with a shortage of those products whose envelopment
constitutes an important source of demand. Thus in India
in 1908 the same drought which caused a shortage of the
food-grains also reduced the length of the jute crop some
inches below its standard six feet.3 Again in 1910-1 the
scarcity of jute was accompanied by a scarcity of some of
its leading clients, — first cotton,4 then American wheat,
then Brazilian coffee : thus manufacturers were squeezed
between high costs and low demand.
Again in the woollen trade, a shortage of supplies of the
raw material led to a continuous rise of prices beginning in
1 Cf. the revival in the volume of yarn exports to the continent
in 1886 and 1894-5, and their decline in 1896.
2 For instance, that of 1909 turned out to be ij million bales
larger and that of 1910 half a million bales smaller than the final
official estimate.
3 Review of Indian Trade, 1908—9, p. 62. There was, however,
a pronounced fall in price, caused by the increase (in response to
the high demand of 1905-7) of acreage under jute from 3-1 m.
acres in the seed time of 1905 to 3-9 m. acres in the seed time of
1907 (cf. Stat. Abs. of India, 1911, p. 126), — an increase which
naturally aggravated the shortage of food-grains (R. of I. T., 1906-7,
PP- 36, 56). This compensatory action benefits Scottish makers,
who are not affected by the conditions of Indian demand, and
who in 1908 were enabled by the heavy Indian exports of raw jute
(17-9 m. cwt. in 1908-9 as against 14-2 m. in 1907-8) to do a toler-
able volume of business at the lower level of values.
4 At the time of the civil war, jute and cotton were mainly rival,
and jute profited largely by the cotton famine ; but at the present
day, when jute has successfully ousted cotton for ordinary packing
purposes, the other relation is considerably the more important.
48 STUDY OF INDUSTRIAL FLUCTUATION.
1868, and prevented the English manufacturing industry
from sharing in the elation of trade in 1872-3 ; while the
breaking in the season of 1904-5 of the long series of Aus-
tralian droughts enabled it to take some share in the boom
of 1906-7. 1
Fluctuations arising from this source will clearly be less
severe the more successful the attempts at inter-local and
, inter-temporal compensation, — in other words, the more
^ numerous the independent sources of supply and the more
efficient the distribution, by a speculative market or other
means, of consumption through time.
With regard to the first point, the position of both the
cotton and the jute 2 trades is exceptionally weak, though
that of the former is no doubt improving. The recent
work of the Lancashire Cotton Growers' Association and
of other private and Governmental enterprise cannot be
ignored ; and it is said that there are now nearly eighty
countries engaged in the cultivation of cotton, and that
the area now coming under the plant in the new cotton
countries is some thirty times as great as the whole cotton
acreage of the United States.3 Nevertheless the difficulties
in the way are well known. 4 When there was a shortage of
American supplies in the (calendar) year 1899, we were able
to obtain 34-1 per cent, of our total imports from other
countries ; but when a similar event arose in 1910, the
corresponding figure is only 25-5 per cent.5 I think we
must be careful for a long time to come not to exaggerate
/ the relief from fluctuation which the industry is likely to
experience from inter-local compensation with regard to
supply.
1 The three chief textiles — cotton, wool and jute — have a!l
suffered from a rise in the cost of material in 1913.
2 India is still for all practical purposes the sole producer.
3 Cf. London Chamber of Commerce Journal, Dec. 1912, p. 367.
4 It is, however, alleged, I know not with what truth, that a great
part of the land hitherto devoted to the cultivation of the poppy
in India is, unlike the rest of India, capable of producing a long
staple cotton suitable for Lancashire consumption.
6 For the two following years the percentages are 23-5 and 23-1.
Computed from figures in Stat. Abs. of U.K. and Ellison Circulars.
FLUCTUATIONS IN COST. 49
With regard to compensation through time the cotton
trade possesses, as is well known, a highly organised specu-
lative market. A glance at the chart will show that the
curve of estimated consumption is considerably more stable
than that of imports,1 i.e., that considerable stocks are held
over in this country from year to year. Moreover, one
further point deserves notice, because it seems to be of
growing importance, — the increased tendency to hold up
stocks in the country of production. For instance, in spite
of the enormous production of the season 1911-2 and of a
large bear influence the price of American cotton did not
fall anything like so much as was anticipated, and spinners
were disappointed. The reason appears to have been that
merchants had made heavy short sales to spinners, and
had covered by future contracts with persons who were in
the main not speculators for the fall but the actual holders
of the surplus of the vast 1911 crop. The volume of such
contracts is said to have been three or four times as numer-
ous as in previous years, and they extend often into 1915
and in some cases even until 1920.2
In this connection mention may be made of the proposal
that has sometimes been put forward for the maintenance
1 It should be noted that the figures, as given in the Fiscal Blue
Book of 1909, p. 153, do not appear to be strictly comparable :
for instance, the continued excess of consumption over net imports
in 1893-6 and 1898-1903 is clearly inconsistent with the figures
of stocks. Messrs. Ellison's net import figures appear to be some-
what larger than the Board of Trade's, but the trend of the two
curves is sufficiently well shown.
a Times Financial Review of 1912. It may be suspected that
throughout the first half of 1913 the same influence was still at
work. " A moderate crop," said Messrs. Ellison in their circular
of October, 1912, " will be amply sufficient for the world's require-
ments, while under a full yield values will have to seek a lower level."
But by December the new crop had been estimated at 14^ million
bales, or well over all previous records except that of 1911 ; yet
spot prices of middling stood at 7-1 id. or i-ozd. higher than the
average for the season 1911-2. "With the near approach of a
new planting season," Messrs. Ellison then said, " a material decline
may easily be witnessed." Yet at the end of June, in spite of
excellent prospects for the crop of 1913, values were still in the
neighbourhood of 6-Sod.
E
50 STUDY OF INDUSTRIAL FLUCTUATION.
of some kind of common cotton reserve in Manchester, the
trustees of which should be obliged to buy when cotton fell
to (say) 4^. and sell when it rose to (say) 6d.1 In any case
it seems clear that in the cotton trade as elsewhere the forces
making for better organisation and more complete inter-
temporal compensation are likely to continue to grow in
strength.
It must be noticed, however, that unless such inter-tem-
poral compensation is perfectly effective, the claim com-
monly made that while increasing the frequency it lessens
the severity of fluctuations,2 is subject to important ex-
ceptions. For instance, in 1895, which was an unsatis-
factory year from the point of view of sales,3 spinning pro-
fits showed a respectable increase ; for the unusually large
stocks which, owing to the slackness of demand, were carried
over from 1894-5 into 1895-6 were suddenly and largely
increased in value at the end of the year owing to the not
altogether anticipated shortage of the 1895-6 crop, and the
credit side of the spinners' balance-sheets consequently
swollen. The same thing occurred in 1900*: conversely
at the beginning of 1898 the spinning trade showed signs
of depression owing to the existence of stocks — not, how-
ever, it would appear very large ones — which were bought
at the comparatively high prices of 1897 but had to be sold
in the form of yarn at the low prices necessitated by the
large crops of the season 1897-8. 5 In the jute trade the
1 Cf. Oualid, Journal d'Economie Politique, Jan. 1912, pp. 58 ff.,
who justly contrasts such schemes with mere " valorisation " (such
as that of coffee by the State of San Paulo), where the holder can
continue to hold on a rising market.
2 Cf. Emery, Stock and Produce Exchanges, p. 122.
3 The rise in values induced a spirit of caution among Indian
importers, and the volume of orders fell heavily (cf . EC., 1896, p. 38).
4 Messrs. Ellison in their circular at the end of the year note
" the refusal of calico consumers to buy goods on the basis of 5^.
to 5|d. per Ib. for raw cotton as freely as they had purchased on
the basis of 3^. to 3%d." : but profits during the year were well
maintained.
6 Ec.t 1898, p. 5. The woollen industry suffered from the same
trouble in 1880. Cf. also the manner in which the crisis of 1847
was precipitated by the very abundance of the English harvests,
FLUCTUATIONS IN COST. 51
increased intensity of fluctuation arising from this cause
is even more serious ; for it would seem that the specialisa-
tion of the function of risk-taking has been carried much
less far than in the cotton industry. The annual reports
of the Dundee Advertiser (quoted in the Economist His-
tories) almost convey the impression that the whole energy
of the spinner and manufacturer is absorbed in securing
snap gains out of variations in the margin between raw
material and cops, and between cops and piece-goods.
And the profits arising from the appreciation of stocks have
the same effect in stimulating investment as profits resting
upon the more solid basis of lower costs, — a good illustra-
tion of the general principle that if one source of fluctuation
is removed while another is left intact, the net result may
in some cases be a decrease in stability.1
Finally, it may be noted that when two industries exercise
a composite demand upon a common material, prosperity
in the one tends naturally to produce depression in the
other. Thus in 1913 the boot trade has been severely handi-
capped by the prosperity of the motor and fancy leather
trades.2 If we adopt the magnitude or relative magnitude
of profits as the criterion of depression, the same applies to
industries which draw upon a common labour supply.
Thus the rise in the price of tin in 1911 was due to the rival
attractions of the rubber plantations to the labourers in the
Malay mines 3 : conversely in Germany in 1900 the textile
industries benefited for a time by the supply of labour dis-
charged from the iron and steel works.4 Again the Scottish
which led to a violent depreciation of the heavy stocks of imported
corn in the hands of Liverpool merchants. (Cunningham, Growth
of English Industry and Commerce, Part II. p. 828.)
1 Cf. Prof. Pigou's remarks (Wealth and Welfare, p. no) on the
effects of ignorance and lack of mobility upon the distribution of
the supply of labour ; an indiscriminate cheapening of the costs of
movement may encourage aimless shifting from job to job (cf.
Edgworth's Review in Econ. Jour., March, 1913).
2 EC. H. of 1913.
3 EC., 1911, i. 206. It seems plausible to suggest that the simul-
taneous rise in coffee may have been due in part to an analogous
situation in Brazil.
4 Lescure, op, cit., p. 206.
52 STUDY OF INDUSTRIAL FLUCTUATION.
mineral oil companies lose (e.g., in I907-8)1 from the pros-
perity and gain (e.g., 1 908-9) 1 from the depression of the
coal trade : for their chief expense — shale miners' wages —
fluctuates not with the conditions of their own industry
but with the wages of coal-miners, — an arrangement which
the men sought in vain to modify in 1903.
On the other hand, where there exists a condition of joint
supply, the increased prosperity of a trade making use of one
of the products leads to an increased prosperity, due to
lowered costs, in the trade making use of the other. The
most important illustration is the export trade in coal,
which profits by the large supplies of tonnage required to
assist in a boom of agricultural or other imports.2 Especi-
ally favourable conditions seem to be such as prevailed, for
instance, in 1891 and again in 1894, when the volume of
total imports rose while that of exports fell, or rose to a
much less extent, so that the hope of good homeward
freights led to a large demand for coal with which to run
out in ballast.3 Thus freights from Odessa and Karachi,
which stood at us. and 175. on January i, 1891, reached
355. and 235. by November and December respectively ;
cotton freights from New Orleans advanced during the year
from 42s. 6d. to 475. 6d, and wood freights from Archangel
1 Years ending April 30.
2 In a rather different fashion the American pig-iron industry
has been known, especially in the winter of 1897-8, to benefit at
the expense of its English neighbour by a big cotton harvest, owing
to the low rates for pig-iron to be carried as ballast in the cotton
ships (Ec., 1898, p. 3).
3 Economist Index (quoted in 3rd Fiscal Blue Book, Cd. 4954
of 1909, p. 53) (1900 = 100)—
Imports. Exports.
1890 75 98
1891 77 93
1892 . . . . 79 90
1893 77 88
1894 ..... 83 .. 91
Exports of coal, etc. —
1890 30,142,839 tons.
1891 ....... 31,084,116
1893 29,031,955 „
1894 33,073,698 „
FLUCTUATIONS IN COST. 53
from 405. to 505. ; * but coal freights declined in every
direction.2 Similarly the remarkable advance in the coal
export trade in 1897 must be attributed in part to the large
shipbuilding of 1896, coupled with a rise in the volume of
imports and a fall in that of exports. 3 Moreover, owing to
the length of the period of gestation, this reflex prosperity
tends to survive that which gave it birth : thus in 1885,
1887, 1902-4 and 1908 — years of poor demand for the ser-
vices of the shipping trade — we again find coal maintaining
or actually extending its export markets with the help of
the competition of superfluous tonnage.4
§ 2. COSTS AND CONSTRUCTIONAL INDUSTRY.
Instances may also be found in constructional indus-
tries of the natural effect of increased costs in producing
depression. For instance, a considerable part of the ex-
penses of production of pig-iron consists in the cost of coal :
hence, since contracts for coal are very rarely of more than
twelve months' duration, we should expect to find some
signs of diminished activity in pig-iron within a year of the
1 Vide tables from Glasgow Herald, quoted in EC. H. of 1891,
p. 36.
2 Vide tables in Cd. 2337 of 1904, pp. 264-6.
3 Economist Index (1900 = 100).
Imports. Exports.
1896 . . . . 92 103
1897 . . . . . 94 . . 102
Exports of coal, etc. —
1896 ....... 34,262,056 tons.
1897 • 37>°96,9i8 „
4 Exports of coal, coke, and manufactured fuel —
1884 . . . . . . . 23,350,230 tons.
1885 23,770,957 „
1886 . .... 23,283,389 „
1887 24,460,967 „
Exports of coal only (excluding bunkers) —
1901 ....... 41,877,081 tons.
1902 ....... 43,159,046
1903 44,950,057
1904 . . 46,255,547
X9o6 . . .... 55»599,77I
1907 ....... 63,600,947
1908 . ... . . . 62,547,175
54 STUDY OF INDUSTRIAL FLUCTUATION.
establishment of a high level of coal prices. And this is in
fact the case. The price of coal began to rise appreciably
in 1882 and the production of pig-iron was curtailed by 1883.
Large rises occurred in 1888-9 and 1898-9, and in each case
the activity of the pig-iron industry began to wane in the
latter year ; and there is no doubt that the famine coal
prices of 1900 were a serious check to the iron industry.
Similarly the price of coal began to revive in 1906, and the
check to pig-iron production came early in 1907. l
Evidence of the stimulating effect of low costs may be
founddn the shipbuilding industry. Thus in the autumn of
1887 one of the impulses to the placing of new orders was
the comparative cheapness of iron after the collapse of the
American railway demand. Again the little speculative
spurts of 1897-8 2 and later igoS-early I9093 are referred
by contemporary observers to the same source. Messrs.
Moss give the same explanation of the increase in 1905*
1 Moreover on all but the first of these occasions, stocks of pig-
iron continued to decrease after the break in production, which
affords some ground for inferring that the check to productive
activity came not from the side of demand but from the side of
costs, since consumption was not reduced to an equivalent extent.
Cf. the following figures of stocks on December 31 (Reports of Brit.
Iron Trade Ass.).
1888 2,588,708 tons.
1889 . ... 1,951,443
1890 ....... 1,393,041
1898 945,3°7
1899 733>989
1900 ....... 456,419
1906 730,752
1907 ....... 221,885
The effect of abnormal increases of cost, such as that caused by
the coal strike in March, 1912, scarcely needs comment.
The British iron trade is hampered also in times of boom by its
dependence on Spanish and other foreign ores : in both these respects
the American industry seems to be in a better position, — an addi-
tional cause of the later restriction of production in that country.
2 Economist History of 1897.
3 " Some far-seeing owners, especially on the Clyde, have been
taking advantage of the very low prices." Moss Circular of July i,
1909.
4 Circulars of July 22, 1905, and February 6, 1906.
FLUCTUATIONS IN COST. 55
and of the aftermath in 1901. 1 It seems possible to invoke
the same cause to help to explain the supplementary booms
of 1891-2, and the temporary recovery in the first quarter
of 1907. It appears, however, that in this respect some
distinction must be made between the owners of tramps
and the great liner companies. Thus the increase of ton-
nage under construction in the first quarter of 1901 is re-
ferred to purchases by the large companies on signs of a
reduction in costs.2 Again, in 1910 : " In the majority of
cases the steamers contracted for are on behalf of regular
lines, which have taken advantage of the situation and
placed their orders at rock-bottom prices."3 But the great
bulk of the orders which go to make up an ordinary ship-
building boom are placed by tramp owners, and it does not
appear that with them costs are usually the decisive factor/
For it does not by any means follow that variations in the
cost of materials are the sole and sufficient explanation of
the fluctuations of constructional industry. Since, however,
this view is very strongly urged by two of the most important
of recent writers on this subject, M. Lescure, the French
theorist, and Mr. Hull, the American iron-master, who regard
it as the key to the whole problem, it appears to be worthy
of very close and detailed examination. Mr. Hull's sug-
gestion is briefly that it is times of low prices of materials
that are the real times of prosperity, for it is then that the
majority of contracts for new construction are placed. But
these contracts are undertaken independently of each other,
and in ignorance of the limits of the available resources of
constructional material. In the low-priced period these
sources have been severely restricted, and the pressure of
demand for the fulfilment of contracts produces a great
rise in the price of materials. This rise discourages the
formation of new contracts, the tide of which begins im-
mediately to ebb. But the rise in prices continues for a
time owing to the necessity of completing existing contracts,
1 Circular of February 6, 1902. By this time costs had fallen 20
per cent, from the highest figures of 1900.
2 Circular of July 6, 1901. 3 Circular of January 2, 1911.
56 STUDY OF INDUSTRIAL FLUCTUATION.
and produces a spurious period of activity long after the
genuine period of activity, as indicated by the volume of
new contracts, has ceased.
Mr. Hull claims that this explanation is based on a long
and " inside " knowledge of the United States iron industry,
and it is therefore somewhat difficult for an outsider to
dispute it. Nevertheless such definite data as are available
are not always in complete accord with his view. For
instance, in the boom of the early eighties he is obliged to
assume, without explanation, a much longer period of
incubation than on other occasions, since while the con-
siderable advance in prices occurred in 1879, the volume
of new construction continued to increase enormously till
1882. * It seems pretty clear that some of the vast railway
mileage of 1882 must have been contracted for at the higher
prices. Again the depression of the early nineties is an
awkward period : for the price of iron declined steadily
from 1890. Mr. Hull is obliged tocut the knot by stating
boldly that there was no genuine industrial depression in
these years, and by the introduction of a veritable Olympus
of financial and political machine-deities. Industry was
all right : it was only the Baring crisis, the 1893 panic, the
Venezuela proclamation, Mr. Bryan's attack on the gold
standard and what not that upset things, — all those extra-
neous circumstances in fact whose inadequacy as the cause
of depression Mr. Hull elsewhere so strenuously asserts.
Again in explaining the collapse of 1907, the chief figures
upon which he appears to rely are those of the building-
permits issued in 1906-7 in fifty-seven cities, and those of
the unfilled orders on the books of the Steel Corporation.
As to the former, it appears that in the panic year 1907 the
decrease in the value of buildings contracted for was about
1 Average price of No. i Foundry New railway
(Philadelphia). mileage.
1878 . $17-67 •• .. 2,665
1879 . . . $21-72 .. .. 4,809
1880 . . $28-48 .. 6,711
1881 . . . $25-17 . . . . 9,846
1882 . . . $25-77 •• -. 11,569
(Stat. Abs. U.S.A., 1912, p. 748.) (Ibid., p. 281.)
FLUCTUATIONS IN COST. 57
12 per cent, on that of the preceding year.1 It is argued,
on the unfounded assumption that the price of all construc-
tion materials rose as rapidly as iron,2 that this represents
a decrease in the volume of building contracts of 34 per cent,
or even, on the further assumption that 58 per cent, of the
total permits in each year were taken out in the months
March-July, of 43 per cent. The inference, however, of the
paramount importance of costs in determining the volume
of construction does not appear conclusive. In the first
place it is not clear from the bare figures that the decrease
in construction was consequent on the very slight net rise
in prices between January and May and not on the con-
siderable fall May-December and the panic in October:
moreover, it is to be noted that the considerable advance
in prices took place in the last four months of 1906. Secondly,
even if we accept Mr. Hull's conclusions, the mere fact that
so large a rise in prices could produce a fall of only 12 per
cent, in the value of the work contracted for would seem
to indicate that the elasticity of the demand for structures
is not enormously great, and that a mere increase of costs
would not suffice to produce those very large declines in
receipts which depression appears to bring to most kinds of
constructional industry.3
1 Industrial Depressions, p. 183.
2 While the average price of iron rose 21-66 per cent. (op. cit.t
p. 184) the average price of all building materials only rose 8-39
per cent., and the average price of brick fell 27-45 per cent. (Report
of Bureau of Labour on Prices, 1910).
3 It may be observed further that according to the returns of
the Geological Survey, for 52 cities, quoted in Stat. Abs. U.S.A.,
1912, pp. 224-5, the value of building permits fell in 1907 not by
12 but by just under 5 per cent. Indeed the whole series of these
returns affords scant support to Mr. Hull.
Average Price of Pig-iron
(computed from Hull,
Value of Permits. op. cu., Ap. 2).
1905 . ..645 m S . . $15-86
$18-34 '
$22-40
$15*82
$16-13
$15-57
1906 . 679
1907 . 646
1908 . 566
1909 - 771
1910 . 726
58 STUDY OF INDUSTRIAL FLUCTUATION.
The figures of the Steel Corporation seem to be even less
conclusive. A comparison of the monthly figures of pig-
iron prices and the quarterly figures of unfilled orders on
the Corporation's books shows that the two move almost
invariably in the same sense — that a fall in prices synchron-
ises with a falling off in orders and conversely. This rela-
tion is sufficiently well indicated by the chart (III) .which
I have drawn on the basis of Mr. Hull's figures. If we
take for instance the expansion culminating in 1907, we
find that the first considerable rise in the price-index (from
137-3 to 162), in the last quarter of 1904, was accompanied
by an increase in the tonnage of contracts from 3,027,436
to 4,696,203 ; and that though a rise of price in the next
quarter coincided with a slackening of the rate of increase
of orders, an appreciable fall of price in the next from 179-6
to 174-3 was accompanied by an actual decrease in con-
tracts of 767,905 tons. For the third quarter of 1905 the
figures are inconclusive, for though they show a rise in
tonnage of 964,278, they do not indicate whether this was
entirely due to the fall in price from 168 to 161 in June to
July, or whether it continued during the steady rise from
161 to 166, between July and September. During the last
quarter, however, it is clear that the rise from 166 in Sep-
tember to 193 in December did not prevent a record in-
crease of orders from 5,865,377 to 7,605,086 tons. Price
broke in January and continued to fall till June, and with
it the volume of contracts : but the latter increased rapidly
in the third quarter of 1906, in spite of a rise in price from
186 to 194-3, and continued to expand, though more slowly,
in spite of the phenomenal rise from 206 in September to
272 in December. During the first half of 1907 prices
remained about stationary, with a tendency to fall ; so did
the volume of contracts. In the second half of the year
prices fell at an accelerating rate : so did contracts. In
spite of a continued price fall through the first half of 1908
It is, however, fair to point out that in 1909 and 1910 the price-
movements determining the character of the year took place in
the autumn.
FLUCTUATIONS IN COST. 59
contracts refused to revive till the third quarter, and con-
tinued to revive in the fourth quarter in spite of a new rise
in price.
I do not think this analysis can fairly be said to support
Mr. Hull's argument in its extreme form. In particular
it seems to suggest two objections. . In the first place if
costs are the sole consideration, why is the revival of in-
vestment so long delayed after the first break in prices ?
If the hard-headed and infallible business man in whom
we are asked to believe found it profitable to increase his
orders in November, 1906, when iron was at $22-93, why did
he not find it profitable to do the same when iron had relapsed
to the same figure in July, 1907 ? And here it must be
remarked once for all that it is no answer to say that on a
falling market he will naturally wait till prices are at a
minimum. This is obviously reasoning in a circle, for a
minimum, like virtue, is one of the things the beauty of
which is only realised when it has been left behind. Prices
only turn out to have been a minimum because business
men have as a matter of fact for some reason chosen that
moment at which to come in and buy, and have therefore
sent prices up.
This objection is supported by a reference to the British
coal and pig-iron industries.
While the price of the production of pig-
coal fell steadily iron did not begin to
from .... 1873- 6 revive till . . . 1876
Do. do. do. 1876-9 Do. do. the end of 1879
Do. do. do. 1884-6 Do. do. do. 1886
Do. do. do. 1891-3 Do. do. do. 1893
It is true that in 1876, and again in 1902-3, a spurt in
pig-iron production intervened in the downward course of
coal, but in neither case was the continued fall in coal prices
able to prevent the subsequent relapse in the activity of
iron : and while coal continued to fall throughout 1903-5,
the production of pig-iron did not take a fresh start till the
end of 1905. Again the price of coal began to decline from
November, 1907, but the production of pig-iron did not
60 STUDY OF INDUSTRIAL FLUCTUATION.
begin to revive till the autumn of 1909. The interval in
this case is shorter than on previous occasions, but even
allowing for the existence of twelve months' contracts at
high prices, it still appears sufficiently long to require ex-
planation. It appears then that though low costs con-
stitute a condition favourable to a revival of investment,
they are not of themselves sufficient to determine the
moment at which such a revival shall begin.
Secondly, even if high costs are ultimately the cause of the
withholding of new orders, is it not clear that they can pre-
vail for a considerable time without exercising any effective
deterrent influence upon investment ? Do not such events
as those of the latter half of 1906 force one to conclude that
people do as a matter of fact continue to buy heavily on a
rising market from choice and not merely from the necessity
of fulfilling contracts ? The study of shipbuilding seems to
confirm this view. We read, it is true, that in 1882 there
was " some tendency towards a slackening of orders/'1
owing not indeed exactly to high costs but to the impos-
sibility of promising delivery under fifteen months. Never-
theless there was a new increase in contracts in the last
quarter of that year and in the first quarter of 1883. Again
in the recent expansion costs were on January i, 1912,
15 per cent., and on January i, 1913, 2 40 per cent, above
those of January i, 1911, yet the tide of orders continued
an almost unbroken increase throughout 1911 and 1912.
We cannot do without the supposition that in all these cases
there was some force at work sufficient to overbear the
deterrent influence of high costs.3
1 Economist History of 1882.
2 Moss Circulars of those dates.
3 The evidence of the copper trade in 1899 is exceptionally inter-
esting. The exactions of the newly-formed Amalgamated Copper
Company, which started by controlling one-half of the American
(i.e., one-quarter of the world's) production, and was soon able to
dictate the world price, raised the price of standard to a maximum
of £jg ios., on May 16 ; and though it declined thence to ^67 in
December, the level was so high that consumption in 1899 fell 16
per cent, in England, 5 per cent, in France, and 8£ per cent, in Ger-
many. (First ten months.) (Ec., 1899, p. 1811.) (If electrical
enterprise was not correspondingly restricted, it was thanks to the
FLUCTUATIONS IN COST. 61
Indeed it seems that on certain occasions rising costs are
actually an attractive force in stimulating new construction.
Thus the first impulse to new shipbuilding in the autumn
of 1879 appears to have been largely fear of the effect upon
the iron market of the rising American demand for railway
material. Similarly the rise in the volume of orders for new
ships in the autumn of 1887 was precipitated largely by the
fact that speculative activity, which had for some months
been amusing itself with copper and tin, had just devoted
its attention with as little apparent reason to pig-iron.
Again the little spurt in mercantile building in 1894 seems
to have been due to the Admiralty demands at the beginning
of the year and the consequent fear of dear material.1
In the main, however, it is confidence in spite of, not fear
induced by, rising costs with which we have to reckon. And
this leads on to a further inquiry : are excessive costs, even
in the end, the determining factor in the ebb of construc-
tional activity ? Is it true, as Mr. Hull implies, that the
rate of construction in a progressive community can never
be excessive, or is it not clear that in a time of boom in any
substitution of aluminium, which though about twice the price
per ton was in some ways more suitable, once initial difficulties
were overcome.) But in Mr. Hull's own country consumption of
copper increased 44 per cent.
1 Cf. EC. Histories of those years. Parallels can be found in
consumptive industry : thus it was a shortage of South American
supplies, aggravated by a sudden speculative rush of French buyers,
which initiated the boom in the English woollen industry in 1886 ;
just as the later boom of 1899 was accentuated by two successive
shortages through drought in the Australian clip (cf. p. 27). Simi-
larly it 'seems to have been partly a fear of shortage instilled by
the strong Continental inquiry that impelled Indian importers
of cotton goods to make their large speculative purchases in 1889
and 1907 (cf. p. 107).
It is odd that Professor E. D. Jones, in his careful study of the
psychological aspect of fluctuations, neglects altogether the import-
ance of fear as an incentive to increased activity (Economic Crises,
chap. ix. pp. 180-219). The point is well emphasised by Mr. G.
Binney Dibblee (The Laws of Supply and Demand), who calls atten-
tion to it as a " reversal of Mill's law of value," e.g. (p. 140) : " A
further rise in price will not diminish demand ; it may very likely
increase it by drawing the attention of remoter buyers to the danger
of a threatened scarcity."
62 STUDY OF INDUSTRIAL FLUCTUATION.
industry there is a very real tendency to over-investment,
and that many undertakings are started which cannot
possibly pay their way ?
To this Mr. Hull might fairly reply : " Yes, but that is
only because the costs of construction were excessive ; if
the costs had remained at a normal level, there would have
been no difficulty in obtaining a remunerative return." But
this answer, it seems, would land him in a dilemma. Either
the excessive costs are paid by what he calls the " holder of
purse-strings," i.e., the ultimate owner of the instruments,,
or they are borne by the various intermediate contractors
and producers of materials. Such figures as those of the
Steel Corporation's contracts in the latter half of 1906 sug-
gest that the former alternative is often the true one. Thus
also Mr. Burton implies that the great depressions in Ameri-
can railroads were enhanced by the fact that the companies
had bought their materials in the years of highest prices
and had to compete with lines built when the prices were
lower.1 But to the extent that this is true the hard-headed
and infallible entrepreneur on whom Mr. Hull's case de-
pends, vanishes into thin air. On the other hand it seems
probable that in many cases the costs are borne by various
contractors and producers, because they have rendered
themselves liable for contracts at the low prices. Thus Mr.
Hull quotes from the report of the Empire Iron and Steel
Company, Catasanqua, Pa., to the effect that its average
profits for the years 1899-1907 were only $1-24 per ton,
although during this period the price of pig-iron was three
times advanced by an average amount of $1171 per ton.2
Similarly it appears that in the English boom of 1872 pig-
iron manufacturers were mostly on contract at the lower
prices and obtained little benefit from the great advance
1 Financial Crises, pp. 84 ff.
2 Mr. Hull makes it $12-70, but his arithmetic seems at fault.
Even so a reference to the figures (op. cit., p. 119) will show that
the argument is far from conclusive, as the average profits over the
nine years are naturally pulled down by the price depressions of
1901 and 1904-5. It' appears that in 1907, while the average price
of No. i Foundry at Philadelphia was $23-49, or 22-4 per cent.
FLUCTUATIONS IN COST. 63
in that year.1 But in this case the costs to the ultimate
entrepreneur have not been excessive, and the fact that
many enterprises have to be abandoned while others fail
to obtain a normal rate of return implies, on this hypothesis
as on the last, that there has been a real over-investment.
This conclusion is, I think, sufficiently supported by the
very facts to which Mr. Hull makes appeal : but it may be
briefly illustrated from other industries. Thus in America
in 1907 the price of copper (electrolytic) tumbled from 25!
cents in March till it reached nf cents in October ; but
" even with the price cut in half, the demand for copper
was sluggish, and it became clear that the diminishing con-
sumption of copper was the fundamental factor in the
situation."2
Again, while it is indeed possible that the initial slacken-
ing of shipbuilding orders in 1889-90, 1899-1900, and 1906
may have been due in part to increasing costs, the final
debacle of these three booms is clearly incapable of such
an explanation, since in each case costs had by this time
fallen. For instance, costs fell 25 per cent, between July,
1901, and July, 19033 : the beginning of the fall was an-
terior to the collapse in new contracts, which its continua-
tion was quite unable to arrest.4
above the average for the nine years, the net profit per ton was
$2-20 or 77-4 per cent, above the average for the nine years : this
does not look as if the company failed altogether to benefit by high
prices ! Yet, in spite of this unfortunate instance, the fact that
makers are in many cases squeezed between high costs and long
contracts at low prices is not, I think, in dispute.
1 Economist History of 1872.
2 Sprague, History of Crises under the National Banking System
(U.S.A. Monetary Commission publication), p. 242. Indeed the
cut in copper, preceding that in pig-iron, was on this occasion the
first indication of the relapse in investors' demand.
3 Circular of July 25, 1903.
4 Indeed, the most cursory glance at the history of shipping
and shipbuilding reveals the inadequacy of the cost theory as in
any sense a complete explanation of their movements. That low
costs are an advantage in construction and high costs a disadvan-
tage is sufficiently plain : but that even tramp owners, whose psy-
chological processes are admittedly obscure, should, however low
the level of cost, maintain for long the rate of construction attained
64 STUDY OF INDUSTRIAL FLUCTUATION.
While, however, a study of the facts suggests that those
writers are wrong who lay exclusive emphasis on variations
in costs as the cause of fluctuations in constructional in-
dustry, it does not follow 'that these variations do not play
an important if subordinate past. It seems probable that
they may be of importance in those tentative oscilla-
tions and false starts which precede the headlong rush of
capital into investment, as well as in those partial recover-
ies which are sometimes observable after the first serious
collapse.1 In particular it seems possible that the bicepha-
lous 2 nature of the American iron and steel expansion
of 1898-1903 was in part due to this cause : that is to say,
that over-investment in 1900 had not been so gross that a
reduction of costs was not able quickly to stimulate a new
flow.3 Again in 1907 the price of pig-iron fell from January
to April, but rose again in May to the highest point hitherto
reached.4 In England also the price broke first in Febru-
ary, but revived in April with a recrudescence in the demand
from Germany and Austria, stimulated by the lower prices,
but not destined to last much beyond the half-year.5 Fre-
quently, however, the influence of lowered costs only ex-
tends to the speculative market in half-finished products,
and does not communicate itself to the purchasers of finished
goods, who are more alive to the real situation.6
in times of boom is clearly neither possible nor desirable. In no
case is it more clear than in that of shipbuilding that there are other
factors at work more important than cost, — the fluctuations of
demand for the final product, and the tendency to rash and ill-
considered investment.
1 Cf. Burton (op. cit., p. 228), who attributes these recoveries to
absorption by wider markets.
2 Cf. p. 96 and Chart II.
3 Thus there is no reason to dispute Mr. Hull's statement (op.
cit., p. 173) that " a greatmany individuals, firms, and corporations
all over the country had plans and specifications of constructive
work which had been prepared for them in 1899 and 1900, and which
they had abandoned or postponed on account of the high prices
then asked," but which were brought out again in 1901.
4 Vide Chart II, and cf. Economist, 1907, pp. 823, 905.
5 Cf. Economist, 1907, pp. 187, 582, 1139.
6 Cf. Times (Financial sheet), July 7, 1913. There has been a
rise in pig-iron prices in New York, but " merchant steel and finished
FLUCTUATIONS IN COST. 65
Moreover, the facts do not preclude us from believing
that the immediate psychological stimulus to the cessation
of new enterprise after the last or longest wave of expansion
is sometimes afforded by high costs. It is quite possible
that the fact uppermost in the consciousness of American
business men in 1907 when they decided not to place new
orders was the high level of costs. This is possible especially
in the case of the initiators of railway enterprise, where the
conditions of demand are particularly hard to gauge.1 In
England in the same year the manufacturers of materials
do not seem to be clear as to whether the relapse in the
demand for their products is primarily due to high costs
or not. Thus in July Barrow simply reports a decline in
the demand for hematite, both on home and foreign account :
while " what Midland manufacturers say is that at present
prices of pig and coal their quotations cannot be reduced
to meet the ideas of consumers." Scottish malleable iron-
makers are chiefly engaged on export orders, for " home
consumers are not inclined to buy except in small quan-
tities at current quotations, which have been raised by dear
coal and pig-iron."
But in any case the argument suggests that on this and
similar occasions the high costs were but a kind of alarum
or pressure-gauge of unsound industrial conditions, which
were bound in any case to make themselves felt before long :
and that to blunt the industrial nerves with the narcotic
of low prices would be as injudicious as it is, according to
Dr. Sims Woodhead, to deaden with alcoholic liquor those
sensations of hunger and fatigue which are nature's warning
of the out-wear of the tissues.2
products have not followed the encouraging lead of the raw material.
There it is not a question of coke or ore, but of orders from rail-
roads."
1 An expert like Mr. J. J. Hill may give it as his opinion that
" Canada has about all the railways eight million people can sup-
port " (Times, July 9, 1913, p. 19), but owing to the stickiness of
prices those actually engaged in railway enterprise are not likely to
be very careful in their estimates of the relation of supply to demand :
that indeed is one of the special dangers of this form of activity.
2 Cambridge Magazine, January 25, 1913.
66 STUDY OF INDUSTRIAL FLUCTUATION.
§ 3. INVENTION.
There is, however, one form in particular of lowered costs
which seems to be of considerable importance both in in-
ducing immediate prosperity and in stimulating the over-
investment which sows the seeds of future depression ; and
that is a lowering of costs due to invention. It is indeed
true that the effect of an invention is by no means always
to raise at once the net receipts of those engaged in the
group of trades in which it is introduced. For in the first
place the depression among those firms which cannot or do
not introduce the new process may outweigh altogether
the prosperity of those firms which do. Thus the intro-
duction in the seventies of the Bessemer process of steel
manufacture led to a prolonged and acute depression in the
puddled iron trade.1 Secondly, the economies of the new
process may be anticipated rather than actual. Thus on
the same occasion those firms which took up steel manu-
facture seem at first to have done little better than the rest,
and the new process, according to the Statist history,
" aggravated a depressed trade by the severe losses arising
from the cost and uncertainties of a new manufacture."
Similarly the adoption of motor-power by the London
omnibus companies in 1905 led at first to a severe depression
in their profits. Thirdly, while the exchange value per
unit of output of the producers' services is raised, the
superiority of the new product in efficiency and durability
may be so great that their aggregate net receipts are con-,
siderably diminished. For instance, in the late seventies a
steel ship of 1,700 tons required 17 per cent, less in weight of
pig-iron 2 than an iron ship of the same dimensions, and was
capable of doing 7 per cent, more work.3 Similarly a few
1 " The rapid development of the production of steel in better
qualities and at lower prices has virtually superseded a large part
of the iron-making establishments in this country," says the Statist
history of 1877.
2 In such cases the depression in the industry furnishing the raw
material is of course still more acute : cf . the reports of the pig-
iron industry in these years.
8 Layton, Introduction to Study of Prices, p. 71. The situation
FLUCTUATIONS IN COST 67
years ago the electric supply companies were hard hit by
the invention of the metallic filament lamp, which at first
reduced the consumption of electricity by two- thirds.1
Again, recent inventions in the building trade, especially
the ferro-concrete process, seem to have reduced on the
balance the incomes of those employed in the trade.2
But the more usual effect of invention is to raise receipts
and increase investment in the trade in which it is intro-
duced.3 There can be no doubt, for instance, that the iron
and steel boom on the continent in the early eighties was
due very largely to the invention of the basic process of steel
manufacture, which first rendered available for industry
the vast iron-ore deposits of Luxembourg and Lorraine,4
seems to have been especially severe in the rail industry, not so
much for this reason (which would not be of importance till the
superior length of life of the steel rail began to make itself felt)
as because of the large quantities of scrap-iron rails thrown on the
American market in 1878. Cf. the figures of English production.
Iron and Steel Rails.
1878 . 633,733 tons.
1879 . .... 519,718 „
(Reports of Brit. Iron Trade Ass.). So also the increased use of
self-actors, and of cotton-warps, which freed a number of spindles
for the production of weft, contributed to the depression of the
wool trade in 1875.
1 EC., 1911, i. 209.
2 The carpentering branch has been especially hard hit, but
bricklayers and stonemasons are also affected (Dearie, Unemploy-
ment in the London Building Trade, pp. 46-8). Where, however,
a piece of work is contracted for at a fixed price, the cheapening of
construction enables more to be spent on panelling and fancy wood-
work, etc., and these branches of the trade therefore prosper (ibid.,
p. 53). In any case profits are of course at first increased.
3 For the view that the harmful effects of invention are rare,
and that as a rule not only the new trade prospers, but the old trade
has ample time to " arrange its own funeral," see article on " In-
ventions and Investment," by Mr. T. C. Elder, in Financial Review
of Reviews, July, 1912, p. 19.
4 " In Germany the process is being steadily carried on at Herde,
Ruhrort, Aix-la-Chapelle, Kaiserslautern and other places. The
Herde Company have erected new and special works for the Gil-
christ-Thomas process, which will be started in a month or so.
Messrs, de Wendel and Messrs. Sturt have also erected new works,
which will be started upon the process early in the spring " (TradQ
Circular apud EC, H, of 1880).
68 STUDY OF INDUSTRIAL FLUCTUATION.
and that the same discovery was a contributory if not a
very important cause of the contemporary boom in this
country.1 The importance of the inventions of electrical
power and oil-fuel in the transport and constructional booms
culminating in 1900, 1907 and 1913 is also, I think, unques-
tioned, but is more conveniently dealt with in detail in other
connections.2 The German iron and steel boom of 1910-11
may here, however, be directly referred in part to the
introduction of electric traction in the Lorraine district
and the application of electric power to the production of
steel direct from its low-grade ores.
1 The Economist quotes an interesting account of the surprise
and interest shown by iron makers at the first successful demonstra-
tions of the new process in 1879 ; for the bowels of England were by
no means innocent of basic ore. But while " the progress of the
Gilchrist-Thomas process for the dephosphorisation of iron has
been both steady and marked during 1880," it appears that it
" has made greater advances on the Continent than in our own
country," and the English steel boom seems to have been due in
the main to other causes.
2 Cf. especially Part II. chap. i. § 5. The importance of legal as
well as physical inventions in stimulating investment is discussed
in chap. i. § 2, sub fin.
The effect on various trades of fluctuations in the cost of credit
facilities is best discussed in connection with the whole question
of our monetary system (Part II. chap. iii. § 3).
B. PHENOMENA OF DEMAND.
CHAPTER IV.
MISCELLANEOUS CHANGES IN DEMAND.
§ i. FASHIONS, WARS, AND TARIFFS.
THE causes of fluctuation considered in the preceding
chapters are all such as might arise in any industry without
any change in the conditions of demand for its products.
In fact, however, we have seen that the course of events is
in many cases impossible to explain completely without
recourse to the hypothesis of changes in demand ; and we
pass on now to consider the nature and causes of such
changes.
Certain preliminary observations may be made. In the
first place it has already become abundantly clear that
fluctuations, due to conditions of supply, in a trade making
consumable goods, involve fluctuations in demand for the
services of trades making instrumental goods. All such
questions concerning the relations between consumptive
and constructional industry are, however, best deferred to
the next book.
Secondly, though a change in demand is usually required
to complete the explanation of a fluctuation, it is not neces-
sary that that change should itself be of the nature of a
fluctuation. Owing to the various temptations to over-
investment, a single permanent elevation of demand is
liable to produce an alternation of prosperity and depression.
Thirdly, if the growth of demand is of the nature not of a
discontinuous leap but of a continuous advance, it may
easily produce not merely one but a series of fluctuations.
69
70 STUDY OF INDUSTRIAL FLUCTUATION.
This proposition follows naturally from the principle of
the discontinuity of the process of investment. Further,
the rise in exchange value in reality caused by the catching
up of demand with permanent investment is liable to be
attributed to an exceptional increase of demand, and so to
accentuate each successive alternation of over-investment
and depression. These considerations seem to be especially
relevant where an initial stimulus to investment has been
given by invention. It may, for instance, I think be plaus-
ibly argued that in Germany the iron and steel boom cul-
minating in 1890 and the electrical boom culminating in
1907 were in large measure simply the harmonics of those
of 1882 and 1900 respectively.1 The two great inventions
of basic steel manufacture and electric traction were too
far-reaching to be satisfied, so to speak, with a single kill :
after they had broken up their quarry of hopes and expec-
tations, they were off again upon a fresh scent. Nor is this
suggestion refuted by the fact that in 1887, though not in
1904, the German revival was apparently precipitated by
that in the United States. For it is possible that the busi-
ness man is not on the look-out for the moment at which
demand can no longer be satisfied by existing structures,
and allows it to pass by. It requires some external stimu-
lus, perhaps of a wholly irrelevant kind, to attract his
attention to it, and by this time investment is in arrear of
demand. And postponement, by increasing the amount
of legitimate investment, accentuates the appearance of an
abnormal increase in demand, increases the temptation to
miscalculation and so aggravates the boom and the subse-
quent depression.
It is, however, clear that in many cases we have to deal
not merely with permanent increases, whether continuous
or discontinuous, but with actual fluctuations of demand :
1 It is for this, among other reasons, that Professor Pigou's
self-imposed limitations in his study of fluctuations (" specific
inventions are like enduring booms in Nature's bounty, and are
not therefore of first-rate importance for the study of fluctuations ")
(Wealth and Welfare, p. 447, note) seems to me particularly regret-
table.
MISCELLANEOUS CHANGES IN DEMAND. 71
and the nature and causes of such fluctuations seem to
deserve a detailed study.1
We may first consider fluctuations arising from changes
in fashion. It is, naturally enough, among the clothing
trades that such fluctuations are most frequent and im-
portant. Thus the Bradford worsted trade was hard hit
in 1881 by the change in fashion, initiated on the Continent,
from lustre fabrics to fine soft wool — a change that did not
touch the woollen districts, such as Huddersfield, which
continued very cheerful. The depression of the same
centre amid the general prosperity of 1898 was ascribed in
part to the development of the taste for bicycling among
the female sex, and the consequent demand for fewer and
shorter dresses ; 2 and in 1912 the Nottingham lace trade
was still suffering from the malignity of the hobble skirt.3
Among the luxury trades conspicuously to the fore in the
recent boom were those of the cinematograph (paying any-
thing up to 125 per cent.) and of the breeding of bull-dogs,
dependent largely on the prosperity of the grinders of
Sheffield.4 The cotton trade benefited by, among other
things, the discovery of a new thing in trousers by the cow-
boys of Mexico, and by the growing sophistication of the
Chinese, who have taken to wearing cotton shirts with cuffs
and collars.5 Conversely the collapse of the Indian demand
for cotton-goods in 1908 was aggravated by the curious
fact that the year was regarded for religious reasons as an
unlucky one, and a considerable number of marriages, with
the large gifts of cotton cloth which the ceremony involves,
postponed until a more propitious season.6
In the main, however, the fluctuations in the demand for
1 I propose to omit altogether discussion of seasonal fluctuations
in demand, as irrelevant to the purpose of this inquiry.
2 EC. H. of 1898.
3 Cf. Daily Mail, Dec. 10, 1912.
4 Id., Dec. 7.
5 Id., Dec. 4. The same cause redounded, I am told, to the
benefit of the sofa-stuffing and allied industries, owing to the importa-
tion from China of large quantities of discarded pig-tails at what
may be called scrap-iron prices.
6 R. of I. T., 1908-^9, p. 68.
72 STUDY OF INDUSTRIAL FLUCTUATION.
articles of luxury seem to be merely the reflex of fluctua-
tions in the profits of constructional and other industry.1
And it will not, I think, be disputed that M. Issaieve is right
in concluding 2 that in view of the fact that France, the
great producer of luxuries, is much less severely affected
by fluctuations than the producers of the great industrial
staples, and that even in France only about ^ J-Q- of the total
volume of normal production is subject to the influences of
fashion, the fluctuations of luxurious industry are not on
the whole of first-rate importance to our main study.
Of more widespread importance are fluctuations in demand
due to the alternations of war and peace and to the vagaries
of foreign tariffs. With regard to the former, the armament
industries occur naturally to the mind ; but here indeed
it seems that the conditions of armed peace bear so close
a resemblance to those of war that we have to deal rather
with an increasing expansion than with a series of fluctua-
tions of demand. In so far, however, as the advance is
discontinuous and spasmodic, the tendency to over-invest-
ment and a consequent slump is present : though indeed
in this respect the indirect effects of Admiralty policy
on the course of ordinary commercial enterprise3 seem to
be more important than the direct effect on those branches
of industry specialising in armament work, who appear
unfortunately to find the growing demands of Governments
impossible to over-estimate.4
The effect of fluctuations in human hatred upon the course
1 The American crisis of 1907 had a serious effect on the export
of tulles from Calais, and threw 70 per cent, of the silk-workers of
St. Etienne out of employment (Issaieve, Journal d'Economie
Politique, 1893, pp. 666 ff.). The silk industry of Japan was also
severely affected, with the result that Japan had to curtail its pur-
chases of Lancashire cotton-goods (Ec. Hist, of 1908). The South
African diamond industry was also hard hit ; the value of the
United States imports of precious stones which was $43,375,000
in December, 1906, sank to $207,000 in December, 1907.
2 Loc. cit. 3 Cf. pp. 6 1 and 159.
4 Of important fluctuations due to a non-military Government
demand, the best instance is perhaps the French constructional
boom of the early '80 's, due in no small measure to the Freycinet
scheme of public works (JLescure, op. cit., p. 149).
MISCELLANEOUS CHANGES IN DEMAND. 73
of fluctuation in the shipbuilding and allied industries
through the medium of the demand for transport is perhaps
more important. Thus the tremendous shipping, coal
and shipbuilding booms in 1900 and the accompanying
over-investment were in large measure due to the huge
demand for tonnage for the conveyance of troops, stores,
etc., to South Africa. Similarly in 1898 Atlantic and in
1912 Near Eastern freights were raised respectively by the
Spanish-American and by the Tripolitan and Balkan wars.
Certain of the textile industries are markedly subject to
similar influences. Thus in 1870-1 the woollen trade, and
in 1900 the khaki manufacturers of Huddersfield, profited
by large orders for army clothing, though on both occasions
the worsted branch was severely injured by the closure of
its normal markets. Moreover, on the first occasion at least
the overflow on the restoration of peace of the dammed-up
arrears of ordinary demand was treated as though it were a
normal stream, and there was a rush of new capital into
the trade.
The jute trade also has long had the reputation of being
a darling of the war-god, and both the South African and
the Japanese Wars brought a flood of orders to Dundee.
But it is worth noting that a sudden military demand of
this nature is not only likely to be short-lived, but also to
create an excitement and an impulse to investment out of
all proportion even to its temporary magnitude. Thus
there is no doubt that in 1904 the impression prevailed that
the jute trade was abnormally prosperous, and extensions
of productive capacity were made from which the industry
suffered for years : yet a reference to the figures will show
that our exports both of yarns and piece-goods underwent
a progressive fall in 1904-5. Similarly with the outbreak
of war there was a hardening of Far Eastern freights and a
flow of tonnage to the East to carry coal and to take the
place of the Japanese merchant ships devoted to warlike
purposes;1 but the results proved to be disappointing,2
1 Moss Circular, Feb. 20, 1904. 2 Id., Sept. 30, 1904.
74 STUDY OF INDUSTRIAL FLUCTUATION.
and by the autumn of 1904 the level of freights was very
depressed. Even to his own minions Mars is often less
generous than he seems.1
The Yorkshire woollen and worsted industry is perhaps
the best example of a trade in which the course of fluctua-
tion is much affected by variations in foreign fiscal policy.
From 1890 to 1898 the export branch of the trade was kept
in a continual state of upheaval by the distracting antics
of the American tariff. A rush of imports before the Me-
Kinley Act in 1890 followed by a large falling off in 1891,
an excessive reaction in the opposite direction in 1892,
another rush in 1895 after the Wilson reduction followed
by a renewed slump — another in anticipation of the Dingley
Act in 1897 with the same sequel — all these led both to rapid
oscillations in the actual volume of business done, and also,
owing to the same liability as in the case of war to mistake
abnormal deferred or anticipatory imports for a genuine
increase in consumers' demand, to a marked extravagance
in investment policy.2 In 1914 the trade was expecting
a rich harvest from the recent tariff reductions.
It is scarcely necessary to add that industrial no less than
international disputes may dry up important sources of
demand. Detailed illustration is superfluous, but we may
instance the effect of the engineering disputes of 1897 and
1908 and of the transport strike of 1911 on the woollen
trade.3
1 A discussion of the broader effects of war on industry in general
will be found in Book II. chap. ii. § 2.
2 Similarly the collapse in 1907 of the Indian " Swadeshi " or
Protectionist movement, whose activity had contributed to the
decline in the volume of cotton imports in the previous year, swelled
the huge speculative imports and investment of 1907. Cf. R. of
I. T., 1906-7, p. 14, and 1907-8, p. 24.
3 The woollen trade was afflicted in 1911 by an extraordinary
combination of miscellaneous disturbances of demand (excessive
heat, transport strikes, European war-scares, Chinese revolution,
hobble-skirts, Japanese tariffs and Turco-Italian War), some though
not all of which were removed in 1912.
CHAPTER V.
CROP VOLUMES, TRANSPORT, AND CONSTRUCTION.
§ i. DIRECT INFLUENCE OF CROP VOLUME.
OF more persistent and widespread importance are
fluctuations in demand arising from variations in the pur-
chasing-power of those whose expenditure upon the pro-
ducts of the trade in question is directly or indirectly
dependent upon the varying bounty of nature. Indeed it is
well known that the most brilliant of all writers upon our
subject, Stanley Jevons, thought to find in such variations
the main origin of the whole phenomenon of industrial
fluctuation.1 The difficulties which are raised by this
hypothesis of the dependence of industry in general upon
crop conditions are well worthy of close examination : they
do not, however, appear to affect the question of the depen-
dence of single industries or groups of industries, and may
therefore be postponed to the next book.
We may consider first such fluctuations as arise solely
from variations in the volume of a particular agricultural
product, irrespective of the effect of such changes in volume
upon the purchasing-power in the hands either of producers
or consumers.
Generally speaking the effect of an increased crop volume
is to increase the demand both for land and sea transport,
and so indirectly for the products of the iron and steel trades.
With regard to land transport, for instance, the general
view 2 is that the volume of the United States wheat crop
1 Investigations in Currency and Finance.
2 Cf. Piatt Andrew, " Influence of Crops on American Business,"
Q. J. of EC., Vol. XX. p. 343 ; and Mr. Finley, President of the
Southern Railroad, in Times Commercial Review, 1912, p. 26
75
76 STUDY OF INDUSTRIAL FLUCTUATION.
has an important effect upon the gross receipts of the rail-
way companies and upon their orders for new rolling-stock
and so forth. This contention is indeed not easy to estab-
lish statistically, and it is hotly disputed by Mr. Hull, who
points out that in the year ending June 30, 1899 (a year,
moreover, of record wheat-crop), agricultural products,
according to the report of the Interstate Commerce Com-
mission, formed only 11-33 per cent, of the total tonnage
carried by the railroads. This evidence, however, seems to
me very far from conclusive,1 and on the whole I see no
reason to doubt the correctness of the primd facie view.
The effect upon United States railway receipts of variations
1 In the first place 11-33 Per cent. seems a not altogether negligible
proportion, and indicates a volume of traffic any large change in
which would have a perceptible influence upon the total. Secondly
the figures quoted are only those for traffic reported as originating
on the road making the return. But it appears that such traffic
forms a less proportion of the total tonnage carried by each line
in the case of agricultural than of other produce. Thus in 1909-10,
for example, the figures are as follows (Stat. Abs. U.S.A., 1912,
p. 298) :—
Received from other
Originating on Road. Roads and Carriers.
Products of agriculture 78,736,507 tons 81,665,186 tons
Products of mines . 544,604,373 ,, 397,461,474 ,,
Total products . . 968,464,009 776,860,819
Moreover, in the absence of ton-mileage statistics for the different
kinds of product, these figures seem to afford a fair indication that
the average ton of agricultural produce is carried further and is
of greater importance to the railways than the average ton of pro-
duce in general. Thirdly, a very large proportion (51-47 per cent,
in the year 1898-9) of the traffic under consideration consists of
the products of mines — coal, ore, sandstone, etc. — for which the
average freight-charge per ton-mile is considerably lower than that
for agricultural produce, so that again the mere tonnage statistics
do not adequately represent the relative importance to the rail-
ways. Fourthly, coal, etc., are carried in open and agricultural
products in covered cars, so that the indirect effect upon the iron
industry of an increased traffic in the latter is greater. Fifthly,
Mr. Hull's figures are for a year in the course of which the indus-
trial revival had made great headway, and afford no proof that at
its inception the relative importance of the crops was not greater,
still less that the increased volume of mineral and manufacturing
traffic was not itself originated in part at least by an increased
volume of crops.
CROP VOLUMES, TRANSPORT, CONSTRUCTION. 77
in the cotton crop (which is carried on the average a much
shorter distance) and in the corn crop l (of which a very large
proportion is consumed on the farm) seems, however, to be
considerably less important than that of variations in the
wheat crop.
An important exception to the general rule is furnished
by India. Thus 1877 was a famine year in India, but our
exports of iron and steel thither increased from 135,725
tons to 153,300 tons owing to an increased demand for rail-
way material. Similarly in the famine years of the middle
'go's, and again in 1902, the imports of iron and steel
exhibited a considerable increase.2 Again throughout the
years of dearth 1907-8 there was a good demand from India
for locomotives and other products of the English engineer-
ing trades.3 The explanation is of course simple : in normal
years a large part of the Indian grain-crops is consumed
locally ; but famine, especially the localised famine which
is typical of India,4 necessitates increased expenditure by
Government in providing for a greatly increased transport
of food-stuffs from district to district. In 1908, for example,
the existing girders, etc., were found unable to bear the
strain, and the Government was obliged at heavy cost to
anticipate largely its renewal programme for
1 On the other hand, an important share in causing the depres-
sion of the U.S.A. railways in 1913 is attributed by the Economist
(May 30, 1914) to the failure of the corn crop.
2 Calendar Year.
Wheat Crop.
Rice Crop.
Iron and Steel
Imports from U.K.
1895
134-7 m- cwts.
497-9 m. cwts.
314,306 tons
1896
106-7
4I5-3 ,
464,377
1897
.
107-1
275-6 ,
494,401
1898
.
143-9
498-3 ,
388,035
1901
141-9
4I3-5 ,
3I7.434
1902
•
I2I-8
384-3 ,
374.906
3 EC., 1909, i. p. 1,323 ; and Histories of 1907 and 1908.
4 Cf. Atkinson, in Stat. Jour., 1897, p. 98.
5 R. of I. T., 1908-9, pp. 25, 26,
78 STUDY OF INDUSTRIAL FLUCTUATION.
The normal effect of a large crop volume in a country
which grows for export is to raise the level of ocean freights
and increase the orders for new ships. Thus the course of
the English shipbuilding trade during the last forty years
seems to have been largely determined by the fluctuations
of the United States wheat crop and the consequent move-
ments of North Atlantic freights.1 It should be noted,
however, that a large United States crop only exercises
its full influence upon freights if accompanied by a shortage
in the less distant sources of supply. Thus the French,
Russian and British shortages of 1873 and the consequent
necessity of an increased export from the abundant crops
of the United States sent up Atlantic freights 2 and seems
to have been largely responsible for the large shipbuilding
output of 1874. Again in 1879 the failure of the French
and English crops coinciding with a continued rise in Ameri-
can production seems to have been largely instrumental in
starting the great shipbuilding outbreak of that and the
following years. But the big United States crops of 1882
and 1884 were accompanied by plenty in Europe, and those
of 1885 were bad. With a bad home harvest and good
United States and Argentine crops in 1886 there was a
slight increase in homeward Atlantic freights in that year
and in tonnage under construction in the last quarter of
1886 and the first of i887.3 But the great impulse to ship-
building in 1888 came from other sources ; homeward rates
from the United States, though they rose slightly by reper-
cussion in 1888, did not advance greatly till the big harvest
of 1889, accompanied by a Russian shortage, led to an
1 The Board of Trade calculate that in the years 1884-1903
50 per cent, of the British tonnage entered with cargo at British
ports was engaged in the North Atlantic trade, and weight their
index-number of freights accordingly (Cd. 2337, C1- I9°4» P- 2)-
2 Average wheat freights from New York to Liverpool rose from
7fd per bushel in 1872 to lo-fcd. in 1873 (Stat. Abs. U.S.A., 1912,
P- 315).
3 It should be noted that a big wheat output in the U.S.A. has
its effect on the tonnage under construction in the last quarter,
and one in the Argentine and in Russia in the first quarter of the
calendar year,
CROP VOLUMES, TRANSPORT, CONSTRUCTION. 79
increased carriage of grain across the Atlantic. But owners
were by now on their guard, and there was no further
increase in the tonnage under construction.
The boom in inward freights was checked by the good
Western European and indifferent American harvests of
1890 ; but though the American crop of 1890 was small, a
considerable proportion of it was exported, and though
there was no rise in freights the activity of ship-builders
seems to have been temporarily stimulated by this fact in
the last quarter of 1890. l With more reason they were
stimulated again one year later, when a phenomenal crop
in America was accompanied by failures in Russia and
France and by a poor out-turn in the United Kingdom. A
poor U.S.A. crop in 1893 and European plenty in 1894-5
helped to keep freights at a low level, but in the cereal-year
1897-8 high rates in the North Atlantic and Eastern trades
consequent on crop failures in Europe and large harvests in
North America and India gave the impulse to the new ship-
building boom ; nor were the efforts of India and the big
European crops of 1898 sufficient to prevent a large demand
being made again on the enormous United States production
of that year. In 1899 with big European crops there was a
relapse in inward freights, reflected in a prolonged decline
of the tonnage figures ; and the great boom of 1900 must
be referred to other causes. So great was the resultant
over-supply of tonnage that even the large exports of grain
from the U.S.A. consequent on their large crop and a Rus-
sian and Indian shortage could not prevent a drop of 74
per cent, in the grain rates from New York to Liverpool in
1901.
In the new century the situation has been changed by the
decreasing proportion of the United States wheat crop
used for export. But the wave of new contracts in the
latter part of 1905 was, I think, precipitated by the large
increase, after three successive years of shrinkage, in the
United States wheat-crop, which gave rise to the very
natural expectation of a large export. This expectation,
1 See last note,
8o STUDY OF INDUSTRIAL FLUCTUATION.
however, was not fulfilled, for only 7-99 per cent, of the
crop was exported, as compared with 18-92 per cent, of the
preceding crop and 41-36 per cent, of that of 1901. * Again
the temporary revival in the first quarter of 1907 z must be
connected with the fact that of the record American wheat-
crop of 1906 as much as 19-95 per cent, was exported — our
own imports thence rising from 14-9 to 36-6 million cwts.
The American exports of grain diminished steadily in
1909-10, and their decreasing importance to shipbuilding
is shown by the fact that in 1909-12 the most significant
increases in new orders have been not as of old in the last
but in the first quarters of the calendar years, i.e., at the
time of the greatest activity in the Eastern trades. Home-
ward freights rose at last in 1911, but mainly by repercus-
sion, for the crop was a poor one, and there was no large
increase in exports ; and though the exportable surplus of
the big crop of 1912 has been considerably greater, the fact
that the hypnotic influence of America, which seems so often
to have blinded the vision of tramp-owners as of other
British business men, was absent during the earlier stages
of the present revival, may, I think, be regarded as a part
explanation of the significant fact already noted that the
main increase in tonnage contracted for in the last year or
two has not been in tramp vessels at all. 3 Whether tramp-
owners will preserve the same moderation in face of the
large American crops of 1913 remains to be seen. The
evidence of 1890, 1891 and 1907 points one way, that of 1889
and 1901 another, but the opinion may be hazarded that
the spell has been broken for ever, and that what has been in
the past the most fruitful source of the violent oscillations
of the shipbuilding industry, has at last by bitter experience
been eradicated.4
1 Cf. Stat. Abs. U.S.A., 1912, p. 563.
2 It should be noted that tramp-owners had learnt their lesson,
and that the flood of new orders was delayed till the large export-
able surplus was no longer a matter of guesswork.
3 Cf. note 2, p. 28.
4 I purposely leave this passage as it was written in the summer
pf 1913, but the opinion here expressed seems to have been justi-
CROP VOLUMES, TRANSPORT, CONSTRUCTION. 81
It must not be supposed however that even in the old
century the trade was insensitive to every agricultural in-
fluence but the American wheat-crop. Thus the increased
cereal exports of Russia in 1888 hardened freights not only
in the Black Sea but also by diversion of tonnage in the
Baltic and Mediterranean trades, and contributed materi-
ally to the shipbuilding boom. Similarly the revivals of
contracts in the first quarters of 1891 and 1894 and in 1910
may be attributed largely to big Russian cereal exports.1
Again the slight revival of contracts in early 1904 may be
referred to the hardening in homeward freights from the
Far East and the River Plate, both of which had two good
crop seasons — a hardening accentuated in the former case
by the record Australian crop of December 1903. 2 Again
the great boom which began in the third quarter of 1909
fied. Our large imports of the U.S.A. wheat crop of 1913 were
unable to stimulate even the freight market (depressed by the poor
exports of corn and cotton), still less the tonnage under construc-
tion.
1 Compare the course of Odessa freights. It has been noted
that a big wheat output in the U.S.A. has its effect on the tonnage
under construction in the last quarter, and one in the Argentine
and in Russia in the first quarter of the calendar year. This con-
sideration furnishes an additional explanation of the greater in-
fluence upon shipbuilding enterprise of an American than of a
Russian increase in wheat production. The Russian crop being
normally shipped later than the American, tramp-owners hang
about in European waters well into the following calendar year
before the shortage is revealed, and it is then too late to be of much
use with the American shipments ; while those disappointed with
the results of the American trade in the autumn have a better
chance of being in time for large Russian shipments in the spring.
The period of transference alluded to in chap. i. § i exerts greater
friction in one direction than in the other.
The expected gain to shipowners from the big Russian crops
of 1913 has failed to materialise, owing to good harvests in the
rest of Europe, and the holding up of the Russian supplies with
the aid of a Government loan (Times History, Jan. 16, 1914).
2 The Indian and Australian shortages in 1896-7 had the same
paradoxical effect upon sea as the former had upon land trans-
port, for they actually turned those countries into considerable
importers, raised outward freights to the Far East, and so were
mainly responsible for the increase in shipbuilding contracts in
the last quarter of 1896 and the first half of 1897.
82 STUDY OF INDUSTRIAL FLUCTUATION.
and progressed almost without interruption for more than
three years seems to have been due initially (on the side of
demand) mainly to agricultural influences in the Far East,
the unexpectedly large proportion exported of the moderate
Indian wheat crop of March 1909, the cereal prosperity of
India and Australia in the two following harvests, and the
springing-up of new trades from the Pacific,1 of which the
most important was the boom in Soya beans from Manchuria
for cattle-feeding purposes.2 It should be observed that
owing to the length of the period of gestation the impact
of several successive independent ripples of demand fre-
quently exercises an effect upon the output of new ships
which resembles that of one continuous and concentrated
breaker.
Any attempt to estimate the relative importance to ship-
ping and shipbuilding of the agricultural and the non-
agricultural demand is necessarily difficult for an outsider.
In 1888-9 high freights for timber from Canada and manu-
factured goods to the Argentine,3 in 1906 the great boom
in European emigration to the States and the carriage of
steel, cement, etc., for the rebuilding of San Francisco, in
1910 the industrial development of Japan,4 in 1912 the
demand for oil-transport5 was a valuable support to agri-
1 Moss Circular, Jan. i, 1910.
2 Cf. Economist, 1909, ii. pp. 691 and 1,144, an<^ I9IO» i- P- 555-
It should be noted that the import of this product was largely stimu-
lated by the very shortage of the American cotton crop which
depressed North Atlantic freights : a good illustration of the truth
that the fulfilment of the law of inter-local compensation does not
necessarily make for the stability of shipbuilding.
3 The rise was sufficient to prevent the usual irruption of tonnage
into the Russian grain trade (cf. note i). Hence freights from
Odessa rose from 11-9$ in 1887 to 15-4! in 1888, or 130-6 per cent,
as compared with a maximum rise of 127-1 per cent, in freights
from New York to Liverpool during the twenty years 1884-1903
— namely, from 2-47^. in 1890 to 3-14^. in 1891. Cf. also Times
Shipping Supplement, Dec. 13, 1912, p. 14.
4 The requirements of the Japanese and Chinese trades had
been more than met for some years by the flood of Japanese mer-
chant shipping and Russian captives available after the war (John
White, Review of 1907).
6 Cf. Part II. chap. i. § 5.
CROP VOLUMES, TRANSPORT, CONSTRUCTION. 83
cultural influences.1 As a rule, however, the non-agricul-
tural seems merely to prolong and intensify the movement
initiated by the agricultural demand.2 I do not think we
shall be convicted of error in assigning directly to agricul-
tural causes a predominant influence over one of the most
important and most fluctuating of British trades. The
importance of this conclusion, in view of the unwillingness
of most recent writers on fluctuation to admit the signifi-
cance of agricultural factors, must be the excuse for the
tedious length of this discussion.
The great encasing or enveloping trade must be treated
together with the carrying trades in connection with the
present argument. It would indeed seem that jute, " the
world's wrapper," which is used for the envelopment of
everything from cement and cotton to submarine cables
and the legs of horses,3 ought to be of all commodities the
most beneficially affected by the soothing influence of the
law of compensation. And indeed a reference to the figures
of exports to the chief markets will show that in 1897 and
1902, for instance, the evil effects of the combined failure
of the Argentine wheat and maize crops were in part —
though only in part — compensated by the increased takings
of Canada and the United States respectively, while in
1903-4 both Canada and the Argentine did their best to
make up for the shortcomings of the United States. But
on the whole the Dundee trade shows no great tendency to
emancipation from the influence of the United States.4 Our
1 Cf. also p. 158, n. 2.
2 Again a study of the" Board of Trade Indices suggests that
outward rates as a whole vary within somewhat narrower limits
than inward ; for a large part of our exports consists in manu-
factured goods which are carried by conference lines, in which the
fluctuations in volume are relatively small, and for which the com-
petition of tramps is unavailing, while the other chief item, coal,
is one for which the demand, though elastic, is relatively invariable,
and the freights for which therefore tend often to depend largely
on the supply of tonnage called into existence by the requirements
of the import trade.
3 Cf. Times Textile Supplement, June, 1913.
1 There has been a trans-cyclical diminution in the importance
of the Argentine market, due to the competition of the Indian mills.
84 STUDY OF INDUSTRIAL FLUCTUATION.
exports of piece-goods thither, though smaller than they
were twenty years ago, still form about the same propor-
tion, i.e., nearly 50 per cent., of the whole. Thus though
America prevented the trade sharing in £he general home
depression of 1901-2 it helped also to prevent it sharing in
the recovery of 1910-11. Nor could the " tolerably large
business done " in 1908 by the home- trade, the Continent,
Canada, the West Coast of South America and other mar-
kets, and the very large demand of the Argentine prevent
the trade from complaining that " there is no new market
to compensate for the loss of the American market." The
export branch of the spinning trade is, however, chiefly
affected by the agricultural conditions not of the United
States but of Brazil, which takes about 40 per cent, of our
yarn exports. A reference to Table V. will show that the
volume of these exports follows very closely the movements
of the coffee-crop discussed in chap. i. § i.1
Finally, the volume of crops exercises in certain cases an
appreciable influence on the coal trade : for by providing
good homeward cargoes, a large crop lowers the outward
freights which must be paid for coal. The extent to which
the trade benefits depends upon the elasticity of demand for
our coal in the country supplying the large crop. The con-
trast between South America and India in this respect is
pointed out by Mr. D. A. Thomas, who remarks that our
exports to the Argentine and Uruguay were larger in 1897
than in 1894 or 1900, though we only derived thence 1-23
per cent, of our wheat en-food supply in that year, as against
15-22 per cent, and 20-80 per cent, in the others : whereas
in the case of India our coal exports show a clear tendency
1 It should be observed that the holding up of the 1907 supplies
did not affect the demand for jute, for it was a prominent feature
of the scheme that the seven million odd bales of valorised coffee
were kept not in Brazil but in the ports of Europe and the United
States. In 1910 the demand furnished by the big current crop
was apparently increased by a considerable exportation of stocks,
for while the world's stocks decreased between December i, 1909,
and December i, 1910, by some 2-8 million bags, the valorisation
trustees only disposed of about 0-5 million.
CROP VOLUMES, TRANSPORT, CONSTRUCTION. 85
to decrease with a decrease in our wheat imports.1 In
respect of elasticity the Russian demand appears to be
similar to the Indian, 2 and the North American, which is at
best inconsiderable, to the Argentine.
§ 2. PSYCHOLOGICAL INFLUENCE OF CROP
VOLUMES.
The volume of crops seems also in certain cases to have
an important indirect effect upon the demand for the pro-
ducts of the constructional industries. Variations in crop
volume, whether or not they alter the volume of resources
available for expenditure upon such products, may increase
or diminish the psychological impulse to the employment
of resources in this direction.
This consideration is of especial importance in the case
of new and rapidly developing countries. In the Argentine
and in Canada particularly, the most important effect of
exceptionally good harvests upon our constructional indus-
try has been through their operation as a psychological
stimulus to English investors — a kind of guarantee or
deposit on the hire-purchase system — a substance of things
1 His figures are (Stat. Jour., 1903, p. 456) : —
Imports of Wheat, etc. Exports of Coal.
1895 . . 440 thousand tons 805 thousand tons
1896 . . . 106 ,, ,, 528
1897 27 „ 195
1898 . 477 „ ,, 331
1899 . 410 ,, „ 433
1900 . . . [Insignificant] 100
There seems, however, to have been an important exception in
1908, when coal freights were low owing to the scarcity of outward
cargoes of manufactured goods due to the failure of the Indian
harvests and demand, the return cargo being, as so often, hypo-
thetical. Cf. Review of Indian Trade, 1908-9, p. 30, and the follow-
ing figures of Indian coal imports : —
1906-7 ..... 257,203 tons
1907-8 308,348 ,,
1908-9 455»8°6 »
2 Thus our large coal exports in 1890-1 (cf. note 3, p. 52) were
conditional on low freights — the large Russian cereal exports of
those years being not apparently accompanied by prosperity to
producers.
86 STUDY OF INDUSTRIAL FLUCTUATION.
hoped for and an evidence of things most emphatically not
seen.
A glance at the curve of our iron and steel exports to the
Argentine shows that maxima were reached in 1884, 1889,
1896, 1901, 1906 and 1909. The first of these booms seems
to have begun in 1883 and to have been mainly a matter of
railway construction.1 The most noticeable feature of this
expansion is that it seems to have both begun and ended
later than the similar movement in other countries. It will
be remembered that these were years of big crops through-
out the world, and the Argentine was no exception.2 Now
while the very universality of the big crops prevented them,
as is argued elsewhere, from stimulating American construc-
tion3 or British shipbuilding,4 in the Argentine, where the
crops were important less for their own value than as an
evidence of the capabilities of the country, no such hindrance
arose ; and that constructional boom whose absence from
America Mr. Jevons so successfully explains 3 broke out,
in the midst of constructional depression elsewhere, in the
Argentine.
Again in the tremendous boom of 1887-9 ^ seems to have
been partly the successful wheat crops of 1 886-8 (exported
in 1887-9) tnat stimulated confidence in the country's pro-
ductivity. A connection of some kind will also be sus-
pected between the enormously increased crop of 1893 and
the import boom of 1895-6, the record crops of 1898-9 and
the minor boom of 1900-1, the very successful harvests of
1902-4 and the boom of 1904-6,5 the failure of 1905 and
1 It was at the time that what is now the great Buenos Ayres and
Pacific Railway was started, and of the country's increase in total
mileage from 150 miles in 1874 to 2,290 in 1884, the greater part
was constructed in the last two years. (Cf. W. A. Hirst, Argentina,
pp. 181, 184.)
2 I have no figures for production before 1887, but it appears
that the exports of wheat rose from 700 quarters in 1881 and 8,000
in 1882 to 280,000 in 1883 and 500,000 in 1884.
3 Cf. p. 153, note i. 4 Cf. p. 78.
5 The railway system increased from 10,285 miles in 1900 to
15,476 in 1908 (Hirst, loc. cit.) ; most of the increase seems to have
been in 1904-6.
CROP VOLUMES, TRANSPORT, CONSTRUCTION. 87
the collapse of 1907, the record crop of 1907 and the boom
of 1909, the failures of 1908-10 and the heavy fall of 1910-2
and between the large crops of 1911-2 and the large increase
in imports in last and (it may be anticipated) the current
year. Now it will be noted that in general the curve of
iron and steel imports tends to lag behind that of wheat
exports by one year (and therefore that of wheat production
by two years). This curious habit would afford food for
reflection to an investigator of the rapidity of working of
the mechanism by which the equation of international
indebtedness is solved ; 1 for our present purpose the
variations from the habit are more significant. In the first
two periods, those of 1882-4 and 1887-9, when the " psy-
chological " element of the boom is likely to have been of
the greatest relative importance, the import curve synchron-
ises with that of wheat exports ; in 1894-6, when the
English investor was still smarting from the memory of
the Baring disaster, and in no mind to listen to the Siren
voices of the wheat-fields, not only was the import boom
much less pronounced, but the period of lag was increased
to two years. The inference that the influence of Argentine
harvests upon the British constructional trades through
the medium of the psychology of the British investor is of
importance as well as their influence through the medium
of the purcjiasing-power of the Argentine farmer is to this
extent confirmed.
1 The country appears to accumulate for a time its claims upon
the foreigner. This is more marked when the increased resources
are ultimately used in railway enterprise than when they are used
mainly for the purchase of instruments, roofing material, etc., for
a period of hoarding is not unlikely to precede the inception of the
new design, and another interval elapses between the latter and
the actual demand for rails. This seems to have been the case in
the middle '90*3 ; the railway mileage increased between 1890
and 1899 from 5,745 to 10,285 miles (Hirst, loc. cit.), and the large
rise in English rail output in 1896 seems to have been largely designed
for the Argentine.
Rail Output.
1894 ...... 598,530 tons
1895 604,338 „
1896 . .... 817,476 „
(Iron Trade Assocn. Reports).
88 STUDY OF INDUSTRIAL FLUCTUATION.
A similar influence must be attributed to the boom in
recent years in the Canadian crops. The results of the
colossal borrowings of Canada on the security of her wheat-
fields are discussed more fully in Part II. chap. ii. ; here it will
suffice to note with regard to Canada in 1910-2 as with
regard to the Argentine in 1887-9 that the more rapid rise
of outward than of homeward freights1 suggests that there
are certain countries the emergence of which as an important
source of our wheat supply has a greater effect upon our
constructional industries than the actual magnitude of their
contribution would seem to warrant.
1 Between 1910 and 1912 average homeward freights from
Canada rose from 345. 6d. to 535. 6d. per ton, or 55 per cent., while
average outward freights rose from 325. to 525. 6d. or 64 per cent.
(Memo, of Committee of Royal Colonial Institute, 1913).
CHAPTER VI.
CROP VALUES AND CONSTRUCTION.
§ i. NORMAL INFLUENCE OF CROP VALUES.
WE now pass on from considerations of the volume of
crops to considerations of their value. Without discussing
at present whether or not an increased crop value represents
a mere transference of wealth from consumer to producer,
we may lay down the general proposition that it involves,
other things being equal, an increased demand for the pro-
ducts of constructional industry.
This proposition is indeed disputed as regards the United
States by Mr. Hull, who asserts l that while consumers
would use increased resources in enlarging their general
purchases, farmers would use it merely in paying off mort-
gages or in " laying field to field." With regard to the first
statement, a fuller discussion of the magnitude and rapidity
of the effect upon constructional industry of an increased
demand for consumption goods must be deferred till later :
but it may be remarked here that it is from the nature of
the case less considerable than the effect of the direct use of
resources in investment ; for in the one case there is merely
a stimulus to the use in construction of resources already
accumulated, whereas in the other there is an actual increase
in the volume of resources available for that use. As to
the second statement, it appears to have escaped Mr.
Hull's attention that the purchase of a field or the paying-off
1 Op. cit.t p. 46. Thus also Mr. Hawtrey (Good and Bad Trade,
p. 86) makes the extraordinary assumption that increased receipts
due to a diminished crop will be simply hoarded, since the farmers'
own immediate demand for labour will not be increased.
£9
go STUDY OF INDUSTRIAL FLUCTUATION.
of a mortgage involves the transfer of a considerable lump-
sum of purchasing power, which is now available for invest-
ment in the hands of the seller or the ex-mortgagee. More-
over, there seems to be good evidence that with a continuous
growth in prosperity, American farmers have become a
creditor rather than a debtor class.1 Again the farmers'
own demand both for repairs and renewals and for new pur-
chases of buildings, agricultural implements and so forth,
is by no means negligible. The importance of these con-
siderations is enhanced when we count among the " pro-
ducers " of corn, not only the actual growers but the whole
tribe of transport agencies, elevator men and merchants,
whose lending and investing capacity is thus increased.
On the other hand it is true that the growth among the
Western populations generally both of the power and the
opportunity to save — both of general wealth and of the
facility for investment upon a small scale, tends to lessen
the discrepancy between the use of resources in the hands
of producers and of consumers. Nevertheless it seems
probable that the increased resources of consumers will
still be used mainly in other kinds of consumption, or that
in so far as they are set aside for investment, the actual
investment will be preceded by a period either of individual
or of corporate hoarding ; so that in either case the effect
upon constructional industry will be less direct and im-
mediate than that of equal resources in the hands of
producers.
It seems indeed that exceptional cases must be admitted.
In the hands, for instance, of the Indian ryot, living in back-
ward up-country districts and making use of primitive
methods of cultivation, or of the Russian farmer, who is
inclined to hoard a considerable part of a good crop, as well
as a considerable part of the proceeds of that part which
is sold, an increased crop value may involve no perceptible
rise in the demand for constructional goods — certainly a
smaller rise than would an increase in the hands of the
American capitalist farmer, whose crop can be turned into
1 Piatt Andrew, op. cit., p. 350,
CROP VALUES AND CONSTRUCTION. 91
money while it is still in the ground, and who will have
ample opportunities to invest his increased earnings.
Similar differences prevail within the United States
themselves. Thus a given increase in the cash-receipts of
wheat farmers is likely to raise the demand for constructional
goods more than an equal increase in those of cotton-
planters.1 For the methods of cultivation of cotton are
from the nature of the plant more or less stationary, so
that the planter is less likely than the wheat farmer to use
increased resources upon the purchase of new machines and
implements, especially as the wheat-farmer is often also a
cattle owner, and likely to spend part of his increased
receipts on wire fences, roofing for steadings, etc. Moreover,
there seems to me reason to believe that the cotton planters,
being partly men of negro origin,2 are likely as a body to be
inclined to spend more on immediate consumption and less
on investment than the white wheat farmers.3 Moreover,
the actual variations in receipts from the wheat crop are
likely to be greater than those in the receipts from other
staple crops. For though there has been a marked trans-
cyclical advance in the value of the cotton crop the year
to year variations are, in consequence of the scarcity of
other sources of supply and the consequent approximation
to unity of the foreign elasticity of demand, comparatively
small.4 As to the corn crop, though the variations in farm
value are very considerable, so large a proportion of the
crop is consumed on the farm that they do not involve a
proportionate variation in cash receipts. Since, however, a
large volume of corn crop cheapens the cost of cattle-
feeding, and since, moreover, it seems likely that the world's
demand for beef is elastic, it might be expected that the
1 On this whole subject, cf. Piatt Andrew, Q. J. of EC., loc. cit.,
and H. S. Jevons, Causes of Unemployment, chap. iii.
* Cf . Coman, Industrial History of the United States, p. 294.
3 For instance, though a large cotton value was able to assist
in staving off depression in the iron and steel trades in 1890 and
1900, it does not appear capable by itself of initiating a revival,
as witness the years 1910-11.
4 Piatt Andrew, op. cit., p. 340.
92 STUDY OF INDUSTRIAL FLUCTUATION.
volume of the corn crop, through the medium of the net
profits of cattle-farming, should have an appreciable though
dilatory effect upon the demand for constructional goods.
§ 2. PSYCHOLOGICAL INFLUENCE OF CROP
VALUES.
The principle enunciated at the beginning of this chapter
is, however, subject to an important modification. At the
beginning indeed of a constructional expansion the " psy-
chological " impulse to investment afforded by a large crop
value is likely to reinforce, or even to compensate for the
absence of the " psychological " impulse due to a large
crop volume. For if the growers of crops find their resources
enlarged and embark upon investment, it is not unlikely
that an infectious spirit of confidence will get abroad which
will impel other people also to enlarge their expenditure
upon instrumental goods. But when the constructional
boom is far advanced, a large crop value may actually
have a deleterious effect upon the demand for constructional
goods.
One reason, at the cost of some anticipation, must be
sought in the relations of constructional and consumptive
industry. At a time when little investment in new plant is
being made by the manufacturers of consumable goods, a
transference of resources from consumers to producers of
agricultural produce will have little effect in checking such
investment, since if the transference had not taken place,
these manufacturers would have been chiefly occupied in
getting more product out of existing plants. But at a time
when manufacturers are investing on a large scale, such
a transference of resources and the consequent decrease in
the demand for consumable goods will have an immediate
reaction, since it will quickly become plain that if the
capacity even of existing plants is too great, there is no
remunerative opening for further investment.
Another reason is to be found in the mechanism of a
credit economy. If the general financial situation is on
CROP VALUES AND CONSTRUCTION. 93
the whole sound, or believed to be sound, advances are
readily made against the crops, which thus create for them-
selves the currency which they require. For instance, in
America in 1891 when there was a very large increase in
wheat value as well as in volume, " the fine crops enabled
bankers and merchants to keep money rates from advancing
further, and to bolster up the market in order to unload their
securities and merchandise." 1 Again in 1890 and 1900
the large sums required for financing the big cotton crops
gave little embarrassment, for in the latter year general
financial conditions were good, and in the former year,
though there was a financial panic in November, the cur-
rency situation was temporarily eased by the passage of
the silver coinage law. In 1903 and 1907, however, the
increased values of wheat, and in 1903 the largely increased
value of cotton, were due to shortage, which diminished
confidence, already shaken by unsound industrial and
financial conditions ; so that the money required for financ-
ing the crops was not created ad hoc but was to some extent
at least withdrawn from other uses. In other words when
bankers are thinking only about their security valuable
crops are able to produce their full effect on the construc-
tional industries : but when they are getting uneasy about
their gold reserves, crop-carrying and loans to other inves-
tors become to some extent rivals.
It is clear that the deterrent effect of a big crop value on
investment is likely to be most severe if the large value is
due to shortage. It is possible, however, that if financial
conditions are very shaky, the net effect upon constructional
enterprise even of large crops which have a high value may
be injurious. During a great part of the summer of 1913
it seemed to be a toss-up whether the large anticipated
crop values would prolong or cut short the industrial boom
in Canada. It was frequently heard that only a continuance
of agricultural prosperity could prolong the expansion :
but it was equally freely urged that the financing of the
large crops would be a very severe strain on the situation
1 R. W. Babson, Business Barometers, p. 148.
94 STUDY OF INDUSTRIAL FLUCTUATION.
unless plentiful assistance were forthcoming from London.
In the end it seems that the financial difficulties were tided
over sufficiently well to allow the big crops to have a per-
ceptibly steadying effect.
§ 3. DETAILED ILLUSTRATION FROM UNITED
STATES AND GREAT BRITAIN.
These • somewhat complicated propositions, as well as
some of those laid down in the preceding chapter, can per-
haps most conveniently be illustrated by a continuous
narrative of the relations between crop conditions and con-
structional enterprise in the United States during the last
thirty years (cf. Chart VI.). In 1886-7 the large volume
and value of the wheat crop, following on two years of
abundant corn, gave a great impetus, mainly through
railway building, to pig-iron production. In 1888 the
wheat volume was small, and the demand for railway
transport correspondingly curtailed ; but the value was
large, and everything points to the fact that the activity
in pig-iron would have continued to expand, but for the
occurrence of that phenomenon, happily peculiar to the
United States, of a " presidential year." I The iron pros-
perity of the following year may be put down partly to the
delayed purchases with these large wheat receipts, partly
to two years of abundant corn (1888-9). By the end of
1890 it seems as if considerable over-investment had taken
place, and the crash that was due was almost precipitated
by the failure of both wheat and corn crops.
As it was, the price and production of pig-iron fell off
considerably in 1891, but a serious depression was postponed
by the artificial stimulus of the McKinley Tariff and the
Sherman Silver Law, by the large and valuable cotton
1 The fact that the years of Presidential election — 1884, 1888,
1892, 1896, 1900, 1904, 1908, 1912 — were all except the last years
either of depression or at least of stress and instability can hardly
be entirely accidental : it seems pretty clear that this disturbing
element, sometimes reinforcing and sometimes counteracting the
influence of agricultural conditions, is of some importance.
CROP VALUES AND CONSTRUCTION. 95
crop of 1890, and by the enormous wheat and corn volumes
and values of 1891. Pig-iron production accordingly took
a new leap in 1892, but in that and the following year cereal
and (to a less extent) cotton production were so bad that
the averted depression was at length precipitated. In the
autumn of 1894 the large volume of cotton, and in that of
1895 the large volume of corn, accompanied in both years by
an increased volume of wheat, produced more or less specu-
lative rises in the rate of pig-iron production, but the
conditions of value were not favourable, and in 1896 there
was a relapse. It was not till 1897 that the very large
wheat value (due, it will be remembered, to abundance
coinciding with a European shortage), accompanied by a
large cotton value, and reinforcing two years of plentiful
corn, was able to produce a genuine industrial revival^
continued in the next year, especially through the medium
of the railways, by the immense volumes of wheat and
cotton. The revival, once on its way, was undeterred by
the wheat and cotton relapses of 1899, but overtrading
began, and in the latter half of 1900 there was a big drop
in iron price and production, accentuated by a further fall
in the wheat crop. The large values of corn and cotton
however kept things going in the winter of 1900-1, and the
enormous wheat volume and value of 1901 gave the impulse
to a new wave of prosperity, which reached its culmina-
tion (with the help of the big corn crop of 1902) in June,
1903. The diminished crops of 1903 refused to support it
longer, and depression followed until relieved by the record
cotton crop and the large wheat value in the autumn of
1904. The new tide of activity was fed by very large
cereal volumes and values in 1905-6, accompanied by a
record cotton value of 1906. But by the end of 1906 it
became clear that things were going too fast, and the
diminished crops of 1907, though accompanied in the case
of the two cereals by larger values, helped to deliver the
coup de grace in the autumn of 1907. The fair crops of
1908 mitigated the depression and the large wheat and
cotton values of 1909 produced a new maximum of pig-iron
g6 STUDY OF INDUSTRIAL FLUCTUATION.
production, which however rapidly fell off with the poor
wheat crop in the autumn of 1910 : and it was not till the
large and valuable wheat crop of 1912 was well in sight
that the constructional prosperity of United States was
fully set upon its legs again. But the shaky financial
conditions of 1913, combined with the failure of the corn
crop, prevented an increased wheat value from prolonging
the constructional boom.1
The evidence thus supplied by the United States is per-
haps sufficient to establish the general principles laid down
in this and the preceding chapters. But a more detailed
study of the effect upon the various British constructional
trades of variations in the purchasing power of certain
groups of agricultural producers seems to be of considerable
historical and practical interest.
For a long time not only upon the shipbuilding but upon
the iron and steel trades in general it was the influence of
the United States that predominated. For instance, though
it was the Russian demand for railway material that ini-
tiated the British iron boom in 1869, it was mainly the
American demand that sustained it in 1870-1. Again in
the autumn of 1879 it was the arrival in Sheffield of an
order from Mr. Vanderbilt for 20,000 tons of rails that
first raised the price of pig-iron. Much the same may be
said of 1886, 1895, 1902 and 1909.
There are certain features, however, about the American
market that call for attention. A glance at the chart of
American construction (II.) will show two solid booms,
culminating in 1872 and 1882, and followed by solid depres-
sions : then an expansion with triple crests in 1887,
1890 and 1892, followed by a depression : then a kind of
anacrusis in 1895, followed by an expansion with double
1 The orders of the railway companies decreased by 50 per cent,
in 1913, and the volume of unfilled orders on the books of the Steel
Corporation declined from October 1912 (Times History, J an. 16, 1914).
The figures given by Mr. Hull in his appendices of the annual rate
of monthly production and consumption of pig-iron appear to me
to corroborate the crop theory by showing that the most character-
istic movements of the curves take place in the harvest months.
CROP VALUES AND CONSTRUCTION. 97
crests in 1899 and 1902, followed by a depression : then a
single expansion culminating in 1907 and followed by a
depression : finally another anacrusis in 1909, followed after
a prolonged dip by a genuine expansion in 1912-3. The
quantity of our iron and steel exports to the United States
rose from 1868 to 1871, fell till 1878 (with a slight recovery
in 1877), rose enormously till 1880, fell in 1881, rose in
1882, fell till 1885, rose considerably to 1887, fell steadily
till 1894, rose in 1895, fell to 1898, rose to 1902, fell till
1905, rose to 1907, fell in 1908, rose to 1910, fell in 1911-2
and rose in 1913. Thus the three solid American booms,
both branches of the double one and the two anacruses, all
had a marked effect on the British iron industry, and were
on each occasion the first influence to wake it into new life.
Of the Geryon of 1887-92 only the first crest made an
appreciable impression on this side of the Atlantic, and the
final expansion of 1912, while it imprinted its kiss upon
England's forehead, was not destined to act the part of
Prince Charming, but found the princess this time already
wide awake. Again in some cases the influence of the
United States, while predominant in the early stages of
English constructional expansion, is before long with-
drawn. Thus in the American booms of 1869-73, 1879-82,
and 1887-92 respectively the English exports of iron and
steel to the States reached their maxima in 1871, 1880 and
1887. On the other hand those of 1898-1902 and 1905-7
continued to exert an influence throughout their length.
These peculiar relations of English and American con-
struction are partly to be explained by the following con-
siderations. In the first place we observe in the earlier
American expansions a marked tendency for the production
of pig-iron to outrun its price. Thus while the price reached
its maximum in September, 1872, 1 the production continued
to increase in 1873 : price reached a maximum in February,
1880, and production in 1882 : price in January, 1887, and
production in 1892. That is to say, the home production
was after the first pressure amply capable of responding
1 Philadelphia prices, quoted by Hull, op. cit., Ap. A (2).
98 STUDY OF INDUSTRIAL FLUCTUATION.
to the new demands, and the resort to English sources
was only temporary. In the later periods the tendency
has been less marked. Thus the price and the annual rate
of the month's production x both reached a maximum in
November, 1899 ; in 1902-3 the interval was seven months
(November, 1902, to June, 1903) and in 1907 four months
(February to June). It may be inferred that on these
occasions, whether owing to natural difficulties or to trust
action, the home supply was not so unrestrainedly respon-
sive, and the pressure on foreign sources likely therefore,
as we actually find, to be more prolonged.
Secondly, the different character of American construc-
tional booms must be borne in mind. In almost every case
the first outburst of energy seems to have been devoted to
railways : but whereas in 1879-82, 1898-1903 and 1905-7
activity in railway building continued till the end of the
boom, in the other periods it reached its maximum in 1871
and 1887 respectively, and the remainder of the expansion
seems to have been mainly concerned with other kinds of
construction. Now it does not appear that we have ever
been able to rely on America furnishing us with a large
market for other kinds of iron and steel goods besides pig-
iron and rails ; our superiority in other kinds of structural
goods has not been sufficient to attract her, provided she
can get enough pig-iron to make them herself. In 1872-3
and 1888-92, as we have seen, she could get more than
enough : hence both the chief sources of the demand for
English goods began to dry up with the slackening of the
railway boom.
But while in a well-developed industrial country like
the United States, fully capable of performing its own
finishing operations, the effect of a revival upon British
constructional industry is concentrated mainly upon the
preliminary stage of manufacture — the production of
pig-iron — in a new and mainly agricultural country the
effect is likely to be more pronounced on the finishing
industries. Thus in 1888-90, when the iron and steel
1 Hull, op. cit., Ap. D.
CROP VALUES AND CONSTRUCTION.
99
boom was largely due to Argentina and the American
demand had subsided, it was the finishing industries that
reaped the chief benefit.1 Again in 1903-4 when pig-iron
was depressed under the influence of reaction in the United
States, both the quantity and value of our exports of such
things as galvanised sheets, miscellaneous railway material,
nails and screws, and almost all descriptions of agricultural,
textile and other machinery, showed a continuous increase. 2
Moreover the revival in 1905, when American demand
" though better was still sluggish " was in no small measure
due to a demand for " thin and thick sheets for wagon-
building and roofing in the colonies " and other distant
countries.3 Again in 1909 certain branches of the engin-
eering trade, such as that of locomotive making,4 seem to
have been fairly prosperous early in the year, while pig-iron
was still depressed, and the year as a whole was a record
one for the export, among other things, of wire, galvanised
1 Exports of : —
Pig-iron.
Steel Ingots,
Bars, and
Blooms.
Railroad
Material
other than
Rails.
Tinplates.
" Other
Sorts."
1882 .
1889 .
Tons.
1,758,072
1,190,371
Tons.
162,774
85,764
Tons.
155,166
331,957
Tons.
265,039
430,650
Tons.
292,419
4H,387
1882 .
1890 .
4,962,185
3,498,568
£
1,861,109
1,175,136
1,245,689
1,974,853
£
4,642,125
6,36i,477
3,930,181
5,041,424
1882
1890
Locomotives. Textile and other Machinery.
£919,988 £7,053,420
£1,848,462 £10,582,603
2 It is true that there were loud complaints of depression in the
engineering trades, and that the unemployed percentage rose from
2-2 in 1900 to 4-4 in 1903 and 6-2 in 1904, but this was due mainly
to depression in the home shipbuilding, textile and other trades.
The demand for more or less finished iron and steel goods from the
new agricultural countries made the real extent of depression in
our constructional industry considerably less than might appear
from concentration upon pig-iron.
3 Economist, 1905, p. 1,546, * Cf. EC., 1909, i. p. 662,
TOO STUDY OF INDUSTRIAL FLUCTUATION.
sheets, " wrought tubes and pipes and fittings," " girders,
beams, joists and pillars " l — records of which all but one
were beaten in the following year, when the temporary
United States boom had collapsed. From the beginning
of 1910, moreover, the engineering unemployed figure began
a prolonged fall, and the Sheffield steel industry entered on
a new career of prosperity.2 Meanwhile, the price of pig-
iron remained stationary from January to April, and then
fell, only rising at the end of the year : for again it was not
the United States but the new wheat countries, Argentina
and Canada, reinforced by Brazil the rubber exporter and
South Africa at last recovering from her troubles, which
among foreign countries were mainly responsible for the
constructional revival of 1910-1.
So also in 1913, after the collapse of the American demand
and the price of .pig-iron, the Sheffield trade " recovered
from the threatened reaction," and business in " tools,
agricultural implements, saws, and foreign and colonial
rail and railway requirements " remained good.
The growing importance of the new corn-raising countries
suggests, therefore, that the current attribution to pig-iron
of an almost sacramental position as the outward and
visible sign of trade conditions in general — even of the
conditions of constructional industry in general may need
to be modified. This suggestion, which is indeed becoming
a commonplace among economists, is confirmed by the
reflection that not only has the other chief source of demand,
the shipbuilding industry, become less sensitive to American
conditions, but at no time has the price of pig-iron been
morbidly amenable to its solicitations. It does not seem
likely, for instance, that all the materials used in the enor-
mous shipbuilding output of 1874 were contracted for at
the prices of 1873 : yet there was a serious relapse in both
manufactured and pig-iron prices in 1874. Again the
aftermaths of 1891-2 and 1901 were not able to restore the
prices of materials. Again while the average price of steel
ship-plates rose from £6 4$, in 1887 to £6 145. yd., in 1888
1 Id., 1910, i. p. 319. 2 Id., History of 1910.
CROP VALUES AND - .oON^TR ACTION 101
and that of angles from £5 145. 2d. to £6 35. 2d., that of pig-
iron actually fell from £i 155. 0-15^. to £i I2s. 5-66^. x Again
in 1896 the price of pig lagged behind that of steel.2 It
seems as if the producers had got into the habit of answering
mechanically to the United States stimulus, and were not
on the look-out to take advantage of revivals in demand
from other directions. Again pig-iron prices seem to have
failed partly in 1910 and wholly in 1911 to profit by the
increased shipbuilding, partly owing to the large quan-
tities of scrap-iron from broken-up tonnage thrown upon
the market.3 Conversely the collapse of shipbuilding
enterprise in 1902 and the consequent fall in the price
of steel plates from £6 4$. $d. to £5 145., and of angles from
£5 195. gd. to £5 us. yd., was not able to depress pig-iron
prices in the face of the American demand.4
I say, therefore, that the traditional conception of pig-
iron as the barometer of trade needs to be temporarily
modified ; yet not perhaps permanently. For side by side
with the decreasing relative importance of America to
English general constructional industry must be put its
decreasing importance to the English pig-iron industry
itself. We have already seen that the greater measure of
control which has been obtained over American production
prolongs the demand upon foreign sources. But the same
factor, implying as it does the maintenance in times of
depression of a larger reserve of unused capacity, and taken
in conjunction with the colossal expansion in capacity in
recent years, seems likely also progressively to lessen the
1 Average Cleveland prices (Report on Wholesale and Retail
Prices, 1903, pp. 28, 18).
2 Cf. Economist history of that year. It is not clear, however,
why the Board of Trade shows an actual fall in its index for the
year : all the varieties which it quotes, except the declared export
values, exhibit a rise. A similar lagging of pig-iron behind steel
prices in 1882 is noticed by the British Iron Trade Association Review
of that year.
3 Economist, 1911, ii. p. 1,359.
4 This is not of course to deny the importance of shipbuilding
to the consumption of pig-iron, nor to its price in such years as
1889, 1899, 1906 and 1912.
102 STUDY OF INDUSTRIAL FLUCTUATION.
urgency of the initial call. In the last decade, while pig-
iron has still been largely at the call of America, and the
rest of constructional industry mainly at that of other
countries, pig-iron has necessarily been somewhat of a
false prophet ; but as it too becomes emancipated, it will
come to reflect, as it did in 1910, more slowly but more
accurately the genuine prospects of English construction
— with something of the irritating serviceableness of a
watch that is never far wrong but generally rather slow.
If our shipbuilding and finished iron and steel trades
are not always affected in the same manner as our pig-iron
industry by cereal conditions, still less is our coal trade.
In the first place, the less highly finished the form taken
by our iron and steel exports, the less the quantity of coal
that goes to their making. Secondly, certain corn-raising
countries, notably the United States, have always remained
impervious to our coal trade : our exports of pig-iron
thither are not followed up by exports of coal to assist in
their further manufacture into instruments of production
and in the operation of these instruments. For both these
reasons those rises in the price of pig-iron which have been
primarily due to American conditions, such as those of
1880, 1895, 1902 and 1909, have not been reflected in the
price of coal : while in 1888-9, wnen the iron and steel
boom was largely in finished goods, and cereal prosperity
reigned not in the United States but in South America and
Europe, our exports of coal were greatly increased, and the
price rose earlier and more rapidly than that of pig-iron.1
1 The importance of Russian agriculture to our coal trade in
these and the following years is indicated by the following figures : —
Russian Wheat Exports. British Coal Exports.
1887 . 10 m. qrs. . . 24
1888 16 . . 27
1889 14
1890 16
1891 13
1892 . 16
1893 ii
1894 15
29
30
31
30
29
33
Again in 1913, the cereal prosperity of Russia, combined with
CROP VALUES AND CONSTRUCTION. 103
Finally there is one very important form of constructional
industry to which the generalisation laid down at the
beginning of this chapter does not seem to be entirely
applicable, and that is the building trade. In the United
States indeed, through the medium of the iron and steel
and railway industries, a conection can be traced between
the prosperity of cereal producers and of those engaged in
the building trades. For first a revival of American con-
structional enterprise involves an immediate construction
of brick buildings, and the price of brick — there being less
reserve productive capacity in existence — rises even sooner
(1898, 1901, 1904) or more markedly (1886, 1909) than that
of iron. Secondly, there follows an immense flow of immi-
gration creating a demand for house-room which leads to
much speculative building.1 But in England the influence
of the prosperity of groups of cereal producers on the iron
and steel trades works chiefly through the export and ship-
building trades, and does not therefore involve a vast
amount of terrene construction. In so far as the English
building trade is affected by agricultural conditions, it
seems to be through the medium of consumers rather than
producers, and its further discussion must therefore be
postponed to the next chapter.
a shortage of home supplies of coal and oil, has been of great benefit
to our coal export trade.
1 It will be noticed (Chart II.) that the curve of the average
price of all building materials seems on the whole to be more sen-
sitive to that of immigration than does that of brick, and may be
inferred to be a better index of domestic building : compare their
movements in 1889-90, in 1895, in 1899-1901, 1904, 1907, 1910-11.
It tends to reach its maximum later than the brick curve, because
the flood of immigration is prolonged at least as long as the iron
boom lasts.
CHAPTER VII.
CROP VALUES AND CONSUMPTION.
§ i. CROP-PRODUCERS AND CONSUMPTIVE
INDUSTRY.
THE demand for certain important kinds of consumable
goods seems also normally to be increased by the transfer-
ence of resources to certain groups of agricultural producers.
Thus there is evidence of a connection between cereal pros-
perity in America and the demand in that country for
imported food-stuffs. The significance of the curve of per
capita consumption of coffee l in each fiscal year (ending
June 30) and of its relation with that of consumption 2 of
pig-iron (see Chart XII.) will be discussednn other connections
in later chapters : here it may be remarked that of the only
four years in which the two curves show unmistakably a
common upward movement, — 1892, 1898, 3 1902, 1907, — the
first three were years in which the farm value of the wheat
crop reached a maximum, and that of the only two years in
which they move together downwards, one — 1894 — was a
year in which it fell to the lowest point reached since esti-
1 On the whole, I think as fair an index of general consuming
power as any single commodity can be. The States absorb more
than half of the world's production, and the per capita consump-
tion is greater than in any other country except Sweden, Norway
and Holland.
a Arrived at by adding the net imports of the fiscal year to the
output of the previous calendar year, but not on the whole (cf.
p. 96, n.) liable to serious error.
* The large coffee " consumption " of this and the previous year
must, however, be referred partly to large importations in anticipa-
tion of the increased duties imposed by the Dingley Tariff late in
the (calendar) year 1897.
104
CROP VALUES AND CONSUMPTION. 105
mates became available. Further the divorce between the
curves is more pronounced when, as in 1901 and 1904, the
small wheat volume which helps to bring about a con-
structional check is accompanied by a larger value to
growers, who thereby gain in purchasing power, than when,
as in 1891, volume and value alike are reduced. The
inference to be drawn is the considerable importance in the
United States of the farming community not only as inves-
tors but as consumers.1
Again the influence of cereal and constructional prosperity
in America on the Yorkshire woollen trade is easy to detect.
The booms of that industry in 1867-70, 1879-81, 1886-90
and 1912-3 were largely export booms due to the American
demand : while conversely the depression of the seventies was
mainly caused by the failure of the American inquiry,2
and in 1903-4 the collapse of the meteoric American demand
of 1902 aggravated the difficulties of the trade. But on
this occasion (as with the engineering trades) our other
export markets, especially Canada, more than made up for
the deficiency : and in 1909 the same agricultural advance
of the new countries — notably Canada, South Africa and
Australia, — which benefited certain branches of the engineer-
ing trade, made itself felt also on a number of miscellaneous
consumptive products, such as chemical fertilisers, plate
glass, leather gloves, paraffin wax, cycles, motors, ropes,
soap, linen piece-goods, straw hats, paper-making materials,
matches, leather boots and shoes and seed oils : of all of
which our exports in this year beat all previous records.3
Finally, there are a few, though only a few, industries to
1 In all respects the period centring about the collapse of 1907 is
an exception : it seems that in this case the industrial boom had
so far outstripped in magnitude the forces which helped to originate
it that the fall in the value of the wheat crop of 1906 was not able
to prevent a rise, nor a rise in the value of the wheat crop of 1907
to prevent a fall, in consumptive activity.
2 The value of our woollen exports to the United States fell from
£6-3 m. in 1872 to ^1-4 m. in 1878 ; the total quantity of our woollen
exports fell from 38-5 million pounds' worth in 1872 to 19-6 million
pounds' worth reckoned at the prices of 1872, in 1878.
3 Cf. Stat. Abs. of U.K. and Economist, 1910, i. p. 319.
io6 STUDY OF INDUSTRIAL FLUCTUATION.
which the prosperity of the home agricultural community
is a factor of leading importance. Its influence, of course,
is not unfelt in the boot and clothing trades, but it is com-
pletely overshadowed by other forces. But the proverbially
eccentric course of the Scottish mineral oil trade seems
often to have been dictated by the demands of home agri-
culturalists for its important by-product of sulphate of
§ 2. CROP-CONSUMERS AND CONSUMPTIVE
INDUSTRY— INDIA.
But the course of English consumptive trade and
industry as a whole seems to be influenced rather by
variations in the available resources in the hands of certain
groups of consumers than of certain groups of producers of
agricultural produce. Among such groups the two most
important are the population of India * and the working-
classes at home.
The influence of the former is of course felt predomin-
antly in the cotton trade. It is well known that the fortunes
of Lancashire depend largely upon the abundance of the
Indian rains and the consequent character of the grain-
crops : but the precise nature of the connection has not
always been made clear. The correct conclusion seems to
1 Used as a top-dressing for the soil. Thus in 1902-3 (the year
ends April 30), in spite of a slight decline in the price of most of
their other products, the Scotch companies made the respectable
profit of ^192,000, while with the advance of " general " depression
in 1903-4 their profits increased by ^93,000. But in 1904-5, while
"general trade " revived, the home harvest was bad, and the com-
panies in spite of a policy of rigid economy and energetic improve-
ment, could only increase their profits by ^9,000. In 1909-10
and 1910-11 the demand for sulphate of ammonia was again their
chief stand-by, and in 1911 it was reinforced by another conse-
quence of the excellent harvest of that year, the growing demand
for agricultural motor machinery. Cf . annual articles in the Econo-
mist : but cf. also p. 52.
2 The exports to British India form about 40 per cent, of our
total exports of piece-goods, and it will be noted that in almost
every year the two curves move in the same direction.
CROP VALUES AND CONSUMPTION. 107
me to be that the most important factor in stimulating
cotton imports is a large production and a low price of rice,
but that the influence of this factor is frequently modified
by other agricultural phenomena and by the speculative
character of the import trade in cotton goods.
Thus the cotton import booms of 1894, 1898-9, 1904 1
and 1910-1 are clearly to be connected with the large
harvests and low price of rice in those years : and the
slumps of 1892, 1897, 1900 and 1908 with shortage and
famine prices. But the import booms of 1889-90, 1896,
1901, 1905-7 and 1912-3 2 seem to have been mainly
speculative : and in 1902-3 and 1907 therefore, and even
in 1895 (when the confidence of importers had not been
without justification in rice conditions 3) stocks were so
large that heavily falling rice prices were prevented from
exercising their due effect. But the speculative behaviour
of importers is not always beyond the reach of explanation.
For 1896 indeed no excuse is forthcoming : 4 but in 1901
and again in 1912 that large production and largely in-
creased value of the exports of wheat, which (though pre-
vented by previous speculation from influencing the move-
ment of the import curve in 1891) had served well in 1886,
1888, 1898, 1904 and 1909-11, was reckoned at perhaps
more than its true importance. Again in 1905 the rice
crop was actually larger than in the previous year, and the
rise in price was due to shortage of the minor rival food
grains.5 In 1906 a large rise in the export value of jute
1 The rice crop indeed was not enormously large in 1904 : but
the price was driven down by the exceptional crops of wheat and
of the minor food-grains.
z The E. H. of 1912 reports that the Indian stocks were very
heavy — those in Bombay had doubled since a year ago.
3 The imports of 1894 were, however, swollen by anticipation
of a new duty.
4 For 1889 and 1907 cf. p. 61, n.
5 Numbers of the poorest inhabitants live habitually on these
cheaper grains (millet, etc.), and regard rice in ordinary circumstances
as a quasi-luxury (cf . Atkinson, Income and Wealth of British India,
Stat. Jour., 1902, p. 213 ; and for the difficulties in the way of
substitution caused by the conservatism of consumers, id. in Stat.
Jour., 1897, p. 96).
io8 STUDY OF INDUSTRIAL FLUCTUATION.
gave confidence in the country's purchasing power : l just
as in 1911 the slight rise in rice prices was due largely to
shortage in China, Java, Japan, etc.,2 and the increased
prosperity of exporters seems to have compensated the
diminished purchasing power of home consumers. On
these occasions, therefore, the confidence of importers seems
to have been not without justification.3
It is worth observing here that the weather conditions
1 Cf. R. of I. T., 1907-8, p. 33, and 1908-9, p. 46 : " For the
third year in succession the salient feature of the Indian rice trade
has been the strong internal demand and the curtailment of the
exportable surplus. This surplus may be diminished either by
poverty of out-turn in the yield itself or by shortage of other food
grains for which rice may be substituted, or by exceptional pros-
perity and enhanced consuming-power among the cultivators of
crops other than food grains, or by a disposition on the part of
cultivators to grow such specially profitable crops in place of rice.
All of these causes have operated in recent years, but such of them
as depend on the exceptionally profitable cultivation of other crops
have in 1908-9 been less operative than in either of the two previous
years, seeing that the price of jute has very greatly fallen."
2 R. of I. T., 1911-12, p. 46. This was a disadvantage to the
Indian cotton manufacturers, whose exports fell in the year 1911-
12 by 1 6-0 per cent, in volume and 18-3 per cent, in value. It seems
possible also (ibid., p. 60) that owing to lack of resources the Indian
industry recovers more slowly from depression than the English,
so that the effect of a grain crop revival on the English industry is
thus enhanced.
3 The assumption, upon which the foregoing account is built,
that the Indian demand for rice is normally inelastic, seems to be
correct, and to be borne out by the following figures, which give the
rice crop of each year multiplied by the average price of that year
(the former in m. cwts., the latter as ratio of the 1873 price — the
actual average prices are not given by the Stat. Abs.) :—
1901 757 1907 1,018
1902 638 1908 951
1903 76° !909 868
1904 641 1910 1,100
1905 758 1911 1,104
1906 922 1912 1,200
Of the exceptional years, 1910 is to be explained partly by export
conditions, 1903 and 1910 partly by the demand of cotton growers,
1908 partly by the collapse of the demand of jute-growers. The
annual computations of the R. of I. T. of the " sum saved " or
" lost " by consumers as compared with previous years is based,
however, on the illegitimate assumption that the elasticity of
demand is actually zero.
CROP VALUES AND CONSUMPTION. 109
of India seem often to be more carefully and intelligently
watched by the Lancashire spinners than by the Indian
importers. The result of this has sometimes been to give
the curve of spinning profits an appearance of sustained
" periodicity " for which neither the period of gestation nor
the alternations of the Indian seasons would have sufficed
alone to account. Thus the speculative prolongation of the
Indian demand in 1889-90 enabled spinners to make large
actual profits, while their own critical observation of the
Indian rains prevented them for a time from annihilating
those profits by over-investment.1 Similarly in 1896 a
good importers' demand was accompanied by a drastic
reduction of spindles dictated by Indian weather condi-
tions, with the result that profits remained almost stationary
in 1896 and leapt up in the great famine year 1897. On
the other hand the over-investment when it came in 1890
had been aggravated by the apparently rational inferences
of spinners from the weather conditions of 1890, — inferences
which were falsified owing to the previous irrationality of
importers. Moreover, just as the successive impacts of
different streams of demand were seen to produce a con-
tinuous boom in shipping, so an increased Continental
demand 2 for yarns or piece-goods or both seems sometimes
(e.g., 1887, 1889, I^95) to dovetail into the interstices of
Indian inquiry and so to prolong the boom. The net
result of all these forces is that while between 1886 and
1908 there were at least six serious failures of the Indian
crops separated by periods of plenty, there were only three
complete upward and downward sweeps of the curve of
spinning profits.
1 N.B., the marked check to the growth in the number of spindles
in 1890.
2 The Continental demand, while mainly a reflex from construc-
tional industry, is also partly dependent on a transfer of resources
to agricultural consumers (cf. especially Messrs. Ellison's Continental
reports in circulars of 1910-1 and 1911-2).
no STUDY OF INDUSTRIAL FLUCTUATION.
§ 3. CROP-CONSUMERS AND CONSUMPTIVE
INDUSTRY —GREAT BRITAIN AND UNITED STATES.
The other group of consumers of food grains in whose
hands an increase of available resources materially raises
the demand for the products of English consumptive indus-
try is the English working-classes. And since there is
abundant reason to believe that their demand for wheat is
on the whole inelastic, here again it is a low price of grain
that is correlated with an increased prosperity of the con-
sumptive trades.
In particular this factor seems to be of the first import-
ance in determining some of the most characteristic move-
ments of the English clothing trades. For while the cotton
trade is mainly dependent on the Indian grain-consumer,
and certain branches of the woollen industry are much
affected directly and indirectly by the fortunes of American
agriculturalists and other foreign customers, the course of
our woollen and clothing trades on the whole is very much
more dependent on the movements of home demand.1
1 According to Sir Charles Macara (letter quoted in EC.,
1909, P- 1323), three-quarters of the product of our cotton spinners
is exported in some form or other : other estimates put the pro-
portion as high as f (Ec., 1910, p. 797), or even -f. I know of no
similar expert estimate for the wool industry, but the following
figures are significant. Those for output are from Table I (a)
(goods made by firms for sale, including work done for them on
commission) of the Preliminary Tables of the Census of Production
(Cd. 4896, 1909), those for export from the Stat. Abs. of U.K. : —
Yarn.
Piece-goods.
Cotton.
Wool.
Cotton.
Woollen.
Worsted.
Output
m. Ibs.
m. Ibs.
m. yds.
m. yds.
m. yds.
1908
I5Q7'4
149
7030
173
181
Export
1907
241-0
83
6298
85
99
1908
214-8
71
5531
76
74
Output
I
i
i
i
i
1908
79-3
18-1
81-9
16-4
15-6
Export
1907
15-4
6-0
81-0
10-3
7'4
1908
12-8
4-6
70-2
9-6
5'9
CROP VALUES AND CONSUMPTION. in
Thus it was partly owing to the sharp rise in wheat prices
in 1872-3 that the woollen trade failed to share in the
general elation of trade in those years. But in 1874-6,
when depression in the great constructional trades was
accompanied by a rapidly falling level of wheat prices, the
home branches of the woollen trade seem to have done on
the whole very well.1 In 1877, however, with a sharp
rise in wheat prices, there was a severe falling off in pros-
perity. Again in 1880-1, when there was a marked reaction
from the low wheat prices of 1879, ^ was the export and
not the home demand that contributed to the prosperity
of the trade.2 But already in 1881 with the slight relapse
in wheat prices, " there has been a continued improvement "
in ready-made clothing at Leeds : in 1882-4, m spite of a
continuous decline in exports, demand was good : and
even in 1885, the classic year of trade depression, it seems
that on the admission of the trade itself " a large volume
of business was done." There can be no doubt that the
immense fall in wheat prices in these years was a very
In the case of piece-goods the figures for export bear a much
higher proportion to those of production in the case of wool than
in that of cotton. In the case of yarns this is not so : but it must
be remembered that the census figures only take account of yarns
sold by one firm to another, so that in view of the much larger pro-
portion of combined spinning and weaving establishments in the
wool (that is, in the woollen branch : in the worsted branch speciali-
sation is the rule : cf. D. H. Weld, Specialisation in the woollen
and worsted industry: Quarterly Journal of Economics, Nov., 1912,
pp. 67 ff.) than in the cotton trade, the figures for output of yarns
are not comparable.
1 " It has not been profitable," said a shrewd man of Bradford
in 1874, " but we have made money." " The producing power of
this district," we learn from Huddersfield, "largely increased as
it has been of late years, has been kept fully employed from first
to last in 1874. And this is the more remarkable in view of the
great depression which has so widely prevailed in other leading
industries." Again in 1875, Messrs. Helmuth Schwarze call atten-
tion in their circular to the " comparative prosperity of the cloth-
ing trade, especially in the commoner kinds of goods " ; and the
rise in the price of the raw material in the second half of 1876 was
partly due to a marked recovery in consumptive demand.
2 At Halifax, in 1880, "home buyers have been little in the mar-
ket " ; in 1 88 1, according to the Leeds Mercury, " the chief increase
has been in the shipping rather than the home trade,"
H2 STUDY OF INDUSTRIAL FLUCTUATION.
considerable benefit to the industry, and that their con-
tinued low level in 1886-90 contributed materially to its
prominence — especially the prominence of that important
barometer the Leeds ready-made clothing trade — in the
revival of those years.1
The explanation of the movements of the industry from
1891 to 1896 must be sought in the main elsewhere : but
its collapse in 1897-8 must I think clearly be connected with
the price of wheat. 1897, it must be remembered, was a
year of rising cotton profits, and of increased production
and price of both iron and coal. Obviously the divergence
of wool in this year is crucial to our whole inquiry. Various
explanations — the great engineering dispute, the warm
weather — suggested themselves to contemporary observa-
tion : but it does not seem that either of these causes is
sufficiently fundamental. Moreover, the depression con-
tinued in the following year, when neither of these causes
can be invoked, when the prosperity of the three other
great industries was even more marked, and when ship-
building had been added to their number. It is, I think, clear
that we must look for the solution in the world's wheat-
shortage of 1897-8, culminating in Sir W. Crooks' wheat-
scare and the Leiter corner, and finally relieved by the
record crops of the season of 1898.
1 For the recovery of 1886, cf. pp. 27 and 61. " There is little
doubt, however," we are told, " that it was not the state of the
wool market alone which caused the revival." Between 1885 and
1889 it is estimated that our export trade increased 10 per cent.,
but our home trade as much as 30 per cent. ; moreover, there were
evidences of a better quality in the goods supplied.
2 The Gazette price of wheat, which had touched 225. lod. per
quarter in 1894, rose from 26s. 2d. in 1896 to 305. 2d. in 1897 and
345. in 1898. The hypothesis that the effect of this rise on the
purchasing power of the masses was of predominant importance
to the woollen industry is borne out by the evidence as to the rela-
tive diminution of the demand for the different qualities of goods.
Thus in 1897 we learn from Leeds that " generally speaking the
home demand for the higher class of goods, such as the superior
worsteds, has been large and steady, for the medium class only
moderate and somewhat fluctuating, and for the lower qualities
distinctly poor. , , , The trade in ready-made clothing has shared
CROP VALUES AND CONSUMPTION. 113
The relapse of wheat prices in 1899 permitted the home
trade to take its share in the boom of that year : but from
1900 to 1903 it was disappointing, and in the second half
of that year took a decided turn for the worse, which was
confirmed in 1904. All sorts of explanations may be in-
voked : in 1900 manufacturers blamed the growing habit
of the people of spending their money on seaside excursions
rather than on clothing, and in 1903 the bad weather, which
made people think that " anything will do for such a
climate." Yet in the fine weather of 1904 things went
from bad to worse : people even showed a tendency to*
refrain from buying whole suits, and to "be content with
odd garments simply." 1 It seems we must look for a
more comprehensive explanation, and again the price of
wheat furnishes some assistance. Even in 1899 it never
nearly touched the pre-i897 level, and thereafter, save for
checks in 1901 and 1906, it rose continuously till 1909.
As a result wool took but little part in the feverish activity
which marked the cotton and constructional trades in
I9°5-7-2 Only in 1906 (in which year, it should be noted,
the depression." Similarly, in 1898, " the lower kinds of fabric
have been the most difficult to sell." Nothing could show more
clearly the unequal incidence of high food prices on the different
classes of the community : nothing could show more clearly, in
view of the general feeling of depression that prevailed in the woollen
industry during these two years, that the predominant factor in
its prosperity is neither the demand from abroad nor that of the
speculative and investing classes at home, but the general consump-
tive power of the mass of the population.
1 This account of 1904 is confirmed from the cotton trade, where
' ' in the home trade buyers have held back owing to poor reports
from retail distributors : at the present time there is no buying
going on in home goods worth noticing."
2 The revival in 1905 was mainly due to the export trade, which
had already begun to improve in 1904 : the home trade in general
was far from brilliant, — in fact, says a Bradford report, " business
in the large centres of population has been most unsatisfactory."
The explanation of the comparative prosperity in 1905 of the Leeds
ready-made clothing trade, which appears on this occasion to have
been a less trustworthy barometer than usual, seems to me to be
that after a considerable period of retrenchment, culminating in
the pathetic state of affairs described in 1904, a moment comes
when a further postponement of expenditure on clothing is impos-
H4 STUDY OF INDUSTRIAL FLUCTUATION.
there was a relapse of wheat prices) do we find anything
that can be called enthusiastic accounts of the trade x :
in 1907 the impression is one of satisfactory, but not, con-
sidering the state of other industries, very exhilarating
activity. The comparative prosperity of 1908-9 indeed
finds no justification in the course of wheat prices ; but
their fall in 1910 doubtless contributed to the boom of the
trade in that year.2
The boot and shoe trade shows unmistakable traces of
the same influence. Thus after two years of good trade,
1895-6, manufacturers of leather in 1897 " were not so
happily situated as merchants, in spite of abundant supply
and reduced competition," for they had only " a variable
consumption of leather to depend on " ; and 1898 was said
to be less active than 1897. In 1899, however, trade was
very good, in 1900 fairly good and " likely to improve," and
in 1901-2 " fairly active," in spite of the handicap of
foreign competition. In 1903 and 1904, however, it was
bad, and at the end of 1905 we learn that " stocks of boots
and shoes at the present time have accumulated, indicating
clearly the poverty of the people generally." The trade of
1906 was " unquestionably disappointing," and that of 1907
only " fairly active," though with " an output quite up to
the average." 3
sible, whatever the state of one's finances. Hence the appearance
of prosperity in the clothing trades in this year, an appearance
which is confirmed by the rise in the profits of ten leading drapery
companies from £363,000 in 1904 to £465,000 in 1905 (Ec., 1906,
p. 307), must not be too readily taken as evidence of a genuine
increase in the purchasing power of the community. Compare
Dr. Marshall's illustrations of the increased consumption of cotton,
in spite of famine prices, in 1864, and of the comparatively early
revival of the boot trade in the United States after the depression
of the seventies.
1 And in 1906 there were more failures in the drapery trade than
in any of the four following years.
2 The factors at work on the trade in the last three years are
too complicated to permit the effect of the continued fall of wheat
prices in 1911, the rise in 1912 and the fall in 1913, to be clearly
seen : cf. especially p. 74. The home consumption of cotton has,
however, been good throughout 1913 (Ec., Ap. 18, 1914).
3 A comparison of these comments with the export figures serves
CROP VALUES AND CONSUMPTION. 115
The per capita consumption of food other than wheat
is largely influenced by the same factor. Mr. G. H. Wood,
indeed, who has reduced to order1 the most important
data for the years 1860-1896, bases on them the conclusion
that the trend of consumption follows closely in an inverse
manner that of the Board of Trade unemployment percent-
age, which again, as is well-known, is affected principally
by the activity of the constructional trades. 2 It appears to
me, however, that the relation is somewhat more complex
to show that in this industry, as in that of wool, the home consump-
tion is of leading importance, and that its fluctuations afford very
similar evidence as to the conditions of general consumptive power.
Number of dozen pairs exported.
1895 . . . 674,620 1904 . . . 675,669
1896 . . . 745,018 1905 • • • 774,061
1897 . . . 642,264 1906 . . . 783,974
1898 . . . 638,920 1907 . . . 786,858
1899 . . . 603,498 1908 . . . 784,197
1900 . . . 630,244 1909 . . . 839,980
1901 . . . 678,543 1910 . . . 1,086,868
1902 . . . 788,838 1911 . . . 1,152,600
1903 . . . 773,858 1912 . . . i,393,096
It will be noted that exports show a minimum in the boom year
1899 and a large increase in the year of accumulating stocks, 1905.
(The figures include footwear of caoutchouc up to 1900, and of
" other materials " up to 1907, but this does not appear to affect
their trend.) It would seem, however, that the demand for boots
is still more urgent than that for clothing, whence the greater
prosperity of the trade in 1900-2 and again in 1908 (cf. note 2, p.
113). This is not contradicted by the earlier revival of clothing
in 1905-6, for the consumption of boots on a large scale having sur-
vived that of wool by some three years, the average state of decay
of the boots being worn in 1905 would be less than that of the
clothing, and so their renewal less urgent. In recent years the
trade has been handicapped in taking advantage of good consump-
tive power by the high cost of material, — alleged to be exceptionally-
important in this case. " The greatest difficulty has been experi-
enced in obtaining highest prices from the public, which has been
accustomed for years to pay a fixed price for its boots " (Ec. Hist.
of 1913). While low food prices may lead to a demand for more
boots at existing prices, they will not apparently induce acquiescence
in a higher level.
1 Stat. Jour., 1899, p. 639.
2 Thus also Mr. Bowley, in commenting on Mr. Wood's paper,
suggested that consumption " appeared to have moved with wages
from 1860 to soon after 1870, but after that to have left wages and
moved with employment."
n6 STUDY OF INDUSTRIAL FLUCTUATION.
than this. Mr. Wood's general index is not in all respects
suitable for our present purpose * : I have therefore drawn
a curve of the per capita consumption of six important
articles of food (tea, sugar, meat, tobacco, currants and
raisins, and coffee), using Mr. Wood's indices up till 1896,
and indices 2 with the average consumption of 1896-1905
as base for 1896 onwards.3
Other aspects of these curves must be discussed later :
in the present context it is sufficient to note that they are
adversely affected by rising wheat prices in 1877, 1880,
1897, 1902-3, 1905, 1908 and 1913, and favourably by
falling wheat prices in 1874-6, 1879, 1881-6, 1894, 1899,
1901 and 1911. It should be noted that while on some
1 In the first place it includes figures of the consumption of raw
cotton and wool, which are indices rather of manufacturing activity
than of ultimate consumptive power ; in the second place it includes
certain kinds of alcoholic drink, the consumption of which seems
to depend partly on rather peculiar psychological factors, and
which it is therefore desirable to treat separately.
2 The figures are from the Statistical Abstracts, except for meat,
which are from the Abstract of Labour Statistics, and based on the
Report of the Dep. Com. on Combinations in the Meat Trade, 1909.
The meat figures are for years ending May 31, and I have been obliged
to assume that half of the consumption indicated took place in
each of the calendar years involved. I originally included also
cocoa : but the great rise in consumption in 1897 and again in
recent years seemed to introduce a disturbing element out of all
proportion to the importance of the article.
3 He takes as base the average consumption in 1870—79. I have
not thought it worth while to translate one curve into terms of the
other, as the object is to show not the trend over a long period,
but the fluctuations between neighbouring years. I have attributed
equal weight to the indices, since such a proceeding, though no doubt
more perilous with such a small number as six commodities than
with Mr. Wood's fourteen, does not seem likely to lead to serious
error : in some cases, however, such as the currant shortage of
1900, we must be on our guard. Moreover, the figures are not so
valuable for year to year comparisons as for showing the general
trend over a number of years, owing to the existence of stocks and
for other reasons. In particular, Major Craigie issues a warning
against such a use of the meat figures, especially as between 1892
and 1893 (cf. Stat. Jour., 1899, p. 672) : nevertheless an appreciable
diminution in 1893 is indicated by Mr. Hooker's figures as well as
Mr. Wood's (i$th Abstract of Labour Statistics, p. 50).
CROP VALUES AND CONSUMPTION. 117
occasions food and clothing share the favours of the eman-
cipated purchasing power, on others (notably in 1910 and
1911) they are rival suitors for -ts hand.
In America also there are evidences of the influence of
cheap grain upon certain other kinds of food consumption.
The curve of consumption of sugar shows a movement
divergent 1 from that of the home production of the relevant
season in the fiscal years 1887 to 1890, 1892, 1902, 1908-10.
In all but three of these years the movement is the same as
that of the coffee-curve, and needs therefore, no separate
discussion 2 : but the curious increases in sugar consumption
in 1887 and 1890 and the still more curious collapse in 1902
(after big home and foreign yields and in a year of appar-
ently all-round prosperity), must, I think, be brought into
relation with the variations in the price of maize 3 and of
beef.4 The consumption of sugar appears to be supple-
1 Only years in which this occurs are relevant, since the esti-
mates of "consumption" (Stat. Abs. U.S.A., 1912, p. 504) are
arrived at simply by adding net imports to home production. I
omit the years 1894-6 and 1897-8 in which the form of the curve
is clearly due in part at least to the Wilson and Dingley tariff re-
visions (the average ad valorem rate was 14 per cent, in 1893-4,
39*77 Per cent, in 1894-5, 42-07 per cent, in 1896-7, 77-47 per cent,
in 1897-8: ibid. p. 500).
2 Cf. p. 104.
3 Cf . the following figures (in cents per bushel) :—
Average Export Average Export
Home Price. Price. Home Price. Price.
1885 • • 54'° 51'0 1892 . . 55-0 54-0
1886 . . 49-8 52-2 1893 . . 53-0 49-9
1887 . . 47-9 48-7 1894 . . 46-0 50-9
1888 . . 55-0 59-2 1895 . . 53-0 47-7
1889 . . 47-4 43-7 1901 . . 46-0 56-7
1890 . .41-8 48-1 1902 . . 68-0 68-4
1891 . . 57-4 70-4 1903 . . 54-0 57-2
The average home prices (till 1889 from Babson's Business Baro-
meters, pp. 84-5, from 1890 New York prices, Stat. Abs. U.S.A.,
1912, p. 549) are for calendar years, and are sometimes (notably
1890) clearly influenced by the harvest of the current rather than
the previous year. The export prices, though a less satisfactory
index of cheapness to consumers, have the advantage for our present
purpose of being calculated for fiscal years.
4 There is a close connection between the price of the raw
material maize and that of the finished product beef.
Cf . the following figures : —
Ii8 STUDY OF INDUSTRIAL FLUCTUATION.
mentary rather than rival to that of corn and meat.1
There is in this country one more important kind. of
consumption, — that of house-room, which is largely affected
by the level of wheat prices. But the connection between
the price of wheat and the activity of the building trade is
obscured by several circumstances. In the first place most
of the indices available until quite recent years (cf. Chart
XIII.) have been of building in general and not of house-
building in particular. Thus the high prices of brick and
timber and the active employment in carpentering in such
years as 1874, 1899-1900, 1906 2 and 1910-1 seem to have
been due to the activity of other trades, notably ship-
building and cotton,3 as much as or more than to house-
Beef extra mess in
N.Y. market.
$973 per barrel.
$9'32
$11-75
$9-03
(Stat. Abs. U.S.A.,
1912, p. 549.)
(Aldrich Report Index, price of
1860 = 100.)
Beef, Salt. Beef, Fresh Roast.
1893 .... 100-3 . 99-0
1894 .... 98-9 - 98-3
1895 99*6 . 98-6
(Bulletin of Bureau of Labour (1910) Index, average 1890-9 = 100.)
1 If this surmise is correct, we may admit a genuine increase in
sugar consumption in 1894, and again in 1903 : cf. above figures.
The assumption in the text that the home demand for beef is in-
elastic does not conflict with our hypothesis (chap. vi. § i sub fin.}
that the world's demand is elastic. The figures of volume and
value of the corn crop for 1887-8-9, 1894-5-6, 1901-2 (cf. Chart
VI.) suggest that in so far as it is put upon the market the demand
for it may very probably be inelastic.
2 The price of brick, however, fell in 1906 (owing partly to the
competition of structural steel), and there were numerous failures
in the timber trade, which declined from September.
3 An important client to the timber trade : cf. Economist, 1909,
i. no. Thus also Mr. Dearie (op. cit., p. 25) attributes the greater
prosperity of the trade in the north than in the south of England
in recent years to the cotton and iron booms.
Calendar
Salt
Loins
Calendar
Year.
Beef.
of Beef.
Year.
1885 .
- 957
XII-I
1900 .
1886 .
. 82-6
105-6
1901 .
1887 .
60-9
100 -0
1902 .
1888 .
. 65-2
1 00-0
1903 .
1889 .
60-9
III-I
0
1890 .
- 56-5
105-6
1891 .
60-9
105-6
CROP VALUES AND CONSUMPTION. 119
building.1 But secondly, even the movements of the house-
building trade do not always follow closely upon the move-
ments of general purchasing power. For, first, the hire,
or still more the purchase for occupation, of a new or better
house represents a discontinuous advance of expenditure
not likely to be undertaken until there has been some
accumulation or at least some repetition of the increased
purchasing power.2 Secondly, building represents the
provision of the instrument of production, not of the
finished service, and its activity is not therefore always
a good gauge of the strength of consumptive demand : on
the one hand it may be speculatively prolonged, on the
other the first wave of emancipated purchasing power will
spend itself in the filling of houses already existing but
unoccupied. The net result is that while house-building
shows, for instance in 1874-6 and 1904-9, unmistakable
traces of the influence of wheat prices, its activity is some-
times, as in 1877 and 1897-8, carried past the rise which is
1 This assertion is confirmed by Table VI. The figures (pub-
lished quarterly in the Labour Gazette from returns made by local
authorities of the estimated cost of the several kinds of buildings
for which they passed plans in each quarter year) suffer somewhat
from not being continuously comparable, owing to variations in
the number of authorities reporting, and from being estimates of
value and not volume of consumption : but one conclusion is clear.
While the figures for factories, etc., show a rise over the corre-
sponding quarter of the preceding year in ten cases out of eleven,
those for total building show a rise in six cases, and those for dwell-
ing houses only twice. The rise in the first quarter of 1911 is
scarcely significant, and it is only in the latter part of 1912 that
housebuilding can be said to have begun to come into its own.
The variations between successive quarters are probably deter-
mined in part by seasonal variations in the rates for money, and
in the weather conditions for starting work ; whence the tendency
to fall in the third quarter.
2 There seems some reason to suspect a trans-cyclical change in
this respect since about 1870 : in the fifties and sixties the build-
ing trades, especially through the channel of the working-class
building societies, were more successful in reaping the firstfruits
of expanding consumptive power than they have since been. The
Land Committee, however (II. 9, i. fL), deny the existence of any
signs of permanent decline in the desire to own one's own house.
120 STUDY OF INDUSTRIAL FLUCTUATION.
fatal to other kinds of consumption ; and it is often late, as
in 1887, 1896 and 1912, in profiting by the fall.1
It only remains to observe that, as might be expected
from the above analysis, the furnishing trades show a rather
closer affinity with the other kinds of consumptive industry.
Thus in 1894, 1899, 1910-1 2 they seem to have benefited
sooner 3 than the building trade by the movement of wheat
prices, while on the other hand in 1897 they were hit more
quickly.4
1 A slight check to the decline in activity in 1902 may also be
referred to the fall of wheat prices in 1901.
2 And perhaps (to judge from imports but not from employ-
ment) in 1883-5 and 1901.
3 It is interesting to contrast America, where, building being more
largely anticipatory and speculative, the curve of furniture prices
is apt to start its pronounced rise one year later than that of build-
ing material prices, e.g., 1899, 1906, 1910.
4 Similarly in America, the furniture index shows clearer traces
than that of building materials of the check to immigration in 1893,
1901 and 1904 (though not in 1889, nor in 1911 ; but n.b., the
earlier falling off in employment in 1910 (Table IX.) and the good
consumptive conditions of 1911 (p. 177, n.).
PART II.
Fluctuations of General Trade.
CHAPTER I.
THE REVIVAL.
§ i. FALSE METHODS .OF APPROACH.
IN the preceding discussion we have been frequently met
by indications that the alleged universality and simultaneity
of the so-called general fluctuations of trade are in part a
figment of the public imagination, — that the vicissitudes
of the several industries and groups of industries are in no
small measure governed by their own individual idiosyn-
crasies of supply and demand. Nevertheless, we are ap-
parently not in a position to deny the existence of something
which may fairly be called an alternation of general ex-
pansion and depression, and of which we have not as yet
succeeded in giving a comprehensive explanation. That
is the task of the present book.
The task is no light one, but it can be simplified to some
extent by certain expedients. First it will be convenient
to neglect for the present such essential modifications of
the course of events as arise from the existence in modern
industry of a monetary mechanism and a system of wage-
labour. Secondly, it will be desirable to follow the practice
of most writers in breaking at some point arbitrarily into
the magic circle of industrial change, and to consider first
the birth of " expansion," and pass on afterwards to the
" crisis " or turning-point and thence to the " depression "
of trade. Thirdly, it will be desirable at various points to
121
122 STUDY OF INDUSTRIAL FLUCTUATION.
clear the ground for further advance by the demolition of
certain popular but misleading arguments which encumber
the path of true analysis.
Let us begin then by attempting to account for the genesis
of an expansion of general trade. And here at the outset
we are met by the necessity for a preliminary consideration
of the explanation which in one form or another does duty
in the majority of writings upon this subject. An expansion,
according to this theory, beginning in some one or more of
the consumptive trades, is communicated by the increased
demand for instruments to constructional industry ; and
is thence spread, by means of the demands of those engaged
in constructional industry, among the other trades making
for consumption. An objection to the first limb of this
theory is sometimes based on the admitted fact that the
fluctuations of constructional industry are disproportionately
large compared with those fluctuations of consumptive
industry of which they are alleged to be the effect. This
objection, however, appears itself to be open to a threefold
answer.
In the first place it is clear that in some industries the
magnitude of the changes in consumers' demand is apt to
be under-estimated. This is particularly true of those
" public utility services " alluded to at the beginning of
Part I. ch. ii. § I.1 In such industries new investment,
though clearly justified by consumptive demand, is not
heralded by a rise in price, because, the service in question
being strictly localized, it is ex hypothesi entirely non-ex-
istent before the new investment takes place, 2 and the only
evidence of the demand for it is the general economic con-
dition of the community in question.
Secondly, M. Aftalion points out 3 that a small pro-
1 This fact is alluded to by M. Aftalion (Journal d'Economie Poli*
tique 1909, pp. 215 ff.), though he does not note the explanation.
2 There is of course an important exception in the case of the
doubling of the track in response to an increased volume of traffic
over a particular route : though even here the demand factor is
apt to be under-estimated, owing to the relative fixity of railway
charges caused by legal restrictions. 3 Ibid., p. 219.
THE REVIVAL. 123
portionate change in the demand for consumption goods
may produce a large proportionate change in the demand
for construction goods : for instance, supposing 10 per cent,
of the annual output of boots is produced by new machines,
a rise of 10 per cent, in the demand for boots will produce
a rise of 100 per cent, in the demand for new machines.1
Somewhat similarly Professor Carver suggests2 that when
working expenses are relatively very large, a small percentage
addition to the value of the product will make a large
percentage addition to the value of the fixed plant, and
therefore to the demand for new construction materials.
This consideration indeed is ex hypothesi most important
in those industries in which fixed capital is relatively small, 3
i.e., probably in those whose demands are of the smallest
absolute significance to constructional industry. Before,
however, refusing to admit its importance, we must remember
that the actual demand for instrumental goods is by no
means the same as the warrantable demand, least of all
in industries where the fixed capital is large : hence in such
industries an increase in consumers' demand may still be
the initial though by no means the sole cause of a larger
proportionate increase in the demand for instrumental
goods.
Thirdly, even if the effect upon constructional industry
of variations in the prosperity of any single consumptive
1 It is not, however, clear that a fairly steady policy of renewals
on this scale always implies that existing plant is working at maxi-
mum capacity, and will not be able to supply some at least of the
new product demanded.
2 Quarterly Journal of Economics, 1903—4, p. 497.
3 It is true that if, as is likely, the fixed capital is not only large
but indivisible and intractable (cf. Part I. chap. ii. § i), the con-
tribution made in time of depression to general costs is likely to be
small, and a given increase in total receipts to make therefore a
large proportionate addition to the value of plant : but the import-
ance of this consideration is minimised when we reflect that its
operation is likely to be most marked when the demand for the
product is inelastic, in which case a given elevation of demand is
likely to produce a comparatively small increase in total receipts ;
and that in any case the effect will only be the emergence of some-
thing more like the anticipated return on invested capital, which
will not constitute a rational incentive to further investment.
124 STUDY OF INDUSTRIAL FLUCTUATION.
trade be relatively small, it is often urged that the need
for construction goods being fundamental to all and not
merely to one consumptive industry, the variability of the
demand for them is naturally likely to be greater than that
of the demand for any class of consumptive goods.1 This
proposition indeed as it stands does not seem to be self-
evident. On the contrary, if there is no reason to suppose
that the variations in the several consumptive industries
are interdependent, it ought clearly to be probable that
the variability of constructional industry as a whole will
be less than if it was dependent on any single client ; 2 and
it ought not to be improbable that the excess of the vari-
ability of that part of it which is devoted to satisfying any
one client over the variability of the whole will be greater
than the excess, if excess there be, of the variability of
that part which is devoted to satisfying one client over
the variability of that client ; so that the variability of
constructional industry as a whole might well be expected
to be less than that of any single consumptive industry.
It follows that this explanation of the large variability
of constructional industry would seem to be invalid, unless
we have reason to suppose a considerable measure of positive
correlation between the variability of the several con-
1 Thus M. Lescure, while in the main finding the origin of fluctua-
tion in the instrumental trades, appears to recognise a subsequent
and supplementary cause of the prosperity of constructional industry
in the demand for materials by the consumptive trades, a reaction
the magnitude of which he explains in the following words : " Dans
une societe capitaliste progressive, il existe en effet une branche
de production particulierement raise a contribution c'est 1'indus-
trie des moyens de production. Les commandes de toutes les
industries se concentrent dans ses usines " (op. cit., p. 503).
2 This thesis is a particular application of the general proposition,
of vital importance throughout economic science, that " the pre-
cision of an average is proportional to the square root of the number
of terms which it contains " : cf. Pigou, Wealth and Welfare, p. 141.
But the Professor, after pointing out the relative stability of raw
material industries, with a large number of outlets for their product,
as compared with the finishing industries with specialised markets,
concludes rather curiously (p. 143) by lumping together iron and
coal with sugar as industries which might be expected to have a high
variability.
THE REVIVAL. 125
sumptive industries. Though, however, it has appeared
in the preceding part that the amount of positive correlation
is considerably less than is often supposed, the probability
of such correlation (whether due to community of agri-
cultural markets or of financial environment or what not)
cannot, so long as we follow the ordinary lines of popular
arguments, be denied. The first limb of the theory under
discussion appears therefore to survive unscathed the prima
facie objection based on the relative largeness of the fluctua-
tions of constructional industry.
Both limbs of the theory are, however, open to a more
fundamental criticism. Does not the whole notion of the
communication of prosperity from one industry to another
in an endless chain imply an elementary confusion of
thought ? Is not the prosperity of the original group of
consumptive trades accompanied by a rise in the exchange
value of their products, involving a fall in the exchange
value of the products of other trades which cancels the effect
upon constructional industry ? Again, does not the sub-
sequent alleged "repercussion" from constructional indus-
tries upon the other consumptive trades merely mean that
those trades which minister to the special needs of con-
structional industry are enabled in the long run to benefit
at the expense of those of which the fall in the value of the
products originally upset the equilibrium ? The whole
theory of repercussion, as expounded so glibly for instance
by M. Lescure, appears to be engaged in making something
out of nothing, and to rest upon a quicksand. We are
forced therefore to penetrate somewhat deeper in our
attempt to analyse the nature of a revival of general trade.
§ 2. THE GROWTH OF PRODUCTIVITY.
The best method of analysis seems to be as follows. An
increased prosperity, in the sense of an increased volume of
consumption, in any producing group can only, in the
absence of miscalculation, result from a rise in the pro-
ductivity, in terms of satisfaction, of the effort expended
by that group. Such a rise may come about in three ways,
126 STUDY OF INDUSTRIAL FLUCTUATION.
— by a rise in the productivity of their effort in terms of
the goods which they themselves produce, by a rise in the
exchange value of their product, or by a rise in the satis-
faction afforded them by the goods for which they exchange
their products. We have to consider these three possibilities
when the producing group concerned consists of industry
as a whole.
The first kind of rise may clearly occur simultaneously
in the totality of trades through a general lowering of the
real costs of production. It must be clearly understood
that we have to deal here not with such a lowering of the
cost of materials, labour, etc., as involves a mere transference
of wealth from one group of producers to another, but with
a real lowering of general costs in the sense of an improved
organisation and equipment of man in his fight against
natural obstacles. Now assuming a condition of " de-
pression " it is pretty clear that there are forces at work
tending to such a general reduction of real cost. Thus
M. Tugan-Baranowsky1 gives a number of citations to
show that the great technical improvements in the cotton
industry have been made during times of depression. Simi-
larly M. Aftalion2 contends that during a depression
" nombre de remaniements s'operent dans 1' organisation
generate de la production, dans le choix des matieres
premieres, dans Tutilisation des sous-produits, dans la
division du travail entre ouvriers, dans la specialisation
meme des etablissements. Une guerre plus vigoureuse est
faite au gaspillage, au coulage, aux frais inutiles." 3 Thus
also Dr. Marshall points out that the average competence
1 Op. cit., p. 312. 2 Op. cit., vol. ii. p. 228.
3 So also Mr. Burton notes that production does not decrease
in an equal ratio with employment, because of the introduction
of improved methods under the stress of hard times (Crises and
Depressions, p. 138) : and M. Lescure insists (op. cit., pp. 422-3)
that the " crise de production et de consommation " is the signal
for a " thorough revision of the costs of production," which includes
among other things an extended use of the devices of concentration
and integration ; and attributes the prolongation of the American
boom beyond 1900 to the unwonted adoption of this policy in
times of rising prices. Cf. also Mitchell, Business Cycles, pp. 563-4.
THE REVIVAL. 127
of business men is likely to be greater in times of depression.1
There can be no doubt that industry possesses a consider-
able automatic recuperative power of this kind : moreover,
it should be carefully noted that an increased prosperity
arising from this cause in any single group of trades is not
necessarily secured at the expense of any other group, and
so may legitimately, if the field over which it extends be
sufficiently wide, be regarded as an explanation of a con-
structional and a general revival. In particular the English
booms culminating in 1890 and 1900 seem to have been in
no small measure due to such an automatic recovery of
general home industry. ' There is evidence," says the
Economist in its review of 1889, " that our home trade
developed more rapidly than did our trade with foreign
countries. . . . Our exports of cotton goods last year
were about 2| per cent, less than in 1888, but our consump-
tion of cotton in the production of manufactures and yarns
was full i6J million Ibs. greater. Our exports of pig-iron
increased by 163,000 tons and our home consumption by
857,000, and by our home market also a large proportion
of our increased production of woollen goods was absorbed."
Similarly in 1890, " the trifling curtailment of our foreign
trade has been much more than counterbalanced by an
expansion of home trade."2
1 " Those who have control of our loan fund are much more
careful about their loans when prices are falling, and . . . are
saved from lending to people who had the capacity for riding upon
the surface of the rising tide, but who had not the capacity for
contending against the stream." (Evidence before G. and S. Com-
mission, C 5512-1, 1888, Q 9816.)
2 Cf. also the evidence of a rapid growth of productivity fur-
nished by the moderate course of non-constructional prices : —
B. of T. Index. Raw Textiles. Food and Drink. Miscellaneous.
1888 70-0 81-8 71-0
1889 72-4 81-1 74-3
1890 . . 72-9 80-6 72-9
and by the progress of the gross receipts of railway goods traffic : —
i m.
1885 . ... 36-8
1886 . . ^6-4 — I -I
'887 . . . 37-3+2-5
1888 . . . . 387 + 3.8
1889 4i.I + 6.2
1890 42.2 + 2-7
128 STUDY OF INDUSTRIAL FLUCTUATION.
Again in the next period already at the end of 1894
" there were signs of improvement " in the ready-made
clothing trade " and happily the forecast then made was
realised by the time the New Year was not more than a few
weeks old." * In 1896 " manufacturers in the Longwood
and Slaithwaite districts, who are chiefly engaged in the
manufacture of goods for the home clothing trade, have
been very busy throughout the year " ; and " merchants in
the home trade record a year of increased properity." " One
of the most, notable features of the iron trade of 1897,"
says the report of the British Iron Trade Association, "has
been that an increase of considerable amount in the output
of finished iron and steel — altogether 376,718 tons, fell
concurrently with a diminution in the construction of
shipping tonnage to the extent of 207,000 tons. This
decline was to some extent compensated for by the increased
exports for the year, amounting to 140,667 tons, but this
advance is so far from explaining the difference that it must
be assumed there was a much greater home consumption,
all the more so that the increased exports were mainly pig-
iron." 2 Similarly in 1899 there is evidence that the advance
in the price of pig-iron was gradual, and due to the demand
of a number of small home industries.3 Nevertheless, in
view of our enforced opposition to the claim of variations
in cost in the ordinary sense to furnish a complete explana-
1 In the cotton trade also in 1895 the " home trade houses were
understood to have had a not unsatisfactory year," and goods
retained for home consumption rose from 249,653 to 271,900 bales.
2 The advance of general home industry in this year is all the
more significant in view of the obstacle presented by high wheat
prices to a rise in consumers' demand. In 1894-6 the two factors
had worked in co-operation.
3 So also the improvement in pig-iron employment in the latter
part of 1909, though due partly to the American spurt, was also
due partly, as was the gradual improvement in iron and steel and
engineering more largely to the gradual growth of demand from
a number of miscellaneous home industries. The increased con-
sumption of food, during the continuance of more or less pronounced
constructional depression, in England in 1878, 1887, 1894, 1904
and 1909, in America in (fiscal years) 1895, 1897 (but cf. note 3,
p. 104), 1905, 1909 may be adduced as additional evidence of the
automatic recovery of industry.
THE REVIVAL. 129
tion of the revivals of particular industries, we may well
be on our guard against allowing such reductions of real
miscellaneous costs the sole or even the predominant place
in determining the revivals of industry as a whole.
§ 3. HARVESTS. THE ELASTICITY OF DEMAND. .
We are led on therefore to consider, with reference to
industry as a whole, the second possible cause of increased
prosperity in any group — a rise in the exchange value of
its products. It should hardly be necessary to insist that
a simultaneous rise in the exchange value against one
another of the products of all trades is a vacant and mean-
ingless phrase. But a rise, due to an increased bounty of
nature, in the exchange value of the products of industry
as a whole against the products of agriculture is clearly by
no means open to the same objection.1 There is, however,
a widespread reluctance among modern writers 2 to admit
its importance as a cause of general revival. This reluctance
seems to be founded partly upon certain irrational habits
of mind, and partly upon more coherent, if not conclusive,
argument. Under the former head we may class, first,
a tendency to infer that if the particular theory of periodicity
connected with the name of Professor W. S. Jevons is proved
invalid, the whole theory of crop influences breaks down
together with it : 3 secondly, the assumption that a lack
of positive correlation between agricultural and other prices
disproves the existence of any causal connection : 4 and
1 This phenomenon might indeed be classified as a reduction
in real costs in one important branch of " industry " ; but the treat-
ment adopted seems to be on the whole more convenient.
2 Mr. Hawtrey is among the most dogmatic ; cf. Good and Bad
Trade, p. 87. " Except in the special instance of an actual famine,
the fluctuations of demand or supply [of corn] produce no fluctua-
tions in trade as a whole."
3 Cf. in particular Tugan-Baranowsky, op. cit., p. 239.
4 Cf. ibid., p. 237. " La comparaison du prix de pain avec
les phases du cycle Industrie! montre avec evidence complete qu'il
130 STUDY OF INDUSTRIAL FLUCTUATION.
thirdly (as in the somewhat analogous case of the " quantity
theory " of money), an excusable resentment that the
advocates of the crop theory have not been at sufficient
pains to explain in detail the method of its operation.1
These nebulous objections need scarcely detain us ; but
the two more definite arguments deserve more careful
consideration. The first asserts that the effect upon general
industry of the contemplated change in the ratio of exchange
would in any case be nil : the second asserts that supposing
the possibility of such an effect, the actual changes in the
ratio of exchange under modern conditions are so small as
to make that effect negligible.
The first argument 2 may be briefly expressed thus.
Suppose there is an abnormally large wheat crop : the
elasticity of demand for wheat is either one or greater than
one or less than one. In the first case the distribution of
purchasing-power will remain unchanged : in the second
case there will be a transference of purchasing-power from
consumers to the producers of wheat, and in the third case
from the producers to the consumers. But in any case
what is lost by one party is gained by the other, and there
is no change in the aggregate volume of purchasing-power
which forms the source of demand for the products of other
industries; there is no reason, therefore, why an increased
output of wheat should lead to any increase of activity in
industry as a whole.
One cannot be surprised that this objection should be
made when one reads such a passage as the following. In
ne saurait etre question de dependence causale entre ces deux
ordres de phenomenes." Still more oddly Aftalion, having re-
marked that high food prices are sometimes correlated with indus-
trial collapse, naively concludes : " Manifestement, 1 'agriculture ne
joue qu'un role de second ordre en notre sujet " (op. cit., Vol. I. p. 25).
1 Thus Professor E. D. Jones (Economic Crises, p. 150) com-
plains that Jevons seems unable to make up his mind whether it
is large crops, by encouraging unrestrained speculation, or small
crops, by producing a dearth, that precipitate a crisis. " The
harvest theory appears to be holding to both horns of the dilemma."
2 Of which neither Mr. Hobson nor Mr. Hull has been innocent.
THE REVIVAL. 131
the event of a deficient crop, says Mr. Burton,1 " the con-
sumer must devote a larger percentage of his expenditure
to the purchase of. necessaries, and will have a smaller
percentage remaining for conveniences and luxuries, the
demand for which is accordingly lessened. On the other
hand the farmer receives no more for his products, for what
he gains by increased price is offset by the decrease of the
quantity which he has for sale." Mr. Burton is indeed
trying to make the worst of both worlds.
The difficulty arises, in my opinion, from a failure on
the part of opponents of the crop theory to realise that
the resources which the consumers of corn 2 are prepared to
expend on commodities in general (including corn) are not a
fixed fund but a flow : and from a failure on the part of its
advocates to realise that a mere alteration in the ratio of
exchange will not increase that flow unless it involves an
increase in the productivity of effort of corn-consumers in
terms of satisfaction, in other words unless the elasticity of
their demand for corn in terms of effort is greater than unity.
If, however, this is shown to be the case, it does inevitably
follow that the increased effort devoted to the acquisition
of corn is not all of it withdrawn from the acquisition of
other things, but is in part a net addition to the volume of
effort expended : in other words, not only is the indus-
trialists' consumption of corn increased, but so is their own
aggregate production, so that their consumption of indus-
trial products is not diminished by the whole amount by
which the agriculturalists' consumption of those products
is increased.
Conversely, if the elasticity of effort-demand for corn is
less than unity, there will be a decrease in aggregate indus-
trial production, accompanied indeed by an increase in
mutual industrial consumption.
The following supplementary propositions will commend
themselves on reflection, but will be rendered much clearer
1 Crises and Depressions, p. 76.
a For the sake of brevity this word is used throughout this dis-
cussion in its Ricardian sense of any agricultural product,
132 STUDY OF INDUSTRIAL FLUCTUATION.
by a reference to the annexed diagram.1 The change in
aggregate industrial production will be greater (i.) the
greater the change in the amount of effort devoted to ac-
quiring corn, (ii.) the faster the rate of change of the marginal
utility of commodities other than corn, (iii.) the slower the
rate of change of the marginal disutility of effort. The
last two factors make also for a small change in mutual
industrial consumption.
Units of effort are measured along OX, units of utility along OY.
UU± = curve of marginal productivity of effort in terms of satis-
faction derived from non-corn commodities.
EEj == curve of marginal disutility of effort.
OA = units of effort expended upon corn.
If the effort demand for corn is elastic, A is moved to the right
to A', the starting point of UUj is moved from the ordinate YtA
to the ordinate Y2A', the point of intersection of the curves is moved
from P to P', and the aggregate of effort expended is increased
from OM to OM'. If the effort-demand is inelastic A7 is moved to
A, and OM7 reduced to OM.
MM' will be greater (i.) the greater is AA', (ii.) the steeper is UUj
(since it is the interval UU' not U'V that is. given), (iii.} the flatter is
EEj. The last two factors also make for the smallnessof (AM-AW).
THE REVIVAL.
133
In estimating the effect of an increased production and
lowered exchange-value of corn, it is important therefore
to inquire (i) whether the elasticity of effort-demand for
corn is likely to be greater or less than unity, (2) whether
the factors determining the amount of the alteration in the
aggregate of effort expended are likely to operate in the
same way when the effort-demand is greater as when it is
less than unity.
i. This depends upon —
(i.) The rate of increase of the marginal disutility of
effort. At a time of slack trade this is likely to be slow.
(ii.) The rate of diminution of the marginal utility of
corn to the consumer. While with certain products and
certain bodies of consumers (for instance, with the majority
of Englishmen in respect of wheat) this rate is rapid, there
is reason to believe that with most products and taking the
world's consumers as a whole it is pretty slow. The general
conclusion is that the industrial demand for corn is on the
whole greater than unity.1
TTni+C
134 STUDY OF INDUSTRIAL FLUCTUATION.
2. (i.) As regards the first factor, the same considera-
tions which make it probable that effort-demand will be
elastic also make it probable that the inelasticity of an
inelastic effort-demand will be less pronounced than the
elasticity of an elastic one. Moreover there is a further
point : it so happens that two of the most important cases
in which there is reason to suspect an inelasticity of effort-
demand — the United States maize crop and the Indian
rice-crop — are cases in which the rate of diminution of the
marginal utility of the crop to the producer is slow, 1 and the
alteration therefore in the ratio of exchange due to a given
crop-increase comparatively small. In such cases the re-
strictive effect upon general production of the inelasticity
of effort-demand is pro tanto mitigated.
(ii.) There seems to me to be a general presumption
that the rate of diminution of the marginal utility of general
commodities will be greater when the effort-demand for
corn is elastic than when it is inelastic, for the following
reason. The current assumption of economists that a
utility schedule which is elastic in one direction from the
existing equilibrium-point is also elastic in the other seems
to me to be true only under extremely " non-frictional "
OA = curve of total utility derived from, expending effort on corn.
OB = curve of total disutility of effort.
OM = units of effort expended, the position of M being such
that PQ is a maximum. Then the slighter the concavity of OA
and the convexity of OB to the axis of X, the more likely is M1 to
lie to the right of M when OA is raised to OA1.
It must be observed that the accuracy of this generalisation
cannot be disproved in any particular case by demonstrating that
the money receipts of producers have been smaller for a large crop
than a small. The very increase of production instigated by the
large crop may, if the supply of currency has not kept pace, have
involved such a fall of non-agricultural prices that the smaller
money receipts indicate a larger volume of commodity receipts.
Professor Lehfeldt, however (Ec. Jour., June 1914, p. 217), estimates
that during the last twenty-five years the elasticity of the world's
demand for wheat has never exceeded 0-75. I am not competent
to appraise the methods by which he reaches this result.
1 The producers are also consumers on a large scale ; cf . Piatt
Andrew in Q, J. of EC., Vol. XX. p. 323, and R. of I. T., 1908-9, pp.
4-5-
THE REVIVAL. 135
conditions, and to neglect an important psychological fact.
Almost everybody is firmly convinced that while he could
very easily do with a great deal more money than he has
got he could only with great difficulty do with any less ;
for since most people's tastes and habits are based more or
less consciously on those of the people slightly above them
in the income-scale, they can be altered much more easily
to suit a rise than to suit a fall in spending-power. From
this fact — the importance of which I think has never been
sufficiently recognized 1 — it follows that those whose effort-
demand for corn is elastic and who are therefore compelled
to reduce the effort devoted to securing other things, will
not reduce it much : while those whose effort-demand for
corn is inelastic and who are therefore enabled to increase
the effort devoted to securing other things, are likely to
increase it a good deal. Moreover, since there is reason to
think that those whose effort-demand for corn is elastic are
as a class poorer than those whose effort-demand is inelastic,
the disinclination to reduce expenditure on other things is
likely to be especially great in their case.
(iii.) There seems primd facie ground for supposing that
the comparative poverty of those groups whose effort-
demand is elastic is due in part at least to an exceptional
restriction of their business activity, and that the rate of
change of the marginal disutility of effort will therefore be
slower with them than with those whose effort-demand is
inelastic.
The general conclusion then is that the aggregate volume
of effort expended, in other words of production, is likely
to be greater if the effort-demand for corn is elastic than
if it is inelastic ; that there is reason to think that on the
whole it is elastic ; and that the increase in general pro-
1 Since a first writing of the above, I find current conceptions
about the utility of income challenged by Professor Chapman in
the Economic Journal, March, 1913, pp. 36 ft. ; but whereas he
argues that certain people live at times on what he calls a " western
slope," I am only claiming that they live on an " eastern slope "
with a steep hill on one side of them and a gentle decline on the other.
136 STUDY OF INDUSTRIAL FLUCTUATION.
duction accompanying a given increase in the commodity-
receipts of corn-producers is likely to be greater than the
diminution in general production accompanying an equiva-
lent decrease. For instance, supposing the commodity-
receipts of producers are increased exactly as much by
Hemichoros a of consumers as they are diminished by
Hemichoros /3, there is reason to suppose a net increase
of general production, and also a net increase in the mutual
consumption by the industrial population of each other's
products. It only remains to inquire whether this con-
clusion is modified in any way by the consideration that
the volume of resources produced is not necessarily the same
as that utilized in exchange and consumption.
Now in the first place the alteration in the ratio of
exchange is likely to affect not only the expenditure of indi-
vidual effort but the utilization of accumulated industrial
stocks. Now in a time of slack trade, the rate of change
of the marginal utility of stocks is likely to be slow, and
consequently the elasticity of commodity-demand for corn
and the increase in aggregate expenditure to be large : 1
while on the other hand, if the elasticity of commodity-
demand for corn should by any chance be small, it will argue
a rapid rate of change in the marginal utility of stocks, and
consequently a smallness in the diminution of aggregate
expenditure.
Secondly, while the arguments which led us to believe
that a given volume of resources will have a greater effect
upon constructional industry in the hands of corn-producers
than of corn-consumers, apply with less cogency to the effect
upon industry in general, there seems some reason to suppose
that, except in certain cases where the agricultural popula-
tion is unusually devoid of economic freedom and dominated
by custom in its habits of expenditure, resources will on the
1 It is indeed true that the actual owners of stocks being as a
rule comparatively wealthy, their demand for corn for their own
consumption is likely to be inelastic ; but since the corn is transfer-
able to those whom they employ, this consideration is irrelevant.
Contrast the conclusions of chap. iii. § i with regard to transport
and constructional goods.
THE REVIVAL. 137
whole be less likely to be hoarded when concentrated in the
hands of corn-producers than when scattered in the pockets
of corn-consumers.
The general conclusion is therefore confirmed, that in
certain circumstances the total volume of industrial activity
may be diminished by an increase in the volume of crops,
but that it is not likely in any case to be diminished much,
and is more likely on the whole to be increased.1
1 Professor Pigou indeed (Wealth and Welfare, Part IV. chap, v.)
reaches by a more direct route the commonsense conclusion that
an increase in the volume of crops is likely to increase the volume
of general business ; but I cannot regard his argument as altogether
valid. Approaching the problem from the point of view of the real
income of the working-classes, he asserts that an increase in the
inflowing stream of natural resources will (in the absence of any
reason to suppose an alteration in the proportion of resources
devoted respectively to storage, consumption by the legal capitalist
owners and the " purchase of labour ") increase the commodity-
income of the working-classes. Now it is perfectly true that it
increases their consumption of corn ; but the further inference
that it increases their aggregate production and consumption of
commodities in general rests on the assumption, tacitly made by
Professor Pigou, that the elasticity of their labour-demand for
corn is greater than one — in other words that the offer of an
increased income of corn will in fact elicit from them a greater
aggregate of effort.
In a correspondence with the present author, Professor Pigou
explains that his analysis is intended to refer to a " homogeneous "
society, from which the phenomena of exchange and the specialisa-
tion of productive function are absent. The former phenomenon,
he says, may affect the direction, while it does not affect the volume,
of the increased " investment in labour " ; while the hindrances
to mobility imposed by the latter may make the increase less than
it would otherwise have been. But the initial effect is " prior to
and independent of " these secondary phenomena.
I cannot regard the argument as convincing. Professor Pigou 's
use of the phrase " the purchase of labour " indicates quite clearly
to my mind that in his published discussion he is, half unconsciously,
taking into account the most fundamental aspect of the phenomenon
of exchange. But as soon as this phenomenon is recognised, the
comparative elasticities of demand for the things exchanged become
obviously a matter not of secondary but of prime importance in
considering the effects upon production and consumption.
The extreme importance of this conclusion both to this and to other
parts of the theory of fluctuations, and the reluctance with which
one differs from such an authority as Professor Pigou upon a point
of analysis, must be the justification for the length and obscurity
of this note.
138 STUDY OF INDUSTRIAL FLUCTUATION.
§ 4. HARVESTS. THE LAW OF COMPENSATION.
We have still, however, to meet the second argument
addressed against the notion of the dependence of industry
in general on crop conditions — the argument which is
founded on the relative smallness of agricultural fluctuations
as compared with those industrial changes which they are
invoked to explain. It may be noted indeed that in certain
forms * this argument seems to be something of an ignoratio
elenchi. The whole art of mechanical engineering is a study
of how, by the economy of force, to produce great changes
of position out of small. A small change in the position of
the piston is able, by means of a system of wheels and cranks,
to effect a large change in the position of the locomotive.
To say that because changes in the volume of crops are
small compared to changes in the volume of industry there
is therefore no causal connection between them, is like saying
that because the force necessary to wind up a motor-car is
small compared to the subsequent activity displayed by the
car, there is in this case also no relation of cause and effect.
Nevertheless it is clear that if the magnitude of agricultural
fluctuations can be shown to be absolutely so small as to
be practically negligible, we shall be wrong in attributing
to them any important influence on the volume of industrial
activity. Now it is urged with much plausibility that in
earlier times the importance of variations in harvests may
indeed have been great ; but that under modern conditions
of world markets and cheap transport and storage so strong
a law of compensation is at work that the variations in the
available supply are no longer sufficient to exercise any
perceptible influence upon industrial conditions.
The forces of compensation appear to work in three
ways — through time, through space, and through the
character of the crop. With regard to the first : the
evidence is clear that the total supply of any crop available
for consumption does not vary so much from year to year
as does the annual harvest ; and it appears probable that
1 Notably in the hands of Mr. Hull (op. cit.t pp. 43-4).
THE REVIVAL.
139
with the better organisation of markets the tendency towards
an equalisation of consumption is on the increase.1
In this connection the following figures of the world's
wheat supplies are of interest : —
World's
Wheat
Crop.
M. Qrs.
Visible sup-
plies of Wheat
and Flour
on Aug. i.
Available for
consumption
in Cereal Year
(Aug. i-
July 31)-
Consumed
in Cereal
Year.
1894
326-9
21-7
348-6
328-9
1895
305'3
197
325-0
309-5
1896
298-4
15-5
3I3-9
304-3
1897
286-4
9-6
296-0
187-3
1898
366-1
87
374-8
358-1
1899
328-4
167
345-1
326-4
1900
332-9
187
351-6
335-1
1901
348-0
16-5
364-5
352-8
1902
396-9
117
408-6
397'6
1903
393-3
II-O
404-0
382-2
1904
362-9
21-8
384-7
364-9
1905
406-6
19-8
426-4
409-9
1906
414-9
16-5
431-5
412-0
1907
368-9
iq-4
388-3
376-I
1908
386-7
12-2
398-9
389-0
1909
451-2
9-9
461-1
447-6
1910
442-8
13-5
456-3
436-8
1911
426-0
19-5
445-5
429-8
1912
441-9
157
457-6
439-8
I9I3
483-9
I7-8
5017
1 For instance, " formerly the practice in Argentina was to ship
practically the whole surplus in the first seven months of the Argen-
tine cereal year ; for instance, in 1908 out of a total of 17,400,000
quarters, 15,200,000 were exported by the end of July. However,
now that futures markets are in existence in Buenos Ayres and
Rosario the grower is no longer solely dependent on the export
market ; if he thinks well of the future he can hold his actual pro-
duce and hedge, if necessary, with futures," i.e., a growing propor-
tion of the Argentine crop is held over into the next cereal year
(Annual Review of the Grain Trade, 1910, p. 43). Cf. Part I. chap,
iii. § i.
2 The first two columns are from the Corn Trade Year-Book of
140 STUDY OF INDUSTRIAL FLUCTUATION.
The third and fourth columns are calculated on the
illegitimate assumption that the invisible reserves remain
constant ; though even so it is easily seen that the percentage
variations in consumption are perceptibly less than those
in production. There is, however, reason to suppose that
a considerable portion of the year's product is kept in in-
visible reserve, and that these invisible reserves vary in the
same sense as the visible supplies.1 It seems probable that
1904, and the Annual Review of the Grain Trade of 1910 by G. J. S.
Broomhall. The other annual reviews are unfortunately out of
print. But Mr. Broomhall has kindly supplied me with the missing
figures.
1 How large these hidden reserves are it seems entirely impos-
sible to estimate. In India there is reason to believe that owing
New Crop.
Carried over
on July i.*
Report on World's
Invisible Reserves.
i893 •
60-0 m. quarters
8' i m. quarters
1894 . 66-2
5-0
—
1895 . 61-2
3-2
—
1896 . 587
—
1897 .
737
2-9
—
1898 .
89-1
2-2
—
1899 -
73'°
8-0
—
1900 .
75-0
6-4
—
1901 .
94'°
3'9
—
1902 .
85-0
6'5
—
1903 .
71-0
5'3
—
1904 .
62-0
4-6
—
1905 .
86-0
—
1906 .
92-0
5-7
Large
1907 .
79-o
6-9
Large
1908 .
83-0
4'2
Moderate to fair
1909 .
92-0
i-9
Very good
1910 .
79.4
4-8
Good
1911 • ! 777
4'8
—
1912 .
86-0
—
1913 •
94 '2
47
Scanty
* For 1893-4 the estimate is that of Mr. F. B. Howard, for the subsequent
years that of the Washington Bureau of Agriculture (Corn Trade Y ear-Book,
1904, p. 152, and Ann. Reps.). Moreover, the fluctuations in these stocks are
greater than those in the visible supplies ; and though on some occasions, such
as the years ending in 1900 and 1902, when the movements in American produc-
tion and in the world production were widely different, they are in a different
sense, they are much more frequently in the same sense.
THE REVIVAL. 141
if we could take into account the whole of these invisible
supplies, the fluctuations in consumption would be very
considerably smaller than the table suggests. An expert
indeed has given me the unofficial estimate that in recent
years a very good world's wheat crop furnishes about
thirteen months' and a very bad crop about eleven months'
consumption. On this assumption, and taking the smallest
and largest crops of the last decade, those of 1907 and 1909,
T2 12
we may estimate the consumption of these years at — and —
respectively of the production, i.e., at 402,387,336 quarters
and 416,498,760 quarters. That is to say, the consumption
of the worst year of the decade was only about 14 m. quarters
or 3-4 per cent, below that of the best year.1
partly to the inadequate facilities for up-country storage the re-
serves in the hands of the small grower are as a rule very small,
since he is unable to hold back his produce from the great merchants,
whose stocks are either held at the ports or shipped westward.
It must be observed, however, that the " visible supplies " com-
prise no estimate of stocks in Indian ports. In general the impres-
sion that one gets from those conversant with the grain trade is
that the hidden supplies are always considerably larger than one
would be inclined to expect. In Russia during the shortage of
1911 each successive rise in price sucked hitherto unsuspected
reserves upon the market. At the time of the Leiter corner in the
spring of 1898, I have been told that farmers in uncouth dress, who
had never previously left the backwoods, crowded from all direc-
tions into Chicago and New York with their little stocks of wheat
to earn the famine prices. Indeed the reserves of American farmers
— the only stocks in growers' hands of which there are any trust-
worthy estimates — are by no means negligible (see table opposite).
1 One curious suggestion of an eminent authority (Corn
Trade Year-Book, 1904, p. 23) seems to require attention at this
point. Though the initial effect of lower prices may be to increase
stocks, their ultimate effect is, it is suggested, to reduce them ;
for the farmer, being intent upon raising just enough money to
discharge his normal fixed expenses, will have now to put con-
siderably more corn upon the market. And conversely with high
prices. If, as is suggested to have been the case in the early QO'S,
the initial lowering is due to large crops, the result will be to dimin-
ish the tendency to equalisation of consumption ; though it must
be observed that in so far as the effect is supposed to be cumula-
tive from year to year, the argument seems to imply the accumulation
of stocks in the hands of somebody. But if, as is hinted of the
early nineteen hundreds, the initial raising of prices is due not to
142 STUDY OF INDUSTRIAL FLUCTUATION.
There can be no doubt that inter-temporal compensation
of this kind tends to diminish the importance of crop-
small crops but to a raising of the demand schedule, in spite of
large crops, the effect would seem to be to increase the tendency
to compensation. The argument, however, assumes (i) a curiously
molluscan policy on the part of farmers, (2) in the first case rather
special conditions of demand on the part of consumers, for neither
of which there appears to me to be much evidence.
Cf. the following diagrams : —
ON iv\
In the event of a fall in price having begun, the argument sug-
gests that DD' is at first inelastic, but ultimately again cuts RH,
the rectangular hyperbola through P, at a point Q, at which point
THE REVIVAL. 143
influences in effecting a general industrial revival. At the
same time it must be remembered that equalisation of
consumption cannot yet be secured without considerable
alterations in the ratio of exchange ; and that except in
so far as the crop is stored by the growers, its arrival (whether
or not it is immediately consumed) leads to an absorption
in exchange of the stocks of other kinds of goods, and to
that concentration of exchangeable resources in the hands
of producers which we have judged to be on the whole
favourable to industrial revival.1
But apart from this inter-temporal compensation it is
urged that there is also in modern times so considerable an
inter-local compensation that the actual volume of pro-
duction of a crop whose sources of supply are widely dis-
tributed over the globe does not vary greatly from one cereal
year to another. This follows partly from the general
proposition that the more numerous the independent sources
of supply of any commodity, the less will be the variability
of that supply, and partly from the deliberate action of
man. For instance, the Argentine wheat-crop, reaped in
December, is sown in July-August, when more or less trust-
worthy estimates of the North American crop for the cereal
year are already to hand ; and the area put under wheat
varies to some extent accordingly.2 And similar variations
sale is restricted ; ON being so large as to necessitate a trenching
upon stocks.
In the event of a rise in price which is due to an elevation of the
demand curve, sale is restricted to ON ; Q being a point of inter-
section of DJY and RH.
It is possible that DD' (Fig. i), DjD/ (Fig. 2), only touches RH
at Q, in which case Q is a point of unstable equilibrium, and the
whole process may begin again.
1 For the sake of completeness, it may be added that the " psy-
chological " influence of good crops in encouraging investment
depends primarily on the volume of production.
2 For instance, the poor U.S.A. crops of 1904 and 1908 were
accompanied by increases in the Argentine acreage under wheat
from 10,550,000 to 12,110,000 and from 14,225,000 to 15,326,000
acres respectively. The remarkable compensatory action of the
Argentine in December, 1907, however, was apparently due mainly
to natural causes, for the acreage only increased from 14,059,000
to 14,225,000 acres (Review of Grain Trade, 1910, p. 42).
I44 STUDY OF INDUSTRIAL FLUCTUATION.
of acreage in other parts of the world tend to render it less
likely that one year of unusual abundance or shortage will
be followed by a second.
There can be no doubt that this compensatory action
has a marked effect in diminishing the fluctuations in the
world's production of wheat, and that with the opening up
of new sources of supply it seems likely to be of increasing
importance ; and I have heard an eminent authority on
the grain trade discourse so eloquently on the fact as to
leave the impression that he regarded it almost as a sufficient
justification for taking a teleological view of the universe.
Yet a reference to the statistics of the world's production
will show that the variations are still sufficiently large,
especially when we consider the fact just noted that the
yields of the several sources are not entirely independent
but are to some extent correlated negatively. It becomes
therefore of interest to inquire whether there is not also
some natural positive correlation calculated to disturb the
compensatory action. There seems to me some reason for
holding that there is..
The original sun-spot theory of W. S. Jevons has indeed,
it is generally admitted, been put out of court by the facts
that the sun-spot period has been discovered to be some-
what longer, and the average length of the so-called " trade-
cycle " has become considerably shorter since he wrote.
Nor do later rehabilitations seem to me quite convincing.
Mr. H. S. Jevons1 indeed asserts that " speaking broadly "
the United States wheat harvests are abundant at times of
sun-spot minimum and unusually deficient at times of
sun-spot maximum ; but it is difficult to reconcile this
statement with the unusually large harvests of 1884, 1891,
1898, 1905, and the unusual deficiencies of 1881, 1888, 1900
and igii.2
One conceivable alternative hypothesis suggests itself.
Mr. Maunder points out that the maximum point is reached
1 The Causes of Unemployment, p. 56.
2 The latest authoritative sun-spot figures, given by Mr. E, W,
Maunder in Scientia, Jan., 1913, are worth reproduction here.
THE REVIVAL.
145
early in the sun-spot cycle, in other words that the upward
limb of the curve is more steeply inclined than the downward.
Is it possible that while the steady decrease in the area of
the spots exercises no considerable influence, their final
almost complete disappearance and again their rapid and
feverish growth produce grave disorders in the sun's con-
stitution which are reflected in violent oscillations in climatic
i.
ii.
iii.
iv.
1867
65
1879 ..
35
1889
53
1901 . .
5
1868
456
1880 . .
440
1890
99
1902 . .
46
1869
907
1881 . .
681
1891
569
1903 . .
340
1870
.. 1,503
1882 ..
1,000
1892
. . 1,214
1904 . .
488
1871
. . 1,091
1883 . .
1,154
1893
. . 1,464
1905 . .
1,191
1872
—
1884 . .
1,079
1894
. . 1,282
1906 . .
778
1873
—
1885 ..
807
1895
•• 974
1907 . .
1,082
1874
604
1886 . .
38i
1896
•• 543
1908 . .
697
1875
248
1887 ..
179
1897
•• 5H
1909 .-
692
1876
126
1888 . .
89
1898
•• 375
1910 . .
264
1877
108
1889 . .
26
1899
in
I9II . .
64
1878
22
1900
75
1879
3
1901
24
1902
16
The figures indicate in millionths of the sun's visible hemisphere
the average surface covered during the year. Till 1871 they are
from the observations of Spoerer and others, since 1874 from those
of Greenwich Observatory. The periods are represented as over-
lapping, since the spots first appear at the poles and fade away from
the equator, the new cycle beginning before the old one is ended.
Even when as in 1870, 1883, 1893 and 1907 on the one hand,
and 1889, 1901 and 1912 on the other, the actual year of maximum
or minimum bears out Mr. Jevons' contention in an apparently
striking fashion, it will be seen that a harvest of the precisely oppo-
site character occurred in the year either immediately or almost
immediately preceding, and sometimes also in the year following.
Indeed the particular form of periodicity detected by Dr. Shaw in
the wheat-yield of the English eastern counties and by Mr. Jevons
in that of the United States, consists in an approximate equality
of the average yields of any two years equally distant from certain
periodically recurring " points of reversal " — the yield of the post-
reversal-point year being sometimes greater and sometimes less
than that of the pre-reversal-point year. It does not, therefore,
seem to indicate such a continuous movement in either direction
as we might expect to be attendant upon such a continuous pheno-
menon as is the enlargement or diminution of the spotted part of
the sun's surface.
L
146 STUDY OF INDUSTRIAL FLUCTUATION.
conditions and consequently in the yields of wheat ? Thus
would be solved in favour of both parties the controversy
between those who, following the Lockyers, believe the
radiant heat of the sun to vary directly, and the investigators
at the Smithsonian Observatory who believe it to vary
inversely with the spotted area.1 I think an examination
of the annexed figures lends some colour, or at any rate some
enchantment, to this view.2
1 H. S. Jevons, op. cit., p. 74.
2 Computed from Corn Trade Year-Book, 1904 (pp. 67, 138, 150),
supplemented by Stat. Abs. U.S.A., 1912.
Solar Half-
Periods.
Character.
Average
Crop.
Average Divergence
in either direction
from Average Crop.
In Bushels.
U.S.A. . .
1866-70
Disturbed
222
32 or 14-4%
1871-77
Quiet
288
29 ,, io-i
1878-83
Disturbed
448
40 „ 8-9
1884-87
Quiet
446
44 M 9-9*
Average
Yield
Average Divergence'
per Acre.
Bushels.
1884-87
Quiet
12-0
0-77 or 6-4%
1888-93
Disturbed
12-5
1-33 ,, 10-6
1894-99
Quiet.
13-4
o-75 „ 5*6
1900-05
Disturbed
13-6
1-05 „ 77
1906-10
Quiet
14-6
0-64 ,, 4-4
Solar Half-
Periods.
Character.
Average
Yield
per Acre.
Average Divergence
from Average Yield.
Bushels.
United
1866-70
Disturbed
27-4
3-66 or 13-3%
Kingdom
1871-77
Quiet
24-9
1-61 „ 6-5
1878-83
Disturbed
24-9
3-27 ., i3'i
1884-87
Quiet
29-9
1-15 „ 3-8
1888-93
Disturbed
28-8
1-91 „ 6-6
1894-99
Quiet
31-2
2-48 ,,79
Eastern
1888-93
Disturbed
29-6
2-8 „ 9-5
Counties
1894-99
Quiet
32-0
2-9 » 9'i
(Dr. Shaw's
figures)
* The steady extension of acreage westward in all probability makes the figure
in this column for each period larger compared with that for the immediately
preceding period than if figures of yield per acre were available.
THE REVIVAL.
147
But perhaps more interest attaches to Mr. H. S. Jevons'
hypothesis of a 3} years' average period, depending on
meteorological conditions as recorded by the readings of
barometric pressure. Any attempt indeed to exhibit the
wheat crop of the whole world as depending in any direct
and simple manner upon some single atmospheric phenome-
non seems to me unlikely to succeed. The requirements
of the crops in the several countries are too complex and
too divergent, and the climatic conditions in the same year
differ too much in different parts of the world to admit of
any such single comprehensive connecting-link. Accordingly
Mr. Jevons' assertion that the curve of the world's wheat
production shows " with admirable clearness " the 3^ years'
period does not seem altogether convincing in view of the
fact that minima occur at intervals of eight years, two
years, five years, three years and four years respectively.
Solar Half-
Periods.
Character.
Average
Yield
per Acre.
Average Divergence
from Average Yield.
Bushels.
France .
1866-69*
Disturbed
15-6
2-15 or 13-8%
1871^77
Quiet
16-4
2-27 , 13-8*
1878-83
Disturbed
16-1
i-53 , 9-5
1884-87
Quiet
17-6
0-25 , 1-4
1888-93
Disturbed
16-5
1-13 , 6-8
1894-99
Quiet
19-8
0-58 , 2-9t
The hypothesis is of course not proved, nor would it in any case
furnish a complete explanation ; but that is more than either we
desire or deserve from any climatic theory of fluctuations. E.g.,
the collapse of 1900, the minor cycle 1902-3, and the initiation of
the 1907 boom all belong to one disturbance-period, the 1907 col-
lapse falling outside it. Similarly the fruits of the wind are reaped
and the seeds of the whirlwind sown in the single disturbance -period
1866-70. But it is suggested that some light is thrown on the long
stretches of comparative industrial quiet occurring in the late-central
70 's, 8o's and go's and followed by storms. Nor would it be right
to conclude without a warning that we have recently entered upon
a disturbance-period.
* No figures for 1870. The yield per acre in 1871 was probably abnormally
reduced by shortage of labour, etc. ; the figures for 1872-7 yield 2*25 : 17*1 or
i2'6 per cent.
t Ignoring the wholly exceptional dearth of 1897.
148 STUDY OF INDUSTRIAL FLUCTUATION.
Nor does it seem at first sight that the attempt to prove
a natural correlation between the crops even of any two
producing countries is likely to be much more fruitful.
The climatic conditions and requirements of the United
Kingdom and France ought to be very similar ; yet an
examination of the curves of the average yield per acre in
the two countries from 1881 to 1903 shows that they do not
move in the same direction more than fifteen times out of
twenty-three. The connection, however, appears to be closer
than that between the United Kingdom and the United
States, for which the curves move together thirteen times,
and than that between France and the United States, for
which they move together ten times. Figures of yield per
acre are not available for other countries, but it may be
noted that in the thirty-one years 1882-1912 the curves of
production of France and Russia moved together seventeen
times, those of France and the United States fourteen times,
those of Russia and the United States only twelve times.
Now these facts are not of themselves sufficient to prove
anything (the movements, for instance, in the same direction
are often very large in one country and almost imperceptible
in the other) and it would serve no purpose to extend the
calculation to the other principal growers. They do, however,
suggest that country A may have a nearer affinity in the
matter both of climatic conditions and of requirements with
country B than with C, so that there is a somewhat greater
probability of the productivity of A and B being similarly
affected in any given year. In particular it seems clear that
though in all countries the crops require a moist seeding and
growing time, and a dry spell for ripening and harvest,
the relative importance of these requirements is not the
same, so that it is possible to say that a year of unusually
heavy rainfall throughout the world will on the whole benefit
some countries, such as Canada, the Argentine, Russia,
Australia and India, and injure others, such as the United
Kingdom, France and the United States.1 Thus the great
1 This hypothesis and classification is suggested by Bruckner's
conclusion, that " The British Isles, France, Germany, Austria,
THE REVIVAL. 149
crop shortage of 1897, for instance, seems to have been
mainly due to the failure of the rain-hating European
countries of the Anglo-French type : while the accompany-
ing cold prevented the northerly crops of Russia from
obtaining the full benefit of the moisture, the rain-loving
countries nearer the Equator — the Argentine, India, and
Australia — did very well. In 1907 and 1911 on the other
hand the shortage was mainly due to the drought-hating
countries, most of which suffered severely ; while the
European countries did fairly well, and in 1907 France beat
all previous and in 1911 England all recent records.
This comparison suggests a conjecture of some interest.
Until recent years the predominant position among the
world's wheat-growers was, with the important exception
of Russia, held by the rain-hating countries of Western
Europe and the United States. Recently, however, the
rain-loving countries have been rapidly growing in relative
importance, and seem likely to continue to do so. Nor are
there wanting indications of apostasy in the United States
themselves. There have been and still are at work in that
country two important trans-cyclical changes in the culti-
vation of wheat. The states and territories whose produc-
tion is increasing most rapidly are many of them, such as
Nebraska, the Dakotas, Colorado, Washington, Idaho,
Montana, Utah, devoted to the growth of spring wheat,
while the cultivation of the older eastern states and of
Scandinavia and the Atlantic coast States of North America have
an ' oceanic ' climate ; Central and Southern Russia, and the
great central area of North America, have a ' continental ' climate.
During cool damp periods the crops of the former countries tend
to be spoiled by too much wet, whilst those of the continental area
are excellent, having just sufficient moisture. On the other hand
during warm dry periods the harvests are on the average more
abundant in ' oceanic ' territory, and deficient through drought
in the hot, dry, continental areas " (Jevons, op. cit., p. 76). On
the connection of the Indian rains with harvests and prices, cf. F. J.
Atkinson, Silver Prices in India, 1897, pp. 95 ff. This suggestion
is borne out by such a detailed life-history of the various crops as is
given in Mr. Broomhall's publications. The following summary,
for instance, of the conditions in certain specimen years seems
legitimate.
150 STUDY OF INDUSTRIAL FLUCTUATION.
western states such as California which grow the wiriter
U.S.A.
Canada.
Russia.
France.
Argentine
Australia.
India.
United
Kingdom.
1901
Warm
Heat and
Dry weather
Winter
Drought
Too dry
Failure
Too cold
winter :
moisture
hinders
first too t and
and hot.
of mon-
and dry
summer
causing
seeding :
wet and 1 heat.
Result
soon
in spring.
dry
some
warm win-
warm,
Harvest
bad.
and
Result
and hot.
rust and
ter bene-
then too
weather
winter
poor.
Result
sprout-
ficial, sum-
cold.
good,
rains.
very
ing.
mer too hot
Great
but Re-
Result
good.
Never
and dry.
heat
sult bad.
poor.
theless
Result poor.
with
Result
storms
very
in sum-
good.
mer.
Result
1902
Some
No
No
Great
Good
Continu-
Good
Too cold
winter
drought.
drought
heat in
rains.
ous
monsoon
and sun-
kill— too
Result
Result very
July:
Result
drought.
and
less in
cold:
very
good.
never-
good.
Result
winter
spring
other-
wise
good.
theless
Result
very
bad.
rains.
Result
and early
summer,
favour-
good.
good.
but good
able.
forcing
Result
weather
very
later. Re-
fair.
sult
ter.
1903
Continu-
Some
Some
Cold and
Frost
Much
Good
Cold and
ous
drought.
drought in
dry : but
does
moisture
monsoon
wet. Re-
heavy
rains.
Result
not so
spring, re-
lieved by
Result
very
little
harm.
(' Minor
com-
and
winter
sult poor.
Result
good.
rans. Heavy
good.
No
plaints
rains.
very
rains in
drought.
of rust').
Result
poor.
July cause
Result
Result
very
anxiety, but
very
very
good.
Result very
good.
good.
good.
1910
Winter
Great
Very wet
Extreme
Severe
Late
Monsoons
Summer
too
severe
heat and
drought
winter,
spring and
wet and
cold.
drought
especi-
sowing
through
abun-
dant.
wet and
cold.
for
in sum-
harvest.
Result
ally in
drought,
Results
Result
winter,
summer
mer.
Result
Quality
poor owing
very
poor.
the
north.
but
heavy
very
good.
disap-
pointing.
too dry
disap-
to early
Esti-
rains
for
pointing.
drought
mates
later.
spring
and harvest
unduly
Result
crop.
rains, but
favour-
very
Result
as to size
able, tf
good.
only
of crop, in
and *
fair.
spite of
Result
gloomy
poor.
estimates,
Result very
good.
Two complicating features must be observed. First the condi-
tions of rainfall are not by any means the same throughout the
world. Thus 1902 was on the whole a wet year, but there was not
too much rain in England or France, and there was continuous
drought in Australia. 1903, again, was on the whole very wet,
but was dry enough to produce good crops in France, while there
was some drought in Canada. 1910 seems to have been dry on
the whole in the western and wet in the eastern hemisphere.
THE REVIVAL. 151
variety is of decreasing relative importance.1 Now it
seems that the fine dry summer weather which brings the
winter crops to maturity may sometimes check the early
growth of the spring-sown crop : this was noticeably the
case in 1910 and 1911. 2 Moreover, these spring-wheat-
growing countries are situated mainly in the great central
area of North America which according to Bruckner has a
" continental climate " and is more liable to damage from
drought than from excessive moisture. America, therefore,
is gradually tending more and more to grow a kind of wheat
that is liable to suffer from drought upon a soil that is liable
to suffer from drought.
In this connection it should be observed that for the
United States at least Mr. H. S. Jevons' hypothesis of a 3^
years' periodicity seems not devoid of foundation. The
conception of a three or four years' period is indeed a dan-
gerous one to play with ; it is apt to mean little more than
that the curve neither presents a regular dog-tooth alterna-
tion from year to year nor moves for many years together
in one direction. Nevertheless it is true that since 1885
wheat crop minima have only once and maxima only twice
been separated by an interval of more than four or less
than three years ; and the fact that Mr. Jevons finds a
correlation coefficient of — 437, with probable error -101,
between the actual movements of his curve of the total
agricultural produce of the U.S.A. and the curve of the
average barometric pressure during April to October at
Cordoba in the Argentine,3 suggests that the close corre-
spondence which he detects between the average periods of
variation of the two curves is not purely fortuitous, but
Secondly, the conditions of rainfall are complicated by condi-
tions of temperature. The dry weather of 1901 was accompanied
in England by cold, and the English crop therefore suffered. Canada
suffers even more from cold than from drought. England abhors
cold more than France.
1 Cf. the figures in Corn Trade Y ear-Book, 1904, pp. 153-4, with
those in Stat. Abs. U.S.A., 1912, pp. 133-5.
8 Cf. Review of Grain Trade, 1910, p. 12, and Economist, 1911, ii,
p. 167. 3 Jevons, op. cit.t p. 63.
152 STUDY OF INDUSTRIAL FLUCTUATION.
that both phenomena are controlled by a real if somewhat
erratic periodic law. This result confirms the ascription
to the United States of a rain-hating character in the past,
for a low pressure inland is connected with an increase in
the sun's heat ; 1 but it must be observed that in 1904-5-6
the curves both of wheat and of total produce moved not
inversely but directly with the Cordoba pressure curve,
and that it is possible that the pressure figures since 1907
would furnish evidence of that change which we have just
seen reason to suspect in the temperament of the American
wheat crop.2
Is it not then possible that the comparative stability of
the world's annual wheat crop which at present excites the
pious wonder of observers is a temporary phase due to the
approximate balancing of two influences, and destined to
vanish as the rain-lovers get a firmer grip upon the rope ?
The future alone can give a convincing answer ; for the
present the general conclusion seems to be that while there
1 Ibid., p. 75.
2 The Cordoba monthly averages do not appear to have been
computed since 1907 ; but at the Meteorological Office in South
Kensington I have obtained those for Santiago in Chili (situated
also on the high ground of South America, though on the other
side of the Andes) up till 1909 ; to judge from the figures from
1894-1906 * they are directly correlated with those for Cordoba.
It will be seen that in 1906-9 they move in the same direction as
the United States wheat crop (falling with it, unlike the Cordoba
figures, in 1907). This seems, so far as it goes, to be important
evidence in favour of the hypothesis of a recent reversal of Mr.
J evens' law.
Average monthly pressure at Santiago in the six months April-
September.
1894 . . 718-28 millimetres 1902 . . ?
1895 . . 718-02 , 1903 . . 718-08 millimetres
1896 . . 717-64
1897 . . 718-04
1898 . . 717-08
1899 . . 716-36
1900 . . 716-46
1904 . . 717-20*
1905 . . 717-16
1906 . . 718-28
1907 . . 717-94
1908 . . 718-30
1901 * . ? 1909 . . 718-56
* Those for 1901-2 are missing. That for 1904 is computed on the assumption
that the average April-September bears the same relation to the average April-
August in 1904 as in 1905 (the figure for September, 1904, being missing) ; it
is therefore liable to error.
THE REVIVAL. 153
has been recently a considerable and hitherto an increasing
tendency to inter-local compensation, it is not clear that
the fluctuations in aggregate annual production are not still
sufficiently large to exercise an appreciable effect on the
course of industry.1
It remains to consider the third form of alleged compensa-
tion— that afforded in the same country in the same year
by diversity of crops. Mr. H. S. Jevons in the work already
referred to has, I think, shown conclusively that in the
United States this inter-specific compensation is far from
complete. If we concentrate upon the three leading crops,
we find that in the twenty-seven years 1885-1911,
Wheat and corn moved together fifteen times, and
attained common maxima three times ;
Wheat and cotton moved together eleven times, and
attained common maxima five times ;
Cotton and corn moved together seventeen times, and
attained common maxima five times.
A comparison of the yields per acre of corn and wheat
for the same years is equally inconclusive ; all that can be
said is that certain years, such as 1889, 1891, 1895, 1898,
1906, seem to have been generally favourable and others,
such as 1890, 1893, 1903, generally unfavourable to cereal
production. It does not seem to me, however, that any more
definite conclusion than this is needed in. order to establish
the probability of the influence of fluctuations in the volume
of crops upon industrial activity.
In India there is considerable diversion of area between
1 Further, it may be noted that if it prevents so large an increase
in total crops that the effort-demand becomes inelastic, the exist-
ence of partial compensation in a year which is good on the whole
may actually aggravate the tendency to industrial fluctuation :
contrast 1882-4, when the very universality of the big crops pre-
vented the outbreak of constructional boom (cf. H. S. Jevons, op.
cit., and Piatt Andrew, op. cit., p. 330). Finally, owing to the differ-
ences in the economic freedom enjoyed by different producers, a
transference of the source of supply does not necessarily leave
industry as a whole any more than any particular industry un-
affected.
154 STUDY OF INDUSTRIAL FLUCTUATION.
one crop and another from year to year, especially between
rice and jute, but the two chief food-grains, rice and wheat,
show a marked tendency to reach their minimum in the
same year, though the maxima often diverge by one year.
It is very rare, however, for a maximum in one crop to be
accompanied by a large diminution in the other.
Again in Germany, the wheat crop reached well-defined
maxima in —
1878, 1882, 1887, 1890, 1892, 1896, 1899, 1902, 1904,
1906, 1908, 1910,
the rye crop in—
1878, 1882, 1887, 1890, 1893, 1896, 1899, 1904, 1909.
The wheat crop shows well-defined minima in —
1879, 1881, 1883, 1889, 1891, 1895, 1897, 1901, 1905,
1907,
the rye crop in —
1880, 1884, 1889, 1891, 1895, 1897, 1901, 1905.
In the Argentine, however, there are fairly clear indica-
tions of a tendency to negative correlation. Mr. Broomhall
thus summarises the results of the (British cereal) years
1887-1903. l
Wheat.
Corn.
Wheat.
Corn.
1887
Success
Failure
1895
Success
Failure
1888
Failure
Success
1896
Failure
Failure
1889
Disappoint-
Success
1897
Failure
Disappoint-
ing
ing
1890
Disappoint-
Failure
1898
Success
Success
ing
1899
Great success
Disappoint-
1891
Disappoint-
Success
ing
ing
1900
Disappoint-
Disappoint-
1892
Success
Failure
ing
ing
1893
Great success
Failure
1901
Disappoint-
Disappoint-
ing
^ ing
1894
Disappoint-
Success
1902
Success
Great success
ing
1 Corn Trade Year-Book, 1904, p. 32.
THE REVIVAL. 155
And from his later figures l we may continue —
Wheat.
Corn.
Wheat.
Corn.
1903
Success
Great success
1909
Failure
Success
1904
Success
Disappoint-
1910
Failure
Failure
ing
1911
Disappoint-
Success
1905
Failure
Great success
ing
1906
Great success
Failure
1912
Success
Success
1907
Great success
Failure
1913
Failure
Success
1908
Failure
Success
If we reckon disappointments as failures, the results
differ in seventeen years out of twenty-seven and correspond
only in ten. This phenomenon I think suggests that corn
being on the whole a heat-loving crop, its conditions of
production are likely to differ more widely from those of
wheat on a rain-loving soil like the Argentine than in an
(on the whole) heat-loving country like the United States.
Even complete inter-specific compensation of volume
would not, however, be sufficient to destroy the theory of
crop influences : for the effect of an increase in the more
valuable and important crop will not be cancelled by an
equal decrease in the less valuable and important.2
1 Review of Grain Trade, 1910, p. 42, and information kindly
communicated.
2 Indeed where, as in the Argentine, the " psychological " im-
portance of crop conditions is great, inter-specific compensation
may actually tend to prolong and intensify an industrial revival.
For while it requires a boom in the more important product to sow
the seeds of confidence, yet when men's minds are already impreg-
nated with optimism, the boom in the less important is sufficient
to outweigh the failure of the more important crop. Thus though
the Argentine wheat crops of December, 1888 and 1889, and again
of December, 1894 and 1895, were bad, the maize crops, which had
been poor in March, 1888, and March, 1893 and 1894, were excel-
lent in March, 1889 and 1890, and in March, 1894 and 1895, and
on each occasion helped to swell the industrial boom. An addi-
tional explanation is thus furnished of the lag of iron and steel
imports behind wheat exports (cf. p. 87).
156 STUDY OF INDUSTRIAL FLUCTUATION.
§ 5. THE INCREASED ATTRACTIVENESS OF
INVESTMENT.
We have thus, by means of a tedious but unavoidable
digression, established the theoretical probability of the
importance of agricultural influences in causing a general
industrial revival. In view of the detailed discussion in
Part I. chaps, v.-vii. nothing more need be said as to the
channels by which they actually operate, and we pass on
to the third possible cause of a general increase in prosperity
—a rise in the satisfaction-bearing power of the goods which
are the subject of mutual exchange. There seems indeed at
first sight no reason to suppose that the estimate which
people form of the satisfaction afforded them by the con-
sumption of goods should be thus subject to variation : but
we have to explain the undoubted fact that at times of
business revival a number of goods which had previously
been withheld in store are drawn into the circle of exchange.
Now an alternative explanation of this fact at once occurs.
While a general rise in the exchange values of all consum-
able goods in terms of each other is clearly impossible, it is
perfectly possible that each group of producers or owners
should expect a rise in the value of its own products, and
consequently be willing to withdraw them from store.
Moreover, the existence of a monetary economy affords a
mechanism by which such an expectation may be raised
simultaneously in many trades. This question, however,
may conveniently be reserved for discussion, along with other
aspects of a monetary economy, in a separate chapter : at
present we are concerned to note that with regard to certain
products which enter into exchange, the suggestion of a
general rise in their satisfaction-bearing power is by no
means without foundation.
For while the marginal utility to them of consumptive
goods is a thing about which most people are capable of
forming fairly accurate and stable judgments, their estimate
of the marginal utility of construction goods is by no means
so likely to be constant. For this estimate depends on the
THE REVIVAL. 157
expected future marginal productivity of those goods ; and
since both this productivity itself is liable to variation, and
also any forecast of it is at best a matter of guess-work, there
is clearly room for considerable variation in the estimates
formed of the marginal utility of construction goods. It is
the author's conviction that it is these variations which
furnish the key to the most important aspects of modern
industrial fluctuations. The most characteristic feature of
a modern industrial boom is the utilisation of an abnor-
mally large proportion both of the past accumulations and
of the current production of consumable goods to elicit the
production not of other consumable goods but of construc-
tion goods.
Among the most important causes of such a revision of
the marginal utility of construction goods may be mentioned
first the confidence inspired by exceptionally good crops in
the capabilities of a given country ; secondly, the wearing
out of an unusually large number of the instruments of pro-
duction in some important trade or group of trades : and
thirdly, the occurrence of an invention i in some important
trade or group of trades. The two first have already been
sufficiently discussed : 2 about the third more remains to be
said.
The magnitude of the effect of an invention upon the
volume of general industrial activity depends upon the
character of the invention and the circumstances in which
it is introduced. It is likely in the first place to be greater
the more prone the community in question to give heed
to the solicitations of permanent investment. We have
already had occasion to refer the very different course of
the German, English, and American electrical booms of
the later go's to the more forward policy of the German
banks : it only remains to add that the same factor helps
to explain how in Germany the investment bacillus spread
1 I use the word to include "legal" as well as physical inven-
tions (cf. Part I. chap. i. § 2), but the discussion will deal mainly
with the latter.
2 Part I. chap. ii. § 2 and chap. v. § 2.
158 STUDY OF INDUSTRIAL FLUCTUATION.
not only to the copper and electrical but to the constructional
industries in general, while in America the latter continued
to wait for the blessing of Nature, and in England to cry
out for the caresses of America and of the sea.
Secondly, the effect is likely to be greater the more
numerous and important the constructional industries
physically involved. Apart from a lowering of their own
cost of production, already discussed,1 the great central
constructional industries of iron and steel may, it would
seem, be involved in two ways. First, the invention may
be concerned with a product which requires carriage over
long distances, and so stimulate the industries making the
instruments of transport. Thus the copper boom in America
in 1906-7 led to a shortage of cars which increased the
demands upon the steel industry.2 Thus also the demand
for the over-sea transport of oil in 1911-2 was one of the
principal causes of the recent boom in British shipbuilding.3
Secondly, the invention may require the direct co-operation
of the iron and steel trades for its application — indeed it is
pretty certain that it will, but it may do so in very different
degrees. In this connection the contrast between the
German iron and steel booms of 1895-1900 and 1904-7 is
instructive. The former was much more intense, and the
crisis correspondingly severe. On both occasions the initial
impulse was given by electrical invention ; but in the former
1 Part I. chap. iii. § 3.
2 In Germany in the same year the demand for steel led to a
shortage of cars, and so to a demand for more steel ; just as in
1880 and again in 1906 the English shipping and shipbuilding boom
was prolonged by the very imports of iron-ore that were destined
to destroy it (cf. EC., 1906, p. 43).
3 Though the world's production of oil increased largely in 1912,
it was only in California that any large accumulation could at first
be built up (Daily Chronicle, June 21, 1913). By August, 1912,
petroleum freights from New York to Europe had reached 705.
a ton, as compared with IDS. or 125. a year earlier (Statist, Aug. 10,
Of 1 08 vessels under construction at the end of December, 1912
forty, with an aggregate tonnage of 231,000 tons, were tankers,
(Lloyd's Register Annual Summary for 1912, p. 4. Cf. Moss Cir-
cular, Jan. i, 1913).
THE REVIVAL. 159
case the leading feature was the application of electricity to
local transport, which required the physical conjunction or
juxtaposition on a large scale of iron or steel and copper in
electrical permanent ways, vehicles and power-generating
stations : while on the latter occasion it was the application
of electricity to processes of manufacture, which required *
the co-operation of a much smaller amount of new iron and
steel construction. Similarly the inverse movement of
pig-iron and copper consumption in Germany in 1910 and
again in America in 1912 suggests that when, as in these
years, the most important application of electricity is in
the electrification of existing long-distance railroads, the
effect upon the iron and steel trades is naturally likely to
be less than when (as in Germany in 1895-1900) it is in the
construction of new local tracks.
In this connection it is naturally of interest to inquire
how far the other great power-invention of recent years —
the discovery of the potentialities of oil — is to be held re-
sponsible for the most recent industrial boom.2 By 1909
the idea of oil-fuel had begun to take hold of the public
mind, and it only needed a contract of the Admiralty with
the Scotch companies in February, 1910, to set the train
alight. The immense extension of supply and the struggle
1 Though, indeed, it seems to have received considerably more
than it required (see below, chap. ii. p. 185).
2 With regard to the boom of 1906-7 it seems pretty clear that
the rise of oil prices in 1906-7 was in part due to phenomena of
supply, and in part a mere reflex of the constructional boom, oper-
ating chiefly through an increased demand for pleasure-cars. The
London Omnibus companies, indeed, with a far-sighted folly recalling
that of the unhappy Lord Revelstoke, adopted motor-power in
1905 ; but it was the demand for pleasure and not for commercial
purposes that first raised the price of motor spirit. In 1906 " the
energies of our manufacturers were absorbed in the demand for
the higher priced cars," and of the 71,000 vehicles estimated to be
in existence at the end of that year, only one-seventh were said to
be for commercial uses (Ec., 1906, ii. 1,864). The commercial
demand, however, continued to expand when the pleasure demand
contracted after the troubles of October, 1907 — in 1908, for in-
stance, there were 200 more motor-buses in commission in this
country than in 1907 (Ec. H. of 1908) — and helped to maintain
oil prices in 1907-8 and to check their fall in 1908-9.
160 STUDY OF INDUSTRIAL FLUCTUATION.
for the markets of the world between the Standard and the
Shell-Royal-Dutch led indeed to a prolonged slump in oil
prices in 1910-1 ; but the subsequent sharp rise l was no
doubt attributed by the public mind to a large increase 2 in
the use of heavy oil for transport-fuel (both in internal
combustion engines and for steam raising), such as
might be expected to have an important effect on the con-
structional industries. Nevertheless the use of the internal
combustion engine does not seem, in fact, to have made great
progress at this time.3 Nor did consumption for steam-
raising make very large strides : the majority of tramp-
owners, from whom most was expected, do not seem to have
been in a position to make the considerable alterations in
boiler-construction, etc., which are required.4 Indeed there
1 The price of petroleum in England rose in 1910 from $%d. a
gallon to 6d. on May 14, whence it fell steadily to 5^. in July, 1911,
rising to $%d. in December, 6fd. in January, and g^d. in March,
1912. One authority indeed (Statist, Oct. 21, 1911) attributes the
rise entirely to cost, ' ' all the conditions [of demand] being in favour
of lower prices," and the conclusion of the oil-war, the crippling
of the Standard by the adverse judgment of the American courts,
the increased cost of handling at the ports, the flooding of the Gali-
cian mines, and (most important of all) the difficulty of transport
were all invoked. But there is no reason to suspect any actual
diminution of oil-carrying tonnage, and Mr. von Ofenheim seems
justified in pointing out that since " crude oil has risen 50 and 100
per cent, on the fields " (Statist, July 20, 1913) the whole respon-
sibility cannot be laid upon freights.
2 Fostered both by the anticipation and the event of the great
coal strike.
3 In 1910, according to the Engineer, the advance of the internal
combustion engine had not been very rapid, and though in 1911
some important experiments were made, and the first Transatlantic
voyage was performed by a Diesel-engined ship, the Toiler, the
aggregate demand for internal combustion purposes cannot have
greatly increased. Of the 655 new vessels classed by Lloyd's
Register in 1912, only six were fitted with oil engines ; and of the
vessels of over 3,000 tons under construction at the end of the
year, only two in England and ten abroad were similarly equipped
(Lloyd's Register Statistical Tables and Annual Return of 1912).
4 By August, 1912, there had been " no increase to any very
material extent in the consumption of oil as fuel for steam-raising
purposes " (Statist, Aug. 31, 1912), and of the steamers built and
under construction in 1913 which were fitted for the consumption
THE REVIVAL. 161
is some reason for thinking that it was not the heavier but
the lighter oils that dictated the rise,1 and for holding with
Mr. Benjamin Taylor 2 that the vast extension of the produc-
tion of heavy oils was unwarranted, and bound to lead to
disastrous reaction. On the other hand there is now abun-
dant direct evidence of the increasing consumption of oil
fuel, and that not only for naval purposes : 3 and a strong
presumption that as a steady and abundant supply is
assured,4 its use will enormously expand. The important
of oil, the greater part were themselves designed for the carriage of
oil for the hypothetical consumption of other vessels.
1 It was the prices of burning oil and motor spirit — American
petrol and Scottish naphtha — which showed the first symptoms of
recovery (Economist, 1912, i. p. 66). The motor industry which,
with the steady growth of commercial uses, had never suffered
severely in the post-i9O7 relapse, took a new lease of life in 1910
with the development of the taxi-cab. On small pleasure and
fishing boats the use of auxiliary motors consuming light spirit
made steady progress (The Engineer, apud Economist history of
1910. Cf. also Economist, 1912, i. p. 903. Cf. also the influence
of home agriculture, discussed on p. 106).
Nevertheless it is perhaps significant that it was the Standard,
which concentrated its attack upon kerosene, and was unable to
compete in the heavy oils, that had to make peace in 1911. More-
over, an increased demand for petrol would naturally, and at first
in fact did (Statist, Oct. 21, 1911), lead to an actual reduction in
the price of fuel oil ; for though a certain amount of crude petro-
leum is used for fuel purposes, its use is restricted by the flash-point
regulations in force in various countries, and the majority of the
oil used as fuel consists of the residual substance left after the dis-
tillation of the lighter products (cf. Encyc. Brit., article Fuel). The
production of motor spirit and of fuel oil are therefore joint rather
than rival operations ; hence the rise in heavy oils seems to need
explanation from the side of demand. (For the prosperity of fish-
ing and agriculture in these years, cf. p. 203, n. i).
2 " The Fallacy of the Oil Boom " (Financial Review of Reviews,
Sept 1912. PP- 35 SO-
The experiences of the Mexican railways, which saved $250,000
in the year 1911-2, almost entirely by the substitution of oil for
coal power, and by the subsidiary economies in handling and car-
riage of fuel and in cleaning, are significant and are not likely to
remain unimitated (Annual report, published in Statist, April 27,
1912).
4 Mr. Taylor (op. cit., p. 45) overlooks this important considera-
tion. " There is probably more petroleum oil in the bowels of
the earth," he states, " than there is coal in its veins ; but while
we have the coal at command we have not the oil. And the quan-
M
162 STUDY OF INDUSTRIAL FLUCTUATION.
conclusion, therefore, emerges that the influence of the oil
invention upon the great central constructional industries
has hitherto been exercised mainly through the subsidiary
channel of demand for the instruments of oil transport, and
that its full and direct effects remain yet to be experienced
in the immediate future.1
In conclusion, it is necessary to issue a word of caution
as to the true nature of a " general " industrial revival.
Neither a revival of the second 2 nor of the third type dis-
cussed in this chapter involves any increase in the consump-
tity of oil now being produced and feverishly prepared is far more
than the industrial and marine world can possibly consume." In
his anatomical enthusiasm he fails to see that the implications of
these statements as they stand are mutually destructive. It is
possible that the consuming power of the world is small simply
because the oil is not yet completely at command ; and that the
" feverish preparation " by increasing our " command " is increas-
ing also the world's consuming power. Thus the prospectus of
the Oilfields Finance Corporation (quoted in Statist, Feb. 10, 1912)
plausibly urges that " the more general adoption of oil fuel has in
the past been retarded by the uncertainty of a regular supply,
which recent developments and the expansion of production neces-
sarily arising from the further large investments of capital in petro-
leum will do much to obviate." In other words, in the case of
products which are of first importance not only in the constructive
but in the operative stage of industry, increased production may
have the apparently paradoxical effect of raising the price. The
apparent Vvrepov irpbrepov of the speculative investors in the low-
priced year 1910 may have been the only far-sighted policy.
1 Cf. EC., 1912, i. p. 66, on the eccentricity of the Scottish oil
trade : "it will probably not be so in the future, for the whole trade
is altering in its relation to the industrial situation."
At first indeed there will be more demand for structural altera-
tion than for new construction ; and the steel trade may even at
first surfer on the balance by a reduced demand for boilers, etc.,
consequent on the adoption of the internal combustion engine.
2 This generalisation does not, however, apply to a single country.
If, for instance, the country is one whose demand for corn is in-
elastic, while its products are in great demand by corn producers,
both those engaged therein is constructional and in consumptive
industry may well be enabled to enlarge their consumption of con-
sumable goods ; cf. England in 1906 and 1912-3. This result
will be modified to the extent that the country's non-corn producers
are dependent on foreign customers whose demand for corn is not
inelastic : contrast the United States, where the agricultural exports
form a large proportion of the whole.
THE REVIVAL. 163
tion of consumable goods by the producers of consumable
goods taken as a whole.1 A revival of the first type does
at first involve such an increase, but so soon as it leads to
a general extension of investment, that increase, owing to
the miscalculation which accompanies the process of in-
vestment,2 is liable to be succeeded by a diminution. Fur-
thermore neither (at this second stage) the first type nor
the second type necessarily involves an increased aggregate
consumption of consumable goods.3 These considera-
tions both suggest some disquieting reflections as to the
true nature of an industrial boom, and explain why (so far
as our evidence may be trusted) a year of constructional
revival is frequently marked by a diminution in the con-
1 Of course certain groups of consumptive producers will benefit ;
N.B., especially the dependence of the Lancashire cotton spinning
trade on continental constructional industry, evidenced especially
by the high level of profits in the early 80 's and by the survival
of spinning over weaving in the summer of 1907.
2 Cf. Part I. chap. i.
3 The third type does involve such an increase, but the increase
may be obscured and delayed. For while we still postpone the
essential modifications introduced by the existence of a money
economy, attention must be called to the way in which it both
interprets and obscures the fundamental process now under dis-
cussion. In one respect it renders clearer the real nature of the
change which is taking place. In a non-monetary economy the
deliberate collection of large stores of consumable goods would be
a necessary preliminary to engaging in constructional enterprise,
and the increased advantage of the possession of consumable goods
would be the first fact to attract public attention. But in a mone-
tary economy, as Mr. Hull rather irritably complains, a great cor-
poration does not prepare to build a railroad " by collecting a huge
quantity of flour, meat, molasses, etc., with which to pay its workers " ;
the provision of increased supplies of consumable goods is post-
poned until the producers of constructional goods are actually
spending their increased money incomes, and it is more plain than
it otherwise would have been that the increased advantage of the
possession of constructional goods is the fact of primary importance.
On the other hand, the same interposition of a money mechanism
obscures and delays that increased aggregate consumption of con-
sumable goods which we have seen to be an essential feature of
the situation ; for the relapse of the money-demand of those who
are turning their attention from consumption to investment is
not immediately compensated by the increased money-demand
of those engaged in constructional industry.
164 STUDY OF INDUSTRIAL FLUCTUATION.
sumption of consumable goods. Thus the curve of English
food-consumption (Chart X.) fell in the years of construc-
tional outburst, 1880, 1888, 1905 and 1910 ; similarly the
wool trade was in difficulties in 1872, 1880, 1905 and 1911.
Thus also the American coffee-consumption curve (Chart
XII.) relapsed in the (fiscal) years of constructional revival,
1887, 1890, 1896, 1906 and 1910. Once the shibboleth of
" repercussion " is discarded, a wide field for scepticism
as to the identity even of " general revival " with universal
prosperity bursts upon the view.
CHAPTER II.
CRISIS AND DEPRESSION.
§ i. AGRICULTURAL SHORTAGE.
WE pass on to an investigation of the crisis or turning
point of the industrial cycle, that is of the forces which put
an end to the progress of general revival. Following the
classification adopted in the last chapter, we may consider
first the steady growth of real costs. Elaborate discussion
is hardly necessary, but special attention may be drawn
to the prevalence, in times of activity, of wasteful methods
of production and organisation, to the use of incompetent
and over-tired labour of all classes, and to the necessity of
resource to less accessible sources of mineral supply. There
seems no reason to doubt that in such ways each period
of expansion carries in itself the seeds of its own dissolution.1
More obvious and catastrophic is the effect of agricultural
shortage : for arguments of the same character as those
which confirmed the primd facie inference of the beneficial
effect of large crops confirm also the primd facie inference
of the harmful effect of small crops upon the volume of
general industrial activity. Returning to the analysis of
chap. i. § 3, we see that the effect of crop-shortage in time
of prosperity depends (i) upon the elasticity of effort-
demand for corn, (2) upon the other factors enumerated on
p. 132. (i) There seems indeed to be a greater probability
than in the case of crop-abundance that the demand will be
inelastic : for though (i.) since at such a time everybody is
likely to be working at full stretch, the rate of increase of
1 For a good discussion, see Mitchell, Business Cycles, pp. 475-
83 ; cf. also Aftalion, Vol. I. Book III. chaps, ii.-vii.
165
166 STUDY OF INDUSTRIAL FLUCTUATION.
the marginal disutility of effort will be fast — a factor mak-
ing for a small expenditure of effort on corn,1 yet (ii.) as with
commodities in general so with corn the rate of change of
marginal utility is likely to be more rapid for a fall than for
a rise in consumption. (2) Though an increase in aggre-
gate activity is thus indicated (cf. also 2, i. in chap. i. § 3) it
must be borne in mind (2, ii.) that since everybody is e%
hypothesi fairly well off, those whose demand for corn is
inelastic and who are therefore obliged to reduce their
expenditure on other things, are likely (since they will not
be obliged to retrench on necessaries) to reduce that ex-
penditure by a comparatively large amount : while those
whose demand for corn is elastic, and who are therefore en-
larging their expenditure on other things are likely (since
they are already living in comparative comfort) to enlarge
that expenditure by only a comparatively small amount.2
Further (2, iii.), since in times of great activity everybody
is likely to be working at full pressure and near the limit
of the possible expenditure of energy, the rate of change of
the marginal utility of effort is likely to be greater to those
who are increasing than to those who are diminishing their
aggregate expenditure. 3 For both these reasons the increase
of aggregate expenditure by the former is likely to be small
compared to the decrease of aggregate expenditure by the
latter. Supposing, for instance, that the commodity-receipts
of corn-producers are increased as much by Hemichoros
a as they are diminished by Hemichoros /3, there is reason
to suspect a net decrease both of general production and also
of the mutual consumption by the industrial population of
each other's products.
Finally, it should be noted that the harmful " psycho-
1 On the other hand, the rate of increase of marginal utility of
stocks is likely to be slow, that utility, at a time of depletion of
stocks, being already almost infinitely high.
2 In terms of the figure (p. 132), the general argument that UUj
is steeper to the left than to the right of the existing equilibrium
point is liable to be modified.
3 EEj is likely to be steeper to the right than to the left of the
existing equilibrium point.
CRISIS AND DEPRESSION. 167
logical " effects of a crop shortage upon constructional
industry discussed in Book I. ch. v. § 2, to which reference
should here be made, are operative (in the absence of any
counterbalancing influence) upon the volume of industrial
activity taken as a whole.
These conclusions give rise to an attractive suggestion.
Is it not possible that the whole cause and meaning of in-
dustrial crisis and depression is a dislocation of the ratio of
exchange between agricultural and industrial products
leading to an increased consumption by agriculturalists and
a diminished consumption by industrialists ? l In particular
this explanation seems to fit well the circumstances of the
collapse and depression of 1907-8. 2 In 1907 in America
the prosperity of the agricultural community was the chief
redeeming feature of the situation ; "if the Wall Street
speculator has suffered cruelly, the farmer of the West, the
Centre and the South is in a better position than ever."3
While the demand for various manufactured articles such
as hardware and shoes,4 as well as constructional goods,
had broken by July, 1907, there was an increasing activity
in cereals, meat, and many other varieties of food products,5
and " a steady request for wool and sugar." 6 Similarly in
England in 1908 " the farmer has hardly felt at all the
effects of the prevailing depression " : 7 and in that year
the great provision companies were among the few prom-
1 Such a dislocation might be due not only to a sudden shortage,
but to a prolonged over-investment in agricultural as compared
with manufacturing enterprise. Cf. a correspondent of the Econo-
mist (1909, p. 1337) '. "The modern capitalist is a great deal more
interested in producing millions of tons of steel per annum or mil-
lions of yards of cotton cloth than in raising food . . . for the
people. . . . Tens of millions of British capital which our soil
is urgently calling for are being blindly frittered away on things of
purely conventional value. . . . They not only waste so much
capital and labour, but direct them from their best use, the produc-
tion of an adequate supply of cheap food."
2 Cf. Lab. Gaz., July, 1907, pp. 204, 236.
3 Lexis, " La Crise Economique de 1907," Revue des Deux Mondes,
December, 1907, p. 821.
4 Economist, 1907, p. 1223. 3 Ibid., loc. cit.
6 Ibid., p. 1360. 7 Economist History of 1908, p. 5.
168 STUDY OF INDUSTRIAL FLUCTUATION.
inent industrial undertakings to show an actual rise in profits
and prosperity.1
Indeed some confirmation of this point of view is afforded
by a study of the whole course in recent years of the com-
parative movements of the prices of our imports (mainly
food and raw materials) and exports (mainly mineral and
manufactured goods). Mr. Bowley's table of the quantity
of exports given in each year in exchange for a constant
quantity of imports shows a maximum in the depression
year 1877 and minima in the boom years 1890 and 1900.
Similarly a computation from the Board of Trade's report
on Exports and Imports at the prices of igoo2 shows a
1 The following figures of net profit are significant : —
1907. 1908. 1909.
Harrod's (year ending Jan. 31) . £162517 £172,837 £198,322
Lyons' (year ending Mar. 31) . £231,849 £248,825 £287,852
Total of shopkeeping companies in Econo- 1907. 1908.
mist returns ..... £1,019,000 £1,048,000
Thus, while most industrial shares fell between September, 1907,
and September, 1908, Lipton's rose from i& to i^-, Lyons' from
6 to 6-jk; Harrod's from 4^- to 4^; and those of the River Plate
Fresh Meat Co. from % to i&.
2 Cd. 6314, 1912 — •
iii.
i. ii. Quantity of Exports
Average value Of Exports. given for a given
net Imports. quantity of Imports.
1900 . 100 -o 100-0 From 100-0
1901 . 96-7 95-2 which 101-6
1902 . 95 '6 90-8 we get 105-3
1903 . 96-9 90-7 106-8
1904 . 97-3 91-8 106-0
1905 . 97-7 91 6 106-7
1906 . 101-8 97-1 104-8
1907 . 106-5 101-9 i°4'5
1908 . 102-5 98-0 104-6
1909 . 103-6 94-3 iio-o
1910 . 109-5 98-4 iii-i
1911 . 106-8 99-9 uo-o
It should be noted that an unduly unfavourable appearance is
given to the whole series by the selection as base of 1900, when the
price of our coal exports was abnormal.
It does not of course follow from these figures of relative values
that the manufacturing demand is inelastic : but this conclusion
is borne out by the Board of Trade figures of volumes of exports
and imports given in the same publication.
CRISIS AND DEPRESSION. 169
maximum in the depression year 1903, and a minimum
in the year of consumptive recovery, 1904. 1 Nevertheless
no light is thrown upon the existence of depression in 1886
and 1892-3, nor (most significant of all) upon the undoubted
relapse in the second half of I9I32 : nor is much support
Volume of net Imports.
Of Exports.
1900
100 -0
100-0
1901
I02-I
IOI-I
1902
105-3
107-2
1903
106-2
IIO-I
1904
107-4
112-5
1905
108-4
123-6
1906
116-6
132-9
1907
113-1
143-6
1908
108-9
132-2
1909
111-9
137-8
1910
114-1
I50-3
1911
117-6
156-2
The volume of imports shows a fairly steady rise, owing to changes
in the demand schedule occasioned by the growth of population
and of the advantages to be gained from international trade : but
the much greater growth in the volume of exports can hardly be
ascribed entirely to the growth in the import of securities or other
" invisible " imports. Nevertheless, in view of our large capital
exports in 1905-7, and again during the rubber and oil booms, we
must be on our guard against over-estimating the inelasticity of
demand.
1 The general conclusion to which these figures, taken as a whole,
lead is that the normal tendency for the ratio of exchange to alter
against the manufacturing and in favour of the agricultural com-
munities was in force in the seventies,was suspended in the eighties and
nineties, and is now once more on the whole triumphing. This is
perhaps the most significant economic fact in the world to-day ;
and, as has been pointed out to me, the figures tend to under-estimate
its gravity, since the rise in the price of exports is in part due to
an antecedent rise in the price of raw materials, so that the altera-
tion of the ratio of exchange to the disadvantage of manufacturing
margins is even more pronounced than the figures indicate. But
the fact is clearly of secular rather than cyclical importance.
z While the prices of the textile materials — wool, jute, and cot-
ton— rose somewhat, the prices of almost all food products — with
the important exception of meat — have been falling steadily. The
alarming rise in meat prices is due chiefly to the increased consum-
ing power of the world in general and the United States (which
has become an importer instead of an exporter) in particular :
but also to the aftermath of the 1911 drought and to the increased
attractiveness of cereals to the Argentine farmer. Increased sup-
170 STUDY OF INDUSTRIAL FLUCTUATION.
afforded to our previous conclusion about 1908. l Both the
evidence therefore and the general probabilities of the case
force us, while emphasising the supplementary importance
of agricultural shortage, to refuse to assign it as the sole
cause of industrial collapse.
§ 2 CONSTRUCTIONAL RELAPSE— DEPLETION
OF STOCKS.
We are led therefore to investigate the possibility of re-
action from the third type of industrial revival discussed
in the last chapter, exhibiting itself in a fall of the ratio of
exchange of constructional as against consumable goods,
and a diminished prosperity of the constructional trades.
One possible cause of such reaction at once presents itself.
It has been suggested that one result of the increased attrac-
tiveness of investment is a considerable absorption into the
vortex of exchange of accumulated consumable stocks.
After a time therefore, unless and until the consumable
goods created by the new instruments appear in sufficient
quantities to compensate for this absorption, it will be phy-
sically impossible for the investment in construction goods
to be maintained on the scale on which it has been begun.
The fundamental cause in such circumstances of the collapse
of constructional enterprise, is thus seen to be not the high
plies are hoped for in time from Mexico and Siberia, but the period
of gestation is some five years (Times, June, 1914) : and if (as I
have been told by a prominent corn-merchant) there are indications
that the Argentine farmer is not likely to put under wheat much
more pasture land, from which he can raise three crops of fodder
per annum, an increased supply from that source will have its
drawbacks for food consumers. On the rise in bacon the follow-
ing comment of a Burnley grocer is suggestive (Daily Mail, Dec.
29, 1912). " Years ago I could not sell bacon costing more than
j\d. a Ib. To-day this quality costs 3d. or gd., but they won't
have that now ; they want the best, and that costs is./' — a criti-
cism which I heard echoed not long since by an observant house-
wife. But there is also no doubt a genuine shortage.
1 The dependence of certain branches of English industry on
certain groups of agricultural producers might theoretically account
for a divergence between English industry as a whole and world
industry as a whole : but a cursory acquaintance with the facts
shows that this explanation will not help us.
CRISIS AND DEPRESSION.
cost of constructional materials,1 but the scarcity of real
capital available for investment.2
Considerable evidence can be adduced in support of this
proposition, and that not only with regard to the actual
months of crisis or turning-point, but also with regard to
certain periods which are commonly regarded as times of
downright " depression."
(i) The first witness, the condition of stocks, is indeed
an unsatisfactory one. Such figures as are available are
mainly either of constructional goods or of agricultural
products whose supply depends largely upon variations in
annual harvests. Nevertheless the information available
suggests, for what it is worth, that the boom period is
sometimes (notably in 1904-6) marked by a progressive
depletion of stocks of consumable goods, and further that
1 Cf. Part I. chap. iii. § 2.
* Those authorities therefore who, following M. Tugan Baranow-
sky, detect the cause of crises in a shortage of capital seem to be
fundamentally in the right. The value of Baranowsky's work
is, however, much impaired by his failure to appreciate that the
investible capital of a community consists essentially not in money
deposits but in its stocks of consumable goods, and his consequent
misunderstanding of the real nature of the phenomenon of which
he treats (op. cit., Part II. chap. iii.). Thus also Aftalion, who
in one place (op. cit., Bk. IV. ch. ii. § 3) makes some suggestive
criticism of Baranowsky's theory, treats elsewhere (Bk. XI. ch. iii.
§ 2) of " fonds de subsistance " or stocks of consumable goods as
something quite distinct from " epargnes." For more detailed
criticism, cf. review of these two works by the present writer in
EC. Jour., March, 1914.
Among Baranowsky's followers, Spiethoff alone appears to recog-
nise clearly the identity of shortage of capital and shortage of con-
sumable goods. His theory of the " ill-balanced production of
industrial equipment and consumable goods " (of which a useful
summary is given in Mitchell, op. cit., pp. 10-11) appears to resemble
closely that elaborated in the text.
Mr. J. M. Keynes, in an unpublished paper read recently before
the London Political Economy Club, develops, especially in regard
to its monetary interpretation, the doctrine of the origin of crises in
a too rapid absorption of consumable accumulations. To this
paper I am much indebted, though Mr. Keynes is, I believe, good
enough to acknowledge a reciprocal obligation : and though he is I
think mistaken in conveying the impression that the relapse of
investment is always due to the physical impossibility of maintain-
ing it upon the existing scale (cf . § 3 below) .
172 STUDY OF INDUSTRIAL FLUCTUATION.
this depletion is sometimes (notably in 1901-3) prolonged
into the period of depression.1
(2) More pregnant is the evidence afforded by the move-
ments of the balance of trade of certain countries upon
certain occasions. That the American collapse of 1907
was due not simply to an abuse of the mechanism of credit
but to an exceptionally rapid absorption of the real capital
of the world2 is clearly suggested by the following figures of
the country's foreign trade.
Per cent. Per cent.
Imports.3 increase on Exports.3 increase on
$ m. previous year. $ m. previous year.
1904 . . 991 . . 1466
1905 . . 1117 127 . . 1518 4-0
1906 . . 1226 9-8 . . 1743 15-0
1907 . . 1434 16-9 . . 1880 7-9
The tendency for exports to increase at a faster rate than
imports was interrupted in 1907 by large capital importa-
tions, America having already absorbed her own accumu-
lations.4 Similarly the increase of the unfavourable trade
balance of Canada from 0-6 million dollars in 1902 to 244
million dollars in 1912 5 argues a colossal increase of recent
years in that country's imports of capital.6 While such
1 Moreover, while statistics of monthly and even of annual prb-
duction, are lamentably deficient, the general consensus of opinion,
supported by such figures as exist, is that the output of consumable
goods increases considerably less rapidly than that of constructional
goods, affording prima facie ground for supposing that the stocks
thereof are at least as rapidly depleted.
* " Extravagant consumption had retarded accumulation, while
the demand of industry and governments for capital was unabated "
(Sprague, EC. Jour., 1908, p. 355).
3 The figures are from the Stat. Abs. U.S.A., from which I have
calculated the percentage.
4 M. Leroy-Beaulieu estimates that the capital demanded by
America in this year amounted to 16,000 million francs, while the
annual savings of the civilised world did not at this time exceed
12,000 million (Levy, Revue des Deux Mondes, 1907, p. 812).
5 EC., February 15, 1913.
6 It makes little difference in principle whether the capital im-
ports take the form of consumable goods or of construction goods
produced in the lending countries by trenching upon consumable
goods.
CRISIS AND DEPRESSION. 173
figures do not of themselves of course afford any evidence
that the savings of other countries are unequal to the strain
imposed upon them, they do afford some prima facie ground
for suspecting that any country which is living thus deliber-
ately and widely outside its income may be involving not
only itself but the whole world in a shortage of real capital.1
That the shortage of real capital is not always confined
to the crisis months is suggested by the course of events in
the United States in 1908-10. The renewed attempts at
investment in the latter part of 1908 2 were hampered by the
scarcity of capital, and in May, 1909, the situation became
acute. Its true nature is made more plain than it otherwise
might have been by the financial policy of the American
railroads. In 1907 they had sought to put off the evil day
by the issue of notes or short-term bonds, repayable in two
or three years' time. 3 In 1909 some 400 million dollars of
these notes became due for conversion, and the situation was
met by renewed importations of foreign capital : in June,
July and August the balance of trade became unfavourable
for the first time since 1897. The situation was temporarily
saved,4 but by the spring of 1910 it again became critical.5
Again the cure was the same ; in February, March and
April the balance of trade became unfavourable : " New
York has been exporting bonds and raising money in Europe
on an unprecedented scale/'6 It was only by swallowing
1 Cf. some suggestive remarks by Hyndman (Commercial Crises
of the Nineteenth Century, p. 155) on the temporary nature of an
export boom caused by the contraction by a foreign country of a
large capital debt.
2 The metal trades unemployment figure (New York State) fell
from 32-3 in May to 18-5 in December : the building figure, how-
ever, after an attempted fall, rose 16 per cent, in the period (vide
Table IX.).
3 Cf. Levy, Revue des Deux Mondes, 1907, p. 812. In the first
half of 1907 286 million dollars of these notes were issued at New
York by railway companies, of which the Pennsylvania Railroad
was responsible for 60 million dollars and the New York Central
for 50 million dollars. Cf. Lescure, op. cit., p. 223.
4 Vide constructional unemployment indices in Table IX.
* The Economist's correspondent (1910, i. 947) speaks truly of
" the inflated prices, which are the critical feature of the present
situation," > EC., 1910, i. 824.
174 STUDY OF INDUSTRIAL FLUCTUATION.
copious doses of the physic of foreign consumable goods
in 1909-10 that America was able to re-establish herself by
1911 on to a firm basis from which to take her subsequent
leap to renewed prosperity.1
(3) The evidence afforded by the comparative course of
constructional and non-constructional prices is also on occa-
sions suggestive. The rapidity with which in the later
stages of the boom the latter gain ground upon the former
affords some indication of the extent to which non-con-
structional stocks have been depleted and of the proximity
and severity of the impending crisis.2 The contrast in this
respect between this country and Canada during the last
boom is very instructive : cf. the following indices3 : —
ENGLAND.
CONSTRUCTIONAL NON-CONSTRUCTIONAL.
Coal and Metals. Food, etc. Textiles. Miscellaneous.
1910 . . 76-6 .. 109-2 136-2 104-3 (1900 =
1911 . . 74-7 . . in -6 128-9 I05'5 100)
1912 . . 84-9 . . 119-9 I:[9>6 II0*1
CANADA.
NON-CONSTRUCTIONAL.
CONSTRUCTIONAL. Fuel and Hides, Tallow,
Metals. Lumber. Lighting. Leather, Boots. Food.
1910
97-6 158-5 . . 103-0
I35-4
136-2 (average
I9II
108-3 165-4 • • I00'5
139-6
137-7 1890-9
1912
117-4 166-5 • • n-3'3
152-4
152-7 =100)
1 Cf.
the following figures : —
Per cent, increase
Per cent, increase
Imports. on previous year.
Exports.
on previous year.
$ m.
$ m.
1909 .
. 1311 + 9'8
1663
— II -2
1910 .
1556 4-18-7
1744
+ 4-9
1911 .
. 1527 - 1-9
2049
+ U-I
2 Indeed we reach the somewhat paradoxical conclusion that
while in the earlier stages of a boom that country is making the
biggest fool of itself in which the divergence is greatest between the
rise in constructional and non-constructional prices, in the later
stages, when the realisation of error has begun, that country's
position is not improbably soundest where the divergence is greatest.
3 Report on Cost of Living, Cd. 6955, August, 1913, pp. 308, 386.
CRISIS AND DEPRESSION. 175
The more rapid rise of non-constructional than of con-
structional prices in 1912 suggests that Canada had made
less successful attempts than had England to remedy the
relative scarcity of consumable goods. Her enormous
capital importations seem to have been used to increase the
supply and to keep within bounds the price of constructional
materials, instead of in attempting to meet the real danger
of the situation. The greater moderation of England in this J
respect, a moderation which seems to some extent to have
deserted her in 1907, and of which she was prevented from
reaping the full benefits in 1901-3, has been without doubt
a most important factor in mitigating the severity of the
recent relapse.
(4) Fourthly, important evidence may be derived from a
study of the divergence on certain occasions between the
fortunes of different groups of consumptive trades. At a
time when the fundamental trouble is a shortage of con-
sumable accumulations, those trades in which the unit of
investment is most imperfectly divisible, — the dysenteric
trades, as we may call them — are pro tanto in a relatively
favourable position. For the capacity to produce largely
from existing plant is now on the whole an enviable one :
moreover, the facilities for investment, which with them is
apt to mean over-investment, are severely curtailed. Ac-
cordingly we find that in England in 1901-3 these trades,
of which wool, boots, ready-made tailoring and hosiery
may be taken as representative, were on the whole, to judge
from the employment figures, the least severely affected.
Thus they showed a marked recovery in the first quarter of
1901, and though from June to September they suffered a
relapse, on the whole they seem to have been less hard hit
in these years than the more dyspeptic trades such as
leather,1 paper, printing and glass. Glass declined unin-
terruptedly, paper from April, leather from June : fur-
nishing became worse than the previous year in June and
printing in September. It is only in the more dysenteric
1 Leather, it is true, improved January- June, 1901, but chiefly
under the stimulus of the war demand.
176 STUDY OF INDUSTRIAL FLUCTUATION.
trades — with the exception of boots — that employment was
fairly well maintained in the second half-year.1
Similarly in America during the capital famine of 1908,
the unemployment index in the clothing trade fell between
May to December from 49-3 to 16-3 or 67 per cent.2 With
the temporary relief of the capital importations, it rose
from n-8 in March to 22-1 in June, from which it only par-
tially recovered in the autumn to relapse again in December
to 20-9. With the New Year, as we have seen, the old
trouble recurred : building was paralysed, metal hesitative,
but the clothing trade began to look up. But with the new
dose of capital imports clothing unemployment rose in May
by all but 100 per cent, on the March figure, while metal
1 The causes of unemployment and " depression " in all trades
in these years are discussed later in § 4 : but it must be premised
here that their existence does not disprove the contention that
shortage is the fundamental evil. A man, to adopt a famous meta-
phor employed by Malthus in another connection, is no less con-
fined within a room because he is not actually knocking his head
against the walls : and the volume of industrial activity physically
possible is none the less undesirably small because the actual volume
is still smaller.
Though the American expedient of large capital importations
was not open to us in 1902—3 the situation was somewhat relieved
by the fact that owing to various circumstances — the great world
demand for coal in 1900, the price policy of the coal cartel in that
and the following year, and the United States iron spurt in 1902 —
the ratio of exchange moved in our favour in 1900, and did not
move as severely against us in the following years as would other-
wise have happened, so that we were to some extent able to obtain
the consumable capital we needed, not like America in 1909-10
by the sale of bonds, but by the sale of instrumental capital for which
we were not yet ready. But the relief was not considerable, and
it was not till 1904 that the slow but steady automatic growth of
productivity (cf. ch. i. § 2) had increased the consumable accumu-
lations sufficiently to render possible the investment of the following
year.
* The recovery of the textile trades in this period is very strik-
ing ; by June, 1908, " owing to the increased demand for textile
goods, there has been a resumption of activity in the New England
textile centres. Manufacturers employing from 35,000 to 40,000
persons increased their hours of labour during the latter part of
May, and many mills which had been on short time for months
started working full time on June j " (L, G., July, 1908, p. 212).
CRISIS AND DEPRESSION. 177
unemployment touched 2-9, the lowest point yet reached in
the period.1
(5) Finally, while adhering to our decision to postpone
the discussion of purely monetary influences, we are justified
in turning for corroboration of our conclusion to the ex-
planation sometimes given in monetary circles of the cause
of crisis. There is some evidence that at such times it is
not only the extent but the nature of the transactions in
which they are involved that causes disquietude to the
powers of the money market. It seems indeed impossible
to determine in what proportions the resources over which
the control is dispensed by the banks are being devoted at
any given time to the furtherance respectively of construc-
tional and consumptive enterprise. The external similarity
between the trade and the finance bill would be an insuper-
able obstacle to the investigator ; nor does the character of
the collateral deposited against a loan necessarily throw
any light upon the use to which that loan is being put. But
the phrase frequently heard at such times in explanation
of the scarcity of " money " for purposes of company flota-
tion, that the " needs of commerce are so great that there
are no further funds available for industry," seems to throw
some light on the true situation. At first sight it would
seem that an abundant supply of genuine trade bills, in
other words a large volume of current consumptive trade,
is precisely the condition most favourable for a recrudescence
of industrial investment and constructional enterpeise.
But beneath and behind its monetary veil the phrase ap-
pears to imply a realisation on the part both of the creators
of currency and of the producers and purveyors of con-
sumptive goods that the supply of consumptive goods is in
arrear of the requirements of those already embarked upon
1 The metal figures 1910-12, however, indicate that mere abun-
dance of capital is not sufficient in America to cause a sustained
recovery in investment, if agricultural influences are absent. The
more " luxurious " consumptive trades, notably printing, reaped
most of the benefit. (There seems, however, to have been a tem-
porary recovery — not entirely seasonal — in building in the summer
of 1910.)
N
178 STUDY OF INDUSTRIAL FLUCTUATION.
constructional enterprise, and that the most profitable use
of increased resources derived from successful trade will be
in eliciting the production not of more construction but of
more consumption goods. This realisation of the true
situation is indeed more or less inextricably mixed in the
minds of bankers, etc., with the conviction that their gold
reserves are small compared with their total liabilities, and
also that the resources already devoted by them to invest-
ment are so far from liquid that the more tightly they hang
on to those gold reserves the better ; and this aspect of the
situation is perhaps the dominant one in their minds. But
there seems justification for suspecting also at least a sub-
conscious recognition that a mere increase in gold reserves
would not permanently solve the essential difficulties of the
position.
It remains to point out certain factors which aggravate
this tendency to a progressive depletion of consumable
stocks. The first is the length of the period of gestation
in the principal trades, with its double consequence of the
absorption of an unwarrantably large proportion of existing
resources in investment, and of delay in the production of
new consumable goods. The matter has been so thoroughly
discussed in the first part that no more need be said of it
here.
The second aggravation arises from the fact that the
invested resources are not as a rule equally distributed
among all industries : in fact, for various reasons elsewhere
discussed, an exceptionally large proportion is absorbed in
investment in the industries of transport, which are not
engaged in the production of " goods," in the ordinary sense,
at all. When a large quantity of assorted goods is invested
in one form or other of railway building, no automatic
provision is thereby made for their replacement, and a
shortage not only acute but prolonged may be expected to
ensue. For instance, there is a good deal of disagreement
between two authorities so much alike in their general con-
clusions as Mr. Hull and M. Lescure as to whether " de-
pression " existed in the United States in 1900-1. The
CRISIS AND DEPRESSION. 179
ambiguity disappears when we remember the abnormally
large part played by investment in the transport industries,
especially in electric traction, in the preceding boom, so
that the ill effects of the over-investment were naturally
enough not felt severely in the form of that " general over-
production " of consumable goods which is M. Lescure's
only criterion of a depression. But that depression, in the
essential sense of a generally curtailed consumption, was
prevalent there seems no reason to doubt.
Thirdly, the pressure on the world's supplies of real
capital may be reinforced by exceptional circumstances.
Thus it seems clear that the crisis of 1900 in England and
the subsequent depression of the " consumptive scarcity "
type was aggravated by the absorption, both by taxation
and by Government borrowing, of large quantities of con-
sumable goods into (industrially) unproductive channels for
purposes of the South African War. Similarly the crash
of 1907 was aggravated by the large toll which had been
taken from the world's real capital in the early stages of
the boom by the Russo-Japanese War and by a series of
disasters of which the most conspicuous was the San Fran-
cisco earthquake and fire. Mars and Vulcan had each
contracted a sterile union with Juno Moneta and the real
" overproduction " of the early twentieth century was an
overproduction of flames and hatred.1 And while those
1 In this connection the figures of capital issues on the London
markets are significant : though, as Mr. Lay ton has pointed out,
these figures must be treated with caution, as they (i) only repre-
sent capital publicly subscribed, (2) include issues floated in London
but ultimately held in part abroad.
British, Foreign
and Colonial Governments. Total.
1898 . £34,236,400 . . £150,173,000
1899* . . . 17,298,600 . . 153,169,000
1900* . . 56,178,200 . . 165,499,000
1901* . . . 91,264,400 . . 159,358,000
1902* . . . 67,143,900 . . 153,812,000
1903 . . 43,960,000 . . 105,463,000
19041 . . . 63,771,200 . . 123,020,000
I9°5t • • • 61,508,000 . . 167,187,000
1906 . . 35,728,100 . . 120,173,780
1907 . . . 36,397,200 . . 123,630,000
* S. African War. t Russo-Japanese War.
i8o STUDY OF INDUSTRIAL FLUCTUATION
writers who, like Mr. F. W. Hirst,1 find in the Balkan War
and the enormous war-loans and armament bills of the
Great Powers the sole and sufficient cause of the ebb of the
recent trade boom, lay themselves fairly open to the charge
of monomania, there need be no question that they have a
considerable measure of right on their side.
§ 3. CONSTRUCTIONAL RELAPSE— ESSENTIAL
CAUSES.
It must not, however, be supposed that a check to the type
of industrial revival depicted in ch. i. § 5 can only be caused
by an actual depletion of the stocks of consumable goods.
In the first place it is possible that in certain periods of
expansion (as perhaps in England in those culminating in
1882 and 1890) the growth of general productivity is so
considerable that no such depletion actually occurs ; sec-
ondly, in normal cases the close of the period of gestation
will bring replenishment. Nevertheless, neither in the first
case is constructional relapse avoided, nor in the second
case is constructional activity immediately restored. The
check to investment arises from the recognition not neces-
sarily that to maintain it upon the same or an increased
scale would be physically impossible, but that it would
involve a sacrifice of present enjoyment disproportionate
to the result. Consumable goods may be abundant, but if
it is known that with the close of the period of gestation
they are about to become far more abundant still, a wise
community will devote them to eliciting the immediate
production of other consumable rather than of construc-
tional goods. The fundamental meaning of over-invest-
ment is failure to attain the ideal distribution of the com-
munity's income of consumable goods through time. Al-
though therefore a country may be actually increasing
both its current consumption and even its accumulation of
consumable goods, it may still be engaged in over-invest-
ment.
1 Cf. Economist, 1912-14, passim.
CRISIS AND DEPRESSION. 181
That the most conspicuous relapses of constructional
industry are not always due to a positive shortage of real
capital is proved by the manner in which the spirit of invest-
ment often flits from trade to trade and country to country
before receiving its final quietus. The apparently most
general collapse is often seen to be not quite general after
all. Thus in the seventies investment, repulsed from con-
struction and transport, continued to flow into the woollen
trade. The productive capacity continued to increase
right up till 1875, which (in view of the shortness of the
period of gestation) indicates a continued investment after
the closing of more obvious fields.1 Similarly in 1883,
1890, and 1900-1 investment seems to have continued on a
considerable scale in the cotton trade2 after it had deserted
constructional industry. The building trade again appears
sometimes to serve as a kind of resting-place for homeless
resources. " Probably owing to the uncertainty of ade-
quate returns in other channels of investment," says the
Leeds Mercury, reviewing the course of 1876, " spare capital
has to a large extent been employed, with assurance of
substantial results, in the building or purchase of property.
On the principle that it is an ill wind which blows nobody
good the building trades have actually reached some advan-
tage from the comparative quietness of other branches of
business. . . . All classes appear ready to invest in house
property/'3
In precisely the same way certain countries, of which
the Argentine Republic is the most conspicuous example,
seem to fulfil the function of a kind of sink for investible
1 The Leeds Mercury estimates that between 1871 and 1875 the
number of spindles had risen in the woollen branch from 2,664,979
to 3,766,703, and in the worsted branch from 1,881,144 to 2,182,792.
2 Cf. p. 21.
3 Similarly a large continuous rise in house-building projects took
place in the first half of 1910 (see table), when the price of pig-iron
was falling : and the building trade has done very well in the first
part of 1914. Indeed, after the main constructional relapse of
1913, there seems to have been a considerable flow of capital, not
only into house building but into general manufacturing enterprise.
Cf . the following figures of London capital applications : —
182 STUDY OF INDUSTRIAL FLUCTUATION.
resources when more familiar channels are already well
filled.1 It will be remembered that the maxima of our
constructional exports to the Argentine occurred in 1884,
1889, 1896, 1901, 1904, and 1909, — years of which all but the
second are commonly looked upon from an English point
of view as years of ebb or at best of hardly perceptible
advance. The history of 1909-11 in particular suggests
that the Argentine is apt to be used as a kind of reservoir of
investment, when other fields have been temporarily
flooded, and to be deserted again on the discovery of a new
toy such as rubber or oil.2
The check to constructional activity, whether catastrophic
or piecemeal, does not, however, necessarily imply that
there has been any miscalculation. If the initial impulse
to investment is either some new invention or a discontinu-
ous leap in general prosperity such as the opening up of a
new market or country, it may in any case, even in the
1912.
1913-
Iron and Steel Companies
£8-8 m.
^6-4 m.
General Manufacturing
Companies .
8-5
10-9
First quarter of
1912.
1913-
1914.
Manufacturing .
. £2-3 m.
^4-9 m.
;£4'2 m.
Electrical
1-6
1-8
4'3
Total
. 48-0
5o-3
97-6
1 Cf. the manner in which, after the relapse of the American
railway-building boom of 1869-71, the German capital by which
it had been largely supported was absorbed in constructional invest-
ment at home.
z It should be observed, however, that while the migration of
investment from trade to trade eases the industrial situation, its
migration from construction and transport in one country to con-
struction and transport in another tends, while affording temporary
relief to the industries in question, to aggravate the essential danger.
The increased opportunities for such migration, to be illustrated
perhaps in the next year or two by the substitution of Rhodesia
and British East Africa for Canada and South Africa as sinks for
British capital, may tend in the future both to prolong the length
of the typical constructional boom and to aggravate the ultimate
reaction. In this as in other cases the action of the law of compen-
satory demand is not, while other disturbing influences remain,
unequivocally beneficial.
CRISIS AND DEPRESSION. 183
absence of all miscalculation, only be of a temporary nature.
Once the new instruments are constructed they will be
capable for a long time of satisfying all demands 1 : the
activity of constructional industry is necessary as a pre-
liminary stage to the prolonged prosperity of industry in
general, but constructional industry itself cannot expect
to share in that prosperity. The fate of Forerunners is
proverbial.
In this connection the differences between the various
inventions which seem to be mainly responsible for the
fluctuations in modern German industry are of considerable
interest. The " invention " which precipitated the great
boom of 1873 was none other than the invention of modern
Germany itself, — a sudden anticipated leap in the whole
conditions of national prosperity and demand.2 Hence
the course of constructional industry was one of dramatic
expansion and collapse. The invention of the late seventies
on the other hand was an improvement of processes within
the iron and steel industries themselves, — equivalent to a
permanent lowering of their costs of production and not to
a temporary increase in the demand for their services : and
the reaction consequently was very slight.3
The invention of electric power did not, like that of
Germany, amount to a general raising of consumers' de-
mand ; but neither was it in the main, like that of basic
1 Compare the manner in which, according to a remark made
in my presence by a manufacturer of agricultural machinery, the
passage of Captain Pretyman's bill enforcing the fencing of the
dangerous parts of such machinery led to a boom in the trade for
some months, followed by a period of slackness. So also with the
more durable kinds of consumptive goods : thus the recovery of
Indian agricultural prosperity in 1910 led to a sudden boom in
copper imports for braziers, followed by a slump (cf. R. of I. T.,
1911-12, p. 19).
2 It was, moreover, accompanied by peculiarly aggravating
circumstances, among others the natural propensity of the human
mind to expect too much in the way of economic advance from
military success and political readjustment.
3 Indeed the production of coal sustained a continuous rise
throughout the decade, and that of pig-iron fell only in 1886 : but
the production of steel fell off somewhat after 1882.
t8'4 STUDY OF INDUSTRIAL FLUCTUATION.
steel, a cheapening of the costs of constructional production.
It involved a permanent lowering of the cost of production
of finished goods and services, a permanent increase there-
fore in the volume thereof produced, but only a temporary
elevation of demand for the instrumental goods required to
produce them. Consequently reaction in the construc-
tional industries was comparatively prolonged and severe.
Here, however, a new distinction must be drawn. The first
wave of electrical development consisted chiefly in the
electrification of the means of transport and illumination,
which must be supplied at once upon the largest possible
scale.1 Hence the progress of investment was particularly
discontinuous, and the reaction on constructional industry
particularly severe. But in the new century electrical
power was more systematically applied to miscellaneous
manufacture,2 — a proceeding which involves not the whole-
1 In England, says the Economist History of 1895, " electric
lighting and transmission have been greatly developed during the
past year ; the various systems have passed the experimental
stage, and trustworthy data are available in regard to cost and
efficiency. Transmission of power up to twenty miles can now be
depended on." The great companies undertaking the supply of
electric light in the metropolitan area were mostly established
by 1897, and in 1896 the principle of electric traction, hitherto used
only in mines, was employed in the new tube railways. The much
slower progress of general electrical engineering is indicated by the
following figures, quoted by the Economist (1898, p. 1489) from
Garcke's Manual of Electrical Undertakings : —
Capital invested in 1896. 1897. 1898
Electric Traction . . £6,084,672 £8,553,173 £14,406,140
Electric Supply Cos. . 5.831,073 6,647,792 " 8,407,628
Municipal Electric Supply
Companies . 1,967,000 3,509,317 5,734,938
Electric Manufacturing
Companies . 6,596,244 8,519,430 10,535,937
In America the abortive boom of 1895 seems to have been largely
due to the substitution of copper for iron wires in the existing
telegraph and telephone plant of the country (Lewis, Circular of
1896)— a process which would neither be very prolonged nor need
repetition.
2 In support of this statement it may be noted that the chief
demands on German industry in 1905 were no longer for transport
but for hoisting, pumping, etc., in iron and coal works (Ec., 1906,
p. 501) : that while the number of workmen employed in the
two great manufacturing trusts of Berlin rose by 10 or u per cent.
CRISIS AND DEPRESSION. 185
sale application once for all of one great invention, but an
infinite variety of miscellaneous separate inventions applic-
able to various trades. " Inventions in one country "
and in one industry — " serve as a basis for new advances in
another," l and the whole process is diffused and prolonged :
the constructive does not march so far ahead of the operative
stage of industry. Consequently the industries of electrical
construction suffered but little in the depression of 1 907-9. 2
in 1905 on 1904, the number employed in " small electrical concerns
engaged in manufacturing one or more special articles " rose about
20 per cent, (ibid.) ; and that the building boom in 1906-7, says
the Economist (I have been unable to verify the reference, and
cannot be certain of the statement), was even more a boom in
factory construction, such as would naturally be generated by the
prospect of cheaper power, than a boom in domestic construction,
such as would naturally follow a cheapening of transport. For
an interesting account of the application of electricity to the tex-
tile, coal-mining, steel, glass, flour and other industries of Lanca-
shire, see Times, Engineering Supplement of July 9, 1913, p. 9.
1 Lewis, Circular of 1905.
2 In 1907 the copper consumption of England, Germany and
France suffered but little diminution, while that of Italy and Austria
increased. In 1908 the consumption of England rose by 1,700
and of Germany by as much as 37,000 tons, while that of France,
Italy, Austria and Russia also increased. The survival was more
marked in Germany (where the conditions relevant to invention
are always of supreme importance) than in England. In England
in 1908 " electrical machinery and equipment makers were com-
plaining of the insufficiency of employment " : and in 1909, while
the copper consumption of Germany rose by 4,000 tons, that of
England fell by 11,000 tons. In America, the copper industry seems
to have been still more at the mercy of the discontinuous type of
demand. Consumption fell by 61,000 tons in 1907 and 33,000 tons
in 1908 ; and in the latter half of 1909 we still find the Engineering
and Mining Journal complaining that " telephone, lighting and
power, etc., have already been exploited, and the cream has been
skimmed off hydro-electrical development" (Ec., 1909, ii. 335), and
that there seem no new worlds to conquer. It was only the cereal
conditions of that autumn that re-established the flow of invest-
ment.
Even in Germany the copper industry was unable to rescue
from depression its iron and steel colleagues. For while its influence
upon them had been less intense than in the 1900 boom (cf. p. 159)
it had also been less warrantable : the iron and steel used for the
factories, steel works, etc., built in 1905-7 only expected to require
the co-operation of copper instead of being physically united with
it, and the check to comparative over-investment was correspond-
ingly small.
1 86 STUDY OF INDUSTRIAL FLUCTUATION.
The most recent phase of electrical development — the
electrification of ordinary railroads — must be expected to
be less continuous in its course, and to involve more serious
constructional relapses. Thus, while in Germany in 1911-12
" with the growing application of electrical force to railways,
the electro-technical industry has received another impetus, ' ' 1
it has not been able to retain its position in more recent
months ; and the drop of some 6,000 tons in English copper
consumption in 1912 and again in 1913 is perhaps to be
connected with the conclusion of the first wave of non-
metropolitan railway electrification, in which Newcastle
and Liverpool took the lead.2
It is clear then that we must be on our guard against
condemning as over-investment what is really only un-
avoidable preliminary investment upon an exceptionally
large scale. We must be careful not to blame the Swiss
Family Robinson for sitting down to make bows and arrows
instead of catching the cassowary — a swift-footed bird — by
putting salt upon its tail. The period of gestation cannot
be shortened except at the cost of a miscarriage. At the
same time there seems little reason to suppose that the
process of investment is ever actually checked at the ideal
moment. It seems clear, to return to the island metaphor,
that as a rule the happiness of the family through time
could be increased if some of them, instead of making bows
and arrows, were to pursue the eggs or the young of the
cassowary, which can be caught with the hand, or even to
dance and sing for the edification of the rest. But they are
1 EC., 1912, ii. 233.
2 Similarly the developments in America in 1909-10, of which
the most notable were the electrification of the New York and
Chicago terminal sections of the Pennsylvania and Illinois Rail-
ways respectively (Ec., 1910, i. n), were followed by a drop of
35,000 tons in copper consumption in 1911. There was a revival
in 1912, but a decrease in 1913 variously estimated at 23,000 and
40,000 tons.
It seems reasonable to suppose that (since it is the vehicle and
not the permanent way that is in question) the " oilification " of
the world's mercantile marine is likely to be less discontinuous and
so to involve smaller fluctuations of constructional industry than
the electrification of railroads.
CRISIS AND DEPRESSION. 187
obsessed by the danger of being unarmed : they keep the
extent of their preparations dark from one another ; and
they forget that (especially if they enjoy the advantage of
electric light) one bow and set of arrows can be made to go
further if they work in shifts than if they all want to use
them at once. Further, there is a tendency among them
to suppose that their existing bow and arrows, with which
they have been accustomed to shoot cassowaries in Cocoa-
nut Grove, will be ineffective now that the herd has migrated
to Silver Creek. Finally, as a matter of fact, a number of the
family are not engaged in making bows and arrows at all,
but in constructing an island railway capable of transport-
ing the corpses of a hundred hypothetical cassowaries yet
unhatched.
§ 4. THE SURVIVAL OF CONSUMPTION.
The main conclusions of the last two sections may be
summarised as follows. First, the relapse in constructional
industry is seen to be due to the existence or imminence of
an over-production of instrumental as compared with
consumable goods. Whether or not this over-production
is indicated by an actual shortage of consumable goods
which renders it impossible to maintain investment on the
scale which has prevailed during the preceding years or
months, and whether it is due to miscalculation or to the
inevitable characteristics of modern large scale production,
its essential nature is the same, — a failure to secure the best
conceivable distribution through time of the community's
consumption of consumable goods. The aggregate satis-
faction of the community over time is thereby diminished,
and the damage to that extent final and irremediable.
But secondly, so far as our argument yet goes, there is
no reason for the consumptive trades as a whole to be
adversely affected by the constructional collapse. The
constructional boom was only the necessary preliminary
stage ; to minimise the damage caused by over-investment,
all that is needed is for the consumptive trades to produce
188 STUDY OF INDUSTRIAL FLUCTUATION.
upon the largest scale for one another's consumption.1 The
classical objection to the possibility of a " general over-
production,"— the objection that since the various products
will form a market for one another there is no sort of neces-
sity for a general accumulation of stocks or restriction of
output, seems so far to be substantially valid.2
1 This argument appears to me untouched by current state-
ments of the doctrine of " repercussion " (cf. Lescure, passim).
It is true that the trades particularly dependent for a market on
constructional producers will be temporarily affected : but no
reason is given why the distress should be either general or per-
manent.
Even Dr. Marshall's mode of expression seems to me to darken
counsel: "In short," he says (Principles, p. 711), "there is but
little occupation in any of the trades which make fixed capital.
Those whose skill and capital is specialised in these trades are earn-
ing little, and therefore buying little of the produce of other trades,"
— a remark which is undoubtedly true as far as it goes, but throws
little light upon the difficulty.
2 This proposition is denied, on apparently inadequate grounds,
by M. Aftalion. " II n'est pas certes trop de produits dans le sens
qu'ils depassent les capacites materielles de la consommation, ou
qu'ils excedent la puissance de nos desirs." Nor is there neces-
sarily any alteration in their mutual values in exchange ; ' ' toutes les
marchandises acheteront bien autant d'autres marchandises qu'au-
paravant. Leur valeur d'echange aux unes et aux autres ne sera peut-
etre modifiee " (Journal d' Economic Politique, 1908, p. 702) . But there
is nevertheless a genuine condition of general over-production due
to a general lowering of marginal utilities. " Par exces des biens
relativement aux besoins, on entend simplement que pour nombre
de marchandises diminue 1'intensite des derniers besoins satisfaits,
par suite aussi 1'utilite finale " (Les Crises Periodiques, ii. 277).
The fallacy of the argument as it stands is obvious : there seems
at first sight no sort of reason why I should be depressed if the
marginal utilities of all commodities I possess are lowered provided
their total utilities are increased, and provided there is no alteration
in the ratio of exchange against my own particular products. It
is true that M. Aftalion subsequently makes the much more fruit-
ful suggestion discussed in the next section ; but it appears to be
added somewhat as an afterthought, and throughout his whole
discussion he constantly succumbs almost without a struggle to
the natural tendency of Man the Producer to regard abundance as
m itself an evil. It is without any sense of paradox that (e.g., in
vol. ii. pp. 182-3) ne speaks of a time in which people have got less
than they want as a time of prosperity, and vice versa, or makes
use of an unfortunate series of metaphors in which the flow of com-
modities is compared successively to a swarm of malignant microbes
and to the flow of calories by which a room is " fatally overheated."
CRISIS AND DEPRESSION. 189
And indeed it seems to the present writer that there is
strong and abundant evidence that such a survival of con-
sumptive over constructional industry is a normal occur-
rence,— that constructional relapse and the close of the
period of gestation tend, as we should expect, to be followed
by an increase in the mutual consumption of each other's
products by consumptive producers. This important
phenomenon has been much obscured, owing partly to the
popular habit of regarding high money prices as the sole
criterion of prosperity, partly to the unsatisfactory character
of most of the available indices of unemployment, in which
constructional industry is greatly over-represented. In
support of this proposition it will perhaps be sufficient to
call three witnesses, — coal, wool and the consumption of
various kinds of food.
A first glance at the curves of production and price of
pig-iron and coal shows so marked a correspondence between
them that it is not uncommon to assume that the relation
that exists is of a very simple nature. Coal is one of the two
principal materials required for the production of pig-iron ;
hence an increased prosperity or depression of pig-iron leads
naturally and immediately to an increased prosperity or
depression of coal. If, however, we go behind the broad
annual figures and examine the actual order of events, this
conclusion can no longer be maintained.
In the first of the boom periods under review, the con-
nection does indeed appear to be of this simple kind. The
immense rise in the price of the two products took place
almost simultaneously in the latter half of 1871, and was
supported throughout 1872 by the German industrial boom
and the increased activity of home shipbuilding. Iron
prices declined on the whole after June, 1873, the price of
coal having broken in February, that is to say, shortly after
the first symptoms of restriction of pig-iron production
became apparent towards the end of 1872. It would seem
that the prosperity of coal was at this time mainly dependent
on the demands of the iron industry.1
1 Some contemporary observers were inclined to attribute the
igo STUDY OF INDUSTRIAL FLUCTUATION.
In subsequent periods, however, there has been a marked
tendency for the prosperity of coal to be independent of
that of iron, and to depend for its continuance upon the
demands of the transport and consumptive trades.1 Thus,
excessive advance in the price of coal to the increase of steam at
the expense of sailing tonnage which was brought about by the
opening of the Suez canal in 1869. There can be no question that
there was such an increase, nor that a reaction in favour of sailing
ships synchronised with the low coal prices of 1874-6. Cf. the fol-
lowing figures of new tonnage (ist Financial Blue Book, Cd. 1761 of
1903, P- 379) :—
Sailing. Steam.
1868 . . 246,358 H5,970
1869 . . . . . 242,562 145,530
1870 ...... 126,461 267,896
1871 ...... 60,260 330,798
1872 58,757 415,961
1873 ...... 89,626 363,917
1874 189,094 414,773
1875 . . . 245,357 226,701
1876 241,088 136,932
Nevertheless, certain facts appear to tell strongly in favour of the
view adopted in the text. In the first place the large output of
steam tonnage in 1874 was quite unable to repair the break in
coal prices. Secondly, the Select Committee of 1873 found that
of the increase of 19^ million tons in the consumption of coal in
1873 over that in 1867 (i23f million as against 104 million tons),
6£ million tons was used by the blast furnaces, 10 million tons in
other kinds of iron manufacture, and only i\ million tons for export
and i million for bunkers (quoted by D. A. Thomas in Stat. Jour.,
1903, p. 485).
1 The change was very marked during the seventies and early eighties.
On the one hand iron had been abandoned as the material of rails
in favour of steel, whose production required only one-fourth as
much coal per ton (Times, Jan. i, 1878). On the other steam had
finally established its victory over sail at sea, and in the vast new
tonnage of 1881-3 steam vessels so largely predominated that the
effect of their demand upon the price of coal was very considerable.
Total. Steam.
1880 ...... 472,896 414,831
1881 608,878 515,110
1882 733.051 635,212
1883 ...... 892,216 744,126
(In this period, however, as in the last, and again in 1890-2,
high coal prices seem to have started a reaction in favour of sailing
ships.)
Moreover, there has been a continuous progress in economy of
coal in the production of pig-iron. Thus : —
CRISIS AND DEPRESSION. 191
while iron prices began to decline from January, 1880, coal
prices remained depressed till 1882, rose slightly in that
and more in the following year, and were fairly well main-
tained in 1884 when iron was slumping. Again in 1900
the first shock to iron came with the cut in American wire
prices in April, while coal prices remained intact till October,
and throughout 1901 the reports of the trade continued
cheerful. Again in 1907 coal continued to boom when the
In 1830 7 tons of coal were used per ton of pig-iron.
In 1866 3 tons 7 cwt. of coal were used per ton of pig-iron.
(Price-Williams, loc. cit.)
In 1 88 1 2 tons 3 cwt. of coal were used per ton of pig-iron.
In 1888 2 tons 0-4 cwt. of coal were used per ton of pig-iron.
(Chisholm, loc. cit.).
In 1901 i ton 12 cwt. of coal were used per ton of pig-iron.
(Jeans, English Iron Industry, pp. 18, 45.)
Mr. Price-Williams estimates that whereas of the coal consump-
tion of 1869 15-21 per cent, was used in the manufacture of pig-
iron, 15-00 per cent, in other kinds of iron and steel manufacture,
3-05 per cent, in steam navigation and 9-10 per cent, for export,
of the consumption in 1887 only 9-44 per cent, was used in the blast
furnaces and 7-02 per cent, in other kinds of iron and steel manu-
factures, while 8-48 per cent, was consumed in steam navigation
and 15-09 per cent, exported (Statistical Journal, 1889, p. 38). This
estimate, however, is arrived at by attributing the same percentage
of consumption (23-58 per cent.) to general home manufactures
in both years ; and in view of the improvement of processes it seems
likely that there should have been some reduction in the latter
year. Mr. Chisholm's estimate of the consumption of the various
iron and steel industries at 33 £ per cent, in 1869 and 20 per cent,
in 1887 is perhaps therefore nearer the mark (Statistical Journal,
1890, p. 567). In the same sense Mr. D. A. Thomas compares the
conclusion of the 1873 Committee cited above with his own estimate
that of the increased consumption of 1900 over 1895 (225 million
tons against 190 million tons) 15^ million tons is to be credited to
the demands of navigation and export, and only 3^ million tons to
the blast furnaces (Statistical Journal, 1903, p. 485).
At the present time the proportion of the yearly product exported
is even greater (in 1913 it was no less than 36 per cent.), nor is it
to be inferred that the coal exported is any more dependent on
the demand of the iron and steel industries than that which remains
at home. Of our exports in January to March, 1903, according
to Mr. Thomas, 82 per cent, was steam coal, mainly for use in
navigation, while only 3-5 per cent, was in the form of coke suit-
able for steel manufacture (Statistical Journal) .
192 STUDY OF INDUSTRIAL FLUCTUATION.
price of iron broke in March,1 and was still rising in Sep-
tember ; for the demands of the world's enormous merchant
fleets were undiminished, 2 and even in November it was
only a " gentle decline " 3 in coal prices that set in. Simi-
larly in 1913-14, not only has the export trade in coal been
abnormally prosperous, but the inland trade also appears
to have " maintained its position."
The evidence of Great Britain is confirmed by that of
the United States. There the maximum of bituminous
coal prices has generally been attained the year after that
of pig-iron prices, to wit in 1874, 4 i88i,5 1888, 1891, 1896,
1903. There is also a maximum in 1893, the year after the
maximum of constructional activity as indicated by pig-iron
production though not by price ; and a substantial rise in
1901, when pig-iron prices fell away.6
In Germany, on the other hand, the coal industry seems in
its natural state to be bound more closely to the wheels of
iron. In the Saar district, for instance, in 1899 it was esti-
mated that 28-04 per cent, of the total product was con-
sumed by the various iron industries,7 and in the Ruhr
district the proportion was said to be still higher.8
1 Economist, 1907, pp. 187, 192. z Ibid., p. 1535.
3 EC. H. of 1907.
4 In this case two years. For the figures and their sources see
Chart II. The figures for 1902-3 are of course largely affected
by the great Pennsylvania strike. I have taken the bituminous
index, as throwing more light on industrial conditions than that
for bituminous and the domestically consumed anthracite com-
bined.
5 In France in the early eighties the consumption of coal survived
that of pig-iron by two years (de Foville, Economic Meteorology,
Jour. Stat. Soc. of Paris, 1888).
6 In 1907 also there was still in October a " rising demand for
coal," long after the constructional slump had begun (Economist,
I9°7> P- 1360). One of the most remarkable features of the curve
is its flatness compared with that of pig-iron during the 1907 boom,
which furnishes additional evidence of the comparatively slight
extent of the direct dependence of coal on the iron industry. The
1907 iron boom was largely a matter of electric transport and of
structural steel for building purposes, neither of which lead to any
very considerable subsequent demand for coal. Coke prices, how-
ever, rose considerably under the influence of steel.
7 F. Walker, Monopolistic Combinations in the German Coal
Industry, p. 12. 8 Ibid., p. 231.
CRISIS AND DEPRESSION. 193
It is true that while the price of iron broke in 1873, that
of coal continued to rise in 1874 1 : but in 1883 and 1890
there are no appreciable signs of similar independence.
The remarkable maintenance of coal prices at the expense
of iron in 1900-1 must therefore be ascribed mainly to
cartel action.2 Yet even so it is noteworthy that while
the Ruhr Syndicate's base price of coking coal, which is
most directly dependent on the iron industry, rose by 50
per cent, in the iron " essor " of 1897-1900, and fell by 5 per
cent, in the following year, that (e.g.) for gas coal rose only
16 per cent, in the same period, and was maintained in 1901
at the 1900 figure.3 Again, though the Coke Syndicate
was able, by means of its " two-year fusion price " deter-
mined in 1899, to maintain the price of coke throughout
1901, the fiscal mines of the Saar were forced to submit to a
reduction of average proceeds per ton of coke from 20-73 m.
in 1900 to 20-36 in 1901, while they raised the price of
" greasy coal " (Fettkohle) from 11-4 m. to 12-5 m.4 Thus
even in Germany we can detect traces of a natural independ-
ence of coal, as opposed to coke, of the demands of the iron
industry.
The conclusion then is that the demand for coal is gener-
ally maintained for an appreciable time after a construc-
tional collapse by the continued prosperity of the consump-
tive trades.5 We must indeed be on our guard against
1 When Ruhr Forderkohle on the Essen Borse reached the long
unbroken record of 15-80 m. (ibid., p. 180).
2 Indeed the tyranny of iron may be regarded as one of the
chief causes of combination in the German coal trade. Germany
more even than any other modern nation has won her way to
industrial pre-eminence through the power of iron, and her coal
producers have neither the vast railway mileage of America nor
the vast mercantile fleet of Great Britain to fall back upon.
3 Vide figures quoted by Walker (op. cit., p. 182).
4 Ibid., pp. 187, 197. Similarly in Upper Silesia the fiscal mines
made heavy reductions in the price of coking coal, but advances
in the price of other kinds in 1901 (ibid., pp. 186, 189).
5 Cf. the manner in which, within the constructional trades
themselves, the maximum of brick prices in the United States is
reached sooner (1886, 1898, 1901, 1906) than that of iron prices
(cf. Chart II.) : the construction of brick buildings is even more
obviously a preliminary and transitory stage than is that of iron
O
194 STUDY OF INDUSTRIAL FLUCTUATION.
misinterpreting the evidence in certain cases. In the first
place, as in England in 1900, the high price of coal may be
due to exceptional causes such as war,1 and be an actual
source of embarrassment not only to constructional but also
to consumptive industry.2 Secondly, a large volume of
railway and shipping traffic, which are among the principal
causes of a high demand for coal,3 do not always indicate
genuine prosperity, but sometimes a large temporary
transport of goods for forced sales, which cannot be expected
to continue, and which is contemporary with a restriction
of output.4 Finally, we have seen in discussing the shipping
trade that the operation of a large quantity of tonnage by
and steel goods, and the iron trade discards brick — kicking away
the ladder upon which it rose — in precisely the same way as it is
itself discarded by trade in general at a slightly later date.
1 General manufacturing industry gave poor support to the
coal trade in 1900: cf. EC., 1900, p. 1533, which attributes special
importance to the slackening of demand from the cotton trade.
2 In Germany also complaints were heard from miscellaneous
industry, but in Mr. Walker's opinion they were not well founded
except possibly in the case of cement and lime (op. cit., p. 244).
8 In 1890 in particular (apart from an increase of i m. tons in
the exports to France) the prosperity of home railways, whose
goods traffics receipt increased by about 4-8 per cent., seems to
have been the most important factor in sustaining the demand for
coal. It should be noted that on this occasion the survival over
constructional industry was more prolonged than might appear
from the pig-iron figures, for the price of pig-iron was unnaturally
sustained by the Scottish strike : the prices of finished iron began
to drop considerably earlier.
4 With the transport industries in this connection must be classed
the enveloping trades. Thus the prosperity of our tin-plate trade
in late 1907 and 1908 was due partly indeed to an increased demand
from Norway, Rumania and Japan, but partly to the enormous
liquidatory exports of the United States. The jute trade might
have been expected to share in this prosperity, but there had
been too much speculation : " As an instance of the state of the
burlap stock of New York it may be mentioned that the warehouse
accommodation was taxed to the uttermost, and thousands of
bales for which room could not be found had to be sent to Brook-
lyn " (Ec. H. of 1908), and demand was cut off "as with a knife/'
The jute trade has been doing remarkably well in recent months :
and the depression in tinplates in 1913 must be attributed entirely
to excessive investment, demand, especially from the United States
and Canada, having been exceptionally active.
CRISIS AND DEPRESSION. 195
no means always implies a large volume of sea-borne trade.
Nevertheless, the evidence seems to be important so far as
it goes.
The next witness is the woollen trade. We have already
had occasion, in connection with the course of wheat prices,
to point out its frequent survivals over the collapse of con-
structional industry. Here it must be added that there
are signs of the same tendency even when the price of wheat
affords no assistance. The trade, for instance, remained
good throughout 1891 ; and in 1908, though it suffered to
some extent from the depression, in particular, of its ship-
building clients and from the disputes in the cotton * and
engineering trades, it was by no means plunged in unmiti-
gated gloom. " Fairly good," " quiet " and " moderate "
are the worst the Labour Gazette has to say of employment
in this year : and by October we hear of " the improved
appearance of factories " and learn that " there are signs
of wholesale buyers being more disposed to consider the
placing of orders on a more liberal scale," that " quite as
many patterens have been taken up as in past years, and
if that is anything to go by, an average trade should be
done," and that " throughout the woollen area a fairly
hopeful feeling over the future prevails." 2 Similarly in
the following March we learn that on the whole the " time
worked and the output of goods have been on a liberal
scale," that in the Dewsbury and Batley district there has
been a " fair amount of prosperity," and at Huddersfield
" the business done, though disappointing in some respects,
has not all round been unsatisfactory " ; while at Leeds,
" some factories have not shown the least sign of depression
* For though the two industries are undoubtedly to some extent
rivals, yet the effect of a dislocation of this kind in upsetting general
business confidence in the neighbouring county seems to be more
immediate and important than the effect in diverting demand, at
any rate when, as in this case, the dislocated industry is suffering
from glut. Moreover the wool industry is affected injuriously by
a shortage of cotton warps.
* Economist, 1908, ii. 51,
196 STUDY OF INDUSTRIAL FLUCTUATION.
during the past year." 1 Throughout 1909 the wool
business was " sound, with almost entire absence of specu-
lation " — a fact which is all the more remarkable in view
of the wheat shortage, which brought the Gazette average
for 1909 up to 365. nd.
Similarly it is worth recording that in 1908 " the boot
and shoe trades were very well employed throughout the
year," and during 1909 trade was active, and manufacturers
" have obtained an advance and are well supplied with
orders."
The last witness is the consumption of various kinds of
food-stuffs, as revealed by the index already discussed in
Book I. ch. vii. § 3. A reference to the curve will show
that it tends to survive the fortunes of construction, and
even sometimes of the wool industry, reaching maxima in
1876, 1885, 1891 and igoi.2 Similarly in the United States
there are clear traces of food consumption surviving con-
struction in the fiscal years 1891, 1901, 1904-5 and 1910 3 :
though in that country the dependence of both kinds of
activity on the prosperity of cereal producers tends to
obscure the issue.
It should be observed that in both countries the consump-
1 Economist, 1908, ii. 751.
2 The failure of food consumption to join with wool in an attempt
at survival in 1908 may be attributed partly (but cf. § i of this
chapter) to the consideration suggested by the remarks of Mr. Wood
and Mr. Bowley (Stat. Jour., 1899, loc. cit.} to the effect that as the
standard of living becomes higher " an average fall of wages of 10
per cent, leads to retrenchment in other ways than would have an
effect on our consumption curve ; but 10 per cent, of the trade
union members being totally unemployed means a necessary economy
on every hand by those who are unfortunate enough to be so un-
employed." That is, the growth of the standard of comfort leads
to a closer connection between the curves of employment and food
consumption.
The consumption of meat, owing to a considerable fall in the price
of bacon and a moderate fall in that of mutton, suffered no diminu-
tion in 1908.
3 The fall in coffee consumption in 1911 was just more than
compensated by an increased consumption of tea : in all the years
from 1884 to 1910 the curves of coffee and of tea-and -coffee move
in the same direction.
CRISIS AND DEPRESSION 197
tion of alcohol shows affinity with constructional rather
than with other forms of consumptive activity. Thus in
England there are maxima in 1873, 1881, 1890, 1899, and
a marked check in the trans-cyclical fall in 1906-7. l The
inference is that the consumption of alcohol is more subject
than that of other things to the psychological influence of
hope and excitement, and is a better index of mental
temperature than (as Mr. Beveridge, for instance, is inclined
to treat it) of genuine prosperity.2
But indeed detailed illustration of the habitual survival
of certain branches of consumptive industry is scarcely
necessary. The most prominent industrial fact of the year
1 I have weighted beer twice as heavily as wine or spirits. The
curve also shows supplementary maxima in years when there is a
large sudden rise in the employment curve, to wit 1887 and 1896,
and afterwards relapses somewhat. In 1876 it also shows a maxi-
mum synchronising with that of general consumption, but in sub-
sequent periods this correspondence is lacking.
Mr. A. D. Webb, however (Stat. Jour., Jan. 1903), asserts a corre-
lation between real wages and the consumption of beer in the fol-
lowing year.
1 have been told that the diminished consumption of beer since
1900 is partly due to the attractions and increased cheapness of
bananas as a food and occupation on public holidays : and the
diminution in the rate of increase of the imports of this fruit in
1906, accompanied by higher prices, the actual decrease in 1907
and again in 1910 is certainly curious when compared with the beer
consumption curve. Cf. the following figures : —
Imports of Bananas.
1904. 1905. 1906. 1907. 1908. 1909. 1910.
3-9 5-7 6-4 6-2 6-4 6-2 6-1 m. bunches.
1-4 1-8 1-9 1-8 1-8 1-8 1-7 m. £
2 It is interesting in particular to compare the figures for beer
and tobacco in, for instance, 1877, the early nineties, the early
nineteen hundreds, and 1908, in the light of the empirical general-
isation that people tend to drink more when they are excited and
to smoke more when they are depressed. In America the only
exception to the rule in the text is the relapse of alcohol consump-
tion (wines and distilled spirits and malt liquors) in 1898-9, which
must be attributed to the increased excise duties on malt imposed
by the act of June 13, 1898. It should be noted, however, that
the constructional depression of 1903-5 is accompanied by a mark-
time and not an actual relapse of the curve, the trans-cyclical trend
of which is in the opposite direction from that for England.
ig8 STUDY OF INDUSTRIAL FLUCTUATION.
1913-4 has been the failure of general industry to follow
in the wake of the relapse of the constructional trades.
The continuance of good harvests and (save for certain
exceptions in India and Canada) of financial moderation
have combined to make the ebb of 1913-4 an almost pure
example of the type of reaction discussed in the last section,
occurring in its mildest form. It must not indeed be
assumed, as will be shown in the next section, that general
trade will remain immune from reaction : it is only a
minority of trades, notably building, printing, glass and
boots, which have been actually in a better position than a
year ago. But the events of the past months ought to
suffice to dispel for ever the legend of that simultaneous
cataclysmic collapse of general industry which is still too
often assumed to be the typical form of industrial decline.1
§ 5. GENERAL DEPRESSION— THE LAW OF
MARKETS.
Nevertheless it is apparent that in spite of the survival
of consumptive industry there are certain years — in Eng-
land, 1892-3 may be taken as the most conspicuous examples
— which are marked by something which may fairly be
called a general depression of trade, and of which cereal
conditions afford an inadequate explanation. We have,
therefore, to discover some satisfactory reason for the
diminished activity and restricted output of consumptive
industry in these years.
And first, postponing as before all complications arising
out of the conduct of industry upon a wage system with the
aid of a monetary economy, let us examine in its most
abstract form the argument which denies the possibility of a
" general over-production." The famous Law of Markets,
which has played so large a part in this controversy, seems
1 For evidence of the condition of industry in 1913-14, cf. the
figures of employment collected by the Lab. Gaz. and summarised
in Table VIII.
CRISIS AND DEPRESSION. 199
to be capable of two quite distinct presentations.1 It may
be so phrased as to assert that it is impossible to conceive
of an excess of aggregate production over the totality of
human wants, and in this form it appears to be beyond
dispute. It is truejrf goods in general, though not perhaps
of particular goods, that however many of them there are
they will not be beyond the physical capacity of the popula-
tion to absorb. 2 Nor is this less true in a regime of division
1 I omit the forms in which it is stated in order to be knocked
down by M. Aftalion and M. Baranowsky respectively. The former
(op. cit., Bk. X. ch. i.) argues at length that an over-production in
one industry does not necessarily imply an 'actual diminution of
output in any other — a proposition never, so far as I know, main-
tained by any one. The " under-production " in other industries
affirmed to be necessary by the Law of Markets of course merely
means a relative under-production, due to a misdirection of pro-
ductive power, and indicated by a rise in the exchange value of
their products.
M. Baranowsky (op. cit., Part II. ch. i. §§ 2 and 4) supposes the
Law to assert that all consumable goods must always be absorbed
in exchange with other consumable goods, and advances over its
dead body to the triumphant announcement of his " paradoxical "
discovery that owing to the accumulation of fixed capital the aggre-
gate production in a society may exceed its aggregate revenue in
consumable goods without a rupture of equilibrium between demand
in general and supply in general. In this progress he is assisted
by his complete misunderstanding of the classical doctrine that
" what is saved affords subsistence to the workman just as much
as what is spent." There seems no reason to believe that the clas-
sical school in its most abstract moments was unaware that while the
consumable capital " saved " is consumed by the makers of the
instrumental goods for which it is in effect exchanged, these instru-
mental goods themselves are consumed, i.e., held in use, by the
capitalist class ; and no reason to suppose that the Law of Markets
in any form excludes consumption of this latter kind from its pur-
view.
2 We need not, I think, delay over the suggestion that a " general
over-production " merely means a comparative over-production
of all those goods which are being produced at all, and so implies
not the satiability of all wants but of all those wants towards satisfy-
ing which any attempt at all is being made. This suggestion is
countenanced rather obscurely by M. Aftalion: "La crise ne con-
siste que dans la moindre desirabilite des biens actuellement exist-
ants. Elle ne suppose pas un amoindrissement de la capacite
humaine d'aspirer toujours a plus de bien-etre " (/. d'E. P., 1909,
p. 89), and appears to involve no theoretical absurdity ; but it is
clearly quite irrelevant to the facts.
200 STUDY OF INDUSTRIAL FLUCTUATION.
of labour and exchange than in a Robinson Crusoe economy.
While it may be to the advantage of particular groups of
owners to destroy the stocks of their product rather than
dispose of them, it can never be to the advantage of owners
in general : if everybody sells freely the various products
will provide a market for one another, and the aggregate
satisfaction of the community derived from the consumption
of goods will be greater than it would otherwise have been.
But the Law of Markets is sometimes invoked in support
of a more presumptuous claim. Since the different products
will find a market in each other, there can be no objection,
it is urged, to continuing production on the largest scale
that is physically possible ; any general restriction of
output is obviously unnecessary and wasteful. The simplest
way of testing this assertion is by the method of reductio
ad absurdum. Supposing it were physically possible for
the human race to maintain its productive efforts through-
out the whole twenty-four hours of every day, would it be
to its economic advantage to do so ? It becomes clear at
once that it would not, and that the second form of the law
of markets is based upon the_elementary fallacy that the
aim of economic endeavour is to maximise the gross satis-
faction derived from the consumption of goods, instead of
the excess of this satisfaction over the aggregate dissatis-
faction involved by all kinds of effort and sacrifice. We
have already seen that the temptations to over-investment
may involve a general rupture between the sacrifice involved
in postponing consumption and the future satisfaction
procured by means of that sacrifice. It remains to add
that a point may be reached where there is a similar rupture
between current effort and immediate satisfaction,1 and
that the operation of industry in general on the full scale
rendered physically possible by the previous orgy of invest-
ment may involve a " general over-production " in a very
real and genuine sense.2
1 In other words, where the effort demand for commodities in
general becomes inelastic.
2 It will at once occur to the reader that the loud complaints
heard during a depression of distress from want of employment
CRISIS AND DEPRESSION. 201
Were the initial over-investment equal in all the con-
sumptive trades, all alike might stand to gain by a restric-
tion of production. In fact, however, the necessity is likely
to be confined to some. Indeed we get what at first sight
appears the rather startling paradox that those trades in
which over-investment has been most severe may be pre-
cisely those which will be least wise to restrain output.
For in a period when the effects of general over-investment
are being felt, the more dysenteric trades — those, that is,
which are producing on a large scale from an expensive
fixed plant, and in which general are high relatively to
special costs — are likely to be the hardest hit : for the
exchange value of their products is likely to be consider-
ably lowered, and the (discounted) satisfaction derived from
the sale thereof to fall particularly short of the satisfaction
surrendered in investment during the boom period. But
on the other hand owing to the fixity of general and the
relative smallness of special costs, they will be the most
likely, provided the demand for their products is elastic,
to lose in net satisfaction by any present restriction of
output below the level of which they are physically capable. 1
render the suggestion irrelevant. Whether this objection is valid
will be considered in connection with the peculiarities of the wage-
system ; here it is enough to insist that a general over-production
in this sense is at least theoretically possible.
The credit for the suggestion seems to belong in the main to M.
Aftalion, who in a very pregnant chapter (op. cit., Bk. X. ch. viii.)
at length finds the essential cause of depression in a rupture of
equilibrium between cost (in the widest sense) and result. " Si
les marchandises . . . se trouvent n 'avoir en fin de compte qu'une
utilite finale inf erieure a celle qui etait escomptee, a celle pour laquelle
on avait consente des sacrifices accrus en travail et en capital, il y
a perte non seulement pour 1'entrepreneur mais encore pour la
societe. ... Le resultat est inferieur a la peine " (op. cit., II.
335)-
1 Diagrammatic representation, though not completely satis-
factory, will perhaps be found useful by some.
Units of (objective) effort are measured along OX, of satisfac-
tion along OY. We may regard OM units of effort as being crystal-
lised in the supplementary costs of the business. UU' is the curve
of satisfaction productivity, EE' of disutility of effort. The pro-
ductivity curve of the units of effort involved in the special cost of
202 STUDY OF INDUSTRIAL FLUCTUATION.
Accordingly we find, for instance, that while the woollen
industry complained loudly of depression in 1892-4, l the
consumption of raw wool increased from 443 m. Ibs. in 1892
to 462 in 1893 and 485 in 1894. In these years, as previ-
ously in 1876, Messrs. Schwarze point out that the price
of the raw product was maintained chiefly by " the demands
of a superabundant machinery." Again even when the
dysenteric trades succumb to the temptation to restrict
output, they are generally early in discovering their mistake
and in assuming the hegemony in the rebirth of prosperity.
Thus in 1886 and again in 1894 2 the wool trade was among
the earliest to recover. Again after the 1908 depression
the woollen and worsted trades improved from November-
December, 1908, and hosiery from February, 1909, while
additional output may be regarded as starting afresh from U 2 ;
and (if output is not so much increased that R on U2U3 lies to the
right of the intersection of U2U3 and EE'), the excess of UEH
over QHU' is increased, or the deficiency diminished, by the whole
amount U2RQ'Q. [In the figure as drawn equilibrium between
utility and special costs of the additional output is just secured,
i.e., R and Q' coincide and U2RQ'Q is maximised.]
1 " Last year the review of the ready-made trade was but a
repetition of the discouraging story previously told " (Leeds report,
1895)-
2 At Leeds in 1894 " the home trade has not been worse than
others, but better than a great many."
CRISIS AND DEPRESSION. 203
(among the more dyspeptic trades) employment in printing
and pottery did not appreciably improve till August, nor
in paper till October, nor in glass till June, igio.1 The
peculiar position of the more dysenteric trades affords the
partial exception which proves the rule of the possibility,
in a genuine and important sense, of a general over-pro-
duction.
In the preceding paragraph it has been assumed that
the demand for the " dysenteric " commodity has an
elasticity greater than one. If, however, its elasticity is less
than one, the situation is entirely altered. In this case
failure to restrict output will diminish aggregate receipts of
producers and therefore the contribution made towards
compensating for the damage done by over-investment.
This fact is indeed generally recognised ; what is not
understood is that in such a case failure to restrict output
may react adversely upon other trades as well. The source
of this misunderstanding is precisely the same confusion
which has obscured the treatment of the influence of agri-
cultural conditions upon industry — namely, concentration
1 This account is on the whole supported by the interesting
address (quoted in E. H. of 1909, p. 5) made to his shareholders
by the chairman of Lloyd's Bank, in which he gave the following
rough classification.
Trades which remained bad Trades which improved appreciably.
in 1909.
Iron. Brass foundry. Armaments. Tinplates.
Cycles. Saddlery. Glass. Pottery (other
Carpets. Jewellery. Hosiery. districts).
Bricks. Timber. Mining. Lace.
Building. Brewery. Iron (some dis- Agriculture.
Pottery (some districts). tricts). Fishing.
Thus also the statistics of failures in various trades collected by
the Bankers' Magazine (Feb. 1913) exhibit a maximum in 1909 for
the leather and coach-building trades, and for that of chemist, and
an improvement in drapery, hardware and furniture among others.
Where, however (as in glass and earthenware), the evidence from
these two sources disagrees with that collated in the Labour Gazette,
I suspect the latter is the better guide.
It must be noted that the dysenteric trades were still subject to
temporary relapses, boots, for instance, in June-August, wool and
worsted in July-August, and lace in April- June, 1909.
204 STUDY OF INDUSTRIAL FLUCTUATION.
upon the alteration in the ratio of exchange instead of upon
the effect of that alteration upon the most profitable level
of production and consumption. A reference to the argu-
ment of chap. i. § 3 will make it plain that if the demand of
an industrial group A for the products of another group B
is inelastic, and if the products of B are cheapened in terms
of those of A, the scale of aggregate effort which yields the
greatest net satisfaction to A will be diminished. Until
or unless output is correspondingly reduced, A will be
suffering from " over-production " in the sense in which
that term has been used in the present section : indeed, in
certain circumstances the net satisfaction enjoyed by A
may be less than it would have been if the alteration in
the ratio of exchange had not taken place. x In such cases,
The diagram refers to group A. Units of effort along OX, of
satisfaction along OY.
CRISIS AND DEPRESSION. 205
if A is unable to restrict output, a restriction of output by B
will be beneficial to both parties.
Now it seems that there are two supremely important
groups of trades which, in a time of " general depression,"
are precisely in the position of our hypothetical group B —
the constructional industries and the industries of trans-
port. In both the unit of investment is very large, so
that they are obliged, or at least strongly tempted, to
maintain output on a very extensive scale. Both, moreover,
find the demand for their products highly inelastic. With
regard to the former, it seems clear that once the public
mind has grasped the reality of over-investment, even a
large reduction in (relative) prices will not greatly stimulate
consumption.1 With regard to the latter the matter is
even clearer : for if there has been an over-production of
transport facilities as compared with transportable goods,
there is (so far as goods traffic is concerned) a physical
limit to the consumption of transport which does not exist
with regard to any kind of tangible commodity. The
dysenteric tendencies, therefore, of these two important
groups are by no means unlikely to give rise to a considerable
measure of over-production, in our special sense, in other
trades. Those writers who, like M. Lescure, find the cause
of general depression in a " repercussion " from construc-
tional industry, while their reasoning is, as a rule, unscientific
and nebulous in the last degree, would seem to have hit
instinctively upon an important truth.
Ul^ = curve of marginal productivity of effort in terms of satis-
faction afforded by non-B commodities.
EEt = curve of marginal disutility of effort.
OK = units of effort expended upon products of group B.
The removal of K to K' removes M to M', and increases net satis-
faction afforded by non-B commodities from U'LP' to UTP. But
if OM is not in fact reduced, this net satisfaction becomes UTP-
PGP'. The flatter EE' and the steeper UU' (i.e. the greater MM'
compared to KK') the smaller is the difference between UTP and
U'LP' ; and the greater also is PGP'. In some cases then the differ-
ence between U'LP' and (UTP-PGP') may be sufficient to outweigh
the gain in net satisfaction from the products of B [not represented
in the figure] which accompanies the removal of K to K'.
* Cf. the conclusions of Part I. chap. iii. § 2.
CHAPTER III.
THE WAGE AND MONEY SYSTEMS.
§ i. THE WAGE SYSTEM AND THE VOLUME OF
PRODUCTION.
FOR the sake of simplicity the argument of the last few
chapters has been so framed as to apply primarily to a
society in which industrial policy is in the hands of co-
operative groups of producers, supplying jointly the needful
capital, enterprise and labour, and exchanging their pro-
ducts directly with one another. Existing society, however,
differs in two important respects — in the differentiation
of the capitalist entrepreneur *• from the wage-earning
workman and in the conduct of exchange by a mechanism
of money and credit. Both these phenomena introduce
important modifications into the theory of fluctuations as
hitherto presented.2
With regard to the first. In view especially of the line
of argument pursued in the last chapter, it is obviously of
considerable importance that the control of industrial
policy does not lie in the same hands as those by which one
important class of the effort necessary for production is
expended. 3 It follows that the scale of production achieved
1 Of course the differentiation is at the present day considerably
more complex. The broad classical line of cleavage seems, how-
ever, to be sufficiently accurate for most of the purposes of the pre-
sent study ; though in some connections we shall have to consider
separately the provider of capital without risk or effort.
2 Those modifications which are the joint result of the two pheno-
mena will be dealt with in the place which appears most convenient.
3 The chief suggestiveness of Professor Wesley Mitchell's recent
monumental work on Business Cycles seems to me to lie in the
206
THE WAGE AND MONEY SYSTEMS. 207
or projected at any given time is not necessarily that which
is expected to afford the greatest net satisfaction to the
community as a whole. The importance of this proposition
becomes manifest when we reflect that owing to the more
unpleasant and monotonous character of their work and
to the greater urgency of their material wants, the effort-
demand of the working classes for commodities in general
is likely to be considerably less elastic throughout its course
than that of those whom I will call for the sake of brevity
the business classes.
Thus there is good evidence that in times of boom in any
trade the high value attached to leisure by the manual
workers is a factor which operates to restrict production
below the level at which the business man desires to main-
tain it. This influence is particularly strong in the coal
trade, where the marginal utility of " a straight back and
the sunlight " is naturally peculiarly high. Thus Mr. D. A.
Thomas repeatedly complains that in times of high prices
and high wages there is a strong tendency among colliers to
work less hard, so that the output per man is diminished ;
and gives the following figures in illustration : —
Great Britain.
South Wales.
Price. Output per man.
Output per man.
1870 .
9-475. per ton 321 tons
320 tons
1874 .
10-985. „ 235 „
222 „
1888 .
8-27S. „ 333 »
1891 .
11-965. „ 296 „
Wage Index.
1899 .
IO'53S. ,, 320 „
100
294 ,>
1901 .
I373S. „ 287 „
140
2541 i.
emphasis laid on this fact. " In practice," he says, " industry and
commerce are thoroughly subordinated to business. . . . The
industrial and commercial processes by which goods are furnished
are conducted by business men in quest of profits. ... An investi-
gation into the ebb and flow of contemporary economic activity
must concern itself primarily with the phenomena of business
traffic — that is, of money-making " (op. cit., pp. 25-6). I cannot,
however, agree with him that this aspect of the question has hitherto
been unduly neglected.
* Article on coal exports, Stat. Jour., 1903. Cf . the same authority
in Fortnightly Review, 1893, p. 301, and his comments on a paper
by the present writer, Stat. Jour., Jan. 1914, p. 174.
208 STUDY OF INDUSTRIAL FLUCTUATION.
The tendency, however, is not confined to the coal-trade.
It has recently been a source of considerable embarrassment
in the Liverpool docks, where it has been intensified by the
self-denying ordinance of the employers as to the restriction
of imported labour. Again in 1889 " towards the end of
the year especially there were frequent complaints that in
some branches of trade the men were not working so steadily
as they did when wages were lower." * And when to Dr.
Rost's attribution of the German depression after 1900
in some part to the laziness of the workpeople when they
could earn as much in four days as previously in six, Pro-
fessor Ashley replies that there is no evidence for this effect
of increased prosperity, his answer does not seem to be in
accordance with the facts.2
It is by no means clear, however, that a restriction of
production arising from this source is necessarily undesirable
in the interests of national welfare. In the case of the
dockers, a notoriously improvident class, it may be so ;
though Mr. Keeling is, I think, justified in remarking that
"it is not to be expected that the docker should suddenly
become anxious to work for six days a week at a laborious
employment, in order to suit the convenience either of
employers or of social theorists. Least of all can this be
expected at times when owing to a system of overtime and
night-work he can earn relatively large sums of money in
a short time." 3 But in occupations where the equalisa-
tion of earnings between man and man and year and year
is better understood it seems even more clearly desirable
that no small part of increased prosperity should be taken
out in increased leisure. The didactic and somewhat
priggish attitude of the other classes in this matter is well
illustrated in the article already quoted from the Econo-
There seems no reason to doubt that this was an important con-
tributory cause of the temporary restriction of coal production,
pending the influx from the new mines, in 1874, 1892, 1901 and 1908.
1 Economist History of 1889.
2 Ashley, Review of Rost " Ueber das Wesen und die Ursachen
unserer heutigen Wirtschaftskrisis," Economic Journal, 1905, p. 230.
3 Economic Journal, March, 1913, p. 15.
THE WAGE AND MONEY SYSTEMS. 209
mist,1 and seems to spring partly from the superior attrac-
tiveness of their own work and partly from their habituation
to that apparent tendency of workpeople to regard work
as an end in itself which is inseparable from a wage-system. 2
But the workpeople appear to remember better than their
critics the truth of the Aristotelian sentiment
At a time, therefore, when the productivity of effort in
terms of commodities is exceptionally high, a curtailment
of output may be in the best interests, if not of the business
leaders, at any rate of the majority of producers. More-
over, since at such a time labour is likely to be scarce, and
since in some trades, especially where piece-work is the
rule, regularity of attendance and maximum intensity of
effort are difficult to secure, it is by no means unlikely that
the interests of wage-earners will be of some weight in
determining the course of industrial policy.
Now it would seem at first sight that a time at which
the effects of a general over-investment are being felt is
precisely such a time as this. The commodity-produc-
tivity of effort is enormously increased, and it might be
supposed that the result would be a general attempt by
the wage-earners to restrict output, partially successful
in its issue, but strenuously resisted by the employing
class. It is well known that the real facts are precisely
the reverse. In resolving this apparent paradox,4 several
considerations must be borne in mind. First, while post-
1 " Something of this kind, however regrettable though it be,
we must always expect to find in times of reviving prosperity.
There is a certain proportion of men who will not bind themselves
to work except under the spur of necessity, and who, when the
price of their labour rises, prefer to take the benefit of it rather
in the form of greater relaxation than of larger earnings."
2 Cf. Macgregor, The Evolution of Industry, p. 24 ; and also Taussig,
Principles.
3 Politics, VII. 14, 13.
* For M. Rist's conclusion (" Relation entre les variations an-
nuelles du chomage, des greves et des prix," Journal d' Economic
Politique, Nov. 1912) that " le chomage volontaire est d'autant
plus fort que le chomage involontaire est plus faible, et reciproque-
ment," though based on the study of a different type of voluntary
employment, is impossible to resist.
P
210 STUDY OF INDUSTRIAL FLUCTUATION.
poning still a full discussion of the direct effects of a mone-
tary system, we may without indiscretion disclose that it
operates at such a time in favour not only of the restriction
of production but of the accumulation in store of a con-
siderable part of what is actually produced. To this
extent the flow of goods available for the consumption of
wage-earners is diminished, and the rise in the commodity-
productivity of their effort is checked — checked perhaps
to such an extent that the point at which their effort-
demand for commodities attains an elasticity less than one
is never reached. Secondly, it should be remembered that
an important cause of the " general over-production "
established in the last chapter is the inelasticity of the
demand for goods transport and for the products of con-
structional industry. But this is a demand which proceeds
almost entirely from the business classes. Hence, while it
is true that the effort-demand of the working-classes for
consumable goods is less elastic than that of the business
classes, it by no means follows that in such periods as those
under consideration the same relation holds between the
effort-demands of the two classes for all those commodities
which are included within their respective economic hori-
zons. For these reasons it is plain that the scale of pro-
duction which commends itself to the business classes may
be smaller than that which commends itself to the working
classes.
Moreover, when the situation is thus reversed, the busi-
ness classes will be in a better position to enforce their
control of industrial policy. For while an employer cannot
easily compel a workman to work more than he wishes to,
he can, through his control of the access to the instruments
of production, effectually prevent him from working as
much as he wishes to. It follows that the complaints of
involuntary unemployment among the wage-earners need
not make us doubt the correctness, as far as concerns the
business classes, and therefore the course of industrial
policy, of our diagnosis of the industrial malady. It
follows also that no solution of the problem can be com-
THE WAGE AND MONEY SYSTEMS. 211
pletely satisfactory which aims merely at the fulfilment
of the policy which the enlightened self-interest of the
business classes would dictate, and neglects the genuine
want of harmony between that interest and the interest
of the working-classes. It is perhaps strange that the
numerous writers, from Sismondi to Mr. J. A. Hobson, who
have found in capitalism as a juridical system the main
cause of industrial depression, should have failed to detect
this particular manifestation of the opposition between the
interests of " capital " and " labour," and have supported a
very tenable position by arguments the validity of which we
shall find it impossible to admit.
§2. MONEY IN THE BOOM.
The influence of a money, especially of a credit money,
economy upon the course of trade is of such obvious import-
ance that it has more or less completely hypnotised all but
a very few of those who have contributed to the discussion
of this problem. The fact that our long, complicated, and
perhaps not unfruitful discussion has been conducted so
far almost entirely without reference to specifically mone-
tary phenomena relieves us of the necessity of a formal
refutation of those who, like Clement Juglar and Mr. Haw-
trey, find in monetary influences the sole and sufficient
explanation of industrial fluctuation.1 It is to be hoped
1 To these writers must now be added Messrs. Bilgram and Levy
(The Cause of Business Depressions), who find the sole cause of
depression, as of most other economic evils, in scarcity. of currency.
Interest, according to these writers, is paid not for the use of capital
goods, but purely for the use of money, the only convenient mechan-
ism by which the business man can make that selection and aggre-
gation of capital goods which he requires. But the quantity of
notes is (in America) directly, and that of bank credits indirectly,
limited by law, while that of gold is limited by nature. Pure interest
is, therefore, a monopoly price exacted by the creators of currency,
for which there is no natural necessity, but which results solely from
antiquated legal restrictions. The free " monetisation of credit "
would result in the disappearance of interest and of trade depres-
sions, and would have the further advantages of facilitating the
speedy transfer of land values to the State and of enabling the exe-
cution of public works on a colossal scale without the inconvenience
of contracting a public debt. The enormous confusions here in-
212 STUDY OF INDUSTRIAL FLUCTUATION.
that it has also saved us from the fate of those who, like
MM. Aftalion and Baranowsky, while professing to dig
below the mere money surface of things, are always prone
to relapse into monetary terms at all critical stages of the
argument. But we are not thereby absolved from con-
sidering in detail the real modifications introduced by a
monetary system into our theory as it stands.
First, then, with regard to the boom. An increased volume
of currency, whether due to an increased confidence in the
breasts of bankers, or to an increased supply of metallic
money, will tend, it need hardly be argued, to raise the
general level of prices. If all prices (including wages)
were equally affected, the result would probably be a
general increase in production beyond the point which is
in fact most advantageous : for it seems to be a natural
tendency of every man to suppose that the product which
he sells will be more rapidly and deeply affected by any
current price-movement than the products which he buys
either for personal consumption or for industrial use.1
volved scarcely need demonstration ; it is sufficient to inquire why
depressions occur in England where the volume of bank credits is
not even indirectly limited by law, and what the evidence is of that
permanent plethora of real capital crying out to be borrowed gratis
which is the basis of the authors' optimism. Cf. review by present
writer in Stat. Jour., July, 1914, p. 15.
Units of effort along OX, of utility along OY.
THE WAGE AND MONEY SYSTEMS. 213
In fact, however, the influence of an increased volume of
currency is considerably more complicated. Even if it
consists originally in an increased supply of gold, it is now
sufficiently well established that it operates through the
medium of the banks, who by lowering the rate of discount
place a larger volume of credit currency in the hands of
business men. Business men thus benefit in two ways.
First, from causes either purely psychological (an uprising
of confidence) or purely technical (an increased gold basis),
the services of bankers (which consist fundamentally in
the transfer of the control of resources from those who
possess but do not wish to use them to those who wish to
use but do not possess them) are supplied to them at a lower
money cost.1 Secondly, with the increase of currency the
goods in which they deal and of which they must dispose to
make repayment, are likely to appreciate in money value
on their hands, so that the real cost of the services of bankers
is lowered to them even more than the money cost.
Nor does this exhaust the gains of business men. Since
the service of saving (in so far as they do not provide it
themselves) is supplied to them at a cost habitually reckoned
in money, and since there is no reason in our present argu-
ment to suspect an alteration in the remuneration which
savers intend to demand, the rising prices will lower the
real cost to them of this service also. Precisely the same
may be said of the service of manual labour, save that the
sanction of habit is not, as in the case of already existing
loans, reinforced by that of law, so that the ga:n to business
men is less prolonged. But all these gains are a genuine
inducement to business men to enlarge the scale of pro-
duction, which is therefore increased still further beyond
EE' = curve of marginal disutility of effort.
A , ( marginal productivity of effort,
AA = curve of actual If J rfsen £ ^ ratio
AtAt = curve of anticipated | < » . * Q
Then total volume of effort expended = ON instead of OM and
total utility enjoyed = AONR, which while <A1ONQ is>AOMP.
1 It is, I think, worth while in the interests of clear thinking to
separate in analysis the payment for this service from the payment
for the service of saving. Cf. J. M. Keynes in EC. Jour., Sept. 1913,
P- 396.
2i4 STUDY OF INDUSTRIAL FLUCTUATION.
the point which would be most advantageous to the com-
munity as a whole than in the simple instance with which
we started this discussion.
The following figures 1 lend considerable support to the
contention that a falling money rate of discount combined
with a rising price level is of importance in the initial stages
of an industrial boom. Professor Fisher's theory of the
Average
Bank of
England
Rate of
Discount.
Prices
(Sauerbeck's
Index).
Average
Rate of
Discount.
Prices.
1870
100-0
100 'O
1896
100 'O
100 -0
1871
93'0
104-2
1897
106-0
101-6
1872
I32-2
H37
1898
I30-5
104-9
1873
154-5
II4-8
1899
I5I-0
in -5
1900
160-1
122-9
1879
IOO-O
100 -0
1880
109-9
100-0
1904
100 -O
100 -0
1881
I39'1
102-4
1905
91-2
103-2
1882
164-2
IOI-2
1906
129-7
no -8
1883
141-7
98-8
1907
146-4
115-1
1887
lOO'O
100 -0
1910
lOO'O
lOO'O
1888
98-4
102-9
1911
94-6
I02'6
1889
115-8
105-8
1912
1027
i09-o
1890
134-6
105-8
1 Collected by Miss England in a criticism of Professor Fisher,
Quarterly Jour, of EC., Nov. 1912, p. 95. The first year of the
series is in each case taken as base. I have supplemented Miss
England's figures by those for 1904-7 and 1910-12.
According to these figures the booms culminating in 1907, 1873,
1912, and 1890 were (in that order of merit) in large measure due to
inflated currency, while those culminating in 1883 and 1900 were
not. This conclusion, as regards the last-named period, is sup-
ported (in opposition to certain popular impressions about the
influence of the Transvaal mines) by the evidence recounted in
chap. i. § 2, and by the figures of the Bank's gold holdings during the
period and of the net imports of gold up till 1898. The large net
imports of 1899 must, I think, be regarded as an effect rather than
THE WAGE AND MONEY SYSTEMS. 215
paramount importance of the rate of interest,1 though we
shall find reason to attack it later, seems here at least to
be on solid ground.
The lagging of money wages behind rising prices is now
so generally admitted as scarcely to require detailed illustra-
tion : 2 but the following table is instructive : —
Money Wages. Wholesale Prices.
1879 .... 83-35 • • 125
1880 . . . 83-27 . . 129
1887 . . . .81-45 . . 98-8
1888 . . 83-58 . 101-8
1896 .... 89-24 . . 88-2
1897 .... 90-05 . 90-1
1905 . ... 95-94 .. 97-6
1906 .... 97-60 . 100-8
(Board of Trade Indices, 1900=100.)
The general result thus far obtained is liable to modifica-
tion in accordance with the features of the industrial situa-
tion. This point will require further elaboration ; at present
it need only be observed that any temptation to over-
investment that may exist will be aggravated by an increase
in currency. For in the first place the transference of
resources to business men is a transference to those who are
most prone to use resources in investment ; and further in
the contingency contemplated those business men who are
most concerned with dealings in constructional goods are
likely to obtain the largest share of the increased volume of
currency. In the second place the fact that the rising
price-level lowers the real rate of interest actually demanded
below the rate which savers mean to demand is particularly
unfortunate at a time when the expected future productivity
a cause of the great increase in the volume of production of goods
in that year.
It should be added that the currency inflations of the early 70 's
and of the i goo's seem to have been initiated by an increase in gold
imports, while the feebler inflation of the late 80 's was of the nature
of a pure credit boom.
1 Cf. Purchasing Power of Money, chap. iv.
2 This lagging of money wages helps to accentuate the fall in
the indices of general consumption in the early stages of industrial
revival ; cf. chap. i. § 5, sub fin.
2i6 STUDY OF INDUSTRIAL FLUCTUATION.
of constructional goods is abnormally high, and when only
the exaction of a high rate could prevent an unduly rapid
absorption of resources in investment.
In Germany in particular, where more than anywhere
else the machinery of the banking system has been de-
liberately used in the development of permanent investment,
the degree of over-investment attained seems to be largely
determined by monetary conditions. Thus there is no
reason to doubt that the boom of 1871-3 and the subsequent
collapse were very much accentuated by the payment of
the 1,000 million francs war indemnity from France. Again
in 1894-5 the enormous increase of gold supplies 1 co-
operated effectively with invention in giving birth to the
over-investment which culminated in 1900. Conversely
the mildness of the boom and relapse of the 8o's must be
attributed in part to the comparative scarcity of gold.2
Gold holdings Net imports Deposits of
of Reichs Bank. of gold. joint-stock banks.
1893 . . . 25-0 1-9 77-9
1894 29-4 12-6 loo-o
1895 . 33-5 -8 II2-2
2 It is interesting to contrast France, where in spite of a steady
diminution in the Bank's holdings, the "invention" of the ore
supplies of Lorraine and the Freycinet scheme of public works
were sufficient of themselves to produce a considerable amount
of over-investment.
Cf. the following figures: —
Cash reserve of Prussian Bank.
i m.
1870 . 4.3
1871 6-0
1872 8-6
1873 n-i
(Lescure, op. cit., p. 131).
Gold holdings Deposits of Gold holdings of
of Reichs Bank. German Banks. Bank of France.
£ m. £ m. £ m.
1879 10-4 26-3 1878 .. 37-9
1880 10-7 26-4 1879 . . 28-5
1881 ... 9-8 33-9 1880 . . 27-9
1882 . . 9-9 31-3 1881 .. 23-3
1882 .. 35-2
(Layton, Introduction to Study of Prices, pp. 137, 140).
THE WAGE AND MONEY SYSTEMS. 217
§ 3. MONEY IN THE CRISIS AND DEPRESSION.
Secondly with regard to the weeks or months of actual
crisis. The exceptional advantages, whether real or ima-
ginary, enjoyed by business men cannot continue for ever.
So far as the gain from rising prices is purely illusory, the
fact is bound sooner or later to be discovered : the antici-
pated productivity will fall till it corresponds with the real
productivity of effort, and the volume of production suffer
restriction. x Nor are the real gains much more persistent ;
the money charges for the provision at least of new savings
and of some kinds of labour will be screwed up to correspond
with the real charges which those who provide them intend
to make. Moreover, the charge for the services of banks
will be increased. In so far as its fall was due to expanding
confidence, that confidence is likely to have been replaced by
distrust due to the increase of resources in reckless or in-
competent hands, and aggravated perhaps by some con-
spicuous failure. In so far as it was due to an increased gold
basis, that source of support is likely to have been removed
by the withdrawal of metallic currency into circulation to
1 If, however, the working-class realisation of error occurs after
the point at which the effort-demand for anticipated commodities
becomes inelastic, it will very possibly set up a reverse movement
towards an enlargement of production.
AA' = curve of anticipated commodity productivity of effort.
2i8 STUDY OF INDUSTRIAL FLUCTUATION.
support the higher level of prices,1 and especially to dis-
charge the higher money wages bill. 2 The result is a drastic
curtailment of credit currency, effected by a more or less
sensational rise in discount rates and sometimes by an
actual refusal of accommodation even at high rates and on
good security.
It must be carefully observed that this argument by no
means commits us to an endorsement of Professor Fisher's
view that the fundamental cause of the crisis consists in the
money rate of interest catching up with the rise in the price-
level, arid so abolishing the bounty received by business
men at the expense of their creditors. Miss England's
figures, quoted in the last section, make it plain that in this
country at all events the rate of discount has always begun
to rise considerably faster than the price level some time
before the collapse comes. It is indeed obvious, if our
reasoning in chap. i. § 5 is correct, that the payment of a high
real rate of interest is not a decisive deterrent to the business
man if he is still convinced of the high future satisfaction
productivity of investment : it needs either a positive
shortage of investable resources, or a growing disillusionment
on his own part, to stem him in his wild career. Monetary
influences, though they aggravate the severity of the crisis,
are not its essential cause.
The relative scarcity of currency makes its influence felt
aa' = curve of actual commodity productivity of effort.
BB' = curve of anticipated satisfaction productivity of effort.
bb' = curve of actual satisfaction productivity of effort.
EE' = curve of disutility of effort.
The further raising of AA' to A1A1', ex hypothesi lowers BB' to
B1B1', but (unless a is raised above A) does not lower, and may
raise, bb'. It is not unlikely then that B1B1/ will fall below bb' ;
in this event production having been restricted from OM± to OM2 will
on the realisation of error be expanded to OM3.
This consideration may, I think, conceivably be of some import-
ance, especially in the coal trade ; but it is impossible to be certain,
1 Cf. Marshall, Evidence before Gold and Silver Commission, Cd.
5512-1, 1888, Q. 9643.
2 The clearness and emphasis with which this point is presented
seems to be the most solid and original contribution made by Mr.-
Hawtrey to this subject. Cf. Good and Bad Trade, p. 75.
THE WAGE AND MONEY SYSTEMS. 219
in two ways. The earlier and less conspicuous is a slight
fall in general prices, especially the prices of (manufactured)
consumable goods. It is this preliminary monetary super-
fluity of consumable goods which has so often darkened
counsel, and distracted attention from that real shortage
of consumable goods which is the essential feature of the
situation. Thus in England in 1900, employment in the
tin-plate and engineering and metal trades did not begin
to give way till June, nor in pig-iron till July, nor in iron
and steel manufacture till October * nor in shipbuilding
till November : while in a number of consumptive trades
it had already broken in April and May. And among these
trades it seems that the relapse began at the end nearest
consumption. Thus boots declined in May, leather not till
July ; printing in April, paper in May. Thus also building
was on the whole well maintained till November, but
furnishing only till June. Similar evidence comes from the
metal and wood-working trades. For instance, in July at
Wolverhampton employment with moulders and bridge and
girder makers is busy, with cycle and motor makers quiet ;
with makers of nuts and bolts and builders' ironmongery
good, of stamped and light hollow-ware moderate, of locks,
keys and latches quiet, of spectacle frames and steel toys
slack. At Redditch it is worse with ordinary than with
machine needles ; at Coventry it is quiet with cycle, good
with tool-makers: while at Birmingham it is already moderate
with tool-makers but still good with iron-founders and
pattern-makers.2 At Nottingham in September it is good
with packing-case makers, mill-sawyers and machinists
among others, but only moderate with basket-makers,
1 The large fall in July was due mainly to holidays and hot weather.
2 Lab. Gaz., Aug. 1910, p. 245. Contrast the conditions during the
latter half of 1902, when the demand on the whole was good except
for structural work (E. H. of 1902). At Birmingham, for instance,
in September employment among ironfounders, pattern-makers,
etc., was by now only moderate, but good among makers of house-
hold articles, hollow-ware, stoves, lamps, pens, etc. (Lab. Gaz.t Oct.
1902, p. 292).
220 STUDY OF INDUSTRIAL FLUCTUATION.
brush-makers and box-makers. 1 In July, however, printing,
in August bespoke tailoring, paper, glass and leather began
to recover and continued fair on the whole till the end of
November ; but the more dysenteric trades — worsted,
boots, ready-made tailoring, hosiery — failed on the whole
to share in this recovery.
In 1907 2 again there was already by June a change in
the woollen, worsted, silk, lace, hosiery, and boot trades
from an increasing 3 to a decreasing wages bill, and employ-
ment began to decline also in furnishing, ready-made tailor-
ing, and the motor and cycle trades.4 Again the relapse
clearly began from the end nearest consumption — at Not-
tingham, for instance, lace and hosiery machine-makers
continued to prosper in June and July, while the trades to
which they ministered suffered — and again, while the
dislocation was in no case severe (the textiles were on their
feet again by August) the dysenteric trades suffered most :
thus glass was not hit till July, while paper, pottery and
hats were untouched. In America in the same year it will
be remembered that boots and hardware 5 were among the
earliest sufferers.6
1 Ibid., p. 308. It is difficult to produce any very coherent
results from the mass of evidence of local and temporary fluctua-
tions collected by the Board of Trade ; but the above instances,
though not, I confess, taken at random, seem to represent the general
trend of events.
2 It must be borne in mind that in this period the situation was
complicated by agricultural influences.
3 That of hosiery began to decrease slightly in May, and that of
lace in April.
4 " The cycle industry showed a considerable decline compared
with a month ago, and discharges of men were reported at Coventry,
Redditch and Birmingham. In the motor trade employment was
fair, but not so good as a month ago " (Labour Gazette, July 1907,
p. 204).
6 Similarly unemployment in the clothing trades in New York
State had already increased before the crisis, from 7-8 per cent,
in June to 10-0 per cent, in September. So also in Canada in 1913
there appears to have been a " contraction in the market " for
general manufactured goods, while the steel industry of the mari-
time provinces was still unaffected (Ec., Jan. 17, 1914).
6 Mr. J. M. Keynes, in a criticism of a former draft of this work,
complains that the argument " tends to confuse a* falling-off in
THE WAGE AND MONEY SYSTEMS. 221
The later and more dramatic effect of a restriction of
currency is to cripple the operations of business men by a
stoppage of the supplies of credit money on which they are
accustomed to rely for financing the current operations of
their business. The natural result may be summarised as
a general restriction of production, accompanied by a general
temporary increase in consumption, due to the unloading of
stocks. The effect of dear money varies considerably, how-
ever, between one industry and another, and seems to permit
of the following generalisations.1
First, those trades are severely hit in which the rate of
profit on the turn-over is small compared with the annual
rate of profit. This seems to be the case with certain
branches of the engineering trade, where valuable contracts
are carried through on a small commission, so that the
margin of profit is easily wiped out by a rise in the. rate of
interest.
Secondly, trades making use of a raw material in which
there is a highly organised speculative market are likely as
a rule to benefit by dear money. Thus in the months of
money-famine October, 1907, to January, 1908, woollen and
cocoa manufacturers frankly welcomed the high rates, and
agricultural buyers benefited by the loosening in rain of the
great nitrate clouds which had been banked up in Chili.
Tanners might have been expected to profit likewise, for
merchants who had been loading and salting their hides
now unloaded them upon the market ; but as a matter of
fact they seem to have held out for a still lower price,
consumptive production with a falling-off in consumption." He
seems inclined to doubt a fall even in the monetary demand price
for consumable goods, and to refer the decline in employment in
the consumptive trades to a diversion of productive energy into
constructional industry. The phenomena under discussion would
thus become a further proof of the argument of chap. ii. § 2. I am
inclined in large measure to agree ; but I think the behaviour oi
the dysenteric trades in particular lends some support to the argu-
ment in the text as an additional and not unimportant explanation
of the course of events.
1 Based mainly on two very interesting articles in the Economist
of 1907, pp. 2,022 ff. and 2,071 ff.
222 STUDY OF INDUSTRIAL FLUCTUATION.
while obliged themselves to unload their stocks of leather,
so that the boot manufacturers were the chief gainers ;
employment in that industry improved steadily, while in
the leather trades it became considerably worse. Similarly
the Bessemer converters and steel rolling mills in the Cleve-
land district 1 seem to have made a temporary gain in
December from the unloading of pig-iron stocks.2
The beneficent aperient effects of dear money are, however,
liable to be cancelled by special circumstances. Thus the
electrical industry failed to profit in 1907 from the forced
sales of copper : for in the first place the high money rates
made it so difficult to remit money for wages, etc., to the
far western mines of Utah, Arizona, and Montana, that
production was very quickly curtailed ; and in the second
place they were a severe check to the enterprise of munici-
palities, on whom the electrical industry largely depends.
Again, though jute manufacturers are not severely affected
by high rates,3 they were prevented by the sudden collapse
in the American demand from reaping the full benefit of the
large supplies " forced off at considerable sacrifice " by the
Indian exporters.4 The sugar trade failed to profit because
the unloaded stocks were to some extent held up by the
insufficiency of water in the continental rivers ; and the
meat trade, because American shippers were induced by
the poor market for hides to kill fewer cattle, being presum-
ably able to recoup themselves for the expense of carrying
their live stock in a time of dear money by squeezing the
growers.5
1 Lab. Gaz., Jan. 1908, p. n.
2 It is odd that Mr. Hawtrey (Good and Bad Trade, p. 62) neglects
altogether this side of the matter, and directs attention only to the
effect of dear money in reducing the stocks of finished goods.
3 According to a Dundee banker (Economist, loc. cit.}. The
employment figure was on the whole well maintained (Lab. Gaz.}.
4 And the manufacturers themselves seem to have been specu-
lating, believing, in the almost ^Eschylean phrase of the Dundee
Advertiser, " that prices were securely established on the higher
platforms not for a long time to be ousted " (E. H. of 1907. Cf.
Part I. ch. iii. § i).
5 The furnishing and wood-working trades would seem at first
THE WAGE AND MONEY SYSTEMS. 223
It should be observed that the stimulus thus given to
certain industries is only a temporary one. Where the
previous high prices have been due to purely speculative
action (as in 1907 with leather and cocoa) its effect may be
beneficial ; but where (as on the same occasion with sugar
and nitrate) the statistical position is genuinely unfavourable
to consumers, the frustration of legitimate efforts at inter-
temporal compensation can only lay up trouble in the
future.
It may be added in this connection that the transport
industries tend to gain from the enforced liquidations.
Thus in the autumn of 1907 there was a temporary recovery
in North Atlantic and other freights,1 and in December
employment at the London docks for the first time for seven
months showed a substantial improvement on the previous
year.
Thirdly, those trades which either make for special order
or for such a wide and speculative demand that there is a
highly organised market in the finished as well as the raw
product suffer more than those which make for a steady and
uniform demand. For the former group, since they do not
enjoy a continuous inflow of receipts, are particularly de-
pendent upon outside financial aid : thus in the money-
famine of 1907 the Labour Gazette reports from the bespoke
are much more unfavourable than those from the ready-
made clothing trades : and the motor trade also was hard
hit, the system of exacting part payment in advance having
recently been abandoned under the stress of competition.
With these trades may be contrasted those which had been
making for a steady market and were in possession of a
favourable bank balance — coal especially, and wool manu-
facture, which was " quite able to handle all the stock
sight to have had a good chance to profit, for not only was the
timber trade severely affected by dear money, but its chief customer,
the building trade, was in a similar condition. It appears, how-
ever, that the stocks of mahogany and other exclusively furniture
woods were not large, and that in any case the furnishing trade is
too closely bound to the wheels of its building colleague to take
much advantage of cheap supplies,
i Mr. John White's Circular, 1907.
224 STUDY OF INDUSTRIAL FLUCTUATION.
required without overdrawing " : glass, paper and tobacco l
were likewise unaffected.2
With regard to the second group the exigencies of mer-
chants work both ways. Thus the cotton industry is not
directly dependent on cheap money, as the continued activity
of the spinning branch in our test months shows ; but we
learn that " the houses engaged in shipping Manchester
goods to foreign countries are finding it rather difficult to
finance their transactions, . . . and doubtless they are in
many cases so arranging quotations as to prevent business
which cannot conveniently be financed going through, and
this naturally reduces the payment for yarn and cloth."
Thus employment in the weaving branch declined appre-
ciably in November. The linen trade was similarly affected :
" owing to financial stringency several American houses
engaged in the trade have cabled requesting that goods on
order for present shipment be deferred/' and employment
was curtailed in November and still more severely in
December.
Fourthly, the more luxurious trades are harder hit than
those making for common consumption. In part this may
be attributed to the fact that they often make for special
order, but in part also to other causes. Thus the managing
director of Humber Ltd., who makes the generalisation from
the experience of the motor industry, points out that the
purchase of a car is apt to be regarded as capital rather
than current expenditure, and to involve the selling of
securities — an unprofitable proceeding when money is
dear. But that the cheaper luxuries suffer as well is sug-
gested by the more rapid fall of employment in these months
in the lace than in the other clothing trades, and by the
immediate drop in November in the fancy leather trade 3
and in the china furniture branch of the pottery trade,4
1 The Imperial Tobacco Co., with bank-rate at 7 per cent., was
able to invest in 16,000,000 Ib. of Kentucky leaf without turning a
hair.
2 Thus also, while the wholesale dealers in tea were hard hit,
the large retail distributors, who collect their money day by day,
were able to profit at their expense.
3 Contrast Labour Gazette, 1907, pp. 339 and 371.
4 Ibid., pp. 341 and 374.
THE WAGE AND MONEY SYSTEMS. 225
while the clay tobacco-pipe branch of the same trade
remained unaffected. Now people do not sell securities
to buy china ornaments, and the phenomenon in question
should be attributed, I suggest, rather to the fact that high
money rates hit in general not so much the manufacturing
as the merchant class, who have no wages bill to help to
bear their losses.
And now with regard to the effect of a monetary economy
when the crisis gives way to depression. The memory of
disaster, combined with the lack of confidence which natur-
ally accompanies the real unprofitableness of further invest-
ment, prevents both business men from demanding and
bankers from supplying any considerable increase in the
volume of credit currency. This shortage of currency,
combined with the influx of the new supplies of consumable
goods, leads to a progressive fall in money prices. As the
divergence between the real and the anticipated productivity
of effort operated during the boom to stimulate production,
so now it operates to restrain it. Moreover, just as business
men were then obtaining a bounty at the expense of savers
and wage-earners, so now the stability of money rates of
interest and wages forces them to pay over a bounty to
these classes ; and to the fictitious inducement to restrain
production a real inducement is added.
The reality of the transfer to wage-earners, until either
the rate of wages is reduced, or production and the oppor-
tunities for employment are restricted, is indicated by the
comparative course on such occasions of the more " neces-
sary " and the more " luxurious " branches of certain trades.
Thus in 1892, while at Leeds (the centre of the ready-made
clothing industry) " there has been a fairly steady volume
of trade," at Huddersfield the failure of many firms making
fancy woollen and worsted goods was recorded. l Similarly
1 In 1901, however, the pressure of war taxation and in 1908
the pressure of dear wheat obscured the issue in the woollen trades ;
so that in 1901 " manufacturers of the superior class of fabrics
have fared much better than their fellows, devoted to the produc-
tion of medium and low-class goods," and in 1908 " the class of
trade a little below the best " (Ec., 1909, i. 661) has suffered less
than the cheaper varieties.
0
226 STUDY OF INDUSTRIAL FLUCTUATION.
in 1900 the " luxurious " printing trade was among the
first to be affected, and ready-made survived bespoke
tailoring by three months. Again in 1908 the luxury trades
seem to have been on the whole the heaviest sufferers —
silk and leather throughout the year, lace till July, motors
and cycles till October. Moreover, it should be remembered
that most of the evidences of the survival of general con-
sumption after the collapse of the " boom " are evidences
in particular of the survival of working-class consumption.1
When, as seems to have been the case in the 8o's, business
men are able immediately 2 to meet an obstinate wage-rate
by a thorough-going extension of the use of machinery,
the inducement to curtail production is limited, and in
spite of increased unemployment and a diminished aggregate
money wages bill, the manual workers, in common with
other classes, may enjoy a continuous increase in con-
sumption.3 Otherwise business men are apt to bear the
loss for an appreciable period, but eventually to make
reprisals by a restriction of production, which reacts upon
the workers both by curtailing employment and by limiting
the fall in prices. 4 It should be observed, however, that the
1 The maintenance of working-class incomes tends to mitigate
the special disabilities of the " dysenteric " trades ; for though
some dysenteric trades are of course " luxurious " (cf. Dr. Marshall's
celebrated aneroid barometers), on the whole the correlation is
surely the other way.
2 For the important effects of this and similar devices at a later
stage, cf. p. 126.
3 Thus while money wages were actually higher in 1885 than
1882 (149, according to Mr. Rowley's index [1850 = 100] as against
147) the aggregate wages bill (Bowley, EC. Jour., 1904, p. 459)
had fallen from ^470 m. to ^440 m. or 6-4 per cent. ; but produc-
tion and consumption continued to increase. The aggregate of
incomes subject to tax fell from ^590 m. to ^580 m. or only 1-7 per
cent. — a curious commentary on Dr. Marshall's opinion (Evidence
before G. and S. Commission) that there was a depression of profits
to the advantage of wage-earners.
4 Cf. the following figures : —
Wage Index. Wages Bill. Taxable Incomes.
£'m, £ m.
1890 . 163 550 640
1891 ... 163 555 635
1892 . 162 545 625
1893 , . ,162 545 630
THE WAGE AND MONEY SYSTEMS. 227
final and most acute stage of depression tends to occur
after a considerable readjustment of wage-rates has taken
place (e.g. in 1878-9, 1886, and 1904 *) : indeed the impulse
given to production by the removal of the tax upon business
men actually enhances the purely monetary and illusory
inducements to restriction.
Such are the aggravations of the phenomena of depression
for which a monetary system must be held answerable. On
the other hand it is only fair to point out that it facilitates
that quest of mutual markets which is the best hope of the
consumptive trades at a time of constructional collapse.
Whatever the ideal scale of production for any given industry
at any time may be, the sheer dislocation caused by the
failure of an important customer, country or industry is
certain, until new adjustments can be made, to restrict the
actual volume of production and still more of sale below
the ideal level.2 But with the assistance of a money
1 Among the results of the similar revision of wage-rates in March
1908 (for which see Lab. Gaz., Jan., 1909, p. 3), were an immediate
fall in employment in the glass bottle branch, which had hitherto
supported the employment figure for the whole industry, and a
more gradual fall in wool and reversal of the relative movements of
ready-made and bespoke tailoring.
2 Thus even the woollen trade, which we have taken as the clas-
sical example of the survival of consumption, seems to have been
nonplussed in 1875 by the depression of its iron and steel neigh-
bour. Again, while neither the initial relapse of the consumptive
trades in 1900 nor their continued slackness in 1901 can be ascribed
mainly to the failure of the demand of constructional industry
(shipbuilding employment in 1901 did not seriously fall off till
September, tin-plates improved rapidly from March, and in pig-
iron there was a decided temporary improvement from May to
September, when the dysenteric trades were at their worst) ; yet
it is perhaps significant that the checks, in February-May and
again in October, to the decline in the activity of iron and steel
manufacture coincided closely with the attempts at revival in the
dysenteric trades, and its renewed decline in June with their collapse
(even the subordinate minimum in April was accompanied by a
relapse in worsted). (That at this time the causation was in this
direction is suggested by the fact that the intermediate engineer-
ing trades do not share the movement.) Again in the first half of
1908 the trades most dependent on the demands of America and
of constructional industry — worsted, linen and lace — were among
the severest sufferers, but recovered, partly owing to the discovery
of new markets, in the latter half of the year.
228 STUDY OF INDUSTRIAL FLUCTUATION.
mechanism it is very much easier to develop new markets
than it would be under a system of direct exchange : though
indeed it would be easier still if the business man could
overcome his distrust of falling prices.
§ 4. GOLD,— MEDICINE, POISON, AND INTOXICANT.
In the preceding sections we have been assuming, in
common with most writers on this topic, that an influx of
new gold, to whatever extent it may be responsible for the
boom, in fact tends to synchronise with that and with no
other phase of the industrial cycle. Neither general reason-
ing, however, nor the evidence of facts supports this view.
It seems that such an influx may occur at any time, and
that its effects may cut across and complicate as well as
merely intensify those of purely industrial phenomena.
Dr. Marshall's famous generalisation, that " if there is an
extra supply of bullion, bankers and others are able to
offer easy terms to people in business, including the bill-
brokers, and consequently there is more money on loan, and
consequently people enter into the market as buyers of
things, as starting new businesses, new factories, new rail-
ways, and so on,"1 seems to be only a partial presentation
of the truth.
What, for instance, are we to make of the following figures
for 1894-6 ?
Gold Holdings of Bank.
Net Imports of Gold.
1893 .
. £26,425,000
£4,670,000
1894 .
34,309,000
11,924,000
1895 .
38,951,000
14,636,000
1896 .
• 44334.000
• -5^37,000
1897 .
• 35.571.000
Zero
In America the dependence of the clothing trades upon construc-
tional industry is still more intimate (partly through the medium
of the ebb and flow of immigration) ; hence their continued depres-
sion, after the readjustment of the balance of trade in 1910, until
the metal boom of 1912 was well on its way.
N.B. also the dependence of the cotton trade on German con-
structional industry, cf. p. 163, n. i.
1 C. 5512-1, 1888, Q. 9677.
THE WAGE AND MONEY SYSTEMS. 229
Yet 1894-6 were years of falling prices, constructional
depression (except for the brief reflex from the U.S.A. in
1895), and absence of speculation. Paradoxically enough,
it was not till some of our large gold supplies had been bled
off, largely to America,1 that prices began to recover. The
solution seems to be that the effect of new gold in the closing
years of depression is purely sedative and medicinal. It
is possible that the mere existence of large gold reserves
and a low rate of discount leads people to think that prices
are about to recover, and so to be less afraid first of buying
other people's goods, secondly of consenting to immediate
actual reductions in the price of their own, which they believe
will only have to be temporary, and thirdly of making for
stock. Thus the steady fall in the unemployment figures
in 1894-6 and the considerable, if quiet, volume of business
done must be ascribed in part I think to the soothing in-
fluence of the new gold supplies. When the revival of the
volume of trade is well on its way, the increase of confidence
leads to an application to the banks for accommodation,
the accumulations of gold are partly used as a basis for
increased credit and partly withdrawn to meet a rising wages
bill, and prices rise. < At the risk of incurring suspicion of
a gross and elementary fallacy, it must be boldly stated
that in such a case the actual rise in prices is the result rather
than the cause of the increase in the volume of trade. In
any event, even when the influence upon prices is not so
long delayed as it was in the 90% its effect at such a time is
not necessarily to precipitate a constructional boom, but,
as a Cambridge lecturer has phrased it, " to cheat " producers
in general " into being sensible."
On the other hand an influx of gold at a time when the
real trouble is still a pronounced shortage of consumable
goods will only serve to postpone the rebirth of genuine
1 U.S.A. net imports of gold.
1894 — ^16,126,000
1895 — 14,114,000
1896 + 9,295,000
1897 ~~ 51,000
1898 + 28,394,000
230 STUDY OF INDUSTRIAL FLUCTUATION.
prosperity. For its effect as a stimulus to production is
outweighed by the fact that it is obtained at the cost of
other and more desirable goods. Thus America's vast
gold imports after the 1907 crisis, though necessary for the
re-establishment of her financial system, involved a vast
compensatory export of agricultural and other produce
that could ill be spared, and thus accentuated the shortage
of real capital and the inflation of prices which was rightly
described as the really disquieting feature of the situation.1
America in 1908-10 was in the position of a man who has
asked for a cocktail before his dinner, and then finds there
is no dinner to follow, but only a succession of cocktails.
Similarly the reopening of the Johannesburg mines after
the war, and the consequent continued large inflow of gold
into this country in 1901-2, aggravated rather than mitigated
the difficulties of the situation ; and it was an advantage to
England and not a disadvantage that in 1903-4 she was
able to take a smaller part of her income in the form of
gold.2
On other occasions the effect of new gold may be partly
medicinal and partly poisonous. Thus the steady resump-
tion of general activity on a moderate scale in this country
in 1909-10 was, I think, partly due to the fact that the
price adjustments necessitated by the 1907 crash were
rendered less severe by the new gold than they would other-
wise have been. But on this occasion, unlike the go's,
there had been no enormous growth in the volume of pro-
duction, and the energies of the new gold could not thus
be so readily absorbed. The inevitable result was specula-
tion, but speculation not this time in constructional enter-
prise, but in raw material. All the evidence shows that
1 Cf. p. 173.
2 Cf. the following figures : —
World's Gold British Net Gold Holdings of
Production. Imports of Gold. Bank of England.
1900 . -377 tons £?>793 thousand ^33-3 million
J901 • • -390 „ 6,750 „ 35-8
1902 • • 425 » 6,219 ,, 35*6
J903 • . 4Sl ., 890 ,, 34-4
J9°4 • • 500 „ 837 „ 34-4
THE WAGE AND MONEY SYSTEMS. 231
the main rises of price in 1909 were not due to a very large
improvement in the genuine industrial demand but to
speculative activity in the earlier stages. In spite of large
supplies the index of British wool rose from 124-8 to 141-2
and that of foreign wool from 98-6 to 100-6 : the manu-
facturing industry indeed, according to the Economist's
reports, was good, but " the raw material bossed the show."
Similarly in the linen trade " everything points to a sharp
advance, owing to an increase in the price of flax."1 The
rise in the price of pig-iron in the latter half of 1909 was
not justified either by home 2 or foreign 3 demand, but was
due mainly to speculation in warrants,4 on the strength of
the American demand. But the largest rise was in rubber,
the average import price of which rose from £14-56 per cwt.
in 1908 to £20-20 in 1909 and £29-76 in 1910. 5
But it was not only to speculation in produce that the
new gold was diverted. The rubber boom on the Stock
Exchange is still fresh in the memory. It is noteworthy
that in 1909, while the amounts cleared at eight provincial
bankers' clearing houses only increased from £651-5 6
million to £673-6 million or 3-4 per cent., those cleared at
the London Bankers' Clearing house, which are much more
subject to Stock Exchange influences, increased from
£12,120 million to £13,525 million or n-6 per cent., while
the amounts cleared at the London house on Stock Exchange
account days increased from £1,672 million to £2,129 million
or 27-3 per cent.
1 EC. H. of 1909.
z It was indeed a source of serious inconvenience to shipbuilding ;
cf. EC., 1910, i. 65.
3 Our total exports of pig-iron fell from 1,296,521 in 1908 to
1,140,695 in 1909 ; though, our exports of iron and steel goods to
America rose from 149,792 to 242,555 tons.
4 EC., 1910, i. 381.
5 The net result was a considerable alteration of the ratio of
exchange to our disadvantage ; " if the prices of 1908 had prevailed, "
calculates the Economist (1910, i. 370), " we should have paid
nearly £6 m. less for our imports, and got £15 m. more for our
exports."
• The figures are from the Stat. Abs. U.K.
232 STUDY OF INDUSTRIAL FLUCTUATION.
When the absence of a genuine investor's demand became
plain, speculative activity turned away from pig-iron, whose
price remained stationary January-April, 1910, and then
fell. But other commodities were ready to take up the
running. In spite of the slackness in building, timber
began to advance in February, and the effect upon prices
of the shortages in the 1909 crops of American cotton and
European sugar was intensified in 1910 by large bull opera-
tions. In the latter part of 1910 the same thing happened
with flax and jute, after crop shortages in Russia and India
respectively ; and the same conditions prevailed throughout
the year in all kinds of vegetable oils. By 1911 rubber was
normal again, but its place had been taken by coffee and tin.
After 1911 speculative dealings in raw produce fell on the
whole into the background before the genuine advance of
consumptive activity and constructional investment. But
it is clear that during the preceding years gold had been a
two-edged weapon. On the one hand it had put artificial
hindrances in the way of manufacturing activity : on the
other hand it must, on my view, be admitted to have facili-
tated the adjustment of prices, the re-birth of confidence,
and so the large volume of exchange in which the more
harmful side of its activity was ultimately swamped. No-
where is the double action better seen than in the leather-using
industries. The boot trade was, as we have seen, among
the most cheerful in 1908-9 : yet in 1909 " boot and shoe
makers have not been able to obtain the full advances which
the price of leather would justify "l : and again in the spring
of 1911, though the boot trade was enjoying a considerable
measure of prosperity, tanners were still being severely
squeezed by the high prices of hides consequent on American
speculation.2
It is interesting to compare the course of speculation in
these years with Professor Cairnes' generalisations as to
the probable effects of the great gold discoveries of the late
forties.3 He concluded (a) that the countries of the world
1 Economist History of 1910. 2 Economist, 1911, p. 210.
3 Essays on Political Economy, cf. esp. p. 65.
THE WAGE AND MONEY SYSTEMS. 233
would be affected in the following order : (i) the producers,
— California and Australia ; (2) the countries with an
elaborate credit system and a high productivity, — England,
the United States, and to a less degree the Continent ;
(3) the irresponsive and absorbent East, (b) The different
classes of commodities would be affected as follows : (i)
those would rise first which were consumed by the " pro-
ductive " classes, especially the working-classes ; (2) while
manufactures might rise rapidly at first, they would soon,
owing to their tendency to dysentery, fall below the general
level ; (3) among agricultural products, the tendency to
surfer from dysentery and a fall would be more conspicuous
among things vegetable than among things animal ; (4)
the last things reached would be those consumed by the
non-productive classes, through a limitation of supply
caused by diversion of capital and labour to other trades.
These expectations seem to have been in the main con-
firmed in 1848-56. x The initial rise, for instance, was
greatest in timber, which Australia had produced for herself,
but now imported in exchange for gold, and in animal
products, — meat and wool, which she had been in the
habit of exporting, but now neglected for the gold-diggings.
But in modern times the conditions under which the gold-
mining industry is carried on and the growth of speculation
and of a credit economy have introduced important modifica-
tions, (a) It is now the commodities purchased by the
gold-receiving and not the gold-producing countries which
are most rapidly affected ; and owing to the international-
isation of currency and the diminishing urgency of the
demand of tropical countries for manufactures, their pro-
ducts are more quickly and fiercely sucked into the rising
stream, — as not only the rubber boom but the generally
more rapid rise of recent years in the Indian than the
European price-level bears witness, (b) I and 4. The
direct effect of gold on retail prices is no longer important :
commodities consumed by the working-classes rise not in
1 See Jevons, Investigations in Currency and Finance, passim ;
and cf . Kemmerer, Money Credit and Prices, pp. 58 ff.
234 STUDY OF INDUSTRIAL FLUCTUATION.
response to an increased demand from those classes but
owing to speculation further back : and since the new
currency passes first into the hands of the profit-receiving
and speculative classes, the commodities which they con-
sume rise faster than those consumed by the working-classes;
(2) seems substantially true and very important, — more
true than it was in 1848-56, when British manufactures
had a virtual monopoly, and when the rise in their price
seems to have dictated that of the raw material.1 But
(3) the growth of speculation and of structural enterprise
has made both mineral and vegetable products more sensitive
to the touch of gold as compared with animal, so that the
order for combined celerity and permanence of effect is
now rather vegetable-mineral-animal than animal- vegetable-
mineral.
The differences in the operation of new gold in the three
chief industrial countries are also worthy of remark. It
seems to be roughly true that in Germany the two main
alternative uses are speculation in foreign securities and the
development of home, especially home constructional,2
industry : the reasons are to be sought in the strictness of
the German laws concerning produce speculation and the
close connection of the German banks with industry. In
America all kinds of speculation are open : but in so far as
such speculation puts money into the pockets of the growers
of food and other raw produce, we have seen that it is likely
to react pretty quickly upon constructional industry, whence
the meteoric recovery of the iron industries in 1909. In
England speculation based on the new gold catches hold of
and intensifies any genuine movement of demand and
supply which it finds to hand, — the temporary American
inquiry for pig-iron in 1909, the shortage of various raw
products in the following year : but it has no necessary
tendency to produce, either directly as in Germany or
indirectly as in America, an attempt at an all-round con-
1 Cf. Jevons, op. cit., p. 52.
2 It is especially instructive to contrast the course of events in
England and Germany in the middle nineties ; cf. Part I. chap. i. § 2.
THE WAGE AND MONEY SYSTEMS. 235
structional revival. Hence England is the country with
regard to which it is most necessary to adopt a sceptical
attitude towards Dr. Marshall's generalisation quoted at
the beginning of this section.
§ 5. THE THEORY OF UNDER-CONSUMPTION.
It has not seemed advisable to interrupt the argument of
this chapter at any point in order to make incidental com-
ment upon the famous theory of " working-class under-
consumption " invoked by most socialistic writers to aid
in the explanation of industrial depression. It is therefore
desirable to give some consideration to it here in connection
with the whole trend of the preceding argument.
The theory, as rehabilitated by Mr. J. A. Hobson with
so much skill in his Industrial System, may be briefly stated
as follows. The receivers of industrial " surpluses/' — rent,
profits, etc., — are enabled to satisfy with part of their income
every conceivable want, and can think of nothing to do
with the balance except save it. The consequence of this
extensive automatic investment is a glut of consumable
goods, which is a real glut, because those who have the
power to consume the goods have not the desire and those
who have the desire have not the power.
It may be observed first, in the light of our ch. i. § 5, that
in attempting to make good this position Mr. Hobson meets
with one difficulty which appears to be of his own creation.
In anticipation of the obvious criticism that since investment
in constructional goods does not continue on an extensive
scale during the depression, the latter cannot fairly be said
to be a period of over-saving, he is at much pains to show
that investment by the wealthy classes continues in various
forms, especially in the form of the purchase of house pro-
perty and land,1 throughout the depression. Mr. Hobson
is here undertaking — and in my judgment with only qualified
success — a work of supererogation which is forced upon
him by his confusion of two quite distinct processes,- —
1 Op. cit., p. 292.
236 STUDY OF INDUSTRIAL FLUCTUATION.
saving and the use of savings in investment. The mere
fact that during the depression large quantities of con-
sumable goods are held up in store instead of being circulated
and consumed is sufficient to prove that the volume of
saving, in the proper sense of the word, is very considerable,
and to suggest that it is greater than the interests of the
community as a whole would dictate.
What then, according to this theory, is the cause of this
persistent accumulation of consumable goods ? Mr. Hobson
appears to hold that it is strictly and literally inevitable.
r< The machinery of industry must become congested and
clogged by excessive goods unable to find an exit in con-
sumption."1 To the argument that a general acquiescence
in a lower price level would result in all the goods being
taken off the market, he replies that " the second check,
the effect of falling prices due to increased production in
stimulating spending and so checking saving, has no true
efficiency. For the first direct effect of falling prices is a
corresponding fall of money incomes, which as we have seen
are derived from retail prices by distribution at the various
stages of production : and if money incomes fall pan passu
with prices, there is no reason to expect that a fall in prices
will stimulate demand for commodities. . . . The play of
elasticity of demand is here inhibited by falling incomes."2
This argument appears to contain a double fallacy. In the
first place while we have seen reason to suppose that there
is in fact as a rule at such a time a decrease in the aggregate
of money incomes, it is a pure confusion to say that this
diminution is caused by falling prices. The aggregate
volume of money incomes depends on circumstances affecting
the supply of currency, which Mr. Hobson makes no attempt
to take into account : if the falling price level results
simply from an increased volume of goods, it is theoretically
quite possible for the aggregate of money incomes to remain
undiminished. In assuming that a fall in prices necessarily
diminishes the aggregate of money received, or in other
1 The Industrial System, p. 52.
2 The Industrial System, p. 286.
THE WAGE AND MONEY SYSTEMS. 237
words expended, Mr. Hobson is assuming as axiomatic the
very point — the inelasticity of money demand — which he
requires to prove. Secondly, even if the aggregate of money
received is diminished, there is no reason to suppose that
the aggregate of consumption will not be increased ; all
that is required is that the percentage reduction in the
price level should not be less than the percentage reduction
in the aggregate of money incomes.
But Mr. Hobson has also occupied a more plausible posi-
tion. If the desire of " surplus-receivers " for consumable
goods is really capable of saturation, it will simply not be
worth their while to take the trouble to unload their stocks.
From sheer lack of incentive to act they will enforce ab-
stinence upon the working-classes. The strength of this
argument lies in its recognition of the comparative rapidity
with which the utility of consumable goods diminishes to
the rich, and of the fact that a period of depression is to some
at least a period of great potential prosperity. Its fatal
weakness lies in its failure to realise that there is no reason
to suppose the desire of the rich for construction goods to be
satiable : and that the complaints (however liberally dis-
counted) of business men, together with the stability of
money wages and the " luxurious " character of the trades
which suffer most during at any rate the earlier stage of a
depression, compel us to conclude that even the consumptive
wants of the rich are not completely satiated. In this case,
whatever may be said of a policy of restriction of output,
a policy of mere accumulation is clearly against the imme-
diate interests of business men in general.
It may, however, still be maintained that accumulation
is in the real interests of business men because they are
right in expecting an ultimate rise in the marginal utility
of the accumulated goods.1 This is indeed one aspect of
1 Mr. Hobson (though making use of course of a different phrase-
ology) seems to expect such a rise from the mere growth of popula-
tion (op. cit., p. 306). This would of course involve a benefit to the
"haves" at the expense of the "have-nots," which would be
exhibited in a fall of real wages. But it seems that in a depression
the growth of population is as a rule accompanied by at least an
238 STUDY OF INDUSTRIAL FLUCTUATION.
Mr. Hobson's central entrenchment — the position that the
unequal distribution of wealth produces a chronic tendency
to what is from a social point of view an uneconomic dis-
tribution of income between immediate consumption and
storage or investment. Now if all this proposition means
is that the intensity of the immediate wants denied by those
to whom the greater part of our saving is in fact due is less
than the contemporary intensity of the immediate wants
of less fortunate classes, the point is so generally admitted
as to need no further discussion. If again it means that a
persistent tendency (exhibited in alternate periods of
accumulation and investment) among western progressive
nations to over-estimate the intensity of future as compared
with present wants is aggravated by the existing system
of ownership and distribution, the whole course of our
argument again leads us to acquiesce. But we must dis-
sent from Mr. Hobson in his implication that the successive
generations of workers reap no advantage from this tendency,
and might not, but for monetary maladjustments, reap a
still greater advantage ; and in his assumption that a more
equal distribution would bring no perils in another direction.
If Mr. Hobson and socialistic writers in general could bring
convincing proof that the abolition of capitalism as a
juridical system would bring no tendency to a chronic under-
estimation of the intensity of future wants,1 their general
position would be immeasurably sounder, though the
arguments by which they support it would remain in large
measure fallacious.
equal growth in productivity : and that such a decline in real wages
as occurs at the beginning of a revival is amply explained by monetary
considerations. The rise in the marginal utility of goods (so far
as it is not purely monetary and imaginary) must be ascribed rather
to a rise in actual or estimated future productivity (cf. ch. i. § 5).
1 Cf. Beveridge, Unemployment, ch. ii..
CHAPTER IV.
CONCLUSION.
§ i. RECAPITULATION.
IT remains to conclude with a brief recapitulation of the
theory of fluctuations presented in the preceding pages,
and with some suggestions as to practical remedies.
The former can best be presented in the form of a con-
tinuous narrative of the successive phases of a " typical "
industrial cycle. It is scarcely necessary to point out that
the study of any particular actual cycle must not be ex-
pected to reveal the phenomena enumerated in either the
same simplicity or the same completeness : or that for the
sake of clearness certain qualifications and complications
already discussed at length have been completely omitted.
Starting then at the nadir of depression, we find that
the aggregate of industrial production begins to increase
beneath the following influences : — (i) A general increase
in the physical productivity of effort due to the adoption of
improved methods, etc., under the stimulus of depression ;
(2) an increase, due to an increased bounty of nature, in the
exchange value of industrial products against the products
of agriculture ; and (3) an increase in the expected future
productivity of constructional goods, due either (i.) to the
wearing out of an exceptionally large number of existing
instruments, (ii.) to the discovery of the industrial possi-
bilities of a new country, or (iii.) to some physical or legal
invention. Production is still further increased by (4) the
expansion, whether owing to an increase of confidence or to
increased supplies of gold, of the volume of credit currency>
which (i.) induces each producer to expect a rise in the
239
240 STUDY OF INDUSTRIAL FLUCTUATION.
exchange value of his own product, (ii.) cheapens the reward
which bankers intend to demand for their services, (iii.)
owing to the tendency of savers of capital, wage-earners
and bankers to demand a reward fixed in terms of money,
affords an additional bonus to those in whose hands the
determination of the volume of production mainly lies.
Finally, the volume of consumption is increased even more
than that of production, for some of the above influences,
in addition to stimulating production, offer an inducement
to the absorption of accumulated stocks into the vortex
of exchange.
But in the course of time (i) the physical productivity
of effort declines owing to the relapse into wasteful methods
of production and to the operation of the law of increasing
cost ; (2) agricultural shortage turns the ratio of exchange
against industrial products ; and (4) the monetary stimulus
to increased production, so far as it is illusory, disappears
with the realisation by each producer that the rise in prices
is not confined to his own product, — so far as it is real, is
reversed by waning confidence, by a depletion of gold
reserves, and by the revision, in accordance with the
rising price level, of the claims of savers, wage-earners and
bankers.
The reversal of (3), as being less generally understood,
demands somewhat fuller recapitulation. Since each new
investment, once it is made, will be capable of functioning
for a considerable period, the rise in the utility of new con-
struction goods will often in any case only be temporary ;
a point will be reached beyond which any further investment
would involve a sacrifice of present enjoyment dispropor-
tionate to the enjoyment which will be afforded by the new
consumable goods which it is proposed to create. In fact,
however, owing to the stress of competition, aggravated by
the length of time which must elapse before the new instru-
ments projected can be brought into working order, invest-
ment is likely to be carried beyond this point. Indeed, it
is possible that the absorption of resources in investment
may be so rapid that owing to the exhaustion of accumu-
CONCLUSION. 241
lated stocks it will for a time be physically impossible to
maintain investment on the scale on which it has been begun.
If the " period of gestation " is prolonged, or if the absorp-
tion of resources is aggravated by war or other exceptional
causes,' or if the resources are absorbed wholly or mainly
in investment in a group of industries producing one or two
particular types of commodity such as transport facilities,
this stage may be considerably prolonged.
In any case, whether or no there is actual over-investment,
and whether or no there is actual depletion of stocks, a decline
in the demand for constructional goods will occur. As to
the consumptive trades, they will, as the period of gestation
closes, be inconvenienced by the difficulty of disposing of
their increased output at existing prices, — a difficulty
enhanced by restriction in the supply of currency, and by
the failure of the accustomed demand from those engaged
in the constructional trades : but the volume of their pro-
duction and mutual consumption will on the whole for a time
be well maintained.
But (i.) the diminished productivity of their effort (in
terms of satisfaction) due to the decline in the urgency of
their desire for constructional goods, will constitute an in-
ducement to the leaders of consumptive industry to restrict
the volume of production. This inducement will be strength-
ened by (ii.) the real disability imposed on business men by
the relative fixity of money wages, etc., and (iii.) the illusory
disability imposed by a falling level of general prices, which
will also lead to an accumulation in store of a considerable
part of what is produced. In the course of time disability
(ii.) wil] be wholly or partially removed, but the others will
remain, and the features of depression continue to prevail
until the forces enumerated at the beginning of this summary
come into play.
§ 2. REMEDIES— THE BOOM.
When fairly faced, the problem of the prevention of in-
dustrial fluctuation becomes nothing less formidable than
the problem of maximising the community's aggregate of
R
242 STUDY OF INDUSTRIAL FLUCTUATION.
net satisfaction through time, — in other words of attaining
the best distribution through time of its income of consum-
able goods which is practicable without undesirable restric-
tion of the total of that income.1
In the light of this definition we may lay down the follow-
ing propositions. The desired aim is likely to be furthered
first (subject to certain modifications)2 by anything which
increases the tendency to inter-local and inter-temporal
compensation in agriculture ; secondly, by anything which,
without sacrifice of efficiency, reduces the necessity for dis-
continuity in the process of investment ; thirdly, by any-
thing which diminishes the tendency to miscalculation
either during the " boom " or the " depression " ; and
fourthly, by anything which mitigates those incidental
effects of over-investment during the boom which prevent it
from being followed by a very large volume of consumption
during the " depression."
With regard to the first point, certain developments and
proposals tending in the required direction, have already
been incidentally discussed, and no more need here be said.
With regard to the second, something may be expected from
such devolution and decentralisation of industry as is likely
to result from the increased use of electrical power : but
in the main this particular source of malady seems to be
one for palliation rather than for prevention.
As regards the third point, the possible remedies may be
summed up under four headings, — improvements in indus-
trial structure, improvements in public knowledge, im-
provements in banking policy, and improvements in mone-
tary mechanism. I propose to discuss these in turn.
(i) The excesses of investment during the boom are ad-
mittedly due in no small measure to the prevalence of com-
petition and the ignorance on the part of each individual
producer of the scale of the preparations which have been
1 A restriction of this total may be desirable if it is more than
compensated for either by an improved distribution through time,
or by the saving of effort : cf. p. 200.
2 Cf. pp. 50 and 153, n. i.
CONCLUSION. 243
and are being made by his competitors. Combination, by
pooling information and prospective markets and so facili-
tating a common investment policy, may be expected
materially to reduce the temptations to over-investment.1
But before acquiescing in the claims of monopoly,2 we
must observe that they are sometimes presented in a different
form which may well give us pause. Combination is often
recommended upon the ground that it will enable prices to
be kept comparatively steady. Those who are disposed to
deny its usefulness in this respect generally base their
opposition either upon the ground that in fact the alterna-
tions of unquestioned monopoly power and of the apprehen-
sion of new competition lead to a considerable unsteadiness
of price policy ; 3 or else upon the ground that steadiness of
prices is dearly purchased by the community if it means
steadiness at a high level.4 It is less common to find an
appreciation of the fact that from the present point of view
it is by no means clear that steadiness of prices is desirable.
1 " Whatever increases the size of business units," says Pro-
fessor Jones (Economic Crises, p. 51), "by reducing the number of
independent unknown elements in the business situation simplifies
the problem " ; and he quotes with approval, though with a warn-
ing that the remedy may be worse than the disease, the Hon. Carroll
Wright's dictum that "if the employers in any industry would
combine under an organisation that should have positive coher-
ence, there would be no difficulty, so far as that industry is con-
cerned, in regulating the volume of production in accordance with
the demand."
2 One prima facie objection to these claims seems to be unsound.
It is true that in fact monopoly is at present most prevalent in those
industries in which the unit of investment is large and indivisible,
and the likelihood of over-investment therefore great. But mono-
poly control is not responsible for the large size of the unit ; the
character of the business would lead in any case to conditions of
large-scale or " monopolistic " competition. Nor would the pass-
ing of small-scale industry under a large unit of control necessarily
involve a change to a larger unit of productive power ; so that
while the extension of monopoly would not diminish it would not
necessarily increase the temptation to over-investment arising
under this head.
3 Cf. Jenks' account of the course of sugar prices in America since
the formation of the Trust (The Trust Problem, pp. 133 fif.).
4 Cf . Macgregor, Industrial Combination, Part III. chap. i.
244 STUDY OF INDUSTRIAL FLUCTUATION.
And first with regard to the raising of prices during the
boom period. The policy recommended to the German car-
tels by their theoretical patrons and apparently carried out
fairly conscientiously in 1895-9 was to " moderate the rise
and maintain reasonable prices with a view to not discour-
aging consumption."1 Thus in each of the years in ques-
tion the price of pig-iron in the Ruhr district was kept
appreciably lower than was necessary to hold the market
against English importations. Coal prices in Germany
rose only some 50 per cent., in England some 80 per cent,
between 1895 and 1900. A similar policy was pursued in
1901 by the United States Steel Corporation, which in May
of that year publicly announced that it would make no
further advance in steel prices.2 This action naturally
wins the admiration of the " costs of production " theorists.
" Plein de moderation," says M. Lescure, " de prevoyance
et d' esprit de suite, il a substitue a 1'ancienne politique
cristallisee dans la formula to make hay while the sun shiness
(sic) la politique du debouche stable par un prix stable."3
But that surely is to beg the whole question at issue. Can
a stable market be secured by means of a stable price ? If
Mr. Hull is right in holding that the maximum rate of con-
struction during a boom is one which could and should be
permanently maintained, perhaps it could. But if we are
right in holding that in a boom too great a proportion of
the nation's resources is being devoted to constructional
uses, it follows that the raising of prices by a constructional
combination is not only excusable but is its bounden duty
if it is to pose as the guardian of the true interests of society.4
1 Lescure, op. cit,, p. 559. 2 Hull, op. cit., p. 117.
3 Lescure, op. cit., p. 569.
4 There seems indeed to be some disagreement among the doctors
as to the actual nature and effects of combination policy in America.
M. Lescure lays emphasis on its moderation in 1897-1903 and on
the mildness of the subsequent depression, and ignores altogether
the depression of 1900-1. Mr. Hull, on the other hand, lays stress
on the " spectacular " advance in the price of iron in 1899, on the
failure of the Steel Trust's self-denying ordinance in 1901, and
on the great severity of the depression which followed in each case.
It would obviously be just as logical to connect the occurrence of
both depressions with the " moderation " in price policy, and
CONCLUSION. 245
A new stage of the problem arises when by some means
over-investment in the sight of God has been converted into
over-investment in the sight of man and the spirit of enter-
prise is dead or dying. In this stage it is urged upon the
combinations that " from death to life they might him yet
recover " by a policy of vigorous price-reduction ; and the
" cost " theorists are loud in their complaints against the
Steel Trust, which after the 1907 crisis refused to submit
to any reduction of prices till February 19, 1909, and against
the Raw Iron and Coal Cartels, which by the device of two
or more year contracts maintained prices throughout 1900-1.
Yet even in this stage it seems quite likely that the snake
of investment-mania has been scotched, not killed : and
that the head if not the heart of the Steel Trust deserves
more credit for attempting to prevent the American people
from wasting their scanty stores of real capital on further
construction in 1908 than for its self-denying ordinance in
1901, or for the lowering of prices in 1909 which prepared
the ground for the meretricious and disastrous little iron-
spurt of that year.1
their comparative mildness with the failure to make this moderate
policy effective. On the whole, Mr. Hull's account of the course
of prices seems to be decidedly the more correct. Cf . the following
figures (Stat. Abs. U.S.A., 1912, p. 550) : —
No. i Foundry Bar Iron Steel Billets
(Philadelphia). (Pittsburg). (Pittsburg).
1898 . . . n-66 .. 23-93 .. 15-31
1899 . . . 19-36 .. 43-75 .. 31-12
1900 . . . 19-98 . . 48-12 . . 25-08
1901 . . . 15-87 .. 40-38 .. 24-13
1902 . . 22-19 •- 43-53 •• 30-57
But even if M. Lescure's account were correct in this respect, we
have seen reason to take exception to his refusal to recognise the
reaction of 1900-1.
1 Competent German observers support this view with regard
to that country in 1900-1, only that in this case it was in the breasts
of the iron producers and not of the " purse-string holders " that
the snake lingered. " It was objected that the situation in the
iron industry was one of over-production and reckless competition,
and therefore any concession made in the price of coal would be
immediately followed by further underbidding in the sale of iron
Herr Kirdorf says that if the prices [of coal] had been lowered
considerably, the crisis would have been much severer. Herr
246 STUDY OF INDUSTRIAL FLUCTUATION.
It should be observed in conclusion that the argument
for the exaction of high constructional prices would remain
valid even if a greater stability in the general price level
were to be secured by monetary reform ; but that in the
absence of such reform an artificially low level of construc-
tional prices is doubly ill-advised.
The general conclusion is that combination would have
it in its power to restrain in certain important directions
the tendency to over-investment ; but that it is uncertain
whether it would in fact do so ; and also whether the advan-
tages to be obtained would in any case be sufficient to out-
weigh the general arguments against monopoly control.
(2) Even, however, without actual combination, a some-
what saner and more centralised investment policy might
perhaps be secured by a greater publicity and diffusion of
information. It may be suggested that a detailed report
of new contracts for structural work or machinery in any
trade should be compulsorily submitted to the Board of
Trade, who should be obliged to prepare in the Labour
Gazette or elsewhere a monthly analysis of such reports.
The experience of the shipbuilding industry, however, leads
one to doubt the efficacy of such knowledge, without the
intelligence to draw inferences therefrom, in restraining the
actions of business men.
Further, either by private or governmental action, the
available information as to the condition of stocks should
be vastly improved. This applies not only to instrumental
goods, but to raw and manufactured consumable goods as
well. It is indeed true that the mere existence of informa-
tion will be of little use unless accompanied by a wider
understanding than at present prevails of the true meaning
of over-investment, and of the real if limited element of
truth still contained in that theory of the wages-fund so
long discredited by professional economists. But proper
information might well be the first step towards the acquisi-
Junghann says that the low prices of iron in Germany are the result
of the over-competition of the German iron producers " (Walker,
Monopolistic Combinations in the German Coal Industry, p. 236).
CONCLUSION. 247
tion of understanding : for the rest, it seems that we must
still wait for some kind of Employers' Educational Associa-
tion which will enable theorists and business men to pick one
another's brains with the same frankness and mutual respect
which is coming to prevail between the former and the
working-classes .
(3) This consideration leads us naturally to the third
remedy — an improvement in banking policy. It seems
clearly on the whole desirable that a banker in making a
loan should have more explicit knowledge, not simply of
the actual security accepted as collateral, but of the prospects
with which the loan is applied for, and the purposes for
which it is used. The existing " no business of mine "
attitude of English banking in this respect has indeed great
incidental advantages, especially in the avoidance of the
suspicion of anything like tied contracts : 1 but it renders
it impossible for the banks to assist as efficiently as their
influence and prestige would otherwise enable them to do,
in preventing an undue absorption in permanent investment
of those consumable resources the title-deeds to which are
entrusted to their control.
(4) The defects of our existing system of currency have
attracted such universal attention in recent years that
little remains to be added here. Briefly it may be said that
from our present as from a more general point of view any
reform which would make the volume of currency vary in
response to the amount of work it has to do and to no other
influence would be on the whole beneficial. For a definite
scheme, reference may be made to the proposals of Pro-
fessor Irving Fisher.2
We must, however, be on our guard against condemning
all the movements at present manifested by the price level
in the course of an industrial cycle as wholly injurious.
Three instances in particular may be given.
(i) The influx of gold and the fact or expectation of a
1 Cf. an able and instructive article by a leading banker in War
and Peace, June, 1914.
3 Purchasing Power of Money, pp. 337-347.
248 STUDY OF INDUSTRIAL FLUCTUATION
higher level of prices hc.s an important influence in the early
stages of the boom in raising the volume of production and
exchanges from an undesirably low level. It is true, how-
ever, that the need for this medicinal influence is mainly
contingent on the misapprehension generated by the pre-
vious fall in the price level : so that a completely steady
level would go far to remove the whole difficulty.
(2) More important is the consideration that the accu-
mulation of stocks during the depression is largely caused by
the falling price level. Without denying that this process
is frequently carried to an extent prejudicial to the best
interests of the community, we must remember that it is
on the existence of such accumulations that the community
largely depends for its power to take advantage of any sud-
den invention. Unless we can count on a uniformity in the
progress of invention and the development of new countries
which seems at present outside the bounds of practical
politics, it is by no means clear that a uniformity in the level
of stocks is desirable. How far the scale of accumulation
is dictated by monetary misapprehension, how far by sheer
satiety, and how far by wider views of industrial prospects,
it seems impossible to determine : but it is at least possible
that a steady price level would have dangers of its own in
rendering the community unprepared to take advantage of
new acquisitions of industrial knowledge and power.
(3) Finally, the falling price level immediately consequent on
the crisis gives a bonus to the working at the expense of the
employing class, which is prima facie desirable in the interests
of society. It has indeed been urged in influential quarters,
notably by Professor H. S. Jevons * and Professor Pigou,2
that labour organisations would be wise in their own interests
voluntarily to surrender the bonus, and to acquiesce more
readily than at present in a falling level of money wages.
On the other hand it must be remembered first that if the
men are employed in constructional industry, the demand
for their labour at such a time is likely to be inelastic, and
the aggregate income of members therefore lessened, even
1 Causes of Unemployment, ch. II. z Unemployment.
CONCLUSION. 249
though unemployment be avoided, by the acceptance of
lower wages : secondly, that there seems reason to believe
that the employer is frequently for a considerable time
unable or unwilling to retaliate by curtailing employment.1
On the whole I cannot help feeling that, in spite no doubt
of errors of judgment, the Trade Unions have known their
own business in this matter better than is always admitted, 2
and that the falling price level at such times is a real source
of social advantage.
§ 3. REMEDIES— THE DEPRESSION.
Finally, we have to consider those incidental effects of
over-investment which deter the community from enjoying
to the full during the depression the large income of con-
sumable goods which over-investment has rendered physi-
cally available.
The difficulty arises from the inelasticity at such periods
of the demand for, in particular, the products of the transport
and constructional trades. We have already seen (p. 205)
that this constitutes an inducement to business men in
other trades to restrict production, and that in so far as
they are unable to do so, a restriction of production by
constructional producers may in certain circumstances be
beneficial not only to themselves but to other business men
as well. We may now add, in the light of ch. iii. § I, that
even if consumptive leaders find restriction easy, such
restriction will not be in the interest of consumptive wage-
earners, and that since constructional restriction will remove
the inducement to consumptive restriction, an additional
1 Cf. p. 226 ; Professor Pigou's whole discussion of sliding scales
(cf. Industrial Peace, Part II.) seems to me to under-estimate the
extent to which the brunt of short-period fluctuations is borne
by the employer.
2 The advantages of a viscous money wage in (a) stimulating
improvements (b) affording a visible token of bargaining strength
must also be considered. M. Baranowsky, however, seems unduly
confident of the advantages of the growing divergence which he
detects (op. cit., pp. 337 ff.) between the English curves of unem-
ployment and of real wages in the 1908 depression, and which he
ascribes to the power of Trade Unionism.
250 STUDY OF INDUSTRIAL FLUCTUATION.
argument emerges in favour of the former. Finally, restric-
tion is clearly in the interests of constructional producers
themselves : and it seems that the increase which it makes
in their aggregate net satisfaction will be likely in any case
to exceed any accompanying decrease in the net satisfaction
of other classes, since their need for consumptive goods is
ex hypothesi more urgent than the need of those other classes
for constructional goods. The classical doctrine of cartel
policy, therefore, which advocates restriction of production
and export of surplus stocks in time of depression, seems
to have more to be said for it than some modern critics are
inclined to admit : and the argument for monopoly control
of constructional industry not only during the boom but
during the depression is thereby strengthened.1
Restriction therefore, while at best only a Sevrepos TT\OUS,
is in the circumstances an effective one : it does not follow,
however, that it is the only one. Its advantage consists in
the maintenance of an artificially high rate of exchange of
constructional against consumptive goods : but this advan-
tage can be obtained in at least two other ways. In the
first place, we may notice the plausible suggestion of Mr.
Hull that by making for stock in bad times, a combination
will be able to prevent inconvenient pressure on the stocks
of constructional materials in the ensuing boom. He justly
remarks on the strength of the " cyclical delusion," preva-
lent in every period of low iron prices, that high prices will
never come again. Yet nothing can be more certain, on the
1 With regard, however, to the third consideration mentioned
in the text, it should be observed that in so far as the combination
has made abnormal gains by maintaining socially beneficial high
prices in time of boom, it may fairly be expected to sacrifice itself
to an equivalent extent to the rest of the community in time of
depression. It is by no means clear, however, that it will have
made such gains on the whole ; in particular in the transitional
period after the first breaking of the boom it seems probable that
while the demand schedule is lowered its elasticity is greater than
unity (cf. pp. 64 and 245), so that the maintenance of beneficially
high prices will involve a sacrifice of monopoly revenue. Moreover,
in so far as the slump in the constructional trades is due not to
miscalculation but to the inevitable discontinuity of investment
(cf. ch. ii. § 3), they seem entitled to special consideration.
CONCLUSION. 251
basis of experience, than that they will come ; and for a
financier with large resources and waiting-power there would
seem to be no safer and more profitable investment than to
buy up all that the furnaces could produce. Yet while
short and feverish bursts of speculation are common enough
on the warrant market, far-sighted and deliberate bull
operations of this character seem to be unknown. And it
is, I have been told, almost impossible to borrow in this
country on the security of iron, unless held in the public
stores. If then either by a clearer realisation of their own
future prospects1 or by the assistance of far-sighted specu-
lators, iron makers could be induced to go on producing for
stock, the waste resulting from restriction of production
could be obviated, while the restriction of sale continued.
But if our analysis is correct, the stocks thus created must
be of a peculiar kind. They will be harmful, from our
point of view, if they resemble the present iron stores, hang-
ing like a cloud over the market, and lowering prices still
further in time of depression : or even if they resemble the
suggested Lancashire cotton reserve, their volume contract-
ing or expanding automatically with the rise or fall of
market price. The kind of store which we require is one
kept upon a mountain out of sight of the market by an
omniscient Olympian, who shall regulate prices in time of
boom in such a way as to keep the ravages of the investment-
bacillus within bounds, and in time of depression in such
a way that he neither gains nor loses at the expense of the
rest of the community on the whole.
Whether this is a function which corporation or cartel
will ever be able to fulfil is a matter of opinion. " It is not
sufficient," says the Wisdom of France, "to be philan-
thropic : it is necessary also to be perspicacious." So far
even the philanthropy of the great combinations has not
always been above suspicion ; and as to perspicacity, their
1 In this connection the recent unsuccessful proposal, emanating
from Messrs. Bolckow and Vaughan, for the formation of a holding
and distributing company by the ironmakers of Cleveland, is of
some interest (cf. EC. Jour., Sept., 1913, p. 463).
252 STUDY OF INDUSTRIAL FLUCTUATION.
occasional excursions into benevolence have perhaps been
more disastrous than their customary pursuit of self-interest.
Yet unless they can fulfil their function in some such manner
as that outlined above, they can hardly claim to have solved
the problem of industrial fluctuation.
But secondly, the desired end can be obtained by the adop-
tion of a system of discriminatory prices. Since the trans-
port industries happen to be among those in which a certain
measure of discrimination is most easily feasible without the
danger of invidious personal distinctions, and also among
those in which the demand in times of depression is least
elastic, this alternative appears to be especially worthy of
attention in their case. It may be suggested with due
respect that the objection on the part of certain authorities,
notably Professor Pigou,1 to admitting its justification is
due to an imperfect appreciation of the importance of the
indivisibility of the initial unit of investment, — to a reluct-
ance to go far enough back in the search for that jointness of
supply which justifies a system of discriminating charges.2
1 Wealth and Welfare, Book II. ch. xiii.
2 For instance, a permanent way built primarily in response
to a demand for wheat transport involves the provision of a num-
ber of potential units of passenger transport as well, since the
energies of the railway are not likely to be entirely absorbed in the
carriage of wheat. In the light of this consideration the controversy
whether units of transport applied to different objects are properly
called homogeneous or heterogeneous seems to me a barren one :
the employment of more than a certain number of transport-units
in the carriage of wheat will be physically impossible, whether or
no the additional units are " homogeneous " with the earlier. But
I confess that the distinction seems to me in any case unimportant :
nor do I understand why Professor Pigou, having decided (op. cit.,
p. 21 1) that " under increasing returns monopoly plus discrimina-
tion of the third degree [i.e., between sub-markets the composition
of which is not wholly within the monopolist's control] may raise
output above the competitive amount, and is more likely to do
this the more numerous are the markets between which discrimina-
tion can be made," concludes on the following page that under
increasing returns monopoly plus such discrimination " may, but
is very unlikely " to yield an output closer to the ideal output than
simple competition yields. It appears to me that wherever a very
imperfect divisibility of the initial unit of investment prevails,
there is a strong prima facie case for discrimination : and that in
the railway industry, where the official and non-official facilities for
CONCLUSION. 253
In so far as monopoly facilitates the adoption of such a
system, its claims to provide a remedy for industrial
depression are strengthened.
Another method of approaching the whole difficulty is
by an artificial elevation of the demand for constructional
goods. The proposal of the Minority Report of the Poor
Law Commissioners that Government contracts for struc-
tural work should be concentrated upon times of bad trade
has found favour in many quarters, and seems to be deserv-
ing of cordial support.1
It must be observed finally that all such proposals for
increasing the volume of consumption during depression
are, like a steady price level and the more equal distribu-
tion of wealth, open to the objection that they will tend to
check that accumulation of consumable goods upon which
industrial progress depends. How much weight we attach
to this objection depends upon more ultimate judgments,
and upon the solution which we are prepared to give of
the ambiguity latent in the definition at the beginning of
§ 2. What is meant by the most desirable distribution of
the community's income through time ? Is the assump-
tion valid upon which western civilisation seems to proceed,
elaborate classification ensure the possibility of forming a large
number of sub-markets, and where the monopolist's power of deter-
mining the composition of those sub-markets in accordance with
his own interests is by no means negligible (cf. Pigou, op. cit., pp.
222-228), the argument is particularly strong.
1 Professor Pigou criticises the Commissioners' detailed scheme
(Wealth and Welfare, pp. 483 ff.) on the ground that it involves
introducing unsteadiness into the demand for one group of workers
in such wise as to balance unsteadiness of the demand for other
groups, and that its efficacy in diminishing unemployment depends
therefore on the mobility between groups. In view of the nature
of the work projected, which is all (with the exception of afforesta-
tion) of the same nature as that required in private industry, I
think he over-emphasises the importance of this point. Mr. Haw-
trey's attack upon the proposal (op. cit., p. 260) scarcely deserves
formal refutation. He asserts that " the Government by the very
fact of borrowing for this expenditure is withdrawing from the
investment market savings which would otherwise be applied to
the creation of capital." The whole point is that in times of depres-
sion savings are not otherwise so applied.
254 STUDY OF INDUSTRIAL FLUCTUATION.
—that it is desirable so to manipulate one's income-stream
that it shall flow in with an ever-rising tide ? From some
points of view the whole cycle of industrial change presents
the appearance of a perpetual immolation of the present
upon the altar of the future. During the boom sacrifices
are made out of all proportion to the enjoyment over which
they will ultimately give command : during the depression
enjoyment is denied lest it should debar the possibility of
making fresh sacrifices. Out of the welter of industrial
dislocation the great permanent riches of the future are
generated. How far are we bound to honour the undrawn
bills of posterity, and to acquiesce in this never-closing
hyperbola of intersecular exchange ? Shall we sacrifice
ourselves as willing victims to the
Urge and urge and urge
Always the procreant urge of the world ?
Or shall we listen to the words of one of the wisest of English
philosophers, who counsels us to eat our grapes downwards,
and who always washed up the knives first in case it should
please God to take him before he got to the forks ? The
question is one of ethics, rather than of economics : but
let us at least remember that we belong to an age which is
apt to forget the ov eveKa among the &>v avev ov, and immo-
late ourselves, if we must, with our eyes open and not as in
a trance.
TABLES
TABLE I.
PETROLEUM OUTPUT IN MILLIONS OF LONG TONS.
U.S.A.
Russia.
Dutch
Indies.
Galicia.
Rou-
mania.
India.
Mexico.
World's
Total.
1899
7-2
8-3
_
•3
_
•I
_
16-8
1900
7'5
9-8
—
•3
•4
•2
—
18-6
1901
8-8
P
•I
'4
•3
•2
—
19-9
1902
II-2
P
•5
•5
•3
•2
—
22-9
1903
12-8
'9-8
2-0
7
'4
*4
—
26-2
1904
14-9
IO
'9
•8
•5
•5
—
28-0
i9°5
I7-I
I9-0
(»
I-I
•8
•6
•6
_ 1
27-0
29'9
1906
I7-9
8-2
1-2
7
•6
•6
—
29-8
1907
23-3
8-4
2'2
1-2
i-i
•6
'3
36-4
1908
24-4
77
2'3
1-8
i-i
•6
•6
387
1909
24-4
8-0
i-5
2-1
i'3
•9
•4
39'2
1910
28-3
9-0
17
17
1-4
•9
•5
44*2
1911
29-0
9-2
1-7
i-5
W
•9
i-o
44'5
1912
29*9
9-2
i'5
1*2
1-8
ro
2-6
48-1
1913
337
9'5
1-5
1*1
1-9
ri
3-6
53'3
Note. — The estimates since 1905 inclusive are from The Mineral Industry.
edited by Charles Of, New York, 1911, p. 557 ; for earlier years they
are those of the Annual British Reports on Mines and Quarries.
255
256
TABLES.
TABLE IA.
PRODUCTION OF PENNSYLVANIA, ETC. (IN BARRELS).
1902
. . . . 32-0 m.
1905 .... 29-5 m.
1903
29-9 m.
1906 .... 28-3 m.
1904
.... 31-4 m.
TABLE II.
ENGLISH RAILWAYS.
(From Whitaker's Almanacks.}
New
Railway
Mileage.
New
Railway
Mileage.
New
Railway
Mileage.
New
Railway
Mileage.
1846
595
1856
427
1866
565
1876
208
1847
909
1857
387
1867
393
1877
205
1848
1,182
1858
448
1868
449*
1878
256
1849
904
1859
460
1869
449*
1879
363
1850
59°
1860
431
1870
392
1880
237
1851
—
1861
436
1871
39
1881
242
1852
—
1862
692
1872
238
1882
282
1852
—
1863
771
1873
268
1883
224
1854
368
1864
467
I874
366
1884
183
1855
226
1865
500
1875
216
1885
—
* Average.
TABLES.
257
TABLE III.
BRITISH TONNAGE.
(From Lloyd's Register.}
On Register,
Broken-up,
Sold Foreign,
Total
January i.
Wrecked, etc.
etc.
Diminution.
100,000
100,000
100,000
100,000
Gross tons.
Gross tons.
Gross tons.
Gross tons.
1887
98-1
3'4
i-5
4-8
1888
98-2
3-o
1-7
4-6
1889
IOI-O
2-4
1-8
4-3
1890
106-6
3-i
1-9
5'0
1891
in-5
3-3
1-6
4'9
1892
116-6
2-9
1-4
4-3
1893
I22-O
2-7
2-5
5'i
1894
124-9
3-i
2-6
57
1895
I28-3
3'i
4-0
7-1
1896
1297
3-2
37
6-9
1897
I3I-4
3'3
4'5
7-3
1898
I3I-6
3-o
6-3
9'3
1899
1337
2-6
6-3
9-0
1900
137-6
2-7
6-7
9'4
1901
I40-3
2-4
4-1
6-5
1902
146-4
1-9
3-2
5-i
1903
153-5
2-6
3-6
6-3
1904
I57'7
2-5
3-6
6-1
1905
I62-8
2-5
6-2
8-6
1906
166-8
2-4
4-9
7*3
1907
174-6
2-4
4-2
6-6
1908
180-8
2-8
2-7
5-5
1909
182-3
2-7
4-1
6-7
1910
184-0
37
5'9
9-6
1911
184-6
2-5
7-7
IO-2
1912
188-1
3-5
7'4
II-O
Total
368-7
737
104-1
1777
TABLES.
TABLE IV.
PROPORTION OF VESSELS, REMOVED FOR FOREIGN TRANSFER
IN EACH YEAR, BUILT BEFORE —
1885.
1890.
1895-
1900.
1905.
1896 . .
75
—
—
—
—
1897 . .
70
—
—
—
—
1898 . .
50
66|
—
—
—
1899 . .
50
60
—
—
—
1900 .
38
55
—
—
—
1901 . .
49
57
—
—
—
1902 . .
45
58
—
—
—
1903 . .
59
71
—
—
—
1904 . .
35
55
80
—
—
1905 . .
43
62
78
9°f
—
1906 . .
36
48
65
80
93
1907 . .
3i
5i
67
78
90
1908 . .
36
5o
65
75
83
1909 . .
27
47
72
8ii
?
1910 . .
19
33i
73*
85i
95
1911 . . 1 14
29i
59
81
91
1912 . .
ii
251
47
65'A
83iV
TABLE V.
Brazilian
Coffee Crop
(Estimates
of " The
Grower,"
quoted in
Economist
Histories) .
Brazilian
Jute-yarn
Imports
from
United
Kingdom
(Annual
Returns of
Trade).
Brazilian
Coffee Crop
(Estimates
of " The
Grower,"
quoted in
Economist
Histories) .
Brazilian
Jute-yarn
Imports
from
United
Kingdom
(Annual
Returns of
Trade).
Thousand
Million
Thousand
Million
bags.
pounds.
bags.
pounds.
1896
16-5
1906
10,277
31-3
IS*9Z
—
23-5
1907
19,663
35-o
1898
—
21-4
1908
10,304
24-2
1899
—
20-3
1909
12,419
21-9
1900
8,971
21 -I
1910
14-944
25*6
1901
IO,25O
28-0
1911
10,548
22-0
1902
15496
31-8
1912
12,464
22-6
1903
12,324
28-0
1913
10,750
14*5
1904
10,407
21-3
1914
? 12, 500
I0'3
1905
9>973
247
TABLES. 259
TABLE VI.
BUILDING PLANS AUTHORISED BY URBAN DISTRICTS.
(From Quarterly Returns in the Labour Gazette.)
I
i
Selling-
houses.
£100,000
Factories,
etc.
,£100,000.
Shops,
etc.
^100,000.
Total.
^100,000.
No. of
Areas.
1909
I
2
21-3
2-3
2-2
347
90
3
21'0
3'9
2-6
39'2
91
4
157
2'5
2-9
28-6
92
1910
i
157
3-o
i'9
28-8
92
2
20-3
3-o
3'i
37-o
90
23'3
4-0
37
43'2
99
3
16-6
3'2
3'5
32-4
91
177
4-1
4-0
36-4
IOI
4
14-4
2-9
2'3
27-2
92
15-4
37
2-6
30-6
98
1911
i
16-6
3'6
4'i
33-8
92
197
5'4
4-6
40-9
99
2
17-4
5'5
47
397
99
18-4
57
4-8
4i-3
100
3
13-1
4-6
37
33'3
IOI
13-6
47
37
34'i
102
4
i3'i
6-0
3-6
34-o
98
12-8
6-0
3'5
33'3
96
1912
i
i5'i
6-8
37
34'9
99
2
14-5
II-2
3'4
40-2
IOO
3
12-6
7'5
3'9
347
IO2
14-4
6-1
3-0
30-4
91
4
12-8
9-0
2-8
35'2
96
I2'I
7-2
2-8
3i-8
93
19*3
i
I4-I
6-3
3'9
34*9
94
2
16-4
87
4'2
44'5
92
3
12-0
7'4
4'9
347
9i
4
n-3
4-0
2-9
29-6
93
1914
i
17-3
6-6
4*5
39'9
94
2
18-0
7-1
7-6
45'6
92
260
TABLES.
TABLE VII.
ANALYSIS OF RESULTS OF TABLE VI.
On Preceding Quarter.
On Corresponding Quarter
of Preceding Year.
Dwellings
Factories.
Total.
Dwellings
Factories
Total.
1909
3 Fall
Rise
Rise
—
—
—
4 Fall
Fall
Fall
—
—
—
1910
i Rise
Rise
Rise
—
—
—
2 Rise
? Stationary
Rise
Fall
Rise
Rise
3 Fall
? Stationary
Fall
Fall
Fall
Fall
4 Fall
Fall
Fall
Fall
Rise
Fall
1911
i Rise
Rise
Rise
Rise
Rise
Rise
2 Fall
Rise
Fall
Fall
Rise
Fall
3 Fall
Fall
Fall
Fall
Rise
Fall
4 Rise
Rise
Rise
Fall
Rise
Rise
1912
i PRise
Rise
Rise
Fall
Rise
Fall
2 Fall
Rise
Rise
Fall
Rise
Rise
3 Fall
Fall
Fall
Fall
Rise
Rise
4 Rise
Rise
Rise
Rise
Rise
Rise
19*3
i Rise
Fall
Rise
? Rise
PRise
Rise
2 Rise
Rise
Rise
Rise
PFall
Rise
3 Fall
Fall
Fall
Rise
Rise
Rise
4 Fall
Fall
Fall
Fall
Fall
Fall
1914
i Rise
Rise
Rise
Rise
Rise
Rise
2 Rise
Rise
Rise
Rise
Fall
Rise
^
J2
8
OM-t O\ •* *x CO «O co cO M N O
VO N. M O co O i-c cotxoo TfV)
M CO CO N CO *x COlOCO-^f^O
+ + £ , + +++M M *
S ££ SJ'S^SS) 5§ S> ?
*^ 95 1 °. °. 9. *"; ^t "^ °° "*£ w
+ ++ + ++++ 7 T
VO
H
||
i
I8ss 'iil'l'l
> a —
•^ •^-4-^4- "a-MtJjd
i '!Hi+"'H!H
v>
H
rf]
II
5|
^COCO -0 CO £££$;*£
CO OlM •<*• CONtXHCO M VO CO
CO NCS» W HMHNM W M Tf-
J*
•N
• S ^
+ + + + 1 s =1 l| =1
X O'-M"^ tJjd-O
O II i o^'^flJ ^o°3
« ++ + ofe£55 5 = f^m
+1*1 +» 1
CO
M
s !
i
S"++ fill 'IN1
QW
!!l+ + Ml= + l|
M
a
1
1
a ^-o -3 a -3
+++•^8 1 1 1 l| "fl
c/5.rt so j 4. «)
0) — «-»
ll+l ' l"flf= 5ll
H
1
opcoptx vo y> ^^p?1?*"^
MHHH H M (t W «t M M «l
ip ip p CO Op w^ CO p\ O N p O
O*) tN IN H M M tN C< CO CO CO CO
<"
1 1 + 1
i
OONOiON 0 0 OOONOiO"
MS w co
o\
i
£
•^J-CONM OO tx Y^P"?Nr>7'"
NNNN f» CO COCOCONCIM
VO VO M »N J» ViOp *^ N CO W «^
CO
|.i
PJOO COM .0 vp 0 copvpf ON
bbow o o oooi-ioo
°\^+ + + | |+++
1-1 VOH O OOCOWONTh M N CO
6 bb b bbboM b b b
1 1 + + +++++ + 1 +
tx
l)
vo o\**m w *^ *s^a>?CTlP
MCOMO< 0 0 OMWMMN
++++ + + ++++++
•<f y> PI N p» M «-> v-iop M o o>
M OW H HMMob O O
+ ++ + +++++ 1 1
VO
|»
s-s
OVO 0 0\ f -0 rxvpoprj-cp^
MMMO O O OOOWHM
++++ + + ++++++
0 MM NNCONONOpVO N 0
N NN M MMHOO HI O M
+ ++ + +++++ + + +
IO
• .4
g
O O> «•> n ^ M OOH T*"VO ^? ^
NN'NN « « WNNNOOCO
O\ C*-* «-> pNO^xON « *x OC
CO CON ij- NCOfOCOCO ij- V vo
Th
iw
CO CO N •* CO </"> ^sOOpNOlThM
NNMN w « MNPJNCOTJ-
M O *r> •* jtcocoinu^ fx co jf
ij- i*- CO CO CO CO CO CO CO CO V ir>
CO
u
^^gS *S 0 S?$S«««
^^^.^ -,)- ^. corococococo
VO fs CO v> •^•THxO>tv M CO OO
OO K. O\ CO •* •*• •*• •* ts. OO ON tx
M WN CO CO CO CO CO CO CO CO CO
•
g'O'ai
.Sai
8%cON vp «o OAppopvpvp
•*"O J, M O O «1^O O> N H w
55+1 1 + 7++I 1 1
o o to p ^pp^^ir>ir> r> 9 f
o N wbboo c* c* w
+ 1 + 1 1 + 1 + 1 1
- 1
d
%%£% * % coco^coc^jo
co c$ W « co co co fO ro co co co
i
ON
M
III! I I
261
°-i II 5 Util 1 1 1
£
£ £«S-?o ££ ££ SS "£
M
|
T TTT TT T? T i
*
J
••£••-
>, -M ^ ^-r^
O "t2 ^O 1 JH & | ^ ^ ''''.L.O
H< _j_ -J-
IT)
ll
CO COOO fx CON NOO OOW txt^
I*-
» I!
.1- >.
CO
M
'3 ^
H
1
1++!'l^l+ '!
a
I
- viiS^^j.- « +1 1 +-S
^|I1|^S I «- ^^s^o1
M
Si
5
3
»« ss?ir « w « ss
0
M
1
!> 1 1 1 i S i !, 4 1 1
3 g j§ a a a i
I
* poop ^^ J.M ,*.£, jo;.
s
,
00
1*
M bb bb bb ob b « "
+ + 1 II +!+++!
i »
CO n -«j-00 COCO MT}- M-«i- MCO
fs
g
b MOO bb wb MO bb
+ + 1 1 1 + ++ ++ + 1
VO
it
co f>.vp M rt- N M c* >ooo \r> M
o ooo oo bb bb bb
+ + 1 + II ++ ++ + I
»._
•I
10 °^ 2 M M
*
ipi
O\OpMfx fvOS ^-^ OO ON
CO
4
OO OOO\O n^ O\O O\O OvOO
co coco-* ^.co coi*- co-* cool
M
O fl &>
*s MpCTivpp OO MOO OM
MMOONV NOCOO
1 +1+ 1+ II 1+
"
*l
OO ID Tj-VO MTj- >O*S >OTj- ThtX
CO CO CO CO COCO COCO COCO COCO
I
j, iii n si is ii
slslili
.c ."J^^g
•stiiii!
•— * ^
's'-s'8sa
^ti'SMi
^tiliil
262
•S5POQ
V V Jovb vb£sbbM^booM
+ I I I I I I I I I I | |
•jpug
5=| |
MvoMMOMMQio<n<o to
1 +++++ 1 1 1 ++ +
>N-*OMOO-*io«<nt-i
+ 1 1+ 1 + 1 1 1 ++
bwbbbbbbbbM c*
I I I I I I I ++++ +
fo o ^ «o o\vo r*s M u^
• 83
i+l1
O\ t* O tV M M O *•> O\0p N O(*p OO O O VO *N »O N «•> f< SO
voWir>V'*ir> £s vbvbvbobobobdb Kob do ob do
•joog
MNNOWsbfoOMN'O Jo MNbcOMfOMMMOHO
1 ++++ 1 1 + 1 + 1 + 1 +++ 1 1 + 1 1+ 1 +
0 J*p H VOVO
M^MCO TfVO
•apeui
-a^odsag
PQ
MThp^-p^Nopvpvp^NM jh t<tHVPZ*'?17*'r'iH y *P S* f
NNwbbVbbbbV co vbMMtowr>Mf«NcofOMf»
1 ++ 1 M 1 + 1 + 1 1 1 1 1 1 M ++++++
1 ++++ 1 1 ++ 1 1 + 1 1 ++ 1 1 1 + 1 M 1
•ajnf
*xp 0 »^-tX M r>0 pNN W Tj-0 0*XW M M
O O HNONO<ONNOMHO
MMtXWQOMMOO
1 +++ 1 1+ 1 1 + 1 +
+I++I+IIII+
Nw-*OMONOOOM Tf txOrONfOOWOHO«OC*
1 +++++ 1 I++ 1 1 1 1 1 + 1 1 ++++++
++++ 1 1 1 ++ 1 + M 1 1 1 1 1 ++ 1 ++
•<J- tx TJ- c< V>VO M VO 1OOO N fO >OOO to fO 1O tx N M
^-H^M^oNboMM b £ b NO H fob H WN
1 ++++ 1 1 +++ 1 1 1 +++ 1 1 1 ++ 1 1 1
•Smpimq
txtXVOVOVO
•SuuaauiSug
tx VO O VO O tvoo <>» vO two t^
T»- fO «0 fO *0 «0 <0 CO TKO •* Tf
ptre noji
« P^^^^?•Hp ro-ptx
OOOOOOMQHNfO
++++ 1 + 1 III
jh .pvp .ocp vp sp ^vp «i «O N N
N NOHWMOOOOOMO
+ 1 1 1 1 1 1 I++ 1 1 +
•uoji-Stj
tx>0 <OiOO »0 moo <OVO N txO ^*NO>O
T*-M--^TOfO M O O\ Cfi O\OO OO VO *N *>.OO OO O\
•^061
263
H-l
PQ
pt
,,000
10 *x n P» ° J-x
1 + +++ +++++++
§§
n S
f>
•jpug
!|+!++!l:':!'S|l
C >>
If
2 c<
P*
•**»*
1= ««. -+i^| '.
s
n
3
•SSBIO
P£ ^P^ 29^5.55
T+ TT+ Tff+r++
M
g"o
0
•Sinusnunj
OtN MOCO Mrj-fxpJQ POOO
S8
++ + 1 1 1 1 1 1 1 1
^g
2
•sinjuuj
POO *?"?"> u^-<l-OvOPO^io
MM bob bbbbboM
++ +++ ++ 1 1 1 1
* fc
"•£*£
°T3
OO
'-**
.b , x + "8
c8 s : :-f" « s : : o is
fa fa a,
3 II
^ 0
co M
^
•i9q}B9l
^n^
>1
VD
H
-,003
CO-M OOP* COM M O 0 M M
++ 1 ++ 1 1 1 1 +++
.Sao!
»O
.,BH JP4
<O •* •* CO CO ThVO 1O N 10 M p»
% V
CO fl
£ §
•apBm
h
§1
T3 ^
oi ^pBsy
+fa
^0
O
CO
M
•ajjodsgg
| s II s 8 |1 | s s s
to Sfa O^wS
o ..•
S|
<l^ co
o a
2
•AjgisoH
M+ +++ ++7+++°
.3 8
II ft.
H-I &.
N <.
M
•90BT
P»M MOP* OMboNob
1 + + 1 1 1 +++++ 1
4J
T3 (H
%B
O
•1ITS
bb cocob cooMOMob
++ +++ + 1 ++ 1 +
^ 0 '
.•&!
vN £3 J
Tt-fx 0 P* N ONOO MVO M COfN
M 8 J
i <o r
*
s;nj[
MO MOM 0^-OOOMO
++ + 1 + ++++ 1 1 1
J>§<
OO
•»«
MCO OOP* pibbcobbb
++ + 1 + ++ 1 ++ !
1«
-
•P^OM
T+ +++ +77++°-?
Sll
rt c
*N<N VO^-PO 10 PO 0 -* N N Tj-
M U-> ^
VO
POM
MCO COP*PO MQMCOMMO
1 + +++ + 1 1 ++ | |
° « £
*
•ampjinq
IF SSI? Iff^saffl
-1S
_ 9
H OO VO "th to MM COOO CO 1OOO
° "o c
•"*•
3TnJ99UI3U3
COP* <MP*p< P*P*MOOO>Os
^^ O "*"
rrt f"
PO
•9}B[dmx
OO tx 1000 O O VO ^* PO O\^O M
•* TJ- T)- PO 10 in •* THO lovo <x
«O to
•*• •*• T)- rj- Tf -<j- -«J- TJ- ^j- rh -»j- •*•
1 r£< P
•pa?S
* S £
PUB UOJJ
+ 1 1 + + 1 +++ 1 +
c §
3 B «
H
•unit-fta
•* CO tx O O\ M 1O CO tx tx •*• M
OO OO OO OO OO OiOO O> O\ O> O O
P»0« P1WP4 OIMNMNCOCO
§ 01 a
S o c
•6061
& ill
II
264
I
s
•f
w
o
1
ep ivfxop c?\ jo *op vp w p
bbbbb bbbboM
1+ 1 1 1 1 ++ 1 + 1
o\
I
•*e»Nrof>w OfOo->Ov>f»
boMbbb bbbbbb
1 1 1 +++ ++ + 1
00
!«
wvpvpopvp^- Y^rir*°PT*'p
fete
tx
M
>>
B
fOfOMrJ-'**-! TfO«">ONO
bbbbbb MOOMOO
1 +++ 1 + 1 +++
VO
M
e
3
N M iri Tfvo f^ y^ N vp covp vp
MOONCOM bbboMO
1 1 +-M 1 IH-+I M
ir>
M
1
^*NOiO>OO OOOOMMVO
OHobbw bboMbb
+1 +11 +++ I ++
M
&8
1 11
KS+U +Il5=i
tr>
1 I
1
$•-4 H!"
, .
M M O«3O 10<O M«COTh"iJ-Tj-
M
Is
OONOOO MOOOOO
+1 +1 + 1 1 1 +1 ++
M
1
o
M 6 H M 6 M O C^ W ^) ^ ^
1 1 1 ++ 1 I ++ 1 + 1
0
1
WWOMCOM txOO^Or-
2
»O>O1OIOTHO VO VO <X*O >0 •*•
<?>
it
s.
voyii^w^-io Mrt^oo'-'
bobbbb bio bob
1 1 1 + 1 1 III +
00
It
M M O «^ ^ « t^vp TJ- -r ^vo
MMHOOH bOMOOb
1 1 1 1 1 1 M 1 1 1 1
tx
• 1
m
O N tNVO ^VO M <M 10 N Tf M
OHOMMti OMOOHW
++ 1 1 1 ++++++
VO
T3 .
p
oo en -^- M \o vo >o woo M ix -rh
bbbbbb OMobbb
1 1 ++++ +++ 1 1 1
10
M
00 0\ 0( 0\00 CO 00 C* 0\ "*0s Th
"1
•S ^
•+
si
MMNNNW NNNC*«fO
CO
M
rOTf-voioooo f^roNOOPOio
cniovOvOt»0 lv.00 ixvo 00 u-j
-
N
m
MHUINOl-l OONNOO
+ 1 1 +++ ++ 1 l+l
M
ac§
E£
SSSSS" £E£g8£
fOfOcn&NN NMNNNN
fO
.
C*
1811
I'll
. ftM «
Hrt ^-i >>
S <u o
"HI-
il
265
TABLE IX.
NEW YORK STATE. UNEMPLOYED PERCENTAGES.
(From Labour Gazette.)
Clothing.
Wood-
working
and
Furniture.
Printing,
Book-
binding,
etc.
Building,
Stone-
working,
etc.
Metal En -
gineerin'^
and Ship-
building.
June, 1907
7-8
15-9
4-0
97
2-5
Sept., 1907
10-0 9-0
5*3
127
6-2
Dec., 1907
41-1 26-2
6-1
40-9
28-8
Mar., 1908
46-2 29-3
I4'5
54-8
29-5
May, 1908
49-3
35-6
17-8
37'1
32-3
June, 1908
45'4
35-o
173
35*1
29-0
Sept., 1908
29-5
19-6
10-4
32-2
22-1
Nov., 1908
21-2
19-6
10-6
35-4
I87
Dec., 1908
l6«3
20-2
I2-I
43-o
18-5
Mar., 1909
ii -8
15-9
6-5
34'9
i5-9
June, 1909
22-1
12-6
9-1
19-1
II -O
Sept., 1909
14-6
8-8
37
12-4
6-3
Nov., 1909
15-0
6-1
4-4
17-2
5-o
Dec., 1909
20'9
9'3
6-4
23-9
6-3
Mar., 1910
15-9
13-2
5-o
22-1
4-0
May, 1910
31-5
10-3
4-2
15-9
2-9
June. 1910
28-0
5'3
3'9
15-5
3'3
Sept., 1910
I7-I
7'5
4-8
9-6
4-8
Nov., 1910
28-6
7-6
0-6
19-9
5'5
Dec., 1910
47-8
14-8
i -i
29-1
6-6
Mar., 1911
16-6
15-3
3-6
38-3
12-4
May, 1911
38-5
17-1
3-3
28-5
9'3
June, 1911
20-7
ii -6
1-2
26-1
12-4
Sept., 1911
9-2
12-7
37
13-1
12-2
Nov., 1911
28-5
16-6
3-5
23-1
147
Dec., 1911
597
18-9
3-6
3i-9
I5-I
Mar., 1912
15-6
16-8
7-1
35'5
7-2
May, 1912
38-0
17-0
1-8
180
9-0
June, 1912
5i-9
17-8
3-i
13-6
8-9
Sept., 1912
7-1
6-1
5'8
5-8
37
Dec., 1912
58-9
17-8
o-o
18-8
8-1
Mar., 1913
14-0
13-6
6-1
25-1
5-2
May, 1913
40-4
17-1
37
15-5
67
June, 1913
36-4
14-6
3*4
20-3
4'5
Sept., 1913
25-2
9'3
3'3
16-7
5-o
Nov., 1913
46-1
22-8
47
267
7'4
Dec., 1913
66-2
23'4
6-8
39'4
14-2
266
jj o o o o o'
l
a
1 !.§•«!
'
bO^ tuo ^ O
.fi iQ 2
4
<x
X*
267
CHART II.— U.S.A. CONSTRUCTION.
1 . Pig-iron production, i = 2 in long tons.
2. Pig-iron price (No. i Foundry, Philadelphia), i = $2.
3. Coal price — Index, i = 10.
4. Brick price — Index, i = 10. Base-line, o.
5. Building materials price — Index, i = 10.
6. Furniture price — Index, i = 10.
7. Number of immigrants, i = 200,000.
8. New railway mileage, i = 2,000 miles.
(Calendar years till 1889 ; years ending June 30 from 1890).
NOTE. — The indices 1890-1911 are those of the Bulletin of the Bureau of Labour, March, 1912 (average
1890-9 = i oo). f| Curve 3, 1869-91, is the Aldrich Report Index (1860 = 100). Curves 4 and 5, 1885-9,
are the same in terms of the Bureau of Labour Index, on the assumption that the prices of 1890 corre-
spond. Curve 6, 1885-91, is the Aldrich^Index for all house-furnishing goods.
"I
I 3
P 6 £
« 8 -
H 7 o
!l o
.-1 £'£
W <£ g
T!
-rt -H *7
s sis
^ -^ §
is
4,
^_
i
269
CHART IV.— WHEAT.
10 million quarters.
1. U.K. Wheat crop
2. French Wheat crop
3. Russian Wheat crop
4. U.S.A. Wheat crop1 )
5. U.K., yield per acre ) _
6. France, yield per acre)
7. U.S.A., yield per acre, i = i bushel.
8. World's Wheat crop, I = 50 million quarters.
1 Mainly commercial estimates from 1891 onwards.
a = S
z
270
H -i
HH o
§ i i § §
i i i i i
*
.a
^
-
3
&
o
a!"
,"- s
o
*> •ag
? Is
10 O I>
271
CHART VI.— U.S./
1. Wheat crop, i = 100 million bushels.
2. Corn crop, I = 500 million bushels.
3. Cotton crop, 1=2 million bales.
4. Farm value of wheat crop, i = 50 mi
5. Farm value of corn crop, i = 100 mi
6. Value of cotton crop, i = 100 million
7. Railway goods traffic receipts — years t
8. Pig-iron production, 1=2 million tor
1869 71 '73 '75 '77 '79 '81 '63 '85 '87 '89
ning July I, I = £50 million.
'93 '95 '97 '99 '01 '03 '05 '07 '09 13
CHART VII.— ARGENTINA.
1. Wheat crop, 1=2 million quarters.
2. Maize crop, 1=2 million quarters.
3. Iron and Steel imports from U.K., i = 50,000 tons.
Base-line, o.
co evi — o.
CO
CHART VIII.— INDIA.
1. Wheat production, i = 10 million quarters.
2. Rice production, i = 100 million cwts.
3. Wheat exports, 1=2 million quarters (year beginning April i).
4. Rice price — Index, i = 20. ] ( 1873 = 100.
5. Average price of seven chief food-grains — Index, i = 20 ) t
6. Cotton piece-goods Imports from U.K.
a. i = 500 million yards.
b. i = 200 million yards.
I873 = 700-
A
5/
\
A
7
V
V
V
J
vv,
1885 *87 "89 '81 *93
*97 "99 '01 '03
TJ7 "09 II T3
275
\
\
I I I I I
CHART IX. — '
1. Net increase in number of spindles eij
2. Cotton price, B. of T. Index (1900 =1
3. American cotton crop of previous aut
4. Average spinning profits, i = ^2,000
5. Total exports of piece-goods, i = i,c|
6. Estimated home consumption of raw
7. Net imports of raw cotton, i = 2 nl
8. Exports of yarns to Europe (excludingj
v\
v
1869 VI
'73
'75 V7
*79 '81
276
'83 '85 '87 '89
:TON.
)yed, i = 500,000. Zero line, 2
)0), I = 10.
i, i = i million bales.
.ero line, 2.
million yards.
on, 1=2 million cwt.
on cwt.
rkey), i = 20 million Ibs.
'93 '95 '97 '99 *OI '03 '05 %07 '09
CHART X.—
Indices, i = 10. Base-line o. For 1869-96, average iSyo-c
i = Tea. 2 = Sugar. 3 = Meat. 4 = Tobacco. 5 = Cur
7 = Unweighted average, i = 20. Base-line — 25. For i
8. Price of foreign wheat, B. of T. Index (1900 = 100).
1869 71 73 75 77 '79 '81 '83 '85 '87 "8<
DD CONSUMPTION.
For 1896-1912, average 1896-1905 = 100.
Raisins. 6 = Coffee.
tverage 1870-9 = 600. For 1896-1912, average 1896-1905 = 600.
7
\
\
•93 '95 -97 '99 '01 '03 '05 '07 '09 II '13
279
Si
l»
,-r
M
t
5?
<D ;§ P<
•s s-s
* 11 »
2 "I «
A
«O ^ IO OJ _
I
7
\
280
V)
.I
^H £-
O
O '^
ii 1
o o<
I
d B
o g
4-> u_,
a °
§ s
8H <4-l
o
"S 8 5
8 U
a
e .SP
a -c 9 PI
d d
Q o
s s a
ft ft
s g
1 1
o o
CO •<*• IT) \O
in -r to N _
<
££
x
A
•P
281
CHART XIII.— U.K. BUILDING TRADE.
1 . Imports of furniture woods, i = 50,000 tons.
2. Imports of other woods, 1=2 million loads.
3. Furnishing Trades Association, percentage employed, i = i. Base-line- 85.
4. Amalgamated Carpenters and Joiners, percentage employed, i = i. Base-
line-85.
5. Price of timber (hewn fir), B. of T. Index, i = 10.
Price of bricks, B. of T. Index, i = 10.
6.
Price of foreign wheat, B. of T. Index, i = 20.
05 oo N SS g 2 tS
7
(i
4
"j
282
INDEX
I. TRADES.
AGRICULTURE, 75-120, 129-55,
165-70
Agricultural machinery, 183 n. I
Alcohol, 197
Armaments, 72, 203
BACON, 169 n. i, 196 n. 2
Bananas, 197 n. i
Beef, 117-8, 169 n. i, 222
Beer, 197 n. i
Boots, 51, 114-5, 196, 198, 219-20,
232
Brick [25, 44, 56-7, 67, 103,
Building) 118-20, 181, 193 n. 5,
198, 219
Bull-dogs, 71
CEMENT, 82, 194 n. 2
Cinematograph, 71
Clothing, 111-4, 175-7, 220, 223,
225-6, 227 n. 2
Coal, 6, 15-8, 32-3, 35, 52-4, 84-5,
189-95, 207, 223, 244-5
Cocoa, 116 n. 2, 221
Coffee, 22, 43, 84 n., 104, 164, 196,
232
Copper, 23, 28-9, 33, 60 n. 3, 63,
183-6, 222
Cotton, 19—22, 25-6, 41, 46—50,
61 n. i, 71, 72 n. 2, 91, 106-9,
in, 113 n. i, 128 n. i, 181, 194
w. i, 195 n. i, 224, 232
Currants, 116
DOCKS, 208, 222
ELECTRICITY, 27-9, 31, 67, 70,
157-9', 183-6, 222, 242
Engineering, 77, 99-100, 219, 221
FOOD, 115-6, 164, 167-8, 196
Furniture, 120, 222 n. 5
GLASS, 175, 198, 203, 220, 224,
227 n. i
HOSIERY, 175, 202, 220
JUTE, 47-8, 51, 73, 84-5, 108, 194
«. 4, 222
LACE, 71, 203 «. i, 220, 224, 227
Leather, 51, 175, 203 n. i, 219-20,
221-2, 224, 232
Linen, 224, 227, 231
MAIZE, 77, 91, 117-8, 134, I53> 155
Metal, see Engineering.
Motor, 159 w. i, 161 n. i, 219,
223-4, 226
NITRATES, 221
OIL, 23-4, 44, 52, 106, 158-62,
186 w. 2
PAPER, 175, 203, 219, 224
Petroleum, sec Oil.
Pig-iron, 15, 32-3, 35, 52 n. 2, 53-
65, 66, 94-102, 127-8, 189-95,
219, 222, 231, 244, 250-1
Pottery, 203, 220, 224
Printing, 175, 177 w. i, 198, 203,
219, 226
Provisions, see Food.
RAILWAYS, 19, 30, 31, 39-41, 76-7,
122, 161 n, 3, 173, 178-9, 182 n. i,
1 86, 194, 205
Rice, 106-9, 134, !54
Rubber, 23, 51, 231
Rye, 154
SHIPPING and Shipbuilding, 18-9,
24, 26, 34-6, 42-3, 52-5, 6, 72-4,
78-85, loo- 1, 186 n. 2, 189-90,
219, 223
Silk, 72 n. 2, 220
Soya beans, 82
Steel, 66, 70, 183-7, 2I9» 222*
See also Pig-iron.
Sugar, 117-8, 222
Sulphate of ammonia, 106
TAILORING, see Clothing.
Timber, see Building.
Tin, 51, 232
Tinplates, 194 n. 4
Tobacco, 116, 197 n, 2, 224
WHEAT, 76-120, 139-55
Wire, 99, 191
Wool, 27, 41, 47-8, 61 n. i, 67, 71,
73-4, 105, 110-4, I28, 164, 175,
181, 195-6, 202—3, 22O» 222> 223»
225, 227
283
284
INDEX
II. AUTHORS.
AFTALION, 14, 36, 122, 126, 165 w. i,
171 n. 2, 188 n. 2, 199, 200, 212
Aristotle, 209
Ashley, W. J., 208
Atkinson, F. J., 107 n. 5, 148 n.
BABSON, R. W., 93 n. i
Baranowsky, Tugan, 2 n. i, 5 n. 2,
10 n. i, 126, 129 wz. 3, 4, 171 w. 2,
199 n. i, 212, 249 n. 2
Bergmann, i
Beveridge, 2 w. i, 238 w. 2
Bilgram and Levy, 211
Bowley, 115 n. 2, 168, 196 n. 2,
226 w. 3
Broomhall, G. J. S., 139 n. 2, 148 n.,
154
Bruckner, 148 n., 151
Burton, 2 w. 3, 64 n. i, 126 w. 3, 131
CAIRNES, 232 ff.
Carver, 123
Chapman, 26 w. 3, 135 n.
Chisholm, 190 n. i
Cunningham, W., 50 w. 3
DEARLE, 25 w. i, 67 w. 2
de Foville, 192 «. 5
Dibblee, G. B., 61 w.
ELDER, T. C., 67 n. 3
Emery, 50 w. 2
England, Miss, 214 n. i, 218
FISHER, Irving, 214, 218, 247
GIFFEN, 33 n. 2
Goschen, 30 n. i
HAWTREY, R. G., cj «. i, 89 n.,
129 w. 2, 211, 218 n. 4, 222 w. 2,
253 w. i
Hill, J. J., 65 w. i
Hirst, F. W., 1 80
Hirst, W. A., 86-7 ww.
Hobson, J. A., 211, 235-8
Hull, 10 n. 2, 15, 38, 55-9, 76, 89,
138 n., 178, 244, 250
Hyndman, 173 n. i
JEANS, 33 n. 3
Jenks, 243 n. 2
Jevons, H. S., 91 n. i, 144-55, 248
Jevons, W. S., 75, 129, 144, 233
n. i
Jones, E. D., 61 n., 130 w. i, 243
n. i
Juglar, 10 n. i, 211
KEELING, F. W., 208
Kemmerer, 171 n. 2, 213 w. i, 220,
233 n. i
Keynes, J. M., 171 n. 2, 213 n. i,
220 w. 6
LAND Enquiry Committee, 45,
119 n. 2
Layton, W. T., 66 >z. 3, 179 w. i,
216 n. 2
Lehfeldt, 133 n.
Leroy-Beaulieu, 172 n. 4
Levy, 35 w. i, 173 n. 3
Lexis, 167 7«. 3
MACARA, Sir C., no n. i
Macgregor, W. H., 209 n. 2, 243 «. 3
Marshall, 5, n, 14, 113 n. 2, 126,
188 n. i, 218 n. i, 226 n. 3, 228,
235
Marx, 36-7
Maunder, E. W., 144
Mitchell, 126 n. 3, 165 n. i, 206 M. 3
OUALID, 50 n. i
PHILIPPOVITCH, 2 w. 2, 8 n. i
Piatt Andrew, 75 n. 2, 90 w., 91 n.
i, 134 w. i, 153 w.
Pigou, 5 w. 2, 51 «. i, 70 n. i, 124
w. 2, 137 n., 248, 249 w. i, 252,
253 n. i
Poor Law Report, 253
Price-Williams, 190 n. i
RIST, 209 n. 4
Rost, 208
SHAW, 144 n. 2
Sismondi, 211
Spiethoff, 4 n. i, 171 n. 2
Sprague, 63 w. 2, 172 «. 2
TAUSSIG, 41, 209 n. 2
Taylor, B., 161
Thomas, D. H., 17-8, 189 n. i, 190
n. i, 207
WALKER, F., 192-3 nn., 245 w. i
Walla?, G., 39 w. i
Webb, A. D'., 197 n. i
Weld, D. H., no n. i
Wood, G. H., 115, 196 n. 2, 220
Wright, Carroll, 2^3 n. i
INDEX
285
III. COUNTRIES.
ARGENTINA, 82, 86-8, 99, 139 n. i,
143, 148 n., 151-2, 154-5, 181-2
Australia, 23 n. 2, 24, 48, 81, 105,
148 n. i, 233
BRAZIL, 22, 84, 100
CANADA, 83, 88, 93, 100, 105, 148 n.,
172, 174-5. 198
FRANCE, 21, 30, 72, 216
GERMANY, esp. 21, 27-8, 30, 67-8,
154, 157-9, 183-6, 192-3, 234,
244-5
INDIA, 47, 71, 77, 81 n. 6, 82, 90,
106-9, 140 n. i, 148 n. i, 153-4
JAPAN, 72 n. 2, 82
MEXICO, 44, 71, 161 n. 3
RUMANIA, 23 n. 2, 194 n. 4.
Russia, 78, 81, 85, 102 n. i, 140 n. i,
148-9
UNITED Kingdom, passim
United States, passim : see esp.
56-9, 74, 78-80, 93, 94-103,
104-5, ii7, 149-53. 167, 172-3,
176, 230, 234
IV. GENERAL.
BANKING system, 93, 177, 212-8,
225, 228 ft., 247
Barometric pressure, 151-2
CARTELLS, 193, 224-5, 250
Combination, 243-6, 251-2
Compensation, 48-51, 138-44,
153-5
" Cost " theory, 53 ft"., 125-9, 165,
244-5
" Crop " theories, 129-55, 1 65-70
DISCOUNT, rate of, 213-225
Discrimination, 252
EXPORTS and Imports, 168-9, 172-
4, 176 n. i, 230, 231 n. 5
FREYCINET scheme, 72 n. 4, 216 n. 2
GESTATION, period of, 13 ff., 178,
1 80
Gold, 214 n. i, 225-35
INVENTION, 66-8, 157-62, 183-7,
248
JOINT-STOCK Acts, 29, 40
LEITER corner, 112, 141
Limited Liability Act, 29, 40
MARKETS, Law of, 188, 198-205
PRESIDENTIAL elections, 94 n. i
Psychology of business man, 8,
38-9, 61, 70, 212
" REPERCUSSION " theory, 122-5,
188 n. i, 205
SPECULATION, 49-51, 138-43, 221-
4, 231, 234, 251
Steel Corporation, 58, 244-5
Sunspots, 144-6
TARIFFS, 74
" UNDER CONSUMPTION," 235-8
WAGES, 206—11, 215, 226-7, 235-8,
249
War, 72-4, 179-80, 183
P. S. King & Son, Ltd., Orchard House, Westminster.
14 DAY USE
RETURN TO DESK FROM WHICH BORROWED
LOAN DEPT.
This book is due on the last date stamped below,
or on the date to which renewed. Renewals only:
Tel. No. 642-3405
Renewals may be made 4 days prior to date due.
Renewed books are subject to immediate recall.
LD21A-20m-3,'73
(Q8677slO)476-A-31
General Library
University of California
Berkeley
U.C. BERKELEY LIBRARIES
Live Music Archive
Librivox Free Audio
Metropolitan Museum
Cleveland Museum of Art
Internet Arcade
Console Living Room
Books to Borrow
Open Library
TV News
Understanding 9/11